United States Office of Research and Development
Environmental Protection Research Triangle Park NC 27711 September 1992
Agency Office of Air and Radiation
Washington DC 20001
vvEPA The 1992 International
Symposium on Radon
and Radon Reduction
Technology:
Volume 5. Preprints
Poster Papers
Session II Federal Programs and Policies
Relating to Radon
Session III State and Local Programs and
Policies Relating to Radon
Session V Radon Measurement Methods
Session VI Transport and Entry Dynamics
of Radon
Session VII Radon Reduction Methods
Session VIII Radon Occurrence in the
Natural Environment
Session IX Radon Surveys
Session X Radon in Schools and Large
Buildings
Session XI Radon Prevention in New
Construction
Session XII Radon in Water
September 22-25,1992
Sheraton Park Place Hotel
Minneapolis, Minnesota
-------
The 1992 International
Symposium on Radon and
Radon Reduction Technology
"Assessing the Risk"
September 22-25,1992
Sheraton Park Place Hotel
Minneapolis, Minnesota
Sponsored by:
U.S. Environmental Protection Agency
Air and Energy Engineering Research Laboratory
and
U. S. Environmental Protection Agency
Office of Radiation Programs
and
Conference of Radiation Control
Program Directors (CRCPD), Inc.
Printed on Recycled Paper
-------
The 1992 International Symposium on Radon
and Radon Reduction Technology
Table of Contents
Oral Papers
Session I: Radon-Related Health Studies
Preliminary Radon Dosimetry from the Missouri Case-Control of
Lung Cancer Among Non-Smoking Women
Michael Alavanja, R. Brownson, and J. Mehaffey,
National Cancer Institute 1-1
Rationale for a Targeted Case-Control Study of Radon and
Lung Cancer Among Nonsmokers
Mark Upfal and R. Demers, Wayne State University and Michigan
Cancer Foundation; L. Smith, Michigan Cancer Foundation I-2
EPA's New Risk Numbers
Marion Cerasso, U. S. EPA, Office of Radiation Programs I-3
Interaction of Radon Progeny and the Environment and Implications as
to the Resulting Radiological Health Hazard
Lidia Morawska, Queensland University of Technology, Australia I-4
Does Radon Cause Cancers Other than Lung Cancer?
Sarah Darby, Radcliffe Infirmary. Oxford, England I-5
Measurements of Lead-210 Made In Vivo to Determine Cumulative
Exposure of People to Radon and Radon Daughters
Norman Cohen, G. Laurer, and J. Estrada, New York University I-6
The German Indoor Radon Study - An Intermediate Report After
Two Years of Field Work
Lothar Kreienbrock, M. Kreuzer, M. Gerken, G. Wolke,
H.-J. Goetze, G. Dingerkus, University of Wuppertal;
H.-E. Wichmann, University of Wuppertal and Center for
Environment and Health; J. Heinrich, Center for Environment
and Health; G. Keller, Saar University, Germany 1-7
in
-------
Session II: Federal Programs and Policies Relating to Radon
EPA's Radon Program
Stephen D. Page, U. S. EPA, Office of Radiation Programs 11-1
Revising Federal Radon Guidance
Michael Walker, U. S. EPA, Office of Radiation Programs II-2
Profile of Region 5's Tribal Radon Program
Deborah M. Arenberg, U. S. EPA Region 5 II-3
The Development of the Homebuyer's Guide to Radon
Paul Locke, Environmental Law Institute, and S. Hoyt,
U. S. EPA, Office of Radiation Programs H-4
Mitigation Standards for EPA's Radon Contractor Proficiency Program
John Mackinney, D. Price, and G. L Salmon,
U. S. EPA, Office of Radiation Programs II-5
Consumer Protection and Radon Quality Assurance: A Picture
of the Future
John Hoombeek, U. S. EPA, Office of Radiation Programs II-6
EPA's Proposed Regulations on Radon in Drinking Water
Janet Auerbach, U. S. EPA, Office of Drinking Water II-7
Session III: State and Local Programs and Policies
Relating to Radon
Radon in Schools: The Connecticut Experience
Alan J. Siniscalchi, Z. Dembek, B. Weiss, R. Pokrinchak, Jr.,
L Gokey, and P. Schur, Connecticut Department of Health
Services; M. Gaudio, University of Connecticut; J. Kertanis,
American Lung Association of Connecticut 111-1
Trends in the Radon Service Industry in New York State
Mark R. Watson and C. Kneeland, New York
State Energy Office III-2
Targeting High-Risk Areas
Katherine McMillan, U. S. EPA, Office of Radiation Programs III-3
How Counties Can Impact the Radon Problem
Jerald McNeil, National Association of Counties, and D.
Willhort, Orange County, NC, Board of Commissioners IH-4
IV
-------
Innovative Local Radon Programs
Jill Steckel, National Civic League 111-5
Session IV: Creating Public Action
Indoor Radon: A Case Study in Risk Communication
Stephen D. Page, U. S. EPA, Office of Radiation Programs IV-1
Activating Health Professionals at the Local Level
Deborah McCleland, American Public Health Association IV-2
Translating Awareness Into Consumer Action
Mary Ellen Fise, Consumer Federation of America IV-3
Radon Testing and Mitigation as Applied in Corporate Relocations
Richard Mansfield, Employee Relocation Council IV-4
Ad Council Radon Campaign Evaluation
Mark Dickson and D. Wagner, U. S. EPA, Office of
Radiation Programs IV-5
Session V: Radon Measurement Methods
The U. S. Environmental Protection Agency Indoor Radon
Measurement Device Protocols - Technical Revisions
Melinda Ronca-Battista, Scientific and Commercial Systems Corp.;
A. Schmidt and T. Peake, U. S. EPA, Office of Radiation Programs V-1
A Performance Evaluation of Unfiltered Alpha Track Detectors
William Yeager, N. Rodman, and S. White, Research Triangle
Institute; M. Boyd, U. S. EPA, Office of Radiation Programs;
S. Poppell, Jr., U. S. EPA-NAREL V-2
An Evaluation of the Performance of the EPA Diffusion Barrier Charcoal
Adsorber for Radon-222 Measurements in Indoor Air
David Gray, U. S. EPA-NAREL; J. Burkhart, University of
Colorado; A. Jacobson, University of Michigan V-3
A Lung Dose Monitor for Radon Progeny
Harvel A. Wright, G. Hurst, and S. Hunter, Consultec
Scientific, Inc.; P. Hopke, Clarkson University V-4
-------
The Stability and Response to Radon of New and Recharged Electrets
William G. Buckman and H. Steen III, Western Kentucky University;
S. Popped, Jr., U. S. EPA-NAREL; A. Clark, City of Montgomery, AL V-5
Design and Performance of a Low-Cost Dynamic Radon Test Chamber
for Routine Testing of Radon Detectors
P. Kotrappa and T. Brubaker, Rad Elec, Inc V-6
Session VI: Transport and Entry Dynamics of Radon
Characterization of 222-Radon Entry into a Basement Structure
Surrounded by Low Permeability Soil
Thomas Borak, D. Ward, and M. Gadd, Colorado State University VI-1
Analysis of Radon Diffusion Coefficients of Concrete Samples
K. J. Renken, T. Rosenberg, and J. Bemardin, University of
Wisconsin-Milwaukee VI-2
Data and Models for Radon Transport Through Concrete
Vem C. Rogers and K. Nielsen, Rogers & Associates VI-3
Simplified Modeling for Infiltration and Radon Entry
Max Sherman and M. Modera, Lawrence Berkeley Laboratory VI-4
The Effect of Interior Door Position and Methods of Handling Return Air
on Differential Pressures in a Florida House
Arthur C. Kozik, P. Oppenheim, and D. Schneider, University
of Florida VI-5
Building Dynamics and HVAC System Effects on Radon Transport
in Florida Houses
David Hintenlang and K. AI-Ahmady, University of Florida VI-6
Radon Entry Studies in Test Cells
Charles Fowler, A. Williamson, and S. McDonough, Southern
Research Institute VI-7
Model-Based Pilot Scale Research Facility for Studying Production,
Transport, and Entry of Radon into Structures
Ronald B. Mosley and D. B. Harris, U. S. EPA-AEERL;
K. Ratanaphruks, ACUREXCorp VI-8
VI
-------
Session VII: Radon Reduction Methods
Durability of Sub-Slab Depressurization Radon Mitigation Systems
in Florida Houses
C. E. Roessler, R. Varas, and D. Hintenlang, University of Florida VII-1
A Novel Basement Pressurization-Energy Conservation System
for Residential Radon Mitigation
K. J. Renken and S. Konopacki, University of
Wisconsin-Milwaukee VII-2
The Energy Penalty of Sub-Slab Depressurization Radon
Mitigation Systems
Lester S. Shen and C. Damm, University of Minnesota; D. Bohac
and T. Dunsworth, Center for Energy and the Urban Environment VII-3
Design of Indoor Radon Reduction Techniques for Crawl-Space
Houses: Assessment of the Existing Data Base
D. Bruce Henschel, U. S. EPA-AEERL VII-4
Multi-Pollutant Mitigation by Manipulation of Crawlspace
Pressure Differentials
Bradley H. Turk, Mountain West Technical Associates; G. Powell,
Gregory Powell & Associates; E. Fisher, J. Harrison, and B. Ligman,
U. S. EPA, Office of Radiation Programs; T. Brennan, Camroden
Associates; R. Shaughnessy, University of Tulsa VII-5
Two Experiments on Effects of Crawlspace Ventilating on Radon
Levels in Energy Efficient Homes
Theodor D. Sterling, Simon Fraser University; E. Mclntyre, Hughes
Baldwin Architects; E. Sterling, Theodor Sterling & Associates VII-6
Session VIII: Radon Occurrence in the Natural Environment
Indoor Radon and the Radon Potential of Soils
Daniel J. Steck and M. Bergmann, St. John's University VIII-1
Nature and Extent of a 226-Radium Anomaly in the Western
Swiss Jura Mountains
Heinz Surbeck, University Perolles, Switzerland VIII-2
Radon Potential of the Glaciated Upper Midwest: Geologic and
Climatic Controls on Spatial Variation
R. Randall Schumann, U. S. Geological Survey VIII-3
VII
-------
EPA's National Radon Potential Map
Sharon Wirth, U. S. EPA, Office of Radiation Programs vill-4
Session IX: Radon Surveys
Comparing the National and State/EPA Residential Radon Surveys
Jeffrey L. Phillips and F. Marcinowski, U. S. EPA, Office of
Radiation Programs IX-1
Radon Testing in North Dakota Day Care Facilities
Arlen L Jacobson, North Dakota State Department of Health ix-2
Ventilation, Climatology and Radon Activity in Four Minnesota Schools
Tim Burkhardt, E. Tate, and L Oatman, Minnesota
Department of Health IX-3
Estimates From the U. S. Environmental Protection Agency's
National School Radon Survey (NSRS)
Lisa A. Ratcliff, U. S. EPA, Office of Radiation Programs;
J. Bergsten, Research Triangle Institute IX-4
Session X: Radon in Schools and Large Buildings
EPA's Revised School Radon Measurement Guidance
Chris Bayham, U. S. EPA, Office of Radiation Programs X-1
Radon in Commercial Buildings
Harry Grafton and A. Oyelakin, Columbus, Ohio Health Department X-2
Iowa Multiresidential Building Radon Study
James W. Cain, Iowa State University Energy Extension X-3
Airflow in Large Buildings
Andrew Persily, U. S. Department of Commerce X-4
Meeting Ventilation Guidelines While Controlling Radon in Schools
Eugene Fisher and B. Ligman, U. S. EPA, Office of Radiation
Programs; T. Brennan, Camroden Associates; W. Turner,
H. L. Turner Group; R. Shaughnessy, University of Tulsa X-5
Radon Reduction in a Belgian School: From Research to Application
P. Cohilis, P. Wouters, and P. Voordecker, Building
Research Institute, Belgium X-6
VIII
-------
Multiple Mitigation Approaches Applied to a School with
Low Permeability Soil
D. Bruce Harris, U. S. EPA-AEERL; E. Moreau and R. Stilwell,
Maine Department of Human Services X-7
General Indoor Air Investigations in Schools with Elevated Radon Levels
Terry Brennan, Camroden Associates; G. Fisher and B. Ligman,
U. S. EPA, Office of Radiation Programs; R. Shaughnessy,
University of Tulsa; W. Turner and F. McKnight,
H. L Turner Group X-8
Comparison of ASD and HVAC System Control in School Buildings
Bobby Pyle, Southern Research Institute; K. Leovic, T. Dyess,
and D. B. Harris, U. S. EPA-AEERL X-9
Effectiveness of HVAC Systems for Radon Control in Schools
Kelly W. Leovic, B. Harris, T. Dyess, and A. B. Craig,
U. S. EPA-AEERL; Bobby Pyle, Southern Research Institute X-10
Radon Prevention in Construction of Schools and Other Large
Buildings - Status of EPA's Program
A. B. Craig, K. Leovic, and D. B. Harris, U. S. EPA-AEERL X-11
Session XI: Radon Prevention in New Construction
The Effect of Radon-Resistant Construction Techniques
in a Crawlspace House
David L Wilson and C. Dudney, Oak Ridge National Laboratory;
T. Dyess, U. S. EPA-AEERL XI-1
Performance of Slabs as Barriers to Radon in 13 New Florida Homes
James L. Tyson and C. Withers, Florida Solar Energy Center XI-2
HVAC Control of Radon in a Newly-Constructed Residence with
Exhaust-Only Ventilation
Michael Clarkin and T. Brennan, Camroden Associates;
T. Dyess, U. S. EPA-AEERL XI-3
A Simplified Analysis of Passive Stack Flow Rate
Pah I. Chen, Portland State University XI-4
Factors that Influence Pressure Field Extension in New Residential
Construction: Experimental Results
Richard Prill, Washington State Energy Office; W. Fisk
and A. Gadgil, Lawrence Berkeley Laboratory XI-5
IX
-------
Evaluating Radon-Resisant Construction Practices in Florida
John Spears, H. Rector, and D. Wentling, GEOMET Technologies XI-6
Laboratory Investigations for the Search of a Radon-Reducing Material
Lakhwant Singh, J. Singh, S. Singh, and H. Virk, Guru Nanak
Dev University, India XI-7
Session XII: Radon in Water
Risk Assessment Implications of Temporal Variation of Radon and
Radium Well Water Concentrations
Alan J. Siniscalchi, C. Dupuy, D. Brown, and B. Weiss,
Connecticut Department of Health Services; Z. Dembek, M. Thomas,
and N. McHone, Connecticut Department of Environmental
Protection; M. v.d. Werff, U. S. EPA Region 1 XII-1
Seasonal Variability of Radon-222, Radium-226, and Radium-228 in
Ground Water in a Water-Table Aquifer, Southeastern Pennsylvania
Lisa A. Senior, U. S. Geological Survey XII-2
Radon in Tap Water from Drilled Wells in Norway
Bjorn Lind and T. Strand, National Institute of Radiation Hygiene XII-3
A Rapid On-site Detector of Radon in Water
Lee Grodzins, Massachusetts Institute of Technology, and
S. Shefsky, NITON Corporation XII-4
-------
Poster Papers
Session II Posters: Federal Programs and Policies Relating to Radon
Radon Measurement Proficiency Program: New Exam and Listing
for Individuals
G. Lee Salmon and P. Jalbert, U. S. EPA, Office of
Radiation Programs IIP-1
Social and Economic Considerations in the School Evaluation Program
Jed Harrison, U. S. EPA, Office of Radiation Programs IIP-2
The Health of the Radon Industry - Survey and Program Results
from Radon Proficiency Program Analyses
James Long, U. S. EPA, Office of Radiation Programs IIP-3
Session III Posters: State and Local Programs and Policies
Relating to Radon
The Radon Health Effects Committee Report and Its Consequences:
Getting Results in Radon Policy Development
Kate Coleman, E. Fox, and F. Frost, Washington State
Department of Health IIIP-1
Washington State's Innovative Grant: School Radon Action Manual
Linda B. Chapman, Washington State Department of Health IIIP-2
Teaming Up on Local Radon Issues
Robert Leker, State of North Carolina IIIP-3
Session V Posters: Radon Measurement Methods
A Decision-Theoretic Model for Evaluating Radon Test Procedures
Based on Multiple Short-term Measurements
Harry Chmelynski, S. Cohen & Associates VP-1
Operational Evaluation of the Radon Alert Continuous Radon Monitor
Emilio B. Braganza, III and R. Levy, U. S. EPA-LVF VP-2
A New Design for Alpha Track Detectors
Raymond H. Johnson, Key Technology, Inc VP-3
XI
-------
Measurements of Indoor Thoron Levels and Disequilibrium Factors
Yanxia Li and S. Schery, New Mexico Institute of Mining
and Technology; B. Turk, Mountain West Technical Associates VP-4
Comparison of Continuous and Occupancy Time Radon Measurements
in Schools Using Programmable E-Perms
Marvin Haapala, C. DeWitt, R. Power, and R. Fjeld,
Clemson University VP-5
Indoor Radon in New York State Schools
Susan VanOrt, L. Keefe, W. Condon, K. Rimawi, C. Kunz,
and K. Fisher, New York State Department of Health VP-6
Session VI Posters: Transport and Entry Dynamics of Radon
Simplified Modeling of the Effect of Supply Ventilation on Indoor
Radon Concentrations
David Saum, Infiltec; M. Modera, Lawrence Berkeley Laboratory;
K. Leovic, U. S. EPA-AEERL VIP-1
Determination of Minimum Cover Thickness for Uranium Mill
Tailings Disposal Cells
Jeffrey Ambrose and D. Andrews, CWM Federal
Environmental Services, Inc VIP-2
A Mathematical Model Describing Radon Entry Aided by an Easy
Path of Migration Along Underground Tunnels
Ronald B. Mosley, U. S. EPA-AEERL VIP-3
Radon Diffusion Studies in Soil and Water
Manwinder Singh, S. Singh, and H. Virk, Guru Nanak
Dev University, India VIP-4
Stack Effect and Radon Infiltration
Craig DeWitt, Clemson University VIP-5
Relative Effectiveness of Sub-Slab Pressurization and
Depressurization Systems for Indoor Radon Mitigation:
Studies with an Experimentally Verified Model
Ashok J. Gadgil, Y. Bonnefous, and W. Rsk,
Lawrence Berkeley Laboratory VIP-6
XII
-------
Session VII Posters: Radon Reduction Methods
Radon Mitigation Systems - A Liability in Cold Climate Homes?
Kenneth D. Wiggers, American Radon Services, Ltd VIIP-1
Why We Like Diagnostics
John W. Anderson, Jr. and J. Bartholomew, Jr.,
Quality Conservation VIIP-2
An Approach to Computer-Assisted Radon Mitigation
Hormoz Zarefar, P. Chen, and P. Byrne, Portland State University;
C. Eastwood, Bonneville Power Administration VIIP-3
Radon Control - Field Demonstrations: Diagnostic and Mitigation
Techniques Used in Twenty-Six Radon Field Workshops
Craig E. Kneeland and M. Watson, New York State Energy Office;
W. Evans, Evanshire Company, Ltd.; T. Brennan,
Camroden Associates VIIP-4
Radon Mitigation at Superfund Remedial Action Sites: Field
Experience and Results
Jean-Claude Dehmel, S. Cohen & Associates; R. Simon,
R. F. Simon Company, Inc.; E. Fisher, U. S. EPA, Office of
Radiation Programs VIIP-5
Dose and Risk Projection for Use of Sub-Slab Radon Reduction
Systems Under Realistic Parameters
Larry Jensen, U. S. EPA Region 5; F. Rogers and C. Miller,
Centers for Disease Control VIIP-6
Session VIII Posters: Radon Occurrence in the Natural Environment
Influence of Meteorological Factors on the Radon Concentration
in Norwegian Dwellings
Terje Strand and N. Bohmer, Norwegian National Institute
of Radiation Hygiene VIIIP-1
Soil Radon Potential Mapping and Validation for Central Florida
Kirk K. Nielson and V. Rogers, Rogers and Associates;
R. Brown and W. Harris, University of Florida; J. Otton,
U. S. Geological Survey VIIIP-2
XIII
-------
Correlation of Indoor Radon Screening Measurements with Surficial
Geology Using Geographic Information Systems
Charles Schwenker, J-Y Ku, C. Layman, and C. Kunz,
New York State Department of Health .............................. VIIIP-3
Analysis of Indoor Radon in New Mexico Using Geographic
Information Systems (GIS)
Richard A. Dulaney, Lockheed Engineering and Sciences Co ......... VIIIP-4
A Radon "Pipe" (?) in the Brevard Fault Zone Near Atlanta, Georgia
L. T. Gregg and J. Costello, Atlanta Testing & Engineering ............ VIIIP-5
Session IX Posters: Radon Surveys
Summary of Regional Estimates of Indoor Screening
Measurements of 222-Radon
Barbara Alexander, N. Rodman, and S. White, Research Triangle
Institute; J. Phillips, U. S. EPA, Office of Radiation Programs ........... IXP-1
Texas Residential Radon Survey
Charles Johnson, G. Ramirez, and T. Browning, Southwest Texas
State University; G. Smith, P. Breaux, and V. Boykin, Texas
Department of Health ............................................ IXP-2
Radon Survey of Oregon Pubic Schools
Ray D. Paris and G. Toombs, Oregon Health Division ................ IXP-3
Quality Assurance in Radon Surveys
William M. Yeager, R. Lucas, and J. Bergsten, Research Triangle
Institute; F. Marcinowski and J. Phillips, U. S. EPA, Office of
Radiation Programs ........................................... IXP-4
Radon in Houses Around the Plomin Coal Fired Power Plant
N. Lokobauer, Z. Franic, A. Bauman, and D. Horvat,
University of Zagreb, Croatia ......................................
A Radon Survey at Some Radioactive Sites in India
Jaspal Singh, L Singh, S. Singh, and H. Virk, Guru Nanak
Dev University, India .......................................... IXP_6
Islandwide Survey of Radon and Gamma Radiation Levels
in Taiwanese Homes
Ching-Jiang Chen, C-W Tung, and Y-M Lin, Taiwan Atomic
Energy Council ......................................... IXP-7
XIV
-------
Session X Posters: Radon in Schools and Large Buildings
Solar Fresh Air Ventilation for Radon Reduction
Monty Holmes, Intermountain Radon Service, and Kelly
Leovic, U. S. EPA-AEERL XP-1
Characteristics of School Buildings in the U. S.
Kelly Leovic, U. S. EPA-AEERL; H. Chmelynski, S. Cohen
& Associates XP-2
Radon in Schools in Wisconsin
Conrad Weiffenbach and J. Lorenz, Wisconsin Bureau
of Public Health XP-3
Investigation of Foundation Construction Details to Facilitate Subslab
Pressure Field Extension in Large Buildings
Michael E. Clarkin, Camroden Associates; F. McKnight,
H. L Turner Group; K. Leovic, U. S. EPA-AEERL XP-4
Radon Measurements in the Workplace
David Grumm, U. S. EPA, Office of Radiation Programs XP-5
Radon Survey of Oregon Public Schools
George L Toombs and R. Paris, Oregon Health Division XP-6
Session XI Posters: Radon Prevention in New Construction
Model Standards and Techniques for Control of Radon in New Buildings
David M. Murane, U. S. EPA, Office of Radiation Programs XIP-1
Combined Ventilation and ASD System
David Saum, Infiltec, and F. Sickels, New Jersey Department
of Environmental Protection XIP-2
Evaluation of Passive Stack Mitigation in 40 New Houses
Michael Nuess, Washington State Energy Office XIP-3
Radon Remediation and Life Safety Codes
Lyle Sheneman, Chem-Nuclear Geotech, Inc XIP-4
A Passive Stack System Study
Geoffrey Hughes and K. Coleman, Washington State
Department of Health XIP-5
xv
-------
Session XII Posters: Radon in Water
Radon in Water Measurements Using a Collector-Bubbler
Robert E. Dansereau and J. Hutchinson, New York State
Department of Health XIIP-1
Measurements of Radon in Water via Sodium iodide Detectors
Paul N. Houle, East Stroudsburg University, and D. Scholtz,
Prosser Laboratories XIIP-2
Continuous Measurement of the Radon Concentration in Water Using
Electret Ion Chamber Method
Phillip K. Hopke, Clarkson University, and
P. Kotrappa, Rad Elec, «nc XIIP-3
Performance Testing the WD200 Radon in Water Measurement System
George Vandrish and L Davidson, Instruscience Ltd XIIP-4
Temporal Variations in Bedrock Well Water Radon and Radium, and
Water Radon's Effect on Indoor Air Radon
Nancy W. McHone and M. Thomas, Connecticut Department of
Environmental Protection; A. Siniscalchi, Connecticut Department
of Heatth Services XIIP-5
XVI
-------
Session II Posters
Federal Programs and Policies
Relating to Radon
-------
IIP-1
Radon Measurement Proficiency Program
••v Bxaa and Listing for Individuals
by
G. Lee Salmon and Phil Jalbert
U.S. Environmental Protection Agency
Office of Radiation Programs
ANR-464
401 M Street, S.H.
Washington, D.C. 20460
ABSTRACT
The U.S. Environmental Protection Agency has developed a
new proficiency listing status for individuals in the Radon
Measurement Proficiency (RMP) Program to assist consumers in
identifying Competent contractors providing on-site residential
radon measurement services. Since its inception in 1986, the RMP
Program has evaluated radon measurement firms and organizations.
The Agency has now developed a measurement exam to better assess
the proficiency of individuals alone.
Contractors Who wish to obtain this new proficiency designation
must pass a written national examination on radon measurement and
meet other requirements. The exam was first offered in January,
1992 and evaluates six major areas of practice including a
contractor's knowledge of radon health effects and risks, types
of measurement devices, special procedures to ensure valid
measurements as part of real estate transactions, interpretation
of test reports, strategies for mitigation work to reduce indoor
radon levels^ and professional standards of conduct and ethics.
Upon obtaining listing status, individuals receive an EPA
identification card to assist consumers in identifying them as
knowledgeable* proficient measurement contractors.
This paper discusses the exam, relevant EPA policies addressed
therein, listing requirements, recommended training through EPA's
Regional Radon Training Centers and the national distribution of
RMP proficient measurement contractors. It will also discuss use
of the exam and fcMP listings as part of requirements in state
radon certification programs, and explore issues relating to
combining KPJi^s proficiency programs under one common
administratiVa structure.
-------
IIP-2
Title: Social and Economic Considerations in the School Evaluation Program
Author: Jed Harrison, U. S. EPA, Office of Radiation Programs
This paper was not received in time to be included in the preprints, and
the abstract was not available. Please check your registration packet for a
complete copy of the paper.
-------
IIP-3
THE HEALTH OF THE RADON INDUSTRY - SURVEY AND PROGRAM RESULTS
FROM RADON PROFICIENCY PROGRAM ANALYSES
By: James Long
U. S. EPA
Office of Radiation Programs (ANR-464)
401 M Street SW
Washington, D. C. 20460
ABSTRACT
From data collected to support the Radon User Fee Rule, it is
possible to characterize the radon industry in terms such as:
o Device availability by type of device and location
o State radon certification programs and
participation in these programs
o Participation levels in state and federal
proficiency programs with keys to how long a firm
has been in the radon measurement and/or
mitigation business
o Etc.
The paper will detail, in terms of statistics, the types of
programs currently in operation and how voluntary and mandatory
programs have impacted the types of services available for
measurements and mitigations. Types of firms and individuals can
be characterized in terms of: how long they have been in
business; what types of services they offer; how many measurements
or mitigations they perform monthly and yearly; what prices they
charge for their services; and who will, likely, be left and what
types of services might be available in the 1990s.
-------
Session III Posters
State and Local Programs and Policies
Relating to Radon
-------
IIIP-1
THE RADON HEALTH EFFECTS COMMITTEE REPORT
AND ITS CONSEQUENCES; GETTING RESULTS IN RADON POLICY DEVELOPMENT
by: Kate Coleman, Ed Fox and Floyd Frost
Washington State Department of Health
Olympia, Washington 98504-7825
ABSTRACT
In the overall effort to develop a public health policy on
radon, the Washington State Department of Health conducted a study
as a background report for the Health Effects Committee of the
state's Radon Task Force. Finding that the results of residential
radon health effects studies are inconclusive, the report warns
that recommendations for action be stated with an explicit
discussion of areas of uncertainty. The Committee finds no current
direct evidence that radon causes or does not cause lung cancer at
concentrations below 4 pCi/1, but recommends that the Department of
Health accept the U.S. Environmental Protection Agency action level
of 4 pCi/1 for residential exposures. The Committee also
recommends the use of the BEIR IV risk estimates, which assume that
all levels of radon pose a risk and that this risk increases with
increasing radon concentrations. The Committee urges the
Department of Health to encourage residential radon testing,
particularly in areas known to have high radon levels. One
important outcome of this work is the state's adoption of interim
radon resistive construction standards for residential buildings.
These standards are effective for two years, during which time the
Department of Health will conduct further research—a state-wide
survey of residential radon exposures, and an evaluation of the
field performance of passive radon vent stacks. New standards to
replace the interim standards will incorporate new evidence and
will attempt to reduce the areas of uncertainty.
The work described in this paper was partially
funded by the United States Environmental
Protection Agency and is in their review process.
Therefore, the contents do not necessarily reflect
the views of the agency and no official endorsement
should be inferred.
-------
THE RADON HEALTH EFFECTS COMMITTEE REPORT
AND ITS CONSEQUENCES: GETTING RESULTS IN RADON POLICY DEVELOPMENT
BACKGROUND
Radon is produced by the decay of naturally occurring uranium.
Uranium is present, at least in low levels, in all soils and is
abundant in soils in certain geologic regions. Radon can accumulate
indoors at levels which some health experts believe may
substantially increase the risk of lung cancer. Radon has been
shown to cause lung cancer among underground miners. Recently,
there has been considerable concern that the accumulation of radon
in homes is a cause of lung cancer.
Lung cancer, primarily caused by cigarette smoking, is now the
leading cause of cancer death in Washington State. Lung cancer
death rates in Washington State have increased from 7.8 deaths per
100,000 population in 1940 to 52.5 deaths per 100,000 population in
1987. Smoking is believed to be the cause of 90% of lung cancer in
men and 78% of lung cancer in women. Radon may also interact with
cigarette smoking to increase the risk of lung cancer among
smokers.
The Radon Health Effects Committee, a subcommittee of the
Radon Task Force, was formed by the Washington Department of Health
(DOH) to review the scientific literature on radon-caused health
effects, evaluate whether radon may be causing human health effects
in this state, and advise the department on a public health policy
for radon. The Committee met three times in late 1990 and early
1991 to develop the recommendations that follow.
FINDINGS AND RECOMMENDATIONS
(1) The Committee finds there is no current direct evidence
either that radon causes or does not cause lung cancer at
concentrations below 4 pCi/1. Some members of the
Committee believe there is uncertainty as to the
existence of health effects at a level of 10 pCi/1.
(2) The Committee believes that for a conservative approach
to public health protection, it is prudent to accept the
risk estimates put forth in reports of major scientific
bodies. More specifically, the Committee recommends
utilizing the National Academy of Science Committee on
Biological Effects of Ionizing Radiations (BEIR IV) risk
estimates.
-------
(3) The Committee recommends that the Washington Department
of Health accept the U.S. Environmental Protection Agency
(EPA) action level of 4 pCi/1 for residential exposure.
However, according to BEIR IV, all levels of radon are
assumed to have a risk, which increases with increasing
radon concentration.
(4) The Washington Department of Health should encourage
occupants, owners or managers of homes and buildings to
test for radon, particularly in areas known to have high
radon levels. Even in areas where average radon
concentrations are low, testing may be advisable since
some houses in these areas have been identified with
levels above 4 pCi/1.
(5) Where indicated, repairs should be made to bring annual
average indoor levels down to or below 4 pCi/1. Follow-up
testing and mitigation information is found in the EPA's
"A Citizen's Guide to Radon".
(6) The Committee recommends that cigarette smokers, whether
or not exposed to indoor radon, be informed that smoking
cessation should be their highest priority for reducing
lung cancer risk.
(7) Since a number of radon health effects studies are
ongoing and the results of these studies may affect the
Findings and Recommendations of this Committee, it is
recommended that the Washington Department of Health
reconvene the Radon Health Effects Committee annually.
The following voluntary participants composed the Radon Health
Effects Committee and endorse the Findings and Recommendations:
MEMBERS AFFILIATION
John Beare Spokane Co. Health District
David Bodansky University of Washington
Fred Cross Battelle, Pacific NW Labs
Kenneth Jackson University of Washington
Jerry Leitch US EPA
Ahmad Nevissi University of Washington
Sam Reed General Public
Maurice Robkin University of Washington
James Matsuyama NE Tri-County Health District
Larry Jecha Washington Association of
Public Health Officials
DEPARTMENT OF HEALTH STAFF
Kate Coleman, Chair, Ed Fox, Floyd Frost,
Pat McLachlan, Sam Milham, Bob Mooney
-------
The attached report reflects the literature review and
discussion of the Committee. All of the comments from the
Committee have not been incorporated, thus the report should be
viewed as a DOH staff report, which formed the background for the
Committee Findings and Recommendations. For organizational
purposes, the report is divided into sections on risk, exposure and
cost.
RADON HEALTH EFFECTS
Alpha-emitting radionuclides can pose a risk to human health,
possibly inducing cancer and birth defects. These radionuclides are
generated by both human and natural sources. Natural sources make
the largest contribution to human exposure, especially through
inhaled radon and radon decay products. This exposure occurs
primarily when radon concentrates indoors. The U.S. Environmental
Protection Agency has estimated that human exposure to naturally
occurring radon is responsible for approximately 16,000 U.S. lung
cancer deaths each year (1). Radon-222 is an inert gas which is
derived from uranium-238. Its abundance in soils varies, depending
on the concentration of uranium-238 in the soil and the
permeability of the soil. Because the air pressure in buildings is
often lower than the atmospheric pressure, radon can be drawn into
buildings from the soil. To a much lesser extent, radon can also be
released from water used in the building and from burning natural
gas which contains low levels of radon. Certain building materials
can also contribute to the indoor radon concentration. Most human
exposure to radon gas and radon decay products occurs from indoor
air exposure.
MINING AND LUNG CANCER
The relationship between mining and elevated lung cancer risk
has been known for over one hundred years. Uranium miners are
heavily exposed to soils and rocks with high uranium
concentrations. Because of this, they receive high exposures to
radon and radon progeny. The evidence for lung cancer resulting
from radon exposure comes from cohort mortality studies of
underground miners who were exposed to differing concentrations of
radon-222 progeny (2,3). The lung cancer hazard for underground
miners was first described in 1879 in Europe (4) . Excess
lung-cancer risk has been observed for uranium miners in the United
States, Czechoslovakia, France and Canada (3,5-7). Excess lung
cancer risk has also been demonstrated for other underground miners
such as Newfoundland fluorspar miners, Swedish metal miners,
Cornish tin miners, Norwegian niobium miners and Chinese tin miners
(8-11).
A casual relationship between lung cancer and radon exposure
was first suggested in the 1930's (5). It was not until after World
-------
War II that radioactivity was seriously considered to be the cause
of the excess lung cancer among miners. During the 1950's and
1960's, several new epidemiological studies were begun to determine
the hazard of exposure to radon progeny in mines (12,13). Unlike
earlier studies, the recent studies more accurately measured radon
exposure. Some also collected data on other lung cancer risk
factors such as smoking. Although studies revealed that underground
miners had a very high smoking prevalence (14), the number of lung
cancers among these miners exceeded the number expected based on
smoking alone (15). One study of Navajo Indian miners, who smoke
very little, demonstrated that radon exposure could increase the
risk of lung cancer even among non-smokers (16). Recently, radon
exposure has also been demonstrated to be a lung carcinogen among
laboratory animals (17).
Since many of the earlier studies of underground miners did
not adequately measure personal radon exposure levels, deriving a
quantitative relationship between exposure and risk was possible
only from the more recent epidemiological studies. These studies
allowed relative risks for lung cancer to be calculated or
estimated per "Working Level Month" of radiation exposure. A
"Working Level" is a measure of radon progeny concentration in the
air where each liter of air has a given amount of potential alpha
energy derived from radon progeny. A "Working Level Month" (WLM)
represents an exposure to a "Working Level" of radon progeny for
170 hours.
DERIVING A RELATIONSHIP BETWEEN EXPOSURE AND DISEASE
Based on the findings in miner studies, concerns were raised
over the hazard of residential exposure to radon and its progeny.
In the 1970's the Public Health Service discovered that high levels
of radon were accumulating in certain houses. The hazards for
people living in houses built over uranium mill tailings or on land
reclaimed from phosphate mines in Florida were discovered. In 1984,
a nuclear-power engineer set off personnel contamination monitors
at his job because of his exposure to high radon concentrations in
his Pennsylvania home (18) .
To assess the risk from indoor radon exposures, a committee of
the National Academy of Sciences (NAS) was formed to evaluate the
evidence for health effects and estimate the risk to citizens. This
committee, known as the Fourth Committee on Biological Effects of
Ionizing Radiation (BEIR IV) (18), developed a model relating radon
exposure to health effects.
In constructing the model or dose-response relationship
between radon exposure and lung cancer risk, BEIR IV used only four
miner cohort studies (18) which had adequate exposure data. The
model assumes that any radon exposure increases lung cancer risk
and that the increasing cumulative exposure linearly increases the
-------
risk. Radon exposure is also assumed to multiply the underlying
risk of lung cancer. Therefore the higher the underlying risk, the
higher the increase in absolute risk.
Because cigarette smokers have a high background risk for lung
cancer, the BEIR IV model estimates that cigarette smokers exposed
to radon will experience a greater absolute risk from radon than
will non-smokers (18). The data supporting this assumption are
limited. Data on tobacco use were available for only two of the
miner cohort studies. Data on the time since cessation of tobacco
use and inhalation practices were unknown. There were also very few
non-smokers in these miner populations. The findings of the
studies suggest that radon exposure and cigarette smoking may
interact by producing a higher risk than simply the sum of the two
risks.
CONCERNS WITH THE RADON RISK EXTRAPOLATION
A number of concerns have been raised about the derived
estimates of risk. Some experts question whether a linear risk
extrapolation is justified. Health effects for miners exposed to
more than 100 "Working Level Month" (WLM) of radon during their
mining work are extrapolated to residents exposed to less than 100
WLM spread throughout a lifetime. It is possible that either the
lower doses or the lower rates of exposure produce fewer cancers
per unit of exposure. Other experts question whether the exposure
estimates for residences, measured during the winter in the lowest
living area accurately estimate lifetime exposure.
The epidemiological studies of underground miners yield data
that relate cancer incidence to radon exposure. However, the
confidence in any one postulated relationship between radon
exposure and cancer incidence is limited. Epidemiological data are
not sufficient to rule out alternative relationships between
disease and exposure. The epidemiological studies are limited by
the study designs and the available data. Some studies restricted
the period of follow-up and thus did not identify disease occurring
many years after exposure. Some studies had uncertain or missing
exposure levels for many individuals. Some were plagued by
unadjusted confounding factors such as smoking. Since smoking is
likely to be the most important cause of lung cancer, even for
miners, not knowing which members of the cohorts smoked and how
much they smoked seriously limits the precision of the derived risk
estimates.
In extrapolating risks for miners to residences, several
adjustments must be made. Radon exposure to underground miners
differs in three important ways from residential exposure. First,
most underground miners have short duration, high intensity
exposure to radon, occurring during the 8 hour workday for 5 days
per week. This pattern of exposure may produce different health
-------
effects than the lower level but longer duration radon exposures
which occur in residences. Second, since mining involves strenuous
physical work, miners breath in more air and thus have a higher
exposure to airborne pollutants. The radon progeny may adhere to
these other pollutants and be more efficiently carried into the
lungs of miners. Third, little is known about the long-term effects
of radon exposure for children.
RESIDENTIAL RADON HEALTH EFFECTS STUDIES
Questions about the BEIR IV risk model would be less important
if the lung cancer risk could be empirically determined for people
living in houses with moderately elevated radon levels. Attempts
have been made to directly evaluate the lung cancer risk from
residential radon exposures. However, since the magnitude of the
expected increased risk is relatively small, the size of the
population in any study capable of detecting such an effect must be
large. To evaluate small elevated risks, epidemiologists prefer
case-control studies rather than cohort studies. A radon
case-control study compares the radon exposure of lung cancer cases
to a comparable people without lung cancer. A radon cohort study
calculates the incidence of lung cancer in a well defined
population with elevated radon exposure. The lung cancer incidence
is then compared to the general population or to a comparable
unexposed population. When studying rare diseases, such as lung
cancer, case-control studies have a much better chance
of detecting an elevated risk.
A 1979 Swedish study examined whether individuals with lung
cancer were more likely to live in stone or wood houses (19) .
Significantly more people with lung cancer lived in stone houses.
A relationship between estimated radon levels and lung cancer risk
was noted in other Swedish case-control studies (20,21). One
cross-sectional study calculated the lung cancer mortality rates
for areas with high and low background radon levels (20). This
study found that areas with high background radon levels had higher
lung cancer mortality rates (21).
A small follow-up study of lung cancers in Maryland revealed
no significant differences in lung cancer risks by housing type
(22). A small cohort study in Ontario, Canada found a twofold
increase in lung cancer risk among residents of houses built with
"radioactive materials" (12,23). This increased risk was not
statistically significant. In New Jersey, a case-control study
found that the lung cancer risk was twice as high for people living
in houses with radon levels above 4 pCi/1 compared to those with
less than 4 pCi/1 (24).
A 1988 Swedish case-control study compared the residential
building characteristics (and thus indirectly radon exposure
levels) of 177 people with lung cancer and 677 controls (people
-------
without lung cancer). A smoking history which included active and
passive smoking was obtained for each case and control. The lung
cancer risk was found to be double for rural residents who lived in
the houses most likely to have high radon levels. No increased risk
was observed for urban residents who lived in houses with high
predicted radon levels (25).
A recent case-control study of 308 Chinese women with lung
cancer was conducted by the U.S. National Cancer Institute (26) .
This study found no association of radon levels above 4 pCi/1 and
lung cancer risk. The lack of association was consistent for both
smokers and non-smokers. No association with radon exposure was
apparent for any lung cancer cell type. This study is the largest
residential case-control study of lung cancer and radon published
to date. The median period of residence was 24 years. The radon
exposures were relatively low with only 20% of the houses having
levels above 4 pCi/1. It is also possible that other exposures
found in China, such as high levels of other indoor air pollutants,
may have obscured a radon effect.
CONCLUSIONS
Without conclusive and consistent findings from studies
relating lung cancer to residential radon exposure, recommendations
for action must be stated with clear, explicit discussion of areas
of uncertainty. Elevated lung cancer risk has been observed
primarily for miners with more than 100 WLMs of exposure delivered
over a relatively short time period. Most lifetime residential
exposures to radon are lower but of longer duration. The strongest
data supporting an elevated lung cancer risk from radon comes from
Sweden (20,21,25). It is possible that the higher residential
exposures commonly found in Sweden differ in risk from the lower
exposures found in most of the U.S. The authors of the Chinese
study report that their findings are not consistent with BEIR IV
risk extrapolations. This may be due to the levels of exposure or
other confounding factors. However, if there is non-linearity in
the risk at low levels, the effect of this non-linearity might be
that low levels of radon exposure would cause less lung cancer than
predicted by the BEIR IV model.
Because all epidemiologic studies are subject to design flaws,
no single study can refute or confirm the risk estimates
extrapolated from underground miner exposures. For this reason,
the risks estimated by BEIR IV and others have not yet been clearly
confirmed or refuted with direct evidence. It should also be noted
that most design flaws or random errors in exposure measurements
will tend to produce lower estimates of risk than are actually
present. These errors and biases must be taken into account when
studying small but potentially important elevated risks.
-------
EXPOSURE ASSESSMENT
An accurate risk assessment requires estimates of both the
health risks per unit radon exposure and the level of exposure.
Since radiation related health effects are believed to depend on
the cumulative radiation exposure, the risk from radon exposure
would depend on the cumulative lifetime radon exposure. Since most
of the time spent indoors occurs at home, the bulk of radon
exposure is received in the home. For that reason, most radon
exposure research has focused on measuring radon levels in homes.
THE EXPOSURE COMPONENT OF RISK ASSESSMENT FOR RADON
Washington State has one of the largest sets of radon
measurements of any state in the U.S. In addition to direct
measurements of radon, accurate geological information pertinent to
radon is also available (27).
GEOLOGY AS A PREDICTOR FOR RADON EXPOSURE
Although nuclides of the uranium-decay series are present in
all soils and rocks, certain geological formations are associated
with high indoor radon levels. Rocks and soils are the primary
source of indoor radon (28). In addition to the soil radium
content, characteristics of density, porosity, and dry gas
permeability of the soil are also critical for prediction. Dryer,
loose soils (often associated with sloping land such as that found
along rivers and streams) facilitates the transport of radon.
Based on the soil radium content and soil characteristics,
predictions of radon problem areas can be made (29) . The most
heavily populated areas of Washington State have soil conditions
associated with low radon levels such as a low radium content and
low permeability due to the wetness of soils. The lower population
areas of northeastern Washington have soil conditions associated
with high radon levels.
RADON TESTING OF RESIDENCES
Testing of residences has focused on single family homes.
Nation-wide, 21 percent of all homes screened had radon levels over
the Environmental Protection Agency's (EPA) action level of 4 pCi/1
(30).
In Washington State, a large number of residences were tested
for radon through Bonneville Power Administration's (BPA) radon
testing program. Eligibility for this program was restricted to
single family homes that were part of a BPA sponsored energy
conservation program. As of January, 1991 BPA tested over 20,000
Washington State homes (31). All the radon measurements were made
-------
by Alpha-Track Detectors (ATD's). The BPA's radon testing program
required that the detectors be placed in residences for a minimum
of three winter months and up to one year (31) . The results of
these tests help establish both a tentative statewide average radon
level in houses and information on the distribution of radon
throughout the state.
The average radon level for Washington State, using only BPA
data, is 1.0 pCi/1 (Fact Sheet, Appendix 3). The percentage of
houses with readings over 4 pCi/1 is 3.3% compared with the
national figure of 21%. This finding suggests that the radon
problem in Washington State is less severe than it is nationally.
SHORTCOMINGS OF BPA SURVEY
It is important to examine the shortcomings of the BPA survey.
The BPA data were collected under the BPA's "Residential
Conservation Programs Radon Monitoring" research program. According
to the BPA:
"These data represent only those readings within the
service areas of utilities (and the State of Washington,
Department of Community Development) who participate in
the Residential Weatherization, Super Good Cents, and
Northwest Energy Code Programs. Several areas in the
Pacific Northwest are excluded in this evaluation. In
addition, the amount of results received is heavily
weighted by the number of homes monitored through large
utility conservation programs located in Western Oregon
and Washington."
The BPA survey was not designed as a random population based
survey of residences in Washington State. Since homes were not
randomly selected and especially since some counties are over
represented in the BPA survey, the results do not provide an
accurate estimate of radon exposure statewide. For example,
Snohomish county, a low radon area, has over 6,000 radon readings
or 29% of all radon measurements in the survey. Snohomish county
has about 10% of the state's population.
WASHINGTON STATE/EPA RESIDENTIAL RADON SURVEY
To improve the reliability of radon exposure estimates and to
provide more county specific exposure estimates, the Washington DOH
is conducting a more valid statewide radon survey of homes. Free
radon testing has been offered to randomly selected households
across the state. This survey will provide a valid profile of
residential radon exposure in Washington State and for selected
regions within the state. The objectives of the project are:
(1) To estimate the distribution of residential radon levels
-------
statewide and
(2) To identify areas of elevated radon levels.
The survey will place activated charcoal canister (CC) radon
testing devices in approximately 2,300 residential structures
throughout the state. A seven day screening measurement will be
obtained. A subset of 10% of the selected residences will receive
additional Alpha Track Detectors (ATD) and additional CCs to be
used as follows:
(1) One ATD will be placed on each floor of the structure (up
to a maximum of four floors) for annual measurements.
There is a minimum of 2 ATDs per home.
(2) Four charcoal canisters will be provided for seasonal
measurements.
The survey began on February 5, 1991, and the first phase,
consisting of approximately 2,300 charcoal canister readings, will
be completed by May 29, 1991. The test protocol maximizes the
estimated radon level for the test homes by requiring that tests be
made in the lowest livable area of the residence under closed home
conditions in the winter months. While this testing strategy is
well founded, its readings cannot be interpreted as accurate
exposure levels for the homes. Reasonable adjustments can be made
to estimate actual annual exposure (32).
SCHOOLS
The Department of Health's Radon Program developed the "School
Radon Action Manual" to be used by school administrators as a guide
to address radon issues. The Department will conduct workshops
statewide on the manual this year.
Some schools in Washington State have begun testing for radon.
The results of these tests are not complete, but some preliminary
findings are available:
(1) Radon levels for schools are high in the same parts of
the state where residential radon levels are high. This
is primarily in the northeastern portion of the state.
(2) Individual schools can have high radon levels even in
counties that have low radon average levels for
residences.
WORLD-WIDE RADON STANDARDS
Since one of the charges of the Committee was to recommend
-------
adoption of a goal for residential radon levels, the Committee
considered the range of standards or action levels adopted
internationally. National radon standards or action levels range
from 4 pCi/1 to 20 pCi/1 for structures. In some countries there
are variations in standards by type of occupancy (residential vs.
workplace), and whether or not the structures are new construction
or existing buildings (33). The International Commission on
Radiological Protection (ICRP) has recommended a standard of 10
pCi/1 for existing buildings and 5 pCi/1 for new construction. It
has been reported that both international recommendations and
individual country standards may soon be revised downward toward
the U.S. EPA action level of 4 pCi/1 (34).
ESTIMATION OF RADON INDUCED LUNG CANCERS - WASHINGTON STATE
The following is a methodology used to calculate excess lung
cancers from residential radon exposure in Washington State.
ESTIMATES OF RISK PER WLM EXPOSURE
BEIR IV estimates that the lifetime lung cancer risk for males
is 506 lung cancers per million males per WLM exposure and 186
deaths per million females per WLM of exposure. For a combined,
equally divided male and female population, the estimated risk is
350 deaths per million population per WLM exposure.
CONVERSION FROM PCI/L TO WLM/YEAR
Assuming that 75% of time is spent in the home, the average
person would spend 3.22 times the number of hours at home than the
average miner spends at work ([75% of 8760 hours/yr]/[170 working
hours per month]). Therefore 12 months of exposure at home results
in (3.22 * 12) =38.6 times the monthly hours of exposure for a
miner. With an equilibrium factor of 0.5 for indoor radon, 1 pCi/1
results in 0.005 WL of exposure. With a duration for yearly
household exposure 38.6 times the duration of a monthly workplace
exposure, an exposure of 1 pCi/1 for one year residential exposure
translates to 0.193 WLM (38.6 * 0.005).
RISK PER PCI/L EXPOSURE
(1) Exposure to 1 pCi/1 for one million males results in the
following risk: (0.193 WLM per pCi/1 * 506 lung
cancers/WLM) =97.7
(2) Exposure to 1 pCi/1 for one million females results in
the following risk: (0.193 WLM per pCi/1 * 186 lung
cancers/WLM) = 35.9
(3) For a combined population of 500,000 males and 500,000
-------
females the risk is: (0.193 WLM per pCi/1 * 346 lung
cancers/WLM) = 66.8
Washington State's average residential radon level, based on
BPA data, is 1.0 picocuries/liter. With a population of 4,657,000
the calculated number of lung cancers due to radon is 311 annually.
COSTS
This section provides current costs of residential testing and
mitigation. The two most common commercially available radon
detectors are the charcoal canister and the Alpha Track Detector.
Charcoal canisters cost $10-15 and Alpha Track Detectors cost $20-
50. The cost of mitigation provided by the U.S. EPA Office of
Radiation Programs is listed below:
TABLE I. ESTIMATED REMEDIATION COSTS
SELF
CONTRACTOR
SIDING
MINOR HOLES & CRACKS $ 50-100 $ 100-150
FORCED VENTILATION $ 50-200 $ 400-800
WALL VENTILATION $ 200-300 $ 1500-2000
SUBSLAB SUCTION $ 200-300 $ 900-2500
Testing for radon is included in the 1990 new energy code passed by the state legislature.
The State Building Code Council (SBCC) has the responsibility for setting construction
standards. SBCC has a contract to develop a cost/benefit model to use in policy making.
When available in the summer of 1991, this cost/benefit model report will address the cost
and benefit of residential radon testing and mitigation.
-------
APPENDIX 1
HISTORIC LUNG CANCER DEATH RATES
WASHINGTON STATE
Lung cancer is the leading cause of cancer death for both males and females in
Washington State. Lung cancer death rates in Washington State have increased
dramatically since the beginning of the century. The rate has increased from 7.8 deaths per
100,000 population in 1940 to 52.5 deaths per 100,000 population in 1987. Time trend
data (Table 1) show a rapidly increasing mortality rate from the late 1940's to the
mid-1950's and again a rapid increase from the late 1960's to the late 1970's. These
increases show the effects of the epidemic initially on men in the 1940's and 1950's and
later on women in the 1960's and 1970's. Each of these epidemics correspond, with a 20
year latency period, to a dramatic increase in cigarette smoking first among men and later
among women.
Since lung cancer is generally fatal, incidence and mortality from lung cancer are
closely related. The fraction of all cases with localized disease at time of diagnosis has
remained stationary over time at about 20%. National data show that one year survival
rate, increased from 20% in 1950 to 31% in the early 1970's.
The lack of a significant number of lung cancers during the early part of the
twentieth century has provided evidence against a significant contribution of radon to the
overall lung cancer risk. Although the diagnosis of lung cancer has improved since the early
part of the century, it is believed that lung cancer should have been easily diagnosed since
the introduction of the X-ray into routine medical practice. This generally occurred during
the 1930's.
Alternatively, a number of factors may have reduced the contribution of radon to
the overall lung cancer risk during the early part of the century. If radon exposure interacts
with cigarette smoke exposure to cause lung cancer, then the combination of the two
exposures produce a greater risk than the sum of the two risks. This means that much of
the increased radon related risk will occur among cigarette smokers and would not have
been observed earlier in the century.
Indoor radon levels may have increased with time. From limited evidence on the
frequency of air exchange in older houses, many believe that less radon would have been
drawn into older houses. As the drafts were sealed, lower air pressure inside relative to
outside would have occurred during the heating season. Therefore, residential radon
exposure would have been lower during the early part of the century. In Sweden,
measurements suggest that indoor radon levels have increased fourfold since the 1950's due
to house tightening and better construction (33).
-------
TABLE 1
Lung Cancer Mortality
(crude mortality per 100,000 population)
(Washington State - both sexes)
Year
1940
1941
1942
1943
1944
1945
1946
1947
1948
1949
1950
1951
1952
1953
1954
1955
Death
Rate
7.8
7.9
8.4
9.2
8.5
7.8
9.3
12.3
11.9
13.3
12.3
13.6
14.8
16.0
16.5
17.9
Year
1956
1957
1958
1959
1960
1961
1962
1963
1964
1965
1966
1967
1968
1969
1970
1971
Death
Rate
19.1
18.5
21.6
19.6
22.5
22.4
22.9
22.9
23.9
23.9
27.2
29.5
31.7
34.0
34.8
37.5
Year
Rate
1972
1973
1974
1975
1976
1977
1978
1979
1980
1981
1982
1983
1984
1985
1986
1987
Deat
34.4
37.9
39.1
38.9
41.1
42.7
41.6
43.4
47.8
46.6
46.5
45.2
49.8
50.5
52.3
52.5
Respiratory cancer defined according to International Classification of Diseases 7th
revision codes 160-164.
-------
APPENDIX 2
LUNG CANCER MORTALITY
WASHINGTON STATE COUNTIES
Age-adjusted death rates were calculated for Washington State counties for the period
of 1950 to 1987. Age-adjustment is necessary when comparing death rates for different
populations. Since the age distribution of populations often differ, the population with
proportionately more old people can be expected to have higher death rates. The question
of interest is usually not whether older people are at a higher risk of death, but rather
whether people of similar age residing in different areas differ in their risk of death. Instead
of comparing the death rates for each age group, an age-adjusted death rate is used as a
standardized summary measure.
To derive an age-adjusted death rate, the age-specific mortality rates by sex for each
county are calculated. These age-specific county rates are applied to the 1940 U.S.
population to estimate how many citizens nationwide would have died in 1940 if the entire
country had the age-specific mortality rates of this county. The age-adjusted death rate is
calculated as the number of these expected deaths divided by 1940 U.S. population counts.
For purposes of county comparisons, death rates for whites were calculated. This was done
because the non-white population of many counties, each non-white racial group has very
different smoking habits. Thus, a rate that groups all non-whites provides little meaningful
information.
For lung cancer, the age-adjusted death rates demonstrate the magnitude and timing of
the lung cancer epidemic in the counties of Washington State (Tables 1,2). During the
period 1950 to 1959, there were eight counties without a single female lung cancer death
during the decade. Between 1980 and 1987, no county was spared a lung cancer death.
Lung cancer death rates have doubled or tripled for women for the past 2 decades, whereas
for men, in many counties, the rates have stabilized. Since 1950, 30,133 white men and
10,513 white women have died of lung cancer. Approximately 4,000 non-whites have died
of lung cancer during this period.
-------
TABLE 1
CANCER OF THE LUNG, TRACHEA, BRONCHUS
(ICD8 162, 163, 165)
WASHINGTON STATE (White-Male)
AGE ADJUSTED DEATH RATES
(# Deaths per 100,000 Pop.)
COUNTY
ADAMS
ASOTIN
BENTON
CHELAN
CLALLAM
CLARK
COLUMBIA
COWLITZ
DOUGLAS
FERRY
FRANKLIN
GARFIELD
GRANT
GRAYS HARBOR
ISLAND
JEFFERSON
KING
KITSAP
KITTITAS
KLICKITAT
LEWIS
LINCOLN
MASON
OKANOGAN
PACIFIC
PEND OREILLE
PIERCE
SAN JUAN
SKAGIT
SKAMANIA
SNOHOMISH
SPOKANE
STEVENS
THURSTON
WAHKIAKUM
WALLA WALLA
WHATCOM
WHITMAN
YAKIMA
WASHINGTON
U.S.
1950-59 1960-69 1970-79 1980-87
RATE(DEATHS) RATE(DEATHS) RATE(DEATHS) RATE(DEATHS)
27.8 ( 9)
17.4 ( 11)
25.5 ( 47)
28.7 ( 57)
20.4 ( 29)
23.5 (101)
37.3 ( 11)
23.3 ( 64)
22.5 ( 12)
32.4 ( 5)
25.9 ( 17)
4.7 ( 1)
17.3 ( 19)
24.3 ( 79)
29.7 ( 21)
22.8 ( 12)
31.8 (1234
27.2 (107)
18.0 ( 21)
22.0 ( 15)
28.5 ( 72)
12.4 ( 8)
18.6 ( 16)
14.3 ( 19)
30.7 ( 28)
24.7 ( 11)
27.7 (376)
19.7 ( 4)
20.6 ( 55)
25.7 ( 8)
19.6 (139)
23.5 (297)
21.2 ( 22)
23.6 ( 59)
22.1 ( 5)
24.1 ( 53)
23.7 ( 89)
19.1 ( 26)
26.3 (172)
26.4 (3331)
29.6
25.8 ( 12)
37.9 ( 29)
40.9 (100)
64.8 (147)
39.2 ( 70)
41.9 (211)
39.1 ( 12)
43.1 (122)
49.3 ( 36)
40.5 ( 5)
42.2 ( 38)
18.0 ( 3)
43.0 ( 61)
46.2 (148)
21.2 ( 21)
37.8 ( 24)
47.7 (2026)
45.8 (203)
34.4 ( 38)
36.5 ( 26)
40.3 (109)
17.2 ( 11)
39.7 ( 41)
36.3 ( 50)
48.4 ( 53)
24.3 ( 10)
45.0 (665)
44.3 ( 13)
28.2 ( 81)
56.8 ( 17)
41.6 (365)
47.3 (639)
32.5 ( 34)
42.0 (129)
55.6 ( 12)
38.3 ( 88)
35.6 (141)
27.4 ( 37)
48.7 (345)
44.1 (6174)
46.8
62.7 ( 31)
62.8 ( 53)
58.4 (184)
71.6 (182)
57.1 (145)
64.1 (420)
49.1 ( 15)
65.8 (218)
67.3 ( 61)
64.8 ( 13)
60.7 ( 68)
40.0 ( 7)
50.6 ( 99)
64.8 (220)
55.9 ( 94)
50.0 ( 49)
63.5 (2974)
70.4 (381)
50.1 ( 58)
63.0 ( 49)
56.4 (162)
48.9 ( 32)
62.6 ( 95)
69.7 (107)
73.1 ( 88)
49.0 ( 23)
69.2 (1180)
22.0 ( 12)
59.4 (194)
43.8 ( 15)
61.5 (701)
60.9 (885)
44.3 ( 52)
66.7 (263)
53.7 ( 11)
51.7 (124)
52.7 (228)
47.5 ( 66)
67.6 (518)
62.6 (10097)
64.0
64.3 ( 30)
55.1 ( 41)
77.0 (217)
75.3 (181)
69.2 (181)
79.5 (523)
34.6 ( 9)
72.6 (220)
73.2 ( 65)
32.8 ( 6)
71.0 ( 76)
84.6 ( 13)
73.7 (147)
74.5 (211)
74.8 (145)
87.6 ( 88)
63.6 (2768)
67.1 (373)
66.9 ( 73)
67.6 ( 49)
67.7 (174)
46.4 ( 28)
75.5 (122)
69.4 ( 95)
85.2 ( 85)
93.4 ( 37)
75.4 (1248)
49.6 ( 25)
70.9 (219)
74.2 ( 21)
67.8 (802)
67.1 (889)
74.5 ( 81)
70.3 (326)
57.0 ( 10)
54.3 (117)
57.6 (245)
44.2 ( 54)
73.2 (537)
68.2 (10531)
-------
TABLE 2
COUNTY
ADAMS
ASOTIN
BENTON
CHELAN
CLALLAM
CLARK
COLUMBIA
COWLITZ
DOUGLAS
FERRY
FRANKLIN
GARFIELD
GRANT
GRAYS HARBOR
ISLAND
JEFFERSON
KING
KITSAP
KITTITAS
KLICKITAT
LEWIS
LINCOLN
MASON
OKANOGAN
PACIFIC
PEND OREILLE
PIERCE
SAN JUAN
SKAGIT
SKAMANIA
SNOHOMISH
SPOKANE
STEVENS
THURSTON
WAHKIAKUM
WALLA WALLA
WHATCOM
WHITMAN
YAKIMA
WASHINGTON
U.S.
CANCER OF THE LUNG, TRACHEA, BRONCHUS
(ICD8 162, 163, 165)
WASHINGTON STATE (White-Female)
AGE ADJUSTED DEATH RATES
(# Deaths Per 100,000 Pop.)
1950-59 1960-69 1970-79 1980-87
RATE(DEATHS) RATE(DEATHS) RATE(DEATHS) RATE(DEATHS)
2
3
4
2
3
8
1
3
2
6
3
16
5
4
6
2
6
4
2
2
6
5
4
3
4
3
6
4
3
1
5
4
5
0
.1
.4
.8
.4
.5
0
.6
.3
0
.8
0
.4
.2
.2
.3
.5
.5
.5
.0
.8
0
.5
.9
.8
.6
.2
0
.5
0
.6
.4
.6
.8
0
.3
.5
.4
.4
.8
.1
( 0)
( 2)
( 6)
( 10)
( 3)
( 14)
( 0)
( 17)
( 1)
( 0)
( 2)
( 0)
( 2)
( 16)
( 2)
( 7)
(224)
( 17)
( 7)
( 1)
( 17)
( 0)
( 3)
( 3)
( 2)
( 2)
( 70)
( 0)
( 12)
( 0)
( 24)
( 37)
( 3)
( 17)
( 0)
( 9)
( 14)
( 2)
( 34)
( 600)
7
6
6
10
8
3
8
2
0
6
3
10
5
12
8
8
4
2
8
6
10
7
9
5
7
2
6
11
6
6
3
5
12
6
5
6
6
7
0
.5
.1
.3
.5
.4
.6
.7
.9
.0
0
.9
.1
.9
.3
.5
.5
.3
.9
.2
.8
.4
.1
.3
.7
.8
.3
.3
.9
.0
.4
.1
.7
.6
.0
.2
.0
.6
.5
( 0)
( 7)
( 16)
( 15)
( 18)
( 48)
( 1)
( 27)
( 2)
( 0)
( 5)
( 0)
( 6)
( 32)
( 6)
( 7)
(447)
( 41)
( 5)
( 2)
( 23)
( 4)
( 10)
( 9)
( 8)
( 2)
(131)
( 1)
( 19)
( 3)
( 59)
(102)
( 3)
( 21)
( 2)
( 14)
( 24)
( 9)
( 49)
(1178)
11.
14.
15.
16.
22.
16.
15.
16.
9.
18.
15.
3.
16.
16.
17.
14.
18.
16.
10.
10.
16.
8.
17.
15.
16.
7.
18.
12.
13.
10.
16.
15.
7.
18.
5.
13.
13.
13.
15.
16.
7.
2
9
9
1
6
4
1
1
9
4
1
3
9
6
1
1
2
2
4
4
5
9
7
5
3
1
3
7
2
7
6
6
0
4
0
4
6
9
1
7
6
( 6)
( 15)
( 61)
( 44)
( 58)
(130)
( 5)
( 61)
( 10)
( 3)
( 19)
( 1)
( 34)
( 61)
( 30)
( 13)
(1119)
(104)
( 13)
( 8)
( 54)
( 6)
( 27)
( 24)
( 21)
( 3)
(380)
( 5)
( 49)
( 3)
(226)
(279)
( 9)
( 98)
( 1)
( 39)
( 69)
( 21)
(132)
(3241)
17
29
31
26
38
31
29
33
22
21
20
19
29
39
31
35
28
29
23
45
35
36
43
25
52
37
29
18
28
36
32
29
32
28
13
15
25
19
21
29
15
.4
.4
.2
.9
.7
.0
.4
.2
.5
.6
.9
.4
.1
.6
.7
.5
.4
.0
.2
.7
.6
.2
.7
.8
.2
.5
.0
.9
.0
.5
.4
.5
.4
.7
.9
.6
.6
.1
.9
.2
.3
( 10)
( 27)
(110)
( 77)
(113)
(242)
( 7)
(121)
( 22)
( 4)
( 25)
( 3)
( 61)
(128)
( 70)
( 33)
(1637)
(195)
( 27)
( 36)
(105)
( 19)
( 74)
( 33)
( 60)
( 15)
(599)
( 8)
(102)
( ID
(454)
(488)
( 36)
(166)
( 3)
( 39)
(127)
( 25)
(182)
(5494)
-------
REFERENCES
1. Puskin, J.S., Nelson, C.B.: EPA's Perspective on Risks from
Residential Radon Exposure. JAPCA, 39 (7):915-920, 1989.
2. Whittemore, A.S., McMillan, A.: 1983. Lung Cancer Mortality
Among U.S. Uranium Miners: A reappraisal. J.Natl. Cancer
Inst. 71 (3):489-499, 1983.
3. Waxweiler, R.J., Roscoe, R.J., Archer, V.E., Thun, M.J.,
Wagoner, J.K., Lundin, F.E.: Mortality Follow-up Through 1977
of the White Underground Uranium Miners Cohort Examined by
the US PHS. In International Conference, Radiation Hazards in
Mining pp 823-830, 1981.
4. Harting, F.H., Hesse, W.:Der Lungenkrebs, Die Bergkrankheit
in Den Schneeberger Gruben. Vjschr. Gerichtl, Med. Offentl.
Gesundheitswesen. 31:102-132, 313-337, 1979.
5. Peller, S.: Lung Cancer Among Mine Workers in Joachimsthal.
Human Biol 11:130-143, 1939.
6. Tirmarche, M., Brenot, J., Piechowski, J., Chameaud, J.,
Pradel J. The Present State of an Epidemiological Study of
Uranium Miners in France in Proceedings of the International
Conference, Occupa-tion Radiation Safety in Mining Vol 1,
Canadian Nuclear Assn, Toronto, Ontario, Canada 1985.
7. Howe, G.R., Nair, R.C., Newcombe, H.B., Miller A.B., Frost,
S.E., Abbatt, J.D. Lung Cancer Mortality 1950-1980 in
Relation to Radon Daughter Exposure in a Cohort of Workers at
the Eldorado Beaverlodge Uranium Mine. J. Natl. Cancer Inst.
77(2):357-362.
8. Morrison, H.I., Semenciw, R.M., Mao, Y., Corkill, D.A., Dory,
A.B., de Villiers, A.J., Stocker H., Wigle, D.T.: Lung Cancer
Mortality and Radiation Exposure Among the Newfoundland
Fluorspar Miners in Occupation Radiation Safety in Mining.
Stocker, H. ed. Canadian Nuclear Assn., 1985.
9. Edling, C., Axelson, 0.: Quantitative Aspects of Radon
Daughter Exposure in Underground Miners. Br. J. Ind. Med.
40:182-187.
10. Fox, A.J, Goldblatt, P. Kinlen L.J.: A Study of the Mortality
of Cornish Tin Miners. Br. J. Ind. Med. 38:378.
11. Solli, M., Andersen, A., Straden, E., Langand, S.: Cancer
Incidence Among Workers Exposed to Radon and Thoron Daughters
at a Niobium Mine. Scan J. Work Environ. Health 11:7-13,1985.
12. Wang, X., Huang, X.: Radon and Miners Lung Cancer. Zhonghua
Fangshe Yixue yu Fanghu Zazhi 4:10-14, 1984.
-------
13. Lundin, F.D., Wagoner, J.K., Archer, V.E.: Radon Daughter
Exposures and Respiratory Cancer, Quantitative and Temporal
Aspects. Joint Monograph No. 1. Washington, D.C.: U.S.PHS.
14. Archer, V.E., Wagoner, J.K., Lundin, F.E.: Uranium Mining and
Cigarette Smoking Effects on Man. J. Occup. Med. 15:204-211.
15. Lundin, F.D., Archer, V.E., Wagoner, J.K.: An
Exposure-Time-Response Model for Lung Cancer Mortality in
Uranium Miners. In Proceeding of the Work Group at the 2nd
Conference of the Society of Industrial and Applied
Mathematics, Breslow, N.E., and Whittemore, A., eds, 1979.
16. Samet, J.M., Kutvirt, D.M., Waxweiler, R.J., Key, C.R.:
Uranium Mining and Lung Cancer in Navajo Men. N.Eng.J.Med.
310:1481-1484, 1984.
17. Cross, F.T, Palmer, R.F., Filippy, R.E. Dagle, G.E., Stuart,
B.O.: Carcinogenic Effects of-Radon Daughters, Uranium Ore
Dust and Cigarette Smoke in Beagle Dogs. Health Phys.,
42:33-52, 1982.
18. Committee on the Biological Effects of Ionizing Radiations,
National Research Council. Health Risks of Radon and Other
Internally Deposited Alpha-emitters, BEIR IV. Washington DC:
National Academy Press, 1988:602.
19. Axelson, O., Edling, C., 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 5:10-15, 1979.
20. Axelson, O., 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,
14:286-292, 1988.
21. Svensson, C., Eklung, G., Pershagen, G., Indoor Exposure to
Radon from the Ground And Bronchial Cancer in Women. Int
Arch Occup Environ Health 59:123-131, 1987.
22. Simpson, S. G.: Lung Cancer and Housing Characteristics.
Arch Environ Health 38:248-251, 1983.
23. Lees, R. E., Steele, R., Roberts, J.H.: A Case-Control Study
of Lung Cancer Relative to Domestic Radon Exposure. Int J
Epidemiol 16:7-12, 1987.
24. New Jersey Department of Health. A Case-Control Study of
Radon and Lung Cancer Among New Jersey Women. Division of
Epidemiology and Disease Control, August, 1989.
25. Svensson, C., Pershagen, G., Klominek, J.: Lung Cancer in
-------
Women and Type of Dwelling in Relation to Radon Exposure.
Cancer Res 49:1861-1865, 1989.
26. Blot, W. J. , Zhao-Yi Xu, Boice, J.D., Jr., Dong-Zhe Zhao,
Stone, B.J., Jie Sun, Li-Bing Jing, Fraumeni, J.F., Jr.:
Indoor Radon and Lung Cancer in China. Journal of the
National Cancer Institute. Vol.82, NO.12. 1025-1030, June
20, 1990.
27. Turk, B. H., et.al: Characterizing the Occurrence, Sources
and Variability of Radon in Pacific Northwest Homes. J Air
Waste Manage Assoc. Vol.40(4):498-506, 1990.
28. Samet, J. M., Nero, A. V.: Indoor Radon and Lung Cancer. New
England Journal of Medicine. Vol.320, No. 9:591-594.
29. Duval, J. S. Otton, J. K. and Jones, W. J.: Radium
Distribution Map and Radon Potential in the Bonneville Power
Administration Service Area. Dept. of the Interior, U. S.
Geological Survey, 1989.
30. United States Environmental Protection Agency. "Radon Week
Report." October, 1990.
31. Bonneville Power Administration. Radon Monitoring Results
From BPA's Residential Conservation Program, January, 1990.
32. Marcinowski, Frank. Analysis of the Relationship of Short
Term Measurements to Annual Measurements in Support of the
Citizen's Guide Revision. Draft, 1990.
33. Committee on Interagency Radiation Research and Policy
Coordination: Federal Programs on Indoor Radon. April 1988.
34. Personal Communication with Fred Cross. March 12, 1991.
-------
IIIP-2
WASHINGTON STATE* S INNOVATIVE GRANT:
SCHOOL RADON ACTION MANUAL
by: Linda B. Chapman, R. S.
Washington State Department of Health
Olympia, WA 98504
ABSTRACT
In 1990 and 1991, the Environmental Protection Agency awarded the
Washington State Department of Health money from the State Indoor
Radon Grants Program to fund an innovative project titled
"Community Support Radon Action Team for Schools". The goal of
the project was to establish cooperative and cost-effective
approaches school administrators could use to assess and mitigate
radon exposure in schools.
During the first year a team of federal, state and local
experts from a number of different fields completed the first
draft of the School Radon Action Manual. During the second year
a full scale field test of the School Radon Action Manual was
conducted in a large suburban school district in Western
Washington.
During the field test, a team of consultants assisted school
district personnel in utilizing the manual and dealing with
community concerns aroused by the testing procedure. At the same
time, the team members evaluated the Manual' s effectiveness. The
results of the field test, and the resulting modifications of the
Manual, will be discussed.
-------
IIIP-3
Title: Teaming Up on Local Radon Issues
Author: Robert Leker, State of North Carolina
This paper was not received in time to be included in the preprints, and
the abstract was not available. Please check your registration packet for a
complete copy of the paper.
-------
Session V Posters
Radon Measurement Methods
-------
VP-1
A DECISION-THEORETTr. MODEL FOR EVALUATING RADON TF.ST PROr.F.nTTRF.S
BASED ON MULTIPLE SHORT-TERM MEASUREMENTS
By: Harry Chmelynski
S. Cohen & Associates
1355 Beverly Road
McLean VA 22101
ABSTRACT
Test procedures based on multiple short-term measurements offer
an opportunity to reduce the uncertainty inherent in a single
short-term radon test result. These procedures were examined by
the EPA in a recent revision of the Citizen's Guide to Radon.
Test performance is estimated by the percentage of homes
correctly classified over (or under) 4 pCi/L. Performance
calculations rely on exogenous estimates of errors due to
temporal and spatial variations in radon concentrations and
measurement error. The joint (multivariate) distribution of
short-term test results and long-term levels is constructed,
reflecting the serial correlation of short-term test results and
the population distribution of long-term levels in U.S. homes
from the most recent EPA survey. Various integrals of the joint
distribution yield expected misclassification rates, costs and
benefits for a wide variety of test procedures, both simultaneous
and sequential.
-------
VP-2
OPF.RATTONAT. EVALUATION OF THF. RADON ALERT
rONTINUOUS RADON MONITOR
By: Emilio Braganza III and Richard A. Levy
U. S. Environmental Protection Agency
Office of Radiation Programs - LVF
P. 0. Box 98517
Las Vegas, NV 89193
ABSTRACT
The U.S. EPA Office of Radiation Programs (ORP) Las Vegas
Facility (LVF) is the principal laboratory for evaluating new
radon/radon decay product measurement devices and developing or
modifying measurement protocols. The devices are evaluated for
environmental sensitivity, precision, and bias, and a detailed
assessment is prepared utilizing appropriate criteria for each
instrument based on its ability to measure radon or radon decay
products.
The Radon Alert Instrument is a new micro-processor
controlled, diffusion type, radon gas monitoring (RGM) device
from Monitor Technologies, Ltd. The RGM has undergone a
comprehensive evaluation at the LVF Radon Laboratory. The
devices were exposed to radon-222 ranging from 4 pCi/L to 29
pCi/L under the following environmental conditions: relative
humidity from 35% RH to 65% RH and temperatures from 13 »C to
29°C. Statistical procedures were used to characterize
individual device performance, environmental sensitivity,
precision, and bias. Device suitability for the EPA "Indoor
Radon and Radon Decay Product Measurement Protocols" and device
eligibility for the Radon Measurement Proficiency program
Application Device Checklist was also determined.
-------
VP-3
A New Design for Alpha Track Detectors
Raymond H. Johnson, Certified Health Physicist
Key Technology, Inc.
P.O. Box 562, Jonestown, Pa 17038
Most alpha track detectors have two sources of uncertainty that may adversely
affect the quality of measurement results. First, they may collect tracks before
deployment for radon measurements. When CR-39 plastic is originally produced, it is
protected from collecting background tracks (from radon and radon decay products in
the air) by a thin plastic membrane on the surface of the CR-39. When the chips of
CR-39 are mounted in alpha track devices, the protective membrane is removed. The
entire device is then enclosed in a metallic/mylar bag to prevent entry of radon and
collection of further background tracks. The success of this approach depends on how
well radon is excluded by the bag. Since most bags will leak radon if the partial
pressure of radon is high enough, or if given enough time, background tracks will
accumulate and may limit the shelf life of the detector.
Secondly, once the "radon proof bag is opened to deploy the alpha track
detector, tracks begin collecting and continue collecting until the device is analyzed in
a laboratory. That is to say, mere is no way to turn off the process once the bag is
opened. A couple of alpha track devices have provisions for placing tape over the
opening in the detector, but this approach seals in the radon at the existing ambient
level. The sealed-in radon continues to produce alpha tracks until the device is
disassembled in the lab. For exposures of several months in homes at low radon levels,
the additional tracks collected by the device in transit to the lab may not be significant.
However, alpha track devices are usually calibrated and performance tested by E.P.A.
at high levels of radon for short exposures to produce high integrated exposure levels in
pCi-days/liter. For these kinds of exposure conditions, the additional tracks collected
in transit could be a substantial fraction of the total tracks.
A new device was designed to correct for both deficiencies of alpha track
detectors. The new device provides a method for starting and stopping the collection of
alpha tracks to define a desired exposure interval. Since the original CR-39 is protected
by a thin membrane, that protection is maintained within the alpha track device until the
point of deployment for radon measurement. To stop the exposure, the CR-39 is once
again covered to prevent further collection of alpha tracks. The new device is also
designed such that the covering material, for stopping the collection of alpha tracks, is
protected from the plating out of radon decay products. This is to.assure that the
covering material itself will not add more tracks to the CR-39.
This new design allows for precise exposure intervals and can be used effectively
for both short- and long-term testing from 3 days to a year. Since background tracks
are maintained at a consistently low level, the lower limit of detection for the new design
is about 5 pCi-days/Liter. The upper limit of detection is about 10,000 pCi-days/Liter.
Furthermore, since the CR-39 is protected from collecting tracks while the devices are
in storage, the shelf-life of the new devices is indefinite.
-------
VP-4
Measurements of Indoor Thoron Levels and
Disequilibrium Factors
Yanxia Li, Stephen D. Schery
Physics Department
New Mexico Institute of Mining and Technology
Socorro, NM 87801
Brad Turk
105 E. Marcy St.
Rm 109
Santa Fe, NM 87501
-------
Abstract
The chemical and physical properties of radon and thoron are similar. However,
because of thoron's short half-life (55.6 s), its transport mechanism, distribution in
the indoor air space, and the behavior of its decay products can be different from
those of radon. We took measurements in nine houses in the state of New Mexico
to investigate diurnal thoron variations and disequilibrium of thoron progeny. The
indoor thoron levels vary significantly with time with a maximum concentration usually
occurring in the afternoon. The equilibrium factor for thoron under normal indoor
conditions ranged from 0.013 to 0.084. Predictions from a conventional well mixed
model gave values for the equilibrium factor that were a little smaller, perhaps
because thoron wasn't well mixed in the rooms.
-------
Introduction
Our previous research has shown that in seven New Mexico houses, indoor
thoron was mainly from soil adjacent to the house, and that radon mitigation by
subfloor depressurization was also successful for reducing indoor thoron
concentrations [1]. Since health effects are more directly attributable to the alpha
particle decay of progeny, it's also important to study the behavior of thoron progeny
in indoor spaces.
Thoron progeny in the indoor and outdoor atmosphere come from decay of
thoron gas, they normally are not transported directly from the soil. They are
removed by radioactive decay, by deposition on surfaces, and washout. For indoor
air spaces, there can be an additional process of removing thoron progeny due to air
exchange between indoor and outdoor air.
The average ground level thoron concentration over continental areas (about 2-
8 Bq m"3 [2]) is comparable with that of radon. In contrast, indoor thoron
concentrations are usually lower than indoor radon concentrations. The most
important contribution to the potential alpha energy concentration from progeny of
thoron, PAEC(Tn), is 212Pb, which has a half-life of 10.6 hr, much longer than the rest
of the short-lived progeny. In the seven houses initially studied in our project [11, the
average radon concentration was about 30 times higher than the average thoron
concentration, but the potential alpha energy concentration for radon was only a few
times higher than that for thoron. This fact indicates that if the concentrations of
radon and thoron are the same in a certain area, the working level (WL) or PAEC(Tn)
-------
of thoron progeny is likely to be greater than that of radon progeny. The equilibrium
factor, F, indicates the extent of disequilibrium between a radon isotope and its
progeny. The more effective the progeny removal processes, mentioned above, the
more the equilibrium factor is reduced. For conditions of complete equilibrium (F = 1),
1 WL corresponds to 3700 Bq nrV3 of radon, but to only 273.8 Bq rrr3 for thoron (1
mWL = 20.8 nJ nY3 PAEC).
A number of measured values of the equilibrium factor F for radon have been
reported in the literature. Typical values of F for indoor radon are in the range of 0.3 -
0.6 [2]. The outdoor values for F for radon tend to be higher due to less likelihood
of contact with surfaces. There are fewer reports for thoron progeny and their F
values. Schery [3] estimated an average indoor value for F in the range of 0.015 to
0.04 based on the limited data and model available at the time of his paper. In this
paper, we will report our most recent measurements of indoor thoron progeny and the
equilibrium factor F.
In addition to the -seven houses mentioned above, two more houses with
identification DRJH and DRSS were tested in New Mexico. House DRJH is an adobe
house with both crawlspace and basement. House DRSS is a single family wood
frame house with crawlspace. During certain periods of testing, blowers were
installed to temporarily mitigate the house by sub-floor depressurization. Average
pressure drops were about 3 Pa. A new thoron detector operating on the principle of
coincidence between decay of 220Rn and 216Po was used in houses DRJH and DRSS
to obtain radon and thoron concentrations. Progeny concentrations were measured
-------
with the Scintrex model WLM-30. As with the other seven houses, the soils at these
two additional houses had uranium and thorium concentrations slightly above, but
close to, the average concentrations for american soils.
Experiments
Diurnal pattern of indoor thoron concentration
The variation of indoor thoron concentration with time depends not only on the
exhalation rate from the sources, but also on variation of ventilation conditions and
air exchange between rooms. These are caused by meteorological conditions (wind,
barometric pressure, temperature) and human activities such as opening windows and
doors. Fig. 1 shows the mean diurnal indoor thoron concentrations for eight tested
houses. House TI42 is excluded because the experimental data are insufficient to
represent the diurnal thoron variations and there is no continuous progeny
*
measurements available. The experimental data were obtained during Jan. 1991 to
May 1991 and in March 1992. The data shown in fig. 1 are obtained by averaging
several days (the number of days for each house is listed in table 1, column 7) of 2-
hour interval (or 1 hour for DRSS and DRJH) semi-continuous measurements with the
mitigation systems turned off. In house DRSS, a fan was used in the room where the
measurements took place to achieve better air mixing. In all the other houses,
measurements were made under the existing air mixing conditions. The average
sampling height for thoron gas is about 1 m above the floor and 0.1 m for progeny
-------
(about 1 m for house DRSS and PE21).
The observed changes of thoron concentration during a day varied by a factor
of 2-10. The maxima of thoron concentrations tend to occur in the afternoon and the
minima in the early morning. The research on thoron emanation rate from soil has
shown that the variation of thoron emanation is often characterized by a morning
minimum with the maximum occurring during the afternoon due to barometric
pressure drop and temperature gradient [4, 5]. The increasing emanation rate could
possibly cause the increase in indoor thoron concentrations, but observed emanation
rate increases tend to be smaller [4, 5], and the houses will tend to insulate the
underlying soil from temperature change. The potential alpha energy concentration
should increase when the parent gas entry is increased. Hence, after adjusting for the
appropriate time delay, PAEC(Tn) can be used to indicate the changes in thoron
sources in the afternoon, assuming deposition rate and ventilation have not changed
significantly. For the above eight houses, the evidence from the PAEC(Tn)
measurements suggests the afternoon maxima of thoron concentration could not be
fully explained by an increase in entry rate. Another possible explanation is that
thoron is not well mixed in indoor space due to its short half-life, and the extent of
mixing can vary with time. If this hypothesis is correct, thoron concentration will be
higher near the sources. For the houses where soil is considered to be the major
source of indoor thoron, the highest thoron concentration would normally be found
near the floor. Therefore, a concentration gradient could exist between the floor level
and the measurement height of 1 m. In the afternoon, if the air mixing process is
-------
faster due to convective processes or human activity, it may cause a better mix of
thoron. Therefore, the detector which is set 1 m from the floor may be able to detect
more decays than at any other time. The thoron data for house DRSS support this
hypothesis about mixing, since this house had the greatest mixing and the diurnal
variation of thoron showed the least change, only about a factor of 2.
As discussed in the literature [3], over the range of normal ventilation rates for
houses, due to its short half-life, the indoor thoron concentration is unlikely to be
affected directly by changes in ventilation. So ventilation changes are an unlikely
explanation of the afternoon maximum.
Disequilibrium Between Thoron and Thoron Progeny
The investigation of disequilibrium between thoron and thoron progeny in indoor
space needs the simultaneous measurements of both thoron and thoron progeny
concentrations. The instruments we used to measure progeny were described in an
earlier paper about indoor thoron sources [1 ]. Because the WLM-30 wasn't available
during the period of testing house TI42, the measurements of the equilibrium factor
in this house are excluded. There is a considerable delay in the response of the WLM-
30 to thoron progeny, and a correction for this delay has to be made in making
estimates of F values.
Thoron progeny are more sensitive to the changes in ventilation rate and aerosol
concentration caused by various house operation and human activities. So the
equilibrium factor F of thoron is not necessarily stable during the day and night. The
-------
radon mitigation systems installed in the houses did not have significant effects on
either ventilation rates or aerosol concentrations, and it is thus not surprising that
there were no changes of the equilibrium factor F observed due to the operation of
the mitigation systems.
Table I lists the mean of measured concentration data for each house. The
concentrations of thoron (C) and progeny PAEC(Tn) are the daily averages obtained
from measurements on the days (column 7) when the mitigation systems were off.
The equilibrium factor Fm is deduced from the means of C and PAEC(Tn). The average
equilibrium factors are low in comparison with typical F values for radon. Variation
in aerosol concentration is one important factor that can cause variation in F values.
An increased aerosol concentration tends to increase the fraction of attached progeny
in the air and cause a decrease in progeny plateout rate. The result is that the
progeny concentration in the air is increased if other parameters are kept unchanged,
and the F value will increase.
Grab sample measurements of aerosol concentrations were made at different
locations of the houses using a TSI model 8510 with the impactor of a size range of
0.01 to 3.5 /;m. A moderate correlation between Fm values and measured aerosol
concentrations is shown in fig. 2. The correlation was probably not stronger because:
1) the aerosol concentrations are from the grab samples which may not be equivalent
to the average over the period when the measurements were taken, and 2) F is
influenced by many factors, for example, ventilation rate and ratio R of surface area
(including walls, floor, ceiling, and furniture) to the volume of the house. The value
8
-------
of R changes from house to house. Houses with bigger R values would be expected
to have smaller F values due to the increased probability of progeny deposition on the
surfaces. Simultaneous measurements of aerosol concentration and ventilation rate
would provide a more accurate interpretation of variation in the equilibrium factor.
Comparison with Model Predictions
The prediction of model calculations is very helpful in interpreting the influence
of the various parameters on indoor radioactivity concentrations. The well-mixed
room model [6] with first order sink and source terms is used in this paper to compare
predicted equilibrium factor (Fc) with the experimental data. The model calculates the
influence of ventilation rate, aerosol concentration, the physical process of plateout
(deposition), and attachment to aerosol on the equilibrium factor. The steady-state
concentrations of the unattached C(j and attached C" decay products are
(1)
and
where the index j = 0, 1, 2, and 3 are for 220Rn, 216Po, 212Pb, and 212Bi, respectively.
The superscript (a) is for the attached progeny, while the superscript (f) designates
free (or unattached) progeny. The total activity concentration of jth decay product Cj
is
-------
The concentration C is in units of Bq m"3. The decay constant h, ventilation rate u,
deposition rate q, and attachment rate X are in hr1. The recoil fraction r is
dimensionless and is nonzero for 216Po only. The attached outdoor concentration is
Cj°a. Equations (1) and (2) assume the unattached progeny concentrations from
outdoors are negligible. The attachment rate X is a function of aerosol concentration
Z (cm'3) and can be written as X=@Z, where /? in cm3 hr1, is the attachment
coefficient.
To simplify the calculation a little further, assume the activity concentration
outdoors for all isotopes is negligible, C°' = Q. This assumption is valid for thoron
because the ventilation rate which determines the source entry rate from outside is
much smaller than its decay constant. The assumption for progeny, although often
true, is only an approximation. For thoron progeny,
PAEC(T/})=69.1C2 + 6.56C3 (4)
where PAEC(Tn) is in units of nJ m"3. The equilibrium factor F of thoron is defined as,
where C0 is the concentration of thoron.
The comparison between the measured and calculated F values, Fm and Fc,
respectively, are shown in table I. The following parameter values were used in the
calculations: qa = 0.2 hr1, qf = 40 hr1, 0 = 0.005 cm3 hr1, r, =0.5, and u =0.75 hr1 [3].
10
-------
Since the aerosol measurements were in units of mass concentration, //g m , a
conversion of 9//g rrf3 corresponding to 20,000 particles cm3, was applied to derive
the attachment rate 0. This approximate conversion factor was obtained for our
study conditions by comparing the TSI 8510 side-by-side with a condensation nuclei
counter (Environmental One model Rich 200). The F values from the model
calculation are a little lower than the measured for most of the cases. The measured
F values may be higher than the true values because of the underestimate of thoron
concentrations due to thoron's inhomogeneous distribution. However, the parameters
q", qf, 0 and u were fixed at values taken from the literature. The value u was verified
as reasonable for our houses, but the remaining parameters can vary, and are not well
known for thoron, so could explain some of the discrepancies.
Conclusions
The experimental data have shown that indoor thoron concentrations vary with
time. The daily variation can be a factor of 2-10 depending on the meteorological
conditions. The maxima tended to occur in the late afternoon possibly due to better
mixing of indoor air during that period. The working level of thoron progeny was
comparable with that of radon progeny even though the gas concentration of thoron
was much lower. The study of the thoron equilibrium factor indicates that in the
indoor air of our houses, the disequilibrium between thoron and its progeny is much
greater than for radon, with an average F value generally below 0.1.
11
-------
Although thoron and its progeny in these houses were not high enough to be
a health hazard, they can be a health concern in other areas. This study provides
information, such as F values, that could be useful for estimating dose in other
situations. If we assume that the primary source of indoor thoron is the soil around
a building and that other variables are equal, then the indoor thoron concentration
should approximately scale with the thorium content of the soil. To better understand
the behavior of thoron and thoron progeny in indoor spaces, future studies would
need to obtain more detailed information on the ratio R, ventilation rate, aerosol
concentration, and outdoor activity concentrations.
References
1. Li, Yanxia; Schery, Steve; Turk, Brad Soil as a Source of Indoor 220Rn. Health
Phys. Vol. 62, No. 5, May 1992: 453-457.
2. UNSCEAR Ionizing Radiation: Sources and Biological Effects. 1982.
3. Schery, Stephen D. Radon Isotopes and Their Progeny in the Indoor
Environment. Encyclopedia of Environmental Control Technology, Vol. 2,
Chapter 23, Gulf Publishing, Houston, Tx, 1989.
4. Druilhet, A.; Guedalia, D; Fontan, J.; Laurent, J.L. Study of Radon 220
Emanation Deduced from Measurement of Vertical Profile in the Atmosphere.
J. Geo. Res. 77, 1972: 6508-6514.
5. Guedialia, D; Laurent, J.L.; Fontan, J; Blanc, D.; Druilhet, A. A Study of Radon
220 Emanation from soil. J. Geo. Res. 75, 1970: 357-369.
12
-------
6. Porstendofer, J. Behavior of Radon Daughter Products in Indoor Air. Had. Prot.
Dosim. 7, 1984: 107-113.
13
-------
40
co
6
cr
PQ
o
•l-H
-P
cd
QJ
O
Ł
o
o
CJ
o
30
10
0
0:00
6:00 12:00
time (MST)
18:00
24:00
Figure 1 Diurnal pattern of indoor thoron concentration
in the houses.
Houses with crawlspaces: * DRJH, * DRSS, • CE11, * PE21.
Houses with slab-on-grade: o AL02, * AL04, * SF31, ° TI41
-------
0.10
0.08
o
cd
0.06
0)
0.04
0.02
0.00
0
5 10 15 20 25 30
aerosol concentration (/xg m~3)
35
Figure 2 The correlation between aerosol concentration
and measured equilibrium factor. The solid line
is fitted with F=0.017 + 0.002*aerosol concentration.
-------
Table I Equilibrium factors, and thoron and thoron progeny data for periods with
mitigation systems off.
HOUSE
AL02
AL04
CE11
PE21
SF31
TI41
DRJH
DRSS
MEAN THORON
CONCENTRATION
C (Bq nr3)
5.7
(3.3-10.4)
1.4
(0.4-2.3)
8.2
(2.2-21.1)
10.7
(5.2-19.1)
6.2
(2.1-8.1)
4.9
(0.4-10.8)
13.5
(8.0-22.9)
26.0
(14.7-30.9)
MEAN
PAEC(Tn)
(mWL)
1.1
(1.0-1.5)
0.1
(0.1-0.2)
1.9
(1.7-2.1)
2.3
(0.3-3.7)
1.0
(0.7-1.4)
1.5
(1.2-2.2)
4.1
(3.8-4.4)
1.2
(0.9-1.6)
AEROSOL
CONCENTRATION
(Mg HT3)
9
9
18
12
—
33
30
6
MEASURED
F VALUE
Fm
0.053
0.020
0.064
0.059
0.044
0.084
0.083
0.013
MODEL
PREDICTION
FC
0.044
0.044
0.051
0.047
0.045*
0.055
0.055
0.038
PERIOD
(days)
5
1
8
2
2
1
4
4
* The aerosol concentration of 10 ng nr3 is used in the calculation.
-------
VP-5
COMPARISON OF CONTINUOUS AND OCCUPANCY TIME RADON
MEASUREMENTS IN SCHOOLS USING PROGRAMMABLE E-PERMS
by: Marvin H. Haapala, Craig A. Dewitt, Robert W. Power, and
Robert A. Fjeld
Clemson University
Clemson, SC 29634-0919
ABSTRACT
A study of indoor radon in a county-wide school district in northwestern South
Carolina was conducted. The objectives of the survey were (1) to compare the results of
continuous (24 hour) and occupancy time (12 hour) measurements and (2) to add to the base
of data on radon in schools. Continuous measurements were made in about 100 schools and
almost 3000 rooms. Simultaneous occupancy measurements were made in over 600 of the
rooms. The study was performed using portable, battery operated, electromechanically
actuated electret radon monitors that permit measurements to be made over pre-specified
periods of time. Continuous measurements were for a three day period of time while school
was in session. Occupancy measurements were from 6:00 am until 6:00 pm during the same
three day period.
The arithmetic mean continuous concentration was 2.4 pCi L' (89 Bq m3) and fifteen
percent of the measurements exceeded 4 pCi L' (148 Bq m'3). Fifty-nine percent of the
schools had at least one room with a concentration in excess of 4 pCi L' (148 Bq m3). For
cumulative probabilities above 25%, the distribution was approximately lognormal with a
geometric mean of 1.8 pCi L1 (67 Bq m3) and a geometric standard deviation of 2.2. A
qualitative comparison of the occupancy and continuous measurements revealed no obvious
systematic differences. Overall, the mean difference between occupancy and continuous
concentrations was not statistically significant. However, in 16 percent of the rooms, the
occupancy concentration was significantly larger than the continuous concentration and in 18
percent of the rooms the continuous concentration was significantly larger than the occupancy
concentration.
The work described in this paper was funded by the U.S. Environmental Protection
Agency's State Indoor Radon Grant for South Carolina. The contents do not necessarily
reflect the views of the agency and no official endorsement shall be inferred.
-------
INTRODUCTION
Since students and staff spend anywhere from six to eight hours a day in school, the
potential exists for them to receive significant radon exposures during the school day. The
Environmental Protection Agency tested 130 schools in 16 states, focusing on areas with a
potential for elevated concentrations. The EPA study included a total of 3000 classrooms.
Nineteen percent of the classrooms had screening measurements in excess of 4 pCi L1 (148
Bq nr3), and three percent of the classrooms had concentrations in excess of 20 pCi L'1 (740
Bq m3). Fifty-four percent of the schools had at least one classroom with a screening
measurement in excess of 4 pCi L'1 (148 Bq m3).
Radon concentrations in schools can show considerable variability with time (1),
possibly more so than in houses. This is because ventilation rates and HVAC system
operation during occupied periods during the day can be quite different from those at night.
As a consequence, the average radon concentration during a continuous measurement may
differ from the average concentration during periods of school occupancy. In a study
involving schools in Iowa, Maryland, and North Carolina, Wiggers gt al. (2) found average
concentrations to vary widely during occupied and unoccupied periods. In some buildings,
concentrations were greater during occupied periods and in others concentrations were
greater during unoccupied periods.
Presented in this paper are results of a study in a county-wide school district in South
Carolina. The objectives of the project were twofold. The first was to compare the results
of continuous and occupancy measurements and the second was to add to the base of data
associated with radon in schools.
METHODOLOGY
The study was performed in the Greenville County School District in northwestern
South Carolina. Previously, between 15% and 25% of the houses in Greenville County have
had radon screening measurements in excess of 4 pCi L! (148 Bq m3)(3,4). The district has
97 schools with 2963 classrooms and other frequently used rooms on or below grade. The
schools are spread over approximately 800 square miles. The survey began in late January
1992 and ended in late May 1992, with approximately 160 to 180 rooms being tested each
week.
Continuous measurements were made in each of the 2963 rooms. The measurement
period was from 6 am Tuesday until 6 am Friday for a total of 72 hours. Twenty percent of
the 2963 rooms were randomly selected for an occupancy measurement. Occupancy
measurements were made in 605 of the rooms. The occupancy measurement period was
from 6 am to 6 pm on Tuesday, Wednesday, and Thursday for a total of 36 hours.
-------
The measurements were made with S-type E-PERMS1 (electret-passive environmental
radon monitors)(5). Each E-PERM was modified to permit intermittent on/off operation of
the monitor with a programmable, battery operated, electronic timer originally intended to
operate a water flow control valve as shown in Figure la. The timer was altered to attach to
the E-PERM and to operate the plunger that turns the monitor on or off. Each monitor had a
bar-coded label which included a school identification number, school name, room number,
and type of measurement. The monitors were placed in small, ventilated, locked metal boxes
to prevent tampering. Rooms designated for occupancy measurements had both a continuous
and an occupancy monitor in each box as shown in Figure Ib. The boxes were delivered to
the schools on Mondays for distribution to the classrooms. They were retrieved on Fridays
and returned to the lab where the electret voltage was read and entered into a database along
with the information on the bar code by means of a computerized reader.2 Processing time
for the 200 to 230 measurements per week (includes continuous, occupancy, duplicates, and
blanks) was two to three hours.
a. Programmable E-PERM
Figure 1. Measurement Apparatus
b. Continuous and occupancy monitors in
metal box
The test program was conducted in accordance with the quality assurance guidelines
of EPA's Radon Measurement Proficiency Program (RMP)(6) and included 230 duplicate
continuous measurements, 60 duplicate occupancy measurements, and 120 control
measurements (blanks). A total of 4002 measurement were made during the course of the
study. At 4 pCi L1 (148 Bq m3) the measurement uncertainty was ± 0.34 pCi L1 (12.6 Bq
m3) for continuous measurements and ± 0.53 pCi L' (19.6 Bq m3) for occupancy
Rad-Elec, Inc., Frederick, MD
: Empac Incorporated, New Haven, CT
-------
measurements. The occupancy measurement lower limit of detection (operationally defined
as the concentration at which the relative error is ± 50%) was estimated to be 0.95 pCi L1
(35 Bq m3).
RESULTS
Of the 97 buildings included in the survey, 57 had at least one room with a
concentration in excess of 4 pCi L'1 (148 Bq nr3). However, most of the elevated readings
were found in a few schools. The majority of these schools were located in the northern part
of the county, which extends into the foothills of the Blue Ridge mountains.
Presented in Fig. 2 is a cumulative log-probability plot of the continuous
measurements. The median concentration was 1.7 pCi L'1 (63 Bq nr3) and the maximum
concentration was 21.4 pCi L1 (792 Bq m3). Approximately 15% of the measurements were
in excess of 4 pCi L1 (148 Bq m3). Data above the 25th percentile are almost linear on the
log-probability plot. Their distribution can be approximated as lognomal with a geometric
mean of 1.7 pCi L'1 (63 Bq m3) and a geometric standard deviation of 2.2. These results
are compared with data obtained in house surveys for Greenville County in Table 1. The
results obtained here for schools are similar to those obtained previously for houses.
Radon Cono«crtrattoo in pCVU
9 7 9 91'
Figure 2. Cumulative log-probability plot of the radon concentrations in the 2992 rooms.
-------
TABLE 1. COMPARISON OF RESULTS OF THIS SURVEY OF GREENVILLE
COUNTY SCHOOLS AND PREVIOUS STUDIES OF HOUSES.
Survey
Schools
Houses (3)
Houses (4)
Number of
Observations
2954
22
89
Arithmetic
Mean
2.4
2.8
3.1
Geometric
Mean
1.8
2.1
1.5
Percentage
> 4 pCi L ( Bq m )
15
23
15
Simultaneous continuous and occupancy measurements were made in 605 rooms. A
graphical comparison of the data is presented in Fig. 3.
20 -
0 4 8 12 16 20
CONTINUOUS CONCENTRATION (pCl/L)
Figure 3. Comparison of occupancy and continuous measurements in 605 rooms.
The ordinate is the occupancy concentration and the abscissa is the continuous concentration.
The data are almost uniformly scattered about the diagonal, and there is no indication of a
systematic difference between the two measurements. However, there were a substantial
number of rooms for which there was a difference between the occupancy and continuous
concentrations. Presented in Table 2 is the number of rooms in which the occupancy and
-------
continuous concentrations differed for various concentration ranges. The decision rule for
this table was that the concentration difference exceed 0.8 pCi L1 (30 Bq m3). This is the
sum of the uncertainty estimates for continuous and occupancy measurements which were
0.3 pCi L-' and 0.5 pCi L' (11 Bq nr3 and 19 Bq nr3), respectively. Overall, the occupancy
concentration exceeded the continuous concentration in 16 percent of the rooms and the
continuous concentration exceeded the occupancy concentration in 18 percent of the rooms.
TABLE 2. COMPARISONS OF OCCUPANCY AND CONTINUOUS
CONCENTRATIONS
Concentration
Range
0 - 1 pCi L '
1 - 2 pCi L '
2 - 4 pCi L '
> 4 pCi L '
No.
Rooms
190
292
259
100
Occupancy
> Continuous
48
43
47
32
Occupancy
< Continuous
18
52
73
26
Of particular interest are the instances in which occupancy concentrations exceeded
continuous concentrations. This is because radon exposures would be underestimated if
based upon the continuous measurement. The extremes are worth noting. In three rooms
the continuous concentration was between 1 pCi L' (37 Bq m3) and 2 pCi L'1 (74 Bq nr3) and
the occupancy concentration was over 6 pCi L1 (222 Bq nr3), in one room the continuous
concentration was 12 pCi L'1 (444 Bq nr3) and the occupancy concentration was 18 pCi L'
(666 Bq nr3), and in ten rooms the continuous concentration was less than 2 pCi L1 (74 Bq
nr3) and the occupancy concentration was over 3 pCi L1 (111 Bq nr3).
Presented in Table 3 are summary statistics for the continuous and occupancy
measurements. The arithmetic means were equal, and the percentage of measurements in
excess of 4 pCi L1 (148 Bq m3) of the two distributions were similar.
TABLE 3. SUMMARY STATISTICS FOR OCCUPANCY AND CONTINUOUS
MEASUREMENTS
Measurement
Type
Occupancy
Continuous
Arithmetic
Mean
2.4
2.4
Geometric
Mean
2.4
1.7
Geometric
Standard Dev
2.4
2.2
Percentage
>4pCiL-' (148Bqnr3)
14
13
-------
Paired difference tests were performed to determine if there was a statistically
significant difference between the occupancy and continuous concentrations. Since the data
more closely followed a lognormal distribution than a normal distribution, the tests were
applied to the natural logarithm of the concentrations. These tests were applied to different
concentration ranges (0 - 1, 1 - 2, 2 - 4, >4 pCi L1 (148 Bq m3)) and different
measurement periods (January-May, January-March, April-May). Overall, the difference
between occupancy and continuous measurements was not statistically significant (p = 0 25)
For some of the measurement periods and concentration ranges, the mean differences were '
significant and for others they were not. For the cases in which the differences were
statistically significant, sometimes the occupancy concentration was larger and sometimes the
continuous concentration was larger. The largest difference was less than ten percent Thus
the statistical analyses support the qualitative observation that there were no systematic
differences between continuous and occupancy concentrations.
SUMMARY AND CONCLUSIONS
Overall, radon levels in the schools were consistent with those in houses in the
county. The arithmetic mean continuous concentration was 2 4 pCi L' (89 Bq m3) the
^••T nlfS, ^i^ ^ and 15 t*™" °f thŁ r°°ms had «»«ntotions in excess
or 4 pci L (148 Bq m). For cumulative probabilities above 25%, the distribution of
continuous concentrations was approximately lognormal with a geometric mean of 1 7 pCi L'
(63 Bq m ) and a geometric standard deviation of 2.3.
Qualitatively, there were no obvious systematic differences between continuous and
occupancy concentrations. This observation was supported by statistical tests performed on
the mean difference between the two measurements. However, when considered on a room
by room basis, significant differences were observed in 34% of the rooms. The occupancy
concentration was greater that the continuous concentration in 16 percent of the rooms and
the continuous concentration was greater than the occupancy concentration in 18 percent of
the rooms. In a few rooms, the occupancy concentration was substantially greater than the
continuous concentration. The implication is that radon exposure in these rooms would be
underestimated significantly if based upon a continuous measurement.
REFERENCES
1. Environmental Protection Agency. Radon Measurements in Schools, An Interim
Report. EPA-520/1-89-010. U.S. Environmental Protection Agency, Office of
Radiation Programs, Washington, DC, 1989.
2. Wiggers, K.D., Bullers, T.D., Zoske, P.A., Leovic, K.W., and Saum D W
Electret ion chambers for radon measurements in schools during occupied and
-------
unoccupied periods. Proceedings of The 1990 International Symposium on Radon and
Radon Reduction Technology. February 19-23, 1990, Atlanta, GA.
3. Fjeld, R.A., Jones, M.D., and Bivens, N. Screening survey of indoor radon in South
Carolina. Health Phys. 59(2): 217, 1990.
4. Fjeld, R.A. and Shealy, H. State Indoor Radon Grant, Progress Report to April 15,
1991. Submitted to U. S. Environmental Protection Agency, April, 1991.
5. Kotrappa, P., Dempsey, J.C., Ramsey, R.W. and Stieff, L.R. A practical E-PERM
(electret passive environmental radon monitor) system for indoor K2Rn measurement
Health Phys. 58(4): 461, 1990.
6. Environmental Protection Agency. Radon Measurement Proficiency (RMP) Program,
Handbook. EPA 520/1-91-006. U.S. Environmental Protection Agency, Office of
Radiation Programs, Washington, DC, 1991.
-------
VP-6
INDOOR RADON IN NEW YORK STATE SCHOOLS
PRELIMINARY REPORT
Susan VanOrt*. Charles Kunz+, Laurence Keefe*,
William Condon*, Kirk Fisher+ and Karim Rimawi*
*New York State Department of Health
Bureau of Environmental Radiation Protection
2 University Place
Albany, New York 12203
+New York State Department of Health
Wadsworth Center for Laboratories and Research
Empire State Plaza
Albany, New York 12201-0509
ABSTRACT
New York State is participating in a project to study radon in schools
funded in part through a grant from the EPA. The program began in the spring
of 1991 and is scheduled to run for three years. Candidate schools are
selected from areas in which existing information suggests there may be a high
risk for indoor radon. These schools are invited to participate in an indoor
radon survey that includes short term screening measurements, one month to
three month follow-up measurements in rooms with greater than or equal to
4 pCi/1, and one year measurements. Follow-up measurements will be made if a
school is mitigated. Twenty-two schools are being surveyed in the first year
(1991/1992) and forty in the second. The soils under and around about one-
third of the schools surveyed will be characterized for indoor radon potential
through soil gas measurements and examined for correlation with indoor radon
concentrations.
The information presented in this paper details initial results of short-
term screening measurements made in twenty-two schools and soil
characterization in three of these schools.
This paper has been reviewed in accordance with the United States
Environmental Protection Agency's peer and administrative review policies and
approved for presentation and publication.
Mention of trade names or commercial products in this document does not
constitute endorsement or recommendation for their use.
-------
BACKGROUND
Indoor radon measurements in single family houses throughout New York
State have revealed many areas with geometric mean radon concentrations
greater than 4 pCi/1. Consequently, the New York State Department of Health
(NYS DON) has become increasingly concerned with the radon-related health risk
to school-age children in the areas with high radon potential. The purpose of
this study is to screen indoor radon concentrations in school buildings in
areas of the state known to have elevated levels of indoor radon. The study's
main objectives are to 1) obtain information regarding the extent and magnitude
of the radon risk in New York State schools, 2) learn more about how large
buildings behave with regard to radon entry and distribution, and 3) provide
instruction about testing procedures and temporary remediation techniques to
school officials so that they will be able to conduct radon tests on their own
in other school buildings.
SCHOOL SELECTION
The two basic criteria used to select candidate schools are that the
schools should be located in high risk areas for above average indoor radon
and that the schools are distributed around the state to provide information to
as many school districts and areas as possible. For the past several years,
the NYS DOH has been studying the geologic factors affecting indoor radon
throughout the state and at the same time the Department has been
accumulating a large database on indoor air radon concentrations in
residences across the state. The geologic data and the extensive database of
indoor radon screening measurements in homes are used to select schools from
areas with potential for above-average indoor radon. Cities, towns and
villages containing schools are matched with the cities, towns and villages
with the highest levels of indoor radon. Schools in areas with high levels of
residential radon are then located on topographical and surficial geology maps
to obtain information on geologic factors such as gravelly, permeable soils
that often correlate with above-average indoor radon. Schools that are
located in areas with above-average indoor radon and/or areas with surficial
geologic factors that correlate with above-average indoor radon are considered
as candidate schools. Generally, only one school per school district is
selected so that schools are distributed throughout the state while still
targeting high-risk areas.
In the spring of 1991, 53 candidate schools were selected and asked to
volunteer for participation in the school survey project. Twenty schools
volunteered and participated in the project. Two additional schools not
included in the 53 candidate schools also volunteered and were measured in the
first year. Several schools declined participation in the first year but
asked to be considered for the second year of the project.
-------
Candidate schools are sent a letter explaining the indoor radon testing
program and requesting their participation. Those wishing to participate in
the program are asked to fill out a questionnaire of general information about
the school building and to return copies of floor plans. This information is
used in the selection process and to develop a measurement strategy for the
selected school buildings.
Once a school building is selected, the school is notified and requested
to designate an individual from the school as the primary contact We prefer
that this individual take an active role in the program from initial screening
tests to post remediation testing. This person will thus, become knowledgeable
enough to carry out or supervise testing in other school buildings in the
district. a
Figure 1 shows the average basement screening radon concentrations for
each County in the state and the number of schools surveyed in each County
during the first year of the project. The twenty-two schools participating
in year one of the project are located in seventeen Counties Nine of the
schools were located in Counties with average basement screening concentrations
above 8 pCi/1, twelve in Counties averaging between 4 to 7.9 PCi/l and one in
a County with between 2 to 3.9 pCi/1.
When testing is about to commence, a box is shipped to the school that
includes detectors, detector deployment instructions, data sheets and a floor
plan that is marked with room location and quality assurance detectors The
detector deployment instructions explain where to place the detectors 'how to
record information on the data sheets, how long to test for and where'to ship
the detectors at the end of the test period. In general, this distribution by
mail has been successful in 1991/1992 and we plan to continue with this format
tor rear Z.
-------
MEASUREMENT PROTOCOLS
All schools have long-term (one year) alpha track detectors1 and
short-term (three day) E-PERMs2 placed in every room at and below grade and
a sampling of rooms on second and third floors. E-PERMs are utilized for
short-term (three day) and follow-up measurements (one and three months).
Table 1 details the follow-up measurement protocols. Additional detectors
are also provided to the school for placing duplicates in about 10% of the
rooms for quality control. An in depth quality assurance plan for E-PERM,
alpha track detectors, charcoal and continous monitoring units is in place.
If radon in a routinely occupied room measures greater than 100 pCi/1,
Health Department personnel visit the school to collect grab samples, conduct
follow-up inspections to identify obvious entry points, deploy additional
short-term electrets and possibly collect some long-term electrets. Based on
these results, the school will be advised regarding possible temporary
remediation, further testing and eligibility for inclusion in the New York
State Energy Office school diagnostics program.
If schools participating in the measurement program undergo any
remediation to reduce indoor radon concentrations, the Health Department will
supply short and long-term electrets for post-remediation measurements
following each phase of the remediation process.
1MAn ATD is a small piece of plastic or film enclosed in a container with
a filter-covered opening. Radon diffuses through the filter into the
container and alpha particles emitted by the radon decay and its products
strike the detector and produce submicroscopic damage called alpha
tracks."(1)
2An E-PERM (Electret-Passive Environmental Radon Monitor) System. "The E-PERM
Chamber is a precision, volumetric flask made of conductive plastic which
provides the ion collection chamber for electret-ion-chamber measurement of
radon. When paired with the electret, these components comprise a working
E-PERM radon measurement unit...Each E-PERM Electret is an electrically
charged wafer of Teflon (Dupont) which has been treated to hold a stable
electrostatic potential."(2)
-------
SCREENING RESULTS
Screening results for the twenty-two schools tested in New York State
during the 91-92 school year are shown in Table 2.
In twelve schools, which each had more than 20% of the rooms tested with
greater than or equal to 4 pCi/1, the percentage of rooms with greater than
or equal to 4 pCi/1 was about the same for the basement (46%) and ground level
(49%). Although only a few rooms were measured on floors above the ground
level, 35% of the rooms tested on the second floor in these twelve schools
were measured with greater than or equal to 4 pCi/1 indoor radon. These
results indicate that for schools with above average indoor radon the
percentage of rooms with greater than or equal to 4 pCi/1, is about the same
for basement and grade levels. The high percent of rooms with concentrations
greater than or equal to 4 pCi/1 for above grade levels could have resulted
from selecting rooms which have high air mixing with lower levels (See Table
3). The average basement, first floor (grade level) and second floor radon
concentrations for all twenty-two schools are 7.1, 3.4, and 2.6 pCi/1
respectively. Figure 3 represents the percentages of rooms with radon levels
greater than or equal to 4 pCi/1 by level. The percentages by floor are as
follows; basement 42%, first floor 39% and second floor 53%. Testing
conducted on the second floor represents only a sampling of the total rooms
on that floor.
Measurements conducted in all twenty-two schools indicate that 345 of 1116
rooms had levels greater than or equal to 4 and less than 20 pCi/1 of indoor
radon and 11 had levels greater than or equal to 20 pCi/1 (See Table 2). In
many cases the rooms with greater than or equal to 20 pCi/1 were storage areas
or basement rooms with little ventilation and low occupancy. We observed a
similar trend in residential indoor concentrations in areas with gravelly soils
in New York State, that is, a high percentage of homes with radon levels near
20 pCi/1 but few homes with greater than 40 pCi/1.
Figure 2 compares radon concentrations by floor for all twenty-two
schools. Seven of twenty-two schools did not have basement readings.
Therefore the total number of schools for first floor readings in each
category is higher. The graph also demonstrates the decrease in the number
with radon levels at higher concentrations.
Due to the high percentage of rooms with concentrations greater than
or equal to 4 pCi/1 on second and third floors, we propose to revise the
measurement protocol for Year 2 and will place alpha track detectors in all
rooms on these floors.
-------
SOIL-RADON POTENTIAL
Soil characteristics relating to the potential for radon infiltration
into buildings will be measured at about one-third of the schools
participating in the survey. Soil measurements include surficial gamma,
radionuclide concentrations (Ra-226, U-238, Th-232), soil-gas radon
concentrations, and the permeability of the soils for gas flow (3).
Measurements are made at various locations around the perimeter of the
schools and at various distances from the school. When accessible,
measurements are also made in the crawlspaces of the schools. It is of
interest to determine if the same soil parameters that cause above average
indoor radon in single-family homes have the same effect on larger buildings.
This information in combination with building design will aid in further
identification of schools with high radon potential.
Initial visits and soil measurements have been made at three schools.
During the summer and fall of 1992 additional field trips will be made to
complete the soil characterization at these schools. Although incomplete, the
results from the initial visits are presented in Table 4.
The soil Ra-226 concentrations listed in table 4 are the averages for 3
to 6 samples collected at a depth of 60 cm from under and around each school.
The mean Ra-226 concentration for soils in the U.S is 1.0 pCi/g (4).
Therefore the soils under and around school ONO-1 in Onondaga County are near
the mean, whereas the soils at COL-1 in Columbia County and WAS-1 in Washington
County have below average concentrations of Ra-226.
Soil-gas samples were collected at a distance of 0.5 m from the schools
foundation, from crawlspaces and from locations more than 10 m from the
school. Since air being drawn through the soil near the foundation and
possibly through the soil in the crawlspaces can significantly dilute the
soil-gas Rn-222 concentration, only samples collected at locations more than
10 m from the school were used to determine the soil-gas Rn-222 concentration.
Most of the measurements used were from samples collected at a depth of 120
cm, however, some of the measurements used were from depths between 45 to
90 cm, due to a high water table or rocks preventing deeper sampling. The
values listed in Table 4 represent soil gas measurements from 5
locations at the school in Onondaga County and only one location for the
schools in Columbia and Washington Counties. Therefore the values listed in
Table 4 cannot be considered as very representative for the soils around the
schools. Even so, the values listed for the three schools are typical for
soils with average to somewhat below average Ra-226.
-------
Since rubble and fill are often found near a school's foundation and in
the crawl spaces only permeability measurements taken more than 10 m from the
school are listed in table 4. Although the values listed in table 2 are
averages for only a few measurements they are useful for an approximate
characterization of the permeability for the soils around the schools. School
ONO-1 in Onondaga County is located in outwash gravel along a creek.
Measurements were made at five locations more than 10 m from the school with
an average permeability for gas flow of 2.9 x 10-6 cm . The soils are quite
gravelly and at one location a high water content in the soil reduced the
permeability by three orders of magnitude. School COL-1 in Columbia County is
located about 400 m from a creek on a small plateau. The permeability
measured at greater than 10 m from the school was 1.8 x 10-7 cm . The soils
contain some stone but are predominantly silty sand and appeared to be well
drained. At school WAS-1 in Washington County the permeability of the soils
for gas flow was measured at 2 x 10-9 cm . The soils were silty with some clay
and were nearly saturated with water at depths greater than one meter.
The most important soil characteristic correlating with indoor radon
concentrations at the three schools studied appears to be the permeability of
the soils for gas flow. At School ONO-1 in Onondaga County where 20% of the
rooms measured had Rn-222 concentrations greater than or equal to 4 pCi/1, the
soils are gravelly and quite permeable. At School COL-1 in Columbia County
where 6% of the rooms measured had Rn-222 concentrations greater than or equal
to 4 pCi/1 the permeability was moderate and the soils were well drained. At
School WAS-1 in Washington County the permeability of the soils was low, the
water table was high and only 1% of the rooms were greater than or equal to
4 pCi/1.
A further indication that highly permeable gravelly soils correlate with
above average indoor radon in New York State schools is obtained by
considering the location of the schools with the highest indoor levels. The
four schools with the greatest percentage of rooms with greater than or equal
to 4 pCi/1 are located in Tioga, Steuben and Cortland Counties (Table 2).
These three Counties are among the seven Counties with the greatest average
residential indoor Rn-222 concentrations in New York State. These three
Counties are located in the south central region of the State. The most
recent glacier formed a large morain just north of these Counties (Valley
Heads Morain) and when the glaciers melted large quantities of water flowed
down through the valleys in these Counties carrying debris from the morains
into the valleys, forming deep gravel deposits in many areas. These highly
permeable gravel deposits are the predominant factor resulting in above
average residential indoor radon in this region of the State (5). It appears
that the gravelly soils are also resulting in above average Rn-222 in the
schools in this region. One of the schools in Steuben County has been visited
and although soil measurements have not yet been made, the soil under and
around the school was observed to be very gravelly.
-------
CONCLUSIONS
The first year of our project has been quite successful. We have
targeted high risk areas of New York State based on our residential data and
have found schools in those areas to also have elevated levels. In some cases
we have found that high concentrations exist at more levels in a school and
therefore we will be implementing more detector distribution at these levels
in Year 2.
The distribution by mail has worked quite well and we will continue it
into Year 2 with some slight modifications in instructional materials for ease
of reading. Additionally, we will be discussing the floor plans in greater
detail in Year 2 to prevent duplication of our efforts.
Since EPERM and alpha track measurements (charcoal if deemed necessary)
and continuous radon monitoring units are being utilized for this study we
cannot overemphasize the need for a strong quality assurance program as an
integral part of this type of study.
In the Fall of 1992, we will be receiving results of our year long alpha
track measurements and additional soil characterizations, which will complete
our assessment of these twenty-two schools.
-------
REFERENCES
1. "Radon Measurements in Schools, An Interim Report." EPA-520/1-89-010,
U.S. Environmental Protection Agency, Washington, D.C., March 1989. p A-6.
2. Rad Elec Inc.," E-PERM Components Electret and Ion Chamber."
3. Kunz, C., "Influence of Surficial Soil and Bedrock on Indoor Radon in NYS
Homes.", NYSERDA Report 89-14, October 1989.
4. Myrick, I.E., Bervin, B.A. and Haywood, F.F., "Determination of
Concentrations of Selected Radionuclides in Surface Soil in the U.S.",
Health Physics, Vol. 45, No. 3, pp 631-642 (1983).
5. Kunz, C., Laymon, C. and Parker, C., "Gravelly Soils and Indoor Radon",
Proceedings EPA 1988 Symposium on Radon and Radon Reduction Technology,
Denver, CO, Oct 1988.
-------
AVERAGE RADON
(Basement
BY COUNTY
Only)
A*orifajr6. 1001
J O.I 19
-40 79
39
15 «*
FIGURE 1. NUMBER OF SCHOOLS PARTICIPATING IN YEAR ONE
-------
3
Ł
cc
a
14-
* *>
12-
4 4%
10-
-
-
-
•
^
.
1
1 — n — 1
1
1
1
1
I
\
1 — i i r'
1
j
^
I
I
i
I
1
-------
60
50
UJ
g 30
UJ
0.
20
10
1 r
BASEMENT
FIRST
FLOOR
SECOND
FIGURE 3. PERCENTAGE OF ROOMS > 4pd/l
-------
TABLE 1.
FOLLOW-UP MEASUREMENTS
SCREENING LEVEL
< 4 pCi/1
> 4 < 20 pCi/1
> 20 pCi/1
FOLLOW-UP MEASUREMENT PERIOD
Three month EPERM
One month EPERM
* These follow-up measurements are made in
addition to one year alpha track detectors
-------
TABLE 2. RADON LEVELS IN SCHOOLS - SCREENING RESULTS
SCHOOL/ AVG RN # OF ROOMS # OF ROOMS # OF ROOMS % OF ROOMS
COUNTY CONC (pCi/1) TESTED . > 4 < 20 ^20 > 4
AL?-1 5.6 90 38 4 46 7
ALBANY
BRO-1 4.5 54 22 0 40.7
BROOME
CHE-1 5.0 36 14 1 41.7
CHENANGO
CHE-2 6.4 44 16 1 38.6
CHENANGO
CHE-3 2.6 37 7 0 18.9
CHENANGO
COL-1 2.3 33 2 0 6.1
COLUMBIA
COR-1 5.8 51 30 1 60.8
CORTLAND
DEL-1 2.0 51 7 0 13.7
DELAWARE
DUT-1 3.6 29 8 0 27.6
DUTCHESS
DUT-2
DUTCHESS 2.2 53 5 0 9.4
-------
TABLE 2 continued
SCHOOL/ AVG RN # OF ROOMS # OF ROOMS # OF ROOMS % OF ROOMS
COUNTY CONC (pCi/1) TESTED > 4 < 20 > 20 ^4
LEW-1
LEWIS
LIV-1
LIVINGSTON
ONO-1
ONONDAGA
ONO-2
ONONDAGA
REN-1
RENSSELAER
SCH-1
SCHOHARIE
STE-1
STEUBEN
TIO-1
TIOGA
TIO-2
TIOGA
ULS-1
ULSTER
WAS-1
WASHINGTON
1.3 34 0 0 0.0
3.7 69 32 0 46.4
2.4 56 11 0 19.6
1.4 56 30 5.4
3.5 48 20 0 41.7
1.0 54 1 0 1.9
8.1 52 37 1 73.1
6.2 76 54 1 72.4
6.1 49 26 2 57.1
3.7 29 8 0 27.6
1.2 84 1 0 1.2
-------
TABLE 2 continued
SCHOOL/
COUNTY
WAY-1
WAYNE
ALL
SCHOOLS
AVG RN
CONC (pCi/1)
1.4
AVG RN
CONC (pCi/1)
# OF ROOMS
TESTED
31
# OF ROOMS
TESTED
# OF ROOMS # OF ROOMS °A
> 4 < 20 > 20
3 0
# OF ROOMS # OF ROOMS
> 4 < 20 > 20
', OF ROOMS
> 4
9.7
% OF ROOMS
> 4
22
3.7
1116
345
11
32
-------
TABLE 3. SCREENING RESULTS
LEVEL # OF SCHOOLS WITH % OF ROOMS > 4 pCi/1
OVER 20% OVER 40% OVER 60%
0
1
2
8
12
6
6
9
4
3
4
4
FOR THE 12 SCHOOLS WITH OVER 20% OF ROOMS ^ 4 pCi/1, THE FOLLOWING
IS TRUE;
LEVEL % OF ROOMS > 4 pCi/1
0 46
1 49
2 35
-------
TABLE 4. SOIL RADON POTENTIAL DATA
SCHOOL SOIL SOIL-GAS
(COUNTY) RA-226 RN-222
(pCi/g) (pCi/1)
PERMEABILITY
2
(cm )
ROOMS
> 4 pCi/1
ONO-1 0.96
ONONDAGA
COL-1 0.68
COLUMBIA
WAS-1 0.73
WASHINGTON
569
589
580
-6
2.9 x 10
-6
0.18 x 10
-6
0.002 x 10
20
-------
Session VI Posters
Transport and Entry Dynamics of Radon
-------
VIP-1
SIMPI.TFTED MODELING OF THE EFFECT OF GUFPLY VENTILATION ON INDOOK
-KADON "
by David S
Infiltec
Falls Church, VA 22041
Mark Modei-a
Lawrence Berkeley Laboratory
Berkeley, CA
Kelly Leovir.
US EPA ORD
RTF, NC 27711
ABSTRACT
This paper investigates a simplified model of the effect of
supply ventilation on indoor radon concenhrat ions. The model
assumes a slab-on-gradc structure, a single well-mixed zone, a
ventilation fan inducing an air flow into the building, winter
ctack effect weather conditions/ and a constant, radon concentration
under the slab. The radon mitigation effect of supply ventilation
is separated into a dilution effert and a change in radon entry
rate due to prcoaurization of the buildinq shell. The modftl is
used to assess f.he radon mitigation impact of increasing the
ventilation rale to meet ASHKAti standard 62-1989 recommended
ventilation levelc. This model was developed to aid in t-.he
understanding the results of 1991 EPA reecarch in Virginia and
Maryland schools into the use of HVAC system modifications as a
radon mitigation technique.
-------
VIP-2
DETERMINATION OF MINIMUM COVER THICKNESS FOR
URANIUM MILL TAILINGS DISPOSAL CELLS
by: D.W. Andrews and J.P. Ambrose
CWM Federal Environmental Services, Inc.
Albuquerque, NM 87106
ABSTRACT
The Uranium Mill Tailings Remedial Action (UMTRA) Project was
initiated by the UMTRC Act passed by congress in 1978 and is
expected to extend into the late 1990's. UMTRA is involved in mill
tailings remediation activities throughout the continental United
States. The remediation project involves the stabilization of the
tailings material into massive disposal cells. These disposal
cells have been designed to endure environmental forces for 1,000
years. Once all the tailings are deposited into the disposal cell
a radon barrier is applied. Prior to placement of the radon
barrier cover on the disposal cell, soil samples are collected at
0.6 meters intervals to a depth of 6.1 meters. These samples are
used to determine the radon emanation fraction, Radium-226
concentration, Thorium-230 concentration, and moisture content.
These measurements along with other field measurements are used in
the RAECOM radon flux computer modeling program to determine
minimum cell cover thickness. Additional flux measurements are
performed to verify compliance with Subpart T, NESHAP regulations.
INTRODUCTION
Public Law 95-604 known as the Uranium Mill Tailings Radiation
Control Act (UMTRCA) was passed by congress in November of 1978.
The Act required the remediation of residual radioactive materials
at some former uranium mill tailings processing sites to aleviate
a potential long term health hazard. The UMTRCA charged the
Secretary of Energy with the remediation of the designated sites
and the Administrator of the Environmental Protection Agency to
establish the standards for clean-up. The Nuclear Regulatory
Commission is required to approve the long-term surveillance and
monitoring plans and certify the disposal sites.
-------
The Department of Energy was to enter into cooperative
agreements with the states and indian tribes to acquire the lands
and perform the remedial actions. Remedial actions involve
demolition and decommissioning of existing mill facilities and
consolidation of all associated residual radioactive materials,
including windblown materials, into a designed disposal cell.
The Environmental Protection Agency established the design
standards for tailings disposal cells. The disposal cell must be
designed in a manner which prevents the background radon levels
from increasing by more than 0.5 pCi/1 or limit the average cell
radon flux to 20 pCi/m2-s. The disposal cell shall be designed to
maintain the above standards for 1000 years and in no case less
than 200 years. Demonstration of compliance is not required.
To achieve the long-term disposal requirements the cells are
constructed of only natural materials. The radon barrier cover
layers are usually constructed of high clay content soils to limit
radon diffusion and to inhibit infiltration of moisture into the
tailings material. Inhibiting the infiltration of water provides
some frost protection of the tailings embankment and prevents
excessive leaching and groundwater contamination. The radon
barrier is then covered with high permeable sand layers to allow
lateral flow of water. The side slopes are covered by layers of
rock with low fracture potential to armor the cell against erosion.
Vegetation is often used on top slopes for erosion protection
although root development can produce conduits for radon transport
to the surface.
The thickness of the radon barrier is not only determined by
the radon diffusion characteristics of the cover material but also
by other design criteria such as frost and erosion protection. The
combination of the above design criteria is used to achieve the
1,000 year disposal requirement. The radon barrier thickness
determination methods presented below are used to determine the
minimum thickness allowed for controlling radon emissions into the
ambient air.
METHOD FOR COVER THICKNESS DETERMINATION
In achieving the design standard prescribed in the federal
regulations for uranium disposal cells, the UMTRA Project utilizes
a radon flux computer modeling program called RAECOM. Given
specific soil parameters and 226Ra concentrations for each layer of
material associated with the disposal embankment, the RAECOM
program calculates radon flux using a one-dimensional steady-state
diffusion equation (1).
-------
PARAMETERS
The radon barrier cover thickness design for a long term
disposal cell is dependent on soil density, specific gravity,
moisture, thickness, M6Ra concentration, 222Rn diffusion, and
emanation coefficients (2). For calculation purposes these
parameters can be measured directly at the specific site or
reference values can be utilized. If reference values are used,
the NRC requires justification and demonstration of conservatism
for their use at a specific site. Generally, the UMTRA Project
uses only measured parameters in the flux calculations.
The parameters used in the RAECOM computer program are soil
layer thickness, porosity, moisture content, 226Ra concentration,
density, emanation coefficient, diffusion coefficient, and ambient
air radon concentration. The modeling system allows the use of
bottom flux which is the radon flux entering the lowest layer of
the model. The bottom flux is usually ignored since twenty feet of
tailings is considered an infinite thickness for calculation
purposes (1).
In calculating cover thicknesses, easily measured parameters
such as soil moisture content, soil density, soil thickness, radon
emanation coefficient, and 226Ra concentration are obtained as the
embankment is constructed. These parameters are determined using
field or on-site laboratory measurements. Diffusion coefficients,
which are more difficult to measure, are obtained by sending soil
samples to geotechnical laboratories for analysis. Ambient air
radon concentration is measured in the vicinity around the disposal
site.
As the tailings embankment and radon barrier cover are being
constructed, soil density measurements using the standard sand cone
method1 or the nuclear density gauge method2 are performed to
demonstrate the quality of embankment construction. These density
measurements are also used for the RAECOM calculations.
Porosity is calculated using soil density and specific gravity
measurement information. The following equation is used to
calculate porosity:
P - d, / (s, * dj
where: P = Porosity
d, = Density of Soil
'ASTM Standard Test Method D-1556, Density and Unit Weight of Soil
in Place by the Sand-Cone Method (3).
2ASTM Standard Test Method D-2922, Density of Soil and Soil-
Aggregate in Place by Nuclear Methods (Shallow Depth) (3).
-------
s, = Specific Gravity of Soil
dw = Density of Water
Percent moisture content of tailings material is measured in
the field using the oven-dry method1. Percent moisture content for
cover materials are calculated using a long-term moisture content
equation or measuring the moisture equilibrium potential for a
particular soil using the 15 bar moisture method2.
Radon emanation coefficients for the tailings material are
obtained at the on-site laboratory by analyzing samples collected
at 0.6 meter intervals in the disposal cell soil column. These
samples are sealed, stored to achieve equilibrium, analyzed, opened
and reanalyzed after a venting period. The results of this
analysis are used to calculate the emanation fraction. The 226Ra
concentration of the tailings material samples is obtained from the
first analysis prior to venting.
As disposal embankments are constructed soil samples are
collected at 0.6 meter interval elevations from established grid
locations. Soil sample collection is initiated at an elevation 5.5
meters below the final design elevation of the top of the tailings
embankment. Sample material is consolidated over a 0.6 meter
depth, homogenized and then split into two samples. Sample
collection continues as new layers of tailings are placed in the
embankment until the final elevation is attained at each location
which forms a sample matrix of the top 6.1 meters of the completed
disposal cell. If portions of the tailings embankment exist when
sample collection is initiated borehole samples are collected at
required elevations below the existing surface.
After the collection process, soil sample cans are sealed and
leak tested as soon as possible to minimize soil moisture loss and
to prevent radon gas leakage. Of the two samples collected at each
matrix grid point, one is analyzed for 230Th and the other is stored
for at least 20 days to allow 222Rn to achieve equilibrium with 226Ra.
After the storage period the sample is analyzed by gamma
spectrometry for 2I4Bi activity. The sample is then opened, oven
dried and allowed to vent for 4 to 24 hours. Bismuth-214 was
chosen for analysis since it produces an easily measurable gamma
peak and reaches equilibrium with 222Rn in a relatively short time.
'ASTM Standard Test Method D-2216, Laboratory Determination of Water
(Moisture) Content of Soil and Rock (3).
2Two capillary moisture methods are used depending on soil particle
size: ASTM Standard Test Method D-2325, Capillary-Moisture
Relationships for Coarse- and Medium-Textured Soils by Porous-
Plate Apparatus or ASTM Standard Test Method D-3152, Capillary-
Moisture Relationships for Fine-Textured Soils by Pressure-
Membrane Apparatus (3).
-------
The sample is then resealed and immediately reanalyzed for 2l4Bi
activity. The emanation coefficient is then calculated using the
following formula:
EC = (RA, - RA2) / RA,
where: RA, = Wet equilibrated 214Bi activity
RA2 = Dry unequilibrated 214Bi activity
EC = Emanation coefficient
Emanation Coefficients for the radon barrier material are measured
at an off -site laboratory using the same method described above.
The ^^a concentrations of tailings material are obtained by
dividing the RA, activity from the above equation by the dry weight
of the sample. Radon barrier 226Ra concentrations are obtained at
an off-site laboratory when samples are analyzed to determine the
emanation coefficients.
The 226Ra concentrations at a 1,000 years must be estimated
before the RAECOM calculations are performed. The existing 230Th
concentration is used in the above estimated 226Ra concentration.
The 1,000 year 226Ra concentration is determined by using the
following equation:
* 0.65) + (TH0 * 0.35)
where: RA10oo = Estimated 226Ra Concentration in
1,000 years
RAo = Initial 226Ra Concentration
TH0 = Initial 230Th Concentration
The equation given above is a reduced version of a secular
equilibrium equation used to decay parent /progeny combinations (8) .
Diffusion coefficients are obtained at an off-site laboratory
by conducting transient diffusion measurements (9) . Radon is
allowed to diffuse though a compacted soil column. Activity
measurements are performed at the opposite end of the column at
different time intervals until the they reach an equilibrium point.
OPTIMIZATION
The radon barrier thickness is optimized by utilizing the RAECOM
radon flux program and parameters representing worst case
conditions. Optimization is carried out by varying the radon
barrier thickness to achieve a resultant radon flux of just less
than 20 pCi/m2-s. The optimized radon barrier thickness is the
minimum thickness allowed in the disposal cell design.
-------
RADON FLUX MEASUREMENTS
Subpart T of the National Emission Standards for Hazardous Air
Pollutants (NESHAPs)1, requires owners of uranium mill tailings
piles to demonstrate compliance with radon flux emission standards.
Under these standards the UMTRA Project is required to dispose of
the tailings such that the measured flux on the surface of the
tailings embankment does not exceed 20 pCi/m2-s. The flux
measurements are performed under the prescribed conditions and
according to the regulatory method.1
Disposal embankment design is based on a worst case situation
over a thousand year period. The radon flux measurement cannot be
used to demonstrate the effectiveness of the 1,000 year disposal
embankment design. However, the radon flux measurement can be
utilized as an indicator of design effectiveness.
0.5
Ifl
Ł
\
r-(
(J
a
0.4-
0.3
0.2-
o. i-
0 10 20 30 40 50 60 70 80 90 100
LOCATION NUMBER
AUERAGE
Figure 1. Radon Flux Measurements for Lowman, Idaho Site
'The NESHAPs are part of the requirements of the Clean Air Act.
-------
Figure 1 illustrates the results of radon flux measurements
from the completed Lowman, Idaho UMTRA disposal cell. The highest
flux measurement was just above 0.4 pCi/m2-s and the average was
approximately 0.05 pCi/m2-s. Flux measurements are typically well
below the standard because the cell is designed for worst case
conditions. Moreover, the radon barrier may be thicker than the
optimized thickness due to other design requirements such as frost
and erosion protection. This was the case for the Lowman disposal
cell. The Lowman worst case situation produced an optimized radon
barrier thickness of only 13 cm, whereas the actual minimum design
thickness was 46 cm (10).
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. Rogers, V.C., Nielson, K.K., Kalkwarf, D.R. Radon Attenuation
Handbook for Uranium Mill Tailing Cover Design. NUREG/CR-
3533. U.S. Nuclear Regulatory Commission, Washington, D.C.,
1984.
2. U.S. Department of Energy, Technical Approach Document,
Uranium Mill Tailings Remedial Action Project, Albuquerque, NM
87108, May 1986.
3. American Society for Testing and Materials, 1992 Annual Book
of ASTM Standards, Section 4, Construction, Volume 04.08,
Philadelphia, Pennsylvania, 1992.
4. Turner, J.E., Atoms, Radiation, and Radiation Protection.
Pergammon Press Inc., Maxwell House, Fairview Park, Elmsford,
New York, 1986. p. 65.
5. Nielson, K.K., Rich, D.C., Rogers, V.C., and Kalkwarf, D.R.,
Comparison of Radon Diffusion Coefficient Measured by
Transient Diffusion and Steady-State Laboratory Methods.
NUREG/CR-2875. U.S. Nuclear Regulatory Commission,
Washington, D.C., 1982.
6. Lowman Idaho, Final Design for Construction - Calculations,
Prepared by Morrison Knudsen Corporation, Environmental
Services Group, for the Department of Energy, Uranium Mill
Tailing Remedial Action Project, Albuquerque, New Mexico,
1991.
-------
VIP-3
A MATHEMATICAL MODEL DESCRIBING RADON ENTRY AIDED BY AN
EASY PATH OF MIGRATION ALONG UNDERGROUND TUNNELS
Mosley, R. B.
USEPA
Air and Energy Engineering Research Laboratory
Research Triangle Park, NC 27711
ABSTRACT
Most houses are connected to the soil through underground
tunnels of higher permeability materials surrounding such things as
sewer lines, drain lines, water lines, electrical entrances, etc.
In some cases, these buried lines may provide effective migration
paths for radon to approach or enter the houses. These easy
transport paths are likely to be most important when the house is
surrounded by relatively low permeability soil such as dense clay.
A model is presented to evaluate the enhanced migration rate of
radon as a result of these transport tunnels.
-------
VIP-4
RADON DIFFUSION STUDIES IN SOIL AND WATER
By: Manwinder Singh
Surinder Singh
H. S. Virk
Department of Physics
Guru Nanak Dev University
Amritsar 143 005, India
ABSTRACT
In soil, the contribution comes by way of diffusion and
transport of radon and thoron from a distant source and radium and
thorium present in the immediate vicinity. In the case of natural
waters, radon is contributed by migration and dissolved radium^
content in the water. In the current investigation, a simulation
study has been carried out in the laboratory on the behavior of
radon diffusion through soil and water. The diffusion rate of
radon through soil has been studied by measuring the rate of
buildup with time at varying thicknesses and porosities of the
soil. The diffusion coefficient, D, and mean diffusion lengths
range from 0.0027-0.0072 cm2/sec and 65-82 cm for porosities of 30
to 50 percent, respectively. The radon flux follows the
exponential fall with the increase in overburden on the source.
The experimentally-observed values at different thicknesses are
lower than the theoretically-calculated values. The radon
activity may be suppressed partly due to the diffusion leakage and
partly because of the back diffusion.
-------
VIP-5
Stack Effect and Radon Infiltration
by: Craig DeWitt
Agricultural and Biological Engineering
Clemson University
Clemson, SC 29634-0357
(803) 656-4041
(803) 656-0338 (fax)
ABSTRACT
Infiltration of outside air as a result of pressure
gradients caused by thermal buoyancy, or the "stack effect",
plays a factor in bringing radon gas into a home. In some
instances, reducing infiltration can actually increase the radon
levels or potential radon problems in a house. By taking
advantage of the principles behind the stack effect, reducing
infiltration can be an effective radon mitigation technique and
significantly reduce the forces drawing radon into a house.
Sealing openings in the ceiling or upper locations of a
building will lower the neutral pressure level within a building.
Lower neutral pressure levels result in reducing infiltration
rates as well as lower pressure difference across infiltration
openings. Depressurization fields of SSD and ASD systems will
also increase in strength.
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.
INTRODUCTION
Radon emanation from naturally occurring soils into the soil
air, which can subsequently be transported into a building is a
source of indoor air pollution. Two conditions must exist for
radon-laden soil-air entry into a building: 1) there must be
openings in the building which couple the soil-air with indoor
air and 2) there must be a driving force that results in a flow
of the radon-laden soil-air. Major strategies for reducing or
preventing high levels of radon in structures include 1) dilution
through increased ventilation, 2) sealing entry pathways into the
structure, 3) building pressurization, 4) soil depressurization,
and 5) sub-slab or sub-membrane depressurization (1).
-------
The dilution through ventilation theory is based on a simple
model; if the radon entry rate Qrn, is assumed to be constant and
set equal to the removal rate, we have;
Qrn = V * Crn
where V is the ventilation rate and Crn is the radon
concentration. Based on this relationship, any decrease in radon
concentration must involve a proportional increase in ventilation
rate. To halve the radon concentration would require doubling
the ventilation rate. Problems encountered with this strategy
include 1) knowing the existing ventilation rate, and 2) doubling
or quadrupling the ventilation rate is about the maximum
acceptable level in terms of energy use and human comfort.
The other radon reduction strategies listed above attempt to
reduce the radon entry rate. This radon entry rate is a function
of the source strength, leakage area, and pressure difference
across the leakage area as the driving force, as related by:
Qrn=f(Sm, ELA, AP) (2)
where Spn is the source strength, ELA is the equivalent leakage
area, and AP is the pressure difference. Since changing the
source strength is not usually a viable option, the only feasible
options are to attempt to alter leakage areas or the pressure
differences. Soil, sub-slab, sub-membrane depressurization and
house pressurization strategies attempt to reduce (or even
reverse) the pressure differences across the leakage area or
radon entry sites.
Crack sealing as a mitigation/prevention strategy attempts
to reduce the leakage area of the radon entry sites. This
technique has had inconsistent results probably due to the
inability to effectively locate and seal all entry points. It
has been observed that even small openings are sufficient to
allow unacceptable radon levels (2) . Crack sealing has become a
part of other mitigation strategies though, by helping to extend
the pressure fields developed in the soil by sealing re-
entrainment paths and preventing "short-circuiting" through slabs
and other building shell membranes (3) .
Both soil depressurization and sealing strategies attempt to
reduce the inflow of radon through manipulation of leakage areas
(cracks) and pressure differences. These strategies though, do
not make the best use of the physics behind the relationship
shown in equation 2. This paper will discuss the physics
involved and resultant, more efficient mitigation techniques.
-------
THE STACK EFFECT
FLOW EQUATIONS
From Bernoulli!'s equation, we can get a theoretical expression
for mass flow rate of a fluid through an opening:
M' = p * A * J2 * AP/p (3)
where M' is the mass flow rate, p is the density, A is the
leakage area, AP is the pressure difference across the opening.
For complex structures, the dependency of the pressure difference
is more complicated (2). Therefore an empirical power law
function is used:
M' = K * v/p *v/5" *A * AP"
e
= K * A0 * APn (4)
The exponent of the flow equation, n, varies theoretically from
0.5 for fully developed turbulent flow to 1.0 for laminar flow
(4). A represents a free area equivalent to the real crack. K
is an airflow coefficient, defined as the flow rate at a pressure
difference of 1.
Leakage area
Finding and sealing cracks in unfinished basement slabs may
be straight forward and relatively easy. Locating cracks in
slabs of finished basements, slab-on-grade foundations or in
floors over crawl space foundations can be next to impossible due
to finish floor materials, wall partitions, plumbing fixtures and
appliances. As such, effectively eliminating or reducing leakage
area may be impossible. As well, new cracks may develop with
time.
-------
Pressure differences
The pressure differences responsible for flow of soil-air
into a building arise from 1) wind pressures on the building, 2)
forced ventilation of the building from kitchen and bath-type
exhaust fans, 3) chimneys and flues from combustion appliances
such as furnaces and water heaters, 4) fireplaces and woodstoves,
and 5) air leakage due to the buoyancy, or the stack effect.
As wind hits a building, pressures develop as kinetic energy
is reduced. Pressures on the windward side increase while
pressures on the downwind side decrease. The interior of the
building will reach a pressure somewhere in between, depending
upon the location of openings. Openings concentrated on the
windward side will tend to pressurize the building, while
openings on the downwind side will tend to depressurize the
building.
Mechanical exhausts and chimney stacks exhaust air from the
building lowering the pressure within the building.
STACK EFFECT
Another physical phenomena that influences the pressure
differences in and around a building is the effect of buoyancy,
or stack effect. This is due to the density differences between
inside and outside air or between two zones within a building,
such as downstairs and upstairs.
Warm air is less dense (lighter) than cold air. This warmer
air will tend to "float" on the colder air. The result is a
pressure difference between the inside air and the outside air,
and a flow through any available opening. If openings exist at
different elevations within the building, warmer (less dense) air
will force its way out the upper openings to be replaced by
cooler (more dense) air through the lower openings.
The density of air is mainly a function of temperature and
moisture content. As temperature increases, density decreases.
Similarly, as the moisture content increases, density decreases.
Many researchers and lay people relate the density differences to
just temperature. For a typical winter condition with outside
air at 0°F, 50% relative humidity (RH), and inside air at 70°F
and 50% RH, the change in temperature accounts for about 90% of
the density difference while the change in moisture content
accounts for about 10%. At high temperatures, moisture content
can result in a significant density change. For example, inside
air at 75°F/60% RH is less dense than outside air at 80°F/20% RH.
At higher temperatures, the effects of moisture content are more
pronounced.
-------
The pressure difference between across an opening in a
building is proportional to the distance of the opening from a
neutral pressure level (NPL) and the density difference between
inside air and outside air. The pressure difference across an
opening due to the stack effect is:
= ( P0-P,-) * g * (H-h) / 2
(5)
where p is density, o indicates outside, i indicates inside, g is
gravitational force and h is the distance from the NPL, H.
THE NPL
The NPL is the elevation at which the pressure inside the
building is the same as outside the building. Above this NPL,
inside air is at a higher pressure than outside; air will tend to
leak (or be forced) out of the building. Below the NPL, inside
air is at a lower pressure than outside; outside air will be
drawn into the house. The further the opening from the NPL, the
larger the force pushing or pulling the air (see Figure 1). The
larger the temperature difference, the larger the force pushing
or pulling the air.
Neutral Pressure Level
Figure 1. The NPL in a typical building with uniformly
distributed openings. The length of the arrow indicates the
pressure difference between inside and outside. The direction of
the arrows indicate airflow.
-------
NPL Location
The location of the neutral pressure level is a function of
the location of the openings and their distribution throughout
the building. The NPL is derived from a solution to the
continuity equation of volume (4):
Ł/„
H-h
where H is the elevation of the NPL, h is the elevation of
opening j, A is the area of opening j.
For a simple enclosure with two openings having areas A1
(lower) at elevation h,, and A2 (upper) at elevation h2, the NPL
H, is given by:
A1+A2
(6)
From equation 6, if A, = A2, the NPL will be at a level
midway between h, and h2. Where A, « A2 (openings are
concentrated in the upper part of the building), the NPL will be
close to the top as in Figure 2. Where A, » A2 (openings are
concentrated in the lower part of the building), the NPL will
also be located close to the bottom of the building, as in Figure
3.
Buildings, as we construct them, have leakage areas through
the side walls also. But as builders utilize better doors and
windows and infiltration barrier house wraps, the leakage through
walls will decrease significantly. At the same time, many
weatherization techniques reduce the leakage rates through doors
and windows. These buildings have effectively become better
chimneys, with a greater potential for drawing radon-laden soil
gas into the building. Openings in the floor and ceiling will
become more critical.
-------
Figure 2. The NPL in a building, with most low openings sealed,
is located near the top of the building resulting in a lower
pressure inside than outside throughout the building. The
remaining low openings are at an even greater pressure
difference, resulting in more flow through the remaining
openings.
Neutral Pressure Level
Figure 3. Sealing openings at the top of the building causes the
NPL to be lowered, resulting in a decrease in infiltration
through low openings.
-------
THE NPL AND RADON
Since locations below the NPL are under a lesser pressure
than outside, air will try to enter the house through these
openings. The farther from the NPL, the greater the pressure
difference, or force pushing the air into the house. Conversely,
the higher above the NPL, the greater the force pushing air out
of the house.
At the same time, the location of the NPL depends on the
distribution of the openings. With uniformly distributed
openings, the NPL will be close to the center of the building.
But with openings concentrated at the top of the house, the NPL
will also be located close to the top of the house. Openings
concentrated low in the house result in a NPL low in the house.
Cavallo et. al. (6) found that opening windows in a basement
(to increase ventilation) also reduces the rate of radon entry.
This is a result of increasing the ratio of openings below the
NPL to those above the NPL. They effectively lowered the NPL and
decreased the pressure differences across the infiltration
openings.
One primary radon mitigation or prevention technique is to
seal cracks and openings in the floor of a basement or ground
floor in an attempt to reduce radon entry. This technique has
had mixed results, and in some cases even resulted in increased
levels of radon in the living space. By sealing openings in the
floor (below the NPL), the remaining openings are concentrated
higher in the house. This pushes the NPL up, creating an even
greater pressure difference across any openings in the floor
which did not get sealed. A greater pressure difference then
potentially causes more (not less) radon-laden air through the
cracks into the house.
SEAL THE CEILING FIRST
Conversely, if cracks and openings in the ceiling or
locations above the NPL are sealed, the NPL moves downward. The
result is a lesser pressure difference (or driving force) across
the lower openings, resulting in less radon-laden airflow through
these openings.
Figure 4 shows typical infiltration/exfiltration locations
in a typical house. Exfiltration sites are typically above the
NPL and are listed in Table 1. Infiltration sites, typically
located below the NPL, are listed in Table 2. Sealing
exfiltration sites will result in a lowering of the NPL and
resultant lowering of the pressure differences across
infiltration openings.
-------
Figure 4. Typical leakage sites in a residential structure.
Under Stack Effect pressures, sites located high in the house
will have air leaking out, while low sites will have air leaking
in.
-------
TABLE 1. EXFILTRATION SITES AS SHOWN IN FIGURE 1. SEALING OR
REDUCING FLOW THROUGH THESE OPENINGS WILL REDUCE THE PRESSURE
DIFFERENCE ACROSS INFILTRATION OPENINGS.
Number Exfiltration Site
1 Ceiling/wall/beam joints
2 Ceiling mounted light fixtures
3 Recessed light fixtures
4 Bathroom exhaust vents
5 Whole house fans
6 Attic access hatchways/stairways
7 Inside electric/plumbing penetrations
8 Flue/chimney penetrations
9 Wall/ceiling perimeter joints
10 Windows and doors
11 Kitchen exhaust vents, range hoods
12 Fireplace/stove dampers & doors
13 Wall/floor baseboard joints
14 Wall mounted heaters
15 Leaking supply ducts
16 Combustion furnaces/water heaters
TABLE 2. INFILTRATION SITES AS SHOWN IN FIGURE 1. SEALING OR
REDUCING FLOW THROUGH THESE OPENINGS MAY ONLY INCREASE FLOW
THROUGH OTHER INFILTRATION SITES.
Number Infiltration Site
10 Windows and doors
13 Wall/floor baseboard joint
17 Exterior wall penetrations
18 Interior crawl space hatches
19 Floor cracks and joints
20 HVAC registers/returns
21 Band joist/mud sill joint
22 Plumbing penetrations
23 Electric/tele/TV penetrations
24 Cracks/pores in concrete walls
25 Floor/wall joints
26 Sump openings
27 French drains
29 Cracks in slab floors
30 Columns/stairs penetrating slab
-------
Other Benefits of Lowering the NPL
Lowering the NPL not only directly reduces the forces
drawing radon into a building, but will also increase the
efficiency of sub-slab depressurization systems. The suction
under the slab does not have to be as large to prevent radon
entry. This would also allow further extension of an adequate
suction field under the slab (as measured relative to the
basement).
Lowering the NPL also reduces infiltration into the house in
general. Reducing this infiltration will have an effect on the
energy use to heat this infiltrating air.
Exfiltration will also decrease with a lowered NPL. In the
past, houses were almost exclusively heated with combustion
appliances, resulting in a NPL at or above the ceiling. The high
NPL causes infiltration through every opening in the house. As
such, cold dry air outside was drawn in through the building
shell. With the use of heating systems other than combustion
appliances such as electric strip heaters and heat pumps, the NPL
has come down to somewhere around the midlevel of the house.
Exfiltration of warm moist air has occurred. This exfiltration
has subsequently led to condensation and moisture problems in
walls and ceilings. Lowering the NPL by sealing openings above
the NPL would reduce this exfiltration and potentially reduce
these moisture problems.
Words of caution
Sealing any openings in a building may reduce the overall
air change rates, resulting in indoor air quality problems. This
situation has to be evaluated more thoroughly to prevent
problems.
Reducing air change rates will almost certainly result in
increased moisture content in the indoor air. This will affect
the density of the indoor air and as such result in a larger
stack effect pressure between inside and outside.
Powered ventilation systems such as kitchen exhaust fans, as
well as chimneys for furnaces and fireplaces, tend to raise the
NPL. In some instances, so much air is vented that the NPL is
above the ceiling, placing the entire house under a vacuum. In
these cases, adding an outside air supply directly to the
appliance is necessary.
Wind forces on the exterior of the building also affect the
inside/outside pressure relationships. Large opening upwind
-------
relative to downwind lower the NPL. Larger openings downwind
tend to raise the NPL.
CONCLUSIONS
Two conditions must exist for radon-laden soil-air entry
into a biulding: 1) an opening and 2) a pressure difference.
Sealing openings in the floor in an attempt do reduce radon entry
increases the pressure difference across any unsealed openings in
the floor. Conversely, sealing openings in the ceiling or upper
locations of a building will lower the neutral pressure level
within the building. This results in a lower infiltration rate
as well as a lower pressure difference across infiltration
openings. Less radon-laden air will be drawn in through openings
in floors or below-grade walls. Depressurization fields of SSD
and ASD systems will increase in strength relative to the
basement.
REFERENCES
(1) U.S. EPA & New York State Energy Office, Reducing Radon In
Structures/Training Manual.
(2) Brennan T, (1990) Evaluation of Radon Resistant New
Construction Techniques. Preprints of the 1990 International
Symposium on Radon and Radon Reduction Technology, Vol. 5,
p.l.
(3) Nuess, R.M. and Prill, R. J. (1991) Radon Control - Towards
a Systems Approach. Preprints of the 1991 International
Symposium on Radon and Radon Reduction Technology: Vol. 4,
VII-4.
(4) Albright, L.D. Environmental Control for Animals and Plants.
ASAE Textbook #4, August 1990, Pamela DeVore-Hansen, Ed. pp.
319-345.
(5) Blomsterberg, A. K. and D.J. Harrje, (1978), Approaches to
Evaluation of Air Infiltration Energy Losses in Buildings,
ASHRAE Transactions, Vol 85, Pt 1, PP. 797-815.
(6) Cavallo, A., K. Gadsby and T.A. Reddy, Natural Basement
Ventilation as a Radon Mitigation Technique. Preprints of
the 1991 International Symposium on Radon and Radon
Reduction Technology, Vol. 2, IV-6.
(7) House Tightening Manual For Homeowners and Weatherization
Contractors. Bonneville Power Administration, October 1985,
DOE/BP/13301-1.
-------
VIP-6
RELATIVE EFFECTIVENESS OF SUB-SLAB PRESSURIZATION AND
DEPRESSURIZATION SYSTEMS FOR INDOOR RADON MITIGATION; STUDIES
WITH AN EXPERIMENTALLY VERIFIED NUMERICAL MODEL
by: Ashok Gadgil, Yves Bonnefous, and Bill Fisk
Indoor Environment Program
Lawrence Berkeley Laboratory
University of California at Berkeley
Berkeley, CA 94720
ABSTRACT
The performance of sub-slab-ventilation (SSV) systems has
been studied with a numerical model, which was earlier
successfully compared with experiment. The parameters explored
in this study are the permeabilities of the soil and the sub-slab
gravel, the magnitude of pressurization (or depressurization)
applied by the SSV system, and the mode of SSV application (i.e.
pressurization (SSP) or depressurization (SSD)).
The mechanisms contributing to the successful performance of
SSP and SSD systems are identified. The numerical modeling
demonstrates that placement of a sub-slab gravel layer
substantially improves the SSV system performance. Except in the
case of highly permeable soils, SSD systems are predicted to
perform better than SSP systems. This prediction is consistent
with anecdotal experience. The numerical model is used to
elucidate the reasons for this difference in performance.
1. Background
The health risk from indoor exposure to radon progeny is the
largest of the health risks arising from indoor pollutants (1).
SSV systems are commonly used to reduce elevated levels of indoor
radon in houses with basements (2). The mode of operation of
sub-slab pressurization (SSP) and sub-slab depressurization (SSD)
systems can be qualitatively described as follows.
SSP systems are used to ventilate the soil in the sub-slab
region. This reduces the radon concentration in the soil-gas in
the sub-slab region and is intended to reduce the rate of radon
entry into the basement. The air entering the basement from the
sub-slab region in presence of SSP system operation is primarily
fresh air injected into the sub-slab region by the SSP system,
carrying along with it a small amount of radon picked up during
the sub-slab transit (Figure 1).
-------
SSD systems invert the pressure-gradient across the basement
slab, which normally drives entry of radon bearing soil gas into
the basement. Owing to this inversion of the pressure gradient,
basement air is sucked into the sub-slab region through cracks
and joints in the basement slab. This air, and also radon-
bearing soil gas, are sucked into the SSD system pipe and
expelled to the outdoors by the SSD system (Figure 2).
Experiments and field measurements of SSD and SSP system
performance provide additional details which can be summarized as
follows.
1. In a majority of cases SSD systems perform better than SSP
systems in reducing radon entry (2). SSP systems were found
to perform better in one field study in Spokane with
basement houses built on highly permeable soil (3).
2. A good connection of the SSV system to the sub-slab region
is important for good SSV system performance. Factors
ensuring such connection are: (a) the presence of a highly
permeable gravel layer beneath the basement slab, (b)
absence of a partitioning of the sub-slab gravel layer with
internal footers (this can be accomplished e.g. by having
either discontinuous internal footers, or embedding shot
lengths of pipes transverse to the internal footers), for
good pressure communication throughout the sub-slab gravel
layer (4-6).
3. ssy performance is improved by the presence of a pit at the
point of penetration of the SSV pipes through the basement
slab (7) .
4. Sealing of the cracks and joints in the basement slab
improves SSV system performance by reducing short-circuiting
of the extension of pressure imposed at the SSV system pit,
throughout the sub-slab region (7).
5. A large value for the ratio of the permeability of the sub-
slab gravel to the permeability of the soil increases SSV
system performance (7).
Most of the findings enumerated above have not been
rigorously proven by field studies. Becuase the uncontrolled
factors in field studies, these findings are only tentative. A
numerical modeling study of SSV systems was undertaken to obtain
a quantitative understanding of how these factors influence the
system performance.
-------
2. Model Description
Models of radon entry have been recently reviewed by Gadgil
(8). The Darcy law, relating pressure gradient to the bulk
velocity in a porous medium, has been commonly used in a number
of numerical and analytical studies of soil gas entry into
basements. However, this law is not a valid description of the
flow when the Reynolds number (based on the gravel diameter) of
the flow becomes large. The literature suggests an upper limit
of 1 to 70 for the Reynolds number beyond which the Darcy law no
longer applies. Since the Reynolds number of the flow in sub-
slab gravel greatly exceeds this upper limit during SSV
operation, a model based on non-Darcy flow must be used to study
the system performance. In a previous study, Gadsby et al. (9)
used an analytical model of SSV system performance based on the
exponential form of the non-Darcy flow. However, the conclusions
of their study are limited by the assumptions that there are no
cracks in the basement slab (so there is no flow between the
basement and the gravel layer), and that the boundary between the
sub-slab gravel and the soil is also impermeable.
For this study, a numerical model of SSV system operation
was written. The model, "Non-Darcy STAR" (Non-Darcy Steady-state
Transport of Air and Radon), is a three dimensional finite-
difference model on a rectilinear coordinate system, based on the
SIMPLE (Semi Implicit Method for Pressure Linked Equations)
algorithm developed by Patankar (10). The model incorporates the
Darcy-Forchheimer (11) equation for flow through permeable media,
and the equation of continuity:
Vp = -—(1 +
V.V = 0 (2)
where p is the disturbance pressure (i.e. the change in soil-gas
pressure owing to the depressurization of the basement and the
pressure applied by the SSV system) , ji is the dynamic viscosity
of soil-gas, K is the permeability of the porous medium (i.e.
soil or gravel), c is the Forchheimer term, and V is the bulk
velocity of soil-gas. The soil block is assumed isothermal for
the present study (although it would be straight forward to
incorporate the effects of buoyancy changes on soil gas flow in
the model). The computational domain is bounded from below, at a
depth of 10 m below the basement slab, by a no-flow boundary, and
from three sides by no-flow vertical surfaces at 10 meters
horizontal distance from the basement walls. The fourth vertical
surface represents the plane of symmetry of the problem and
vertically bisects the basement and an internal footer (Fig. 3).
-------
This fourth vertical surface is also a no-flow boundary owing to
symmetry. The computational domain is .bounded from above by the
basement slab, the wall and footers (with a joint defined between
the slab/wall and the footers), and the soil-surface outside the
basement. The slab is impermeable except at the joint defined
along its entire periphery. The soil-surface outside the
basement is defined to be at zero disturbance pressure.
Once the velocity field for the soil-gas is calculated, the
radon concentration field is calculated with the following
equation:
V.(^ VcRn)-V.VcRn+S(S-XRn) =0 (3)
where Ł is the bulk diffusivity of radon in the porous medium,
cRn is tne concentration of radon in the soil-gas, S is the
release rate of radon into the soil gas per unit volume of the
porous medium, /.Rn is the inverse of the time-constant of radon
decay, and e is the porosity of the medium. The model is
described in detail by Gadgil, et al. (12). The solution
approach used for solving eg. (3) is similar to that used by
Loureiro (13). The rate of radon entry into the basement is
calculated in the model by integrating over the crack area, the
product of soil-gas entry rate and the local radon concentration
in the soil-gas. Finally, the radon concentration in the
building is evaluated using the following two simplifying
assumptions: (1) The air in the building is perfectly and
instantaneously mixed and (2) air-exchange rate in the building
is obtained by summing in quadrature (14) a fixed typical air
exchange rate and the air exchange rate induced by operation of
the SSV system. To simplify the model and make results easier to
understand, we have neglected diffusive transport of radon
through the concrete and the radon in outdoor air.
3. Description of Parametric Study
The parametric study was defined on a "typical" single
family residence with a basement; with an axis of symmetry; the
latter permitted us the modeling of only one of the two symmetric
parts of the building and the soil block surrounding it. The
floor plan and a cross-section of the modeled basement are shown
in Figure 3. The layout and dimensions are based on common
single family houses, but are not based on a statistical survey.
The cracks in the basement slab are represented with a
single equivalent crack of width 1 mm, located at the joint of
the slab with the wall and footer, and has an L shape in cross-
section. In the model, the exterior surface of the basement
-------
walls is assumed to be displaced outwards to line up with the
footer edge (see Figure 3, basement section). This allows some
reduction in the number of computational control-volumes required
in the model, and a corresponding reduction in computational
time. The basement walls are assumed perfectly impermeable, so
is the top surface of the soil outside the basement that is
covered by the slab of the garage. The permeabilities and
Forshheimer terms for three kinds of gravels are those measured
in the laboratory (12). These three gravel types represent the
range of types used in housing construction in the state of
Washington (12). The backfill region is assumed to have the same
permeability as the soil, and is assumed to be in firm contact
with the exterior surface of the basement wall. (The influence
of gaps between the footer and soil, the permeability of the
backfill, and the cracks in the basement walls on the SSV
performance is under study). A pit (25 cm radius) is present
where the SSV pipe penetrages the basement floor (see Figure 3)
(12) The pressures imposed with the SSV system at this pit are
either 60 Pa (the lower limit of SSV system after e.g. system
degradation with age), or 250 Pa (the upper limit for systems
installed in houses with a sub-slab gravel layer). The basement
is assumed to be under a fixed depressurization of -10 Pa. The
model does not account for an upper limit on the flow-rate
capacity of the SSV fan.
In this study, we assume a radon concentration in the deep
soil-gas of 58,830 Bg/m3, equal to three times the smallest deep
soil-gas radon concentration measured by Turk et al. (16) in the
region of Spokane, Washington. The sub-slab gravel layer is
assumed to release no radon into the soil-gas. The building
volume is taken to be 122.5 m3, and has a fixed value of air
exchange rate (in the absence of SSV system operation) of 0.4
ACH.
The choice of the degree of resolution of the computational
domain into control volumes is determined by the trade-off
between computational effort and residual errors owing to
inadequate resolution of regions with large gradients of velocity
and pressure. The computational domain in most of the parametric
runs had 37,260 control volumes. We estimate that the error in
radon entry rate resulting from inadequate grid resolution is
less than 20%.
4. Results
The performance of SSV systems is evaluated in this study
with two criteria: (1) a good extension of pressure field in the
sub-slab region, and (2) the degree of reduction of indoor radon
concentration achieved in the building described in the previous
sections.
-------
a
4.1. Indoor Radon Concentrations in Absence of SSV
In absence of SSV, the model predictions of indoor radon
concentrations for different combinations of soil and sub-slab
gravel,^are shown in Figure 4. We note that our model
predictions show a substantial enhancement in indoor radon
concentration resulting from placement of a highly permeable sub-
slab gravel layer. This effect has been noted also by Revzan et
al. (17), using a different numerical algorithm and a distinct
computer model to solve equations (1-3). Placement of an
impermeable plastic membrane below the sub-slab gravel
substantially reduces this detrimental increase in radon entry
(9), and is being studied in more detail.
4.2. SSD System Performance
Figures 5 and 6 show model predictions for SSD system
operation with a depressurization at the pit of -60 Pa and -250
Pa respectively. Each figure displays several curves, each for
different permeability of sub-slab gravel. The horizontal axes
display soil permeability. The maximum sub-slab pressure at 30
cm from the wall, a measure of sub-slab pressure extension, is
plotted along the horizontal axes. This measure is chosen so
that pressure extension can be experimentally assessed without
being influenced by small-scale variations in the width of the
wall-slab-footer crack. The indoor radon concentration after SSD
operation, expressed as percentage of the indoor radon
concentration before SSD system operation, is shown on the
figures in parentheses near each simulation result.
The detailed results from these simulations of soil-gas
velocity and soil-gas radon concentration fields provide
additional insights into SSD operation. SSD system operation may
be satisfactory owing to either of two mechanisms.
1. If the extension of the depressurization in sub-slab gravel
layer is such that the sub-slab gravel layer is at a lower
pressure than the basement at all points, then the direction
of pressure-gradient across the cracks in the basement slab
is reversed, and the system functions in the desired fashion
(Fig. 2).
2. If_the flow drawn by the SSD system is large, the sub-slab
soil adjacent to the gravel layer is "ventilated" by the
fresh outside air entering the outdoors soil surface and
traveling to the SSD pit. In this case, even if the
pressure field extension in the sub-slab region in
inadequate to reverse the pressure-gradient across all
cracks in the basement slab, the large reduction in radon-
-------
concentration in the soil gas entering the basement can
prpvide an effective system performance.
In Figures 5 and 6, the simulation results show elimination
of radon entry through the basement cracks for the results
displayed below the horizontal line representing -10 Pa (the
pressure in the basement), owing to mechanism (1) above. The
large reduction in indoor radon concentrations, above the line at
-10 Pa, and particularly at large soil permeabilities (near the
top right hand corner of each figure), are a consequence of the
mechanism (2) described above.
Results presented in Figures 5 and 6 support the following
conclusions regarding SSD system operation:
1. Placing of a sub-slab gravel layer is necessary for best
performance of an SSD system (note the curve in both figures
for simulation results without sub-slab gravel). On the
other hand, an increase in the applied depressurization from
-60 Pa to -250 Pa allows satisfactory reduction in indoor
radon concentrations for all cases studied in the
simulations, even in the absence of the gravel layer, by
ventilation of the sub-slab soil.
2. An increase in sub-slab gravel permeability from 2 x 10~8 m2
to 3 x 10~7 m2 at least triples the minimum subslab
depressurization predicted at a distance of 30 cm from all
the walls. This improvement in the sub-slab pressure field
extension provides a large margin of safety for assured good
performance of the SSD system. In addition, the increase in
permeability of the sub-slab gravel permits a reduction in
the applied depressurization. This could be significant in
some cases when system components degrade over the years
(e.g. caulking around the SSV pipe penetration in the slab
may develop cracks), or can be translated into savings in
first cost by installing a lower power fan in the system.
3. For ratios of soil to gravel permeabilities larger than
1000, it is possible to interpret the results as being
equivalent to the soil being completely impermeable—
practically all the air drawn by the SSD system originates
from inside the basement and enters the SSD pipe by crossing
the sub-slab gravel layer after passing through the cracks
in the slab. For these situations, good SSD system
performance is achievable with quite low values of applied
depressurization (theoretically, even for an applied
depressurization of — 10.1 Pa for a basement depressurized
to -10 Pa). This model prediction encourages further
exploration of applicability of passive SSD systems (based
on depressurization of the SSD pipe by buoyancy forces).
-------
4.3 SSP System Performance
Figures 7 and 8 show predicted performance of SSP system for
a range of soil and gravel permeabilities. The axes and
presentation symbols have the same interpretations as described
for Figures 5 and 6. Figure 7 is for an applied pressurization
of +60 Pa, and Fig. 8 for that of +250 Pa.
In absence of a gravel layer, an SSP system can not provide
a complete elimination of convective radon entry from soil. For^
lower pressurization (+60 Pa) and low permeability soils (K = 10
11 m2), the operation of an SSP system is predicted to lead to an
increase in the indoor radon concentration; the dilution of the
radon concentration in the soil gas entering the basement does
not compensate for the net increase in entry under these
conditions. The placement of a sub-slab gravel layer is crucial
for successful operation of an SSP system.
The simulations were undertaken with the (commonly
justifiable) assumption that the radon emanation from the sub-
slab gravel layer is negligible. The detailed simulation results
of the pressure, velocity and radon concentration lead to the
following understanding of the several mechanisms influencing
effective SSP system performance.
1. A good extension of pressure field ensures that no region of
the sub-slab gravel layer is under absolute depressurization
(p < 0 Pa). So long as this condition is satisfied, radon
can not enter the sub-slab gravel layer by convection. This
condition is not satisfied for the gravel of low
permeability (k = 2 x 10~8 m2 in our study) at an applied
pressure of +60 Pa (see Fig. 7).
2) At the same time, the system can have a good performance
even in the presence of a poor extension of the pressure
field, if the flow of air injected in the soil by the SSP
system is large (as is the case for very permeable soils).
This large flow essentially washes away the radon in the
near-basement soil and the concentration of radon in the
soil gas entering the gravel slab (owing the poor pressure
field extension) is reduced.
3) The soil must be sufficiently permeable that the flow
between the pressurized sub-slab gravel layer and the sub-
gravel soil must not be too small. If this flow is
adequately large, the air passing from the gravel into the
soil pushes away the radon that may otherwise enter the sub-
slab gravel layer from the sub-gravel soil by diffusion
across the large concentration gradient. In case of highly
impermeable soils (e.g. k < lO"11 m2), the flow of air
across the interface of soil and gravel is close to zero
-------
(the soil can be practically considered impermeable),
allowing radon to diffuse into the air passing along the
gravel layer from the adjacent soil. This radon then enters
the basement from the sub-slab gravel along with the air
flow that occurs through the cracks during all SSP system
operation.
4) Enhancement of the pressure imposed by the SSP system at the
pit can reduce the diffusion of radon from the sub-gravel
soil into the sub-slab gravel only at the cost of an
increase in the air exchange rate in the building (by
dilution). This can have a significant energy cost. In
conclusion, for a proper functioning of the SSP system,
highly permeable soil and gravel are required.
5. Discussion
Our parametric simulation study of a typical house shows
that generally an SSD system is predicted to perform better than
an SSP system, except in the case of extremely permeable soils (k
> 2 x 10~8 m2). For the combinations of highly permeable gravels
(k > 10~7 m2) and permeable soils (10~10 m^ < k < 10~9 m2) the
two systems have equivalent performance. This is summarized in
Table 1.
A permeability in excess of 2 x 10~8 m2 is a unusually high
value for soil permeability. The simulations show that only in
such cases the SSP system performance is better than the SSD
system performance. However, ongoing research (18) shows that the
numerical models under-predict a soil-gas flow entering basements
(in absence of an SSV system) by approximately a factor of eight.
The phenomenon of the "washing" of the radon from near-building
soil is based on the large rate of air-flow from the SSP system
into the soil. Therefore, we suspect that SSP system performance
may exceed SSD system performance at lower soil permeabilities
than predicted by Non-Darcy STAR.
Except for cases of the SSP system performance in the
absence of a gravel layer and for soil permeabilities less than
or equal to 10~10 m2, an increase in the value of the imposed
pressure at the SSV pit from 60 to 250 Pa allows a satisfactory
reduction in the indoor radon concentration. However, this
increase in the imposed pressure can translate into a large
increase in the air exchange rate of the house, with the
associated energy cost for heating (or cooling) the house. As an
example, for an SSP system functioning at an imposed pressure of
250 Pa in presence of a highly permeable gravel (k = 3 x 10~7 m2)
and a soil with low permeability (k < 10"*1 m2), the rate of air
exchange increases from 0.4 ACH assumed in the absence of SSP to
1.95 ACH in the presence of SSP operation. The energy costs
-------
associated with this increase is generally significant. For the
system operating in highly permeable soils, the flow associated
with a depressurization of -250 Pa exceeds the maximum practical
flow rates of SSV fans (about 0.2 m3/s).
A plastic membrane below the gravel layer can be beneficial
to the performance of the two systems, and would permit a better
extension of the pressure field (for both the SSP and SSD) and
also eliminate the diffusive entry of radon from sub-gravel soil
into the sub-slab gravel during SSP system operation.
Besides having a generally poorer performance than the SSD
system, the SSP systems have two major inconveniences: 1) the
flow of air entering the house induced by the system operation
can entrain other soil-resident pollutants such as termicides,
and 2) the air injected by the system is at outdoor temperature,
which, in severe winters, could lead to freezing of sub-slab
pipes. An SSP system based on injecting house (i.e. indoor) air
is not a viable option owing to the continuous increase in indoor
radon concentration as the radon diffuses into the air during its
passage through the sub-slab gravel.
6. Conclusion
A numerical study of SSV systems has been carried out using
a model "Non-Darcy STAR". This study has led to the
identification of mechanisms contributing to the success of these
systems: (1) for SSD systems, inversion of the direction of
pressure gradient across the basement slab and the reduction in
the radon concentration in the soil and (2) for SSP system
pressurization of the whole of the sub-slab gravel layer
eliminating convective entry from sub-gravel soil into the sub-
slab gravel, reduction in the radon concentration in the soil,
and air flow across the interface of soil and the sub-slab gravel
suppressing the diffusion of radon into the gravel layer.
A layer of highly permeable gravel is necessary for best
performance of the SSV systems. Selection of a highly permeable
gravel (k = 3 x 10~7 m2) permits considerable improvement in the
SSV system performance and also provides a margin of safety in
case of system deterioration. In certain cases (e.g. installation
of SSD system in soils with permeabilities of less than or equal
to 10"11 m2), it might allow successful operation of passive
systems.
The SSD systems are generally better at reducing indoor radon
concentrations than SSP systems with the exception of
installations in soils of extremely high permeabilities (k > 10~8
m2) .
-------
7. Table Caption
Table 1: This table shows which system (SSD or SSP) has the
better performance in radon reduction for various
specific combinations of gravel and soil
permeabilities, as predicted by the model Non-Darcy
STAR.
-------
8.
List of Figures
Figure 1: Conceptual representation of the functioning of a sub-
slab pressurization (SSP) system.
Figure 2: Conceptual representation of the functioning of a sub-
slab depressurization (SSD) system.
Figure 3: The plan and section of the "typical" house modeled for
the parametric simulations with Non-Darcy STAR. Owing
to the assumed plane of symmetry of the house, only
half of the house is shown in the plan and the section.
The modeled space includes the soil block surrounding
the house extending 10 meters to all sides of the
house, and 12.5 meters below the outside soil surface.
Figure 4: Indoor concentration of radon in an idealized perfectly
mixed house of volume 122.5 m3, and with an air
exchange rate of 0.4 ACH. The assumed concentration of
radon in the deep soil is 58830 Bq m~3.
Figure 5: Predicted SSD system performance for an applied
pressure at the pit of -60 Pa. The pressure in the
basement is assumed to be -10 Pa. Percentage reduction
in the indoor radon concentration resulting from SSD
operation is shown in parenthesis for each of the
simulation points on the figure.
Figure 6: Predicted SSD system performance for an applied
pressure at the pit of -250 Pa. The pressure in the
basement is assumed to be -10 Pa. Percentage reduction
in the indoor radon concentration resulting from SSD
operation is shown in parenthesis for each of the
simulation points on the figure.
Figure 7 Predicted SSP system performance for an applied
pressure at the pit of +60 Pa. The pressure in the
basement is assumed to be -10 Pa. Percentage reduction
in the indoor radon concentration resulting from SSD
operation is shown in parenthesis for each of the
simulation points on the figure.
Figure 8: Predicted SSP system performance for an applied
pressure at the pit of +250 Pa. The pressure in the
basement is assumed to be -10 Pa. Percentage reduction
in the indoor radon concentration resulting from SSD
operation is shown in parenthesis for each of the
simulation points on the figure.
-------
9. References
(1) Nero, A.V. "Controlling Indoor Air Pollution", Scientific
American, Volume 258, No. 5, May 1988.
(2) Henschel, B. "Radon Reduction Techniques for Existing
Detached Houses — Technical Guidance, third edition", U.S.
EPA Office of Research and Development, Washington, D.C.
20460, First Draft, January 1992, Documents in preparation.
(3} Turk, B.H.; Prill, R.J.; Fisk, W.J.; Grimsrud, D.T; Moed,
( B A. and Sextro, R.G. "Radon and Remedial Action in Spokane
River Valley Homes. Volume 1: Experimental Desl9nTandA^ta
Analysis". Lawrence Berkeley Laboratory Report, LBL-23430,
December 1987.
(4) EPA "Radon Reduction Techniques for Detached Houses,
Technical Guidance (second edition)", U.S. EPA, Report
EPA/625/5-87/019, 1987.
(5) WSBCC [Washington State Building Code Council] "Washington
State Ventilation and Indoor Air Quality Code Chapter 5
Radon Resistive Construction Standards". Published by the
State of Washington, 1990.
(6) Nuess, M. "Northwest Residential Radon Standard Volume 1:
Project Report". Bonneville Power Administration Report,
Portland, OR, 1989.
(7) Bonnefous, Y.C.; Gadgil, A.J.; Fisk, W.J.; Prill, R.J.;
Nematollahi, A. "Field Study & Numerical Simulation of
Subslab Ventilation Systems." to appear in Environmental
Science and Technology, 1992.
(8) Gadgil, A.J. "Models of Radon Entry: A Review". Accepted
for publication in Rad. Prot. Dosimetry, 1992. Also
Lawrence Berkeley Laboratory Report LBL-31252, Berkeley, CA,
August 1991.
(9) Gadsby, K.J.; Reddy, T.A.; Anderson, D.F.; Gafgen, R; Craig,
A B "The Effect of Subslab Aggregate Size on Pressure Field
Extension. In: Proceedings of the 1991 International
Symposium on Radon and Radon Reduction Technology, vol. 4,
April 2-5, Philadelphia, PA. Published by the U.S. EPA.
(10) Forchheimer, P.H. Z. Ver. Dtsch. Ing. 45, pp. 1782-1788,
1901.
(11) Patankar, S. Numerical Heat Transfer and Fluid Flow. McGraw-
Hill Book Company, New York, 1980.
-------
(12) Gadgil, A.J.; Bonnefous, Y.C.; Fisk, W.J.; Prill, R.J.;
Nematollahi, A. "Influence of Subslab Aggregate
Permeability on SSV Performance". Lawrence Berkeley
Laboratory Report, LBL-31160, Berkeley, CA 1991.
(13) Loureiro, C.O.; Abriola, L.M.; Martin, J.E.; Sextro, R.G.
"Three Dimensional Simulation of Radon Transport into Houses
Under Constant Negative Pressure", Environmental Science and
Technology, Vol. 24, pp. 1338-1348, 1990.
(14) Sherman, M.H. "Superposition in Infiltration Modeling",
Lawrence Berkeley Laboratory Report, LBL-29116, Berkeley,
CA. Submitted to Indoor Air, 1992.
(15) Palmiter, L.; and Brown, I. "Northwest Residential
Infiltration Survey: Analysis and Results". Ecotope, 2812
East Madison, Seattle, WA 1989.
(16) Turk, B.H.; Harrison, J.; and Sextro, R.G. "Characterizing
the Occurrence, Source, and Variability of Radon in Pacific
Northwest Home". J. Air Waste Manage. Assoc. 40, pp. 498-
506, 1990.
(17) Revzan, K.L.; Fisk, W.J.; and Gadgil, A.J. "Modeling Radon
Entry into Houses with Basements: Model Description and
Verification". Indoor Air, Vol. 1, No. 2, pp. 173-189,
1991.
(18) Garbesi, K.; Sextro, R.G.; Fisk, W.J.; Modera M.P.; and
Revzan, K.L. "Soil-Gas Entry into an Experimental Basement:
Model-Measurement Comparisons and Seasonal Effects".
Lawrence Berkeley Laboratory Report, LBL-31873, Berkeley,
CA. Submitted to Environmental Science and Technology,
1992.
-------
Perm.
m2
2x10
1 x 10
"9
1x10
-10
<1 x10
No gravel
SSP
SSD
SSD
SSD
2x 10
"8
SSP
SSD
SSD
SSD
1x10-7
SSP
SSD
3x10 '7
SSP
SSD
-------
gravel
Figure 1.
-------
gravel
Figure 2.
-------
Plane of
symmetry
Figure 3.
Modified Wall location for
easier computer description
\
basement
/ x x y x /
y\pOOOOQt^
pit f
i^:
i*
X X
rr
i\K
J$]
Cross section A-A
-------
10000
cr
EL
c
o
+3
CO
v.
*-•
C
CO
o
c
o
o
c
o
T3
CO
O
o
•o
c
1000 :
100
10
EPA guideline
, Mr\ rtrov/ol
INU yiclVUI
— Gravel No. 1 , k = 2
Gravel No. 2, k = 1
- - Gravel No. 3, k = 3
x 10'8 m2
x 10"7 m2
x 10'7 m2
10
,-11
10
,-10
10
,-9
10-
Soil permeability [m2]
Figure 4.
-------
(0
= 0.0
CO
CO
E
o
E
o
o
CO
+"»
CO
0)
3
CO
(0
0)
Q.
c
3
E
"5
03
•10.0
•20.0
•30.0
-40.0
-50.0 -J-
(42)
(33)
Basement pressure
-•- No gravel
-+-Gravel No 1, k = 2 x 10"8 m2
-*-Gravel No 2, k = 1 x 10'7 m2
"•"Gravel No 3, k = 3 x 10"7 m2
(0)
(0)
Impermeable
membrane
Soil permeability
10
•11
Figure 5.
-------
03
Q.
I
CO
0.0
-20.0
-40.0
-60.0
-80.0
2
O
O
CO
to -100.0
o
^ -120.0 1
to
8 -140.0 :
g -160.0 j
•| -180.0
CO
-200.0 +-
Basement pressure
-No gravel
'Gravel No 1,k = 2x 1(T8m2
'Gravel No 2, k= 1 x lO^m2
'Gravel No 3, k = 3 x 10"7 m2
(2
Impermeable
membrane
10
n
(0)
I i i i ni| 1 i i i mi
10'
Soil permeability [m2]
Figure 6.
-------
CO
= 50.0 t~
03
I 40.0
E
o
* 30.0
o
o
CO
13
0)
3
(0
(0
Q)
Q.
E
3
E
!E -10.0
20.0
10.0
0.0
(54)
No gravel
Gravel No 1,k = 2x 10'8 m2
-Gravel No2,k= 1 x 10"7 m2
Gravel No 3, k = 3x 10-7m2
Impermeable
membrane
1 O'11 1 O'9
Soil permeability [m2]
10
-7
Figure 7.
-------
CO
0.
i
(0
o
o
o
CO
+•»
CO
CO
CD
o.
E
3
E
160.0
140.0
120.0
100.0
80.0
60.0
40.0
20.0
0.0
•No gravel
-Gravel No 1, k = 2x 10"8 m2
-Gravel No 2, k=1 x 10'7m2
•Gravel No 3, k = 3x 10'7m2
(42)
(46) (14)
Impermeable
membrane
10'11 1 O'9
Soil permeability [m2]
(4) (0)
10
r7
Figure 8.
-------
Session VII Posters
Radon Reduction Methods
-------
VIIP-1
RADON MITIGATION SYSTEMS — A LIABILITY IN COLD CLIMATE HOMES1?
by: Kenneth D. Wiggers, Ph.D.
American Radon Services, Ltd.
Ames, IA 50010
ABSTRACT
Homeowner observations and our studies have shown that there
is not enough air available in many relatively air-tight and
"average" homes to allow safe operation of natural draft combustion
appliances (furnaces, water heaters, and woodburners) and other
air-consuming appliances such as clothes dryers, exhaust fans and
radon mitigation system fans. Radon mitigation system fans remove
an unpredictable amount of conditioned air from homes. A certain
amount of available household air becomes dedicated to the radon
system; the air consumed by the radon system is no longer available
for other uses. It is recommended that all active soil
depressurization (ASD) and submembrane depressurization (SMD) radon
mitigation systems installed in cold climate housing with natural
draft appliances be recalled. The recall would consist of using a
blower door to determine the amount of air available in the homes
at different house pressure differentials. It is recommended that
a blower door be used to measure the air leakage characteristics of
a home before and after the installation of all radon mitigation
systems in cold climates.
'Presented at The 1992 International Symposium on Radon and Radon Reduction Technology, September
22-25, 1992, Minneapolis, MN
-------
INTRODUCTION
This study was prompted in part due to homeowner observations
(or perception) that radon mitigation systems caused natural draft
furnaces and water heaters to backdraft. The homeowners have
become aware of the backdrafting by installing tissue paper on the
draft diverters of their furnaces and by the observation that their
thermally-sensitive backdraft indicators had turned black. After
each radon mitigation system is commissioned, backdraft indicators
are installed on natural draft furnaces and water heaters and
passive carbon monoxide detectors are installed in a highly visible
place (usually on the refrigerator door).
The potential for radon mitigation system-induced backdrafting
is discussed with the homeowner during the on-site design phase
visit and in the subsequent "System Design for Radon Reduction"
report. The backdraft indicators give the homeowner a method of
detecting backdrafting. One young couple reported that their
backdraft indicator on the water heater indicated backdrafting
(flue gas spillage) and questioned whether their recent lethargy
and short-term memory loss could be associated with the
backdrafting of the water heater (and hence the radon mitigation
system).
The accompanying question was "Are the health effects
permanent?" In this example, the radon mitigation system consumed
(dedicated) 37 in2 of the 149 in2 premitigation available equivalent
leakage area (ELA) and 44 cfm of the 129 cfm premitigation air
available at one pascal pressure differential. The cfm of air
available at one pascal pressure dropped from a premitigation value
of 129 cfm to 85 cfm postmitigation. A furnace requires
approximately 70 cfm air for proper combustion and drafting. The
"numbers" show there is not much tolerance (forgiveness) and
therefore installers of air-consuming appliances must address the
issue "Is there adequate air for this additional air-consuming
appliance?"
MATERIALS & METHODS
Approximately 25 homes that had radon mitigation systems installed
one to two years ago were blower door tested to determine the
amount of equivalent leakage area and cfm of air available at one
pascal pressure differential. The blower door testing (BDT) was
conducted with the radon mitigation system fan shut off and
terminus of the PVC pipe taped shut (assumed pre-radon mitigation
conditions). The BDTs were also conducted with the radon
mitigation system in operation. All homes were retested for radon.
BDT data have been collected on all radon mitigation systems
installed in 1992. The equivalent leakage area (ELA) and the "cfm
available at 1 pascal" appear to be the most sensitive indices to
use for determining the effect of a radon mitigation system on
household air availability. The ELA sums all the leakage area in
-------
the house into one number. The "cfm available at 1 pascal" is
calculated by the blower door computer program and represents the
cubic feet per minute of air available for household use at one
pascal pressure differential between the inside and outside of the
house. The "n" value is the slope of the line drawn through the
air leakage data points.
The "leakage ratio" is the ELA divided by the total 100s of
ft2 of surface area (four sides plus top and bottom). The ACH50 is
the air changes at 50 pascal pressure differential. The cfm (air
flow) at any pressure differential (measured in pascals) can be
calculated by the following equation: air flow = (cfm at 1 pascal
pressure differential)(pascal").
RESULTS & DISCUSSION
The variability of air leakage characteristics of three
typical Iowa homes is illustrated in Table 1. The values shown are
typical of the variability encountered in Iowa housing stock.
Table 1. Air Leakage Characteristics of Three Iowa Homes (Pre-
mitigation)
City Address
Sioux City
Clive
Rock Rapids
cfm air
available at
one pascal
401
228
47
"n" value
0.554
0.652
0.561
ELA (in2)
422
243
62
The uncertainty of the amount of available air in housing
compounded with the uncertainty of the amount of conditioned air
removed by an ASD or SMD system creates an uncertainty about the
adequacy of combustion air for proper drafting of natural draft
appliances and hence a concern about the safety of the occupants.
Table 2 shows the "pre" and "post" radon mitigation cfm air
available and ELA.
-------
Table 2. Pre- and Post Radon Mitigation Blower Door Data
City
Sioux City
Clive
Rock
Rapids
Premiti-
gation
cfm air
available
401
228
47
Postmiti-
gation cfm
air
available
(pre - post)
294 (107)
172 (56)
35 (12)
Premiti-
gation
ELA (in2)
422
243
62
Postmiti-
gation ELA
(in2)
(pre - post)
294 (128)
212 (31)
53 (9)
The numbers enclosed in parentheses in the third and fifth
column show, respectively, the cfm air available at one pascal
pressure differential and the ELA dedicated (consumeed by) to the
radon system. The reduction amount (value in parentheses) in
available air and ELA is dedicated to the radon reduction system
and is no longer available for use by natural draft water heaters
and furnaces. Kitchen and bathroom exhaust fans, clothes dryers,
and other powered exhaust systems further reduce the amount of air
available for natural draft water heaters and furnaces. The amount
of air removed from the house by ASD and SMD radon mitigation
systems is unpredictable and thus each house must be blower door
tested before and after the installation of ASD and SMD systems to
determine (best guess, there are no standards) if there is adequate
air in the home.
HOUSE ILLUSTRATIONS
Representative Case Study — Marion. Iowa
A typical example of the problem follows. The owners of this
house were the first to call us to tell us that "their radon
mitigation system made their furnace backdraft." An alpha track
detector showed a radon concentration of 5.2 pCi/L in an ordinary
Marion, Iowa home. An ASD system was installed into a subslab
drainage tile using 4" ID PVC pipe and an in-line fan capable of
moving 270 cfm at zero inches static pressure.
The homeowner disconnected the fan for the family's ASD radon
mitigation system shortly after installation because it made the
furnace "run more" and because it made the furnace backdraft.
Blower door testing under various house conditions was conducted to
determine the effect of the ASD system (4" vs 2" PVC) and the
effect of the clothes dryer on air leakage parameters.
The house in the "radon fan off" condition theoretically
represents the house before the radon system was installed; the
electricity to the radon fan is disconnected and duct tape is
placed over the discharge end of the radon pipe. Table 3 shows the
-------
air leakage characteristics associated with two air-consuming
appliances, a clothes dryer and a radon system with two sizes of
PVC pipe.
Table 3. Air leakage characteristics — house in Marion.
House
condition
Radon fan
off
Radon fan
on (4H
PVC)
Radon fan
on (4"
PVC) +
dryer on
Radon fan
off +
dryer on
Radon fan
on (2")
Radon fan
on (2") +
dryer on
ELA,
(in2)
133
105
89
113
127
104
cfm
available
at 1
pascal
130
80
61
99
122
86
"n" value
0.543
0.653
0.700
0.590
0.551
0.614
Leakage
ratio
(inVlOO
ft2)
3.8
3.0
2.5
3.2
3.6
3.0
ACH50
5.0
4.8
4.3
4.6
4.8
4.4
The house had an apparent 133 in2 of ELA and 130 cfm of air
available to it at 1 pascal pressure differential in the assumed
pre-existing condition (before the radon mitigation system was
installed). Operation of the radon fan reduces the apparent ELA by
28 in2 (133-105) and the cfm available at 1 pascal pressure
differential by 50 cfm (130-80). The installation of the radon
mitigation system reduced the availability of household air and
thus has the potential to contribute to the backdrafting of the
furnace.
A 4/2" PVC bushing adaptor was installed on the effluent end
of the 4" PVC to reduce effluent flow (the fan was not changed).
The ELA and the cfm available at 1 pascal were respectively, 127
in2 and 122 cfm, after the 2" adaptor was installed. Reduction of
the 4" ID PVC to 2" ID PVC reduced the dedicated ELA from 28 in2 to
6 in2 and reduced the dedicated air available at 1 pascal pressure
differential from 50 to 8 cfm.
The dryer in this house and other houses tested consumed
approximately 20 in2 of the total ELA. It is noteworthy that the
-------
ACH50 fluctuated little in response to demands on the household
air. It seems that "cfm available at 1 pascal" pressure
differential is the most sensitive index to use to determine
adequacy of air in the home. The radon test data are shown in
Table 4.
Table 4. Pre- and Postmitigation radon data — four vs two inch
PVC.
Location
Basement
family room
Basement
utility room
Main floor
family room
Main floor
master bedroom
Main floor
children's
bedroom
Pre-
mitigation,
pCi/L
4.5
5.2
4.8
5.8
4.5
Post-
mitigation,
pCi/L (4" PVC)
0.4
0.4
0.4
0.5
0.5
Post-
mitigation,
pCi/L (2" PVC)
< 0.3
< 0.4
< 0.3
< 0.3
< 0.3
The 2" PVC was as effective in inhibiting radon entry as was
four inch PVC. The 2" PVC is preferable to the 4" PVC because
the likelihood of radon mitigation system-induced backdrafting of
the natural draft appliances and heating/cooling penalty is less
than when a 4" PVC pipe is used.
Representative Case Study — Newton, Iowa
The homeowner installed his own drain tile depressurization
(DTD) system. The homeowner was aware that the clothes, dryer
caused the furnace to backdraft before the radon system was
installed. The awareness of the backdrafting problem was
discovered when Mrs. Homeowner was nearly overcome by furnace
fumes while drying clothes. The problem has been partially
solved by opening a window next to the clothes dryer when
operating the clothes dryer.
The first 32 day screening test conducted in March 1989 gave
a radon concentration of 75.5 pCi/L. Short-term tests conducted
in March 1990 and December 1990 showed radon concentrations of
17.3 and 23.8 pCi/L, respectively. Postmitigation radon
concentrations were as follows (in pCi/L): crawl way, 0.6; family
room, < 0.5; living room, < 0.5; large bedroom, < 0.4; and < 0.5
in the small bedroom.
-------
Mr. Homeowner installed a drain tile depressurization (DTD)
radon mitigation system and discovered that he had exacerbated
his existing backdrafting problems. Mr. Homeowner found that it
was necessary to leave the window open all the time to avoid
backdrafting. (The furnace, water heater, and dryer are in an 8
by 15 foot utility room.) The homeowner has also noted that
weather conditions affect the backdrafting response to operation
of the radon mitigation system and the clothes dryer; under
certain weather conditions it is not necessary to have a window
open to prevent backdrafting. Table 5 presents the air leakage
characteristics of the Newton home.
Table 5. Air leakage characteristics — house in Newton.
House
condition
Radon fan
off
Radon fan
on (4"
PVC)
Radon fan
off (4"
PVC) +
dryer on
Radon fan
on (4"
PVC) +
dryer on
Radon fan
on (2"
PVC)
Leakage
area,
(in2)
82
55
47
34
63
cfm
available
at 1
pascal
67
27
22
14
35
"n" value
0.624
0.849
0.861
0.937
0.795
Leakage
ratio
(inVft2)
2.2
1.4
1.2
0.9
1.7
ACH50
3.3
3.2
2.8
2.3
3.4
There is not adequate air leakage area in the home to
provide air for the clothes dryer, radon system, furnace, water
heater, nor probably adequate ventilation to provide "fresh" air
for the occupants. The homeowner is looking for a permanent
solution to the backdrafting problems. The homeowner wants a
solution that anyone can live with; they are concerned that
someone housesitting while they are on vacation may inadvertently
kill themselves by drying the wash in the clothes dryer!
Representative Case Study — Luana Iowa
Two short-term tests conducted over a two-year period showed
a radon concentration of 9.4 and 9.6 pCi/L for the two respective
testing periods.
-------
A one-hole SSD system was installed in this home. Pre-
mitigation soil gas communication testing showed good soil gas
communication to all corners of the slab from the point chosen
for the SSD hole (under the flue chase). Two inch ID PVC pipe
was used in part because of the relatively small air leakage of
the home and because a 2" PVC pipe would fit up the flue chase
alongside the 3" ID Class B metal flue for the natural draft
water heater. A inline fan capable of moving 150 cfm of air at
zero inches static pressure was used. The homeowners have a
sealed combustion furnace. Table 6 shows the air leakage
characteristics of the home.
Table 6. Air leakage characteristics — house in Luana.
House
condition
Radon fan
off
Radon fan
on (2M
PVC)
Radon fan
off +
dryer on
Radon fan
off +
kitchen
fan on
Radon fan
off +
kitchen
fan +
dryer on
Leakage
area,
(in2)
118
113
105
97
85
cfm
available
at 1
pascal
111
102
92
73
57
"n" value
0.557
0.577
0.593
0.655
0.704
Leakage
ratio
(in2/ ft2)
3.3
3.2
3.0
2.8
2.4
ACH50
4.5
4.5
4.3
4.4
4.1
There were two problems with the installation: (1) it was
discovered by the homeowner that the water heater would backdraft
when the kitchen exhaust fan was operated (if the door to the
basement was open) and (2) the radon concentration did not go
below 2 pCi/L as promised in our contract. Table 7 shows the
pre- and postmitigation radon testing. We believe that the
backdrafting was a pre-existing condition discovered by our
backdraft indicator and the increased awareness we created. It
is also clearly evident that the radon mitigation system
exacerbated (consumes 5 in2 of the leakage area) a likely pre-
existing condition. The kitchen exhaust fan consumes 21 in2 of
the 118 in2 available leakage area in the house.
-------
Table 7. Pre- and Postmitigation radon data — Luana home.
Location
Main floor
master bedroom
Main floor
northeast
bedroom
Main floor
living room
Basement north
side
Basement south
side
Pre-
mitigation,
pCi/L
4.6
5.3
not available
11.1
12.3
Post-
mitigation,
pCi/L (2" PVC)
1.4
1.6
1.4
3.9
4.5
The radon concentration did not decrease to below 1 pCi/L as
we usually find (the subslab communication is excellent to all
four corners). The residual radon must be entering through the
block walls. We see two choices as methods to reduce the radon
concentration to below 1 pCi/L: (1) install wall suctions or (2)
install a forced air ventilation (pressurization) system (FAVS).
There are some additional considerations. The children in
the home have allergies and window condensation appears in cold
weather. We think the FAVS approach would be most appropriate
because it will provide needed fresh air, should solve the radon
problem, and should prevent the natural draft water heater from
backdrafting. The FAVS may obviate the need for the SSD system.
DISCUSSION
Many homes are time bombs in the sense of having inadequate
ventilation; radon mitigation systems shorten the fuse by
dedicating a specific amount of a home's leakage area to radon
mitigation. The Spandex concept of "one size fits all" may work
for socks but does not work when installing ASD and SMD radon
mitigation systems in houses. There is too much variability in
the housing stock to assume that a conventional ASD radon
mitigation system consisting of four inch ID PVC and an inline
centrifugal fan capable of moving 270 cfm of air at zero inches
static pressure can be used safely in all houses.
Coyne1 states that "it is absolutely critical that anyone
seeking to repair ducts be aware of, and consider, the
interactions among components of the house. Combustion appliance
venting, indoor air quality, moisture, comfort, and occupant
behavior must all be factored in." "Proper training is
-------
essential, and repair work must proceed with caution to avoid
'solutions' causing problems of their own."
Coyne's recommendations apply directly to radon mitigation.
Radon mitigation is a science with too many unknown parameters.
We must as radon mitigators, appliance installers (clothes dryers
and exhaust fans), heating contractors, plumbers, builders,
consultants, researchers, etc. treat the house as a "system" and
not as an entity with disconnected functional units. The health
and safety of home dwellers are at stake.
Coyne recommends that "pressure differentials and spillage
should be tested before and after sealing ducts. . . to leave
the home safe and healthy." Radon mitigators must do likewise.
The air leakage of a home should be measured before and after
installing a radon mitigation system. The reduction in air
leakage area and cfm of air available at various pressure
differentials should be determined. Research needs to be
conducted to develop standards for the amount (cfm) of air
requisite for proper operation of the air-consuming appliances
and requisite for a healthy occupant environment at different
pressure differentials.
Some of the Environmental Protection Agency's Radon
Contractor's Proficiency training suggests using smoke with all
exhaust fans running to determine if proper drafting occurs.
Passive combustion air in the form of ductwork from the outdoors
to near the combustion appliances is suggested as a method of
providing air. This training gives a false sense of security.
The CMHC3 reports that "the provision of additional supply
air is not likely to be effective as a remedy for pressure-
induced spillage of combustion products if the supply air is
introduced unaided through a building envelope opening of any
size likely to be considered practical." Some homeowners
recognize that environmental conditions affect drafting. A smoke
test at a point in time is not a reliable predictor of adequate
air for drafting of natural draft appliances.
RECOMMENDATIONS
All radon mitigation systems installed in Iowa (and other
cold climates) homes should be "recalled" and evaluated; i.e.,
air leakage area and cfm air available should be measured with a
blower door and a determination (best estimate) made if the homes
have adequate air available for the air-consuming appliances.
Air leakage characteristics should be determined before and after
installing radon mitigation systems in cold climate homes.
-------
The Canada Mortgage and Housing Corporation2 (CMHC)
suggests that "the most appropriate material for indicating a
temperature rise caused by spillage from a gas furnace was found
to be temperature sensitive labels or dots." Backdraft
indicators should be installed on all natural draft gas furnaces
and water heaters. Electronic carbon monoxide detectors should
be installed in "suspect" homes.
Due to the complexity of installing radon mitigation
systems, do-it-yourselfers should not be encouraged.
Relatively fool-proof procedures need to be developed to
allow a systematic approach to the installation of a safe radon
mitigation system in cold climates.
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
Coyne, B. "A Million Miles of Ducts: Duct Sealing Update,"
Home Energy. 9 (2):14-20 (1992).
Residential Combustion Venting Failure - A Systems Approach
Country-Wide Survey; Development and Testing of Spillage
Detectors. Canada Mortgage and Housing Corporation, Jan.
1987, p 11.
Residential Combustion Venting Failure - A Systems Approach
Summary Report. Canada Mortgage and Housing Corporation,
July 1987, p 117.
-------
VIIP-2
WHY WE LIKE DIAGNOSTICS
By: John W. Anderson, Jr. and Jack C. Bartholomew, Jr.
Partners in
Quality Conservation
E. 5805 Sharp, #A8
Spokane, Washington 99212
ABSTRACT
We think building diagnostics are valuable .when bidding,
designing, and installing radon mitigation systems.
Diagnostics start with the homeowner interview and mapping the
plan of the house in contact with the soil, air flow study with a
smoke gun, pressure measurements, radon sniffing, and sub-slab
diagnostics, and end with system pressure documentation, testing,
and customer satisfaction.
These techniques enable us to design and install radon systems
that are not oversized and require less energy to operate, and
with less than a 1% call-back.
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.
-------
WHY WE LIKE DIAGNOSTICS
RADON SYSTEM INSTALLATION DIAGNOSTICS
Some people swear by them.
Some people swear at them.
Some people say what?
And some do them only if there are problems with an installed
system, which is mostly the case since 90% of radon systems are
installed with no diagnostics before or after except for the
short-term radon tests.
Some contractors have said they only have to return to modify
10% of their systems without using diagnostics. We like our
return rate of less than 1% with diagnostics and it helps
customer satisfaction, referrals, and the, "You want it done
when?", time line.
Diagnostics help design and install systems that will work the
first time with few or no call backs.
This type of building investigation is 50% science and 50% art.
It is the sum of observation, measurement, and intuition which
comes from familiarity with diagnostics and radon system
installation performance.
DIAGNOSTICS BEGINS
This process begins with the homeowner interview and is
followed by a thorough inspection of the house.
We look for:
1. Type of furnace system
2. Type of ventilation system
3 . Type of foundation
4. Slab floor characteristics
-------
5. Crawlspace access (if it has one) and headroom
6. Attic access and headroom
7. Stack and mechanically induced pressure differences
8. Air movement using a smoke gun
9. Possible entry points by radon sniffing
10. A route out of the building for the radon vent pipe
This inspection takes about 1 to 1 1/2 hours.
Diagnostics performed once the job begins are:
1. Sub-slab sniffing through test holes drilled in the floor
2. Sub-slab communication testing
These two procedures add about 1/2 to 1 hour to the system
installation time and help determine radon sources under the
slab, and sub-slab pressure reach. Most of these holes should be
drilled to test the installed system performance anyway.
WHY DO WE LIKE DIAGNOSTICS?
A HOMEOWNER INTERVIEW
Meeting the homeowner helps answer questions they may have
about radon and entry control, uncovers concerns about radon and
system routing, may direct the investigation, and establishes a
time-line for the system installation.
Asking for floor plans and construction pictures (if
available), can help locate hidden utility chases for pipe
routing or attic bypasses, can indicate if there may be gravel
beneath the slab, and can speed the foundation plan measurement
process especially in complex houses. Unfortunately, very few
homeowners have these plans or pictures, but it's worth it to
ask.
-------
A THOROUGH INSPECTION AND MAPPING OF THE HOUSE
This cannot be done sight-unseen.
A careful eye and knowledge of construction practices helps
locate possible entry points and pipe routes out of the building.
It has uncovered details that other contractors have missed such
as dirt floored areas under stairs and furnace return plenums
that pull air from under slabs. Finding details that others miss
really helps in landing the job, and may uncover details that
make the mitigation system easier to install, less expensive for
the homeowner, or may cause you to bid high and someone else gets
the job.
TYPE OF FURNACE SYSTEM
Noting the type of furnace system...gas, oil, electric, solid
fuel; atmospheric burner, draft induced, or sealed combustion?
Will backdrafting be a concern, or is it already a concern?
Is there a need for make-up or dedicated combustion air. What
about draft inducing atmospheric gas burners for furnaces and
water heaters?
Is the system ducted, baseboard, or radiant heat in the slab
floor or ceiling?
How will the heating system type effect the radon system
installation?
TYPE OF VENTILATION SYSTEM
What type of ventilation system is there (bath fans, down
draft range vent, exhaust only, or balanced)?
How will its operation effect negative or positive pressures
in the building? It is important to test mechanically induced
pressure differences with major exhaust fans on and off (200
CFM+).
Some of the newer / tighter houses can be depressurized to 50
pa (0.2" WC) with just 200 CFM of air exhausted from the
building.
-------
TYPE OF FOUNDATION
The type of foundation...concrete, block, rubble, or wood?
What is special about each of these foundation types and how will
it effect the installed system?
How much of the foundation is accessible?
SLAB FLOOR CHARACTERISTICS
What condition is the slab in? Are there cracks, holes,
carpet, or vinyl. Is the cold joint between the slab and
foundation accessible?
CRAWLSPACE ACCESS
Accessibility of crawlspaces and use, and how may the
mitigation strategy effect the house?
1. Pipes. Are they insulated? Can they be freeze
protected?
2. Ducts. Are they sealed and insulated? Can they be
sealed and insulated?
3. Furnace. Provisions are needed for access.
4. Storage. If stuff is stored in this space, then a
durable membrane is needed.
5. Insulation. Is there any of this itchy stuff? Is it at
the perimeter or under the floor?
We prefer using sealed sub-membrane ventilation because it has
a lower energy penalty than sucking on the crawlspace (especially
true if there are uninsulated heat ducts,pipes and floors),
however we have treated several inaccessible crawlspace houses
with unprotected pipes, by pressurizing the crawlspaces with
house air.
ATTIC ACCESS
Accessibility of attics. You want the fan where? We try to
route pipes through closets to keep as much of the system as
possible in the heated envelope, and place fans as near an attic
access as possible because there are electricians who are 6'5"
and 250 pounds who don't do well in attics. Also, anything
mechanical will eventually need replacement and should be easy to
service.
-------
MEASUREMENT OF STACK AND MECHANICALLY INDUCED PRESSURE
DIFFERENCES
Do these forces contribute to radon entry and what will be
their effect on a mitigation system?
What effect does furnace fan operation have? What effect do
exhaust fans have?
Are these significant and need to be dealt with?
Fifteen pascals positive has been measured under slabs when
the wind blows on some houses. We know of a 9 story building in
which expansion joints were sealed with gun grade urethane and
the material kept bubbling from the slab joints.
Does a forced air furnace create negative pressure in the
basement and how much? Some house basements have been
depressurized by 10 pascals (O.OA inch WC) when the furnace fan
runs, which is enough to overwhelm most sub-slab suction systems.
Commercial buildings can be much worse.
The bottom line is these pressure forces are the major reason
radon enters a building and can, in some cases, be great enough
to counteract the effects of a pressure field under a slab. We
think they need to be measured so an effective sub-slab
ventilation system is installed the first time. Well, 99% of the
time.
SMOKE GUN ANALYSIS
Watching how air behaves with mechanical systems on and off
helps to tell which way air is flowing, and if pressure
differences are being created. Does air flow from under the slab
into the basement through cracks and pilot holes? What effect do
mechanical systems have on air flow and pressure differences?
A small, teflon bottle filled with fiberglass and 5 cc of
titanium tetrachloride, with a tight fitting cap can be really
easy to carry in a pocket.
LOCATION OF POSSIBLE ENTRY POINTS
Knowing what's hot and what's not helps in directing the
mitigation approach.
-------
In a crawlspace / basement combination house, sniffing can
determine if the pollution is coming from the crawlspace, the
basement, or both. We have found crawlspace / basement houses
where the crawlspace is not the problem and have successfully
lowered radon levels by applying sub-slab suction to the basement
slab and tightening the wall between the basement and the
crawlspace. Conversely, we have found crawlspace / basement
houses that only needed the crawlspace treated. A stepped
approach can save the homeowner big bucks.
In marginally polluted houses, sniffing may indicate that the
sealing of significant entry points may control radon levels.
This has been found with untrapped floor drains that have geysers
of air into the house 3 feet high and test with a sniffer at 300
pCi/1.
Radon sniffing can confirm elevated radon levels.
SUB-SLAB COMMUNICATION TESTING (SLABS.,.ALWAYS)
It is valuable locating sumps, number needed, and to tell if a
sump is large enough (would a larger sump, trench, another hole,
or stitching work better?) Sometimes the preferred location for
a sump has no reach or maybe under just half the floor where an
alternate location 10 feet away shows reach everywhere
(difference in hole locations has been as small as 3 feet on the
same side of a footing).
Sub-slab communication testing is simple to do during
installation, and maybe adds 1/2 to 1 hour to the installation
time. We call the suction point the main test hole (1 1/4" to
fit the vacuum) and the test holes around the perimeter (1/2"
holes) the pilot holes. There is the initial testing through the
main test hole(s) to judge pressure reach and sump location, and
then several tests of the sub-slab communication while digging
the sump helps to see if it is large enough. The pilot holes at
the perimeter are needed to check performance once the system is
installed anyway.
Sub-slab communication testing can help size system fans. The
measurement of sump pressure 1 cubit from the main test hole (the
distance from your elbow to your finger tips or as far as can be
easily dug out) shows sump pressure needed to get the same
results at the pilot holes. Measurement of air flow shows how
much air needs to be moved. The combination of sump pressure and
CFM indicates what size fan is needed (F-100, F-150, F-160,
F-175, etc.) once the static losses in the pipe are accounted
for.
-------
Even in houses that for some reason didn't let us drill all
the holes wanted, and didn't submit to good communication, the
result is large sumps that obtain a lot more reach than if the
system were installed by eyeballing it and using some rules-of-
thumb for pit size. Uh...how's that look George... George says it
looks good so it must be good. Yup... George is my friend...!
will follow him anywhere...
SUB-SLAB SNIFFING
Sniffing is often a useful diagnostic procedure for locating
sources under the floor and directing the focus of the sub-slab
ventilation system toward the areas of highest radon
concentrations. Sub-slab sniffing must be done before
communication tests to sample the soil "as is" because once
negative pressure has been applied to the soil under the slab,
the distribution of radon changes.
Here's a question as in a familiar test.
Sub-slab sniffs of 150 to 300 pCi/1 everywhere except for a
corner at 6000 pCi/1.
Where would the system pressure field most definitely have to
extend to?
1. The fireplace
2. Next to the floor drain
3. The corner with 6000 pCi/1
4. The clothes hamper
There is no clue if sources haven't been sniffed out.
WE RARELY USE OUR DOOR FAN IN RADON WORK
It is helpful to have the ability to test building tightness.
A door fan test can asses the viability of ventilation strategies
and how much air will be needed, and can locate bypasses (ie
mechanical chases that contribute to stack effect.
We have used a door fan to simulate winter stack effect on a
warm summer day to help determine entry points and sources.
We proposed taking care of radon entry in an older
inaccessible crawlspace house by pressurizing the crawlspace with
house air. We didn't feel easy about exhausting air from the
crawlspace because of unprotected water pipes. A door fan test
of the crawlspace told how much air was needed to dump into the
-------
crawlspace to pressurize it to 5 pascals, and helped locate air
leaks that could be sealed to reduce the amount of air needed
from the house. The house was then door fan tested to see what
effect 150 cubic feet of air per minute exhausted from the house
would have. The effect was found to be about the same as "C" or
the flow through the house at 1 pascal pressure (in this case "C'
was 150 CFM). Through two severe winters the occupants have
remarked how much more comfortable their home is.
The door fan has been used to test the pipe tunnels under a
school for the amount of air needed to be exhausted to create a
negative 25 pascals in the tunnels. The radon mitigation system
fan was sized with the help of this information. Asbestos was
not a problem but definitely would be if present.
THE "GEE WHIZ" ASPECT OF DIAGNOSTIC EQUIPMENT
It can help sell the job. What, more flashing lights and
buzzers? Neat! It shows you care enough to analyze radon entry
and behavior to assure control of radon levels.
CASE STUDIES
A HOUSE
The house has an indoor lap pool (12 X 76), basement, and two
crawlspaces. When we were first called by the homeowner, the
concrete deck around the pool had not been poured.
Upstairs radon was measured at 24 and 31 pCi/1. The pool room
was 193 pCi/1.
Four inch ADS perf was looped around the pool and stubbed up
for an easy exit from the house. The pool deck was poured, slab
joints were sealed, and the house was retested in seven places.
Of the seven test locations the 2 crawlspaces had the lowest
readings of 0.7 and 8.0 pCi/1. Basement readings were 19, 27,
and 38 pCi/1, and upstairs were 9 and 12 pCi/1.
The beauty of this installation was that the crawlspaces were
investigated to see if they were a source, and were not mitigated
because they did not have to be. Que the happy homeowner and
another good reference!
-------
ENGLAND
We were asked by a company in England to travel across the big
pond and visit their country for three weeks in May '91 to teach
them how to do diagnostics and install radon systems.
Familiarity with diagnostics really saved our cookies or is it
crumpets? There we were. The same language was spoken, but
somehow it was different. Cars were driving by on the left side
of the road, but there were no drivers, just passengers. And the
people we met were very courteous.
They usually build their homes with masonry exterior and
interior walls, and each wall has its own foundation and footing.
Each room is effectively cordoned off from the others by
foundations. Five homes were mitigated.
Two were a combination of basement, suspended timber floors,
and slabs on grade, two were slabs on grade, and one was a
suspended timber floor with and a filled-in basement combination.
Ages of the houses was 25 to 300 years.
Sub-slab and crawlspace communication testing proved helpful
in locating suction points. A combination basement, timber
floor, and basement house was mitigated with a central suction
point that accessed four rooms. This location was found through
sub-slab communication testing that required 4 different main
test holes.
Sub-slab communication testing helped judge pressure field
extension and locate cold joint leaks. Cold joint short circuits
in the 2 slab on grade houses needed to be sealed before
communication was achieved. The sub-slab communication testing
helped que us to the slab/foundation cold joint. One house was
mitigated with one centrally located sump and the other was fixed
with 2 sumps, one at each end of the house.
A radon system was installed in a house with a basement floor,
two slabs on grade, and a suspended timber floor. Monitors were
left and results were obtained a few days later. The basement
was still hot. Sniffing found sources at the basement foundation
walls.
SCHOOLS
Stevens
We described Stevens Elementary School in our bid to the
school district as a mitigator's nightmare... It was.
-------
This building has a footprint of about 12,000 square feet, was
built around the turn of the century with rubble foundation
walls, and half the basement was dirt floored. Finished basement
areas were used as classrooms.
Initial sniffing for radon sources showed the foundation walls
not to be a source, except in room 11. A couple of pipe tunnels
were sources, and so were the dirt floors.
Room 11 was door fan tested to see what pressures and CFMs
were needed for flow reversal through the cracks in the
foundation walls. We then pressurized room 11 to 50 pascals
which helped us locate leaks in the rubble foundation and seal
them to limit radon entry.
The dirt floored areas were prepared for sub-slab ventilation
with 4 inch ADS perf and then capped with 4 inches of concrete.
Have you ever tried moving concrete trucks through a playground
full of kids? Score that day...Cement Trucks 0, Little Kids 6.
The sub-slab ventilation systems were completed and the accesses
to the pipe tunnels were sealed.
All rooms in the basement were then tested. Radon had been
controlled in all rooms except for 2 adjoining rooms, numbers 10
and 12 which were now between 4 and 16 pCi/1.
More diagnostics! Where was the source? Now with radon
levels much lower throughout the basement, the foundation walls
in these 2 rooms showed up hot.
The sub-slab ventilation system for these 2 rooms was reversed
to pressurize under the floor and reduced radon levels below 4
pCi/1, but it also pumped a lot of moisture into the rooms. The
system was again modified to depressurize under the slab
(original configuration).
The 2 rooms were door fan tested to check tightness, entry
point flow reversal pressure, and cubic feet of air per minute
needed to do it. The door fan testing also revealed a couple of
mechanical chases that led to the attic.
The rooms were tightened to reduce the amount of air needed to
be vented into them to reverse entry point flow. The exterior
foundation walls were not suited for crack sealing because of
crumbling mortar.
A 300 CFM fan was mounted in the transom over the door to
pressurize rooms 10 and 12 with air from the hallway.
Subsequent testing has shown levels stay down as long as the
-------
door to the rooms is closed at night, indicating that 8 to 12
hours of room pressurization a day is enough to help control
radon entry through the foundation walls in these 2 rooms.
Clover Park
One western Washington school district tested all their
buildings and discovered most rooms OK with just a few of them
coming back over A pCi/1. Upon inspection we found a close
association with inadequate fresh air and radon. We checked the
damper operation of the unit ventilators and found most not
working, and in those that were working we found blocked air
intakes. Carbon dioxide was also found to be around 2000 ppm.
Increasing the amount of outside air delivered to the rooms would
have the double impact of reducing radon levels and providing
enough fresh air to keep the students and teachers awake for a
better learning environment.
Greenacres
We diagnosed and prepared a mitigation plan for a slab-on-
grade elementary school built in 1979. Politics of the situation
were that, if during Christmas vacation, we were not able to
lower radon levels substantially, the school would have to be
closed.
There were 5 HVAC systems with return ductwork under the slab
and radon test results from each room showed a strong correlation
to each HVAC system.
Radon sniffing was done in the return ducts at each air
handler. Radon sniffs of the main HVAC system showed the highest
levels and the area served by this system tested the highest in
the building. The return ductwork had to go.
Could the sub-slab return ductwork be used as a sub-slab
ventilation network to control radon entry?
The next step involved drilling 15 test holes through the slab
to test pressure field extension from the return ductwork. The
return grills were plugged, and the return fan was carefully
brought up to 1.5 inches WC hoping the ductwork would not
collapse. 7000 CFM was measured being drawn through the return
air ductwork. Most of the test holes registered between -2.0 and
-30.0 pascals.
7000 total CFM (measured)
1200 floor grill leakage (measured)
-700 above grade return duct leakage (assumed)
5100 total CFM needed to create same pressure field
-------
Since 5100 CFM produced a large and consistent pressure change
under the slab, we figured one half that amount would still
control radon entry.
There was a plan! The return ductwork under the slab would be
abandoned, modified to be used as the subslab ventilation system,
and return ductwork installed in the ceiling of the building.
-------
VIIP-3
An Approach To Computer-Assisted Radon Mitigation
by
Hormoz Zarefar, Pah Chen, Patricia Byrne
Portland State University
Mechanical Engineering Department
P.O. Box 751
Portland, OR 97207
and
Charles Eastwood
Bonneville Power Administration
P.O. Box 3621-RMRD
Portland, OR 97208
ABSTRACT
A computer aided approach to radon mitigation is discussed.
The proposed methodology involves the development of a knowledge-
based system as well as procedural computational tools to aid the
mitigators in the selection and design of mitigation systems for
existing homes. The computer software (RnX) requests pertinent
data about the house, processes this data, and recommends an
appropriate mitigation method. For sub-slab suction techniques,
the system assists in determining the number of needed suction
points and performs a fan selection and cost estimation. The
development of this prototype software involved the incorporation
of knowledge in existing literature, communication with field
experts, and the creation of rules for representing, converting,
and applying data in order to provide recommendations to the end
user.
The software offers many tools to facilitate various aspects
of radon mitigation such as record keeping, compliance with
established EPA protocol, consistent fan sizing and cost
estimation.
-------
INTRODUCTION
Radon mitigation can be considered a multi-faceted, knowledge-
intensive problem solving process. Generally, the mitigators rely
on their prior on-the-job experience, rules of thumb and common
sense to design and install a mitigation system. At the same time,
they must follow current EPA guidelines on the design of mitigation
systems for existing homes.
The current project formalizes some of the aspects of
mitigation method selection as well as provides the mitigators with
the methodology used to select the required mitigation method,
sizes the fan (if needed) and provides a detailed cost estimate.
This paper describes the development, goals, significance, and
characteristics of the prototype hybrid knowledge-based expert
system for indoor radon mitigation. Some features of user-
interaction and the capabilities of the system including the use of
the system as a potential tool for training novice mitigators are
discussed.
KNOWLEDGE-BASED (EXPERT) SYSTEMS
Knowledge-based (expert) systems are computer programs which
result from incorporating and codifying heuristic and expert
knowledge. Knowledge-based systems applications have been reported
in various fields including diagnostics, monitoring, planning,
trouble shooting, and design. There have been several attempts in
the application of knowledge-based systems to building design (1,
2), and efforts in applying expert system technology to assist in
radon mitigation have also been reported. An initial attempt was
made by Mosley in 1987 (3) , and a demonstration system on a
Macintosh computer was developed by Brambley in 1990 (4). In
addition, an interactive system was developed for a" Macintosh by
Brennan in 1990 (5). The demonstration system illustrated the
usefulness of user-directed point-to-point hypertext when working
with large amounts of textual information. The interactive system
was designed to assist in the training of mitigation contractors.
These ventures demonstrated the capability of expert systems in
dealing with radon mitigation.
A knowledge-based expert system is typically comprised of four
main components; an inference engine, knowledge base, working
knowledge, and user interface. The relation between these elements
is shown in Figure 1. The inference engine implements the problem
-------
solving strategy. The knowledge base contains the knowledge
pertaining to the solution of the problem. The working knowledge
represents data relevant to the problem at hand. This data is
extracted from the user through the user interface, which may
utilize a variety of techniques for querying the user (6).
User
Interface
Inference
Engine
Knowledge
Base
Working
Knowledge
Figure 1. Major components of an expert system.
The heuristic knowledge that characterizes the field of radon
mitigation is generally a surface (experiential) knowledge.
Surface knowledge is heuristic, experience-based information that
is the result of successfully solving a large number of similar
problems. On the other hand, there exists a body of knowledge
which are well defined and are based on scientific investigations
and mathematical and physical sciences. They embody the segment of
the radon mitigation task which includes ducting analysis, fan
sizing, and costing. These are deep knowledge which, if well
defined and formalized, can be programmed via algorithmic or
procedural routines (7) . We define a hybrid system as an
integrated computer program containing knowledge-based and
algorithmic subprograms (8).
Expert systems are suitable programming environments when
experts are available in the subject, but are in short supply, thus
dissemination of knowledge becomes a crucial task. The knowledge
base must be fairly narrow and should not vary (fluctuate) . Expert
systems are also suitable where the task takes a significant amount
to do manually and there are many factors involved in a decision;
a poor decision will make a significant difference in the outcome
when the competitors are performing the job consistently. These
classes of programs can handle incomplete and/or uncertain data by
making reasonable assumptions (similar to an expert) and proceed to
come up with an answer.
-------
There are several advantages associated with the knowledge-
based (expert) systems. Knowledge-based systems allow experts to
dedicate more time to tasks that require human ingenuity and
creativity. The programs can be employed in training novices while
simultaneously providing a collective pool for sharing and
propagating knowledge. Expert systems also permit the
standardization of processes and techniques.
RNX - A RADON MITIGATION EXPERT SYSTEM
Considering the potentials of knowledge-based development and
the issues of concern in radon mitigation, a hybrid knowledge-
based/algorithmic approach was selected. The computer program,
which is regarded as an advisory system, is a modular program
composed of several segments. A modular structure was adopted to
limit the size of the separate components, and to facilitate error
detection and correction, and potential expansion of the system.
The modules, which include a house investigation summary, fan
selection, and cost estimation, are activated hierarchically, as
shown in Figure 2.
House investigation
information
HOUSE INVESTIGATION
SUMMARY MODULE
Ducting information.
Number of points
H
i
FAN SELECTION
i
COST ANALYSIS
I
Figure 2. A representation of the system modules.
The knowledge-base was constructed by employing a rule-based
architecture, in which the knowledge is stored as a series of IF-
THEN statements. This particular architecture was deemed most
-------
appropriate after careful observation of structure of information
presented in the EPA publications and analyzing the sequence of
events and decisions which lead to the selection of a particular
radon mitigation scheme.
The computer program was developed using the expert system
development environment LEVELS/OBJECT running under Microsoft
Windows Graphical User Interface. This particular combination
enhances the presentation of information to the end user and makes
it possible to incorporate graphical as well as textual
information. Each module is composed of its own demons, methods,
classes, attributes, and instances (9). Demons and methods are the
IF-THEN rules of the system. Demons are the heuristic rules and
are different from methods. Methods are purely procedural, like a
simple computer code. Variable names and their related properties
are grouped in classes and attributes. A class encompasses the
attributes and instances, which are a group of related objects.
For example, the class "House" may include an attribute such as
"foundation type." An instance of foundation type may be
"crawlspace." So, to specify the foundation type of the house in
an object oriented fashion, it would be stated as: "foundation
type OF house IS crawlspace."
Some of the modules use the same classes and attributes. The
values of these attributes are obtained from the user by one of the
modules, and passed between the modules during execution. The
method used to pass the data was to write the information to a text
file which in turn is accessed by a subsequent module.
The hybrid nature of the entire system is evident in its
integration of heuristic and procedural methods. The heuristic
portions of the rule base pertain mostly to the mitigation method
selection. The procedural methods are exclusively for the
numerical computations associated with ducting analysis, fan
selection, and cost estimation.
PROGRAM STRUCTURE
RnX begins by performing data acquisition. The first module,
which is the House Investigation Summary module, extracts the
characteristics of the house. This module was designed to be
analogous to the house investigation summaries that mitigators
often use to gather pertinent data about a particular house (10,
11, 12). The user is prompted to either create a new file or to
edit an existing file. This user-named file contains the input
data. If the user has opted to edit an existing file, the data
from that file is loaded into the system and all previous entries
are displayed and may be changed. Once data gathering is
concluded, the new or newly-edited file is saved. Finally, the
module assesses the input data and recommends one or several
mitigation methods. This is followed by an explanation of the
-------
logic that was followed in determining the recommendation. A
sample explanation screen is shown in Figure 3. If the recommended
mitigation method(s) is or includes sub-slab suction, the module
proceeds to recommend the number of slab penetrations for the
ventilation system. This number is based on data pertaining to the
degree of sub-slab communication, the area of the slab, and the
presence of footings that partition the sub-slab area and could
interfere with the pressure field extension. At the conclusion of
this module, the user may choose to exit the system, or proceed
with either a fan selection or a cost estimation.
RnX- House Investigation Summary
File
Explanation for the recommended mitigation method
The highest measured radon concentration was specified as 40 pCi/L. The type of foundation
that was specified included a crawlspace. The highest measured radon level is less than or
equal to 40 pCi/L and does not necessitate an active ventilation mitigation method, so
passive (natural) ventilation of the crawlspace is recommended. Also, since there are not any
utilities present in the crawlspace, there is little danger of freezing. The type of foundation
specified included a basement and/or a slab. It has been specified that there are not exterior
or interior footing drains present so draintile methods are not applicable. The sub slab
communication was not very good, or was not available, but the foundation walls are mostly
poured concrete, so block wall ventilation will not be applicable. Because of this, sub slab
communication is recommended.
Figure 3. An example of the explanation facility.
The Fan Selection module performs a ducting analysis and fan
selection for sub-slab suction techniques. This module can analyze
ducting systems with up to three branches. For a great majority of
U.S. houses, three branches (or three suction points) are
sufficient for reducing the radon level. At the time of this
writing, the module supported duct diameters of 3, 4, and 6 inches,
and modifications were being effected to enable it to also support
diameters of 2 and 5 inches. This module requests the data
necessary to perform the calculations for the ducting analysis and
subsequent fan sizing. The user is presented with a screen
displaying the total air flow of the system along with the total
system friction loss (Figure 4) . In addition to these fan
requirements, the user is shown the maximum velocity that is
achieved in the duct, and for multiple branch systems, the user is
also shown the ratio of the branch friction losses. The optimum
value of the ratio should be around 1.0 to provide equal pressure
-------
drop to the last pipe junction point. Otherwise, less flow from
the branch with the higher pressure drop will render that branch
ineffective in reducing the radon level. To achieve the ratio of
1.0, one can change the pipe size, fitting type, or add pipe length
or a flow damper to balance the pipe network.
RnX- Fan Selection Module
file
An appropriate fan will meet the following flow and pressure requirements:
95 cfm 0.332 WC ( 82 Pa)
For the given diameter(s) and flow rate,
the maximum velocity in the duct is:
Ratio of pressure drops
611 fpm
Two branch system, branch 1 to branch 2 : Not Applicable
Three branch system, branch 1 to branch 2 : 1.08
Three branch system, intermediate branch to branch 3 : 0.75
More on
pressure drop
ratio
More on
maximum
velocity
Next
screen
Figure 4. A screen from the Fan Selection Module.
A small database of brand name fans has been incorporated into
this module. The system determines the particular models of the
brand name fans that are capable of satisfying the fan requirements
that were presented earlier. These models, with their respective
purchase costs and power consumption ratings, are displayed in
order for the user to select one for use with the cost estimation.
At the conclusion of this module, a temporary file containing the
data concerning the diameter, length, and number of fittings used
in the ducting system as well as the fan data is written to the
disk. This is considered a temporary file because subsequent runs
of the module will overwrite the file. Lastly, the user may
proceed with a cost estimation.
-------
There are currently three modules that perform cost
estimations. These modules are customized to the tasks required
for either block wall suction, sub-membrane suction, or sub-slab
suction. If the user is entering the cost estimation directly from
the fan selection process, the cost estimation reflects the tasks
that are needed for sub-slab suction. The tasks are itemized and
the user can enter the amount of time each task will consume along
with the number of times that the task must be performed. This is
used to determine the number of hours that are spent on the job.
The labor rate is entered by the user. The cost estimation also
covers material costs. The module also determines the material
costs for the ducting system by accessing the data file generated
by the fan selection module. The total cost for labor and the
total material cost is considered to be a one-time installation
cost. The annual cost to the home owner includes energy costs and
maintenance. Energy cost is determined by the power consumption of
the fan, regional cost per kilowatt hour, and the number of hours
that the fan is in operation (usually year-round). The cost for
maintenance varies regionally. The user can incorporate a discount
and/or an overhead factor. The module is equipped with default
values for the various costs, and all of them may be changed.
All of the modules incorporate some form of error checking.
The system checks for unusual entries and for conflicting data.
For example, if the user specifies that the house is 200 years old,
or that the ceiling of the basement is 88 feet tall, these may be
considered unusual entries. The user is notified of the situation
and may either verify or correct the value. If conflicting entries
occur, the user must make a correction or the program will not
proceed. In some cases the user may not have all of the data that
the system requests. For example, sub-slab communication testing
may not have been performed for some reason or another. If that is
the case, the user selects the response "none observable or no
testing performed." By assuming the worst case, the system will
produce a recommended solution that is conservative.
Once the House Investigation Summary module has been executed
for a particular case, subsequent modules can be repeatedly
executed without invoking the previous module. This arrangement
makes it possible for the user to reactivate individual modules as
many times as is desired. An additional and completely independent
module is the Fan Tutorial, which is similar to the Fan Selection
module. This module provides on-line assistance and defines the
notation that is used for the Fan Selection module.
RnX recently has been reviewed by members of the Northwest
chapter of AARST. After the review comments were received, the
system was modified to eliminate difficulties and incorporate
features which had been overlooked during development. In addition
to debugging, the user-friendliness and the error-detection
capability of the interface were increased.
-------
ADVANTAGES OF A COMPUTER-ASSISTED APPROACH
One of the most attractive features of a computerized process
is the capability to create and store systematic records. The
records are easily updated and have other potential uses. As
previously mentioned, RnX has the capability to create and maintain
a file which contains the house investigation summary data. This
file is created in the initial data run, and may be updated by
employing the editing feature.
Another important feature is that computerized assistance
lends consistency and reliability. The mitigation method selection
process uses pertinent information only. The recommended method is
followed by an explanation which furnishes the user with the logic
followed in determining the recommended mitigation method. An
explanation facility reveals that the solutions are deduced in a
systematic fashion, and shows which data were used to arrive at the
solution. It may also enhance the experts' own ability to suggest
consistent solution methods. In addition, significant attempts
have been made to comply with present EPA protocol.
The benefits in terms of computational efficiency are obvious.
The numerical nature of the fan selection and cost estimation
modules are especially suited to coding. The additional features
of the fan selection module (velocity and pressure drop checks)
provide a pre-design refinement capability. The advantage here is
that the necessary calculations for alternative designs may be
carried out in rapid succession until a suitable design is reached.
When entering the cost estimation directly from the fan selection
module, the user is spared the task of recording some of the
material cost data, since this is transferred via the temporary
file that is created by the fan selection module.
ADVANTAGES OF RNX
As the preceding discussions illustrate, the proposed
knowledge-based advisory system (RnX) can be beneficial to the
mitigators in terms of record keeping, observance of established
EPA protocol, consistent fan selection, and cost estimation.
The ability to store the interactive house investigation
sessions allows the mitigators to keep, update and store the
information as business records as well as allowing for in-depth
analysis such as observing trends in radon concentration levels and
the ability to infer suitable mitigation methods for similar data
files. Furthermore, the advisory system has the potential to be
interfaced with a database. This would facilitate data management
and would make it possible for mitigators as well as health and
regulatory agencies to keep a database of cases for a particular
-------
region. The information contained in such a database may be of use
to planners and researchers involved in radon related studies.
The advisory system provides a form of quality assurance for
mitigators just entering the field. The advisory system in its
final form will have incorporated the suggestions of field experts,
and will reflect their collective opinion. Therefore, the novice
user will have the assurance of the expert opinion, and the client
will be reassured that the recommended mitigation method complies
with established protocol and professional practice.
The Fan Selection module alleviates the amount of computation
that must be performed during the design of a ducting system.
Since the module also informs the user of potential noise problems
and imbalances, it encourages fine-tuning of the ducting system
before the actual installation. This may help to manage the amount
of time that is spent on installation, and may reduce the
possibility of the need for follow-up corrections to ducting
systems. Another benefit of the module is that it provides
consistent results for similar cases and will prevent possible
oversizing of the fan. An oversized fan consumes more energy and
raises operating cost.
The Cost Estimation module is an efficient way for the
mitigator to provide a quick estimate to the client. It will also
assure the client that the quoted price is justified. The module
is flexible enough to include unforeseen costs involved in the
installation and maintenance of a mitigation system. When used in
conjunction with the Fan Selection module, the cost estimations for
several alternative designs may be presented to the client. This
would also be a benefit to the client who is deciding between
several possible ducting configurations.
Aside from assisting established professional mitigators, the
advisory system illustrates the potential for a knowledge-based
system to serve as an interactive training tool for novice
mitigators. A modified version of the software can be developed
with the eventual goal of training novice mitigators. A successful
training tool should have a simple and highly visual method of
communicating the knowledge to the novice, and a graphical user
interface such as the one employed in the advisory system is a
fitting representation of this technique. An additional
consideration is that the effectiveness of a computerized training
tool is reasonably dependent upon the availability of the computers
needed to run the software. Current trends in the PC market are
evidence that advanced technology is becoming increasingly
available at a lower cost. In addition to analyzing single family
residences, the present system could be expanded to incorporate
other building types such as schools and commercial offices.
-------
SUMMARY
The feasibility of applying expert systems methodology to the
problem of indoor radon mitigation has been illustrated by this
work. The prototype RnX hybrid advisory system addresses various
facets of the radon mitigation problem, from the selection of a
mitigation method to the determination of necessary building
materials and cost estimation. Efficient modification and the
implementation of experts' opinions is facilitated by the modular
structure. To date, the system has received favorable reviews
after demonstrations at several conferences and meetings. At the
time of this writing, the remaining tasks for the completion of
project includes the addition of recommendations submitted by radon
mitigation experts who have been contracted to review the software.
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
The authors acknowledge the Bonneville Power Administration
for its funding of this work under research contract IAG-07967.
We wish to thank the members of the Northwest Chapter of AARST
for their review of RnX.
-------
REFERENCES
1. Hitchcock, R.J. "Knowledge-Based System Design Guide Tools."
To be included in ASHRAE Transactions, V. 97 Pt. 2, 1991.
2. Mayer, R., Degelman, L.O., Su, C.J., Keen, A., Griffith, P.,
Huang, J., Brown, D., Kim, Y.S. "A Knowledge-Aided
Design System For Energy-Efficient Buildings." To be
included in ASHRAE Transactions, V.97, Pt. 2, 1991.
3. Mosley, R.B. Personal communication, 1990.
4. Brambley, M.R., Hanlon, R.L., Parker, G.B. "Expert Systems:
A New Approach To Radon Mitigation Training And Quality
Assurance." Proceedings, Indoor Air Conference, Toronto,
Canada, August 1990, pp 483-487.
5. Brennan, T. , and Gillette, L.M. "Interactive House
Investigation and Radon Diagnostics Computer Program."
Proceedings of the 1990 International Symposium on Radon
and Radon Reduction Technology, January 1990.
6. Luger, George F. , and Stubblefield, William A. Artificial
Intelligence and the Design of Expert Systems. The
Benjamin/Cummings Publishing Company, Inc., 1989.
7. Dym, Clive L., and Levitt, Raymond. E. Knowledge-based
Systems in Engineering. McGraw-Hill, Inc, 1991.
8. Zarefar, H. "An Approach to Mechanical Design Using a Network
of Interactive Hybrid Expert Systems." Ph.D. Thesis,
University of Texas at Arlington, 1986.
9. LEVELS OBJECT User's Guide and Reference Guide. Information
Builders, Inc., 1990.
10. Mosley, R.B., and Henschel, D.B. Application of Radon
Reduction Methods. EPA/625/5-88/024, Air and Energy
Engineering Research Laboratory, Research Triangle Park,
NC, August 1988.
11. Brennan, T., and Galbraith, S. Practical Radon Control for
Homes. Cutter Information Corp., Arlington, MA, 1988.
12. Henschel, D.B. Radon Reduction Techniques for Detached
Houses, Technical Guidance (Second Edition). U.S. EPA,
EPA/625/5-87/019, Research Triangle Park, NC, January
1988.
-------
VIIP-4
RADON CONTROL: FIELD DEMONSTRATIONS
DIAGNOSTIC AND MITIGATION TECHNIQUES USED
IN TWENTY-SIX RADON FIELD WORKSHOPS
By: Craig E. Kneeland and Mark R. Watson,
New York State Energy Office
Two Rockefeller Plaza
Albany, NY 12223
Wade Evans
Evanshire Company, Ltd.
Rome, NY 13440
Terry Brennan
Camroden Associates
Oriskany, NY 13424
ABSTRACT
The New York State Energy Office conducts two-day radon field workshops as follow-up training
for attendees of its three-day "Reducing Indoor Radon" workshop. These workshops also meet the
training requirements for the EPA's Radon Contractor Proficiency Program. Field workshop sites are
single family homes whose owners have participated in an energy conservation program. During the
workshops students perform a variety of diagnostic tests, and design and install mitigation systems. To
date, twenty-six workshops have been conducted. Pre-mitigation test results from charcoal canisters
placed in the basements of the houses used in this program range from 20 to 522 pCi/l, with an average
of 101 pCi/l. Data collected during radon diagnostics and mitigation installation include: house
construction features, sub-slab radon concentrations; soil communication tests; and basement ambient
air radon concentrations before, during and after mitigation. The mitigation techniques used in these
houses include sub-slab depressurization (23 of the 26 systems installed), sub-membrane
depressurization (in three houses, in conjunction with sub-slab depressurization), sub-slab
pressurization/dilution (3 of the systems) and various sealing measures (all of the systems). Post-
mitigation radon measurements are made using a charcoal canister (2-5 days), continuous radon monitor
(7 days) and an alpha track detector (one year). The average post-mitigation radon measurements were
1.6 pCi/l for charcoal canisters (CC) and 1.4 pCi/l with a continuous radon monitor (CRM). Alpha track
detector results for the houses with depressurization systems average 0.6 pCi/l. Two of the three houses
with sub-slab pressurization/ dilution systems have been problematic. Pre-mitigation (CC) measurements
for these houses were 163 pCi/l and 298 pCi/l. Short-term post-mitigation measurements were
satisfactory: in one house the readings were 3.3 pCi/l (CRM) and 1.8 pCi/l (CC); in the other house the
readings were 1.5 pCi/l (CRM) and 0.8 pCi/l (CC). Alpha track detector measurements, however, were
19.0 pCi/l and 22.2 pCi/l. A case study of the workshop and post-mitigation activities in one of these
difficult to mitigate houses follows a discussion of test procedures and data collected in all workshop sites.
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.
-------
BACKGROUND
The New York State Energy Office (NYSEO) has been involved in radon training and education
since 1983. At that time radon was a major component in the indoor air quality unit of the agency's
•Better Builders' workshops. In 1986 NYSEO received an EPA grant to develop the original "Reducing
Radon in Structures' training manual and three-day radon workshop. Then, in 1987, NYSEO was given
a mandate by the New York State Legislature to provide training for radon assessment specialists (those
who perform radon diagnostics and design radon mitigation systems), utilizing Petroleum Overcharge
Restitutionary (POOR) funds. NYSEO was chosen for these tasks because of its extensive experience with
radon and other training programs for builders, developers, heating contractors, architects, engineers and
others.
With POOR monies, the NYSEO completely revised the original radon training manual and slides,
and developed a videotape on radon diagnostics and mitigation. In addition to providing radon-related
instruction, these workshops supplied a vehicle for addressing the perceived connection between tight"
houses and high radon levels.
RADON PROGRAM
The foundation of the NYSEO radon training program is the revised and updated three-day
workshop entitled 'Reducing Indoor Radon'. The training program also includes two one-day workshop
series which focus on radon-resistant new construction and radon and real estate issues, and a two-day
hands-on mitigation field workshop.
The radon field workshops are conducted as follow-up training for attendees of the three-day
workshop. The goal of the field workshops is to reinforce the diagnostic and mitigation information
presented during the classroom instruction of the three-day course. This is accomplished by providing
hands-on training in radon diagnostics and mitigation system design and installation.
NYSEO field workshops are conducted in single family homes whose owners have participated
in an energy conservation program. The training is conducted by personnel with extensive experience
in radon services, including testing, diagnostics and mitigation. NYSEO began offering these training
sessions in January, 1989; in January of 1991 the EPA approved these workshops for the continuing
education requirement of the Radon Contractor Proficiency Program (RCPP).
RADON FIELD WORKSHOPS
PRE-WORKSHOP ACTIVITY
The night before each training session begins a continuous radon monitor (CRM) is set up in the
house to take hourly measurements. The CRM is used to record radon levels immediately prior to
mitigation, monitor students' exposure during the workshop, and measure the effectiveness of the
mitigation system after it has been installed.
DAY ONE
The goal for the students on the first day of this course is to design an appropriate and effective
mitigation system. To this end, the first part of the morning of day one is spent completing a house
investigation summary form. This form is used to guide students through the appropriate steps for a
thorough radon assessment. Students are directed toward identifying potential entry routes, thermal by-
passes, and sources of negative pressure, while paying close attention to house-specific construction
-------
features. After this careful visual inspection of the house, students conduct a homeowner interview in
order to record the testing history of the house, its unobservable construction features and to determine
what the owners' future plans are for the house. This helps to determine the type of mitigation system
selected and, if necessary, the pipe routing and fan location. The remainder of the morning is spent
performing a battery of diagnostic tests to determine the feasibility of certain mitigation approaches.
These tests can include grab samples, sniffer measurements, sub-slab and block wall communication
tests and whole house and basement fan door tests.
The results of the diagnostic survey are discussed by students in a two hour post-lunch meeting.
Students are left alone to develop a detailed mitigation plan which is expected to include a cost estimate
for labor and materials needed to complete the installation. Students are also expected to list appropriate
post-mitigation tests which can be conducted both during and after the workshop. The plans and
specifications are reviewed by instructors. Differences between students' and instructors' plans are
discussed to ensure that students understand all the factors that experienced contractors consider when
designing effective mitigation systems. If time allows and the mitigation strategy warrants it, students
return to the house to begin the installation of the mitigation system.
DAY TWO
The goal of day two is to complete the installation of the selected mitigation system in accordance
with EPA Mitigation Guidelines. When this is finished, tests are conducted to measure the effectiveness
of this system. These post-mitigation tests include sub-slab/block wall communication, and air flow and
radon measurements in the ventilation stack. To complete the installation, all components of the system,
including monitoring devices, are labeled. The operating principles of the mitigation and monitoring
systems are explained to the homeowner, and key components of the installation are identified.
Post-mitigation measurements are used to determine the success of the installed system. A short-
term charcoal canister and a long-term alpha track detector are left with the homeowner to set out 1-2
days after the workshop ends. The CRM is also left on site for a week after the workshop. Before leaving
the site, the class restores the house to its 'as found1 condition.
Along with their letters of attendance, students receive the post-mitigation test results from both
the CRM and charcoal canister. SEO staff record all testing, diagnostic and post-mitigation results, and
write detailed reports for each workshop.
It is expected that students will leave this course with:
a greater understanding of diagnostic procedures
the ability to select an effective, energy efficient mitigation plan using diagnostic test
results
the ability to design a mitigation system which will reduce radon levels and fit in with the
homeowner's present and future lifestyle
specific information on appropriate materials, equipment and installation methods that
assure high quality work
an awareness of procedures available to maintain worker safety
resources for answering questions that arise after the course.
WORKSHOP RESULTS
To date the New York State Energy Office has conducted 26 radon field workshops locations
throughout the State. These training sessions, which are limited to ten students and one representative
-------
from the New York State Department of Health per house, were attended by a total of 249 people. Eighty
of these attendees are Radon Contractor Proficiency Program (RCPP) listed for New York State. This
number represents 64% of the New York State RCPP contractors. Other trainees are providing services
in states throughout the nation, including Maine, New Hampshire, Vermont, Massachusetts, Rhode Island,
Connecticut, New Jersey, Pennsylvania, Ohio, Virginia, Iowa, and California
PRE-MITIGATION MEASUREMENTS
There are a number of pre-mitigation charcoal canister results for twenty-five of the field workshop
houses (one house has only pre-mitigation ATD results). Using the highest basement measurement for
each house, the results range from 20.0 to 522.1 pCi/l, with an average reading of 101 pCi/l. These tests
were conducted by the homeowners, using detectors received through the New York State Department
of Health free detector program. Most of the tests (81 %) were conducted during the heating season
(October to March), with the detectors exposed for three to four days.
Pre-mitigation alpha track results are available for twenty-two of the twenty-six houses used in this
project. Eight of these testing devices were exposed in the basement, ten on the first floor, and four in
an unspecified location. The duration of these tests ranges from two to twelve months, and all but two
results include at least two months' exposure during the heating season. These measurements range
from 4.0 to 222.7 pCi/l, averaging 40 pCi/l.
SITE SELECTION
A number of factors were taken into account when selecting houses for use in this project. In
addition to elevated radon levels and participation in an energy conservation program, selection criteria
included house construction features and the two-day time frame. The first house chosen had poured
foundation walls, floor cracks, a french drain, and an open sump pit. As our experience with workshops
increased, we began to select houses with more features, such as crawlspaces and difficult pipe routes,
that may have needed to be addressed.
Of the twenty-six houses used for this project, fifteen had block wall foundations and three more
had interior block walls. The block tops were sealed in seven of the homes with block wall foundations
and one of the homes with an interior block wall. In the remaining houses, the tops of the walls were
inaccessible. Ten of the project sites had crawlspaces, seven with dirt or stone floors. There were eleven
houses with french drains, and nine with sump holes.
MITIGATION SYSTEMS
Twenty-three of the houses mitigated have active sub-slab depressurization (ASD) systems.
Among these houses, there are six one-point systems, twelve two-point systems, three three-point
systems, and two four-point systems. Active sub-membrane depressurization was used in conjunction
with ASD in two houses having two-point systems and one house with a single point system.
The other three houses involved in this program have two-point active sub-slab pressurization systems.
The average cost of materials only (labor is supplied by students) for these mitigation systems,
including electrical inspection, fire stop(s) and, in one case, a licensed electrician, was $625. The cost
of the systems ranged from $389 to $949.
It should be noted that the nature of this project sometimes dictated a more extensive mitigation
system than may have been necessary to reduce radon levels to or below 4 pCi/l. At times this decision
was made for instructional purposes, since the primary goal of the workshops is to teach by providing
-------
hands-on experience. This sometimes caused in-depth discussions with students about the necessity
of sealing floor cracks or block tops, for example. Another consideration was liability, which led to the
installation of new sump pumps in each house with an existing pump, and numerous self-sealing floor
drains. We also realized that it was less expensive to install what may have been extra materials during
the workshop, rather than returning later.
POST-MITIGATION RESULTS
Extensive post-mitigation testing was conducted in every house used in this project. Continuous
radon monitor measurements are available for each house. These measurements, which were made in
the basement on an hourly basis, from the time the mitigation system fans were turned on and continued
for at least seven days, averaged 1.4 pCi/l. The test results ranged from 0.4 pCi/l (in three houses) to 3.9
pCi/l. The distribution (in pCi/l) was: 0.1 to 1.0 (11 houses), 1.1 to 2.0 (8 houses), 2.1 and 3.0 (4 houses),
and 3.1 to 3.9 (3 houses). Each of the houses with CRM results over 3.0 pCi/l is being studied further.
Post-mitigation charcoal canister results, all from basement readings, are available for 23
workshop sites. Of these results, 15 are below 1 pCi/l, 4 between 1 and 2 pCi/l, and the remaining four
are between 2 and 3 pCi/l. The range of results is from 0.2 to 2.9 pCi/l, with an average of 1.6 pCi/l. In
all but one house the charcoal canister was exposed near the CRM in order to compare the two results.
In the houses for which we have data from both devices, the two measurements usually (81% of the time)
varied by less than one pCi/l.
Currently, we have post-mitigation alpha track detector results for seventeen houses. The
homeowners were instructed to set the ATD's in the same location as their pre-mitigation alpha tracks (if
they had one). The actual location of these long-term testing devices was either in the basement (6
houses), first floor (5 houses) or unspecified (6 houses). The average of these results in houses with
active sub-slab depressurization systems is 0.6 pCi/l, with a range of 0.1 to 1.5 pCi/l. In each of these
houses the ATD results are less than or equal to the short term measurements. We are satisfied that
these houses have been successfully mitigated.
Of the three houses where sub-slab pressurization was used, however, the alpha track results are
acceptable in only one instance (1.5 pCi/l). The short term test results in all three houses were
satisfactory, but the long term results in two of them, 19.0 and 22.2 pCi/l, were not. The remainder of this
paper will discuss the work done in one of these houses.
CASE STUDY
HOUSE DESCRIPTION
Field workshop #13 was conducted in a four year old raised ranch located on a hilltop. The
house has a block wall foundation with a walk-out basement and an attached slab-on-grade garage. The
basement walls, framed out with 2 x 4's, are insulated and paneled. The basement floor is unfinished
concrete, except for vinyl tile in the bathroom. The house has approximately 3,000 square feet of surface
in contact with the soil (1,500 s.f. basement floor, 480 s.f. garage floor, 1,020 s.f. basement walls). The
house has its own well, cistern and septic system. According to the homeowners the well driller went
through 85 feet of gravel before hitting rock.
-------
PRE-MITIGATION MEASUREMENTS
This house was first tested for radon in August, 1988. A charcoal canister in the basement from
the 18th to the 22nd measured 49.4 pCi/l. An alpha track detector left in the living room (first floor) from
August 18 to November 19, 1988 measured 85.5 pCi/l.
On March 7, 1989 NYS Department of Health Radiological Health Specialists visited this house
to conduct on-site testing. Grab samples were taken in the ambient air and at suspected radon entry
points. The working level grab samples in the basement air (0.3 WL) and living room air (0.2 WL)
compare with radon grab samples of 88.2 pCi/l (basement) and 61.2 (first floor) for an equilibrium ratio
of approximately 0.3. The radon concentration at floor cracks in the basement ranged from 119.2 pCi/l
to 208.7 pCi/l, while the sump hole measured 504.9 pCi/l. The well water radon content was 343 pCi/l.
After the homeowner attempted mitigation by sealing the floor cracks and open sump pit, the
house was retested. Charcoal canisters exposed in the basement and living room from January 10-12,
1990, measured 298.3 pCi/l and 164.2 pCi/l. Further testing, conducted March 6-8,1990, measured 165.4
pCi/l in the basement and 93.4 pCi/l in the living room. A water test (3/28/90) found 165 pCi/l in the well
water.
PRE-WORKSHOP ACTIVITY
The night before the workshop started, a continuous radon monitor (Pylon AB-5 with a passive
radon detector) was set up in the basement to print out hourly readings. The windows and doors, which
had been open because of the warm weather, were closed at the start of the monitoring (6:30 PM,
8/15/90). When students and instructors arrived on site at approximately 8:15 the next morning the radon
level, which had climbed steadily through the night, was 13.6 pCi/l. Before students began their radon
assessment of the house the instructors opened the basement windows and set up a fan to blow fresh
air into the basement.
DIAGNOSTICS
Students began the radon assessment process by conducting a visual inspection of the house.
Exhaust appliances they found were an electric dryer vented to the outside and a wind-driven turbine on
the roof. Observed thermal by-passes were limited to kitchen soffits and recessed lighting fixtures.
Students noted that space heating is provided by electrical thermal storage units; the domestic hot water
is also electrically heated.
Potential entry routes discovered by students included utility entrances and exits through the
block walls, open block tops, plumbing penetrations through the slab, and hollow support columns. After
the visual inspection the students conducted a homeowner interview. The homeowner verified the
inspection results, provided photos of the house under construction, explained his future plans to finish
off the basement, and gave the students the radon testing history of his house.
After completing the visual inspection, students took sniffer measurements in the sub-slab soil and
block walls. These measurements were made to locate areas with high radon concentrations. Pilot holes
(1/41 diameter) were drilled in five locations in the basement floor (see floor plan). Sub-slab radon
measurements at FB (514.6 pCi/l), FD (489.6 pCi/l), FE (446.7 pCi/l), and FF (564.5 pCi/l) were similar.
At FC the radon level dropped to 131.2 pCi/l. A cistern is located in the corner of the house where FC
was drilled. It was theorized that the backfill in this area was not compacted, thereby allowing radon easy
access to the surface. Samples in the walls above FC and FA were approximately 15 pCi/l each. Two
jack posts were measured, with one containing 23 pCi/l and the other 5 pCi/l.
-------
The final diagnostic procedure was sub-slab communication testing. Before conducting this test
the students, using a micromanometer, measured the pressure differential between the basement and the
sub-slab soil under normal conditions. With the exception of W2 (+0.001' WC), in the common wall
between the basement and garage, there was no measurable pressure differential at any of the pilot
holes.
Next, a 1 1/4" hole was made at FA so that a vacuum cleaner hose could be inserted. Another
pipe was attached to the vacuum exhaust port and vented outside through a basement window. Then
the vacuum was turned on and pressure differentials were measured at each pilot hole. At FB, 20 feet
from FA, the measurement was -0.007" WC, at FC, 30 feet from FA, it was -0.003" WC, and at FF, 20 feet
from FA, the pressure differential measurement was -0.010" WC. There was no measurable difference at
FD, 55 feet from FA, FE, 40 feet from FA, W1, 30 feet from FA or W2, 3 feet from FA.
At this point the diagnostic portion of the workshop was completed. After extensive discussion
of the diagnostic test results, students and instructors decided that a two-point sub-slab depressurization
system would be used to mitigate the house.
SYSTEM INSTALLATION
Students and instructors returned to the house, bringing in the tools and materials needed to
begin the installation of this system. The students used a core drill to make holes for the suction points
at the mid-point of the long walls (see Floor Plan). The optimum location for the suction points would
have been at the mid-point of the short walls, but the homeowners' future plans for the basement
precluded this option. The holes were located approximately six inches from the floor-wall intersection
to take advantage of the settling which often occurs around footings. Five to ten gallons of sub-slab soil
were removed at each suction point in order to extend the pressure field and reduce the resistance to
airflow.
The students cut risers to length from 4" schedule 20 PVC, attached elbows to the top end of
each, and inserted them in the suction points. The pipe from suction point #1 (see Floor Plan) was run
directly to the garage rim joist in a joist bay. Suction point #2 was connected to this manifold with a tee"
and a 45° elbow. Where the exhaust pipe penetrated the garage wall, the students installed a firestop,
consisting of intumescent wrap and a restricting collar. From the rim joist the pipe was run up the rear
wall, across the garage between the trusses and into the attic. There an in-line centrifugal fan rated at
270 cfm in free air was connected to the pipe with rubber couplings. The pipe above the fan penetrated
the roof near the ridge, was cut off approximately twelve inches above the roof line, and given a rain cap.
Students installed a roof boot to make this penetration weather-tight. They also painted the exposed
portion of the pipe to match the shingles and retard deterioration due to exposure to ultra-violet rays.
The students insulated the pipe from the fan to the rim joist, to reduce noise and condensation.
To monitor the system an electronic pressure gauge was installed on the back wall of the garage. Service
switches were mounted adjacent to the gauge and fan.
Most of the sealing this house needed had already been done by the homeowner. He had
chiseled out and caulked the floor cracks and filled the sump hole with concrete. This concrete shrank
a little, so the students used a grinder to widen the crack, put gun-grade polyurethane in the bottom of
the crack and filled the remainder with pourable polyurethane caulk. The suction point holes were sealed
with mortar and pourable polyurethane around the pipe and the edge of the mortar patch. Students used
expanding foam to seal the hollow support posts.
-------
Before the fan was permanently installed in the garage attic it was temporarily mounted to the end
of the pipe extending from the rim joist. This was done so that the students could check the extension
of the pressure field and determine whether a third suction point would be needed. At FA, 24 feet from
the second suction point, the micromanometer measured -0.041" WC; at FB, approximately 27 feet from
each suction point the reading was -0.062* WC, at FD, 25 feet from the first suction point, the reading was
-0.007' WC; at FE, 25 feet from the first suction point, the pressure differential was measured at -0.010"
WC; and at FF the measurement was -0.066" WC.
POST-MITIGATION TESTING
The pressure measurements seemed to indicate that a third suction point would not be needed,
so the system installation was completed. This involved an additional 25 feet of pipe and four fittings and
moving the fan to the attic of the garage. The communication test was then repeated with the following
(in ' WC) results: FA: -0.037, FB: -0.053, FC: -0.048 (FC was inaccessible during the previous
communication test), FD: -0.006, FE: -0.008, and FF: -0.054. The pressure field developed by the system
did not extend into the walls during any of these tests.
The continuous radon monitor which had been set up before the workshop and was making
hourly readings during it, was left in the basement for one week after the conclusion of the workshop.
The average measurement during that week was 1.52 pCi/l. A charcoal canister exposed in the basement
for four days during the same week measured 0.8 pCi/l.
An alpha track detector placed in the basement at the conclusion of the workshop was due to
be returned for analysis in August, 1991. The homeowner decided to mail the ATD to the laboratory in
March because he wanted winter-only post-mitigation data. This alpha track detector, which was exposed
from 8/24/90 to 3/7/91, measured 22.2 pCi/l. When the homeowner received this information, he called
the Energy Office to report it.
FOLLOW-UP PROCEDURES
Although the homeowners' waiver, which is signed by all participating homeowners, expressly
states that post-mitigation results are not guaranteed, the Energy Office responded to the homeowner's
request for assistance.
The homeowner had conducted a short-term test with a charcoal canister from May 10 to May
15,1991, which measured 0.3 pCi/l. This reinforced the initial assumption that the ATD had been analyzed
incorrectly. To confirm the long-term test results, a continuous radon monitor was installed in the
basement of the house for five days (5/16-21/91). This monitor was set up by an employee of the State
Health Department who works in the area and also happened to be an attendee of the workshop
conducted in this house. The hourly readings during this time frame ranged from 0.1 to 6.0 pCi/l, with
an average of 2.3 pCi/l.
The top graph in Figure 1 shows the post-mitigation CRM measurements in this house when the
sub-slab soil was being depressurized. Line A is the data collected immediately after the system was
installed (August 1990). The data in line B was gathered after the ATD results were returned (May 1991).
On May 21,1991 a contractor and a New York State Energy Office representative, who were the
instructors of the field workshop conducted in this house, visited the house to check the CRM
measurements and assess the operation of the system. After examining the radon measurements made
earlier in the week, we realized that the ATD results could not be dismissed as inaccurate, and began to
think about what options were available. The two alternatives considered were adding a second fan to
-------
the system (in series), and reversing the fan so that it would 'pressurize' the sub-slab soil. Following a
lengthy discussion, we decided to reverse the fan, continue monitoring the radon level in the house, and
then determine whether additional work would be required to bring the radon levels below 4 pCi/l. The
decision to try pressurization was based on several factors, including high soil permeability and the
energy cost of adding a second fan.
The measurements made with the fan reversed, from 5/21 to 6/1/91, averaged 0.9 pCi/l, with
minimal variation (0.4 to 1.4 pCi/l). These results are plotted out on line A of the bottom graph in Figure
1. In order to determine whether these test results were influenced by the weather we reviewed local
data This review showed that the weather patterns during each mode of this test (depressurization and
pressurization) were similar, including the amount of rainfall. In addition, the homeowners were on
vacation, so closed house conditions were maintained for the duration of the testing.
Follow-up CRM measurements, from 1/6-13/92, showed much more variation than the May
measurements (1.5 to 6.0 pCi/l), but they averaged 3.1 pCi/l (see line B in the bottom graph in Figure
1). This average is encouraging, and we believe that the house has been satisfactorily mitigated. Again,
ATD results will be used as the final determinant. Because of a misunderstanding with the homeowner,
this testing did not begin until January 1992, so the results are not currently available.
If the long-term test results are unsatisfactory, the options being considered include adding a third
'drop' (floor penetration), converting the system back to sub-slab depressurization, and stacking fans.
Short-term testing will be conducted after each option is completed to determine its effectiveness. Long-
term testing will follow satisfactory short-term measurements.
-------
Table 1 Field Workshop #13
Pre-mitigation Radon Measurements
Communication Tests
Location
first floor
bftsofnont
firat floor
first floor
WON "WAIST
Date Tested
r 8/88-11/19/88
1/10-12/90
1/10-12/90
3/6-8/90
3/28/90
Method
AID
CC
CC
CC
Liq.Sctnt
Results (pCi/L)
85.5
298.3
1642
93.4
165
Post-mitigation Radon Measurements
Locstion
basemont
paMinont
basement
basement
basement
i Date Tasted
8/17-24/90
8/20-24/90
8/24/90-3/7/91
5/21-6/2791
1/6-13/92
Method
CRM
CC
ATD
CRM
CRM
Results (pCi/L)
1.5
0.8
222
0.9
3.1
Location
FB
TO
FD
Pro-mitigation
dPTW.C.)
(Vacuum)
-0.007
-0.003
+0.000
I1 —
FE | -0.000
..
FF
W1
W2
-0.010
+0.000
+0.001
Foot front
Suction Point
27
48
63
51
25
50
4
•SnHfef*
pCi/L
515
131
490
447
564
15
15
Poet-mitigation
dP fW.C.)
(Fan)
-0.037
-0.053
-0.006
-0.006
-0.054
-0.001
+0.000
4T
• FD
ISink
Suction Point «1
• FE
24'
Field Workshop #13
Drawing Not To Seal*
Suction
Point «2
Windows
(dowdOff)
Storage
FC* fl
W1
Support Poets
FB«
FF«
Garage
-------
FIGURE 1
FIELD WORKSHOP #13 POST-MITIGATION DATA
SUB-SLAB DEPRESSURIZATION (SSD)
HOURS
. 8/17-24/90; immediately after installation; avg: 1.5 pCi/1
5/16-21/91; post-mitigation follow-up; avg: 2.3 pCi/1
5 -
FIELD WORKSHOP #13 POST-MITIGATION DATA
SUB-SLAB PRESSURIZATION (SSP)
HOURS
5/21-6/2/91; immediately after fan reversed; avg: 0.9 pCi/1
1/6-13/92; post-mitigation follow-up; avg: 3.1 pCi/1
-------
Field Workshop Installation Summary
Site System Description
#1 Waterford: 1/89
single point SSD
trench drain
sump hole
floor cracks
utility penetrations
Measurements
Device Location
#2 Olean: 1/89
two point SSP
floor cracks
utility penetrations
#3 Poughkeepsie: 2/89
two point SSD
floor cracks
block tops
utility penetrations
#4 Manlius: 2/89
one point SSD
french drain
floor cracks
utility penetrations
sump hole
#5 Binghamton: 3/89
one point SSD
floor & wall cracks
#6 Painted Post: 3/89
two point SSP
floor cracks
support columns
utility penetrations
Pre-mitigation: 78.0 pCi/l CC basement
Pre-mitigation: 15.9 pCi/l AID 1st floor
Post-mitigation: 0.8 pCi/l CRM basement
Post-mitigation: 0.5 pCi/l CC basement
Post-mitigation: 0.3 pCi/l AID 1st floor
Pre-mitigation: 522.1 pCi/1 CC basement
Pre-mitigation: 12.7 pCi/l AID 1st floor
Post-mitigation: 2.7 pCi/l CRM basement
Post-mitigation: 2.9 pCi/l CC basement
Post-mitigation: 1.5 pCi/l AID 1st floor
Pre-mitigation: 43.0 pCi/l CC basement
Pre-mitigation: 4.0 pCi/l AID 1st floor
Post-mitigation: 0.8 pCi/l CRM basement
Post-mitigation: 1.1pCi/ICC basement
Post-mitigation: 0.8 pCi/l AID unspecified
Pre-mitigation: 143.0 pCi/1 CC basement
Post-mitigation: 0.4 pCi/1 CRM basement
Post-mitigation: 0.7 pCi/1 CC basement
Post-mitigation: 0.2 pCi/1 AID 1st floor
Pre-mitigation: 66.1 pCi/1 CC basement
Pre-mitigation: 13.1 pCi/l ATD unspecified
Post-mitigation: 2.4 pCi/1 CRM basement
Post-mitigation: 2.1 pCi/1 CC basement
Post-mitigation: 1.3 pCi/l ATD 1st floor
Pre-mitigation: 163.0 pCi/l CC basement
Post-mitigation: 3.3 pCi/l CRM 1st floor
Post-mitigation: 1.8 pCi/l CC basement
Post-mitigation: 19.0 pCi/l ATD basement
-------
#7 Voorheesvllle: 6/89
two point SSD
two point SMD
french drain
block tops
utility penetrations
#8 E. Aurora: 6/89
one point SSD
one point BWD
block tops
floor cracks
utility penetrations
#9 Manlius: 6/89
two point SSD
one point SMD
french drain
floor cracks
sump hole
utility penetrations
#10Rexford: 11/89
three point SSD
french drain
floor cracks
floor drain
#11 Goshen: 11/89
two point SSD
perimeter crack
floor & wall cracks
support columns
utility penetrations
Willlamsville: 3/90
three point SSD
block tops
sump hole
plywood deck suction
Pre-mitigation: 58.0 pCi/l CC basement
Pre-mitigation: 4.9 pCi/l AID 1st floor
Post-mitigation: 0.4 pCi/1 CRM basement
Post-mitigation: 0.5 pCi/1 AID unspecified
Pre-mitigation: 196.8 pCi/l CC basement
Pre-mitigation: 222.7 pCi/l AID basement
Post-mitigation: 0.8 pCi/l CRM basement
Post-mitigation: 0.7 pCi/l CC basement
Post-mitigation: 0.5 pCi/l AID unspecified
Pre-mitigation: 70.8 pCi/l CC basement
Pre-mitigation: 66.2 pCi/l AID basement
Post-mitigation: 0.6 pCi/l CRM basement
Post-mitigation: 1.8pCi/ICC basement
Post-mitigation: 0.6 pCi/l ATD 1st floor
Pre-mitigation: 56.9 pCi/l CC basement
Pre-mitigation: 21.2 pCi/l ATD basement
Post-mitigation: 0.6 pCi/l CRM basement
Post-mitigation: 0.7 pCi/l CC basement
Post-mitigation: 0.2 pCi/l ATD unspecified
Pre-mitigation: 37.9 pCi/l CC basement
Pre-mitigation: 38.8 pCi/l ATD basement
Post-mitigation: 1.2 pCi/l CRM basement
Post-mitigation: 1.1pCi/ICC basement
Post-mitigation: 0.4 pCi/l AT 1st floor
Pre-mitigation: 39.4 pCi/l CC basement
Pre-mitigation: 7.5 pCi/l ATD unspecified
Post-mitigation: 2.3 pCi/1 CRM basement
Post-mitigation: 2.3 pCi/l CC basement
Post-mitigation: 1.2 pCi/l ATD basement
-------
#13 Naples: 8/90
two point SSD
floor cracks
support columns
two point SSP
#14 Washingtonville: 8/90
two point SSD
block tops
floor cracks
#15 Syracuse: 10/90
two point SSD
french drain
floor drain
#16 Voorheesville: 10/90
two point SSD
french drain
floor cracks
sump hole
#17Cortland: 12/90
two point SSD
block tops
french drain
floor cracks
Oneonta: 5/91
three point SSD
block tops
support columns
floor cracks
#19 Marilla: 5/91
two point SSD
floor cracks
support columns
sump hole
Pre-mitigation: 298.3 pCi/l CC basement
Pre-mitigation: 85.5 pCi/l AID 1st floor
Post-mitigation: 1.5 pCi/l CRM basement
Post-mitigation: 0.8 pCi/i CC basement
Post-mitigation: 22.2 pCi/l AID basement
Post-mitigation: 0.9 pCi/l CRM basement
Post-mitigation: 3.1 pCi/l CRM basement
Pre-mitigation: 52.5 pCi/l CC basement
Post-mitigation: 1.1 pCi/l CRM basement
Post-mitigation: 0.8 pCi/l CC basement
Post-mitigation: 0.4 pCi/l ATD unspecified
Pre-mitigation: 31.8 pCi/l CC basement
Pre-mitigation: 12.0 pCi/l ATD unspecified
Post-mitigation: < 1.4 pCi/l CRM basement
Post-mitigation: 0.2 pCi/l CC basement
Post-mitigation: 0.2 pCi/l ATD basement
Pre-mitigation: 11.7 pCi/l ATD basement
Post-mitigation: 0.7 pCi/l CRM basement
Post-mitigation: 0.8 pCi/l CC basement
Post-mitigation: 0.3 pCi/l ATD basement
Pre-mitigation: 73.8 pCi/l CC basement
Pre-mitigation: 37.2 pCi/l ATD basement
Post-mitigation: 0.5 pCi/l CRM basement
Post-mitigation: 0.4 pCi/l CC basement
Post-mitigation: 0.1 pCi/l ATD unspecified
Pre-mitigation: 152.4 pCi/l CC basement
Pre-mitigation: 42.8 pCi/l ATD basement
Post-mitigation: 1.3 pCi/l CRM basement
Post-mitigation: 2.9 pCi/l CC basement
Pre-mitigation: 33.3 pCi/l CC basement
Pre-mitigation: 32.3 pCi/l ATD unspecified
Post-mitigation: 2.8 pCi/l CRM basement
Post-mitigation: 0.3 pCi/l CC basement
-------
#20 Fayetteville: 6/91
two point SSD
floor cracks
sump hole
utility entrances
Pre-mitigation: 105.5 pCi/l CC basement
Pre-mitigation: 102.5 pCi/l AID basement
Post-mitigation: 3.3 pCi/l CRM basement
Post-mitigation: 0.9 pCi/l CC basement
NOTE: POST-MITIGATION CRM MEASUREMENTS AT FW 19 & 20 MAY BE INACCURATE BECAUSE
OF EQUIPMENT ERROR.
#21 Mountainville: 6/91
four point SSD
floor cracks
block tops
perimeter crack
#22 Manlius: 7/91
one point SSD
french drain
floor cracks
sump hole
EPDM in crawlspace
#23 Loudonville: 7/91
two point SSD
french drain
floor cracks
sump hole
floor drains
Pre-mitigation: 51.9pCi/ICC basement
Post-mitigation: 1.1 pCi/l CRM basement
Post-mitigation: 0.6 pCi/l CC basement
Pre-mitigation: 20.0 pCi/l CC basement
Pre-mitigation: 24.1 pCi/l AID 1st floor
Post-mitigation: 1.2 pCi/l CRM basement
Post-mitigation: 0.7 pCi/l CC basement
Pre-mitigation: 57.9 pCi/l CC basement
Pre-mitigation: 4.1 pCi/l AID 1 st floor
Post-mitigation: 0.4 pCi/l CRM basement
Post-mitigation: 0.4 pCi/l CC basement
#24 Florida: 7/91
one point SSD
two point SMD (2 crawls)
floor cracks, patching
floor drain
Pre-mitigation: 43.7 pCi/l CC basement
Pre-mitigation: 21.1 pCi/l AID unspecified
Post-mitigation: 0.6 pCi/l CRM basement
#25 Monroe: 12/91
four point SSD
floor cracks
#26 Schenectady: 12/91
two point SSD
french drain
block tops
floor drains(2)
well cover
Pre-mitigation: 43.8 pCi/l CC basement
Pre-mitigation: 8.7 pCi/l ATD 1st floor
Post-mitigation: 1.7 pCi/l CRM basement
Post-mitigation: 1.6pCi/ICC basement
Pre-mitigation: 74.8 pCi/l CC basement
Post-mitigation: 61.0 pCi/l ATD basement
Post-mitigation: 3.9 pCi/l CRM basement
-------
VIIP-5
RADON MITIGATION AT SUPERFUND REMEDIAL ACTION SITES
FIELD EXPERIENCE AND RESULTS
by: Mr. Jean-Claude Dehmel, CHP
S. Cohen & Associates, Inc.
McLean, VA 22101
Mr. Ronald F. Simon
R.F. Simon Company, Inc.
Barto, PA 19504
Mr. Eugene Fisher
U.S. Environmental Protection Agency
Office of Radiation Programs
Washington, D.C. 20460
The Environmental Protection Agency (EPA) has initiated a
radon mitigation project in Montclair, West Orange, and Glen
Ridge, NJ. In these communities, numerous properties have been
contaminated with radium tailings which were initially introduced
around homes as backfill and used as construction materials. This
practice has long since been discontinued, but many residential
properties remain contaminated. These homes have been listed in
the National Priority List under the CERCLA and Superfund acts
and are currently being remediated by the EPA's Region II
Offices. In 1983, the EPA mitigated a number of homes with radon
levels above 4 pCi/L. The mitigation technology relied on
forced-air ventilation systems. The old systems required frequent
repairs, were noisy, exhibited temperature fluctuations causing
excessive humidity and condensation, and many of the systems were
also not successful in reducing ambient radon levels below the
EPA guideline.
In this project, old mitigation systems were removed in 28
homes and replaced with depressurization systems designed to fit
the construction features of each home. The effectiveness of each
mitigation system was determined by conducting field inspections,
pre-operational tests, and by using continuous radon monitors and
alpha track detectors. Test results revealed that all radon
levels were well below the EPA guideline. Reduction in radon
levels ranged from 99% to 45%. Reduction gains were more
significant for those homes that had initially higher radon
concentrations (above 20 pCi/L). For those which had only
marginally elevated levels (above 4 pCi/L), the resulting gains
were significantly lower, ranging from 45 to 83%.
-------
The work described in this paper was funded by the U.S.
Environmental Protection Agency. This paper, however, was
exempted from Agency and administrative review. Therefore, the
contents do not necessarily reflect the views of the Agency and
no official endorsement should be inferred. The work described in
this paper was conducted under a cooperative agreement between
the Office of Radiation Programs and Region II's Emergency and
Remedial Response Division.
PURPOSE AND SCOPE
Old mitigation systems were removed in 28 homes and
replaced with state-of-the-art depressurization systems (1).
Radon diagnostic techniques and mitigation strategies developed
by the EPA in its House Evaluation Program (HEP) were applied in
the planning stages. The new systems were assembled on site and
designed to fit the construction characteristics of each home.
Each assembled system was subjected to a complete assessment of
its performance, including pre- and post-mitigation radon
testing. The monitoring results were used to evaluate and
document the effectiveness of each system in reducing indoor
radon levels. Finally, the old ventilation systems were removed
and appropriate restoration work was conducted under a negotiated
agreement with each homeowner.
In general, diagnostic testing procedures, selection of
remediation system designs, and system installation relied on
techniques and methods which have proven to be successful. The
field work was divided into five major phases:
1) Pre-mitigation activities,
2) Installation of mitigation system,
3) Evaluation of mitigation systems,
4) Mitigation activity closure, and
5) Removal of old ventilation systems and restoration (R&R) of
affected areas.
Because of the presence of soils contaminated with radium,
the work was performed under the umbrella of site specific health
and safety plan. A QA/QC plan was also developed for the purpose
of ensuring that all radon monitoring methods and techniques used
in the course of the project complied with EPA guidelines.
Descriptions of field and project management activities were
described in three documents, a Work Plan, Health and Safety
Plan, and a Quality Assurance Plan (2, 3, 4).
-------
Following the completion of each mitigation and
restoration, a completion report was prepared and submitted to
the EPA Region II and EPA-ORP offices. Each completion report
contains information and documentation regarding the negotiations
held with the homeowner, diagnostic and pre-mitigation test
results, mitigation system design plans and specifications,
information about the installation activities of the new
mitigation system and removal of the old ventilation unit,
summary of all health and safety survey results, and data
regarding the placement of 3- and 12-month alpha track detectors
(ATD).
The project was initiated on April 27, 1990 following the
issuance of the work assignment. The full mobilization of the
field equipment and personnel was completed in early June. All
field work and support activities were also started at the same
time. All project activities were closed on September 30, 1991.
PROJECT ACTIVITIES
PRE-MITIGATION ACTIVITIES
Homeowner Participation
Homeowner participation varied significantly. Some owners
were quite cooperative in signing a release form and starting
the diagnostic and mitigation work. For homes equipped with older
pressurization and ventilation systems, some owners acknowledged
their frustration and inconvenience caused by the need for
frequent service and repairs. System breakdowns often involved
lengthy delays before repairs could be made resulting in elevated
radon levels. For some residents, the resulting elevated radon
levels were of specific concern. Some homeowners were interested
in selling their homes in the near future, and therefore were
looking for documentation that radon was no longer a concern.
Other homeowners were very skeptical and uncooperative and
responded only after persistent inquiries. In a few instances,
participation was obtained only after the intervention of the EPA
Region II's representative. Some homeowners seemed frustrated by
years of surveys, testing, and disruptions, and expressed doubt
that any final resolution to the problem would ever be found.
Others, particularly elderly owners, verbalized disbelief and
confusion as to the nature and extent of the potential health
risks associated with radium contaminated soils (elevated radon
and gamma exposure levels) given that they had lived there for
several decades.
-------
Other homeowners expressed anger and fear given that four homes
remained vacant for so many years only to be recently demolished.
Initially, much effort and time were devoted in obtaining
early homeowner approvals. Less emphasis was devoted to less
cooperative homeowners or to homes with relatively lower radon
concentrations. Homes which had just been mitigated were used as
referrals to hesitant homeowners. Neighboring homes were also
used, with prior approval, as model installations and hesitant
homeowners were invited to inspect finished installations and
talk with their owners. This approach was used to enlist the
participation of the remaining homeowners.
Diagnostic Work
Diagnostics and pre-mitigation measurements were conducted
in all homes prior to the installation of any mitigation system.
Before any work was performed, a signed agreement was obtained
from each homeowner. In some instances, diagnostic work was
initiated based on prior EPA agreements obtained during earlier
site visits. Discussions were held with each homeowner regarding
the extent and types of diagnostic activities and intentions to
replace the old mitigation system with a new one. Discussions
also addressed the significance of earlier radon test results and
those obtained during this round of diagnostic testing. A
walk-through was also conducted to go over system installation,
pipe routing in living areas, system operations, operating costs,
maintenance needs, electrical hook-up, etc.. Initial health and
safety surveys were normally performed in conjunction with all
diagnostic activities.
Pre-Mitiaation Radon Measurements
Radon measurements were made prior to the installation of
any new mitigation system. A continuous radon monitor was placed
in each home and left running for a period of approximately one
week. During such tests, the older radon mitigation system was
rendered inoperable. These measurements were made in the lowest
"livable" area (generally the basement) or the lowest ground
floor, in the case of split-level homes. The tests were started
at the conclusion of the diagnostic activities when it was
possible to schedule the mitigation for the following week. When
scheduling would not permit this type of arrangement,
pre-mitigation tests were started at a later date and scheduled
to end just before the start of the mitigation work.
-------
This approach was used to ensure that pre- and post-mitigation
measurements were performed under similar seasonal conditions. At
the onset of the mitigation work, the old system was re-activated
to reduce ambient radon levels for health and safety purposes,
especially in some homes where radon concentrations were at times
greater than 100 pCi/L.
Pre-mitigation measurements were performed under closed-
house conditions whenever possible, given that some tests were
performed during summer months. Since none of the homes had
central air-conditioning and only a small percentage of homes
possessed room air-conditioners, summer time measurements were
not always performed under true "closed-house" conditions.
Instead, attempts were made to have homeowners maintain "closed
basement" conditions, as best as they could. Only one pre-
mitigation measurement was made with an existing pressurization
and ventilation system left running, since the owner expressed
concern for health reasons about rising radon concentrations
during the conduct of the testing period.
Radon Monitoring Methods
All radon measurements associated with pre-mitigation,
follow-up, and post-mitigation testing were performed utilizing
Femto-Tech continuous radon monitors, Model R210F. The continuous
radon monitors were calibrated by the manufacturer every six
months. A total of ten monitors were deployed during the course
of the project. As monitors were removed from homes (particularly
from those homes with elevated radon levels), these units were
returned to the field office (where radon concentrations
approaching outdoor ambient had been documented) until sufficient
flushing had occurred. Pre-mitigation measurements involved using
both the monitor and its data-logger, which automatically records
hourly readings and running average radon concentrations.
Radon measurement results were routinely scrutinized to
identify the possibility of power interruptions which would
re-set the data-logger. Duplicate measurements were also
performed in several tests and while instruments were placed in
storage at the field office. One monitor failure occurred during
the conduct of a final post-mitigation measurement. This
measurement was later repeated using another monitor. A second
monitor failure occurred while performing an interim evaluation
of an installation. A third failure was discovered while
performing duplicate radon measurements at the field office.
-------
Diagnostic radon measurements were conducted using Pylon
AB5 units. The Pylons were used to characterize radon levels in
walls, under basement slabs, in crawl spaces, in exhaust stacks,
etc. The measurements were performed only for the purpose of
assessing relative rather than absolute radon concentrations. The
instrument were initially calibrated by the manufacturer or
EPA-ORP, and were later re-calibrated at the U.S. Department of
Energy's facility located in New York City. Pylon measurement
results were not used to assess the effectiveness of the
mitigation systems.
Short- and long-term radon measurements were performed
using Radtrak alpha track detectors (ATDs) supplied and read by
Tech/Ops Landauer. ATD measurement results were used to assess
the effectiveness and to document the installation of radon
mitigation systems. A group of ATDs were spiked at the EPA's
NAREL facility and sent to the supplier for processing. In a few
instances, radon measurements were also performed using E-PERMs
for comparative purposes. As before, E-PERM measurement results
were not used to assess the final effectiveness of the mitigation
systems.
MITIGATION ACTIVITIES
System Designs and Descriptions
Radon mitigation systems were designed after reviewing data
collected during diagnostic activities and files from earlier
characterizations. The formal designs and floor plans were then
presented to the homeowners (and the EPA Region II and the ORP
Work Assignment Manager) for review and approval. At times, minor
modifications were made to the original system designs to satisfy
specific homeowners requests. Mitigation system designs and
installations varied very little from house to house. Although
the work plan allowed flexibility in selecting any mitigation
approach, standard sub-slab depressurization systems were
designed and installed in all homes. These systems involved one
or more floor taps depending upon the extent of sub-slab
communication. Areas with elevated gamma radiation levels were
generally avoided when locating floor taps and fan boxes. When
required, crawl spaces were also depressurized either by tapping
through knee-walls or directly down and through crawl space
floors. Table 1 summarizes some of the important system design
features.
-------
TABLE 1. MITIGATION SYSTEMS INSTALLATION VARIABLES
Installation Variables*
Home Fan No. of Amount Finish Crawl Basem't Work
Address LOG. Taps"1" Sealing Basem't Space Access Condition
1 Alan box
2 Alan attic
13 Amelia box
15 Amelia box
18 Amelia attic
145 Carteret box
149 Carteret box
151 Carteret box
153 Carteret box
26 Fremont attic
30 Fremont attic
32 Fremont attic
34 Fremont box
2 James box
6 James box
8 James box
21 Lorraine attic
26 Lorraine attic
53 Nishuane attic
56 Nishuane attic
62 Nishuane box
64 Nishuane attic
66 Nishuane box
26 Virginia box
28 Virginia attic
30 Virginia attic
31 Virginia roof
35 Virginia attic
3
1
2
1
3
4(lw)
2
3 (1cm)
3(2cs)
3(lcm)
2
3(2cs)
3
1
1
2
4(lcs)
3
2(lsog)
3(lw)
3(lcs)
4(lcs)
4(lcm)
3
3(lw)
3
1
3
avg.
less
less
less
avg.
avg.
avg.
less
less
more
avg.
more+
avg.
less
less
avg.
avg.
avg.
less
less
less
avg.
avg.
avg.
less
less
less
more
no
no
no
yes
yes
no
yes
no
no
no
no
no
no
semi
yes
yes
no
no
semi
no
yes
no
no
no
no
no
yes
no
none
none
none
none
large
small
none
small
small
small
small
small
small
none
none
none
small
none
none
none
avg.
avg.
small
none
none
none
none
small
med.
good
poor
good
good
avg.
avg.
avg.
avg.
poor
poor
poor
poor
good
good
good
avg.
avg.
good
avg.
avg.
avg.
avg.
poor
poor
good
poor
poor
avg.
good
poor
good
avg.
poor
avg.
poor
poor
poor
avg.
avg.
avg.
avg.
good
good
avg.
avg.
avg.
poor
avg.
avg.
avg.
avg.
poor
avg.
poor
avg.
* Definition of variables: Fan - where the fan was located;
Taps - total number of locations tapped for each SSD system;
Sealing - approximate amount of sealing that was involved;
Finish - whether basements were finished or not; Crawl -
presence of a crawl space; Access - ease or difficulty in
accessing various basement areas; and Condition - ease or
difficulty in working in basement areas due to the
availability of open spaces.
-------
TABLE 1. MITIGATION SYSTEMS INSTALLATION VARIABLES, CONT'D
+ Key to system tap designations: "lw" - denotes that one of the
taps was a wall tap; "1cm" - denotes that one of the taps was
installed to depressurize a crawl space with a membrane cover;
"Ics" - denotes that one of the taps was installed to
depressurize a crawl space over an existing floor (generally
involved coring through knee-walls); and "Isog" - denotes that
one of the taps was installed to depressurize a slab on grade
area (generally involved coring through knee-walls).
Various types of equipment, supplies, and materials were
used equally in both mitigation and restoration work. Materials
used for most of the 28 installations were fairly consistent with
a few exceptions. Every job required the following items:
o An in-line exhaust fan
o Pressure flow indicator and audio-visual alarms (as negotiated)
o PVC piping, joints, dampers, and fittings
o Sealants, caulks, and PVC cements
o Electrical switches, wiring, and wiring boxes
o Exhaust caps and roof flashing
o Miscellaneous hardware and supplies
o Maintenance and warranty package and operating instructions
Extra materials were at times used when installing fans outside
or in attic areas; including:
o PVC fan box (outdoor installations only)
o PVC down-spouts with fittings (used to direct exhaust above
roof levels for outdoor installations only)
o Sound and thermal insulatory material for pipes and fan boxes
The placement of exhaust fans can be divided into two
groups, conventional installation and outdoor fan box. In 14 out
of 28 homes, installations involved locating the fan in attic
spaces. For one home with an attached garage (there were only two
such homes), the fan was located in the garage attic. In another
home, a roof fan was installed, however. Homes where fans were
located in attic spaces required the construction and
installation of "box-ins" in closets and/or rooms on the first
and second floors for aesthetic reasons. The "box-ins" were
finished to match or approximate their surroundings, since
replacement moldings were sometimes difficult to obtain and
accurately match those typically found in such older homes.
-------
Because of the difficulty in routing a manifold up through
the first and second floors, the remaining installations involved
locating the fan outside near a basement wall. Most of the homes
were of "Dutch Colonial" designs with basements, first and second
floors, and unoccupied attic spaces that did not facilitate the
direct routing and installation of exhaust manifolds. The fans
were encased in insulated PVC boxes for weather protection and
noise control. The top of the fan box was located at grade level
to permit access for maintenance and servicing. Exhaust stacks
were made of PVC down-spouts rising from the fan box and
discharging above the roof line.
Sealing and caulking were performed with each sub-slab
depressurization system installation. Sealing and caulking
enhance the effectiveness of the sub-slab depressurization system
by minimizing basement air losses while maximizing negative
pressures under the slab. The amount of sealing varied from home
to home. A few homes required very little sealing because new
floors had been recently installed or there were only a few
joints, penetrations, cracks, etc.. Conversely, in one home, the
entire basement floor had to be re-finished in order to isolate
and depressurize the sub-slab areas.
System Installations
Labor time and material expenses balanced one another out
when compared to the efforts required to install a fan either
indoors or outdoors. Box-in materials were similar in cost to the
materials used to manufacture the fan box. Labor hours were
approximately the same to build "box-ins" as opposed to building
the fan box and excavating the fan box hole. Typically, health
and safety implications were more significant in excavating fan
box holes, as there always was the risk of finding radium
contaminated soils. Table 1 summarizes some of the most
significant variables found to impact work productivity. For the
28 mitigated homes, installation times ranged between one and
eight days. This effort does not include pre-site preparation
work, follow-up activities, and time spent off-site in support of
installation work or health and safety functions.
Work productivity increased by about 30% after completing
the first few installations, as work procedures and skills were
being fine tuned. Scheduling problems also affected productivity
as some homeowners were confused about making arrangements to
have someone at the home for several consecutive days.
-------
Follow-up visits to support system modifications or touch-up work
also tended to increase the overall level of effort. As mentioned
earlier, the level of effort associated with sealing and caulking
varied significantly among homes.
Health and safety requirements generally reduced work
productivity between 30 and 50 percent. This was particularly
true when coring and excavating floor taps or fan box holes in
areas suspected of containing radium contaminated soils. All work
areas had to be evaluated for each home since radiological
conditions varied from house to house. Homes with contaminated
soils in immediate work areas were of concern since soils could
be re-suspended or entrained in otherwise non-contaminated areas.
In such instances, cautionary measures were implemented to
complete the work as efficiently as was possible while ensuring
that all health and safety requirements were met. In seven homes,
the basement floors were found to be covered with lead shielding
and carpeting. In five other homes, sand bags were used as
shielding to cover exposed soils in crawl spaces. For these
homes, the mitigation phase took much longer since the shielding
had to be first removed to drill floor or wall taps and later
modified to permit its re-installation. Additional time was also
spent conducting radiation surveys to verify the integrity of the
shielding at the completion of each installation.
Lack of access was also found to have a significant impact
on productivity. Basements with doors and windows, however,
allowed easy entry and offered convenient ways with which to
ventilate all work areas. However, other basements with narrow
doors, steep and narrow stairs, low head-room, etc. proved to be
difficult in staging the work and required more elaborate
ventilation schemes. Some basements were virtually empty or
otherwise afforded ample space to store tools and equipment.
On the other hand, some basements were packed with boxes,
clothes, furniture, tools, and other assorted items requiring
frequent moving and re-location. Most basements, however, were
found to be between these two extremes. Similar problems were
encountered in "living-spaces". Homes with crawl spaces usually
required more time depending upon the size of the space and the
work that needed to be performed. Although most basements were
unfinished, a few were clearly finished and were used as
"lived-in" areas, while a few others were "semi-finished". Work
in finished spaces always required extra precautions and,
therefore, more time and effort.
-------
Table 1 summarizes some of the major installation variables which
had an impact on productivity and project resources. It should be
noted that not one single variable can be used as an indicator of
the ease or difficulty encountered during any of the mitigations.
Finally, the level of effort is difficult to break-out by work
phases, since in many situations, mitigation work occurred
concurrently with removal and restoration activities. In such
situations, time expenditures evened out between both types of
activities since some work could equally be defined as mitigation
or restoration.
All mitigation activities were performed by the project
team, with the exception of electrical hook-ups to circuit
breaker panels and utility poles. A licensed electrician was
sub-contracted to perform such tasks and to secure the required
permits. All new sub-slab depressurization systems were wired to
dedicated electrical meters. Finally, in re-flooring the basement
of one home, a plumbing company was sub-contracted to re-hook
sink drains to the waste line.
Radon measurements were made following the installation of
each new mitigation system. Continuous radon monitors were
deployed in each home for a period of approximately two weeks.
During such tests, the older radon ventilation systems were
rendered inoperable. The measurements were made in the lowest
"livable" level (usually the basement, except for split-level
designs) within a relatively short-time after the mitigation work
had been completed, generally within 24 hours after the
depressurization system became operational.
Evaluation of Mitigation System Effectiveness
The effectiveness of each mitigation system installation
was based on evaluating pre- and post-mitigation measurement
results. Short-term measurements were most often utilized to
perform this function, but long-term radon measurements are
better indicators in confirming actual reductions. Post-
mitigation measurements were performed under closed-house
conditions whenever possible. Since none of the homes had central
air-conditioners and only a few homes possessed room
air-conditioners, summer time measurements were for the most part
not performed under true "closed-house" conditions. Instead,
attempts were made to have residents maintain "closed basement"
conditions. Successful mitigations were considered to be those
with measurements that were below the EPA guideline of 4 pCi/L.
-------
Table 2 presents pre- and post-mitigation test results with the
associated reductions. The results reveal that all radon levels
were well below the EPA guideline. Reduction in radon levels
ranged from as high as 99% to as low as 45%. Typically, reduction
gains are more significant for those homes which had initially
higher radon concentrations (above 20 pCi/L). For homes which had
only marginally higher initial radon levels, when compared to the
EPA guideline, the gains were significantly lower, ranging from
45 to 83%.
Removal and Restoration (R&R) Activities
The R&R activities usually started with an interview to
discuss and list those tasks needed to restore the basement to
its original conditions, i.e., prior to the installation of the
old ventilation system. Soliciting the cooperation of these
homeowners into the removal and restoration (R&R) program was
generally much easier since the project team had already
established some credibility during the mitigation work. The EPA
Region II office had also informed homeowners that they could
retain the equipment providing that they assume all energy and
maintenance costs. Few homeowners chose this option when informed
of the typical energy bills associated with the operation of
these older systems. A few others, however, retained basement
window air-conditioning units. A few homeowners, were not pleased
with losing their older ventilation systems despite the
advantages that the new mitigation systems offered. Typically,
these owners used the old units to cool and heat the basement
areas, thereby off-setting heating and cooling costs in the rest
of the house. Fortunately, this situation was an exception as
most homeowners were pleased with the removal of the old systems.
The R&R work was scheduled after such negotiations were
finalized in an agreement. Generally, most R&R activities took
one day followed by a few short return trips to conduct minor
touch-ups, e.g., application of a second coat of paint or stucco
and dry-wall finishing. Some R&R activities, however, required
several days (up to three), when more extensive agreements had
been reached with homeowners. Wall openings left by the removal
of the old systems and duct work were replaced with new windows
and/or filled in. In some instances, new windows and screens were
custom made in order to match existing ones. All windows were
made operable or fixed shut as was specified by the homeowner.
Window fans that were part of the old system were also removed
and the windows were restored.
-------
TABLE 2. PRE- AND POST-MITIGATION RADON MEASUREMENT RESULTS
House
1 Alan
2 Alan
13 Amelia
15 Amelia
18 Amelia
145 Carteret
149 Carteret
151 Carteret
153 Carteret
26 Fremont
3 0 Fremont
3 2 Fremont
34 Fremont
2 James
6 James
8 James
21 Lorraine
26 Lorraine
53 Nishuane
56 Nishuane
62 Nishuane
64 Nishuane
66 Nishuane
26 Virginia
28 Virginia
30 Virginia
31 Virginia
37 Virginia
I.D.I
211
212
112
113
221
312
313
314
315
141
142
143
144
241
242
243
321
322
163
164
165
166
167
173
174
175
176
178
Short-Term*
Pre Post
144.8
2.8
18.7
8.0
7.3
3.0
27.7
25.8
22.6
5.4
30.4
25.3
11.8
32.6
13.7
1.8
17.7
3.1
13.7
25.4
1.1
145.4
4.0
11.1
7.0
12.9
15.3
136.3
1.3
1.0
0.5
2.4
3.6
1.4
1.2
1.8
0.5
0.9
1.2
1.3
2.2
2.1
0.6
0.6
1.3
0.6
0.7
0.5
0.6
2.4
0.7
0.8
2.1
0.5
0.9
2.3
Percent
Reduction
99%
64%
97%
70%
51%
53%
96%
93%
98%
83%
96%
95%
81%
94%
96%
67%
93%
81%
95%
98%
45%
98%
83%
93%
70%
96%
94%
98%
Long-Term Post"*"
ATD E-PERM
0.8
0.3
ND&
ND
ND
ND
ND
ND
0.6
ND
ND
0.8
ND
ND
<0.2
ND
<0.2
<0.2
0.6
<0.2
ND
0.8
ND
ND
1.1
ND
ND
ND
0.7
0.4
-
-
-
-
-
-
-
-
0.7
-
-
<0.4
-
0.4
0.7
3.0
-
-
0.4
-
0.8
All pre- and post-mitigation tests were performed in the
"lowest livable" area (basement) using continuous radon
monitors (Femto Tech R210F); pre- and post-tests were
performed within several weeks of each other and under closed
house conditions as best as possible. All radon results
are expressed in pCi/L. See text for detail.
Long-term test results listed above are for basements or
lower floors under "normal house conditions". E-PERMs were
deployed only in nine homes. See text for detail.
ND signifies "no data" as the ATDs were still deployed at
the completion of field activities.
-------
Only conventional health and safety precautions were implemented
during R&R activities, since no intrusive work was being
performed in radium contaminated soils.
Construction materials typically included concrete blocks,
mortar, windows or glass panes, finish trims, and paint and
caulking compounds. Several homes required the installation of
wooden storm and/or screen windows, which often were custom made
since replacements were difficult to find for such older homes.
In other instances, the overall level of effort was significantly
higher, especially for those homes requiring more extensive
restoration work and those involving the use of sub-contractors.
Finally, discarded equipment and debris were disposed
through the local waste carting company. Refrigerant gases
[chlorofluorocarbons (CFCs)] contained in the old HVAC systems
were drained and collected for proper disposal. Each unit
contained up to five pounds of Freon gas. Arrangements were made
with a specialized firm for the collection and proper disposal of
refrigerant gases. Such efforts were made to abide by the intent
of the recently enacted EPA Clean Air Act Amendments of 1990.
Health And Safety (H&S) Activities Summary
Health and safety monitoring results revealed that by
adopting simple protective measures, personnel exposures were
maintained well below occupational standards and, in some
instances, at the threshold of measurement detection limits. Some
of the applied protective measures included working in well
ventilated areas, judicious uses of local exhaust ventilation at
the source of contaminants, application of dust suppression
techniques, use of the functional sections of mitigation systems
to minimize radon exposures and resuspended particulates while
completing its installation, use of containment methods to
minimize the spread of contaminants, and restricting personnel
traffic out of the work areas. The use of monitoring equipment
was shown to be helpful in detecting trends in ambient radiation
exposure rates and radon levels. Routine surveillance of all work
activities has also allowed the timely detection of potentially
problematic situations.
A review of the survey results revealed that all exposures
were well within occupational radiation protection standards and
OSHA criteria. Survey measurement results varied depending upon
pre-existing conditions and type of mitigation work. Typically,
average radon levels varied from 0.4 to 32.5 pCi/L; radiation
exposure rates ranged from 6 to 460 uR/h; surface contamination
-------
were below the detection limits of 17 dpm/100 cm2; long-lived
radionuclides concentrations were <6.8 x 10~13 uCi/mL; asbestos
fiber concentrations varied from <0.002 to 0.016 fibers/cm3;
total suspended airborne particulate concentrations varied from
<0.01 to 0.65 mg/m3; and organic vapor concentrations ranged from
<0.36 to 13 ppm-TWA for compounds typically found in caulking
compounds and PVC cements.
In support of the field work, several soil samples were
analyzed for the presence of U-238, Ra-226, and Th-232.
Contaminated soils and tailings originated from the extraction
and purification process of uranium ores to produce radium
luminous paints. The analyses were performed by GeLi gamma
spectroscopy. The maximum soil concentrations were noted to be 27
pCi/g for U-238, 123 pCi/g for Ra-226, and 5.9 pCi/g for Th-232.
Background soil concentrations for these nuclides in Northern New
Jersey are typically <1.0 pCi/g (5).
CONCLUSIONS
The project revealed that the installation of radon
mitigation systems in homes with contaminated soils can be
effectively performed, even when the distribution and
concentrations of radium are not well known. The installation of
standard mitigation systems, and, at times, variant designs to
meet specific needs, proved to be successful in achieving radon
reductions as high as 99% and in meeting the EPA guideline of 4
pCi/L for all 28 mitigated homes. When compared to other
traditional mitigation activities, this project involved more
extensive efforts, since it was required to remove old
ventilation systems and restore affected areas to their original
conditions. In some cases, the work was further delayed or
complicated by demands made by the homeowner. In all instances,
negotiated agreements had to be reached with each homeowner. The
experience gained during this project may prove to be useful in
other similar mitigations, since there are still a few hundred
homes with elevated soil radium concentrations and radon levels
in Northern New Jersey. Results from health and safety monitoring
activities revealed that by adopting simple protective measures,
personnel exposures were maintained well below occupational
standards and, in some cases, at the threshold of detection.
Excluding the presence of radium contaminated soils, the health
and safety monitoring methods used in this project could be
applied during the installation of mitigation systems under
conventional conditions.
-------
REFERENCES
1. Environmental Protection Agency, Office of Radiation
Programs, Work Assignment No. 1-39, House Evaluation Program
Applied to Superfund Sites, S. Cohen & Associates, Inc., EPA
Contract No. 68D90170, April 27, 1990, amended October 26,
1990.
2. Work Plan, House Evaluation Program Applied to Superfund
Sites - Montclair, West Orange, and Glen Ridge, New Jersey,
Prepared by S. Cohen & Associates, Inc. for the U.S.
Environmental Protection Agency, Office of Radiation
Programs, Work Assignment No. 1-39, Contract 68D90170, May
1990, revised November 1990.
3. Health and Safety Plan, House Evaluation Program Applied to
Superfund Sites - Montclair, West Orange, and Glen Ridge,
New Jersey, Prepared by S. Cohen & Associates, Inc. for the
U.S. Environmental Protection Agency, Office of Radiation
Programs, Work Assignment No. 1-39, Contract 68D90170, May
1990.
4. Quality Assurance Plan, House Evaluation Program Applied to
Superfund Sites - Montclair, West Orange, and Glen Ridge,
New Jersey, Prepared by S. Cohen & Associates, Inc. for the
U.S. Environmental Protection Agency, Office of Radiation
Programs, Work Assignment No. 1-39, Contract 68D90170, Rev.
1, August 1990.
5. Camp Dresser & McKee, Inc. Supplemental Feasibility Study
for the Montclair/West Orange and Glen Ridge Radium Sites,
Vol. 4, prepared for the U.S. Environmental Protection
Agency, Region II, Edison, NJ, April 3, 1989.
-------
VIIP-6
Dose and Risk Projection for Use of Sub-Slab Radon
Reduction Systems Under Realistic Parameters
Larry Jensen
U.S. Environmental Protection Agency
Chicago, Illinois 60604
Felix Rogers
Centers for Disease Control
At1anta, Georgia
Charles Miller
Centers for Disease Control
Atlanta, Georgia
Structures found to have elevated indoor radon levels that require
reduction are often fitted with sub-slab depressurization systems
that remove the radon soil gas from beneath the foundation before
it can enter the home. Such systems generally discharge the radon
soil gas to the atmosphere, relying upon mixing to reduce
concentrations to levels indistinguishable from that of the ambient
air. However, if a community where a significant fraction of the
residences that exceeded the U.S. Environmental Protection Agency's
radon action level were to install these systems then the
collective radon emission to the atmosphere potentially might be
consequential. To estimate the impact of large-scale installation
of sub-slab depressurization systems a hypothetical scenario was
modeled using realistic housing densities and the consequences,
measured as increased radon concentration in the community and its
surroundings, were projected.
1992 International Symposium on Radon &
Radon Reduction Technology
Minneapolis, Minnesota
September 22-25, 1992
-------
PROBLEM STATEMENT
One standard system to mitigate homes that exceed radon guidelines
is sub-slab depressurization. Piping is inserted through the floor
slab, into a sealed sump or into drainage tile to draw the soil gas
out and vent it to the atmosphere through a chimney (He88). The
standard assumption is that the atmosphere will rapidly mix and
dilute the radon, making it indistinguishable from ambient levels.
This method might work well with one or a few homes, but what if
entire communities installed these systems, thereby introducing a
substantial volume of radon to the atmosphere through large numbers
of stacks? The purpose of this paper is to test the hypothesis
that community-wide utilization of soil gas venting will not lead
to an appreciable increase in the ambient radon levels nor a
significant increase in ambient risk.
HOUSING DENSITY
In order to be realistic, census data giving a breakdown of housing
units in 4 suburban and 3 central city census tracts in a major
midwestern city, along with the area of the tract, were used to
derive single family housing densities (Br92). This density was
used to construct a square, area emission source with a uniform,
rectangular, housing array. Single story housing densities ranged
from about 250 structures per square kilometer in suburban areas to
about 1575 structures per square kilometer in the central city,
with areas from 4 to 1.4 square kilometers, respectively.
ANALYSIS
The Atmospheric Turbulence and Diffusion Laboratory (ATDL) model
(Ha72) was used to calculate the concentration of radon within and
in the area surrounding the hypothetical community. The community
(source) was a 2000 meter square with multiple emission points,
rectangularly arrayed, equal to the housing density. Receptors
were placed at the center of 2000 meter square grids surrounding
the area source. No sources were in the receptor grids.
The transit time of the radon from the soil to the atmosphere is so
rapid that it was assumed that only radon, no decay products, were
emitted from the emission points. Drawdown in the soil was
believed not to occur to any significant degree based upon
discussions with a major mitigator-investigator (Si92)and an EPA
reseracher (He92) and, therefore, was not considered. Radioactive
decay during diffusion from the stack was included. Wind speeds
were set at 2 and 5 meters per second with 2 m/s representing calm
conditions. An actual wind rose from a neighboring community was
used. Calculations were made for each of the seven census tracts
at two wind speeds with the highest and lowest results reported in
this paper.
-------
RADON CONCENTRATION
A normalized value for the radon emission rate was calculated.
Assuming the low end of the range for radon soil gas is about 100
picocuries per liter (pCi/L) (NCRP87) and a common value for fan
volumes is about 200 cubic feet per minute (cfm) (He88), then an
order of magnitude emission rate is about 104 picocurie per second
(pCi/sec) per house. The emission rate range, over background, for
the communities used was from 2.4 to 14.9 picocuries per square
meter per second (pCi/m2-sec). This compares with natural soil
emission rates of 0.0054 to 1.89 pCi/m2-sec (UNSCEAR88).
RESULTS
Ambient concentrations, above background, for the highest and
lowest cases are shown in Figures 1 and 2. With a 5 meter per
second (m/s) wind speed the concentration is about 57 picocuries
per cubic meter (pCi/m3, 0.057 pCi/L) at the community and
diminishes rapidly from the source. Under calm conditions, the
community concentration is about 860 pCi/m3 (0.86 pCi/L). For
other conditions, results can be scaled. For example, with an
elevated soil radon concentration of 10,000 pCi/L (10' pCi/m3, 100
times higher than the base calculation), an effective fan speed of
50 cfm (1/5 the base calculation), and a radon reduction factor of
1/3 (from soil to stack) (Si92) the net result is to multiply the
base result by 6. This could lead to an elevated community
concentration of about 860 (6) = 5160 pCi/m3 (5.16 pCi/L) for a
wind speed of 2 m/s.
CONCLUSIONS
The result of these calculations indicate that, for low range
assumptions, community radon concentrations would be elevated
slightly. In radon problem areas where these systems would most
likely be installed, the resulting levels could be elevated to
higher concentrations based upon local conditions and mitigation
system characteristics. Inclusion of occupancy times, an outdoor
to indoor reduction factor and an equilibrium factor would lead to
a more precise determination of the community impact. Further
modification of the estimation process would be to include all
communities of the city area. This would raise the projected
ambient concentration. Overall, it would appear that installation
of these systems in radon problem areas could reintroduce
sufficient radon to the ambient air so that the utility of the
mitigation system would be offset.
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.
-------
BIBLIOGRAPHY
Br92 Brugman, B.L., 1992, Personnel correspondence
Ha72 Hanna, S.R. , 1972, Description of ATDL Computer Model for
Dispersion from Multiple Sources, National Ocenaic and
Atmospheric Administration, Oak Ridge, TN (April)
He88 Henschel, D.B., 1988, Radon Reduction Techniques for
Detached Houses, Technical Guidance, 2nd ed. , U.S.
Environmental Protection Agency, EPA/625/5-87/019,
Research Triangle Park, NC (January)
He92 Henschel, D.B.; Personnel communication; June 1992.
NCRP87 National Council on Radiation Protection and
Measurements, 1987, Exposure of the Population in the
United States and Canada from Natural Background
Radiation, NCRP Report No. 94, Bethesda, MD (December)
Si92 Simon, R.; Personal communication; June 1992.
UNSCEAR88 United Nations Scientific Committee on the Effects of
Atomic Radiation. Sources, Effects and Risks of Ionizing
Radiation. United Nations, NY; 1988
-------
0.79
0
0.24
0.17
0.17
0
0.20
0
1.08
0.33
0.23
0.24
0.27
0
0.87
1.19
1.91
1.40
0.48
0.65
0.48
0.5
0.68
4.26
60.46
5.83
1.44
1.06
0.4
0.54
0.76
3.25
1.62
1.21
0.89
0
0.43
0.54
0.60
0.70
0.92
0
0.32
0
0.40
0.44
0.52
0
0.67
Figure 1: Lowest case: Radon concentrations (pCi/m3)
surrounding community for a wind speed of 5 m/s. Blocks are
2000 meters, center to center.
-------
15.11
0
4.53
3.17
3.32
0
3.78
0
20.64
6.19
4.33
4.54
5.16
0
16.62
22.70
36.38
26.56
9.10
12.38
9.16
9.52
13.00
81.13
1151.08
110.97
27.45
20.09
7.55
10.32
14.55
61.85
30.93
23.11
16.92
0
8.25
10.32
11.35
13.41
17.54
0
6.04
0
7.55
8.31
9.82
0
12.84
Figure 2: Highest case: Radon concentrations (pCi/m )
surrounding community for a wind speed of 2 m/s (calm
conditions). Blocks are 2000 meters, center to center.
-------
Session VIII Posters
Radon Occurrence in the Natural Environment
-------
VIIIP-1
INFLUENCE OF METEOROLOGICAL FACTORS ON THE RADON
CONCENTRATION IN NORWEGIAN DWELLINGS
Terje Strand and Nils H. Bohmer
National Institute of Radiation Hygiene
P.O.Box 55, N-1345 Osterfie
NORWAY
ABSTRACT
Owing to variations in different meteorological factors, ventilation conditions and
human behaviour, the radon level in indoor air can vary by more than an order of
magnitude over a few days period. In Norwegian houses, the radon concentration is
usually much higher In the winter season, when the temperatures are well below
freezing point, than in the summer. In our etudy, short-term and long-term variations
in the radon concentration, and influencing factors, were investigated. Over a period
of one year an extensive measurement program were undertaken in four typical single
family houses with elevated levels of radon (between 1500 - 4000 Bq/mJ). In one of
these houses, the level in the summer was found to be about twice the level In the
winter and the variation pattern was almost opposite to the other three houses.
Between February 1987 and March 1989. measurements of radon in Indoor air were
made in a total of 7500 randomly selected dwellings from all parts of Norway. These
data were correlated with meteorological data from different parts of the country for
the some period. The results of these studies are reported in this paper.
-------
VIIIP-2
SOIL RADON POTENTIAL MAPPING AND VALIDATION
FOR CENTRAL FLORIDA
by: K. K. Nielson and V. C. Rogers
Rogers and Associates Engineering Corporation
Salt Lake City, Utah 84110-0330
R. B. Brown and W. G. Harris
University of Florida Soil Science Department
Gainesville, Florida 32611
J. K. Otton
U.S. Geological Survey, MS-939
Denver, Colorado 80225
ABSTRACT
Maps of soil radon potentials are being developed to provide a possible geographic
basis for implementing radon-protective building construction standards in Florida. The
maps are being developed from soil properties, independent of institutional boundaries or of
particular present radon limits. The radon potentials are defined from the calculated rates
of radon entry into a hypothetical house modeled over soil profiles for each map polygon. Two
approaches are tested in prototype radon maps developed for Alachua County. The first
approach defines soil profiles and associated transport properties (air permeability, radon
diffusion, moisture) from county soil survey data for 65 map units occurring in 15,000 map
polygons. Radon source strengths are based on 323 radium and radon emanation
measurements, averaged by geologic unit, from archived samples collected at reference pedon
sites in the soil survey. The second approach defines soil profiles and transport properties
from the state-wide Statsgo soil maps and data, which include 30 map units occurring in
approximately 200 polygons in Alachua County. The second approach uses NURE
aeroradiometric data and geologic classifications to define individual radon source strengths
for each map polygon.
The two approaches were compared with ground-truth data that included surface
radon fluxes, soil-gas radon concentrations, and indoor radon data. The mapped soil radon
potentials were related most precisely to the radon fluxes, followed by the soil gas radon
concentrations. Indoor radon data exhibited more variation, as expected, due to house
variability. A mapping precision of ± 1 tier in a four-tier approach appears possible. The
higher resolution of the first mapping approach did not demonstrate any advantage in
mapping precision when comparing the maps with ground truth data. The second approach
is proposed for mapping broader areas of the state because it utilizes data and samples that
are already available throughout most of the state.
-------
This project is funded by the U.S. Environmental Protection Agency (EPA) and the
Florida Department of Community Affairs (DCA). 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.
INTRODUCTION
BACKGROUND
Radon (222Rn) gas is formed by decay of radium (226Ra), which occurs naturally in
virtually all soils at varying concentrations. Radon formed in the top few meters of soil can
enter indoor environments through pores and cracks in building foundations. If sufficient
radon enters a building and if its dilution by outdoor air is small, it can accumulate to levels
that pose significant risks of lung cancer with chronic exposure. Indoor radon levels in the
U.S. average about 1 pCi L"1, with approximately 1 to 3% of homes exceeding 8 pCi I/1 (1).
The U.S. Environmental Protection Agency (EPA) attributes 5,000 to 20,000 lung cancer
deaths per year to indoor radon exposure, and suggests that levels averaging 4 pCi L"1 or
higher warrant remedial action (2).
The Florida Department of Community Affairs (DCA) is developing radon-protective
building standards for new construction that will help reduce public health risks from indoor
radon (3,4). The standards are to be integrated into the state-wide uniform building code,
but may add an incremental cost for constructing new buildings when certain radon-
protective measures are required. To minimize economic burdens and still provide the
intended health protection, requirements for extra-cost radon protective measures can be
based on the potential of the building site for causing elevated indoor radon accumulation.
Although elevated indoor radon occurrences are highly variable, regional trends and
geographic clustering (1,5-6) suggest the possibility of defining geographic criteria for certain
radon-protective construction requirements.
State-wide mapping of radon potentials in Florida has been proposed as one means
of estimating regional needs for radon-protective construction features (7). The radon
potential maps assume that soils are the primary source of indoor radon, a condition that is
generally acknowledged (2). The specific relationships between indoor radon levels and soil
properties are complicated and often unclear, however, despite a sound theoretical basis and
a large body of empirical evidence (8-26).
Soil radon potentials primarily depend on soil radium concentrations, radon emanation
fractions, soil moisture, air permeability, diffusivity, and density. The exact relations to these
parameters are obscured by the frequent variations in indoor pressures and house ventilation
rates. Other, invariant house parameters such as floor properties and foundation
construction details also affect the relation between indoor radon and soil radon potential.
Despite the dependance of indoor radon on static and dynamic house properties as well as
soil properties, it is possible to partition the effects of soil properties for mapping soil radon
potentials.
-------
Numerous radon maps have been compiled previously, as reviewed in a DCA-EPA
radon mapping workshop and related feasibility studies (7,27). The maps have mainly
presented empirical correlations of indoor radon measurements or related parameters with
various institutional units such as state, county, or township boundaries, ZIP Code areas, or
occasionally geologic or physiographic regions. They most commonly present multi-tiered
geographic classifications of areas correlated with indoor radon concentrations. Numerical
radon indices and other, surrogate parameters related to radon potential also have been
mapped, including aeroradiometric gamma activity, uranium mineralization zones, and
surface outcrop areas of geological formations with elevated radon potential. Although these
approaches indicate where elevated radon has been observed, they tend to be indirect or
imprecise predictors of indoor radon for new construction, and they are difficult to relate to
the needs for or results of using radon-protective construction features. Maps aimed at
optimizing testing programs or locating areas of highest observed indoor radon are already
available for Florida (28).
OBJECTIVES
The mapping approach for implementing radon-protective building standards in
Florida differs from previous mapping efforts. The new maps concentrate on radon source
potentials of soils to satisfy the basic objectives for the DCA radon-protective building
standards. The radon mapping objectives include:
• Identify as precisely as possible regions that require radon-protective
building features to attain prescribed indoor radon concentrations.
• Avoid political and institutional boundaries that are unrelated to radon
potential.
• Avoid restrictive association to a particular radon standard (i.e., 4 pCiL ).
• Minimize uncertainties related to variations in time, house design, and
occupancy.
To attain these objectives, an approach was developed to separate soil radon source
parameters from other parameters affecting indoor radon concentrations. A numerical radon
transport and entry model was used to compute radon source potentials for soil profales
occurring throughout the regions to be mapped. The soil profiles are defined from Soil
Conservation Service (SCS) soil survey data. Maps were developed by attributing the
calculated radon potentials to the areas represented by each soil profile. An initial, detailed
mapping approach was compared with an alternative approach that provided more complete
coverage throughout Florida. Field measurements of radon fluxes and soil gas radon
concentrations were conducted to evaluate the mapping approaches. Indoor radon data from
a prior land-based survey (28) and a University of Florida survey (29) also were compared
with the maps of calculated soil radon potentials. This paper presents an interim report on
the prototype mapping effort, which presently is in progress.
-------
THEORY
PARTITIONING OF RADON SOURCE PARAMETERS
Indoor radon distributions are subject to variations in source parameters, to variations
in house (and occupancy) parameters, and to time variations of these parameters. Although
indoor radon concentrations are the ultimate concern in this mapping program, it is their
geographic variability, independent of house and time variations, that is of particular
interest. The large uncertainties normally associated with populations of indoor radon
measurements therefore can be reduced by considering only the geographic distribution of
radon source parameters. Time variations are eliminated by using only invariant or long-
term average parameters in developing the maps. House variations similarly are eliminated
by using average or typical parameters to represent a constant, reference house for all
geographic locations. The reference house, a slab-on-grade single-family dwelling, is defined
and modeled for radon entry calculations using soil profile properties at each source location.
The source variations throughout the county or state thus are characterized from calculated
radon entry rates, independent of actual house variations. Once the distribution of radon
source potentials is determined, the additional variations due to house variability also can
be assessed by statistical comparisons with indoor radon data. Actual house parameters are
only required to relate the potentials to particular indoor radon concentrations. This study
addresses only the radon source characterization.
RADON ALGORITHM
The algorithm to compute radon entry into the reference house on each soil profile is
developed using the unified-theory representation of multi-region, multi-phase radon
generation and transport by both advection and diffusion (30). The algorithm is implemented
in 2-dimensional numerical calculations by the RAETRAD model (RAdon Emanation and
TRAnsport into Dwellings, 31). The steady-state radon balance equation solved by
RAETRAD for the map calculations is :
V.faDV(Cb/fs) - V.(K/uX(yfs) VP - XCb + RpXE = 0 (1)
where
V = gradient operator
fa = p(l-S+SkH)
p = soil porosity (dimensionless: cm pore space per cm bulk space)
S = soil water saturation fraction (dimensionless)
kH = 222Rn distribution coefficient (water/air) from Henry's Law
(dimensionless)
D = diffusion coefficient for 222Rn in soil pores (cm2 s"1)
Cb = fsCa = 222Rn concentration in bulk soil space (pCi cm"3)
C = 222Rn concentration in air-filled pore space (pCi cm"3)
-------
f. = p(l-S+SkH)+pka
p = soil bulk density (g cm"J, dry basis)
ka = k^extf-bS) .
k ° = dry-surface adsorption coefficient for Rn (cm g )
b* = adsorption-moisture correlation constant (g cm )
K = bulk soil air permeability (cm )
ji = dynamic viscosity of air (Pa s)
VP = air pressure gradient (Pa cm" )
X = 222Rn decay constant (2.1xlO"6 s"1)
R = soil 226Ra concentration (pCi g"1)
E = total 222Rn emanation coefficient (air + water) (dimensionless).
This equation applies to gas-phase advective transport of radon, and to combined gas-
phase and liquid-phase diffusive transport of radon. The combined-phase diffusive transport
is characterized by appropriate moisture- and porosity-dependent values of the pore-average
diffusion coefficient, D (32,33). This approach is important to correctly characterize radon
diffusion in unsaturated soil pores that may have small intermittent water blockages, but
that still may transmit significant radon flux (32,34). Liquid-phase advective transport of
radon is not addressed because it typically is negligible. The radon fluxes between different
soil layers and at the top surface are calculated as
F = -D fa VCa + (K/u) VP Ca (2)
where
F = bulk flux of 222Rn (pCi cm"2 s"1).
A 2-dimensional form of equation (1) is used in modeling the reference house with
elhptical-cylindrical symmetry over the specific soil profiles that are being mapped (Figure
1) Two-dimensional modeling has been found previously to adequately represent houses for
calculating indoor radon entry. Soil horizons defined by county soil survey date are used to
define the vertical soil profiles. Radial uniformity of the soils is assumed. The house is
modeled to have a foundation crack near its perimeter for permitting advective transport of
radon by pressure-driven flow. It also permits radon transport through the foundation slab;
however this transport is dominated almost completely by diffusion because of the low air
permeability of concrete. To approximate rectangular house geometry, a skewing factor is
applied to the radial gradient term that results from the 2-dimensional gradient operator in
Equation (1).
-------
Concrete
Floor
Pressure-Driven
Air Flows;
Advective Radon
Transport
Radon Gas Diffusion
5 m
House Center Line
(symmetry assumed)
RAE-103688
Figure 1. Cylindrically-symmetric house and foundation soil
profiles used to model radon entry rates.
RADON MAPPING METHODS
MAPPING APPROACHES
Two radon mapping approaches (A and B) were devised and tested for the Florida
radon mapping program. Approach A was developed first (35,36), using a digital, high-
resolution (1:24,000) county soil survey map, intersected with geology map units, to define
individual radon map polygons. A total of 65 map units were defined to occur in
approximately 15,000 polygons in the Alachua County map. Radon source potentials were
calculated for the area represented by each polygon with the numerical model using existing
soil physical and hydrologic data to define the radon transport properties. New
measurements of soil radium concentrations and radon emanation coefficients, averaged by
geologic unit, were used to define radon source strengths of the soils.
Approach B was developed afterward to utilize lower-resolution digital Statsgo soil
maps (37) that are available with broader coverage over nearly all of Florida. Radon map
polygons for Approach B were defined from the intersection of the Statsgo polygons with
geology map polygons. Thirty map units were defined to occur in approximately 200 polygons
for the Alachua County map. Radon source potentials for Approach B used NURE
aeroradiometric data (equivalent uranium), averaged over each radon map polygon, to
estimate average (equilibrium) radium concentrations. Radon emanation coefficients were
estimated as a function of radium concentration from the new measurements.
-------
For both approaches, radon source potentials were calculated as the rate of radon
entry into the hypothetical reference house that was modeled on the soil profiles of each map
unit. The radon entry rates were expressed in long-term units (mCi y" ) to emphasize their
long-term average nature. The radon potentials then were grouped into four tiers of similar
numerical values for color-coded display of the soil radon potentials associated with each map
polygon. The soil radon potentials can be related to indoor radon concentrations for the
reference house using a simple ventilation-rate model,
C = 114Q/(VhXh) (3)
where
C = average indoor radon concentration (pCi L" )
114 = unit conversion (pCi L"1 h"1 per mCi m^ y"1)
Q = average radon source potential (mCi y" )
Vh = house volume (m3)
Xh = house ventilation rate (h" ).
The reference house used to represent Florida housing with Approach A was described
previously (35,36). The reference house used with Approach B was nearly identical, and
consisted of a 28 ft. x 54 ft. rectangular slab-on-grade house with nominal dimensions and
characteristics as summarized in Table 1. Its volume was based on that of a median U.S.
family dwelling (38), and is similar to that of typical Florida houses (39). A nominal 8 ft
(2.4m) ceiling height was used to estimate its area, which also is similar to other estimates
of Florida floor slab areas (39). Its ventilation rate corresponds to the nominal median U.S.
house ventilation rate (38), although even lower values sometimes have been used to
represent Florida houses (39). The floor crack location is chosen near the slab perimeter to
approximate a slab/footing shrinkage crack. The stem wall footing is assumed to be 3 ft (91
cm) deep, penetrating 2 ft (61 cm) into the natural terrain. A 1 ft (30 cm) layer of fill soil
beneath the floor slab is comprised of material identical to the surrounding surface soil. The
indoor pressure is typical of that used previously to generically model indoor thermal and
wind-induced pressures in U.S. houses (40), and to represent Florida housing in particular
(39). Concrete slab permeabilities and diffusion coefficients were estimated from data
measured on Florida floor slabs (41).
TABLE 1. NOMINAL VALUES OF PARAMETERS USED IN RADON ENTRY
CALCULATIONS FOR THE REFERENCE HOUSE
House Area 143 m2 Fill Soil Thickness 30 cm
House Length/Width 1.9 (ratio) Indoor Pressure -2.4 Pa
House Volume 350 m3 Concrete Slab Thickness 10 cm
House Ventilation Rate 0.5 h-1 Concrete Slab Porosity 0.22
Floor Crack Width 0.5cm Concrete Slab 226Ra*Emanation 0.07 pCi g
Floor Crack Location slab perimeter Exterior Footing Depth 61 cm ^ ^
Crack Area Fraction 0.002 Concrete Air Permeability 1x10' cm
Concrete Rn Diffusion Coeff. 8x10"* cnT s"
-------
SOIL PHYSICAL PARAMETERS
Layers of the soil profile for each map unit were defined for modeling throughout the
top 5 m. Soils in the top 2.0 to 2.5 m were characterized explicitly from SCS soil survey data,
which typically represented the various occurrences of the A, E, B, and C soil horizons and
their subdivisions in about six layers. An additional layer was defined beneath the SCS-
characterized layers to represent the remaining deeper soils to a depth of 5 m. This layer
was defined either as an extension of the lowest layer from the SCS-characterized horizons
or as a layer of the Hawthorne formation. The selection of these deeper layers was made
according to a geologic map of surface occurrences of the Hawthorne Formation in Alachua
County. Horizontal uniformity was assumed in the distributions of the radon source and
transport parameters in the vicinity of the reference house.
The parameters used to define each soil layer included the soil density, porosity, water
content, radon diffusion coefficient, air permeability, radium concentration, and radon
emanation coefficient. Most of the physical parameters were defined or derived from data
in the Alachua County soil survey report (42) or from more detailed data files maintained by
the University of Florida Soil Science Department. Soil densities (bulk, dry-weight basis)
were taken directly from the soil survey report for each horizon in each soil map unit. Soil
porosities were calculated from soil density and specific gravity as
P = 1 - P/Pg (4)
where
o
pg = soil specific gravity (nominally 2.7 g cm ).
Soil water contents were estimated from soil water drainage data in the Alachua
County soil survey (42). For Approach A, the water contents were estimated as drained,
field-capacity values corresponding to -0.1 bar matric potential for sands, and to -0.33 bar
matric potential for clays (36). Later theoretical water balance calculations with the
FEMWATER code (43,44) indicated that sub-slab matric potentials are well-approximated by
the position above the water table for wet climates and shallow water tables as in most of
Florida. Therefore the water contents for Approach B were interpolated from the soil survey
water drainage data at matric potentials corresponding to the height above the water table.
Water table depths were computed as time-weighted averages from the high water table
limits and durations reported in the soil survey (42), and deeper limits for other seasons as
documented with Approach A (36). Soil water contents in volume percent and weight percent
units in the soil survey report were converted to fractions of saturation as:
100 S = My/p = pMw/p (5)
where
= soil water content (volume percent)
Mw = soil water content (dry weight percent).
-------
Soil radon diffusion coefficients were estimated from the water contents and porosities
of the soils using a predictive correlation that is based on 1073 laboratory measurements of
radon diffusion in recompacted soils at moistures ranging from dryness to saturation (33).
The soil textures ranged from sandy gravels to fine clays, and their densities covered the
range of most of the Florida soil densities. The correlation exhibited a geometric standard
deviation (GSD) between measured and calculated values of 2.0, and had the form
D = Dj) exp(-6Sp - 6S14p) (6)
where
D = diffusion coefficient for 222Rn in soil pores (cm2 s'1)
D0 = diffusion coefficient for 222Rn in air (l.lxlO'1 cm2 s'1).
Soil air permeabilities were estimated similarly from the water contents, porosities,
and grain diameters of the soils using a predictive correlation that was based on more than
a hundred in-situ field measurements of soil air permeability, including measurements in
Florida (33). This correlation exhibited a GSD between measured and calculated values of
2.3, and had the form
K = 104 (p/500)2 d4* exp(-12S4) (7)
where
K = bulk soil air permeability (cm )
d = arithmetic mean soil particle diameter, excluding >#4 mesh (m).
SOIL RADIOLOGICAL PARAMETERS
Soil radium concentrations and radon emanation coefficients used with Approach A
were obtained from new 226Ra measurements on 323 samples and radon emanation
measurements on 131 samples. Most of the emanation measurements were made on samples
with 226Ra concentrations exceeding 1 pCi g"1, since these provided the best measurement
precision. More than 280 of the soil samples were obtained from an archive of samples
collected at the reference pedon sites by SCS during the original Alachua County soil survey.
The remaining samples were obtained from new borings in May 1991 by the U.S. Geological
Survey and and Florida Geological Survey at selected supplementary sites in Alachua
County. Individual results of these analyses were reported previously (36). Radium
concentrations exhibited a geometric mean of 0.8 pCi g"1 and a geometric standard deviation
of 3.3. Radon emanation coefficients had an arithmetic mean and standard deviation of 0.48
± 0.16.
Soil radium concentrations in Approach B were estimated for the top 2.5 m layer from
the mean equivalent uranium concentration as averaged for each Statsgo map unit from
NURE aeroradiometric data (45). Equilibrium was assumed between the uranium and
radium activities. A weak correlation observed between the measured radium concentrations
-------
and emanation coefficients (Figure 2) was used to estimate emanation coefficients from the
radium concentrations for Approach B. The relationship from Figure 2 was approximated
as
E = min(0.15R + 0.20, 0.55), R<8 pCi g'1
E = 0.50, R>8 pCi g -1
(8)
Radon Emanation Coefficient
D p p p p -
D M ** b> b» c
(
* 4
•
7
•
i
:
& i
>m
*:
B
i
4 ^
% '
• 1
•
1
m •
i i
•
•
<
•
• i
•
•
..
•
•:
0123456 10 20 30
Radium-226 Concentration (pCi/g)
40
RAE-103638
Figure 2. Relationship between measured radon emanation
coefficients and radium concentrations.
VALIDATION MEASUREMENTS AND RESULTS
Mapped soil radon potentials were compared with two sets of empirical ground-truth
measurements and two sets of indoor radon measurements to estimate the degree of success
in mapping particular land areas in Alachua County and to compare mapping Approaches
A and B. The comparisons were made by locating each measurement location on the
Approach A or Approach B maps from their latitude/longitude coordinates, and then
associating the color-tier of the map location with the numerical measurement. Each of the
four map color tiers was assigned to an integer (l=blue, 2=green, 3=orange, 4=red), which
was used to estimate mapping precision. The measurement data and map tiers from each
set were then sorted by measurement, and the number of measurements in each map tier
was assigned successively to the sorted data set. The resulting tier classification errors then
were computed by subtracting for each measurement the sorted tier assignment from the
actual tier definition. This approach defined the maximum tier classification precision for
each data set. The results of these analyses were plotted as frequency histograms, and the
standard deviation of each set of tier classification errors was computed to estimate overall
precision of the approach.
-------
To account for potential imprecision in map polygon line positions, a 500 meter fringe
was imposed in ranked order on the polygons of alternative versions of the Approach A and
Approach B maps. The fringe was applied first to extend the size of the polygons of greatest
radon potential (red), followed next by orange, and then green. The fringe could expand the
size of a polygon only into an area occupied by a lower-order color. In this way, points of
elevated potential were included in a higher tier if they were within 500 m of its boundary.
The first ground-truth data set was collected by USGS personnel in May 1991, and
consisted of 66 soil gas radon measurements at 1 m depth at different locations throughout
Alachua County. Additional measurements also were collected at other depths, but only the
1-m measurements were analyzed here to assure consistent comparisons among different
locations. The detailed methods and results of this study are reported separately (46).
The second ground-truth data set was collected in April 1992 by Rogers & Associates
personnel, and consisted of 81 surface radon flux measurements. The measurements were
made in triplicate at 10 m spacing at each of 27 sites throughout Alachua County. Sampling
was conducted over a 24h period with small charcoal canister radon samplers, and analyses
were conducted on the sealed, retrieved samplers by spectrometric gamma assays. The
procedure for these measurements and a comparison of the procedure with EPA method 115
(47) has been reported previously (48).
The first set of indoor radon measurements used for comparisons with the radon
mapping tiers was taken from the state-wide land-based radon survey conducted in Florida
in 1986 (28). The radon measurements utilized charcoal canister radon samplers, and were
located in 102 areas throughout Alachua County. The second set of indoor radon
measurements was conducted by University of Florida (29), and utilized alpha track radon
detectors in 35 homes throughout Alachua County.
The precisions of the tier classifications for mapping Approaches A and B are
illustrated in Figure 3 for the 81 individual radon flux measurements. Conservative
(positive) errors resulted from including low radon measurements in a high radon-potential
tier, and non-conservative (negative) errors resulted from including high radon measurements
in a low radon-potential tier. As expected, the 500 m fringe made the classifications slightly
more conservative for both mapping approaches. Although Approach B exhibited more data
points with correct classifications, it also had slightly more non-conservative errors at the -2
level Numerical summaries of the data in Figure 3 (means and standard deviations of each
distribution) are presented in Table 2. They indicate that Approaches A and B have similar
precisions for the basic map definitions, but that the 500 m fringe improves the precision for
Approach B and degrades it for Approach A.
-------
TABLE 2. MEANS AND STANDARD DEVIATIONS OF MAP TIER
CLASSD7ICATION ERROR DISTRD3UTIONS
a
b
c
d
Data Set
RnFluxa
RnFluxa
Rn Flux meansa
Rn Flux means8
Soil Gas Radonb
Indoor Radonc
Indoor Radond
Indoor Radon
This study.
Otton et al., (46).
Nagda et al., (28).
Roessler, (29).
Approach
A
B
A
B
B
B
A
B
n
81
81
27
27
66
102
35
35
No Fringe
0.00 ± 1.21
0.00 ± 1.22
0.00 ± 1.11
0.00 ± 1.11
0.00 ± 1.37
0.00 ± 1.22
0.00 ± 1.39
0.00 ± 1.73
Fringe
0.59 ± 1.43
0.07 ± 1.16
0.59 ± 1.42
0.07 ± 1.03
0.12 ± 1.23
0.29 ± 1.27
—
~~~
00
to
I 50
^ 45
« 40
(0
| 35
x 30
u.
c 25
o
"8 20
a:
"o 15
S 10
1 s
Z n
n s 81 measurements
Non-Conservative
m
j
P
l_ '•
fc :
Hi!
/
" ^
/
^
s
j
r*
$ C3 A, No Fringe
\ 0 A, 500m Fringe
^ • B, No Fringe
\ 0 B.SOOm
:ringe
# Conservative
^ _— i- •s^
s
t
f
s
/ R^_
? I/
/ ', #
t ',
f : S
/ '.
/ '. S
1
^^
jp
^
1
/
,/
P'
Number of Radon Flux Measurements
3 O1 O Ul O
n = 27 measurement
locations
|: /
Non-Conservative ;j
'', ED A, No Fringe
0 A, 500m Fringe
• B, No Fringe
E2 B, 500m Fringe
Conservati
ve
/
/
(
-
-3-2-10123
Tier Classification Error
RAE- 1040W
Figure 3. Map tier precisions from
comparisons with individual
radon flux measurements.
-3-2-101 2
Tier Classification Error
3
RAE- 1040U
Figure 4. Map tier precisions from
comparisons with means of
radon flux measurements.
-------
Identical analyses were performed for the means of the radon flux data at each of the
27 sites to test the effects of local heterogeneity and the possible benefits of replicate
sampling. Figure 4 summarizes the analyses with averaged radon flux data, which exhibit
slightly better precision in all cases. The numerical summaries in Table 2 indicate standard
deviations of 1.11 compared to about 1.21 for the individual measurements. Including the
500 m fringe for the averaged data made little difference for the Approach A map, but
reduced the standard deviation for averaged data with Approach B to 1.03 compared to 1.16
for the individual measurements.
The precision of the tier classifications for the USGS soil gas radon data is
summarized in Figure 5 for Approach B. The 500 m fringe again caused a slight conservative
shift in the distribution. The numerical summary in Table 2 suggests larger tier errors for
the soil gas radon measurements than for the radon flux measurements. This is consistent
with theoretical calculations, which suggest that radon fluxes provide a more precise
indicator of soil radon potential (as computed here with the RAETRAD model) than soil gas
radon concentrations.
The precisions of tier classifications for the Geomet indoor radon data with Approach
B are summarized in Figure 6. As expected, they exhibit greater tier errors due to the
contributions of house variability, along with variations in soil radon potential. The 500 m
fringe again causes a slightly conservative shift; however the numerical summaries in Table
2 indicate that the fringe slightly reduces the mapping precision.
Ł
3
(0
CO
-8
(0
QC
(0
m
O
"5
CO
n - 66 measurements
afl m depth
c
-------
The precisions of the tier classifications for the University of Florida indoor radon data
with Approaches A and B are summarized in Figure 7. They also exhibit greater tier errors
then for radon fluxes and soil gas radon due to the contributions of house variability. The
numerical summaries of these indoor radon data (Table 2) indicate greater variation than for
the Geomet data set, particularly for Approach B. This may result in part from the smaller
number of houses in this set or from location differences between the two sets.
-3-2-10 1 2 3
Tier Classification Error _._ ,„„.
Figure 7. Map tier precisions from comparisons with University of
Florida indoor radon measurements.
DISCUSSION
The present analyses suggest that the lower-resolution Approach B for mapping soil
radon potentials does not significantly degrade mapping precision compared to Approach A.
This is probably due in part to the improved method for defining soil radium concentrations
from area-specific aeroradiometric data instead of relying on soil classifications to define
radiometric similarities among soils throughout the county. It is still recognized that
localized anomalies probably occur within many map polygons, and that these will be largely
undetectable using the NURE aeroradiometric data (6-mile flight-line spacing, (45)). Based
on the limited ground-truth comparisons made to date, however, mapping Approach B may
provide maps with a tier classification uncertainty of approximately one tier if a four-tier
classification system is used.
The development of soil radon potential maps from soil (Statsgo) data, Geologic maps,
NURE radiometric data, and soil radon emanation measurements is presently in progress for
twelve counties in central Florida. The maps will utilize Approach B for consistent coverage
of parts of the state where digitized high-resolution soil maps are unavailable. The maps will
be displayed as soil-defined polygons with a colored-tier system to illustrate general areas of
high and low soil radon potential. For more complete interpretation, numerical values of the
-------
calculated soil radon potentials will be displayed or provided for each map polygon. These
can be used to estimate indoor radon levels for the reference house, or for other specific house
conditions if more detailed house parameters are available. For the reference house, the
indoor radon concentration is estimated from the simple ventilation-rate model in equation
(3), which gives a radon concentration of 4 pCi L"1 for a soil radon potential of about 6 mCi
y*1. Statistical estimates of variability within each polygon will be available from the
standard deviations of the NURE-based radium estimates. These may potentially provide
a basis for estimating confidence intervals on the radon potentials for the map polygons.
REFERENCES
1. Nero, A. V., Schwehr, M.B.,Nazaroff,W.W.,andRevzan, K.L. Distribution of Airborne
Radon-222 Concentrations in U.S. Homes. Science 234:992-997; 1986.
2. EPA. A Citizen's Guide to Radon. Washington D.C.: U.S. Environmental Protection
Agency and U.S. Department of Health and Human Services; report OPA-86-004;
1986.
3. Sanchez, D.C., Dixon, R, and Williamson, A.D. The Florida Radon Research Program:
Systematic Development of a Basis for Statewide Standards, proceedings of the 1990
International Symposium on Radon and Radon Reduction Technology. Atlanta, GA:
February 1990.
4. SBCCI. Florida Code for Radon-Resistant Construction and Mitigation," Southern
Building Code Congress International, Inc., Rev. 1, January 1990.
5. Cohen, B.L. A National Survey of 222Rn in U.S. Homes and Correlating Factors.
Health Physics. 51:175-183; 1986.
6. Peake, R.T., Gundersen, L.C.S., Ronca-Batista, M., James, J., and Schumann, R.R.
Determination of Radon Geologic Provinces in the United States, proceedings of the
1990 International Symposium on Radon and Radon Reduction Technology. Atlanta,
GA: February 1990.
7. Nielson, K.K. and Rogers, V.C. Proceedings of the Florida Workshop on Radon
Potential Mapping. Salt Lake City, UT: Rogers & Associates Engineering Corp.;
report RAE-8945-4; 1990.
8. Brookins, D.G. Indoor and Soil Rn Measurements in the Albuquerque, New Mexico
Area. Health Physics 51:529-533; 1986.
9. Peake, R.T. and Hess, C.T. Radon and Geology: Some Observations," in Trace
Substances in Environmental Health XXI, D. D. Hemphill, ed., pp. 186-194, Columbia:
Univ. of Missouri, 1987.
-------
10. Sextro, R.G., Moed, B.A., Nazaroff, W.W., Revzan, K.L., and Nero, A.V. Investigations
of Soil as a Source of Indoor Radon. In: Radon and Its Decay Products. P. K. Hopke
ed., Washington, DC: Am. Chem. Soc., ACS Symposium Series 331, pp. 10-29, 1967
11. Gundersen, L.C.S., Reimer, G.M., and Agard, S.S. Correlation Between Geology,
Radon in Soil Gas, and Indoor Radon in the Reading Prong, in Geologic Causes of
Natural Radionuclide Anomalies. M. A. Marikos and R. H. Hansman, eds., Missouri
Dept. of Nat. Res. Spec. Publ. 4, 91-102, 1988.
12. Gundersen, L.C.S., Reimer, G.M., Wiggs, C.R., and Rice, C.A. Radon Potential of
Rocks and Soils in Montgomery County, Maryland. U.S. Geological Survey map MF-
2043; 1988.
13. Nazaroff, W.W., and Nero, A.V. Radon and Its Decay Products in Indoor Air.
Chapters 1 and 12; New York: Wiley & Sons; 1988.
14. Nazaroff, W.W., Moed, B.A., and Sextro, R.G. Soil as a Source of Indoor Radon;
Generation, Migration and Entry. In: Radon and Its Decay Products in Indoor Air,
W.W. Nazaroff and A.V. Nero, eds., New York: Wiley & Sons, p. 60-65; 1988.
15. Otton, J.K, Schumann, R.R., Owen, D.E., Thurman, N. and Duval, J.S. Radon
Potential of Rocks and Soils in Fairfax County, Virginia. U.S. Geological Survey map
MF-2047, 1988.
16. Buchli, R. and Burkart, W. Influence of Subsoil Geology and Construction Technique
on Indoor Air 222Rn Levels in 80 Houses of the Central Swiss Alps. Health Physics
56:423-429; 1989.
17. Duval, J.S., Otton, J.K, and Jones, W.J. Estimation of Radon Potential in the Pacific
Northwest Using Geological Data. Portland: Bonneville Power Administration, 1989.
18. Kunz, C., Laymon, C.A., and Parker, C. Gravelly Soils and Indoor Radon,
proceedings of the 1988 Symposium on Radon and Radon Reduction Technology, Vol.
1, EPA600/9-89/006a (NTIS PB89-167480), paper. 5-75; 1989.
19. Muessig, K.W. Correlation of Airborne Radiometric Data and Geologic Sources with
Elevated Indoor Radon in New Jersey, proceedings of the 1988 Symposium on Radon
and Radon Reduction Technology, Vol. 1, EPA60019-89/006a (NTIS PB89-167480),
paper 5-1; 1989.
20. Reimer, G.M. and Gundersen, L.C.S. A Direct Correlation Among Indoor Rn, Soil Gas
Rn and Geology in the Reading Prong Near Boyertown, Pennsylvania. Health Physics
57:155-160; 1989.
21. Smith, D.L. and Hansen, J.K Distribution of Potentially Elevated Radon Levels in
Florida Based on Surficial Geology. Southeastern Geology 30:49-58; 1989.
-------
22. Yokel, F.Y. Site Characterization for Radon Source Potential. Washington DC:
National Institute of Standards and Technology report NIST-IR 89-4106; 1989.
23. Gregg, L.T. and Coker, G. Geologic Controls on Radon Occurrence in Georgia,
proceedings of the 1990 International Symposium on Radon and Radon Reduction
Technology. Atlanta, GA: February 1990.
24. Laymon, C.A. and Kunz, C. Geologic Factors and House Construction Practices
Affecting Indoor Radon in Onondaga County, New York, proceedings of the 1990
International Symposium on Radon and Radon Reduction Technology. Atlanta, GA:
February 1990.
25. Otton, J.K. and Duval, J.S. Geologic Controls on Indoor Radon in the Pacific
Northwest, proceedings of the 1990 International Symposium on Radon and Radon
Reduction Technology. Atlanta, GA: February 1990.
26. Schumann, R.R., Peake, R.T., Schmidt, K.M., and Owen, D.E. Correlations of Soil-Gas
and Indoor Radon with Geology in Glacially Derived Soils of the Northern Great
Plains, proceedings of the 1990 International Symposium on Radon and Radon
Reduction Technology. Atlanta, GA: February 1990.
27. Nielson, K.K. and Rogers, V.C. Feasibility and Approach for Mapping Radon
Potentials in Florida. Salt Lake City, UT: Rogers & Associates Engineering Corp.;
report RAE-8945/3-1; 1990.
28. Nagda, N.L., Koontz, M.D., Fortmann, R.C., Schoenborn, W.A., and Mehegan, L. L..
Florida Statewide Radiation Study. Germantown, MD: Geomet Technologies Inc.
report IE-1808; 1987.
29. Roessler, C.E., University of Florida Environmental Engineering Department, GRU
data, private communication, 1992.
30. Rogers, V.C. and Nielson, K.K. Multiphase Radon Generation and Transport in
Porous Materials. Health Physics 60:807-815; 1991.
31. Rogers, V.C. and Nielson, K.K. Benchmark and Application of the RAETRAD Model,
proceedings of the 1990 International Symposium on Radon and Radon Reduction
Technology. Atlanta, GA: February 1990.
32. Rogers, V.C., Nielson, K.K, and Merrell, G.B. Radon Generation, Adsorption,
Absorption, and Transport in Porous Media. Washington D.C.: U.S. Department of
Energy report DOE IERI60664-1; 1989.
33. Rogers, V.C. and Nielson, K.K. Correlations for Predicting Air Permeabilities and
222Rn Diffusion Coefficients of Soils. Health Physics 61:225-230; 1991.
-------
34. Nielson, K.K., Rogers, V.C., and Gee, G.W. Diffusion of Radon Through Soils: A Pore
Distribution Model. Soil Science Society of America Journal 48:482-487; 1984.
35. Nielson KK., Rogers, V.C., Brown, R.B., Harris, W.G., and Otton, J.K., Prototype
Mapping of Radon Potentials in Florida, Rockville, MD: The 1991 Annual AARST
National Fall Conference: Conference Preprints, Vol. 1, 145-162, 1991.
36. Nielson, K.K. and Rogers, V.C., Development of a Prototype Map of the Soil Radon
Potentials in Alachua County Florida. Salt Lake City: Rogers & Associates
Engineering Corp. report RAE-9127/3-1, October 1991.
37. SCS, State Soil Geographic Data Base (Statsgo) Data Users Guide, Lincoln, Nebraska:
National Soil Survey Center, Soil Conservation Service, U.S. Department of
Agriculture, draft report, 84pp, 1991.
38. Nazaroff, W.W., Doyle, S.M., Nero, A.V., and Sextro, R.G. Radon Entry via Potable
Water, pp. 131-157. In: Radon and Its Decay Products in Indoor Air, W. W. Nazaroff
and A. V. Nero, eds., New York: Wiley & Sons; 1988.
39. Acres. Measurement of Crack and Opening Contribution to Radon Entry (Feasibility
Study). Vol. Ill of Radon Entry Through Cracks in Slabs-on-Grade, Acres
International Corp., report P09314; 1990.
40. Nazaroff, W.W., Lewis, S.R., Doyle, S.M., Moed, B.A., and Nero, A.V. Experiments on
Pollutant Transport from Soil into Residential Basements by Pressure-Driven Airflow.
Environmental Science and Technology 21:459-466; 1987.
41. Nielson, KK. and Rogers, V.C. Radon Entry Into Dwellings Through Concrete Floors.
in Proceedings: The 1991 International Symposium on Radon and Radon Reduction
Technology. Philadelphia, PA: paper V-3; EPA-600/9-91-37c, April 1991.
42. Thomas, B.P., Cummings, E, and Wittstruck, W.H. Soil Survey of Alachua County,
Florida. Gainesville, FL: U.S. Department of Agriculture, Soil Conservation Service;
1985.
43. Yeh, G.T., FEMWATER: A Finite Element Model of Water Flow through Saturated-
Unsaturated Porous Media - First Revision, Oak Ridge, Tennessee: U.S. Department
of Energy report ORNL-5567IR1, 1987.
44. Sullivan, T.M., Kempf, C.R., Suen, C.J., and Mughabghab, S.M., Low-Level
Radioactive Waste Source Term Model Development and Testing, Washington DC:
U.S. Nuclear Regulatory Commission report NUREG/CR-5204, 1988.
45. EG&G Geometries, Aerial Gamma Ray and Magnetic Survey Gainesville and Daytona
Beach Quadrangles Florida. Final Report. Grand Junction, Colorado: U.S.
Department of Energy report GJBX-101, 1981.
-------
46. Otton, J.K., Asher-Bolinder, S., Reimer, G.M., and Henry, M., Prototype Mapping of
Soil Radon Potential in Alachua County Florida: Geologic and Radiometric Support
Investigations, Denver, Colorado: Draft USGS report to Florida Department of
Community Affairs, December 1991.
47. EPA. Monitoring for Radon-222 Emissions, Method 115, Appendix B of National
Emission Standards for Hazardous Air Pollutants; Radionuclides, 40 CFR Part 61,
Federal Register 54:51654-51715,1989.
48. Rogers, V.C., Nielson, K.K., Sandquist, G.M., and Rich, D.C., Radon Flux
Measurement and Computational Methodologies, Albuquerque, NM: U.S. Department
of Energy report UMTRA-DOE/AL-2700-201, 1984.
-------
VIIIP-3
CORRELATION OF INDOOR RADON SCREENI.NG_lffiASUREMENTS WITH SURFICIAL GEOLOGY
USING GEOGRAPHIC INFORMATION SYSTEM?
by: Charles Schwenker, Jia-Yeong Ku*,
Charles Layman and Charles Kunz
N.Y.S. Department of Health
Wadsworth Center for Laboratories and Research
Laboratory of Inorganic and Nuclear Chemistry
Empire State Plaza
Albany, N.Y. 12201-0509
*N.Y.S. Department of Health
Bureau of Environmental Radiation Protection
2 University Place
Albany, New York
ft
Present Address
National Aeronautics and Space Administration
Marshall Space Flight Center
Huntsville, Al 35812
ABSTRACT
The State of New York has a growing database of over 50,000 homes for
which radon levels have been measured. These data include information on home
construction, type of heating, and location by address. We have looked at
correlations with many home variables but have not been able to do a thorough
analysis of radon-level relations to surficial geology, which requires
accurate spatial relation of home location with surficial geology maps. We
are doing this by using a Geographic Information System mapping program (CIS),
to which we will add boundary maps of surficial geology, derived by digitizing
existing maps. The indoor radon measurement data for Albany County have been
linked to latitude and longitude coordinates using the U.S. Bureau of the
Census Topologically Interpreted Geographic Encoding and Referencing (TIGER)
database. The surficial geology boundary map has also been prepared for
Albany County and relationships of indoor radon to surficial geology are
discussed. Comparisons are made for mapping indoor data using county-wide
averages, zip code averages and individual measurement mapping.
-------
INTRODUCTION
The State of New York has a growing database of over 50,000 homes for
which radon levels have been measured. These data include information on home
construction, type of heating, and location by address. Mapping the average
or geometric mean indoor-radon concentrations for the counties or by zip code
is useful, providing information on the regional distribution of indoor-radon
concentration. More information is obtained when each measurement is
specifically located and mapped.
Computer-based Geographic Information Systems (CIS) are being developed
and utilized to map various spatially located information such as roads,
buildings, and geologic features. Identifying the latitude and longitude
coordinates for the individual radon measurements would make it possible to
map the indoor-radon data utilizing a Geographic Information System.
The geologic features that affect indoor-radon concentrations often vary
over smaller spatial dimensions than county, zip code, township, or other
politically defined areas. Therefore it is sometimes difficult to associate
indoor-radon concentrations with environmental factors such as surficial
geology. When the individual radon measurements are located by latitude and
longitude, the distribution of indoor concentrations can be correlated with
areas of any size or shape. Surficial geology maps have been drawn for New
York State at a scale of 1:250,000. These maps are being digitized into a GIS
so that the surficial geology can be readily correlated with spatially located
radon data. This will help facilitate the identification of surficial geology
associated with above-average indoor radon. The most direct method for
determining the potential for above-average indoor radon is to measure indoor
concentrations in a number of homes in the area in question. Finding a
correlation of high radon levels with a county or zip code is often not of any
value as a predictor of what to look for in another county or zip code.
However, if correlations can be developed with surficial geology for certain
soil types, this could serve as an indicator of a high probability that homes
in similar unmeasured areas will also have above-average radon. There are
other situations in which spatially located radon measurements can be useful.
For example, when the individual data points within a county or zip code area
are mapped, it may be possible to identify high and low risk areas within the
county or zip code area. Counties with average or below average
concentrations of indoor radon may in fact contain areas with above-average
indoor radon. The radon potential of a specific location, such as a school or
building site, can be estimated by averaging the indoor measurements within a
specified radius of the location and by examining the radon potential for the
geology at that location. When the individual radon measurements are mapped,
the spatial distribution of the data will identify those areas where most of
the measurements have been made so that proper interpretation and weighting
can be applied to the area-wide averages. Clearly, the utility of indoor-
radon data is considerably increased when they are specifically located by a
coordinate system.
-------
LOCATING RADON MEASUREMENT DATA
The key element in this project is to find an efficient means of
assigning coordinates to large numbers of indoor radon measurements. When
dealing with less than a few hundred radon measurements, it may be possible to
use street maps to locate each measurement. For a large database such as the
more than 50,000 homes measured by the New York State Department of Health, it
is not feasible. The Albany County database selected for this study consisted
of 1080 homes. This database included 150 homes which were P.O. box or rural
delivery, leaving 930 homes with radon measurements and street addresses. At
this point it is not possible to match a P.O. box or rural delivery address
with latitude and longitude coordinates using the TIGER database. This
hinders our ability to match coordinates with addresses in rural areas. Each
radon measurement is associated with a name, address, home construction, and
home operation information. Radon readings at different levels of the house
are available for many homes. Only the readings in the basement are used in
this study.
We are investigating two approaches to link latitude and longitude
coordinates with the addresses for the indoor radon measurement data. One
approach is to use the US Bureau of the Census Topologically Integrated
Geographic Encoding and Referencing (TIGER) database. The second is to use
the New York State Department of Equalization and Assessment (NYS DEA) tax
assessment database. The tax assessment database contains detailed
information about each household, but not every locality's tax assessment
database has latitude and longitude information in it. The geographical
location of the radon measurement is assigned by matching the common link
between the radon data and the TIGER and NYS DEA data, which is the name and
address field.
We were able to locate spatial coordinates for 73% of the 930 radon
measurements in Albany County that had street addresses. The radon
measurements that we were unable to locate include homes in newly developed
areas. Other reasons for not matching measurements with coordinates are being
investigated.
INDOOR RADON DISTRIBUTION BY COUNTY
The average basement screening indoor-radon concentrations for each
county in New York State are shown on figure 1. Various regional differences
are evident and have been discussed elsewhere(l). The counties with the
highest average indoor radon are located in the southwestern part of the
state, generally in an arc south of the Finger Lakes. Gravelly glacial out-
wash deposits forming highly permeable soils in the valleys of this region is
the principal factor resulting in above-average indoor radon (2). The
counties with the lowest average indoor-radon concentrations are located in
the Adirondack Mountains and on Long Island. The soil in both these areas has
below-average radium and is of moderate permeability. Regional differences
for indoor-radon concentrations can usually be understood from the general
-------
characteristics of the regional geology. However, mapping and interpretation
at this scale requires sufficient field data at specific locations to
determine the radon potential of the surficial geology in the area.
Additional information is obtained by looking more closely at the distribution
of the indoor radon measurement data.
ZIP CODE MAPPING
The average basement screening concentration in Albany County is 3.8
pCi/L which is less than the statewide average of 5.5 pCi/L. The geometric
mean for Albany County is 1.7 pCi/L and the statewide geometric mean is 2.5
pCi/L. Average basement screening data distributed by zip code are shown for
Albany County in figure 2. A total of 836 screening measurements were used to
determine the zip code averages. Less than 10 measurements were made in 8 of
the 25 zip code areas; however, in the more populous areas sufficient
measurements were made to obtain more reliable averages. Geometric means were
only calculated for zip codes with 10 or more measurements. Most of the
measurements were made in and around the city of Albany in the northeastern
part of the county, with few measurements in the rural areas.
The surficial geology for Albany county is shown in figure 3. The map
for Albany County has been digitized into a CIS from surficial geology maps
prepared by the NYS Geological Survey. The NYS Geological Survey is in the
process of digitizing the surficial geology maps for the entire state.
Although the surficial geology in Albany county is varied and complex,
particularly in the area around the city of Albany, it is possible to obtain
some correlation between zip codes, average indoor concentrations, and the
surficial geology. In the northeastern region of the county, there are fairly
large areas of sand dunes and lacustrine (lake) sand (figure 3). Several soil
measurements relating to indoor radon potential have been made in this
area(3). The sandy soil in the dunes has below-average Ra-226 (0.5pCi/g),
below-average soil gas Rn-222 (250pCi/L at a depth of 1.2m) and moderate
permeability resulting in below-average potential for indoor radon. The
average concentration of Ra-226 in soils in the U.S. is 1.0pCi/g(4) and
average soil-gas Rn-222 at a depth of 1.2 m in New York State is about
700pCi/L (3). The several zip code areas located in the region of sandy soils
have geometric-mean indoor-radon concentrations of less than 1.0 pCi/L which
is below the County mean of 1.7 pCi/L and well below the statewide mean of 2.5
pCi/L. The zip code areas in the sandy region are quite populous and the
number of measurements in these zip codes is sufficient for reasonably
reliable average values.
The geometric mean indoor radon concentrations for several of the zip
code areas are above the county mean. Two of these areas will be examined in
more detail.
-------
INDIVIDUAL MEASUREMENT MAPPING
There were 1080 screening measurements for Albany County at the time the
paper was being prepared. From this database of 1080 screening measurements,
680 homes were matched with latitude and longitude coordinates and are plotted
in figure 4. The size of the dot representing each measurement is
proportional to the indoor radon concentration. The measurement density is
greatest in and around the city of Albany. Few screening measurements were
made in the rural areas of the county; in addition few coordinate matches were
made in the rural zip codes, resulting in few or no data located by
coordinates for the rural areas. Some regions with above-average levels can be
seen from the distribution of measurements.
Zip code 12211 and the surrounding area are shown in figure 5. The
measurement data are superimposed on the surficial geology. It can be seen
that most of the higher levels are located in or near the kame area while the
measurements in the dunes are below average. It should be noted that the
boundaries for the different surficial geologies are not as sharp as pictured.
The New York State Geological Survey Surficial Geology maps were drawn to
conform to USGS maps, in which some terrain features are not accurately
located. This means that surficial geology areas are not always plotted to
their exact locations. Plotting and transcribing errors are magnified when an
area is expanded for closer examination. In addition, there is normally a
gradation from one surficial geology to another while the maps show sharp
boundaries. These factors can combine to give errors of up to 500 meters in
actual surficial geology boundary locations. The soils in the kame area were
measured for Ra-226 concentration, soil gas Rn-222 concentration and
permeability for gas flow at the location indicated by the circle of figure 5.
The soil Ra-226 concentration was measured at 0.7 pCi/g, which is below the
U.S. average of 1.0 pCi/g. The soil-gas Rn-222 concentration at a depth of
1.2m was 600 pCi/L which is about average for soils in New York State; the
permeability for gas flow was measured at 1.7xlO-7cm"2 which is moderate.
These soil characteristics suggest average indoor radon. However, in the U.S.
Geological Survey topographical maps it was noted that a gravel pit is located
in this kame deposit area (figure 5), indicating that the soils in portions of
this area can be gravelly and with higher permeabilities than observed at the
measurement location. More soil measurements are required to resolve this
question but it appears that the above-average indoor radon in this kame area
results from gravelly soils and fairly high permeabilities for gas flow.
The zip code area with the highest average indoor radon in Albany County
is zip code 12186 (figure 2). The northeastern part of this zip code area is
shown in figure 6 with the individual home measurements superimposed on the
surficial geology. Before the indoor radon screening program in New York
State was initiated, a home with above-average indoor radon was identified in
this area and a special study was conducted, which included measurements in
about 80 homes. Subsequently, when the screening program became available, a
number of residents in this area requested measurements, resulting in a
disproportionate number of measurements from an area identified as an above-
-------
average area. Both the screening and special study measurements are shown in
figure 6. A number of soil measurements have been made in this area of above-
average indoor radon. The soil Ra-226 averaged l.OpCi/g, the soil-gas Rn-222
675 pCi/L and the permeability for gas flow averaged 7xlO-6cm"2(2). The soils
in this rather small area of 1 to 2 square miles are very gravelly and are
highly permeable. The gravels are over 10m deep for most of the area.
Although the soils have average levels of Ra-226 and soil-gas Rn-222, the high
permeability of the deep gravel deposits is the soil characteristic resulting
in above-average indoor radon. The measurements shown in figure 6 indicate
that, although the indoor levels are considerably above average in the area
bounded by the two railroad tracks, the levels are higher in the region closer
to the point where the two tracks cross.
There can be considerable variation of the characteristics relating to
indoor-radon potential such as soil Ra-226, soil-gas Rn-222, and permeability
within a specific surficial unit. Variability is greater for some units such
as lacustrine delta, which is characterized by coarse to fine gravel and sand
and kame deposits, which includes kames, eskers, kame terraces, and kame
deltas, with coarse to fine gravel and sand. Within classifications such as
lacustrine delta and kame deposits the permeability for gas flow can be highly
variable depending on the proportion of coarse and fine particles. In
surficial units such as dunes, lacustrine sand, and lacustrine silt and clay
the permeability is generally moderate to low and average to below-average
indoor radon is expected.
All of the measurements in Albany County falling within the boundaries
of a particular surficial soil unit such as dunes were combined to determine
the geometric mean indoor radon for that particular surficial geology (table
1). There are sufficient data for six surficial units to calculate geometric
mean screening levels. The geometric mean screening levels for New York State
and Albany County are 2.5 and 1.7pCi/L respectively. The geometric mean levels
for the dunes, lacustrine sand and lacustrine silt and clay range between 0.9
and 1.1 pCi/L, which is considerably below the mean for both State and County.
The geometric means for the lacustrine delta and kame deposits are 3.6 and 3.2
pCi/L, which are above the State and County means. The 17 measurements used
to calculate the mean for the lacustrine delta are from the screening
measurements and do not include the special study data. However 13 of the 17
measurements were located in and around the area identified as an above-
average area. Clearly, the data for the lacustrine delta are heavily weighted
toward a gravelly region and cannot be interpreted as an accurate average for
this surficial unit.
CONCLUSIONS
This pilot study was limited to Albany County as a demonstration
project. As the process of locating measurement data with a coordinate system
is extended throughout the state, our understanding of the relationship
between indoor radon and surficial geology will be improved. A CIS allows for
overlaying additional spatial information for correlation such as topography,
-------
NURE data and more detailed soil information from the U.S. Geological Survey
Soil Conservation Service maps. Clearly, the value and utility of the indoor
radon data are increased when they are spatially located and analyzed with a
CIS. In the future thought should be given to facilitate coordinate location
of measurement data as they are acquired.
Finally care must be taken in presenting and using the data to maintain
measurement confidentiality. The scale and street information should be such
that specific homes cannot be identified.
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.
ACKNOWLEDGEMENT
We would like to acknowledge discussions with Bob MacHaver and Andy
Silfer in 1989, who first suggested the use of the tax assessment database to
locate home measurement data.
REFERENCES
1. Layman, C., Kunz, C. and Keefe, L., Indoor Radon in New York State:
distribution, sources and controls, New York State Department of Health,
Technical Report. Nov 1990
2. Kunz, C., Layman, C. and Parker, C., Gravelly Soils and Indoor Radon:
Proceedings EPA 1988 Symposium on Radon and Radon Reduction Technology,
Denver, CO. Oct 1988.
3. Kunz, C., Influence of Surficial Soil and Bedrock of Indoor Radon in New
York State Homes, New York State Energy Research and Development
Authority Report 89-14. Oct 1989.
4. Myrick, T.E., Bevin, B.A., and Haywood, F.F., Determination of
Concentrations of Selected Radionuclides in Surface Soil in the U.S.,
Health Physics, Vol.45, No.3, pp631-642 (1983).
-------
(Basement Screenings Only)
As of October 1. 1991
0.1 - 1.9
4.0 - 7.9
Illlllllll 2.0 - 3.9
8.0 - 15.9
Figure 1.
Headings are measured In pCl/1
County averages from about 50,000 basement screening
measurements in New York State.
-------
A bony County
Geometric Mean Indoor Radon Concentration by Zip Code
Radon pCi/l
0 to 0.49
0.5 to 0.99
1 to 1.49
1.5 to 1.99
2 to 2.49
2.5 to 4
Missing
Miles
10
Figure 2. zip Code averages for Albany County. Averages for zip codes with less than 10 measurements are not included.
-------
Albany County
Surficial Geology
Map Legend
Bedrock
Dunes
Kame
Kame moraine
Lacustrine Delta
Lacustrine sand
Lacustrine silt and clay
Outwash sand and Gravel
Swamp deposits
Recent deposits
ill
Till-moraine
Miles
E
5
i (!
Figure 3.
Surficial geology for Albany County. Adapted from surficial geology maps prepared by the ms Geological
Survey, State Education Department.
-------
Albany County - Zip Codes
Radon Measurements Matched Data Points
Map Legend
Zip Codes
• Datapoints
Radon pCi/l
31.2
62.5
125.0
Miles
Figure 4. Radon screening measurements for Albany County. The size of the dot is proportional to the indoor radon
concentration.
-------
Albany County - Zip 12211
Surficial Geology and Indoor Radon Concentrations
Map Legend
Bedrock
Dunes
Kam e
Kome moraine
Lacustrine Delta
few! Lacustrine sand
Lacustrine silt and clay
Outwash sand and Gravel
swamp deposits
Recent deposits
Till
Till moraine
Gravel pits
Radon in Soil test site
Datopoints
Radon pCi/l
7.5
15.0
30.0
Miles
o
0.5
Figure 5. Surficial geology and indoor radon screening measurements for zip code 12211 and surrounding area in
Albany County.
-------
Albany County - Zip 12186
Surficial Geology and Indoor Radon Concentrations
Map Legend
Bedrock
Dunes
Kame
Kame moraine
Lacustrine Delta
Lacustrine sand
Lacustrine silt and clay
Outwash sand and Gravel
Swamp deposits
Recent deposits
Till
Till-morame
— Highways
**-* Railroads
X Schools
• Datapoints
Radon pCi/l
31.2
62.5
125.0
Mi 6S
0.5
Figure 6. Surficial geology and indoor radon measurements for the northeastern part of zip code 12186 in Albany County,
Indoor radon measurements include both screening and special study measurements.
-------
TABLE 1. GEOMETRIC MEAN INDOOR RADON LEVELS FOR
SEVERAL DEPOSITS IN ALBANY COUNTY
Surficial
Geology
No. of
Measurements
Geometric
Mean
(pCi/L)
Min.
(pCi/L)
Max
(pCi/ L)
Dunes 213
Lacustrine Sand 258
Lacustrine Silt & Clay 53
Lacustrine Delta 17
Recent Deposits 36
Kame Deposits 35
Surficial Deposit Explanation*
0.9
0.9
1.1
3.6
1.5
3.2
0.1
0.1
30
34
10
58
16
74
Dunes: Fine to medium sands, well sorted, stratified, non-calcareous,
unconsolidated, generally wind-reworked lake sediments, permeable, well
drained, thickness variable (1-10 meters).
Lacustrine sand: Sand deposits associated with large bodies of water,
generally a near-shore deposit or near a sand source, well sorted, stratified,
generally quartz sand, thickness variable (2-20 meters).
Lacustrine silt and clay: Generally laminated silt and clay, deposited in
proglacial lakes, generally calcareous, potential land instability, thickness
variable (up to 100 meters).
Lacustrine delta: Coarse to fine gravel and sand, stratified, generally well
sorted, deposited at a lake shoreline, thickness variable (3-15 meters).
Recent deposits: Generally confined to floodplains within a valley, oxidized,
non-calcareous, fine sand to gravel, in larger valleys may be overlain by
silt, subject to frequent flooding, thickness 1-10 meters.
Kame deposits: Includes kames, eskers, kame terraces, kame deltas, coarse to
fine gravel and/or sand, deposition adjacent to ice, lateral variability in
sorting, coarseness and thickness, locally firmly cemented with calcareous
cement, thickness variable (10-30 meters).
* Taken from surficial geologic maps of New York, University of the State of
New York, State Education Department.
-------
VIIIP-4
Analysis of Indoor Radon in New Mexico Using
Geographic Information Systems (GIS)
Richard A. Dulaney
Lockheed Engineering and Sciences Company
Environmental Programs Office
1050 E. Flamingo Road #120
Las Vegas, NV 89119
(702)798-3158
FTS: 545-3158
ABSTRACT
Geographic Information Systems (GIS) are a powerful computer based information management
tool for spatial, or geographic, data. This technology, which blends traditional computer
cartography with state-of-the-art database management, is being employed to analyze indoor
radon in New Mexico. The project, being undertaken in support of EPA Region 6, looks at
both state-wide and local patterns of indoor radon. The geographic distribution and spatial
relationships of over 1000 indoor radon samples, taken from the EPA/State random survey
conducted in 1989, are examined. The correlation of these radon samples with geology, soils,
aerial radiometric and population data is explored using GIS spatial overlay techniques.
-------
VIIIP-5
A RADON "PIPE" (?) IN THE BREVARD FAULT ZONE NEAR ATLANTA. GEORGIA
L.T. Gregg, P.G., and John Costello, P.G.
Atlanta Testing & Engineering
11420 Johns Creek Parkway
Duluth, Georgia 30136
ABSTRACT
The Brevard fault zone (BFZ) extends from the Smith River allochthon in southern Virginia
southwestward to the Fall Line unconformity in central Alabama. The BFZ contains a sequence of
retrograded (low-grade) metamorphic rocks that has been affected by major thrust and lateral-slip
faults that moved under both ductile and brittle conditions through the Late Paleozoic Era.
Compression-related movement is believed to have ceased following the Alleghanian orogeny (ca. 300
ma). The adjacent Palmetto Granite (ca. 325 ma) in Southwest Atlanta is sheared by Brevard
faulting, and the BFZ is cross-cut by Mesozoic extensional fractures and dikes constraining the latest
movement. Bedrock lithologies include mylonite, phyllonite, and cataclasite. Regional strike of the
BFZ in the Atlanta area is N45°E and regional dip of foliation is, in general, 20 to 60 degrees SE.
The dip angle on individual fault planes is generally about 15 to 20 degrees. BFZ rocks have mostly
weathered by chemical processes, to a locally thick saprolite residuum, typically up to 50 to 75 feet
or more.
In March, 1990, a soil-gas radon survey was conducted at a new building site located in the
BFZ southwest of Atlanta. Radon concentrations up to 13,050 picocuries per liter (pCi/1) were
measured at original grade in near-surface soil. After site development, which involved undercutting
the original grade up to 15 feet, another soil-gas survey in October, 1990 showed radon
concentrations up to 1,130 pCi/1. By separately contouring the measured concentrations from both
surveys and overlaying the resulting maps, a preliminary geologic model was developed. The data
suggest a very localized upward-spreading "pipe" through the soil/saprolite residuum. Geologic
controls on this "pipe" probably include northeast-trending foliation and/or (most probably) brittle
faults and joints which may be cross-cut by northwest-trending Mesozoic extensional fractures. The
intersection of these permeable zones may cause or control the "pipe".
-------
INTRODUCTION
During the period March-October, 1990, a rather curious near-surface geological feature was
mapped in subsurface soils of the Brevard fault zone (BFZ) at a new building site some 13 miles
southwest of Atlanta, Georgia. We have named this feature a radon "pipe" because of certain
morphological (not diagenetic) similarities to kimberlite pipes in southern Africa and elsewhere. The
"pipe" was delineated by two soil-gas radon surveys. From what is known about the geology of the
BFZ in this area, we have hypothesized a set of geologic controls on occurrence and orientation of
the "pipe".
GEOLOGY OF THE BREVARD FAULT ZONE
APPALACHIAN REGION
The BFZ is a conspicuous NE-SW-trending linear feature on satellite images and small-scale
topographic and geologic maps of the southern Appalachian crystalline core. At least 31
interpretations of Brevard zone geology (1) have been advanced since the pioneer mapping of Keith
(2) identified the Brevard Schist, named after exposures of dark schistose rock near Brevard, NC.
Since that time, most geologists have interpreted various types of faulting to explain BFZ rock
textures (e.g., mylonite, phyllonite, etc.), discontinuity and repetition of mappable internal rock units,
and the overall linearity of the belt. Conventional Brevard zone characteristics are summarized by
Hatcher (3):
The zone is a linear fault zone throughout most of its extent, but it appears
to splay toward its ends. Some geologists (4; 5) have questioned the existence
of faulting along segments of the zone.
Foliation dips moderately over much of its length.
The zone is stratigraphically controlled at least along part of its extent.
Local slices of comparatively unmetamorphosed platform-type carbonate rocks
occur within the zone.
The zone occupies several rock units along strike, but the same lithologies
(metagraywacke, schist, amphibolite, aluminous schist, and quartzite) that
comprise several major, locally named lithostratigraphic sequences (Sandy
Springs Group, Tallulah Falls Formation, Ashe and Alligator Back
Formations, and Lynchburg Formation) lie northwest of the zone, except in
the Grandfather Mountain, NC area.
The BFZ has been affected by at least one early ductile deformation and at
least one later brittle deformation. Most likely, more than one of each type
of event has occurred.
Bobyarchick and others (1) summarize BFZ movement history throughout the Late Paleozoic
Alleghanian orogeny as beginning with a ductile, dextral (right lateral) shear zone. Structures related
-------
to this shearing crosscut Early to Middle Paleozoic fold-generated fabrics. Later Alleghanian thrust
faulting, accompanied by retrograde metamorphism, reactivated the deep part of the older ductile
shear zone to emplace the Blue Ridge-Inner Piedmont crystalline thrust sheet. The latest
Alleghanian continent-continent collision reactivated the Blue Ridge-Inner Piedmont thrust and
generated a high-angle, brittle thrust splay off of the Blue Ridge-Inner Piedmont thrust fault that
extended upward through the older ductile fabrics generated during dextral shearing. This last thrust
event emplaced the platform carbonate slices and brecciated older mylonitic rocks.
PIEDMONT PROVINCE OF GEORGIA
The BFZ in northern Georgia (including the Atlanta area) separates the eastern Blue
Ridge/northern Piedmont Geologic Province from the Inner Piedmont Geologic Province. In the
Atlanta region, the BFZ is bounded to the north by the Sandy Springs Group (6) and to the south
by the Atlanta Group (7). Throughout most of the Atlanta region, the Chattahoochee River is
entrenched into the BFZ and adjacent strike ridges in the Sandy Springs Group (8).
METROPOLITAN ATLANTA AREA
Typical BFZ rocks in the Atlanta region include protomylonite, mylonite, blastomylonite,
button schist, and phyllonite (generally in order of decreasing grain size), produced by compressive
deformation of various metasedimentary and igneous parent rocks under ductile conditions. Older
mylonitic textures are overprinted by later brittle fracturing and brecciation, enhancing rock porosity
and permeability. Local evidence of normal faulting (most likely also brittle) in the BFZ is reported
by McConnell and Abrams (9).
WEATHERING OF IGNEOUS/METAMORPHIC ROCKS
Brevard fault zone rocks throughout the Atlanta region are more or less chemically weathered
in natural surficial exposures. The weathered mantle above bedrock typically exhibits a compositional
and textural spectrum ranging from shallow, bioturbated residual soils, through an intermediate layer
of saprolite (completely weathered bedrock that retains original rock textures and structures), to a
deep, partially weathered zone of interlayered saprolite, partially weathered rock, and bedrock. The
thickness of this weathered mantle is dependent on bedrock mineralogy, concentration of joints
and/or other open, water-bearing structures, and local erosion rates.
PREVIOUS RADON STUDIES IN APPALACHIAN SHEAR ZONES
A recent and highly valuable compendium of radon studies in Appalachian shear zones is
presented in Section 2 (pp. 39-64) of (10). In particular, Gundersen reports on measured soil gas
radon concentrations in shear zones at or near Boyertown, PA, Brookneal, VA, Glen Gardner, NJ,
and Montgomery County, MD, and Agard and Gundersen discuss detailed soil geology and
geochemistry in the Boyertown-Easton, PA area. In these studies, Gundersen and her co-workers
showed very good correlation between mapped shear zones and high measured values (> 2,000
picocuries per liter, pCi/1) of radon in soil gas. The geology of these shear zones consists basically
of mylonites emplaced or occurring in hornblende and biotite gneisses, granodiorite, hornblende
granite, amphibolite, and various schists, gneisses, mafic rocks and ultramafic rocks. In these mylonite
-------
zones, Gundersen and her co-workers measured radon in soil gas up to 4,000 pCi/1, although the
average values were somewhat lower.
In addition, Gundersen and Gregg collaborated in 1989-90 (unpublished) on a series of
informal, small-scale measurements of radon in soil-gas in the BFZ on the west side of Atlanta, GA,
along 1-285 near its crossing of the Chattahoochee River. This limited work, in which Gundersen
took the lead and Gregg and others served primarily as observers and consultants, showed isolated
readings (1989) as high as 35-40,000 pCi/1; these were not replicated in the 1990 revisitation to the
site, but readings as high as 10,000 pCi/1 (Gundersen) and 8,500 pCi/1 (Gregg) were recorded by two
side-by-side teams with separate instrumentation. These limited data are nowhere near diagnostic,
but we feel they indicate the potential for high radon concentrations in soil (saprolite) overlying BFZ
rocks.
SOIL-GAS RADON SURVEY
SITE LOCATION, CHARACTERISTICS, AND DEVELOPMENT
The site is located about 13 miles southwest of downtown Atlanta, GA. Original site
topography, as taken from visual observations and the U.S.G.S. 7.5-minute topographic quadrangle
map, showed site drainage to be moderately sloping to the southwest. At the time of the first survey
(March, 1990), the site was ungraded and generally devoid of surface vegetation. Surface soils were
observed to be primarily clayey sandy silts and silty sands derived from phyllonite. As part of the
geotechnical investigation, 5 borings were drilled on original topography by hollow-stem auger to
depths up to 15 feet. Subsurface soils (saprolite) encountered were primarily micaceous silty clays.
Site development called for excavation to final grade up to 15 feet, to provide a slab-on-grade
building footprint of approximately 160,000 square feet. The planned structure was a one-story
building with no basement or below-grade structure. Grading was completed in August, 1990.
INSTRUMENTATION AND SURVEY LAYOUTS
Because the BFZ is known to locally have high concentrations of both Radon-222 (Rn-222)
and Radon-220 (Rn-220) in near-surface soil, the client requested a soil-gas radon survey. The initial
survey was conducted in March, 1990, utilizing an EDA RD-200 Radon Detector (alpha
scintillometer) and associated peripheral equipment. Soil-gas was sampled and a-counted at 14
stations, at a depth of 14 to 18 inches below ground surface. Sample station locations were taken
from architect's drawings of building column locations, and were established by pace and compass as
well as by visual observations of topography.
Because of the high concentrations of both radon isotopes (discussed below) that were
measured in the initial survey, a second survey was recommended. The client agreed, and the second
survey was conducted in October, 1990 at final grade. This soil gas sampling was done using the same
procedures and protocols as the first survey. The locations were selected to bracket the two locations
(i.e., sampling stations) which reported the highest radon concentrations in the initial survey.
SURVEY RESULTS
-------
The results of the first survey are shown in Table 1 and Figure 1, while the results of the
second survey are shown in Table 1 and Figure 2. Note the change in contour interval from Figure
1 to Figure 2. The highest Rn-222 value measured in the first survey was 13,050 pCi/1 (Station 12),
while the highest value measured in the second survey (Station 15, a short distance northwest of
Station 12), was 1,130 pCi/1. The reduction in highest measured value, over 15± feet of vertical cut,
was about 11.5:1. The Rn-222 isorad contours in Figures 1 and 2 have been drawn on the assumption
of linear gradients between sampling stations, which may not be correct.
The instrumentation used in these surveys is designed to primarily measure concentrations of
Rn-222, but it also measures Rn-220 concentrations. A linear regression analysis of Rn-222 (x) vs.
Rn-220 (y) values showed the following correlation coefficients (r/r2):
1st Survey: r = 0.998; r2 = 0.9%
2nd Survey: r = 0.999; r2 = 0.998
The conclusion from the measured data and the r/r2 values is that Rn-222 and Rn-220
concentrations are highly correlated. From that conclusion, we draw a further conclusion, viz. that
both uranium and thorium are present in anomalous amounts in the BFZ bedrock/saprolite/soil at
this site. This latter conclusion is generally borne out by the results of a reconnaissance study (11)
of the Tyrone Granite (Phase) some 14 miles south of the site, where anomalous values of both
uranium and thorium were reported using gamma scintillometry. The exact geologic relationship
between the Tyrone Granite (Phase) and the BFZ is murky at best. Current thinking (7) is that the
Tyrone Phase is related to the Ben Hill/Palmetto Granites, which are locally exposed in outcrop and
are hypothesized to be cupolas of a deep batholith emplaced up to 25 ma(?) before the last
compressional movement along the BFZ (ca. 300 ma). The Ben Hill/Palmetto granites have been
mapped over a large areal extent on the south side of the BFZ. In any event, we believe there is a
relationship, whether syngenetic or epigenetic, between the anomalous radionuclide concentrations
in BFZ rock/saprolite/soil and those to the south in Tyrone rock/saprolite/soil.
INTERPRETATION AND GEOLOGIC MODEL
As is well-known to practitioners, exact geologic interpretation of soil gas radon data is often
problematic at best and uncertain at worst (10; 12). The geologic, geochemical, and geophysical
controls on radon movement through saprolite and soil are only now becoming somewhat understood,
or at least enumerated (13; 14).
With this apologia, we offer the following model for the site geology and radon concentrations
measured at this site:
• Dominant lithology is low-grade (retrograde metamorphosed to greenschist
facies) phyllonite and minor mylonitized granitic rocks.
• Strike of mylonitic foliation is approximately N45°E and dip is 20 - 60°SE.
• Cross-cutting Mesozoic (Triassic?) extensional joints and fractures strike
generally NW; dip is generally steeply inclined to vertical.
-------
TABLE 1. MEASUREMENT OF RADON IN SOIL-GAS;
BFZ SITE, SOUTHWEST ATLANTA, GEORGIA
Sampling
Station
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
March, 1990 Survey
Rn-222.
pCi/l
120
BDL
500
10
BDL
BDL
BDL
180
1,020
BDL
9,030
13,050
BDL
40
—
—
—
—
...
Rn-220.
pCi/l
—
BDL
130
10
BDL
BDL
20
80
160
30
6,330
8,660
BDL
BDL
—
—
—
—
—
October, 1990 Survey
Rn-222.
pCi/l
—
—
—
—
—
„.
—
—
—
...
—
—
—
—
1,130
1000
BDL
40
60
Rn-220,
pCi/l
—
—
—
—
—
—
—
—
—
—
—
—
—
—
200
260
BDL
BDL
BDL
Notes:
1. pCi/l = picocurieslliter
2. BDL = Below Detection Limit
3. Measured values rounded to nearest 10 pCi/l
-------
LU
LT\
-3-
BUILDING FOOTPRINT
CONTOUR INTERVAL = 2000 pCi/1
7• SOIL GAS SAMPLING STATION
—I t CROSS SECTION
A' 105 210 FT.
FIGURE 1. RADON-222 ISORADS IN SOIL GAS AT ORIGINAL SURFACE
-------
LT\
-3-
19
10.
6.
BUILDING FOOTPRINT
CONTOUR INTERVAL =200 pCi/1
?• SOIL GAS SAMPLING STATION
CROSS SECTION
A A1
9
105
210 FT.
FIGURE 2. RADON-222 ISORADS IN SOIL GAS AT FINAL GRADE
-------
Strike of the short axis of the Rn-222 anomaly shown in Figure 1 is
approximately E - W; strike of the long axis is approximately N39°E. The two
axes are obviously not orthogonal. There is also a discernible tertiary (?) axis
striking approximately N45°W, somewhat elongated to the NW and
foreshortened (?) to the SE.
. Strike of the long axis of the Rn-222 anomaly shown in Figure 2 is
approximately N8°E. The short axis strike appears to be approximately the
same as in Figure 1, viz. E - W.
• Figure 3 presents a N - S cross-section through the anomalies contoured in
Figures 1 and 2. We interpret Figure 3 to represent an upward-spreading
"pipe" that widens as the original surface is approached. Geologic controls,
we believe, are northeast-trending mylonitic foliation coupled with similarly
oriented brittle faults and joints that have been cross-cut by northwest-
trending Mesozoic extensional fractures and joints. These structural features
extend from bedrock upward into the saprolite. The intersection of these two
"permeable" zones (foliation plus structural features) may cause or control the
"Pipe-"
Bedrock depth is uncertain, due to the lack of bedrock outcrops near the site
and the limitations of the geotechnical engineering survey that was conducted;
however, we suspect that bedrock (geologically, unweathered rock) may occur
at about 20 to 30 feet below final grade, which was the grade at which the
second radon survey was conducted.
CONCLUSIONS
While we have advanced a structurally-oriented model (relict foliation, fractures, and joints
in saprolite) for the occurrence and overall lateral geometry of this radon "pipe", we do not have a
plausible hypothesis for its vertical (apparent upward spreading) geometry. To develop such a
hypothesis would have required an extremely detailed study of soil and saprolite physical and chemical
properties, as well as measurement of radon concentrations in soil gas at several different elevations
(i.e., between original grade and final grade) and on a closely-spaced lateral grid pattern. This was
not possible within the constraints of the client-funded work reported in this paper.
A detailed micro-model of radon movement in Piedmont soil/saprolite would require
consideration and measurement of a large number of parameters. As is well known, once radon is
"liberated" (i.e., emanates) from its parent radium source, it will migrate by diffusion and convection
to zones of lower concentration (diffusion) and pressure (convection), e.g., vertically toward the
surface. As Gregg and Coker (14) have pointed out, radon concentration and migration in saprolite
are influenced by the lithology of the parent rock, the amount and degree of jointing and fracturing
and interconnection, the degree of water saturation, permeability and porosity, thickness, zonation
(whether the saprolite is structured or massive), and the distribution and extent of nanopores (pores
less than one micron in width). In surface and near-surface soil, the principal influences are
thickness, zonation (A, B, and C zones), moisture content (8 to 15 percent has been suggested as
optimum for radon emanation), permeability and porosity of the soil, and finally the temperature
-------
r~
A
15±FT
NE
ISORAD CONTOUR (pCi/1)
0 55 FEET
VERTICAL/HORIZONTAL EXAGGERATION » 3-7
A'
GRADE FOR
FIGURE 1.
GRADE FOR
FIGURE 2.
FIGURE 3- CROSS SECTION OF RADON "PIPE"
-------
gradient from the surface, which determines the water vapor pressure of the soil. Finally, there are
meteorologic and topographic effects on radon migration that can be enumerated but are not at all
well understood.
The "upward-spreading pipe" model suggested above has ramifications for other studies of
radon in soil gas. Except for a few client-funded studies such as this one, there is rarely if ever the
opportunity to revisit a site and measure radon at different depths (i.e., elevations below original
grade). Yet these revisitations and different sampling depths may be critical in trying to understand
the subsurface migration of radon.
The U.S.G.S. (10) has proposed (and uses) sampling soil gas for radon at a nominal depth of
70 to 75 cm. below ground surface. In the Piedmont Province, this depth presumably puts one into
the C soil horizon, or lowermost B and uppermost C horizons. We suggest that the relationship
between radon concentrations in the C horizon (or any soil horizon) and those at final grade
(Finished Floor Elevation, or FFE) should be site-specific, and that site-specific (i.e., FFE)
measurements should be taken before actual construction begins, so that (if necessary) the architect
can incorporate either passive or active radon mitigation features into the final design. As is well
known, saprolite is not soil and it is not rock, yet it has some physical and chemical characteristics
of both soil and rock. It needs to be intensively studied in its own right (14) for its contribution to
and influence upon the vertical and lateral migration of radon from source (radium atoms) to surface.
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. Bobyarchick, A.R., Edelman, S.H., and Horton, J.W., The role of dextral strike-slip in
the displacement history of the Brevard zone. In Secor, D.T., ed., Southeastern geological
excursions: Geological Society of America Southeastern Section guidebook, 1988. p. 53.
2. Keith, A., Description of the Mount Mitchell quadrangle. U.S. Geological Survey Atlas Folio
124, Washington, D.C., 1905. 10 pp.
3. Hatcher, R.D., Jr., Tectonics of the western Piedmont and Blue Ridge, southern
Appalachians: review and speculation. American Journal of Science, 278:276. 1978.
4. Crawford, TJ. and Medlin, J.H., The western Georgia Piedmont between the Cartersville and
Brevard fault zones. American Journal of Science. 273:712. 1973.
5. Hurst, V. J., Geology of the southern Blue Ridge belt. American Journal of Science 273-643
1973.
6. Higgins, M.W. and McConnell, K.I., The Sandy Springs Group and related rocks of the
Georgia Piedmont: nomenclature and stratigraphy. In Platt, P.A., ed., Short contributions
to the geology of Georgia. Georgia Geologic Survey Bulletin 93, Atlanta, Georgia, 1978.
p. 50.
-------
7. Higgins, M.W. and Atkins, R.L., The stratigraphy of the Piedmont southeast of the Brevard
zone in the Atlanta, Georgia area. In Wigley, P.B., ed., Latest thinking on the stratigraphy
of selected areas in Georgia. Georgia Geologic Survey Information Circular 54-A, Atlanta,
Georgia, 1981. p. 3.
8. Higgins, M.W., Geology of the Brevard lineament near Atlanta, Georgia. Georgia Geologic
Survey Bulletin 77, Atlanta, Georgia, 1966. 49 pp.
9. McConnell, K.I. and Abrams, C.E., Geology of the greater Atlanta region. Georgia Geologic
Survey Bulletin 96, Atlanta, Georgia, 1984. 127 pp.
10. Gundersen, L.C.S. and Wanty, R.B., eds., 1991, Field studies of radon in rocks, soils, and
water. U.S. Geological Survey Bulletin 1971, Washington, D.C., 1991. 334 pp.
11. McConnell, K.I., and Costello, J.O., Uranium evaluation of graphitic phyllites and other
selected rocks in the Georgia Piedmont and Blue Ridge. Georgia Geologic Survey Open -
File Report 80-5, Atlanta, Georgia, 1980. 127 pp.
12. Gregg, L.T. and Holmes, J.J., 1990, Radon detection and measurement in soil and
groundwater. In Ward, S.H., ed., Geotechnical and Environmental Geophysics. Vol. I:
Review and Tutorial. Society of Exploration Geophysicists, Tulsa, Oklahoma, 1990. p. 251.
13. Tanner, A.B., Methods of characterization of ground for assessment of indoor radon
potential at a site. In Gundersen, L.C.S., and Wanty, R.B., eds.. Field studies of radon in
rocks, soils, and water. U.S. Geological Survey Bulletin 1971, Washington, D.C., 1991. p.l.
14. Gregg, L.T., and Coker, G., Geologic controls on radon occurrence in Georgia. In Bearce,
D.T. and Neilson, M.J., eds., Case studies in applied geology in the Southeastern United
States. Georgia Geologic Survey Bulletin 122, Atlanta, Georgia, 1990. p.40.
-------
Session IX Posters
Radon Surveys
-------
IXP-1
SUMMARY OF REGIONAL ESTIMATES OF INDOOR SCREENING
MEASUREMENTS OF 222pn
by: Barbara Alexander, Nathaniel Rodman, and S.B. White
Research Triangle Institute
Research Triangle Park, N.C. 27709
Jeffrey Phillips
U.S. Environmental Protection Agency
Office of Radiation Programs
Washington, D.C. 20460
ABSTRACT
As part of an EPA/State cooperative program, a random sample of 54,851
houses from 38 of the 48 conterminous states has been screened for 222pn
over a six year period. The number of houses in each state was
sufficiently large to allow meaningful estimates to be derived for
geographic areas within each state. Summary statistics (e.g., the
arithmetic mean and the percentage of houses exceeding 4 pCi/L) and
associated 95% confidence intervals have been computed for 225 geographic
regions within the 38 states. This paper summarizes the 225 regional
estimates in tabular and graphical form. The results can then be used to
identify "hot spots" within the 38-state area. For example, there are 21
regions within seven states where more than 50 percent of the target
population houses are estimated to have screening measurements greater than
4 pCi/L.
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.
-------
INTRODUCTION
In response to the growing concern about potential health risks associ-
ated with indoor radon exposure, the U.S. Environmental Protection Agency
(EPA) began a program to provide assistance to states to measure radon
concentrations in homes. Since the winter of 1986-87, EPA has assisted 38
of the 48 conterminous states in conducting probability-based surveys of
indoor 222Rn. Short-term screening measurements were made in the lowest
livable level of over 54,000 randomly selected houses during winter heating
seasons. The 38 state radon surveys included in this paper were carried
out during six years of the program as listed below.
Year 1, 1986-87 heating season, eight surveys:
Alabama (AL) Rhode Island (RI)
Kansas (KS) Tennessee (TN)
Kentucky (KY) Wisconsin (WI)
Michigan (MI) Wyoming (WY)
Year 2, 1987-88 heating season, seven surveys:
Arizona (AZ) Missouri (MO)
Indiana (IN) North Dakota (ND)
Massachusetts (MA) Pennsylvania (PA)
Minnesota (MN)
Year 3, 1988-89 heating season, seven surveys:
Georgia (GA) Ohio (OH)
Iowa (IA) Vermont (VT)
Maine (ME) West Virginia (WV)
New Mexico (NM)
Year 4, 1989-90 heating season, eight surveys:
California (CA) Nevada (NV)
Idaho (ID) North Carolina (NC)
Louisiana (LA) Oklahoma (OK)
Nebraska (NE) South Carolina (SC)
Year 5, 1990-91 heating season, six surveys:
Arkansas (AR) Mississippi (MS)
Illinois (IL) Texas (TX)
Maryland (MD) Washington (WA)
Year 6, 1991-92 heating season, two surveys:
Montana (MT) Virginia (VA)
-------
The goals of the state radon surveys were twofold. Some measure of the
distribution of radon levels among residences was desired for major
geographic areas within each state and for each state as a whole. In
addition, it was hoped that each state radon survey would be able to
identify areas of potentially high residential radon concentration ("hot
spots") in the state, enabling the state to focus its attention on areas
where indoor radon concentration might pose a significant health threat.
METHODS
Two-day deployment of open-face charcoal canisters was used by 22
states during the first three years of the state radon survey assistance
program (1). During these years a diffusion barrier charcoal canister was
developed specifically to be less sensitive to the effects of humidity and
air flow than the open-face canister (2). Two-day deployment of barrier
canisters was used by eight states in Year 4 of the program. The exposure
period for the barrier canister was increased from two days to seven days
in Years 5 and 6. The performance of the charcoal canisters was monitored
periodically through the use of unexposed canisters, canisters exposed to
known levels of 222Rn, and collocated canisters.
To ensure the discovery of elevated radon concentrations within a home,
the charcoal canister was exposed under closed-house conditions during the
winter and was placed on the lowest livable level. Thus, the estimates of
indoor radon concentration provided by the survey reflect a worse-case
scenario and maximize the likelihood of identifying residences with high
radon concentrations. This method was used in the state surveys because it
is the one EPA recommends to homeowners for determining whether additional
tests are needed (3).
For each state, a probability sample of listed residential telephone
numbers was selected from a sampling frame constructed from telephone
directories for all communities in the state. Probability sampling permits
the extrapolation of survey results to the sampled population and enables
the calculation of measures of precision for the estimates. Because one of
the goals of each survey was to characterize the distribution of
residential radon levels for the state as a whole, the use of probability
sampling was imperative. In addition, probability-based surveys permit
valid comparisons of results from one state with those from another.
The target population for the surveys consisted of owner-occupied
residences with a permanent foundation, at least one floor at or below
ground level, and a telephone number published in the latest directory.
(Mobile homes with permanent foundations and airtight panels/skirts were
included beginning in Year 3.) The statistical estimates generated from
the survey data apply to a target population of over 28.6 million homes
from 38 states.
-------
Each state was divided into strata based on population density,
potential for high 222pn, and geographic areas for which separate
statistical estimates were desired. For convenience in selecting the
sample of telephone numbers, county boundaries were used to delineate the
geographic reporting regions. Each stratum was then sampled at a different
rate to ensure a wide dispersement of the sample across the state and to
enhance the chances of finding areas with elevated radon.
Interviewers from each state placed telephone calls to random
subsamples of the residential telephone listings. The interviewer first
screened for survey eligibility, which required that the dwelling qualify
as a member of the target population. Eligible households agreeing to
participate in the survey were provided with a charcoal canister and
instructions for placing it on the lowest livable level of their home.
Participants were instructed to return the canister to the EPA analysis
laboratory after exposing the canister under closed-house conditions for a
designated time period.
Because telephone numbers in different strata were selected at
different sampling rates, it was necessary to assign sampling weights that
counterbalanced the unequal selection probabilities. The weights assigned
were the inverse of the sample selection probabilities. An additional
weight adjustment was made to compensate for nonresponse so population
aggregates could be easily estimated from the sample data. All data
analyses were carried out using properly weighted data that reflected the
full complexity of the sample design. This permitted the generation of
unbiased statistical estimates (4).
RESULTS AND DISCUSSION
Based on program needs and available resources, each state determined
how many houses would be tested. The actual number of houses that provided
valid test data are shown in Table 1 and ranged from 376 in Rhode Island to
2,680 in Texas. In 30 of the 38 states, more than 1000 houses were tested.
The number of houses tested in each state was large enough to provide
reliable estimates of 222Rn concentrations for subpopulations of each state
(e.g., groups of counties formed by political or geologic boundaries).
During the course of the surveys, state personnel were asked to
identify geographic regions for which estimates of 222Rn levels were
important. These regions were composed of one or more counties (not
necessarily contiguous) with a combined sample size of at least 100 valid
screening measurements. Table 1 shows the number of geographic regions
defined for each state which ranged from one region in Rhode Island to 13
regions in Texas. The sample size in the regions ranged from 47 to 1,215
measurements, and only 4 of the 225 regions had fewer than 75 screening
measurements. Summary statistics (e.g., means, percentiles and
proportions) and associated 95% confidence intervals were computed for the
-------
TABLE 1. RANGE OF STATE RADON REGIONAL ESTIMATES FOR 38 STATES
STATE
AL
AR
AZ
CA
GA
IA
ID
IL
IN
KS
KY
LA
MA
MD
ME
MI
MN
MO
MS
MT
NC
ND
NE
NM
NV
OH
OK
PA
RI
SC
TN
TX
VA
VT
WA
WI
WV
WY
OVERALL
SURVEY
YEAR
1
5
2
4
3
3
4
5
2
1
1
4
2
5
3
1
2
2
5
6
4
2
4
3
4
3
4
2
1
4
1
5
6
3
5
1
3
1
NO. OF
HOUSES
TESTED
1,180
1,535
1,507
1,885
1,534
1,381
1,266
1,450
1,914
2,009
879
1,314
1,659
1,126
839
1,989
919
1,859
960
833
1,290
1,596
2,027
1,885
1,562
1,734
1,637
2,389
376
1,089
1,773
2,680
1,156
710
1,935
1,191
1,006
777
54,851
RANGE OF REGIONAL ESTIMATES
NO. OF
REGIONS
8
6
3
9
3
9
7
3
5
6
6
4
11
4
7
4
5
6
6
3
5
6
5
4
8
4
7
10
1
4
11
13
5
5
4
10
3
5
225
ARITHMETIC
MEAN (PCI/L)
0.7- 4.4
0.5- 1.8
1.1- 1.7
0.6- 1.5
0.8- 2.1
7.4-10.3
2.0- 5.1
2.0- 4.6
2.8- 4.2
1.1- 4.8
1.1- 4.5
0.3- 0.6
1.7- 4.6
0.7- 5.5
2.3- 5.6
1.5- 5.9
3.0- 6.3
1.9- 3.8
0.4- 1.2
4.1- 8.3
0.4- 3.4
4.8- 8.9
3.0- 7.1
2.2- 4.7
1.1- 3.2
3.2- 7.0
0.6- 1.6
2.3-17.8
3.2- 3.2
0.5- 1.5
1.0- 5.1
0.4- 3.4
0.6- 3.8
2.0- 3.2
0.4- 7.7
2.6- 4.8
1.9- 4.7
2.5- 5.4
0.3-17.8
% OF HOUSES
> 4 PCI/L
0.3-25.1
0.0- 9.1
1.2- 7.9
0.5- 5.5
0.3-10.3
61.4-86.9
8.9-30.5
9.6-35.3
18.5-36.2
3.4-41.2
2.2-34.5
0.0- 1.6
2.8-37.5
4.4-38.3
16.6-40.9
4.8-44.7
17.5-62.3
11.1-29.6
0.9- 3.8
36.4-52.5
0.6-17.9
46.1-72.4
25.5-62.9
11.2-41.6
3.8-24.7
18.3-48.4
1.1- 7.5
11.3-74.0
20.6-20.6
0.0- 7.0
0.7-29.9
0.0-22.4
0.7-27.5
10.8-19.8
1.3-45.0
14.4-44.3
10.7-31.2
12.6-51.0
0.0-86.9
-------
225 geographic regions within the 38 states. This data base was then
sorted in descending order of the various parameter estimates to examine
patterns in the regional estimates.
As noted earlier, all households in a given state were not selected
with equal probability. To counterbalance these unequal selection
probabilities, appropriate sampling weights were calculated and used in all
data analyses to assure unbiased statistical estimates.
ARITHMETIC MEAN
Table 1 shows the range of regional estimates for two of the parameters
computed for each of the 38 states. Previous papers utilizing similar
parameter estimates at the state level (5,6) have shown that 222Rn
concentrations vary widely from state to state. It is evident from Table 1
that radon levels vary even more from region to region. For instance
estimates of the arithmetic mean (AM) for the 225 regions range from 6.3
pCi/L in Louisiana to 17.8 pCi/L in Pennsylvania. Twenty-three of the 38
states have at least one region where the estimated AM is greater than 4
pCi/L and all the regions in three states (Iowa, Montana, and North Dakota)
have AMs above 4 pCi/L. Figure 1 graphically identifies regions with
II ^ ^ than 2 pC1/L' 2> AMs between 2 and 4 pCi/L, and 3) AMs greater
than 4 pCi/L. The lower two groups contain 83 and 82 regions respectively
and there are 60 regions with AMs greater than 4 PCi/L. In keeping with
the goal of identifying areas with elevated levels of radon or "hot spots"
the 60 regions with AMs greater than 4 pCi/L are examined in greater detail
in Table 2 and Figure 2. Table 2 gives, for each of the 60 regions, the
number of houses tested, summary statistics (i.e., AM and percentage of
houses exceeding 4 pCi/L) and associated confidence intervals, and a
listing of the counties included in the region. The regions in Table 2 are
listed in descending order according to the arithmetic mean. Figure 2
shows, graphically, the 60 regions broken down into three groups according
to the level of the AM. There are 36 regions with AMs between 4 and
6 pCi/L, 12 regions with AMs between 6 and 8 pCi/L, and 12 regions with AMs
greater than 8 pCi/L. The 12 regions with the highest mean concentrations
consist of seven regions in Iowa, three regions in Pennsylvania, and one
region each in Montana and North Dakota.
PERCENTAGE OF HOUSES EXCEEDING 4 PCI/L
Figures 1 and 2 focus on a region's arithmetic mean which is only one
point on the distribution of 222Rn screening measurements. Another point
of particular interest on this distribution is the percentage of houses
within a region that have screening measurements exceeding 4 pCi/L. For
the 225 regions, this percentage varied from 0% in Arkansas to 86.9% in
Iowa. Twenty-one of the 38 states have at least one region where more than
30-s of the houses have screening measurements exceeding 4 pCi/L. More than
-------
Arithmetic Mean in pCi/L
AM<=2 2
-------
Arithmetic Mean in pCi/L
4
-------
TABLE Z. STATE RADON REGIONAL ESTIMATES AND CONFIDENCE INTERVALS FOR REGIONS HITH ARITHMETIC MEAN VALUES > 4 PCI/L
STATE
PA
PA
IA
IA
IA
IA
NO
IA
IA
MT
PA
IA
NO
IA
NA
IA
REGION
2
3
6
Z
1
3
4
9
8
2
7
7
1
5
1
4
NO. OF
HOUSES
TESTED
258
270
179
160
144
128
470
133
143
266
207
169
423
187
708
138
ARITHMETIC MEAN
ESTIMATED
PCI/L
17.8
10.8
10.3
10.2
9.8
9.1
8.9
8.8
8.7
8.3
8.3
8.1
7.8
7.7
7.7
7.4
9SX CONF.
INTERVAL
12.3-23.4
8.4-13.2
8.8-11.7
9.1-11.3
8.3-11.2
7.9-10.2
8.1- 9.6
7.1-10.5
6.6-10.8
6.0-10.6
5.2-11.3
6.6- 9.6
6.8- 8.7
6.3- 9.0
6.3- 9.2
6.4- 8.3
X OF HOUSES
ESTIMATED
PERCENT
74.0
51.3
81.0
82.1
86.9
71.7
72.4
61.7
67.6
52.5
40.4
61.4
62.4
62.5
45.0
67.5
> 4 PCI/L
95X CONF.
INTERVAL
68.9-79.0
49.0-53.7
75.1-87.0
71.8-92.3
82.4-91.4
63.6-79.8
68.6-76.2
55.5-68.0
52.1-83.1
44.0-61.0
37.0-43.9
55.3-67.4
56.7-68.2
57.3-67.6
43.1-47.0
63.6-71.4
LIST OF COUNTIES INCLUDED IN THE REGION
CUMBERLAND, DAUPHIN, LANCASTER, LEBANON, PERRY, YORK.
ADAMS, BEDFORD, BLAIR, CARBON, CENTRE, CLINTON, COLUMBIA,
FRANKLIN, FULTON, HUNTINGDON, JUNIATA, LYCOMING, MIFFLIN,
MONTOUR, NORTHUMBERLAND, SCHUYLKILL, SNYDER, SULLIVAN,
UNION.
BOONE, DALLAS, GREENE, GRUNDY, HAMILTON, HARDIN, JASPER,
MARSHALL, POLK, STORY, MEBSTER.
AUDUBON, CALHOUN, CARROLL, CRAWFORD, GUTHRIE, HARRISON,
IDA, MONONA, SAC, SHELBY, HOODBURY.
BUENA VISTA, CHEROKEE, CLAY, DICKINSON, EMMET, LYON,
O'BRIEN, OSCEOLA, PALO ALTO, PLYMOUTH, POCAHONTAS, SIOUX.
ADAIR, ADAMS, CASS, FREMONT, MILLS, MONTGOMERY, PAGE,
POTTAHATTAMIE , RINGGOLD, TAYLOR, UNION.
BARNES, CASS, GRAND FORKS, HIGHLAND, SARGENT, STEELE,
TRAILL.
DAVIS, DES MOINES, HENRY, JEFFERSON, KEOKUK, LEE, LOUISA,
MUSCATINE, VAN BUREN, HAPELLO, WASHINGTON.
APPANOOSE, CLARKE, DECATUR, LUCAS, MADISON, MAHASKA,
MARION, MONROE, WARREN, WAYNE.
BEAVERHEAD, BROADHATER, DEER LODGE, GALLATIN, GRANITE,
JEFFERSON, JUDITH BASIN, LEWIS AND CLARK, MADISON,
MEAGHER, PARK, POWELL, RAVALLI, SILVER BOH, SWEET GRASS,
HHEATLAND, YELLOWSTONE NAT PARK.
BRADFORD, MONROE, PIKE, POTTER, SUSQUEHANNA, TIOGA, HAYNE,
WYOMING.
8ENTON, CEDAR, CLINTON, IOWA, JACKSON, JOHNSON, JONES,
LINN, POWESHIEK, SCOTT, TAMA.
ADAMS, BILLINGS, BOWMAN, GRANT, HETTINGER, MORTON, OLIVER,
PEMBINA, SLOPE, STARK.
ALLAMAKEE, BLACK HAWK, BREMER, BUCHANAN, CHICKASAW,
CLAYTON, DELAWARE, DUBUQUE, FAYETTE, HOWARD, WINNESHIEK.
DOUGLAS, FERRY, GRANT, LINCOLN, OKANOGAN, PEND OREILLE,
SPOKANE, STEVENS.
BUTLER, CERRO GORDO, FLOYD, FRANKLIN, HANCOCK, HUMBOLDT,
KOSSUTH, MITCHELL, WINNEBAGO, WORTH, WRIGHT.
-------
TABLE 2. STATE RADON REGIONAL ESTIMATES AND CONFIDENCE INTERVALS FOR REGIONS HITH ARITHMETIC MEAN VALUES > 4 PCI/L
STATE
NO. OF
HOUSES
REGION TESTEP
ARITHMETIC MEAN
ESTIMATED 95X CONF.
PCI/L INTERVAL
'/. OF HOUSES > 4 PCI/L
ESTIMATED 95X CONF.
PERCENT INTERVAL
LIST OF COUNTIES INCLUDED IN THE REGION
PA
PA
NE
OH
ND
NO
MN
NE
MI
MN
ME
MT
PA
MO
HY
2
5
268
273
358
428
210
121
142
190
4
B
211
264
279
316
135
7.3
7.3
7.1
7.0
6.9
6.6
6.3
6.1
5.9
5.7
5.6
5.6
5.6
5.5
5.4
6.0- 8.6
6.5- 8.1
6.0- 8.2
5.6- 8.4
4.8- 9.0
5.6- 7.7
j.i- 6.6
5.6- 6.6
4.6- 7.1
5.2- 6.1
4.5- 6.8
3.8- 7.4
3.0- 8.2
4.6- 6.4
4.0- 6.8
44.2
48.4
39.9
36.4
27.5
39.3-49.1
43.9 SB.1-49.7
57.9 55.0-60.8
46.0-50.7
60.8 54.3-67.2
65.3 57.5-73.1
62.3 54.4-70.3
62.9 58.9-66.8
44.7 38.8-50.6
55.4 48.7-62.1
34.2-45.7
30.7-42.0
17.0-38.0
38.3 32.8-43.8
51.0 39.0-63.0
BERKS, BUCKS, CHESTER, DELAWARE, LEHIGH, MONTGOMERY,
NORTHAMPTON.
BEAVER, BUTLER, LAMRENCE.
ANTELOPE, BOONE, BOYD, BURT, CEDAR, CUMING, DAKOTA, DIXON,
HOLT, KNOX, MADISON, PIERCE, STANTON, THURSTON, WAYNE.
ATHENS, BELMONT, FAIRFIELD, FRANKLIN, GALLIA, GUERNSEY,
HOCKING, JACKSON, LAMRENCE, LICKING, MEIGS, MONROE,
MORGAN, MUSKINGUM, NOBLE, PERRY, PICKAWAY, PIKE, ROSS,
SCIOTO, VINTON, WASHINGTON.
BENSON, CAVALIER, DICKEY, EDDY, FOSTER, GRIGGS, LA MOURE,
NELSON, RAMSEY, RANSOM, STUTSMAN, WALSH, WELLS.
DUNN, EMMONS, GOLDEN VALLEY, MCKENZIE, MCLEAN, MERCER,
SIOUX.
BLUE EARTH, BROWN, DODGE, FARIBAULT, FILLMORE, FREEBORN,
GOODHUE, HOUSTON, LE SUEUR, MOWER, OLMSTED, RICE, STEELE,
WABASHA, WASECA, WINONA.
BUTLER, CASS, COLFAX, DODGE, DOUGLAS, FILLMORE, GAGE,
JEFFERSON, JOHNSON, LANCASTER, NEMAHA, OTOE, PAWNEE,
PLATTE, POLK, RICHARDSON, SALINE, SARPY, SAUNDERS, SEWARD,
THAYER, WASHINGTON, YORK.
HILLSDALE, LENAWEE, WASHTENAW.
AITKIN, BENTON, BIG STONE, CARLTON, CARVER, CHIPPEWA,
CHISAGO, COTTONWOOD, DOUGLAS, GRANT, ISANTI, JACKSON,
KANABEC, KANDIYOHI, LAC QUI PARLE, LINCOLN, LYON, MARTIN,
MCLEOD, MEEKER, MILLE LACS, MURRAY, NICOLLET, NOBLES,
PINE, PIPESTONE, POPE, REDWOOD, RENVILLE, ROCK, SCOTT,
SHERBURNE, SIBLEY, STEVENS, SWIFT, TRAVERSE, WATONWAN,
WRIGHT, YELLOW MEDICINE.
CUMBERLAND, YORK.
CASCADE, CHOUTEAU, FLATHEAD, GLACIER, HILL, LAKE, LIBERTY,
LINCOLN, MINERAL, MISSOULA, PONOERA, SANDERS, TETON,
TOOLE.
CAMERON, CLARION, CLEARFIELD, CRAWFORD, ELK, ERIE, FOREST,
JEFFERSON, MCKEAN, MERCER, VENANGO, WARREN.
ALLEGANY, FREDERICK, GARRETT, WASHINGTON.
LINCOLN, NIOBRARA, SHERIDAN, WESTON.
-------
TABLE Z. STATE RADON REGIONAL ESTIMATES AND CONFIDENCE INTERVALS FOR REGIONS HITH ARITHMETIC MEAN VALUES > 4 PCI/L
STATE
ID
NO
TN
KS
ME
NO
HI
ME
NM
OH
PA
MV
IL
MA
MN
REGION
1
3
4
3
7
5
1
3
2
1
9
1
2
5
ft
t.
NO. OF
HOUSES
TESTED
66
131
245
195
10Z
241
101
95
422
445
261
211
539
219
160
ARITHMETIC MEAN
ESTIMATED
PCI/L
5.1
5.1
5.1
4.6
4.8
4.8
4.8
4.7
4.7
4.7
4.7
4.7
4.6
4.6
4.6
95X CONF.
INTERVAL
3.4- 6.9
4.7- 5.6
3.9- 6.3
4.7- 5.0
4.0- 5.7
4.2- 5.3
4.1- 5.5
1.9- 7.5
4.3- 5.1
4.1- 5.3
3.7- 5.6
3.2- 6.2
3.7- 5.5
3.9- 5.2
3.9- 5.2
X OF HOUSES
ESTIMATED
PERCENT
30.5
46.1
29.9
38.9
40.9
47.4
34.9
33.7
41.6
36.7
27.4
31.2
35.3
37.5
41.7
> 4 PCI/L
95X CONF.
INTERVAL
27.4-33.5
37.7-54.6
24.3-35.6
33.3-44.5
31.9-49.9
40.4-54.4
24.6-45.2
27.2-40.1
36.9-46.3
33.3-40.0
23.3-31.6
23.4-39.0
29.2-41.3
34.0-41.0
31.9-51.5
LIST OF COUNTIES INCLUDED IN THE REGION
BENEWAH, BONNE R, BOUNDARY, KOOTENAI, SHOSHONE.
BOTTINEAU, KIDDER, LOGAN, MCHENRY, PIERCE, ROLETTE,
TOWER.
DAVIDSON.
CLAY, CLOUD, DICKINSON, ELLSWORTH, GEARY, JEHELL, LINCOLN,
MITCHELL, MORRIS, OSBORNE, OTTAWA, REPUBLIC, RILEY,
RUSSELL, SALINE, SMITH, WASHINGTON.
AROOSTOOK.
BURKE, BURLEIGH, DIVIDE, MCINTOSH, MOUNTRAIL, RENVILLE,
SHERIDAN, WARD, WILLIAMS.
MARATHON, PORTAGE.
FRANKLIN, OXFORD, SOMERSET.
COLFAX, HARDING, MORA, RIO ARRIBA, SAN MIGUEL, SANTA FE,
TAOS, UNION.
ALLEN, AUGLAIZE, CRAWFORD, DEFIANCE, DELAWARE, ERIE,
FULTON, HANCOCK, HARDIN, HENRY, HURON, KNOX, LUCAS,
MARION, MERCER, MORROW, OTTAWA, PAULDING, PUTNAM,
RICHLAND, SANDUSKY, SENECA, UNION, VAN WERT, WILLIAMS,
WOOD, WYANDOT.
ALLEGHENY.
BERKELEY, GRANT, GREENBRIER, HAMPSHIRE, HARDY, JEFFERSON,
MERCER, MINERAL, MONROE, MORGAN, PENDLETON, POCAHONTAS,
RANDOLPH, SUMMERS.
ADAMS, BROWN, CASS, CHAMPAIGN, CHRISTIAN, CLARK, COLES,
CUMBERLAND, DE HITT, DOUGLAS, EDGAR, FORD, FULTON,
HANCOCK, HENDERSON, HENRY, KNOX, LOGAN, MACON, MASON,
MCDONOUGH, MCLEAN, MENARD, MERCER, MORGAN, MOULTRIE,
PEORIA, PIATT, PIKE, ROCK ISLAND, SANGAMON, SCHUYLER,
SCOTT, SHELBY, STARK, TAZEWELL, VERMILION, WARREN,
WOOD FORD.
WORCESTER.
BECKER, BELTRAMI, CASS, CLAY, CLEARWATER, CROW WING,
HUBBARD, ITASCA, KITTSON, KOOCHICHING, LAKE OF THE WOODS,
MAHNOMEN, MARSHALL, MORRISON, NORMAN, OTTER TAIL,
PENNINGTON, POLK, RED LAKE, ROSEAU, STEARNS, TODD, WADENA,
WILKIN.
-------
TABLE 2. STATE RADON REGIONAL ESTIMATES AND CONFIDENCE INTERVALS FOR REGIONS HITH ARITHMETIC MEAN VALUES > 4 PCI/L
NO. OF
HOUSES
STATE REGION TESTED
ARITHMETIC MEAN
ESTIMATED 95X CONF.
PCI/L INTERVAL
'/. OF HOUSES > 4 PCI/L
ESTIMATED 9SX CONF.
PERCENT INTERVAL
LIST OF COUNTIES INCLUDED IN THE REGION
NY
KY
HI
AL
KS
TN
IN
NE
PA
ID
IN
MA
MT
4
4
7
6
11
3
5
5
5
2
6
3
108
143
118
156
190
ZOO
448
298
463
249
70
456
391
303
4.6 3.1- 6.0 35.9 25.9-45.9 ALBANY, CONVERSE, GOSHEN.
4.5 3.8-5.2 34.5 28.0-41.0 ANDERSON, BOURBON, BOYLE, BRACKEN, CLARK, FAYETTE,
FRANKLIN, HARRISON, MADISON, MERCER, MONTGOMERY, NICHOLAS,
ROBERTSON, SCOTT, SHELBY, WOODFORD.
4.5
4.4
4.4
4.3
4.2
4.2
4.2
4.2
4.1
4.1
4.1
4.1
3.6- 5.4
0.0-10.1
3.8- 4.9
3.7- 5.0
3.5- 4.9
3.8- 4.6
4.0- 4.5
3.6- 4.8
2.4- 5.7
3.7- 4.6
3.8- 4.5
3.7- 4.6
44.3 38.8-49.9
7.3 2.8-11.8
39.8
29.6
33.7
34.7-44.9
24.0-35.3
27.4-40.0
40.1 33.<»-*6.9
40.8 36.3-45.3
29.4 23.7-35.2
22.9 14.6-31.2
36.2 30.5-41.9
25.2 20.3-30.0
37.6 34.3-40.9
KENOSHA, RACINE, NALNORTH, HAUKESHA.
AUTAUGA, CALHOUN, CHILTON, CLAY, CLEBURNE, COOSA, ELMORE,
RANDOLPH, TALLADEGA, TALLAPOOSA.
CLARK, COMANCHE, EDHARDS, FINNEY, FORD, GRANT, GRAY,
GREELEY, HAMILTON, HASKELL, HODGEMAN, KEARNY, KIOHA, LANE,
MEAOE, MORTON, NESS, PAWNEE, RUSH, SCOTT, SEMARD, STANTON,
STEVENS, WICHITA.
CARTER, GREENE, HANCOCK, HAHKINS, JOHNSON, SULLIVAN,
UNICOI, WASHINGTON.
BLACKFORD, BOONE, BROWN, CLINTON, DEARBORN, DECATUR,
DELAWARE, FAYETTE, FOUNTAIN, FRANKLIN, GRANT, HAMILTON,
HANCOCK, HENDRICKS, HENRY, HOWARD, JAY, JOHNSON, MADISON,
MARION, MONTGOMERY, MORGAN, OHIO, OMEN, PUTNAM, RANDOLPH,
RIPLEY, RUSH, SHELBY, SWITZERLAND, TIPTON, UNION, MARREN,
WAYNE.
ANOKA, DAKOTA, HENNEPIN, RAMSEY, WASHINGTON.
ADAMS, BUFFALO, CLAY, FRANKLIN, GARFIELD, GREELEY, HALL,
HAMILTON, HOWARD, KEARNEY, MERRICK, NANCE, NUCKOLLS,
SHERMAN, VALLEY, WEBSTER, WHEELER.
ARMSTRONG, CAMBRIA, FAYETTE, GREENE, INDIANA, SOMERSET,
WASHINGTON, WESTMORELAND.
BLAINE, CAMAS, CASSIA, GOODING, JEROME, LINCOLN, MINIDOKA,
TWIN FALLS.
ADAMS, ALLEN, CASS, DE KALB, ELKHART, FULTON, HUNTINGTON,
KOSCIUSKO, LAGRANGE, MARSHALL, MIAMI, NOBLE, STEUBEN,
WABASH, WELLS, WHITLEY.
MIDDLESEX.
BIG HORN, BLAINE, CARBON, CARTER, CUSTER, DANIELS, DAWSON,
FALLON, FERGUS, GARFIELD, GOLDEN VALLEY, MCCONE,
MUSSELSHELL, PETROLEUM, PHILLIPS, POWDER RIVER, PRAIRIE,
RICHLAND, ROOSEVELT, ROSEBUD, SHERIDAN, STILLWATER,
TREASURE, VALLEY, WIBAUX, YELLOWSTONE.
-------
one-half of the houses in 21 regions (from seven states) have screening
measurements greater than 4 pCi/L. Figure 3 graphically identifies regions
where the estimated percentage of houses with screening measurements
greater than 4 pCi/L is 1) less than 15%, 2) between 15 and 30%, and
3) greater than 30%. Of the 225 regions, 103 fall into the lowest group,
66 regions fall into the middle group, and 56 regions have at least 30% of
the houses with screening measurement greater than 4 pCi/L. In order to
make a more meaningful identification of "hot spots", the 56 regions in the
upper group are examined in more detail. Figure 4 shows the top 56 regions
broken down into three groups. There are 35 regions where 30-50% of the
houses have screening measurements greater than 4 pCi/L, 15 regions where
50-70% of the houses have screening measurements greater than 4 pCi/L, and
there are six regions where more than 70% of the houses have screening
measurements exceeding 4 pCi/L. These six regions have estimated AMs
greater than 8.8 pCi/L and appear among the first seven regions listed in
Table 2.
CONCLUSIONS
Over a six year period, 38 of the 48 conterminous states completed
statistically-designed surveys to characterize the distribution of indoor
radon. These surveys have produced screening measurements in over 54,000
randomly selected houses. An important element of each state survey was
the inclusion of enough test houses to permit the distribution of radon to
be characterized within geographic regions within each state. As a result,
the 38 states were divided into 225 regions.
This paper examined the screening measurements in these 225 regions for
the purposes of identifying specific locations or patterns of locations of
elevated radon levels. A listing is provided of the 60 regions (and the
counties making up the regions) having arithmetic means exceeding 4 pCi/L.
The first 24 regions listed are particularly important because they make up
the top two groups highlighted in Figure 2 (i.e., arithmetic means over
6 pCi/L). Collectively, these 24 regions may be characterized as having
1) 2.8 million houses which had a positive chance of being tested,
2) an arithmetic mean (of short-term measurements) of 8.9 pCi/L,
3) a geometric mean (of short-term measurements) of 4.9 pCi/L,
4) a median (of short-term measurements) of 5.1 pCi/L, which
translates to 1.4 million houses exceeding 5.1 pCi/L,
5) 1.6 million houses (57.3%) exceeding 4 pCi/L, and
6) 240,000 houses (8.6%) exceeding 20 pCi/L.
-------
Percent >4 pCi/L PCT<15 •Mii 15<-PCT<30
White areas not tested
30< "Pa-
3. Distribution of Percentage of Houses with Screening Measurements >4 pCi/L in 225 Regions
-------
Percent >4 pCi/L
30<>PCT<50 50<=PCT<70 MBBM 70<-PCT
White areas not in top regions or not tested
Figure 4. Distribution of Percentage of Houses with Screening Measurements >4 pCi/L in the Top 56 Regions
-------
REFERENCES
1. Gray, D.J. and Windham, S.T. EERF standard operating procedures for
radon-222 measurement using charcoal canisters. EPA-520/5-87-005, U.S.
Environmental Protection Agency, Washington, DC, 1987.
2. Gray, D.J. and Windham, S.T. NAREL standard operating procedures for
radon-222 measurement using diffusion barrier charcoal canisters. EPA-
520-5/90-032, U.S. Environmental Protection Agency, Washington, DC,
1990.
3. U.S. Environmental Protection Agency, U.S. Department of Health and
Human Services, and U.S. Public Health Service. A citizen's guide to
radon (second edition): the guide to protecting yourself and your
family from radon. EPA 402-K92-001, U.S. Government Printing Office,
Washington, DC, 1992.
4. Shah, B.V. Software for survey data analysis. Am. Stat. 38:68-69,
1984.
5. White, S.B., Bergsten, J.W., Alexander, B.V., and Ronca-Battista, M.
Multi-state surveys of indoor 222Rn. Health Phys. 57(6):891-896, 1989.
6. White, S.B. Bergsten, J.W., Alexander, B.V., Rodman, N.F., and
Phillips, J.L. Indoor 222Rn concentrations in a probability sample of
43,000 houses across 30 states. Health Phys. 62(1):41-50, 1992.
-------
IXP-2
Texas Residential Radon Survey
Charles Johnson, Ph.D., Gilbert Ramirez, Dr.P.H., and
Terry Browning
Southwest Texas State University
San Marcos, Texas 78666
Gary Smith, Ph.D., Paul Breaux, and Vonya Boykin
Texas Department of Health
Austin, Texas 78756-3189
ABSTRACT
Exposure to elevated levels of indoor radon in residential
structures has been suggested by many researchers to pose a public
health risk and be related to a potential increase in the
incidence of lung cancer. To thoroughly examine this issue, the
Texas Department of Health, with the help of an EPA grant,
commissioned a statewide survey of indoor residential radon to
determine the extent of the problem in Texas, and to identify
potential "hot spots." This report examines the radon measurements
from over 2900 randomly selected Texas homes measured during the
winter of 1991. Texas homes, when viewed on a statewide basis,
have a relatively low level of radon, averaging 1.2 pCi/1 (pico
curies per liter). Such levels are not a major public health
concern, as it would be extremely costly and difficult to achieve
lower average residential levels on a statewide basis. This Texas
average is within the national norms, where U.S. homes have been
reported to have average indoor radon levels between 1.0 and 2.0
pCi/1. However, when examined on a county basis, several areas of
Texas are identified where local geology is suspected of
contributing to elevated levels of indoor radon. The Panhandle
area of Texas, especially those counties clustered in a band
through its center, is shown to have the highest potential for
indoor radon. This area of the state is the only area to report
any sizable number of homes with radon over 20 pCi/1.
Correspondingly, it's also the area of the state with the greatest
number of homes measuring over 4 pCi/1. Other areas of the state
with a potential (based on geology) for elevated radon levels
include the Big Bend area (also based on survey results), Llano
Uplift, and the uranium mining areas in South Texas.
Since indoor radon in Texas is a localized problem, efforts
to educate citizens about the potential dangers of radon can be
focused most effectively in those counties with elevated radon
potential. For the most part, the areas of Texas where radon
levels are highest are also areas of lower population density,
minimizing the public health risk and maximizing the potential to
find and correct any threats to the public health.
History of the Radon Question
For many years, radiation scientists and epidemiologists have
noted a strong correlation between exposure to elevated
-------
concentrations of radon (including radon decay products) and
increased risk for lung cancer among underground uranium miners
(5,6,12). More recently, studies by other scientists have found
higher than expected radon concentrations in homes in various
parts of the United States (1,2,4,7,8,11). This has led public
health specialists to a concern that radon exposure within our
homes may be a harmful health risk comparable to that experienced
by many underground miners. The EPA has suggested that indoor
radon is the most serious environmental carcinogen which the EPA
must address for the general public (9). While cigarette smoking
is recognized as the principle agent responsible for lung cancer,
accounting for about 85% of the lung cancer deaths in the United
States; exposure to radon gas has been suggested as a major agent
involved in the remaining fifteen percent (3). In response to the
public concern about the potential harmful effects of radon, the
Environmental Protection Agency (EPA) began a campaign in 1986 to
determine the average radon exposure for homes within the United
States and to encourage all citizens to test their homes for
indoor radon.
In 1986, the Texas Department of Health (TDH) began
distribution of indoor radon information in response to requests
from Texas citizens. In June 1989, the Governor of the State of
Texas designated the Texas Department of Health, Bureau of
Radiation Control (BRC) the lead agency for evaluation and further
analysis of the potential for indoor radon in Texas. Following
that designation, the BRC applied for additional funding from the
EPA to conduct further analysis and testing of indoor radon in
Texas homes. An EPA grant was awarded to TDH in April, 1990, in
part to fund a statewide survey of indoor radon. The results from
that statewide survey of indoor radon are the subject of this
report. The survey was designed to address two questions:
(1) What is the average radon concentration in Texas homes, and
(2) Are there any Texas regions of higher radon potential ("hot
spots")?
The Texas Survey Design
The overall objective of the Texas state survey of indoor
radon is to respond to the public's questions and concerns about
indoor radon exposure. The potential risk associated with long
term exposure to elevated radon concentrations is an increased
risk for lung cancer. A statewide screening survey was designed to
define; (1) the statewide average indoor radon concentration in
homes, and (2) the regional average indoor radon concentrations
(to identify "hot spots") by county and other geographical regions
where appropriate.
The Survey Sample
To simplify survey procedures, only owner-occupied houses,
-------
selected at random, were surveyed in this study. Rental houses
were excluded because of the need to simplify survey procedures
and avoid the problems of gaining permission from house owners to
conduct radon measurements. In addition, high-rise structures,
apartments, and group quarters were excluded from the study in
order to create a uniform sampling population. These exclusions
did not materially affect the statistical basis of the survey, and
had the advantage of making a complex undertaking feasible in a
reasonable amount of time.
Survey Method
Indoor radon measurements were made with charcoal canisters
(CO supplied by the EPA. Measurements were made under
closed-house conditions using EPA screening survey protocols. In
addition, long term alpha track detectors (ATD) are being used to
measure twelve-month radon levels for a subset of the survey
sample. Both the canisters and ATD detectors are returned to the
EPA, which is responsible for the analysis of radon levels.
Regional Sampling Plan
A regional sampling plan, based largely on geological
potential for indoor radon, was produced to aid in the random
allocation of radon detectors, and to help provide a basis for
interpretation of survey results. Survey staff examined available
geological and population data for the state of Texas and produced
a map which grouped all Texas counties with respect to their
potential for indoor radon. Contiguous counties with a similar
potential for radon were grouped into regions. Large metropolitan
areas were designated as regions to control their dominance of the
sample due to larger populations. Figure 1 shows the regional
sampling areas developed in consultation between the EPA, Research
Triangle Institute, and the Texas researchers.
Telephone lists were generated randomly, with all residents
within a defined region having an equal chance of being chosen.
With the total number of homes needed for the survey known, it
then became a matter of choosing a proportional sampling plan to
insure that the large metropolitan areas did not dominate the
survey. Each wave of fifty names contained residences from
throughout the state in proportion to the percentages for the
regions designated by the sampling plan. The large metropolitan
areas were sampled at a lower percentage to insure that rural
areas would have adequate numbers in the survey. Even with 13
sampling regions (more than any other state), we may have had to
include some sparsely populated counties in a region with more
populous counties; leading to the probability of some counties not
being sampled. Financial constraints prevented more regions from
being used in the sampling plan.
Home owners were contacted by telephone in each region on a
random basis at a frequency determined by the county's radon
-------
potential and population density. The results of the indoor radon
survey were analyzed and reported at the county level in order to
better understand identified "hot spots."
Figure 1. Regional Sampling Plan
1 - Southwest Texas
2 - El Paso
3 Big Bend 10
4 - West Texas Shales 11
5 - North Texas 12
6 - Dallas/ Fort Worth 13
7 - East Texas
Llano Uplift
• Central Texas (Austin - San Antonio)
- Tertiary Sands Cresent
Harris County (Houston)
- Gulf Coast
- Texas Panhandle
-------
Sample Size and Allocation of Detectors
Survey staff initially placed over 4000 charcoal canisters
throughout the state for the measurement of indoor radon, and over
300 Alpha Track Detector's (ATD's) for the determination of radon
levels for a twelve-month period. Those homes receiving the twelve-
month ATD measurement devices will also receive a charcoal
canister for each climatic season, or a total of four charcoal
canisters for the year. The routine placement of charcoal
canisters was accomplished through the use of "waves" or lists of
50 telephone numbers (with names and addresses) supplied by the
Research Triangle Institute under contract to the EPA. In order to
avoid introducing statistical bias, all telephone numbers in each
opened wave were called in a search for working numbers and
eligible participants. Initial refusals to take part in the study
were called again, at different times in hopes of better
explaining the importance of the study and gaining the homeowner's
cooperation. Up to six tries were made to convince home owners of
the importance of participation in the survey.
Although the placement of detectors was accomplished by using
a random telephone list, the actual number of canisters placed in
any given geographical region depended on the regional potential
for radon and the population density. To insure samples in rural
areas of the state and identify these rural "hot spots," it was
necessary to sample at a higher density in these rural regions
which may have a higher geological potential for radon, but a
lower population density; and at a lower frequency in densely
populated metropolitan areas. Because of this objective to
identify rural "hot spots," the statewide sample cannot be
considered a true proportional sample of the state, nor truly
representative of the state as a whole. Therefore, since the rural
areas of the state were sampled in greater numbers dis-
proportionately to the metropolitan areas, the statewide
percentages cannot simply be multiplied to determine the number of
homes in the state with elevated indoor radon. Percentages
measured within individual counties, however, can be used as
determinants of the potential radon problem for that county, since
all residents within a specific county had an equal chance of
being chosen for the survey. Readers should be cautioned however,
that counties with fewer than five measurements are still
tentative at best.
Quality Assurance/Quality Control
Precision for radon analysis was established by placement and
analysis of duplicate samples in survey staff member's homes.
Accuracy estimates were provided by the EPA's Montgomery, Alabama
laboratory as part of their routine QA/QC program. Blank samples
were submitted for radon spiking and analysis at the rate of 2
percent of the total canisters placed, and were selected from
throughout all canister shipments.
-------
Staffing for the Survey
The Texas Department of Health contracted with health
researchers in the Department of Health Administration at
Southwest Texas State University. Working from a central location
on the SWT campus near Austin (the location of the Department of
Health headquarters), part-time graduate and under-graduate
students operated the telephone banks, which were the heart of the
survey. On-site supervision was provided by SWT faculty; and
project oversight was the responsibility of TDK.
SWT health researchers presented a proposal based on the
placement of up to a maximum of 4500 CC's and 450 ATD's. TDH staff
evaluated the proposal and negotiated an interagency contract with
SWT. Personnel from both SWT and TDH received appropriate radon
survey training from the EPA and Research Triangle Institute. EPA
guidelines for a random statewide survey were followed. Potential
survey participants were first contacted by mail, which was
followed by telephone interviews to confirm eligibility. SWT staff
mailed out radon detectors to eligible participants and confirmed
their use. Under TDH agreement, SWT received copies of all
analytical reports from EPA (TDK received duplicates), encoded
them in a computer database, and mailed results to all survey
participants.
Standard Forms
Standard form letters were used to announce the house survey
to potential participants and to return radon analytical results
to participating home owners. EPA-supplied questionnaires were
used to gather demographic and other relevant information from
eligible participants. These data were entered into machine
readable form by the staff of the Research Triangle Institute,
under contract to the EPA, and are due to be returned to the SWT
health researchers for further analysis.
Data Management and Analysis
Texas Department of Health worked with SWT and the EPA to
develop an information storage and management system for all radon
related data. Specifically, relevant questionnaire responses and
radon analytical results were stored in a computer database.
Radon analyses were done by EPA's National Air and Radiation
Environmental Laboratory in Alabama and finished data sets were
returned to TDH and SWT.
Results and Discussion of the Texas Survey
Charcoal canisters were mailed to over 4000 homeowners
throughout Texas. Of these canisters, 2692 valid winter
measurements were returned. The attrition in the sample was due to
a variety of reasons, including some homeowners who delayed
-------
conducting the tests until past the winter season, home owners
deciding to not take part after agreeing to the test, or canisters
being lost in the mail. Results from those homeowners who delayed
their home measurements and results delayed for other reasons will
be addressed in a future report.
Statewide Results
When examined statewide, Texas has a low level of indoor
radon in homes, with an arithmetic mean of 1.2 pCi/1 reported.
The percent of Texas homes tested during this survey with a radon
level above 4.0 pCi/1 (the threshold of concern by EPA definition)
is 4.7 percent (Table 1). Furthermore, 0.2 percent of the Texas
homes tested in this survey have a radon level above 20 pCi/1.
These results should be examined carefully because of the decision
not to use exact proportional sampling.
Disproportionate sampling was used to insure that less
densely populated areas of the state (such as West Texas) were
over sampled; so that potential hot spots would not be missed.
Many of these potential hot spots were in rural areas and called
for disproportionate sampling because of the low population
density and widely separated areas between towns or homes. Because
of the sampling strategy used, the survey findings, as reported in
this report, cannot be generalized to make a statement about all
homes in Texas. These preliminary findings are useful, however,
to help identify which geographical regions in the state have a
higher potential and concern for indoor radon.
Arith.
Mean
pCi/1
1.2
Geo.
Mean
oCi/1
0.6
Table
Median
pCi/1
0.7
1. Texas Statewide Results
75th
Percentile
DCl/1
1 .4
90th
Percentile
pCi/1
2.7
% Houses %Houses
>4 oCi/1 >20 DCi/1
4.7 0.2
County Results
When statewide data were analyzed on a county basis,the
geological potential for radon becomes more evident.The sampling
plan used in the survey was a balance between geological potential
for radon and population density factors.In many cases, counties
in Texas were grouped into larger regions identified by their pre-
survey geological potential for radon, and weighted dis-
proportionately according to their population density.Within these
defined larger groups of counties, all homes had an equal chance
of being selected for participation in the survey.In some cases,
more heavily populated counties within these regions overshadowed
-------
the less populated counties, causing some counties to have low
numbers or no homes participating in the survey. Due to the large
number of Texas counties, it was necessary to group them into
these larger regions for sampling cost considerations. In those
counties where measurements were not made as part of the random
sample reported in this report, additional canisters were placed
on a volunteer basis, and will be reported and discussed in a
future report.
In general, the Texas counties which have higher potential
for residential radon are found in the West Texas Panhandle
region; the Big Bend area; the Llano Uplift area; and inland from
the coastal bend in South Texas where underground formations of
ancient Mesozoic beach sands, rich in uranium, can be found. All
the counties where higher levels of radon were found have geology
which supports their higher potential. Valid radon measurements
were collected from 229 out of the 254 Texas counties.
Counties where a calculated average level of radon exceeds
the 4.0 pCi/1 threshold level of concern as identified by the
Environmental Protection Agency include Carson (1 of 4
measurements above 4.0 pCi/1), Hale (7 of 14 measurements above
4.0 pCi/1), Randall (7 of 20 measurements above 4.0 pCi/1),
Sherman (4 of 4 measurements above 4.0 pCi/1), and Swisher (2 of 5
measurements above 4.0 pCi/1). All of these counties are found in
the Central Panhandle region of Texas.
Examination of the Texas counties map (Figure 2) for
percentage of homes with radon measurements above 4.0 pCi/1
clearly shows a greater potential for elevated indoor radon in the
Texas Panhandle region.
The counties of Jeff Davis (3 of 16 measurements above 4.0
pCi/1), Presidio (8 of 44 measurements above 4.0 pCi/1), and
Brewster (10 of 60 measurements above 4.0 pCi/1) are all found in
the Texas Big Bend region, and have subsurface geology which
support a higher potential for indoor radon. The counties of Mason
(2 of 19 measurements above 4.0 pCi/1) and Llano (7 of 47
measurements above 4.0 pCi/1) are both in the Llano Uplift region,
and also have local geology supportive of radon production.
-------
Figure 2 Texas Counties Percent of Homes > 4.0 pCi/1
SSSS
ssss
s/ss
Less than 10 %
10.1 to 20 %
20.1 to 50 %
More than 50 %
One or no valid measurements
Other counties showing elevated potential for indoor radon
have only a few measurements, but they are reported here since they
have surface and subsurface geology which could provide a source
for radon, and there is the likelihood that some homes in these
counties have elevated radon levels. Because of these geological
sources and the likelihood that some homes in these counties have
elevated radon levels, the counties are shown on the map. However,
because of the small number of measurements, they should be taken
as inconclusive without further and more numerous measurements.
These counties include the Coastal Bend and South Texas counties of
La Salle (1 of 2} measurements above 4.0 pCi/1), Karnes (I of 3
-------
measurements above 4.0 pCi/1), Victoria (1 of 8 measurements above
Figure 3. Texas Counties Percent of Homes > 20 pCi/1
.
SSfS.
SSSS*
ssss.
ssss
Less than 10 %
10.1 to 20 %
20.1 to 50 %
One or no valid measurements
-------
4.0 pCi/1), and Lavaca (1 of 8 measurements above 4.0 pCi/1). In
each of these counties, only one measurement was found above 4.0
pCi/1. Further volunteer placements will be reported in a future
report in an effort to better define the extent of indoor radon for
these counties.
Three additional Texas counties of Bastrop (1 of 10
measurements above 4.0 pCi/1), Brown (2 of 6 measurements above 4.0
pCi/1), and Taylor (3 of 26 measurements above 4.0 pCi/1) had at
least 10 percent of their measurements above 4.0 pCi/1. Examination
of Texas county map for the percentage of homes over 20 pCi/1
(Figure 3) shows only two counties with potentially more than 10
percent of the homes at this level. Hale (2 of 14 measurements
above 20 pCi/1) and Carson (1 of 4 measurements above 20 pCi/1) are
both in the earlier identified Texas Panhandle region of higher
radon potential.
Conclusions and Recommendations
The measurements taken in this survey followed EPA guidelines
for closed home testing during winter months, thus leading to a
higher than normal accumulation of indoor radon. More
realistically, the true annual radon levels and occupant exposures
for these Texas homes would be lower than measured. The numbers
reported here represent the higher radon potential for these homes,
not necessarily the radon levels experienced by the home owners on
an annual basis. It should be remembered that the EPA guidelines
for health risk are based on a 70-year lifetime exposure to an
average radon concentration, and a 75 percent house occupancy rate.
Less exposure would result in a lower estimate of lung cancer risk.
Given that the survey protocol for the data reported in this report
required that the canisters be placed during the winter months (in
closed-house conditions), and that in all likelihood average indoor
radon levels on an annual basis would be up to 25 percent less, as
reported by Rood, et al (10), the annual risk levels for the homes
measured in this survey are less than the initial data would
suggest.
For this preliminary report, geology is the only factor
examined in relation to indoor radon levels. A future report will
examine other factors such as type of dwelling or presence of a
crawl space. The Panhandle High Plains area of Texas has the
highest potential for indoor radon, with some counties such as
Hale, Swisher, Randall, Carson, and Sherman reporting at least 25
percent (or higher) of the homes with indoor radon levels above 4.0
pCi/1. These findings are consistent with surface and subsurface
Permian shales with higher potential for uranium.
Another part of the state with elevated potential for indoor
radon is the Big Bend area. The three counties of Brewster, Jeff
Davis, and Presidio all had at least 15 percent of the homes
measured with radon levels above 4.0 pCi/1. Radon in these counties
most likely results from uranium in Tertiary igneous intrusions and
-------
shales from Paleozoic periods. The Davis Mountains are noted for
their Tertiary basaltic lavas and volcanic mud flows. Numerous
faults and fractures are present in this region, having contributed
to the upward movement of molten rock and volcanic vents. These
faults and fractures act as conduits for upward migration of deeper
sources of radon.
Other parts of the state with elevated potential for indoor
radon include counties above a crescent of Tertiary sands which
parallels the Texas Gulf coast. This crescent extends from deep in
South Texas over toward the Texas-Louisiana border. Fortunately,
most of the counties did not exhibit elevated measurements for
indoor radon. Counties above these sands, which did show some
elevated potential, include La Salle and Karnes. Karnes county has
been the site of commercial uranium mines for years. For a very
limited number of measurements, other counties in this region show
a small potential for elevated indoor radon. Lavaca and Victoria,
both, report one out of eight canisters above 4.0 pCi/1.
Overall, Texas does not have as great a problem with elevated
indoor radon as other states in the Midwest and Northeast United
States. The fact that some areas of Texas report indoor radon
levels higher than other areas is justification for further study
and analysis. The goals of this survey were to establish a
statewide radon average and to identify areas of the state which
may have potential "hot spots." Those goals have been accomplished.
Further research and surveys should now concentrate in the areas of
the state where elevated indoor radon was reported.
References
I.Cohen, B. L., and R. S. Shaw, "Mean radon levels in US homes by
states and counties," Health Physics, 1991, 60, 243-259.
2.Cothern, C. R.. 199G. "Indoor air radon," Review of Environmental
Contaminants and Toxicology, 111, 1-60.
3.Council on Scientific Affairs, American Medical Association,
"Health Effects of Radon Exposure," Archives of Internal Medicine,
1991, 151, 4, 674-677.
4.Ganas, Michael J., "Radon Contamination in Dwellings." In
International Journal of Environmental Studies, 1989, Vol. 32, pp.
247-260.
S.Harley, N. H., and J. H. Kariey, "Potential lung cancer risk from
indoor radon exposure," CA, 1990, 40, 5, 265-275.
6.National Research Council. 1988. Health Risks of Radon and Other
Internally Deposited Alpha-Emitters, BEIR IV. National Academy
Press, Washington,
D. C.
-------
7.National Research Council. 1991. Comparative Dosimetry of Radon
In Mines and Homes, National Academy Press, Washington, D. C.
S.Nero, A. V., M. B. Schwehr, W. W. Nazaroff, and K. L. Revzan.
1986. "Distribution of Airborn Radon-222 Concentrations in U.S.
Homes," Science, 234, 992-997.
9.Puskin, Jerome S., and Christopher B. Nelson. 1989. "EPA's
Perspective on Risks from Residential Radon Exposure," JAPCA, 39,
7, 915-920.
10.Rood, A. S., J. L. George, M. D. Pearson, and G. H. Langner,
Jr., "Year-to-year variations in annual average indoor 222Rn
concentrations," Health Physics, 1991, 61, 3, 409-413.
11.U.S. Environmental Protection Agency. 1988. Radon Reduction
Techniques for Detached Houses.. Office of Research and
Development, Washington, D.C., EPA/625/5-87/019, January 1988.
12.Woodward, A., D. Roder, A. J. McMichael, P. Crough, and A.
Mylvaganam, "Radon daughter exposures at the Radium Hill uranium
mine and lung cancer rates among former workers, 1952-87," Cancer
Causes Control, 1991, 2, 1, 213-220.
-------
IXP-3
ABSTRACT
Radon Survey of Oregon Public Schools
by
R.D. Paris and G.L. Toombs
Oregon Health Division
To assist in addressing the concerns of potential elevated radon levels in
public schools nationwide, the Oregon Health Division conducted a limited
study to determine the average radon levels in the schools in Oregon.
Thirty-one schools were selected at random out of a population of 1,190
statewide to participate in this study during the 1990-1991 school year.
Long-term alpha track detectors were placed in each of the ground-floor and
basement classrooms to obtain the average radon levels this entire nine-
month school year.
The results of this study showed that the mean radon concentrations in
Oregon public schools statewide was 1.1 pCi/liter. This compares to a mean
of 1.2 pCi/liter for indoor radon in homes in the state. Two of the schools
surveyed had many rooms above the 4 pCi/liter EPA guidelines.
Follow-up with these schools and their options for lowering the levels in the
elevated rooms are discussed.
-------
IXP-4
QUALITY ASSURANCE IN RADON SURVEYS
by: William M. Yeager, Robert M. Lucas, and Jane W. Bergsten
Research Triangle Institute
Research Triangle Park, NC 27709
Frank Marcinowski, and Jeffrey Phillips
Office of Radiation Programs
U.S. Environmental Protection Agency
Washington, DC 20460
ABSTRACT
This paper describes a variety of quality assurance (QA) activities that
were performed throughout the planning and execution of three U.S.
Environmental Protection Agency (EPA)-sponsored radon surveys -- the National
Residential Radon Survey (NRRS), the National School Radon Survey (NSRS), and
the State Residential Radon Surveys (SRRS) — to assure that data would be of
known quality and would satisfy survey objectives. These activities included
the preparation of comprehensive QA plans, training programs for interviewers,
performance evaluations of radon detectors, technical systems audits of
detector processing facilities and of Research Triangle Institute (RTI) data
entry, and audits of data quality. The use of spiked, blank, and duplicate
detectors is described. As a result of these activities, several problems
were identified and resolved, and the performance of the detectors over the
range of concentrations encountered in the surveys was characterized.
The final precision and bias of the radon measurements will be of interest
to those who wish to utilize the results of the NRRS, the NSRS, and the SRRS.
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.
-------
1.0 INTRODUCTION
This paper describes a variety of quality assurance activities that were
performed throughout the planning and execution of three U.S. Environmental
Protection Agency (EPA)-sponsored radon surveys. These activities were
designed to provide confidence that the surveys would satisfy identified
goals. Integrating quality assurance into all phases of these projects
allowed us to recognize and resolve problems as well as to document and defend
the quality of the data produced.
1.1 QUALITY ASSURANCE/QUALITY CONTROL
Quality assurance (QA) activities include
• Defining appropriate measures of quality,
• Setting objectives for those measures,
• Specifying operating procedures capable of meeting those objectives,
• Assessing actual performance,
• Improving performance, and
• Documenting
what was planned,
what was actually done,
how it was done, and
how well it was done.
Quality control (QC) activities include
• Verifying that the specified procedures are being followed, and
• Monitoring the measures of quality to confirm that the QA objectives
are being met.
1.2 IMPORTANCE OF QA
Many people believe that because they are competent and committed to doing
good work, they do not need a formal QA program. Consequently, they may
resent the time and resources required to develop and implement a QA program
for their work. A good QA program, however, is an investment that will pay
off in several ways. It is a useful complement to project planning and test
design. The earlier in the course of a project that one asks, -What can go
wrong with this procedure? How can I recognize a problem if it occurs? How
could I change the procedure to eliminate this possibility or make it easier
to recognize?", the more likely the project will produce useful data. QA
makes it easier to recognize and resolve the problems that inevitably occur.
QA also provides documentation about the procedures used and the quality of
the data produced that is valuable for defending the results. This documen-
tation may be helpful for training new staff, clarifying planning with the
client, etc.
-------
1.3 MEASUREMENT METHODS
Two types of radon detectors were used in these surveys. Short-term
measurements (2 to 7 days) were made with EPA-supplied charcoal canisters,
which were processed by EPA laboratories. Long-term measurements (5 to 12
months) were made with alpha track detectors (ATDs), which were purchased from
and processed by commercial laboratories.
1.4 EPA-SPONSORED RADON SURVEYS
1.4.1 State Residential Radon Surveys (SRRS)
The SRRS measured radon in 42 states and 6 Indian lands between 1986 and
1992. These surveys had two goals: 1) to estimate the statewide frequency
distribution of residential screening measurements, and 2) to determine high
risk areas within states. These were really 48 separate surveys. The EPA
provided the radon detectors, the reading of those detectors, technical
assistance in the areas of geology and physics, and assistance with the survey
research methodology through a contractor. Research Triangle Institute (RTI).
RTI designed and selected the sample of houses, developed the forms and
data collection procedures, trained the interviewers (who were provided and
paid by the states), processed the survey forms, calculated the sampling
weights, and analyzed the data. In addition to providing the telephone
interviewers, the states solicited the cooperation of households selected into
the sample, distributed related literature and the radon detectors to partici-
pants, and reported the results of the measurements to the participants. The
states mailed each participant a charcoal canister and instructions for
placing it according to the EPA protocols for screening measurements. The
canister was placed on the lowest livable level of the house during the
heating season. During the first four years of the SRRS, canisters were
deployed for two days. During the last two years, canisters were deployed for
seven days. After the first year of the survey, approximately 10 percent of
the participating houses were selected to receive alpha track detectors (ATDs)
in addition to the charcoal canister; one ATD was placed on each level of the
house and remained there for 12 months.
The target population of the SRRS was owner-occupied houses with 1) a
permanent foundation, 2) at least one floor at or below ground level (this
includes houses over a crawlspace), and 3) a telephone number published in the
latest directory. This survey did not include rental units, mobile homes, or
upper-floor condominiums.
1.4.2 National Residential Radon Survey (NRRS)
The NRRS measured annual average radon concentrations in a probability
sample of residences around the U.S. The goal of the survey was to estimate
the frequency distribution of annual average radon concentrations in occupied
housing units. Two explicit precision constraints were defined as part of the
study objectives: 1) the national estimated fraction of residences with
concentrations > 10 pCi/L should have a standard error Ł 50%, if the fraction
-------
were > 0.005, and 2) for an EPA Region, the estimated fraction of residences
with concentrations > 4 pCi/L should have a standard error < 50%, if the
fraction were > 0.07. A probability sample of 11,400 homes was selected. An
RTI field interviewer visited each home to determine eligibility and solicit
participation. Housing construction, heating, ventilating, and air
conditioning characteristics were determined during on-site interviews with
the residents. ATDs were placed in 7118 households (84% of the 8444 survey-
eligible households) in the summer of 1989 by the residents, with guidance
from the RTI field interviewers. Detectors were retrieved from 5694 of these
homes (89% of the 6419 households that remained eligible during the entire
monitoring period) after approximately 12 months. Only detectors exposed for
11 to 13 months were used in the statistical analysis.
The target population of the NRRS was housing units (including rental
units) and their permanent residents in all 50 states and the District of
Columbia. To avoid including vacation homes, a housing unit had to be 'lived
in' at least nine months during the year to be survey-eligible. Also excluded
from the NRRS were all places of residence on military bases and all institu-
tional residences (e.g., prisons, nursing homes). Note that this survey
included housing units without ground contact, such as upper-floor apartments.
1.4.3 National School Radon Survey (NSRS)
The NSRS measured radon concentrations in a probability sample of schools
around the U.S. The goals of the survey were to determine 1) the frequency
distribution of radon measurements in schools nationwide, as well as in
schools located in EPA-identified high risk areas; 2) the relationship between
short-term and long-term measurements in schools; 3) the relationship between
ground-floor and upper-floor radon measurements; and 4) the correlation
between radon levels and construction or ventilation characteristics. Short-
term radon concentrations were measured in 928 schools using charcoal canis-
ters provided and processed by EPA. Canisters were shipped to participating
schools, placed by a school contact person, exposed for 7 days, and shipped
via overnight express to an EPA laboratory for processing. Additional long-
term measurements were made in 101 of these schools using ATDs that were
placed by RTI field interviewers in December 1990 and retrieved in late May
and early June, 1991.
The target population for the NSRS consisted of all public schools located
in any of the 50 states or the District of Columbia, and with one or more of
grades K through 12. For each survey-eligible school, all school buildings
containing one or more survey-eligible rooms were included in the survey
population. Most occupied school rooms (e.g., classrooms, offices, cafe-
terias, gymnasiums) were considered survey-eligible. Three types of rooms,
however, were specifically excluded: 1) "wet' areas (e.g., lavatories,
showers, kitchens), 2) areas used strictly for passage (i.e., hallways), and
3) areas used strictly for storage. Survey-eligible rooms were further
classified according to whether or not they had ground contact. Ground-
contact rooms either had a portion of the floor or a wall directly contacting
the ground, or were separated from the ground by a crawlspace that was com-
pletely enclosed. Both canisters and ATDs were placed in all survey-eligible
-------
rooms of the 101 "canister-ATD" schools. In the 827 "canister-only" schools,
canisters were only placed in survery-eligible rooms with ground contact.
1.5 ORGANIZATION OF THIS PAPER
This paper discusses the QA activities for these three radon surveys.
Section 2.0 reviews the importance of planning for quality assurance early in
the project. Section 3.0 discusses a variety of activities that supported
data collection. Section 4.0 defines data quality indicators for these
surveys. Section 5.0 describes the QA audits performed during the surveys.
Section 6.0 presents the data quality indicators for the radon measurements.
Section 7.0 summarizes our QA experience with these surveys.
2.0 QUALITY ASSURANCE PROJECT PLANS
Quality assurance should be integrated into all phases of a project,
including the initial planning. Recognizing this, the U.S. EPA requires a
Quality Assurance Project Plan (QAPjP) for all projects that make environ-
mental measurements. This planning document identifies the QA requirements of
a particular project, describes how those requirements will be satisfied, and
specifies how the quality will be documented. It should also describe the
planning, monitoring, assessing, and documenting of the quality of data from
the project. A QAPjP was prepared for each of these surveys by the QA
coordinator and senior project staff. '2f3
2.1 PLANNING
The QAPjPs included a description of the objectives, design, and schedule
for each survey. Scheduling can have a critical effect on the quality of
radon surveys. Two of these surveys required that measurements be made during
the heating season. This meant that schedules could not slip a month or two
without severely compromising the goals of the surveys. Key activities, the
QA components of those activities, the personnel responsible, and the lines of
authority among those personnel were identified. The QAPjPs also defined the
target populations and described the sample designs.
Goals for data quality were specified. This involved identifying critical
measurements and documentation (e.g., items on the questionnaires), specifying
critical concentrations, describing the measurement methods, defining indica-
tors of data quality (e.g., completeness, precision, bias) and specifying
goals for those indicators. For each survey, the goals for data quality
indicators were related to the overall goals of that specific survey.
Procedures for collecting and documenting both questionnaires and radon
measurements were described. An important part of any large survey is
tracking the questionnaires, samples/detectors, and associated documents in
the field, office, laboratory, and archive. Procedures for documentation and
data processing were described. This included the editing of the question-
naires, statistical analyses, and reporting of the measures of quality.
Laboratory procedures for processing the radon detectors were also described.
-------
2.2 MONITORING
The QAPjPs described procedures and frequency for calibration of the radon
detectors by the processing laboratories. QC checks performed by RTI were
also described. These checks included verifying that the specified procedures
for sampling, documenting, and detector processing were followed. An impor-
tant part of monitoring data quality was designing and implementing sets of
spiked, blank, and collocated (duplicate) radon detectors to collect data on
the quality of the radon measurements. This is discussed in section 4.0.
2.3 ASSESSING
Each QAPjP described how to assess the quality of data and included
procedures to calculate the data quality indicators. Three types of audits
were described: 1) technical systems audits, 2) performance evaluation
audits, and 3) audits of data quality. These will be discussed in more detail
in section 5.0.
2.4 DOCUMENTING
The QAPjPs described how data quality would be documented. Written
reports on all QA audits were prepared. Circumstances that would trigger
corrective action were specified, and QA reports to management were outlined.
These QAPjPs also included the QA plans of the detector processing labora-
tories as appendices.
3.0 SUPPORT OF DATA COLLECTION
3.1 DETECTOR PLACEMENT GUIDELINES
Detector placement guidelines were developed for each survey. These were
consistent with the appropriate EPA guidance,4 but tailored to the objectives
of the survey. In the SRRS, one charcoal canister was placed on the lowest
livable level of each participating house. In the NRRS, an ATD was placed on
each level of the home with a minimum of two ATDs at different locations in
single-level homes. In the NSRS, one radon detector was place in each survey-
eligible room with ground contact. Detectors were not placed where they would
be subject to drafts (e.g., near windows, doors, air vents, or in hallways) or
to high levels of moisture (e.g., kitchens, bathrooms, school locker rooms).
They were to be exposed to air that people would breathe, and not placed in
cupboards, closets, or drawers.
3.2 TRAINING OF FIELD PERSONNEL
Field personnel (RTI interviewers, SRRS state coordinators, SRRS state
telephone interviewers) were trained to use the survey instruments and
procedures for each survey. This training included basic information about
radon and the radon detectors used in that survey, as well as a review of the
detector placement guidelines.
-------
For the SRRS, RTI developed survey forms and data collection procedures.
In addition to presenting a detailed description of the SRRS procedures at the
annual orientation program for the state coordinators, RTI conducted in-state
training sessions for the telephone interviewers and clerks that were hired by
each state. RTI also edited the completed survey forms as they were received
from the states and provided immediate feedback to the state coordinators so
that any deviation from prescribed data collection procedures could be
corrected immediately.
The NRRS involved administration of a complex questionnaire to capture
housing unit construction, as well as heating, ventilation, and air condi-
tioning characteristics; occupant demographic data; and the amount of time
each occupant spent at different levels of their home. Radon detectors were
also placed in each home using a uniform protocol. More than 150 field staff
were thoroughly trained and tested in the proper administration of the ques-
tionnaire, radon detector placement protocols, and other survey procedures.
The field staff were provided with training materials and tested before the
training sessions to identify any specific areas that needed emphasis. Upon
completion of a 2-day training session, the field staff were tested again.
Only field staff who tested satisfactorily were allowed to work on the survey.
Radon detectors were deployed by school or school district personnel in
about 90% of the NSRS sample schools and by trained RTI interviewers in the
remaining 10%. A training videotape was prepared by SC&A, Inc., to demon-
strate how to deploy the charcoal canisters in school rooms. Detailed written
instructions were prepared to supplement the videotape, and both were sent to
participating school districts and to the RTI interviewers. Most of the field
interviewers for the NSRS had worked on the NRRS the previous year. RTI pre-
pared detailed written instructions for placing ATDs in school rooms. In
addition, RTI survey specialists conducted interviewer training sessions by
telephone.
3.3 PANEL MAINTENANCE
Panel maintenance involved recontacting survey participants with long-term
(ATD) radon detectors periodically during the exposure period. They were
asked if there had been any problem with the detector (e.g., it disappeared,
it fell down and the dog chewed it) or if their survey-eligible status had
changed (e.g., if they were planning to move). In addition to providing
information about the detectors and the participants, panel maintenance
reinforces the participants' impression of the importance of the survey.
Starting in year 2 of the SRRS, in most of the state surveys a 10%
subsample of homes was designated to receive ATDs in addition to charcoal
canisters. In these subsamples, two to four ATDs were placed in each
participating home for a 1-year period. States were encouraged to recontact
these participants periodically during the year, but little panel maintenance
actually occurred in the year 2 surveys. Beginning in year 3, seasonal short-
term charcoal canister measurements, as well as 1-year ATD measurements, were
made in the homes in the ATD subsamples. Both ATDs and canisters were
deployed in the winter and additional canisters were sent to each home during
-------
the spring, summer, and fall. Thus, homes in the ATD subsample were contacted
approximately every three months, ending one year after the initial ATD and
canister deployment with a request to return the exposed ATDs.
In the NRRS, every participant was recontacted at roughly 3-month
intervals. For each recontact, the participant was mailed a letter that
contained a postage-paid card, which they returned to RTI verifying the status
of their detectors. Participants who did not return these postcards were
contacted by telephone.
For the NSRS, field interviewers placed ATDs during December 1990 in a
sample of 101 schools and retrieved them 5 to 6 months later. Although no
special panel maintenance program was developed, about 4 to 6 weeks after
placing the ATDs the interviewers returned to the schools to deploy charcoal
canisters for a 7-day period. The interviewers were instructed to check the
status of each ATD and to replace any that had been damaged or removed.
4.0 DATA QUALITY INDICATORS
Current EPA guidance5 requires that projects address the five data quality
indicators (DQIs): precision, accuracy (including both systematic and random
error), completeness, representativeness, and comparability. The first three
DQIs are quantitative; the last two, qualitative. The QAPjPs specified goals
for the quantitative DQIs, which were related to the specific objectives of
each survey.
Precision is the random error of the measurement process and can be deter-
mined by replicate measurements. The precision of field measurements is best
assessed using collocated (duplicate) detectors. In both the NRRS and the
NSRS, 5% of the homes or schools were designated to receive collocated
detectors. Two detectors were placed side-by-side at each measurement
location. The reported values of each collocated pair were compared to
determine the precision of the detectors for several ranges of radon
concentrations (see Section 6.0). During the first year of the SRRS, a
subsample of participants received two detectors to be collocated.
Unfortunately, many of the participants did not appreciate the purpose of
collocated detectors, so they placed them in different locations or not
simultaneously. In subsequent years of the SRRS, only people working on the
survey, who had attended training sessions where the importance of collocated
measurements could be explained, received two canisters.
Bias is the systematic error of the measurement process and can be deter-
mined by blind analysis of spiked detectors. For each survey, RTI arranged
for some detectors purchased for that survey to be reserved for quality
control. Some of these received known exposures at EPA radon chambers and
were then mixed in with field-exposed detectors and sent to the processing
laboratories. The chamber exposures were designed to span the range of
exposures expected in the field. In each survey, measurements near 4 pCi/L
were considered especially important, so some detectors were exposed at this
-------
level. The reported values for the spiked detectors were divided by the
monitored values in the radon chamber to calculate their recovery; an ideal
detector would have a recovery of 1.0.
Accuracy is a measure of the total error, both systematic and random, in a
measurement. When the QAPjPs for these surveys were prepared, the EPA's
Radon Measurement Proficiency Program was using the mean absolute relative
error (MARE) to assess the accuracy of commercially available radon
detectors.0 RTI specified an objective for the MARE in each QAPJP.
The method detection limit (MDL) is the smallest value that can be
reliably distinguished from zero. It is related to the standard deviation of
detectors exposed at or near zero concentration. In each survey, unexposed
(blank) detectors were mixed with the field-exposed detectors sent to the
processing laboratories. The standard deviation of the collocated detectors
in the lowest exposure range was another indicator of the MDL.
Completeness is a measure of the number of valid data points obtained
versus the number that were planned to meet the established statistical level
of confidence for the project. The precision of a survey estimate depends on
the completeness of the survey. For example, one of the objectives of the
NRRS was to estimate the fraction of U.S. homes with annual average radon
concentrations > 10 pCi/L with a relative standard error of < 0.50. RTI
projected that a final sample size of 5000 homes would be necessary to meet
this objective. RTI then projected the fraction of addresses that would
produce survey-eligible homes, the fraction of these homes that would be
willing to participate in the NRRS, and the fraction of residents that would
move during the course of the survey. These estimates led to an original
sample size of 11,400 homes. The actual fractions were carefully monitored
during tbe time when field interviewers were contacting prospective
participants and during panel maintenance.
Representativeness is the degree to which data accurately and precisely
represent the characteristic of a population. Each of these surveys used
measurements in carefully designed samples of the target population to satisfy
the objectives of the survey. It is important to keep the specific objectives
of the survey and the definition of the target population in mind when consi-
dering the results of a survey. For example, only the NRRS was designed to
provide data on the time people spent where radon was monitored.
Comparability is the confidence with which one data set can be compared
with another. One must consider the measurement method and the objectives of
the measurement program, as well as documented precision and bias, when
comparing data collected by different surveys or programs. For example,
although both the SRRS and the NRRS have measured radon concentrations in U.S.
homes, their results are not directly comparable because they had different
objectives, target populations, and measurement methods.
-------
5.0 QA AUDITS
A QA audit is a formal review of facilities, procedures, and documen-
tation. Ideally, the audit is performed by personnel organizationally
independent of the project management to minimize the risk of any conflict of
interest. Because the QAPjP is the basis of a QA audit, it is important that
the plan describe all important aspects of data collection and processing and
that it be accurate and current. Three types of audits were performed by RTI
during these surveys: technical systems audits (TSAs), performance evaluation
audits (PEAs), and audits of data quality (ADQs). After each audit, a written
report describing the audit objectives, procedures, and findings was prepared.
A TSA is a qualitative on-site evaluation of a measurement system. Its
objective is to assess and document all facilities, equipment, recordkeeping,
data validation, calibration procedures, and QC procedures. Any undocumented
or unauthorized deviations from the QAPjP should be noted in the audit and
included in the audit report. Although a TSA does not provide quantitative
information, the information collected during these audits helps the auditor
quickly determine whether or not the quality of data is likely to be adequate
for its intended use.
RTI conducted five TSAs for these surveys. Two commercial ATD processing
facilities were audited, as well as EPA's primary canister processing faci-
lity. In addition, RTI's data processing and data management for the NRRS and
the NSRS were audited. Written reports for each audit were prepared and the
majority of the auditor's recommendations were implemented by project
management.
A variety of problems or potential problems were identified by these TSAs.
For example, calibration control charts for canister counting stations could
not be readily produced, although this had been recommended in a previous
audit. When these charts were finally produced, they indicated a problem with
one of the counting stations. Also, ATD processing laboratories should rou-
tinely calibrate at exposures > 4 pCi-y/L, since this is a critical value in
mitigation decisions. Calibrations at high exposures also help to separate
the effects of the background correction from the response of the sensitive
material in the detector. For a large project, data processing and
management system should be developed before data collection begins. This
allows data editing to begin as soon as the first data is available, thereby
providing feedback to correct some problem patterns before all the data have
been collected.
A PEA is a quantitative evaluation of a measurement system. This usually
involves analysis of prepared samples or spiked detectors. Blank and collo-
cated (duplicate) samples/detectors may also be used. As described above, RTI
evaluated the performance of all the radon detectors used in these surveys
using blank, spiked, and collocated detectors. In most cases, these QC detec-
tors were mixed with field-exposed detectors and processed 'blindly by the
laboratories. These results were tabulated and compared with the goals for
DQIs specified in the QAPjPs. Formal PEA reports were prepared for each
-------
survey. In some cases, the laboratories identified and corrected problems
with calibration or processing. In other cases, the values of the field-
exposed detectors reported by the laboratories were adjusted based on analyses
of the spiked and blank detectors.
An ADQ involves the assessment of the methods used to collect, interpret,
and report the information required to characterize data quality. RTI
performed an ADQ for each of these surveys. These audits were overall reviews
of the survey implementation, detector performance, and data processing. They
involved detailed reviews of the recording and transfer of raw data, data
calculations, documentation of procedures, and the selection and discussion of
appropriate DQIs. Written reports were prepared for each audit.
6.0 RESULTS
6.1 MEASUREMENT ERROR
Measurement error was determined from the spiked, blank, and collocated
detectors described in section 4.0. Table 1 shows the results of the spiked
ATDs from the NRRS, the NSRS, and year 4 of the SRRS. Recovery is the average
reported exposure divided by the monitored exposure +_ the standard deviation
of the reported exposures divided by the monitored exposure. For each of
these surveys, the DQI goals for ATDs were for recovery to be between 0.85 and
1.15 (i.e., |bias| < 15%) and the MARE < 0.25. Note that the NSRS ATDs with
monitored exposures > 0.9 pCi-y/L had recoveries around 0.80 +_ 0.15. A linear
regression of the monitored exposures on the reported exposures gave a formula
to adjust the reported exposures of the field-exposed detectors to correct for
this bias:
Spikes: Monitored = (1.30 ± 0.04)*Reported - (29.4 ± 8.6).
Field-exposures: Adjusted = (1.30)*Reported - 29.4
Figure 1 is a plot of the reported recovery versus the monitored exposure;
Figure 2 shows the adjusted recovery versus the monitored exposure.
Although Table 1 shows the effect of a similar adjustment on ATDs from the
SRRS, the reported values were within the QA objectives specified in the QAPjP
and the values reported to participants were not adjusted.
Table 2 shows the results of the spiked charcoal canisters from the NSRS
and year 5 of the SRRS. The goals for DQIs of spiked charcoal canisters were
the same as for ATDs: recovery was to be between 0.85 and 1.15 and the MARE <
0.25. These diffusion barrier canisters were simultaneously exposed for one
week in EPA radon chambers. The NSRS canisters were shipped via overnight
express to the processing laboratories. The SRRS canisters were shipped via
overnight express to the state survey coordinators, who then sent the canis-
ters to the processing laboratory via U.S. Mail. This table includes only
those SRRS canisters that were processed within 7 days of exposure. The NSRS
canisters were counted the next day after exposure and have better precision.
-------
TABLE 1. RECOVERY OF SPIKED ATDs
Monitored
Exposure
(pCi-y/L)
4
10
1.
2.
4.
8.
1.
2.
4.
8.
0.
0.
1.
4.
0.
0.
1.
4.
2
1
3
5
2
1
3
5
4
9
7
1
4
9
7
1
Survey
NRRS
NRRS
SRRS4
SRRS4
SRRS4
SRRS4
SRRS4*
SRRS4*
SRRS4*
SRRS4*
NSRS
NSRS
NSRS
NSRS
NSRS*
NSRS*
NSRS*
NSRS*
N
302
286
68
68
68
68
68
68
68
68
36
36
36
36
36
36
36
36
Recovery
Mean+Std
0
0
1
0
0
0
1
0
0
0
0
0
0
0
1
0
0
1
.974
.925
.08
.98
.95
.93
.04
.99
.99
.97
.95
.77
.79
.81
.05
.91
.99
.04
+
±
_f
^+
•»•
±
+
+^
•f
±
+
+
^
jf
±
+_
_+
+_
.Dev.
0
0
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
.093
.070
13
14
07
07
14
15
08
07
32
17
14
12
40
21
17
14
MARE
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
.074
.084
.12
.08
.07
.08
.12
.09
.06
.07
.27
.25
.21
.19
.31
.19
.13
.12
* The reported values were adjusted based on a linear
regression of the monitored on the reported values.
N = number of detectors
-------
I
4J
Vl
a
05
1
u
new
Ficfure 1.
Monitored Exposure (pCi-y/L)
Reported recovery of spiked ATDs in the NSRS.
-O
o
4->
•H
c
a
f->
a
HI new
O old
Monitored Exposure (pCi-y/L)
Figure 2. Adjusted recovery of spiked ATDs in the NSRS.
-------
TABLE 2. RECOVERY OF SPIKED DIFFUSION BARRIER CHARCOAL CANISTERS
Monitored
Concentration
(pCi/L)
2.7
3.9
4.1
6.2
8.6
2.7
3.9
4.1
6.2
8.6
2.7
3.9
4.1
6.2
8.6
Survey
SRRS5
SRRS5
SRRS5
SRRS5
SRRS5
NSRS
NSRS
NSRS
NSRS
NSRS
NSRS
NSRS
NSRS
NSRS
NSRS
N
10
24
5
19
20
16
16
16
16
16
16
16
16
16
16
Recovery
Mean+Std.Dev.
0.94 + 0.08
1.02 i 0.11
0.91 + 0.07
0.87 i 0.06
0.94 +. 0.09
0.96 + 0.06
1.10 ± 0.07
1.02 i 0.06
1.02 + 0.05
0.96 + 0.04
0.87 i 0.11
1.09 + 0.09
0.96 i 0.09
0.98 i 0.07
0.97 + 0.07
MARE
0.09
0.08
0.09
0.13
0.09
0.06
0.10
0.06
0.05
0.05
0.15
0.11
0.08
0.06
0.06
Processing
Laboratory
NAREL
NAREL
LVF
N = number of detectors
Each survey mixed some blank (unexposed) detectors in with the field-
exposed detectors. In the NRRS, two blank ATDs were included in each etch
batch. RTI asked the processing laboratory to report their calculated
exposure for each detector, no matter how low, although their normal procedure
was not to report a specific value when the calculated exposure was < 30 pCi-
d/L (0.082 pCi-y/L). The NRRS QAPjP specified that all blank ATDs should have
reported exposures Ł 0.16 pCi-y/L. The average reported exposure of the 158
processing blanks was 0.11 +_ .13 pCi-y/L; 59% had reported exposures > 0.082
pCi-y/L and 23% were > 0.16 pCi-y/L. This suggests that the laboratory may
have underestimated the background correction of the NRRS ATDs. The effect on
12-month exposures was not considered significant, however. RTI also distri-
buted blanks to field supervisors around the U.S. to evaluate the potential
for exposure of detectors during shipping. Eight field blanks were mailed to
each of the 11 field supervisors in September 1990, shortly after most of the
field-exposed detectors had been retrieved. The supervisors were instructed
to open the foil bags, apply the adhesive foil seals, and return these
detectors immediately to RTI. There was no significant difference in the
reported values for any location; their average reported exposure was 0.15 +,
.12 pCi-y/L.
In the NSRS, two blank detectors were sent to each school with instruc-
tions that they were to be stored, unopened, while the other detectors were
being exposed. After the other detectors had been retrieved, the blank
-------
detectors were opened, immediately sealed, and packaged with the field-exposed
detectors for shipment to the processing laboratory. The NSRS QAPjP specified
that all blank ATDs should have reported values < 45 pCi-d/L (0.12 pCi-y/L)
and all blank canisters should have reported activities < 0.5 pCi/L. The
average adjusted exposure of the 188 field blank ATDs was 0.06 i .10 pCi-y/L;
4.3% were > 0.12 pCi-y/L. Only 0.6% of the 1485 charcoal canisters had
reported activities > 0.5 pCi/L; their average activity was -0.05 +. .18 pCi/L.
There were no true field blanks in the SRRS; that is, all detectors sent
to houses were exposed. Some unexposed ATDs, however, were mixed with spiked
ATDs and field-exposed ATDs when they were sent to the processing laboratory.
Also, 2% of the canisters sent to the states for distribution to participants
were mailed unexposed to the EPA processing laboratory. In year 4 of the
SRRS, 40% of the 68 blank ATDs had reported exposures > 0.16 pCi-y/L, but only
9% of the adjusted exposures were this large. During year 5 of the SRRS, all
of the 166 blank canisters had reported activities < 0.5 pCi/L, which was the
goal specified in the QAPjP. The average activity was -0.12 ± .27 pCi/L.
Precision of the radon detectors was assessed using collocated, field-
exposed detectors. The mean value (M^) and difference (D^) of each collocated
pair (A^, B^) were determined:
The mean value was used to group pairs into exposure ranges. The pooled mean
(Mp) , standard deviation (SDp) , and coefficient of variation (CVp) were
calculated for each range:
Mp = (!, M^/N
SDp={2; Di2/(2N)]1/2
CVp= SDp/Mp= t(2N)I(Ai-Bi)2]1/2/[Z(Ai-(-Bi)]
The pooled standard deviation can be used to calculate confidence intervals.
Table 3 shows the precision of the ATDs used in the NRRS, the NSRS, and
year 4 of the SRRS. The goal in each survey was for the CV of pairs exposed
near 4 pCi-y/L to be < 0.20. Since the NSRS detectors were exposed for only 5
months, there are few pairs in the higher exposure ranges. Their precision,
however, did not satisfy the goal specified in the NSRS QAPjP. A review of
individual pairs with poor precision showed that their precision varied by
lot. Two lots of ATDs were used for the NSRS: one lot of 5000 was purchased
in the fall of 1990, and another lot of 500 ATDs, left over from the NRRS, was
used to supplement these detectors . The NRRS detectors were the same model as
the ATDs purchased for the NSRS, but were purchased in the spring of 1989.
These detectors were distributed to RTI field interviewers around the U.S.
during the summer of 1989, but were not placed in homes. The field
interviewers returned the unused ATDs to RTI where they were stored in a low
(but not zero) radon environment. It is possible that some of the radon-proof
foil bags received minor damage during shipment in the summer of 1989 and that
-------
these detectors accumulated some exposure during storage for over two years.
When the precision of the NSRS ATDs was calculated by lot, the performance of
the 'new" ATDs was much better than that of the 'old* ones. Less than 10% of
the field-exposed ATDs were left over from the NRRS, but 42% of the spiked
ATDs were "old." The linear regression of monitored on reported values for
the spiked ATDs was repeated for each lot and the lot-specific regression
coefficients were used to adjust the values reported to participating schools.
TABLE 3. PRECISION OF COLLOCATED ATDs
Average Exposure
(pCi-v/L)
0.00
0.25
0.75
1.50
3.00
0.00
0.25
0.75
1.50
0.00
0.25
0.75
1.50
- 0.25
- 0.75
- 1.50
- 3.00
- 6.00
- 0.25
- 0.75
- 1.50
- 3.00
- 0.25
- 0.75
- 1.50
- 3.00
N
44
47
32
28
17
139
78
68
30
113
60
20
3
Mp ± SD
(pCi-y/L)
0.11 i 0.08
0.46 _+ 0.19
1.17 i 0.21
2.06 + 0.13
4.22 i 0.26
0.08 + 0.08
0.42 ± 0.14
0.98 HK 0.35
2.22 ± 0.49
0.07 i 0.07
0.45 ± 0.10
1.00 ± 0.18
2.01 ± 0.22
CV- Survey
0.754 NRRS
0.336
0.177
0.063
0.060
0.96 NSRS All*
0.33
0.36
0.22
0.96 NSRS New*
0.22
0.18
0.11
N = number of pairs
* Includes some ATDs left over from the NRRS
** Includes only ATDs purchased for the NSRS
Table 4 shows the precision of the diffusion barrier charcoal canisters
exposed for one week in the NSRS and year 5 of the SRRS. The last row shows
the precision for canisters exposed near 4 pCi/L and processed the next day
after exposure. The goal in each of these surveys was for canisters exposed
at or above 4 pCi/L to have a CV < 0.10.
As described in section 4.0, the goal was to make 12-month measurements in
at least 5000 homes. RTI estimated the fractions of homes contacted that
would be eligible and willing to participate in the NRRS, and that would
actually return detectors. RTI also estimated the fraction of those detectors
that would be usable (i.e., undamaged with good placement information). Based
on those estimated fractions, RTI determined that 11,400 homes should be
contacted. Table 5 shows the completeness goals and actual results for the
NRRS. Only 74% of the housing units contacted were eligible for the survey
(continuously occupied for at least 9 months during the year with no plans to
-------
TABLE 4. PRECISION OF COLLOCATED CHARCOAL CANISTERS
Average Cone .
(pCi/L)
0.00
0.25
0.75
1.50
3.00
0.20
0.25
0.75
1.50
3.00
6.00
3.50
- 0.25
- 0.75
- 1.50
- 3.00
- 6.00
- 0.25
- 0.75
- 1.50
- 3.00
- 6.00
-
- 4.50
N
9
14
12
8
1
107
665
343
239
64
24
30
Mp ± SDp
(PCi/L)
0.11 ± 0.13
0.50 i 0.09
1.00 +. 0.32
2.13 + 0.65
3.43 + 0.42
0.23 + 0.16
0.47 i 0.15
1.07 +. 0.21
2.09 i 0.37
4.12 + 0.33
9.77 ± 0.46
3.90 i 0.35
ŁVp_
0.19
0.32
0.31
0.12
-. ___
0.32
0.20
0.18
0.08
0.05
0.09
Survey
SRRS5
NSRS
NSRS
N = number of pairs
move within 12 months) and about 10% of these did not remain eligible
throughout the 12-month monitoring period. Hence, only 68% of the sample
housing units were eligible for the NRRS. More of these units than
anticipated were willing to participate in the survey and returned usable
detectors, however, so the final results included data from 5700 homes. With
this higher response rate, the potential for bias due to missing data was
substantially reduced. Recontacting the participants every 3 months (panel
maintenance) was an important factor in the high response rate.
TABLE 5. COMPLETENESS OF THE NRRS
Fraction
Eligible
Willing
Returning detectors
Usable detectors
Overall
Estimate
0.79
0.80
0.69
0.98
0.43
Actual
0.74
0.84
0.80
0.999
0.49
The SRRS QAPjP specified a cumulative usable rate of at least 40% for
charcoal canisters. Eligibility requirements for the SRRS were stricter than
for the NRRS, because the house had to be owner-occupied with a permanent
foundations and ground contact. The estimates listed in Table 6 below were
based on experience during years 2 and 3 of the SRRS.
-------
TABLE 6. COMPLETENESS OF YEAR 4 OF THE SRRS
Fraction Estimate Actual
Eligible 0.63 0.59
Willing 0.88 0.90
Returning usable detectors 0.82 0.78
Overall 0.45 0.41
The NSRS Q.APJP specified three completeness objectives: 1) that 80% of the
schools contacted would agree to participate and would actually place
detectors, 2) that 80% of the canisters placed in schools would be returned
and would produce usable readings, and 3) that 70% of the ATDs placed in
schools would be returned and would produce usable readings. A higher
attrition rate was anticipated for the ATDs because they were placed for
several months rather than one week. Actual attrition during placement was
much less than anticipated for both types of detectors {see Table 7).
TABLE 7. COMPLETENESS OF THE NSRS
Canisters
Fraction
Goal
Eligible and
Returning
Overall
willing
usable
detectors
0
0
0
.80
.80
.64
Actual
0
0
0
.78
.91
.71
ATDs
Goal
0
0
0
.80
.70
.56
Actual
Not
yet
avail
6.2 SAMPLING ERROR
For each of these surveys, sampling weights that accounted for the unequal
probabilities of selection were calculated so that unbiased population
estimates could be generated. Because 100% response rates were not obtained
in any of the surveys, the sampling weights of the participants were increased
to account for sample cases that failed to yield usable information; the
resulting analysis weights were used in analyzing the survey data.
Weighted data were used in the analysis for each of these radon surveys.
Sampling errors of survey estimates of means, proportions, and totals were
computed using SUDAAN, a computer software program developed by RTI for
analyzing survey data with complex error structures. Thus, the estimated
sampling error for each of the survey estimates took into consideration the
full complexity of the sample design.
This procedure implicitly included the random measurement errors
(measurement precision) in the calculation of the sampling errors. The SUDAAN
software, however, does not routinely incorporate systematic errors (such as
measurement bias) in estimating these errors. Statistical techniques do exist
-------
to adjust the parameter estimates for measurement bias, but they greatly
increase the complexity of the analysis. The QA data allowed us to estimate
the magnitude of the measurement bias and judge whether the bias was
sufficiently large to warrant efforts for more complex analysis. In the
NRRS, the measurement bias was small enough that the more complex analysis was
not necessary. In the SRRS and the NSRS, some of the ATD values reported by
the processing laboratory were adjusted to reduce the bias using the
regression procedure described in section 6.1.
6.3 UNCERTAINTY IN THE FREQUENCY DISTRIBUTIONS
Estimating the frequency distributions is a special case of estimating the
proportions alluded to in section 6.2. The frequency distribution was
estimated by grouping the radon measurements into intervals and estimating the
proportion (or percent) of housing units with radon measurements in each
interval. The standard error for the percent in each interval was estimated
using our SUDAAN software.
In this special case, the measurement precision may bias the estimates of
the percent in each interval. Radon measurements near the interval boundaries
may be assigned to the incorrect interval because of either systematic or
random measurement error. Bias arises when misclassifications due to
measurement errors do not "balance out." Because the QA procedures controlled
both the measurement bias and precision to satisfactorily low levels, the
effect of measurement error on the uncertainty in the frequency distributions
was judged not to be significant.
The sample design, its implementation, and the QA procedures were
instrumental in achieving the primary objective of the NRRS: an accurate
estimate of the frequency distribution. The relative standard error of the
fraction of homes with radon levels greater than 10 pCi/L was 28%, well below
the goal of 50% established during the design of the NRRS.
7.0 SUMMARY
Quality assurance activities were incorporated in all phases of three EPA-
sponsored radon surveys. Project planning included preparation of
comprehensive QA project plans based on the specific objectives of each
survey. These project plans included goals for data quality and specified
operating procedures capable of meeting those goals. QA audits were conducted
during each survey to verify that these procedures were followed. The
performance of the radon detectors was monitored with spiked, blank, and
collocated detectors. The QC detectors were processed "blindly' along with
the field-exposed detectors. This helped us to identify and resolve some
calibration and processing problems. In addition to detector performance, the
QA/QC activities documented the procedures for collecting and processing
questionnaire data.
-------
The precision of ATDs exposed to at least 3 but not over 6 pCi/L for 12
months in the NRRS was about 6%. The precision of the diffusion barrier
charcoal canisters exposed to at least 3 but not over 6 pCi/L for one week in
the NSRS was about 8%. The precision of charcoal canisters with similar
exposures in the SRRS was slightly worse (about 12%), probably because they
were returned via U.S. Mail rather than by overnight express.
8.0 REFERENCES
1. Yeager, William M., et al. State Radon Assessment Program: Quality
Assurance Project Plan. RTI/4658/16F-03F, Research Triangle Institute,
Research Triangle Park, NC, December 12, 1990.
2. Daum, Keith A., et al. National Residential Radon Survey: Final Quality
Assurance Project Plan. RTI/4240/10-01F, Research Triangle Institute,
Research Triangle Park, NC, September 6, 1989.
3. Yeager, William M., et al. National School Radon Survey: Quality Assurance
Project Plan. RTI/4658/24F-01, Research Triangle Institute, Research
Triangle Park, NC, January 30, 1991.
4. Office of Radiation Programs, U.S. Environmental Protection Agency.
Indoor Radon and Radon Decay Product Measurement Protocols. EPA 520-1/89-
009, U.S. Environmental Protection Agency, Washington, DC, March 1989.
5. Quality Assurance Management Staff, U.S. Environmental Protection Agency-
Interim Guidelines and Specifications for Preparing Quality Assurance
Project Plans. QAMS/005-80, U.S. Environmental Protection Agency,
Washington, DC, 1980.
6. Gearo, J.R., et al. The Growth of the National Radon Measurement
Proficiency (RMP) Program. In: Proceedings of the 1988 Symposium on Radon
and Radon Reduction Technology, Vol. 2, EPA-600/9-89-006b, U.S.
Environmental Protection Agency. Research Triangle Institute, Research
Triangle Park, NC, March 1989.
7. National Council on Radiation Protection and Measurements. A Handbook of
Radioactivity Measurements Procedures. NCRP Report 58, 1985.
-------
IXP-5
RADON IN HOUSES AROUND THE PLOMIN COAL FIRED POWER PLANT
Lokobauer N.. Franie Z., Bauman A. and Horvat D.
Institute for Medical Research and Occupational Health.
University of Zagreb
Zagreb, Croatia
ABSTRACT
The paper presents the results of investigations of radon
activity concentrations in the old houses, assumed to be built
from materials containing slag and ash of coal combusted in the
Plomin Coal Fired Power Plant. Measurements were performed using
solid state nuclear track detectors (Kodak Pathe LR-115 films,
type III mounted inside open and filter cups. Detectors were
installed during three months at 15 locations in 30 houses. In
order to determine radon emitting source (building material,
soil, radioactive airborne particulates) as precisely as
possible, films were also installed in certain number of houses
built more recently. The radon activity concentrations inside
houses varied between 370 Bqm~3 (old houses) and 16 Bqm~3
(recently built houses).
-------
INTRODUCTION
Radon emitted from building materials which usually contain
more uranium/radium than soil, may cause a high indoor radon
concentration, particularly in winter, when there is a greater
tendency to seal houses to conserve heat. In the area of the
impact of the Plomin Coal Fired Power Plant (CFPP), it was of
special importance to investigate the concentration of radon, and
possible health risk to the population in hoxises assumed to be
built from ash and slag.
Plomin CFPP. located on the Istrian Peninsula in Croatia,
uses coal containing elevated uranium series concentrations. The
impact of the Plomin CFPP on the environment has been studied for
years by the Department for Radiation Protection of this
Institute. In the eighties uranium activity concentrations
measured in the coal was 10 to 15 times higher than the average
world values.
Ash and slag remaining after coal combustion contain uranium
and its decay products which cause redistribution of natural
radioactivity from the soil to the locations at which it may
produce a significantly impact to the environment and health of
the inhabitants. In the vicinity of Plomin CFPP the need for
radon measurement was noted since ash and slag were used in the
past as building materials for the construction of houses.
Another possible source of radon emission for local population
could be cement produced in a highly developed industry based on
regional mineral resources.
MATERIAL AND METHODS
Radon activity concentrations were measured by Kodak Pathe
LR-115 films, type II, mounted inside open and filter cups. The
detectors were exposed during three months (December 1991 to
February 1992) at thirty houses. The track densities were counted
under a microscope, after etching the films under precisely
controlled conditions (2.5 N NaOH, 60 °C, 90 min). Bare detector
was calibrated in National Radiological Protection Board,
Chilton, Didicot, and detector sensitivity coefficient was 10.3
BQ m~3 d per tracks cmr2.
-------
RESULTS AND DISCUSSION
Detectors for radon activity measurement were installed in
old and recently built houses on selected locations in the radius
of 60 km around the Plomin CFPP. The values of mean radon
activity concentrations measured in winter in thirty houses
around the plant are given in Table 1.
TABLE 1. THE MEAN RADON ACTIVITY CONCENTRATIONS* IN HOUSES
AROUND THE PLOMIN COAL FIRED POWER PLANT (DEC. 1991 -FEB. 1992)
Activity concentration (Bqm~3)
Location Old houses More recently built houses
1
2
3
4
5
6
7
8
9
10
11
12
13
14*
15*
209
270
77
90
160
105
122
102
370
152
179
107
132
118
61
+
+
+
+
±
±
+
±
±
±
±
+
±
±
+
14
16
9
9
13
10
11
10
19
12
13
10
11
11
8
35
23
16
48
60
66
66
64
43
25
79
39
48
16
22
+
+
±
±
±
+
±
±
±
±
±
+
+
+
+
6
5
4
7
8
8
8
8
7
5
9
6
7
4
5
* Activities reported as ± one sigma error.
* Location of cement industry
The data show that radon activity concentrations were higher
in the old houses which points to conclusion that recently
constructed houses were built from material containing lower
uranium series radionuclide concentrations. On basis of the
obtained values it can be noted that impact of Plomin CFPP as
well as nearby cement industry could be neglected with respect
to radon concentrations in the air. Building material was
therefore major radon emitting source in old houses.
-------
In 12 old houses radon activity exceeded the value of 100
Bq m~3 and in 3 old houses radon activity was even higher than
200 Bq m~3. According to recommendations of International
Commission on Radiological Protection this is the level which
requires protective measures (2).
Investigation of track densities (tracks cm-2) of the bare
and filtered detectors installed in selected houses around the
plant gave equilibrium factor F of radon and its daughters
ranging from 0.2 to 0.8.
CONCLUSION
Investigations carried out in the area of the impact of the
Plomin CFPP including old houses built mostly from ash and slag
of the coal combusted in the plant showed that indoor radon
concentrations may cause risk to human health. In order to
protect people inhabiting this area it should be necessary to
continue the investigations and assess the effective dose
equivalent, to the population.
The work described in this pap>er 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. Marovic G. Enhanced Natural Radioactivity around a Coal
Fired Power Plant. M.Sc. Thesis. Technological Faculty,
University of Zagreb, 1985. (In Croatian)
2. International Commission on Radiological Protection,
Publication 39, Pergamon Press 1984.
-------
IXP-6
A RADON SURVEY AT SOME RADIOACTIVE SITES IN INDIA
By: Jaspal Singh, Lakhwant Singh, Surinder Singh,
and H. S. Virk
Department of Physics
Guru Nanak Dev University
Amritsar-143005, India
ABSTRACT
From studies carried out all over the world, there is
scientific agreement between the incidence of excess lung cancer
among underground miners and exposure to radon and radon
daughters. The epidemiological studies also support the fact that
there are increased incidences of bronchial carcinomas among
individuals exposed to relatively high levels of radon and its
daughter products. Increased incidences of lung cancer have also
been found in homes even at moderate values of radon activity.
The current study has been carried out for a complete year at two
sites in Himachal Pradesh, India. In one of the areas under
study, the average radon activity has been found to be 26.41 pCi/L
in summer and 31.73 pCi/L in winter, giving an annual exposure of
47.09 mSv. In the second area, these levels were found to be
16.54 pCi/L and 18.52 pCi/L, respectively, which gives an annual
exposure of 28.39 mSv. The study is being conducted for the
second year. The complete data collected for the two years will
be presented in the paper.
-------
IXP-7
Islandwide Survey of Radon And Gamma Radiation
Levels in Taiwanese Homes
Ching-Jiang Chen, Chi-Wah Tung and Yu-Ming Lin
Taiwan Radiation Monitoring Center, Atomic Energy Council
823 Cherng-Ching Road, Kaohsiung, Taiwan 833, ROC
ABSTRACT
An islandwide survey of Taiwanese homes was conducted to determine the average annual effective
dose equivalents to Taiwan population from exposure to radon and i radiation. The radon concentration
was measured using cellulose nitrate films (CN film) as the solid state nuclear track detectors, while 7
radiation dose was measured using thermoluminescent dosimeter (TLD). The CN films were put in a
plastic cup with a filter on the top and mailed to approximately 250 randomly distributed householders and
replaced for every two months. The TLD were sent to 42 selected homes for 3 months exposure. The
-3
average Rn concentration in Taiwanese homes measured over 1.5 year is 10 Bqm while the average 7
dose rate is 0.12 u Gy h . Using appropriate conversion factors, the annual average effective dose
equivalents to the Taiwan population were determined to be 0.5 mSv and 0.59 mSv for Rn and 7
radiation exposure respectively.
-------
1. INTRODUCTION
There are many papers deal with radon recently. According to the UNSCEAR report (1) the annual per
caput dose from natural radiation in the world is 2.4 mSv. Fourty-six percent of it comes from radon ( Rn ->
"> 14
Po) and 33% of it comes from external gamma radiation which include cosmic rays. In a lifetime of human
beings about 80% were spent indoors, so indoor radon and indoor gamma dose are the most important items in
assessing population doses. From the UNSCEAR 1988 report, there are 27 countries that performed large scale
indoor radon survey including 96,469 houses. Most of the data show a log-normal distribution with medians
ranged from 10-140 Bqm . Most of the data came from developed countries located in the middle and high
latitude of northern hemisphere. However these data cannot stand for the world average. Some of the newly
published data show somewhat different from the UNSCEAR report. New large scale and nationwide survey and
reassessment are still undertaken (2,3,4,5).
Tai%van island is located in lower latitudes of northern hemisphere (22' -25 ° N), along the west coast of
Pacific ocean, which is an area with oceanic subtropical climate. Both temperature and humidity are high in
Taiwan. Geologically, sedimentary rocks formed of shale and sandstone dominate in the whole island. Its
natural radioactivity is quite normal as shown in table 1. More than 90% of the houses in Taiwan are build with
concrete and bricks. Most of the building materials are made of domestically produced river sand, cement,
gravels and bricks made of clay. Their floors are made of concrete, tile, sheet vinyl, marvel, ground aggregate
and granite. Because the temperature is warm and tlie windows are large and opened frequently, the ventilation
rate is normally high in Taiwan. Owing to the promotion of living standards, the air-conditioning facilities are
popular and there are many skyscrapers with central air-conditioning systems in city areas. Islandwido survey of
indoor radon level has been an urgent project for Taiwan Radiaiton Monitoring Center.
2. EXPERIMENTAL DESIGN
Sampling
One thousand householders with profession as science teachers in junior high schools were randomly selected
and invited to participate in the survey, and 250 householders responded positively. A questionnaire including 9
questions about the basic information of each house was sent to each participants. The format of questionnaire
is shown in Table 2. Part of the participants terminated the participation during the survey. About 212
householders participate in this study from the beginning to the end.
Indoor radon measurement
Two cellulose nitrate (CN) films in a plastic cup was mailed to each participant together with a letter
describing how to install the measuring device. In order to avoid the pollution of aerosol in the air, a whatman
GFC glass fiber filter was attached to the cap of the cup. Thirty small holes were drilled on the cap to let the
air diffuse through the filter. Figure 1 shows the structure of the measuring cup. Two pieces of CN film cut into
30 x 45 mm were adhered on the inner wall of the cup.
Indoor 7 dose measurement
The pen-type TLDs of Panasonic UD-200S CaSO4:Tm were \ised to measure the cumulate indoor t dose.
Fourty two houses were measured for three months exposure. The TLDs were put in the bottom of radon
measurement cup. All the dosimeters were mailed through post office.
-------
building material and structure of the building may have positive effect on the radon level.
Figure 6 shows indoor radon level. According to the measured result, it is found that the indoor Rn level
-3 -3
follows log-normal distribution. There are 30% between 3-6 Bqm , 28% between 6-9 Bqm and 20% between 9-
-3 -3 -3
12 Bqm . Hence 79% houses are between 3-12 Bqm . As a whole, the geometry mean is 8.5 Bqm and the
geometry standard deviation is 0.59.
Seasonal variation
Figure 7 shows the seasonal variation of indoor radon. Data were collected and analyzed in every two
months. The mean concentration fluctuates between 8.4-11.1 Bqm with no apparent seasonal variation. In
medium and high latitude countries, winter is cold and window is often closed. Radon concentration becomes
higher due to poor ventilation. In the summer, radon concentration becomes lower in good ventilation rate. In
Taiwan, the seasonal variation of indoor Rn is not significant.
Relationship between indoor radon concentration and gamma dose rate.
-i
Results of TLD measurement shows that indoor gamma dose rate are between 0.066-0.189 u. Gyh and the
average is 0.121 u Gyh (including contribution of cosmic rays). The result is high as compared with other
countries. It is because that more concrete and brick is used as the building materials, which contains more
natural radionuclides. Figure 8 shows the correlation between indoor gamma dose rate and indoor radon activity.
The regression curve based on 42 pairs of data is y = 0.000562X+0.113032, and the correlation coefficient, i is
equal to 0.3. The result shows that the indoor radon concentration is in correlation with gamma dose rate at
0.95 confidence level.
Correlation of indoor radon and building factors
In order to explore the factors that might contribute significally to the level of Rn concentrations in
Taiwanese houses, the following information was sought through the questionnaires^!) building style; (2) main
building materials; (3) floor materials; (4) wall materials; (5) in which year the house was built. Most of the
householders had forgotten the year of construction, so this factor could not be analyzed. The other factors were
analyzed as shown in Table 4.
The first row in Table 4 listed 4 types of house including the number of house and the average radon
concentrations for each type of house. There are no correlation between radon concentration and type of house.
Average radon concentration of each type of house is very close to the gross average.
The second row in Table 4 shows 4 types of main building materials. There are no wooden house in this
study. Wooden house is hardly seen in Taiwan nowadays. The others include 4 houses with an average of 19.4
Bqm which is double that of gross average. These houses are the old style and were built with local soil and
bricks. Their flooring were just natural soil that might possess higher radon exhalation rate. This is an
interesting finding.
The third row shows the effect of 7 types of floorings. The items of wooden floor and others show higher
radon concentration than the other 5 types of flooring. The fourth row shows the effect of wall surface
materials.There are no much difference between the five types of wall materials.
Kinman district
Kinman isle is close to the seashore of Mainland China which is famous for its miltary factility. There are
granite rocks on this isle. Natural radioactivity in granite is 3-4 times higher than Taiwan rocks. RaJon
-------
Outdoor radon measurement
In order to measure the outdoor radon level, 145 participants was selected all over Taiwan. The measuring
device is the same as indoor and the cup was put in balcony or under eaves to get rid of rainfalls and sunshine. In
order to enhance the efficiency of CN films, the exposure period was extended to 3 months.
3. Experimental
Measurement of radon
Measuring methods of radon in the air can be catagorized into active and passive methods. For long term
observation of radon concentration in air, the passive alpha track method was selected in this survey. CN film
method is a popular method in large-scale indoor radon survey. The CN film used is the LR-115 Type I ,
strippable which is a product of Kodak Pathe, France . The optimized etching condition is 60 °C for 70 min.
under 2.5N (10% by weight) Na(OH) solution. After chemical etching, the films were rinsed in clean water for
10 min. and then dried in air. The stripped films were counted by a spark counter. Figure 2 shows the
etching device, and Fig.3 shows the structure of spark counter. To enlarge the holes on the CN films, the films
were presparked three times under 600V HV. Six times spark counting under 450V can get the a tracks on
each film in a good statistics. Figure 4 shows the pulse of spark counter from an oscilloscope.
Calibration of CN film
For each batch of films to be etched, 10 calibration films were inserted during chemical etching to get the
individual response factors. The calibration films were exposed in a radon chamber of 3.0m x 1.8m x 2.4m with
radon concentration mainteined at around 1000 Bqm . The radon concentration in the chamber was monitored
-3 -2 -3
by a scintillation cell. The response factor of CN films is 1.02 ± 0.12 x 10 track . cm /Bqm .h obtained from
10 exposed films.
Measurement of indoor t dose
Indoor i dose was measured with Panasonic UD-200S TLD and UD-512 TLD reader. Its feasible dose
range is from 0.1 mR to 20 R. While in reading, the TLDs were heated with hot nitrogen for 10 s. from room
temperature to 430 °C . The CaSO4:Tm was sealed in glass tube and shielded with copper cap to reduce the low
g
energy response. The system was calibrated with a 3.7 x 10 Bq radium-226 source which is close to the
characteristics of natural radiation field. Its fading effect is negligible.
4. Results and Discussion
Analysis of the results
The results of radon survey in Taiwan during Dec. 1989 through June 1991 is shown in Table 2. The
Taiwan island is divided into 24 districts and data are collected every two months. Each district contained 6-29
homes in the survey. Buildings in granite rocks are different from others and were considerd as a special
district.Indoor to outdoor concentration ratios and indoor/outdoor mean concentrations are also listed in Table 2.
The indoor mean concentration is 9.93 + 4.08 Bqm while the outdoor is 4.0 + 3.2 Bqm .
Figure 5 shows the indoor and outdoor radon concentrations and the mean ratio is 2.5. Nantou County
(district-lO)has higher indoor and outdoor radon levels as compared with the other counties. The Nantou county
is far from the coastal line and surrounded by mountains where radon is easier to accumulate in air and leads to
higher level. In addition, Nantou county has many brick-made houses that were included in this survey. The
-------
concentration in underground buildings of granite is 36-64 Bqm with an average of 51 Bqm . Its 5 times higher
than the average of Taiwanese houses. The last row of Table 2 shows that the radon levels in houses of Kinman
are also higher than that of Taiwan.
Comparison with other countries
The New York State data show that there are 70% homes between 0-37 Bqm and 17.6% between 37-74
Bqm . Table 4 lists the data of some countries in the world. Radon level in Taiwan is low as compared with
, . (4)
other countries .
Dose assessment
other countries . The result is close to Japan and far below New York State of U.S.A.
There are many dose assessment models for radon-induced lung dose( 1,7,8). According to the dose model of
James, the dose conversion factor of indoor radon is 0.05 mSvy per Bqm for adults. The calculated per caput
dose for Taiwan population is 0.5 mSv/y and the collective dose is 10 manSv/y.
The average indoor 7 absorb dose rate is 0.12 u. Gy/h. Suppose that 80% of the time is spent indoor, the
induced per caput dose is 0.85 mGy/y. By applying ihe conversion factor 0.7 for absorbed dose to effective dose
equivalent (l),the induced average effective dose equivalent to Taiwanese population is 0.59 mSv/y. The collective
dose in Taiwan is 1.19 X 10 man • Sv/y.
5. Conclusions
Indoor radon level is strongly dependent on the ventilation rate and radon exhalation rate. Radium-226
content is moderate in concrete of Taiwanese house and the exhalation rate from concrete is not high as
compared to that of soil. The warm temperature makes the ventilation rate in Taiwanese house higher than that
in temperate region. So the radon level in Taiwan is lower as comapared with that in temperate region countries.
From the previous discussion, we conclude:
(1) The average radon level based on the survey of 250 Taiwanese homes is 9.9 ± 4.1 Bqm" which is much lower
than the action level of U.S. EPA. Even the maximum indoor Rn concentration is still much lower than 150
Bqm . The outdoor Rn level is 4.0 ± 3.2 Bqm .
(2) The ratio of indoor to outdoor radon level is 2.5.
(3) Both Nantou County and Kinman County possess higher radon level because of their geographical and
geological characteristics.
(4) The traditional Taiwanese houses have higher indoor Rn level than modern concrete houses.
(5) Indoor t dose rate varied from 0.066 to 0.189 u. Gy/h with an average of 0.121 u Gy/h which is higher
than temperate countries. The annual per caput dose from i radiation is 0.59 mSv.
(6) The correlation coefficient of indoor y dose rate and Rn level shows a significant correlation at 0.95
confident level.
(7) Indoor Rn level shows no seasonal variaion in subtropical Taiwan.
(8) Building type and wall materials show no significant correlation with indoor Rn level. Traditional building
materials showed higher indoor radon level.
(9) Building in granite rocks shows high radon level even in good air ventilation system.
(10) The effective dose equivalent due to indoor radon is about 0.5 mSv/y in Taiwan.
-------
REFERENCES
1. United Nations Scientific Committee on the effect of Atomic Radiation. Sources, effects and risks of ionizing
radiaiton. New York, United Nations, 1988.
2. Cohen, B.L. and Gromicko, N. Radon-222 levels in low income households. Health Phys. 56:349, 1989.
222
3. Langroo, M.K., Wise, K.N., Duggleby, J.C. and Kotler, L.H. A nationwide survey of Rn and 7 radiation
levels in Australian homes, Health Phys. 61:753, 1991.
222
4. Rahlenbeck S.I. and Stolwijk, A. Indoor Rn level in New York state, North Carolina, and South Carolina.
Health Phys. 61:879, 1991.
5. Cohen, B.L. and Shah, R.S. Radon levels in United States homes by states and counties. Health Phys.
60:243, 1991.
6. Tommasion, L. Etch track techniques in radiation dosimetry, Radia. Protect. Dosim. 17:135, 1986.
7. 1CRP. Lung cancer risk from indoor exposures to radon daughters. Oxford: Pergamon Press: ICRP
Publication No.50, 1987.
8. James, A.C. Lung dosimetry In: Nazaroff, W.W., Nero, A.V., eds. Radon and its decay products in air, New
York Wiley, 1988. P.259.
-------
Fig. 1. Passive measuring device for environmental radon. CN films were adhered on the inner wall of the CUD.
Fig. 2. Chemical etching device for CN films.
-------
CN film
Positive electrode Negative e|ectrode. °
Heavy material
Thick cloth
Aluminum mylar
Base
Fig. 3. Structure of the electrode of spark counter.
Fig. 4. Pulses of spark counter.
-------
( 1989.1E - 1991.06 )
15
c
cr
&
i 2 3 4 5 6 7 8 9 10 11 12 13 W 15 16 17 13 19 20 21 22 23
Districts
Fig. 5. Indoor and outdoor radon levels of 23 districts in Taiwan. The average indoor/outdoor ratio is 2.5.
i"
PERCENT
30
28
26
24
20
18
16
14
12
10
S
6
4
2
0
30.29
.07
~|
\
\
\
\
\
\
\
\
\
\
\
s
\
\
\
\
28.2
\
\
\
\
\
\
\
\
\
\
\
\
\
\
\
2
20.33
\
\
\
\
\
\
\
\
\
\
\
7.05
VI
4.56
\ r\i
\ \
^ *^ u 83D 83% ^ "^" 0 83
\ \ Fvi 1X1 IXI n f\l An ° ° Pvl ° p^aAo ° ° rU
Rn ACTIVITY (Bq nr
Fig. 6. Distribution of radon concentrations in Taiwanese homes.
-------
E
$
<
c
W
O
UJ
15 -r
14 -
13 -
12 -
11 -
10 -
9 -
8 -
7 -
6 -
5 -
4 -
3 -
2 -
1 -
0 --
11.06
10.53
9.14-
9.73 9.6
9.6
9.27
8.39
89.12 90.02 90.04 90.06 90.08 90.10 90.12 91.02
YEAR AND MONTH
Fig. 7. Time variation of indoor radon level in Taiwan.
O
3
O
Q
o
0.19
0.18
0.17
0.16
0.15
0.14
0.13
0.12
0.11
0.1
0.09
0.08
0.07
0.06
20 40
Fig. 8. Correlation of indoor radon and gamma dose.
r =0.3 for 42 houses.
60
-------
Table 1. The mean natural radionuclide concentrations and their ratios in rocks
Radionuclide concentrator
Type of rocks
Igneous
Andesite
Basalt
Dunite
Gabbro
sedimentaries
Conglomerate
Sandstone
Shale
Limestone
Metamorphics
Slate
Green schist
Black schist
Phyllite
Gneiss
«.,
1.11
1.35
0.13
0.05
0.88
1.75
2.70
0.26
2.80
1.07
2.96
3.13
1.90
Th(ppm)
4.18
4.86
0.10
0.15
4.09
7.48
12.40
1.26
13.50
1.62
13.40
16.30
9.70
U(ppm)
1.20
1.27
0.10
0.07
0.87
1.53
2.51
1.36
2.62
0.43
2.47
2.81
1.85
Th(ppm)/
3.77
3.60
0.67
2.75
4.65
4.27
4.58
4.85
4.84
2.14
4.51
5.19
5.09
Raitos
U(ppm)/
1.08
0.94
0.82
1.25
1.15
0.87
0.93
5.23
0.94
0.62
0.83
0.90
0.97
Th(ppm)/
U(ppm)
3.
3.
0.
2.
4.
4.
4.
0.
5.
4.
5.
5.
5.
48
82
82
20
05
89
92
93
17
22
40
79
23
*K: 308 (Bq kg"')/K%
232Th: 4.07(Bq k
238
;U :12.62(Bq kg-'j/ppm
-------
Table 2 Questionnaire for indoor radon survey
1. Your name:
2. Your address:
Phone No.:
3. Type of house:(l) Appartments (2) Bungalow (3) Detached (4) Others
4. Main building materials:(l) Concrete (2) Brick with plaster (3) Wooden (4) Wood and bricks
(5) Others
5. Flooring materials:(l) Concrete (2) Sheet vinyl (3) Marble (4) Wooden (5) Tile (6) Grind
aggregates (7) Others
6. Wall materials:(l) Tile (2) Paint (3) Wall paper (4) Wooden (5) Others
7. Construction year:
8. Measured room:(l) Living room (2) Bed room
9. Dimension of measured room: m long, m wide m height.
-------
Table 3 Indoor and outdoor radon concentration in 24 districts of Taiwan
district
01 Keelung
02 Hualien *
03 Taitung *
04 Ham *
05 Taipei
06 Taipei *
07 Taoyuan *
08 Hsinchu *
09 Miaoli
10 Nantau *
1 1 Taichung *
12 Taichung
13 Clianghua *
14 Yuanlin *
15 Chioyi
16 Tainan *
17 Tainan
18 Kaohsiung *
19 Fungshan
20 Kaohsiung
21 Pingtung *
22 Penghu *
23 Kinman *
24 Kinman #
indoor Rn Cone.
range (Bqm"3)
5.7 - 9.9
5.5 - 12.6
5.4 - 12.8
6.2 - 13.9
4.7 - 19.0
4.4 - 24.1
5.1 - 13.6
6.3 - 17.3
7.2 - 14.5
9.8 - 22.0
8.1 - 15.3
5.7 - 20.4
4.7 - 11.8
4.6 - 18.0
5.0 - 23.8
5.8 - 14.9
8.3 - 21.9
4.7 - 10.5
6.3 - 10.1
6.4 - 22.1
5.5 - 17.8
5.3 - 12.4
9.0 - 24.8
36.3 - 63.5
mean indoor Rn
cone. (Bqiri3)
8.4 ± 1.5
8.2 ± 2.1
8.9 ± 3.1
8.0 ± 2.9
10.6 ± 4.3
10.2 + 5.6
9.0 ± 2.5
10.1 ± 3.0
9.9 ± 2.3
14.4 ± 3.8
10.4 ± 2.6
10.8 ± 5.4
7.8 ± 3.0
8.0 ± 4.1
9.8 ± 6.4
8.2 ± 3.6
13.1 ± 5.0
8.3 ± 2.5
8.5 ± 1.7
10.6 ± 3.8
7.2 ± 3.9
8.3 ± 2.3
14.4 ± 4.0
50.8 ± 13.4
outdoor Rn cone.
( Bqm-3)
3.3 ± 1.8
4.3 ± 1.4
*** ***
*** ***
3.0 ± 1.7
5.5 ± 1.5
5.0 ± 2.7
4.3 ± 4.9
3.4 + 1.5
5.5 ± 1.5
5.4 ± 4.9
4.0 ± 4.2
2.2 ± 1.8
3.8 ± 4.5
1.8 ± 1.3
* *# _L. * * *
3.1 ± 2.3
4.4 ± 4.9
5.4 ± 5.3
3.1 + 3.9
5.1 ± 3.0
4.3 ± 3.7
3.5 ± 3.6
3.5 ± 3.6
Remark : 1. Indoor Rn mean cone. 9.9 ± 4 Bqm"3 .
2. Outdoor Rn mean cone. 4.0 ± 3 Bqm'3 .
3. * county.
4. # under ground passage .
-------
Table 4. Dependence of indoor radon level on building characteristics
Type of building Main building materials
Appartments Concrete
10.8 (58) 9.9 (159)
Bungalow Brick with plaster
9.4 (33) 8.9 (42)
Detached Wooden
9.3 (90) — (0)
Others Others
10.6 (31) 19.4 (4)
Flooring materials Wall materials
Concrete Tile
10.3 (29) 10.6(10)
Sheet vinyl Paint
9.2 (20) 9.5 (158)
Marble Wall paper
8.6 (14) 10.4 (19)
Wooden Wooden
14.5 (8) 10.7 (6)
Tile Others
9.5 (74) 12.1 (19)
Grind aggegates
9.6 (62)
Others
16.9 (5)
-------
Table 5. Radon concentration obtained in some
recent larg-scale indoor surveys
Country
or area
Argentina
Belgium
Denmark
Germany
Italy
Ireland
Japan
Norway
United
Taiwan
Number of Type of Date of
dwellings sampling completion
112 Radon, passive 1985
79 Radon, passive, 1984
one-year exposure
400 Radon, passive 1935
3-month exposure
in summer and
in winter
5970 Radon,passive 1984
3-month exposure.
1 in bedroom
1 in living room
1000 Radon, passive, 1984
3-12 month exposures
736 Radon passive 1987
6-month exposure
250 Radon, electrostatic 1988
integrating, 2 months;
total 1.5 year
exposure
1500 Radon, passive 1985
1-week exposure
552 various 1934
250 Radon, passive 1991
Average vaule
(Bq/m3)
37 (median)
41 (median)
50 (median)
40 (median)
49 (mean)
43 (median)
37 (median)
10 (mean)
90 (mean)
35 (median)
61 (mean)
10 (mean)
8.5 (median)
-------
Session X Posters
Radon in Schools and Large Buildings
-------
XP-1
SOLAR FRESH AIR VENTILATION FOR RADON REDUCTION
by: Monty Holmes
Intermountain Radon Service
P. O. Box 3
Salida, CO 81201
Kelly W. Leovic
Air and Energy Engineering Research Laboratory
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
ABSTRACT
This paper discusses the construction and evaluation of a
solar fresh air ventilation system installed in two classrooms of
a Ranches de Taos, New Mexico, school. The project was initiated
because school radon mitigation research has shown the need for
improved indoor air quality in addition to reduction of elevated
radon levels. Additionally, there are certain types of school
buildings where subslab approaches to radon reduction are
impractical or expensive.
A datalogger recorded continuous radon concentration, carbon
dioxide concentration, room temperature, and subslab-to-classroom
differential pressures in two ventilated and one control classroom
during the February to April 1992 testing period. While the
building was occupied the solar ventilation system did not provide
much additional radon reduction because occupants' frequently
opened windows in the school. However, the classrooms with the
ventilators did have slightly lower carbon dioxide levels compared
to the control classroom, indicating some benefit of ventilation.
To evaluate the effect the ventilation system on radon levels
without the confounding factor of the natural ventilation provided
by open windows, the system was also evaluated during spring break
when the building was unoccupied. Results during this period
indicate lower radon levels in the two classrooms with the solar
ventilators as compared with the control classroom.
This paper has been reviewed in accordance with the U. S.
Environmental Protection Agency's (EPA) peer and administrative
review policies and approved for presentation and publication.
-------
INTRODUCTION
School radon mitigation research has shown the need for
improved indoor air quality in addition to reduction of elevated
radon levels (1). To address both of these needs, a solar fresh
air ventilator was developed by Monty Holmes (one of the authors)
and received an "honorable mention" in the 1991 Innovative Radon
Mitigation Design Competition. The design was a fresh air
ventilation system utilizing solar tempered outdoor air to dilute
(and possibly pressurize) a school classroom. This presents an
alternative radon mitigation technique that provides conditioned
outdoor air to help reduce radon levels and improve indoor air
quality at a relatively low operation cost.
The system was designed to use an electrically powered fan to
continually supply outdoor air during occupied periods to reduce
radon levels and improve indoor air quality. The system supplies
"tempered" outdoor air to the classrooms and is not intended as a
heating source. The Environmental Protection Agency's (EPA's) Air
and Energy Engineering Research Laboratory (AEERL) evaluated the
solar ventilator since it offered a radon mitigation alternative
for buildings with the following characteristics:
• elevated radon levels
• sunny climate (for solar capability)
• occupied primarily during daylight hours (to take
advantage of solar capability)
• active soil depressurization (ASD) was not reasonably
applicable (low permeability subslab fill)
• existing heating system did not supply sufficient outdoor
air
Radon measurements in New Mexico school buildings indicated
that over 20 percent of the 125 schools tested had at least one
classroom with elevated radon levels. In Taos County, every school
had rooms where weekend radon measurements following current EPA
protocols (2) exceeded 4 pCi/L.* Taos has a sunny climate; however,
it also has over 7,000 degree days of heating load. As a result,
building plans from a number of Taos schools under consideration
for this project were reviewed. In a majority of the buildings
investigated, slabs were poured on compacted adobe or dirt with no
subslab aggregate. It was felt that an alternative to ASD would be
necessary in these types of buildings.
* 1 pCi/L (picocurie per liter) *= 37 becquerels per cubic meter
-------
SCHOOL BACKGROUND INFORMATION
A school located in Ranches de Taos, New Mexico, was selected
for this research project because the school has moderately
elevated radon levels (between 5 and 15 pCi/L) in half of the
classrooms, slab-on-grade construction on compacted adobe, and no
conditioned outdoor air in most classrooms. Selection of an
appropriate building where school officials were also amenable to
the project was more difficult than anticipated. Consequently,
when the school in Ranches de Taos was selected for this project,
school officials and the researchers agreed that the installation
would be temporary for the purpose of evaluating the solar fresh
air ventilation system. School officials would have the option of
keeping the system once the project was completed.
Constructed in 1965, the school is a single-story cinder block
building with a 2 percent slope built-up asphalt roof. The
exterior finish is stucco, and the interior walls are painted.
There are 15 classrooms that average about 25 occupants and about
900 square feet* in area. The eight perimeter classrooms are heated
by a gas fired hot water radiant heating system with one thermostat
controlling all the classrooms. Each perimeter classroom has two
operable 40 x 24 inch" single glazed awning type windows. Except
for the north-facing rooms, the rooms in the school were very warm.
A floor plan of the school is displayed in Figure 1.
Classrooms 3 and 4, located on the south- and west-facing
sides, respectively, were selected for evaluation of the solar
ventilator. Room 5, on the west side, was selected as the control
room. January 1992 radon measurements using 2-day charcoal
canisters over a weekend were 9.7, 11.5, and 11.0 pCi/L in
classrooms 3, 4, and 5, respectively. Alpha track detector (ATD)
measurements by the New Mexico Environment Department from February
to May 1991 were 3.6 pCi/L in room 3; 5.6 pCi/L in room 4; and 2.4
pCi/L in room 5. Lower average radon levels would be expected
during this 3.5 month period since windows and doors were probably
open during the spring.
Subslab radon concentrations in classrooms 3, 4, and 5 were
measured in February 1992 through a 0.375 inch diameter hole using
a Pylon AB5 in a "sniff" configuration. Classroom 3 measured about
800 pCi/L; classroom 4 measured about 600 pCi/L; and classroom 5
measured highest at about 1250 pCi/L.
*1 square foot = 0.093 square meter
**1 inch = 2.54 centimeters
-------
SOLAR FRESH AIR VENTILATOR DESIGN
In order to evaluate the solar ventilators in two classrooms,
two 64 square foot solar collectors were constructed for this
project. A schematic of the ventilator design is displayed in
Figure 2 and discussed in Reference 3. Each collector warms
outdoor air as the air moves through a 30 square inch* cross-
sectional area serpentine channel 20 feet" long. The air moves
under a 5 mil"* black chrome selective surface absorber plate. A
1.5 inch dead air space between the absorber and tempered glass
glazing decreases heat loss from the collector.
To provide the necessary ventilation air to reduce radon, an
airflow of 150 cfm"" was targeted. The volume of air in each
classroom was about 7500 cubic feet*"". An outdoor air supply of
150 cfm would be slightly greater than 1 air change per hour, at
9000 cubic feet per hour. A 265 cfm (free air) blower was initially
selected for each of the ventilation systems, but these were
replaced with 495 cfm (free air) blowers because of the high static
pressures in the systems.
The collectors were built on the school site in February 1992
using the following materials:
preassembled wooden framework
ductboard, 1 inch thick and 4 x 10 feet in area
1.25 inch metal "zee" flashing
3.625 and 1.625 inch metal stud "c" channel
24 x 94 inch selective surface absorber sheets
0.15625 x 46 x 96 inch low iron tempered glass
miscellaneous screws and rivets
high temperature silicone caulking
Material costs for these "state of the art" all-metal
collectors were about $500 dollars each. Costs for additional
materials (such as fans, ducting, diffuser grille, insulation, and
wiring) totaled about $300 for each system. Each system required
about 25 person hours to build and 25 person hours to install. The
collectors are easy to build and have a 20 year life expectancy.
The two collectors were mounted at 70 degrees from the
horizontal at a south-southwest compass orientation, adjacent to
classroom 3. (This installation was considered temporary for the
*1 square inch = 6.45 square centimeters
**1 foot = 0.305 meter
*"l mil = 25.4 micrometers
****! cubic foot per minute (cfm) =0.47 liter per second
***"*! cubic foot = 0.028 cubic meter
-------
purposes of evaluating the ventilators' potential for radon
reduction. For a permanent installation, a roof mount would be
preferred.) The collector support used for this project was a 4 x
4 inch redwood beam supported 4 feet on center with 4 x 8 x 16 inch
solid cinder blocks and secured to the ground with 36 inch steel
stakes. Angle brackets (12 x 12 inches) secured the top of the
collectors to the outside wall of classroom 3.
Outdoor air enters the collector through an 8 inch diameter
furnace filter covered with 0.25 inch grid hardware mesh. A blower
supplies 100 percent outdoor air into a 4 x 12 x 7 inch round sheet
metal boot at the collector inlet. The air is warmed as it moves
through a 20 foot long serpentine pathway with a 1.5 x 20 inch
cross-sectional area under the absorber.
The heated air exits the collector via a 4 x 14 inch metal
boot connected to an elbow, and then flows into 8 inch diameter
insulated flexible ducting. (In a permanent installation, sheet
metal ducting would be preferred.) The flexible ducting is
insulated with 1 inch foil-faced duct wrap to provide additional
insulation and weather resistance. The ducting from the collector
to classroom 3 is about 20 feet long, and the ducting from the
collector to classroom 4 is about 60 feet long. From the ducting,
the air enters the classroom through a 2 x 2 foot diffuser
installed vertically in the outside wall. The system was
originally designed with a remote bulb thermostat to activate a
1250 watt in-line duct heater when air temperature dropped below 60
degrees Fahrenheit;* however, this additional heat was not needed
so the heater was removed since it contributed to pressure loss.
A timer was used to control fan operation. The timer was set
to operate Monday through Saturday, 7:00 a.m. to 4:00 p.m. On
Sunday the system was off to obtain background data with the
classrooms unoccupied.
Teachers in classrooms 3, 4, and 5 were instructed to minimize
open doors and windows during the testing period. Impromptu site
inspections indicated that this instruction was not consistently
followed. This was because the existing heating, ventilating, and
air-conditioning (HVAC) system in the school often keeps the rooms
on the south and west sides of the building too warm.
System airflow measurements were made on April 1 and repeated
May 21, 1992. The airflows into classroom 3 averaged 165 cfm, and
the airflows into classroom 4 averaged 130 cfm. These airflows are
consistent with the longer duct run to classroom 4 (60 versus 20
feet).
*°F = (9/5 °C) + 32
-------
Indoor-to-outdoor differential pressure measurements were made
with a micromanometer. Baseline measurements made in February
averaged - 0.008 inch water column" in perimeter classrooms 1
through 8. These measurements were repeated in April and May with
the ventilators operating. However, conditions were windy at both
times, and it was difficult to ascertain the effect of the
ventilator on the indoor-to-outdoor differential pressure. The
effect of the ventilator on the room-to-subslab differential
pressure (typically less sensitive to wind effects) is discussed
later in this paper.
The measured airflows to the classrooms, together with the fan
manufacturer's performance curve data (airflow vs. static
pressure), indicate that the static pressures are 0.94 and 0.98
inch water column in classrooms 3 and 4, respectively. Static
pressures in the system are caused by:
(1) The resistance to air movement through the collector
baffles;
(2) 90 degree turns in the air pathways in the ducting;
(3) Resistance of the spiral duct walls of the flexible
ducting;
(4) Resistance of the electric coil heater in the airstream
of the duct heater (this was subsequently removed);
(5) Flexible ducting from collector to diffuser (in a
permanent system, sheet metal ducting would be used); and
(6) The diffusing air grille in each classroom.
EVALUATION OF SOLAR VENTILATOR
Classrooms 3, 4, and 5 were instrumented with a datalogger to
collect continuous radon concentration, carbon dioxide
concentration, temperature, and differential pressure measurements
from February to May 1992. The datalogging system used by AEERL is
described in Reference 4.
Data collected during February and March 1992 indicated that
radon levels were frequently below 4 pCi/L in both the ventilated
rooms (classrooms 3 and 4) and the control (classroom 5) whenever
the school was occupied. Site inspections and analyses of the
continuous data (for example, only slightly higher carbon dioxide
levels in classroom 5 during occupied periods) implied that the
windows were frequently opened during occupied periods, interfering
with the evaluation of the solar ventilator. As a result, the most
appropriate data for simulating the effect of the ventilators under
typical winter conditions were collected during spring break in
April when the building was unoccupied. The measurements during
this week form the basis of much of the discussion below.
1 in. H,0 = 2.5 kPa
-------
Figure 3 shows that background radon levels — measured on
April 20 after the ventilators had been off for about 32 hours —
were about 5 pCi/L in classrooms 3 and 4. Background levels during
this same period in classroom 5 (the control) were about 6 to 7
pCi/L. It is probable that the radon levels in classrooms 3 and 4
were not able to build back up to background levels in the 32 hours
that the ventilators were off. Unfortunately, long-term continuous
radon data were lost.
During the day, while the ventilators are operating and the
building is closed, radon levels in the two ventilated classrooms,
3 and 4, are reduced to below 2 pCi/L. Radon concentrations in
classroom 5 were always above 5 pCi/L, and sometimes were above 10
pCi/L during this week.
For comparison, radon data for the following week — while
school was in session — are shown in Figure 4. Radon reduction in
classrooms 3 and 4 presumably occurred due to operation of the
ventilator and by opening windows, as radon levels during the day
were well below 1 pCi/L. Radon reduction in classroom 5 (the
control) was probably a result of open windows during the day. On
Saturday, May 2, radon levels were still well below 1 pCi/L in
classrooms 3 and 4 while the ventilator was operating and the
school unoccupied. Levels in the control room were about 8 pCi/L
during this time.
The radon data in Figures 3 and 4 often show that, when the
ventilators are operating in classrooms 3 and 4 and the school is
not occupied, radon levels in classroom 5 increase. This is
particularly apparent observing the radon peaks in classroom 5 on
April 20, 21, 25, and May 2. It is not known whether these peaks
are coincidental or whether the ventilators in classrooms 3 and 4
actually increase radon levels in classroom 5.
The indoor-to-outdoor differential pressure measurements
indicated that there was very little pressurization in the
ventilated classrooms. This is not surprising since the
ventilators supply only about 130 and 165 cfm of outdoor air to
classrooms 4 and 3, respectively. With these relatively low
differential pressures, other factors such as wind speed and
direction, room occupancy schedules, and barometric pressure can
also greatly influence the indoor-to-outdoor differential pressure.
The continuous classroom-to-slab pressure differentials for
the week of spring break under closed building conditions are shown
in Figure 5. These data indicate that room 3 was slightly
pressurized relative to the subslab when the ventilators were
operating. The room-to-subslab pressure in room 3 decreased on
Sunday when the ventilator was off. Room 4, with less outdoor air
supply, was not consistently pressurized, maintaining a pressure
close to neutral most of the time. The control room was
consistently under a negative pressure throughout the week.
-------
Analyses of the carbon dioxide data for the week of March 23
(Figure 6) show slightly higher levels in classroom 5. These data
also show that the carbon dioxide levels took longer to level off
at the end of the day in classroom 5 than in classrooms 3 and 4
(with the ventilators). This is probably because some dilution
occurred in all three classrooms during the day when the windows
were frequently opened; however, classrooms 3 and 4 also had the
additional ventilation provided by the solar ventilators so that
carbon dioxide levels fell rather quickly once the source (the
occupants) was removed. This was because the ventilators operated
for about 1 hour after the occupants left.
Background carbon dioxide levels indicate some variation
between the three rooms. Since background carbon dioxide levels
should be the same in all three rooms, these data imply a
calibration problem. The manufacturer of the carbon dioxide
monitor recommends calibrating the instruments to 300 ppm in
outdoor air. If the background levels for all three rooms are
adjusted so that background levels are 300 ppm, these data still
show that carbon dioxide levels are higher in classroom 5. In
fact, spikes above the 1000 ppm guideline frequently occurred in
classroom 5 and only occasionally in classrooms 3 and 4,
demonstrating the added benefit of this controlled outdoor air
supply.
The temperature data are not presented here since the room
temperatures in the classrooms are influenced by many factors in
addition to the ventilators. These include: existing heating
system, south-southwest room orientation, and opening of windows
during occupied periods.
CONCLUSIONS
An objective evaluation of the ability of the solar fresh air
ventilation system to reduce indoor radon levels and improve indoor
air quality was complicated by occupants' opening the windows in
both the ventilated and control classrooms. During the occupied
periods, radon levels were well below 4 pCi/L in all three
classrooms because occupants frequently opened windows in the
school, even during cold weather. Although the open windows do
have a positive effect on radon reduction, this is not a cost-
effective long-term radon (or thermal comfort) control strategy.
Although radon levels were consistently low in all three
classrooms, carbon dioxide levels showed that some additional
ventilation benefits were achieved in the two classrooms with the
ventilators.
To simulate the effect of the ventilators under closed school
conditions — that one would normally expect during the winter in
a cold climate — radon levels for all three rooms were compared
during spring break when the building was unoccupied and windows
were closed. Results during this period indicate consistently
8
-------
lower radon levels in the two classrooms with the solar ventilators
than in the control room.
In retrospect, this school was not an ideal candidate to
evaluate the solar fresh air ventilation system because the
occupants frequently opened the windows even without the
ventilators. The teachers were asked to keep the windows and
classroom-to-hallway doors closed as much as possible during the
testing, but certainly not to compromise the comfort of the
students. Switches on the operable windows and doors would have
helped to monitor this condition. As a long term solution to the
over-heating problem (caused by the existing heating system), the
researchers recommended that the school maintenance personnel
adjust the building HVAC system to better adjust to temperatures
throughout the building. For example, rooms on the south and west
sides are typically very warm, whereas rooms on the north side are
sometimes cold. It was also recommended that a permanent supply of
conditioned outdoor air be delivered to all occupied rooms in the
school. This will help to address comfort, radon concentration, and
indoor air quality.
In order to approach the ASHRAE guideline of 15 cfm per person
(about 375 cfm per classroom) , the air supply of the existing
ventilators would need to be increased. Using larger diameter
sheet metal ducting instead of flexible ducting, decreasing duct
length, and using a larger fan would all help to increase airflow
into the classrooms.
The use of single classroom solar ventilators in a school with
several rooms with elevated radon levels is not the best
application of this system. One 64 square foot collector per
classroom would be too expensive, cumbersome, and aesthetically
unpleasing to use for radon control throughout the building. If a
school has special construction or site conditions that preclude
ASD, and only a few rooms with high radon levels, this system might
be more practical.
A better application for this system might be to supply
tempered outdoor air to the return plenum of a central HVAC system
serving several classrooms. The limited ability to preheat outdoor
air is a real problem in schools in cold climates. As a result,
outdoor air supplies are often cut off or restricted during very
cold weather to avoid freezing the heating coils and/or to reduce
energy costs. Use of the solar collector to preheat this air would
help to maintain a consistent outdoor air supply throughout the
year.
-------
REFERENCES
1. Leovic, K.W., A. B. Craig, and D. W. Saum, The Influences of
HVAC Design and Operation on Radon Mitigation of Existing
School Buildings. IN: Proceedings of ASHRAE IAQ '89. The
Human Equation: Health and Comfort. San Diego, 1989. (NTIS
PB89-218762).
2. Radon Measurements in Schools: An Interim Report. U.S.
Environmental Protection Agency's Office of Radiation
Programs, EPA-520/1-89-010 (NTIS PB89-189419). March 1989.
3. Kornher, Steve and Andy Zuagg. The Complete Handbook of Solar
Air Heating Systems. Rodale Press, Emmaus, PA, 1984.
4. Harris, D.B. and B.E. Pyle, Data Logging Systems for
Monitoring Long-Term Radon Mitigation Experimental Programs in
Schools and Other Large Buildings. Presented at the 85th
Annual AWMA Meeting, Kansas City, MO, June 21-26, 1992.
ACKNOWLEDGEMENTS
The authors would like to express their appreciation to the school
officials in the Taos Municipal School System in New Mexico.
Appreciation is also extended to EPA and Acurex staff for their
invaluable assistance.
10
-------
C.R. #6
C.R. #7
_J
C.R. #6
Exit
A
I
<5
CD
o
o
<«Exit
Closed Corridor ToGym»»»» ,'
C.R. #5
1
C.R. #4
C.R. #9
Library
C.R.
Mech.
Sto.
Girts
/N*/Vk*1
C.R. #11
Boys
M W Nurse
Special Ed.
Reading Lab
Sio.
3
rincipa
Reception
C.R. #12
Workroom
C.R. #3
C.R. #2
:T
Exit>» Covered Entry
C.R. #1
N
\
Figure 1. Floor plan of school,
11
-------
Southwest Facing
Wall of School
Electric Heater
2 x 2 foot (optional)
diffuser
\_f
8 inch diameter
flexible ducting
Classroom Windojc
Fan
•'>
«
,-*
\
D
VN
\
i
64 square
^ foot solar
"^ coll
ector
(arrows show
air flow)
Grade level
Figure 2. Plan view of solar fresh air ventilator.
12
-------
20
Ventilator On Time, x-axis
(Mon to Sat, 7:00 a.m. - 4:00 p.m.)
0
04/20/92 04/21/92 04/22/92 04/23/92 04/24/92 04/25/92 04/26/92
Mon Tues Wed Thur Fri Sat Sun
DATE (Midnight)
Figure 3. Continuous radon levels under
closed/unoccupied conditions.
13
-------
12
10
8
w
5
Ventilator On Time, x-axis
(Mon to Sat, 7:00 a.m. - 4:00 p.m.)
2 -
0
04/27/92 04/28/92 04/29/92 04/30/92 05/01/92 05/02/92 05/03/92 05/04/92 05/05/92
Mon Tues Wed Thur Fri Sat Sun Mon Tues
DATE (Midnight)
Figure 4. Continuous radon levels under occupied conditions.
14
-------
0.3
0.2
0.1
I
W
•-0.1
g
-0.2
-0.3
-0.5
-0.6
Room 3-to-Subslab
Room 4—to—Subslab
Room 5_to-Subshb
CMMI •>•
04/20/92 04/21/92 04/22/92 04/23/92 04/24/92 04/25/92 04/26/92
Mon Tues Wed Thur Fri Sat Sun
DATE (Midnight)
Figure 5. Continuous differential pressure measurements
under closed/unoccupied conditions.
15
-------
3000
-"*-*""•-^-——L
03/23/92 03/24/92 03/25/92 03/26/92 03/27/92 03/23/92 03/29/92
Won Tues Wed Thur Fri Sat Sun
DATE (Midnight)
Figure 6. Continuous carbon dioxide measurements
under occupied conditions.
16
-------
XP-2
CHARACTERISTICS OF SCHOOL BUILDINGS IN THE U.S.
by:
Harry J. Chmelynski
S. Cohen & Associates, Inc.
McLean, VA 22101
Kelly W. Leovic
Air and Energy Engineering Research Laboratory
Environmental Protection Agency
Research Triangle Park, NC 27711
ABSTRACT
A subsample of 100 schools from the Environmental
Protection Agency's (EPA's) National School Radon Survey were
visited to obtain information on building structure, location of
utility lines, and the type of heating, ventilating, and air-
conditioning system. Information for each school was entered
into a database to determine the relative proportions of physical
characteristics of the U.S. school building population. The
results will be used by EPA to guide future radon mitigation
research in schools. The building characteristics will also be
correlated with school radon levels to identify any relationships
between the physical characteristics and radon levels.
This paper has been reviewed in accordance with EPA's peer
and administrative review policies and approved for presentation
and publication.
-------
INTRODUCTION
The Environmental Protection Agency's (EPA's) Radon
Mitigation Branch (RMB) conducts research and development on
reduction of indoor radon levels. To help guide future radon
research in schools and better focus technical guidance
documents, RMB conducted a literature search to find information
that quantifies the physical characteristics of U.S. school
buildings. Information specific to radon mitigation research in
schools was not found in any existing reports or databases. In
fact, according to a 1989 publication sy the Education Writers
Association (1), "Nationally, not even a marginally adequate data
base about school facilities exists.... Several national groups
have conducted surveys of school facilities, but these tend to be
either outdated or incomplete." As a result, RMB chose to
characterize the U.S. school building population using a sample
of schools from EPA's National School Radon Survey (NSRS).
The schools are a nationally representative random sample
selected for the NSRS by EPA's Office of Radiation Programs
(ORP). To record the necessary information, a building
characteristic profile sheet was completed for each of a sample
of the schools by RMB staff engineers and selected contractors
during 1991 and 1992.
This paper discusses the random sample selection procedures,
describes the information collected on the building
characteristic profile sheets, summarizes some of the results
recorded on the school profile sheets, compares the results with
those observed in RMB's research schools, and presents the
statistical limitations of this study. All analyses from this
project will ultimately be summarized in an EPA report.
SAMPLE SELECTION PROCEDURES
The NSRS consists of two independent samples: (1) a large
sample of approximately 1,000 schools where all-ground contact
rooms were measured with charcoal canisters, and (2) a smaller
sample of 101 schools where all occupied rooms were measured with
both alpha track detectors (ATDs) and charcoal canisters. This
smaller sample was selected independently of the larger sample.
The schools were drawn randomly from lists of schools in 25
geographical areas called Primary Sampling Units (PSUs). These
25 PSUs were randomly selected for the NSRS from the 125 PSUs
used previously by EPA for the National Residential Radon Survey.
ORP's use of these residential PSUs in selection of schools for
the NSRS is intended to permit comparison of residential and
school building radon concentrations in these PSUs.
-------
The 125 PSUs used for the Residential Survey were selected
from a list of counties or county-equivalents covering the entire
U. S., except for portions of Alaska and all territories and
possessions. This list was partitioned in 22 strata, developed
to guarantee proportioned sample sizes in each of the 10 EPA
regions. Within each region, counties were assigned to one of
three radon potential categories: High, Medium, or Low. The
assignment of states and substate areas to radon potential
categories is summarized for each region in Table 1. The number
of residential PSUs selected for the Residential Survey is shown
in the far right column.
Within each of the 25 NSRS PSUs selected randomly from Table
1, approximately 5 public schools were randomly selected for
inclusion in the NSRS ATD/canister sample, resulting in a total
of 125 schools. This small sample of schools represents a random
sample of the 78,715 U. S. public school population in 1988 (2).
For the NSRS, radon was measured (using both ATDs and
charcoal canisters) in 101 of the 125 schools in the sample. The
remaining schools either refused to participate or were unable to
decide to participate within the time frame allotted for
placement of the ATDs. One of the 101 schools did not
participate in the profile, resulting in a sample of 100 schools
for our study. The locations of the 100 participating schools are
shown in Table 2.
DESCRIPTION OF PROFILE SHEETS AND DATA ENTRY
A three-page profile sheet was developed for this project
for on-site characterization of the structure, utility
penetrations, types of heating, ventilating, and air-conditioning
(HVAC) equipment, and other building features pertinent to radon
diagnostics and mitigation. Because many schools have several
contiguous structures often constructed at different times and
each with its own unique characteristics, the profile sheet was
completed separately for each structure. In a few cases, where
the structures are not contiguous but are campus-style school
complexes, profile sheets were completed for each distinct
structure in the school, unless all were of the same vintage and
construction type.
Where available, building plans were examined to determine
structure and HVAC system information that is not always
available through on-site observation. Following inspection of
the building plans, the school was visited to verify information
on the plans and to collect any additional profile sheet
-------
information that was not on the plans. Complete sets of
construction plans were available for only 40% of the structures.
When the plans were not available, the profile sheet was
completed based on discussions with school personnel and the
judgement of the researchers.
Distribution of the profile sheet responses into the
categories used for data analysis required reducing detailed
responses to shorter/ categorical responses for many of the
profile sheet questions. The original responses for each school
were entered into a DBase IV file along with the shorter
categorical responses used for the statistical analyses.
Because many of the schools have a number of distinct
structures, it is difficult to generally describe the entire
school for a given characteristic, except rarely when all
structures have the same characteristic. For example, in a
school with two additions to the original building, two of the
three buildings might be slab-on-grade and the third building a
basement. Each of the individual buildings would be treated
separately on the profile sheet. Therefore, no attempt is made
to calculate percentage distributions based on the number of
schools in each category. Instead, distributions are calculated
both in terms of the number of sample structures and in terms of
structure area.
A sample of the results is discussed in the following
section. Statistical limitations of the study are contained in
the final section. Detailed results of the complete analysis for
this project will be included in a final project report.
PRELIMINARY RESULTS OF SCHOOL BUILDING PROFILE
The sample of schools selected for this profile are
nationally representative. However, due to the small sample
size, extrapolation of estimates based on the sample statistics
to the national population of schools involves some degree of
sampling error. The standard deviations due to sampling errors
for reported percentages range from 2.5 to 5.5 percentage points
for population estimates of 5% and 50%, respectively. The 95%
confidence intervals for these estimated population percentages
thus range from +/- 5% points to +/- 11%. Due to the large
confidence intervals, small differences (less than 10 percentage
points) in reported population percentages may not be significant
at the commonly used 95% level of significance. These
statistical limitations are discussed in detail in the next
section.
-------
The results presented in this paper are the actual
proportions of the school characteristics for the nationally
representative sample of 100 schools. For most of the
characteristics, the results are presented both in terms of the
percentage of the number of structures and in terms of area. The
discussion is grouped into structural characteristics and HVAC
system characteristics.
Where available, comparisons from RMB's 47 research schools
are presented. Although the RMB research schools do represent a
biased sample in that they are located in radon prone areas,
comparisons of these two samples are helpful in observing trends.
Structural Characteristics
The schools used for this study typically contain two or
three unique structures. The distribution of structures by year
of construction is shown in Figure 1. Nearly half of the school
structures were built between 1950 and 1969, with about 20% built
before and 30% built after. This distribution is consistent with
the survey conducted by the Education Writer's Association that
found that more than 50% of the schools in use today were
constructed during the 1950s and 1960s (1). By comparison, 46%
of the schools in our profile were constructed during this
period.
Over 90% of these school structures have a conventional
classroom design, with a corridor that has classrooms on either
side. Approximately 5% have a campus-type design, with a number
of individual buildings.
The distribution of school structures in terms of total area
is shown in Figure 2. Approximately 45% of these structures are
less than 10,000 square feet*, probably because many of the older
buildings have had additions to the original building.
Approximately one out of eight structures (12.3%) have more than
50,000 square feet, ranging to over 600,000 square feet in one
school structure.
For radon reduction research, the substructure of a school
is of interest. As seen in Figure 3, slab-on-grade substructures
are most prevalent, accounting for 72.6% by structure and 51.6%
by area. Crawl spaces and basements account for 10.3% and 6.7%
of the structures, respectively. These results are consistent
with RMB's research schools which are 70% slab-on-
* 1 square foot = 0.093 square meter
-------
Figure 3 also shows that about 10% are combination
s, such as slab-on-grade and crawl space in the same
*h« r^o <- ComParin9 the Percentage by number of structures with
the percentage by area, there is a tendency for a crawl space to
be constructed in conjunction with either a slab-on-grade or a
basement. These two categories account for approximately 8% of
the number of structures, but almost 35% of the area. More than
two-thirds of all school structures consist of only one floor?
Location of subslab footings and the presence of subslab
aggregate are very important in designing a subslab
depressurization (SSD) system for radon mitigation. As seen in
Figure 4, gravel (which improves the SSD system effectiveness)
was indicated on the plans for about 45% of the structures with
information available. Many of the structures did not indicate
the subslab material on the plans or the plans were not
available. The remaining structures indicated fine-grained
material (such as sand or earth) under the slab. The location
SS ?USbSr °* sub?lab footings is also important in determining
subslab barriers for SSD systems. Figure 5 shows that over half
of the structures have no internal footings (typically post-and-
beam construction, facilitating SSD). However, 24% hive footings
between classrooms and along the corridor, complicating a SSD
system installation.
Location of utility lines is also important since utility
lines located under the slab or in a subslab tunnel can serve as
a major radon entry route. The data in Figure 6 show that about
Hr^oif? *K® structures (and area) have overhead utility lines.
However, a third (a quarter by area) also have utility lines in
either a tunnel or subslab. Utility tunnels were present in one-
third of RMB's research schools (3), and tended to be more
prevalent in certain school districts than others.
HVAC Svstem Characteristic
Research on the use of HVAC systems for radon reduction
includes a large portion of both RMB's and ORP's radon research
Jvn^S S? m*7^AS \resuit' i1:.is important to quantify the various
types of HVAC systems found in existing U.S. school buildings.
The distribution of types of HVAC equipment in the sample
schools is shown in Table 3. The categories in this table are
mutually exclusive. Only one-third of all schools have a single
type of equipment in all structures. Most often, this is a
,
fl K Radiant heat only <6*) or fan coils only
(8%) or both (2%) are present in 16% of the surveyed schools,
indicating that the other 84% of schools have either central HVAC
or unit ventilators capable of delivering conditioned outdoor
-------
air. The remainder of the schools have various combinations of
central HVAC, unit ventilators (UVs), fan coils (FCs), and
radiant heat (RAD). In some schools, other radiant heat systems
have been abandoned (RAD-NU) for heating, but their presence must
be considered from a radon perspective.
In Table 4, the distribution of the four basic types of HVAC
systems is tabulated by number of schools, count of structures,
and structural area. These categories are not mutually
exclusive, due to the occurrence of combinations of HVAC systems
within a school structure. Central HVAC is the predominant
system, occurring in 71% of the schools and 52% of the
structures, either alone or in combination with other equipment.
Radiant heat, including abandoned systems, is the second most
common system, when counted by schools (56%) or by structures
(44%). In terms of structural area, radiant heat systems are as
prevalent as central HVAC systems. Unit ventilators and fan coil
systems (with no ventilation capability) are less common than
central HVAC and radiant heat systems, each occurring in
approximately 30% of all structures and 40% of all schools.
Considering the combinations of HVAC systems within a given
school, 45% of RMB's research schools have central air handling
systems; 43% have unit ventilators; 30% have radiant heat; and
11% have fan-coil units (3). Only radiant heat (11%) or only
fan-coil units (6%) are present in 17% of the research schools,
indicating that the other 83% have some type of installed HVAC
system that can deliver conditioned outdoor air.
The school profile sheets contain more detailed information
concerning the location of air supply and return ducts, the
location of unit ventilators and fan coils, and the types of
radiant heating systems. The most common location of air supply
and return ducts in structures with central HVAC systems is in
the ceiling or suspended overhead. However, ducts located in
corridors, basements, or tunnels occur more often in larger
structures.
The most common location of unit ventilators and fan coils
is along the outside wall. Radiators are used in most
structures, but baseboard systems amount for more structural
area.
STATISTICAL LIMITATIONS
Because of the random selection of NSRS ATD/canister schools
within the 25 selected residential PSUs, the sample of profiled
schools is nationally representative. However, extrapolation of
-------
the survey estimates to the national population of schools must
reflect the magnitude of sampling errors expected for a survey of
this size. Sampling errors should also be considered when the
relative proportions of two response categories are compared.
Clustering of the sample schools within the residential PSUs
results in some loss of sampling efficiency compared to a truly
random sample of schools for this survey. An additional loss of
sampling efficiency arises due to non-response adjustments to the
sampling weights, which will be made when the final weights are
provided by ORP. At this time, the sampling weights for the NSRS
ATD/canister sample have not been determined.
The loss in sampling efficiency can be explained in terms of
a design factor (DF) for the survey, defined as:
DF = »
n
where N represents the actual sample size (100) and n represents
the reduced effective sample size for this design. The effective
sample size is defined as the required size for a truly random
sample to generate the same sampling errors. Because of the
random selection of residential PSUs and the random selection of
schools within these PSUs, we estimate that a worst case DF would
not exceed 1.25. For this assumption, the effective sample size
is approximately 80.
The standard error (SE) of an estimated population
percentage (P) is given by:
SE =
P ( 1 - P )
n
N
-=-=, = effective sample size
DF
Knowledge of the standard errors of the estimated
percentages permits determination of approximate 95% confidence
intervals (CIs) for the reported estimates. An estimated
population percentage P has a 95% CI extending approximately two
standard errors on either side of the estimate. Thus, the
approximate 95% CI for a population percentage estimate P would
be the interval (P - 1.96SE, P + 1.96SE).
The estimated 95% CIs for various estimated population
percentages are reported in Table 5 for the specified effective
sample size. The 95% CI for an estimate of 5% extends
8
-------
approximately from 0 to 10%. For a population percentage of 20%,
the 95% CI extends approximately from 10 to 30%. Similarly, an
estimate of 80% has a 95% CI ranging approximately from 70 to
90%.
The 95% CIs reported in Table 5 are relatively large, due to
the small effective sample size of approximately 80. The size of
these CIs should be considered when comparisons are made between
the reported population proportions for two different response
categories. For example, if outcome "a" is observed in A% of the
100 schools, and outcome Mb" in B% of the schools, then A + B <=
100% with the inequality applying if more than two outcomes are
possible. To determine if A is significantly less (or greater)
than B, the standard error of the difference (A - B) is
determined by:
SE( A- B) = i " - " - < A -
n
To test the hypothesis that A is greater than B (or A is
less than B), the difference between A and B should be
significantly greater (less) than 0. Hence, the quantity (A - B)
should be more than two standard errors away from 0, indicating
that the difference is significantly positive (A greater than B)
or significantly negative (A less than B). Regions where A is
significantly greater (or less) than B are shown in Table 6. In
this table, the symbol « denotes that A is significantly less
than B, and » denotes that A is significantly greater than B (at
the 95% significance level) as determined by:
| A - B | > 1.96 SE( A - B )
For example, an estimate of 8% is significantly less than an
estimate of 20%, but it is not significantly less than an
estimate of 16% because the SE of the difference between
estimates of 8 and 16% is about 5 percentage points. Thus the
difference of 8 percentage points is less than 1.96 SE, and the
difference is not considered significant at the 95% significance
level.
CONCLUSIONS
The school profile sheets contain many significant findings
concerning the distribution of school building characteristics.
The profile sheets provide evidence of the variety of building
-------
structures and HVAC equipment found in typical schools. The age
of a school, number and size of different structures, type of
substructure, location of utility lines, and types of HVAC
equipment vary widely in the sample schools.
The substructure of a school has important implications for
radon diagnostics and mitigation. Determination of substructure
detail depends on locating building plans, which were available
for only half of the structures. Where identified, subslab
materials were almost evenly divided between gravel and fine-
grained material such as earth or sand. Internal footings are
found in half of the structures, with footings under both
corridor and classroom walls in one-quarter of the structures.
Utility lines may enter the building at a wide variety of
locations, including tunnels, subslab penetrations, and overhead.
In a few schools, older unused radiant heating systems may
provide additional radon entry routes.
Commonly encountered structural characteristics include
slab-on-grade with a conventional school building design with a
single floor. Central HVAC is common, but often combined with
other HVAC systems within a single school. Where applicable,
central HVAC ductwork is usually located in the ceiling or
suspended overhead. Radiant heat, using baseboard or radiator
systems, is the second most common HVAC system. Unit ventilators
and fan coils also present in many of the schools are most often
located along outside walls, but may be in the ceiling, suspended
overhead, along an inside wall, or on the roof.
REFERENCES
1. Education Writers Association, Wolves at the Schoolhouse
Door. Washington, DC, 1989.
2. Quality Educational Data (QED), Inc., Denver, CO, 1988.
3. Leovic, K.W., A.B. Craig, and D.B. Harris, Update on Radon
Mitigation Research in Schools. Presented at the 1991 AARST
Conference, Rockville, MD, October 1991.
10
-------
TABLE 1.
ASSIGNMENT OF RADON POTENTIAL CATEGORIES
FOR RESIDENTIAL SURVEY
EPA
Region
8
10
Radon
Potential
Category
^—s^^^^—^.
High
Medium
Low
High
Medium
Low
High
Medium
Low
High
Medium
Low
High
Medium
Low
High
Medium
Low
High
Medium
Low
High
Medium
Low
High
Medium
Low
High
Medium
Low
State/Substate Area
ME, NH, VT
MA, CT, RI
None
Northern NJ
NY
Southern NJ
PA, Western MD, WV, Western VA
None
DE, Central and Eastern VA,
Eastern MD, DC
Western NC, Western SC,
Northern GA, Northern AL,
Eastern TN
KY, Western and Central TN
Central and Eastern NC,
Eastern SC, Southern GA,
Southern AL, MS, FL
MN, WI, IL, IN, OH
None
MI
NM
OK, Western and Central TX,
Northern AR
LA, Southern AR, Southeastern
TX
NE, IA
KS, MO
None
MT, WY, UT, CO, ND, SD
None
None
NV
None
CA, AZ, HI
AK, ID
None
WA, OR
No. of
PSUs
Selected
=
3
5
0
4
8
2
15
0
2
3
7
30
0
2
2
0
8
2
0
srn
2
5
3
4
3
0
6
0
0
11
-------
TABLE 2. LOCATION AND CHARACTERISTICS OF PARTICIPATING SCHOOLS
IN THE SCHOOL PROFILE SAMPLE
EPA
Region
1
2
3
4
5
6
7
8
9
10
State
Massachusetts
New Jersey
New York
Virginia
West Virginia
Mississippi
Tennessee
Illinois
Ohio
New Mexico
Oklahoma
Texas
Kansas
Nebraska
Utah
Arizona
California
Washington
Total
No. of
Schools
3
5
7
5
5
7
5
4
4
4
5
11
5
4
5
4
13
4
100
Type of Schools*
K-6, K-6, K-6
K-6, 7-12, K-6, K-6, K-8
7-9, K-6, K-6, P-3, K-6,
K-6, K-6
6-8, 6-8, K-6, K-6, K-6
K-6, K-6, 7-12, K-6, 7-12
6-8, P-K, 10-12, 6-8,
K-6, K-6, K-6
6-8, K-8, K-6, K-6, SP-ED
K-8, K-8, K-6, 9-12
SP-ED, 6-8, 7-9, K-6
K-6, K-6, K-6, 6-8
K-6, 7-12, K-6, 6-8, 9-12
9-12, K-6, 6-8, K-6, K-6,
6-8, K-6, K-6, K-8, 9-12,
K-6
9-12, 7-12, K-6, 7-12, K-6
7-12, 6-8, K-6, K-6
K-6, K-6, 6-8, K-6, 9-12
9-12, K-6, K-6, 6-8
K-6, K-6, K-6, 7-9, K-6,
K-6, P-K, K-6, 6-8, K-6,
9-12, K-6, K-6
K-6, K-6, K-6, 6-8
P - Primary
SP-ED = Special Education
12
-------
TABLE 3. DISTRIBUTION OF TYPES OF HVAC SYSTEMS BY SCHOOL
Type of System; No. Schools
Central HVAC only (HVAC) 13
Unit ventilators only (UV) 7
Fan coils only (FC) 8
Radiant heat only (RAD) 6
UV/RAD-NU* 1
HVAC/UV/RAD-NU 1
HVAC/UV/FC/RAD 3
HVAC/UV/FC/RAD-NU 1
HVAC/UV/RAD 16
HVAC/FC 7
HVAC/RAD 8
HVAC/FC/RAD 12
HVAC/UV 7
HVAC/FC/RAD-NU 2
UV/FC 1
FC/RAD 2
UV/RAD 4
HVAC/UV/FC 1
Total Number of Schools 100
* NU = not used
TABLE 4. TYPE OF HVAC EQUIPMENT BY NUMBER OF SCHOOLS, NUMBER
OF STRUCTURES, AND STRUCTURAL AREA
Count by
Count by Count by Structural Area
Type of System Schools Percentage Structures Percentage (square feet) Percentage
Central HVAC 72 71.3 120 51.5 3643604 67.1
Radiant heat 57 56.4 103 44.2 3659727 67.4
Unit ventilators 43 42.6 70 30.0 1734323 31.9
Fan coils 38 37.6 61 26.2 1883987 34.7
Note:' Unknown types are not included in analysis.
k Parcents add to more than 100% due to the possibility
or more than one system for a structure.
13
-------
TABLE 5. APPROXIMATE 95% CONFIDENCE INTERVALS FOR ESTIMATES
Estimated
Population Percentage
Expected
95% Confidence Interval1
P * 5%
P = 10%
P - 20%
P - 50%
or P -
or P =
or P =
95%
90%
80%
(P -
(P -
(P -
(P -
4
6
8
11
.8%,
.6%,
.8%,
.2%,
P H
P H
P H
P H
K 4.
K 6.
h 8.
h 11.
8%)
6%)
8%)
2%)
The actual confidence intervals surrounding the estimates
will not be symmetric except for the case P - 50%.
TABLE 6. REGIONS OF SIGNIFICANT DIFFERENCE BETWEEN TWO
POPULATION PERCENTAGE ESTIMATES, A AND B.
PERCENTAGE B
4 8 12 16 20 24 28 32 36 40 44 48 52 56 60 64 68 72 76 80 84 88 92 96
4
8
12
16
20
24
28
32
< 36
M 40
O 44
Ł 48
P 52
ft 56
60
64
68
72
76
80
84
88
92
96
g
NOTE: The symbol n»" denotes that percentage A is significantly
greater than percentage B at the 95% significance level; the
symbol "«" denotes that A is significantly less than B; and the
symbol "." denotes that A and B are not significantly different
at the 95% significance level.
14
-------
100
00 -
60 -
TO -
60 -
60
40 -
30 -
20 -
10 -
0
23.2X 23.2X
I3.3X
0.4X
I900-O9 I9M-I9 1920-28
1940-49 I95O-59 1960-69 I970-T9 1950-89
TEAR CONSTRUCTED
Figure 1. Distribution of structures by year constructed,
i
I
o
i
100
90
60
70
60
50
40
50
20
10
0
I5.0X
PIX
6.4X 6.B*
Uqfl»O.OB3*4m
0.9X
o.9X 0.5%
r- - r
0-10 10-20 2O-30 9O-40 40-50 SO-60 60-70 7O-M 6O-90 90-XXIOO-2OO330-SOO 600t
ARC A. COO •<) It
Figure 2. Distribution of structures by area.
15
-------
too
80 -
i
T2.6X
30
20
10
0.4X 0.8*
or«wl bimt dab/crawl b»m«/ef»wl vlcb/pm* »Ub/»>»mt b»ml/»1»b
1
crawl
b*ml (lab/craw! b»m«/e*«wl •tab/pUr* alab/b«ml temt/alab
SUBSTRUCTURE TYPE
Figure 3. Distribution by substructure type
(top % by number of structures; bottom % by area)
16
-------
K>0
10
E
I
2
•and earth gravd grvf-a*nd dirl-ll prvt-fl cndy-tol ckider-fl p*a pvl fil-aand
•and «a»tt»
•UBSLAB MATERIAL
Figure 4. Distribution of subslab material
(top % by number of structures; bottom % by area)
17
-------
2
oc
I
M
eterm corridor
SUBSLAB WALL LOCATION
Figure 5. Distribution of subslab wall locations.
18
-------
too
0.8X 0.5% 0.8X
crawl evrhd b*ml •ubslab evlng tunnel nweti rm o* w«I *Ub crvdvrhd *w*l m»ch rrfauled
100 -|
60 -
eo -
TO -
60 -
SO -
40 -
3O -
20 -
10 -
30.3X
cr«w( ovrhd b*m(
iiUvbd
VTUTY UNE LOCATION
Figure 6. Distribution of utility line locations
(Top % by number of structures; bottom % by area),
19
-------
XP-3
RADON IN SCHOOLS IN WISCONSIN
by: Conrad Weiffenbach and John C. Lorenz
Wisconsin Dep't. of Health and Social Services
Bureau of Public Health
1 W. Wilson St., P.O. Box 309
Madison, WI 53701-0309
ABSTRACT
Three-fourths of the school districts in the state have made
three month winter alpha-track detector measurements. More than
35,000 radon screening measurements have been made in classrooms in
this continuing voluntary program. Results in about 1% of the
classrooms were higher than 4 pCi/1. Follow-up studies using
continuous radon monitors show that ventilation systems reduce
radon concentrations in about half of those rooms to below 4 pCi/1
during the hours they are occupied (8 AM - 4 PM) . Most of the
other rooms with elevated screening levels are underventilated.
The administration of the program, the value of continuous radon
monitor data, and mitigation measures taken and their effectiveness
are discussed.
This paper been reviewed in accordance with the U.S.
Environmental Protection Agency's peer and administrative review
policies and approved for presentation and publication.
-------
INTRODUCTION
The Wisconsin school radon program was established in 1989 by
the Radiation Protection Unit in the Bureau of Public Health,
working with Wisconsin's Cooperative Educational Service
Associations and the Department of Public Instruction. The
Cooperative Educational Service Associations (CESAs) serve schools
in 12 regions into which the state is divided, offering (among
other services) the coordination of environmental work among
buildings and grounds personnel of participating schools, which can
include private schools. The Radiation Protection Unit put out a
request for bids for a large number of alpha track detectors, and
the CESAs coordinated detector purchases by the schools through the
supplier selected by the Radiation Protection Unit. An engineer
from the Radiation Protection Unit traveled around the state in the
Fall of 1989 giving training sessions for personnel of schools
participating in the program.
SCREENING MEASUREMENT METHOD
Screening measurements were made by school personnel following
procedures recommended by the U.S. EPA (1) for alpha track
detectors. Many of the school personnel also received training
from the Bureau of Public Health. The Landauer (Terradex) Radtrack
alpha track detectors (ATDs) that were used were placed one per
occupied room, at or below ground level, or one per 2,000 square
feet in large rooms, for exposures of three winter months. The
detectors were typically hung from ceiling light fixtures.
When the results were received, they were entered into a
relational data base at the Wisconsin Geological/Natural History
Survey. The data base was transferred on magnetic media to the
Bureau of Public Health.
QUALITY ASSURANCE
Duplicates
School personnel were instructed to place side-by-side
detector duplicates in ten percent of the rooms as quality control
measures, but in analysis of data we received from the detector
laboratory, where two monitors are listed as having the same
location, it is evident that in many cases that they were not side
by side. Pairs such as (0.3, 4.5), (0.4, 7.5), and (0.6, 5.5)
-------
together in data from one school, and (55, 1.2); (65, 2.5) together
in another school, with the same indicated locations, indicate that
the information in the data base is not suitable for measurement
quality assessment based on duplicates. Multiple detector
placements are expected in areas (gyms, etc.) with over 2000 square
feet, but it seems school personnel made many experiments and
unnecessary measurements with multiple detectors in rooms. A
scatter graph of 98 measurement pairs, that are labeled
specifically "duplicate" in the data or known from personal
communications to be duplicates, is presented in Figure 1. The
average coefficient of variation (standard deviation / average) for
those 98 pairs is 0.08. The reproducability indicated is good.
Control Blanks
Many detectors labeled as controls in the data base showed
significant radon exposures. This is because many of the schools,
following a procedure outlined in EPA literature for schools (1),
opened the control blanks at the beginning of the measurement
period and sealed them with the metallized stickers coming with the
detectors. Unfortunately the stickers are not radon proof.
The few known properly handled control blanks had results below
detection limits of 0.3 to 0.4 pCi/L.
Spikes
Four detectors were sent for exposures in radon chambers, with
three control blanks, to The U.S. EPA National Air and Radiation
Environmental Laboratory, in Montgomery, Alabama. The exposure was
reported by EPA as 447 +/~ 15 pCi/L - days. The detectors were
forwarded, blind with three different names and a false exposure
time of 91 days (which would result in an indicated 4.9 pCi/L +/~
4%), to Landauer Corporation by the Bureau of Public Health.
Landauer reported the readings as 4.7, 4.5, 5.0, and 4.8 pCi/L,
with less than 0.4 and 0.3 pCi/L for the three blanks.
The standard deviations given for these detector readings and for
most of the school measurement results are about 7% at 4 pCi/L and
11% at 1.5 pCi/L.
Landauer Radtrack alpha track detectors have regularly passed
EPA Radon Measurement Proficiency testing. With our duplicate,
blank, and spike results, we believe the precision, accuracy, and
reliability of these detectors has been satisfactory.
-------
SCHOOL SCREENING RESULTS
For 31,999 screening measurement results that were in the data
base as of October 1991, the average radon level was 1.0 pCi/L and
775 were greater than 4.0 pci/L. These included a large number of
elevated results for detectors placed in unoccupied spaces for
examples storage rooms and tunnels. The average screening radon
level in spaces that have been verified to be occupied is less than
i pci/L. Several thousand results from winter of 1991-92 were
received this spring but are not counted in this report.
Schools with the highest screening results or for which many
of the rooms exceeded 4 pci/L were given the first follow-up
investigations. Of 477 results greater than 4.0 pCi/L that had
been investigated by May 1992, 262 were in unoccupied spaces or
otherwise invalid, and 215 were occupied and called for follow-up
measurements. *
Nineteen screening results were reported at greater than 20
pci/L. Seventeen of these were in unoccupied spaces or had start
and stop date data-entry errors. The highest results for occupied
S^SS *l®re:. a locker room at 36 pci/L which could not be confirmed
with Radiation Protection Unit charcoal canisters, though it had
been confirmed by the school with a second alpha track detector;
and a classroom at 29 pci/L, which had had ventilation system
repairs before we could make a follow-up measurement. Both of
these had follow-up measurement results below 4 pCi/L durina
occupied hours. y
FOLLOW-UP MEASUREMENT METHODS
INTENSIVELY STUDIED SCHOOL
When the first screening results arrived in 1990, we selected
one school with many rooms having elevated screening levels in the
4 and 5 pci/L range for intensive study, to develop follow-up
methods and investigate mitigation strategies. The buildings
supervisor for this school district was very cooperative, as was
his replacement when he retired. This experience was to be applied
for follow-up studies and mitigation for other schools in the
state.
This school has been studied and worked on for over two years
with dozens of alpha track, charcoal, and continuous radon
-------
measurements. The levels have now been reduced to average below 4
pCi/L in all rooms during occupied hours, though some are still
close. We have developed considerable experience that has been
useful in follow-up studies for other schools.
The first follow-up procedure that was studied involved
measuring with charcoal canisters with ventilation system on
continuously for two days, and remeasuring with the ventilation
system off for two days, in an elaboration of the procedure
suggested in U.S. EPA literature on radon measurements for schools
(1). Diagnostic grab samples and charcoal canisters in tunnels and
other non-classroom locations were used, as was a continuous radon
monitor. The continuous radon monitor was found to be the most
useful follow-up measurement tool, as it indicated when, how much,
and how fast levels dropped when the heating/ventilation system
came on in the morning.
It was decided that continuous radon monitors would be the
best way to do follow-up measurements if we could get enough of
them, if they were suitably automated as data loggers, and if the
data could be easily uploaded for analysis and graphing with
computers.
CONTINUOUS RADON MONITORS
Ten continuous radon monitors (CRMs), assembled at the Bureau
of Public Health, have been used in follow-up studies in over two-
hundred school rooms during two winters. They are based on
portable nuclear multichannel analyzers (Canberra Corp., model S-
10), which were surplussed to the Bureau of Public Health by
Commonwealth Edison Corporation of Illinois, and diffusion
scintillation cells, of 5.7 cm inside diameter and 12.5 cm long,
with eight 1.9 cm diameter holes circling each end, manufactured by
the Radiation Protection Unit.
Quality assurance for the data from these detectors is based
on continuing efforts of three types.
a.) Calibration is done by running all ten CRMs side-by-side
in a high-radon basement along with a commercial Pylon AB5 monitor
and charcoal canisters. The Pylon CRM has a flow-through
scintillation cell, and is calibrated with a Pylon flow-through
radon source. It's calibration has been verified with two Femto-
Tech 210F CRMs and an identical Pylon CRM, independantly calibrated
at other institutions. The results from the ten CRMs and the Pylon
are consistent with each other, with variations of about +/- 10% in
the raw counts among them, for which calibration factors adjust
during data analysis.
-------
b.) Background corrections are made by running the monitors
plus the flow-through Pylon CRM, calibrated for background with
nitrogen gas, outdoors. This introduces an uncertainty of less
than 1/4 pCi/L in the overall calibration.
c.) Comparisons of CRM data with charcoal canister radon
measurements are made in every school room in which the CRMs are
deployed: more than 200 schoolrooms in the past two years.
„„ ^ Variations of up to 30% from unity for the ratio of the CRM
48-hour average to the 48-hour charcoal canister result are common.
For nine CRMs, the average ratio for from 7 to 10 classrooms each
™ Wlth,ln 10% of unity. For the tenth it was 1.19. For all ten
CRMS and the 76 classroom measurements with the lowest precision
deviation?116 3Verage rati° is 1'01' +/~ 0.22 (one standard
Several factors contribute to the differences between charcoal
canister results and CRM 48-hour averages, though it is seldom
evident which are significant in the data for individual rooms.
The canister results are typically +/- 10% in precision. The CRMs
are +/- 10% in calibration. For open-faced charcoal canisters, the
desorption of radon is reported (2) to be exponential: exp(-t/T)
with a 24-30 hour characteristic time T, so the result can depend
strongly on the radon level in the final 12 or so hours of the
measurement period, while the CRMs give truer averages. Canisters
results are also susceptible to temperature changes and air
currents, while the CRMs are not affected by these.
In a typical field trip, three or more days of data from each
ot the ten monitors is uploaded to a laptop computer, and they are
transported to different schools and started in the next set of
rooms needing follow-up measurements. No operator or student has
damaged a continuous radon monitor in over 200 classroom
placements, in schools of all sorts, all over the state. The
monitors have been operated by personnel from county health
departments, Milwaukee Public Schools, as well as the Radiation
Protection Unit, but school maintenance personnel are not generally
asked to run them. *
FOLLOW-UP RESULTS, BEFORE AND AFTER MITIGATION WORK
A summary of the results for follow-up measurements is
presented in Table 1.
-------
In more than half of the 206 occupied rooms for which follow-
up measurements have been made with continuous radon monitors, the
radon levels were found to be below 4 pCi/L during occupied hours.
At least 20 of these rooms had had ventilation adjustments and
repairs before follow-up testing was done.
Fifty of the 78 occupied rooms that had been confirmed with
follow-up measurements to be over 4 pCi/L during occupied hours
have been mitigated with work on the heating/ventilation systems,
including balancing, adjustments, repairs, and new construction;
bringing radon levels to below 4 pCi/L during occupied hours, as
indicated in further CRM follow-up measurements.
Twenty-eight rooms, in 7 buildings, may still have radon
levels over 4 pCi/L during occupied hours, averaged through the
school year. For some of these, the schools have not finished
heating/ventilation system adjustments and sealing; for others the
Radiation Protection Unit has not been able to return for a second
set of follow-up measurements after whatever work has been done.
For a few, mitigation via work on the ventilation system has not
been effective enough, and study of the problem is continuing.
TYPICAL DATA FOR CLASSROOMS
Figure 2 shows radon versus time in a classroom in which the
radon is greatly lowered during occupied hours by the automatic
setback heating/ventilation system, 'in CRM data graphs, each data
point has been centered on the hour mark between the middle and end
of the hour collection interval, depending on when the CRM was
started. In the evening and over night, the heat and ventilation
is automatically set back to be minimal, and radon levels climb.
In the morning, the heat and ventilation come on, driving down the
radon levels. Radon in most rooms in this building followed a
similar pattern, and many school buildings exhibit a similar
pattern, though without going as high over night.
A correction, for the decay and buildup of radon progeny in
the diffusion scintillation cell, which is largest for rapid
changes with time like this example, is also shown in the figure.
This correction, involving adding various fractions of the radon
counts for several previous hours to the count for a given hour, is
derived from empirical studies of the response of the diffusion
scintillation cell to step function increases and decrease of radon
in a chamber. When the changes of radon concentration with time
are less rapid, the correction is not important.
The radon in the classroom of figure 2 averages less than 2
pCi/L during occupied hours 8 AM to 4 PM, though the full three day
-------
average is close to 10 pCi/L. This school had done repairs and
balancing for the ventilation system before the CRMs were placed in
it. The alpha track screening measurement result for this room was
29 pCi/L.
Figure 3 is for a classroom in which the major source of radon
was a sub-slab return air duct, in contact with the ground. When
the furnace return air fan is on the duct is below atmospheric
pressure. The radon in this classroom was reduced by installing a
ceiling duct for the return air.
Figure 4 is for a classroom for which one of the return air
paths is through a grill in the door to the hall, while another
path is through return air ducting from the room. The door was
kept closed, then open, on alternate days. Because the hall is
quite negative and the door grill is the major impedance between
the room and hall, when the door is open, the pressure in the room
is more negative, and the radon is more elevated than when the door
is kept closed. The door of this room was usually kept open. Many
schools in the state use halls in return air circuits.
CHARACTERISTICS OF BUILDINGS WITH ELEVATED FOLLOW-UP LEVELS
INTENSIVELY-STUDIED SCHOOL
The building has a system of return air tunnels which had
radon levels of about 7.5 pCi/L, 2 to 3 pCi/L higher than the
classrooms. Since air is recirculated to the furnace through these
tunnels and then distributed to all rooms, they were believed to be
important sources of radon for the rooms.
Among the methods used or attempted to lower the radon were:
increasing ventilation in some rooms, starting the heat/ventilation
cycle earlier in day; balancing the ventilation system; blending
more fresh air into the recirculated mix in the coldest season,
when the dampers had been closed maximally; and a moderate amount
of sealing of openings to soil in the tunnels and associated return
air system. Work that was done over one summer to try to seal the
tunnels against radon entry had little or no impact on radon
levels.
It was found that spring and fall radon values in the building
were lower than winter values. This is to be expected, because the
heating/ventilation system automatically blends more fresh air into
the system when the outdoor temperature is warmer. Averaging radon
levels over the academic year was necessary to achieve levels below
4 pCi/L for some of the rooms.
-------
At present all rooms in the intensively-studied school are
below the 4 pCi/L guideline during occupied hours, though some are
near 4. The heating bill for the school has probably been
marginally increased.
The characteristics of the school appear comparable to those
reported (3) by U.S. EPA researchers for a Colorado school with
return air tunnels which was difficult to mitigate.
OTHER SCHOOLS
About a third of all occupied rooms with elevated screening
levels had been converted to offices or small classrooms from
underventilated locker rooms or storage areas not originally
designed to be occupied, or were actual locker rooms.
Rooms occupied for less than a cumulative hour per day were
considered unoccupied if the radon was just a few pCi/L over 4.0.
A locker room occupied for less than fifteen minutes each hour, for
changing clothes at the beginning and end of each of four periods
per day, is classed as not occupied.
Two of the school buildings with elevated or borderline
follow-up results have been closed in the normal course of events,
not because of radon. Administrators of two schools with confirmed
elevated levels would like to move or replace the buildings, mainly
for reasons not concerning radon, but cannot get the communities to
support these plans.
Forty-three occupied rooms with elevated screening levels are
in school systems which have refused the offer of follow-up
measurements. Some have not returned repeated telephone contacts,
and others have simply declared they don't want the involvement of
the state.
Schools generally have been reluctant to call the Radiation
Protection Unit for follow-up measurements. Some have made 9-month
follow-up measurements with alpha track detectors, using a
procedure suggested in the U.S. EPA literature (1) on radon
measurements in schools. With these they often obtained results
similar to the 3-month screenings. In general the Radiation
Protection Unit contacts schools after receiving data showing
elevated screening levels, and explains about the CRM follow-up
measurements that are available.
Most schools appreciate Bureau of Public Health follow-up
assistance when it is explained that it is at no charge; that the
continuous radon monitors often show reduced radon levels during
occupied hours; that the Radiation Protection Unit has accumulated
-------
experience that can help them deal with elevated levels if they are
confirmed; and that action to reduce levels is not compulsory.
CONCLUSIONS
The voluntary measurement program has given schools a common
and inexpensive screening procedure. Although each year additional
schools make the radon screening tests, it doesn't seem that all
schools will have tested for radon until after testing is required.
CRM follow-up measurements reveal significant reductions of
radon levels during occupied times in many schools with automatic
setback HVAC systems, and produce useful diagnostic information in
some cases.
Quite a few schools have made minor and otherwise required
adjustments to ventilation systems that reduced radon levels. For
some of the schools with moderately elevated radon levels, the
common mitigation strategies are unsatisfactory, and the analysis
of the problem and search for a means of radon control continues.
ACKNOWLEDGEMENTS
The authors would like to thank John Micka, Larry McDonnell,
Teri Vierima, Mike Mudrey, and Tom Shepro, who established the
school radon program, and all the school officials who have
graciously permitted measurements to be made in their buildings.
REFERENCES
1. U.S. EPA. Radon measurements in schools: an interim report.
EPA 520/1-89-010. March, 1989.
2. George, A.C., and Weber, T. An improved passive activated C
collector for measuring environmental 222Rn in indoor air. Health
Physics 58: 583-589, 1990.
3. Leovic, K.W., Pyle, B.E., Borak, T., and Saum, D.W. HVAC
complications and control for radon reduction in school buildings.
Presented at the 1991 International Symposium on Radon and Radon
Reduction Technology, Philadelphia, April, 1991.
-------
TABLE 1. RESULTS SUMMARY FOR SCREENING AND FOLLOW-UP MEASUREMENTS
Alpha track screening measurements, by Oct. 1991 31,999
Alpha track screening results > 4.0 pCi/L 775
No. of these investigated by May 1992: 477
Were in unoccupied tunnels., etc. 262
Were in occupied school rooms 215
Rooms with charcoal follow-up measurements only 9
(all were < 4.0 with vent. syst. on continuously)
Rooms with CRM follow-up measurements: 206
Initial results < 4.0 between 8 AM and 4 PM* 128
Initial results > 4.0 between 8 AM and 4 PM 78
Mitigated to < 4.0, 8 AM to 4 PM 50
Not yet mitigated to < 4,.0, 8 AM to 4 PM 28
"Some were mitigated with balancing or repair of ventilation
systems before follow-up measurements were made.
-------
7-
6-
O
Q.
a
a
a
n
n I mi i
a
345
pCi/L, 1STATD
Figure 1. Scatter plot of alpha track duplicate results.
20-
18-
16-
14-
>^ 12-
o
„- 10-
12
n
a
ffl
24 12 24 12
TIME (HOUR OF DAY)
a UNCORRECTED
CORRECTED
24
Figure 2. Effect of ventilation system, and correction for
progeny delays.
-------
35
30
25-
| 20-
z
oc
15-
10-
5-
n
a
a
a
a
D
a
a
a a
a a
a
a
a
cP
ODD,
12 24 12 24
TIME (HOUR OF DAY)
1111111111111111111111111111
12 24 12
Figure 3. Radon in room with sub-slab return air duct,
2-
a
D
a el-,
f~l uj
qSp a a
n & rP a
a off
a, a n
tJ Q CfclD
Q '-' ndP o
a
a
a
a
n
on
n
a
DOOR CLOSED OPEN CLOSED OPEN (SAT.) (SUN.)
1-
0-Uw,
12
12
12
iiiiininiiiiinmniiininniniiiiiniiiiininiiiiiiiiiniiinniii in
12 12 12 12 12
TIME (HOUR OF DAY)
Figure 4. Door effect, return air partially through hall,
-------
XP-4
INVESTIGATION OF FOUNDATION CONSTRUCTION
DETAILS TO FACILITATE SUBSLAB PRESSURE
FIELD EXTENSION IN LARGE BUILDINGS
by: Mike Clarkin
Camroden Associates, Inc.
RD #1, Box 222, East Carter Road
Oriskany, NY 13424
Kelly W. Leovic
U.S. Environmental Protection Agency
Air and Energy Engineering Research Laboratory
Research Triangle Park, NC 27711
Fred McKnight
The H.L. Turner Group
Harrison, ME 04040
ABSTRACT
Pressure field extension (PFE) measurements were conducted in
four recently constructed Maine buildings to determine the effect
of foundation design on radon control in new buildings. The goal
was to evaluate the impact of subslab materials on the strength and
extent of the subslab pressure field. Three schools and one office
building, each incorporating different foundation designs, were
studied.
Results support the following EPA guidelines for installing
subslab depressurization systems in new schools and large buildings
constructed in radon prone areas:
• install a layer of clean, coarse gravel beneath the slab;
• limit barriers to subslab communication or plan features
that will extend the pressure field across unavoidable
subslab barriers;
• install a large suction pit and radon vent pipe; and
• if elevated radon levels are measured, activate the
system with a fan.
The results also demonstrate the need for observing
installation of radon prevention systems before the slab is poured
in order to ascertain system design parameters. Radon prevention
systems in three of the four buildings were much less effective
than anticipated since clean, coarse gravel had not been placed
under the slab.
If gravel is not used, a subslab pipe network will help to
extend the subslab pressure field. However, the PFE developed by
a pipe network (without gravel) is not as evenly distributed as PFE
under a gravel bed. The builder must also consider the additional
expense of purchasing and installing the pipe network.
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.
-------
INTRODUCTION
The U.S. Environmental Protection Agency's (EPA's) Air and
Energy Engineering Research Laboratory (AEERL) began research on
indoor radon in residential buildings in the mid-1980's. In 1987,
AEERL initiated research on developing and demonstrating radon
reduction techniques that could be used in existing and new
schools. AEERL has also conducted limited research in existing and
new large buildings.
One of the most frequently used and effective methods for
radon reduction is subslab depressurization. A subslab
depressurization system maintains the soil gas beneath the slab at
a lower pressure than the air in the building. As a result, air
moves from the building into the soil rather than from the soil
into the building, effectively keeping radon from entering the
building by pressure driven transport. If a fan is used to create
the negative pressure under the slab, the system is referred to as
"active." A system without a fan is called "passive." An active
system is much more effective in achieving and maintaining a
consistent negative pressure under the slab.
In radon prone areas, AEERL recommends installing a subslab
depressurization system during building construction in order to
facilitate radon mitigation if needed (1). The objective of this
project was to determine the impact of certain construction
techniques on the effectiveness of subslab depressurization
systems. Addressing these construction techniques during the
design and construction stages will result in a more effective
radon control system at a lower cost than installing a retrofit
system after construction (2). Indoor radon measurements were not
part of this investigation.
Four recently constructed Maine buildings were s .lected for
this study. Investigation of each building included the following
steps:
(1) Review the blueprints and construction documents. The
radon mitigation system design was determined from the
drawings, and served as a road map for the next step.
(2) Thoroughly inspect the building. Assistance of someone
familiar with the building was critical during this step.
(3) Ask maintenance personnel and the builder for information
concerning footings, slab openings, subslab material, and
footing drains.
(4) Determine the airflow and pressure characteristics of the
subslab radon control system using an in-line centrifugal fan
to draw different airflows. The resulting pressure field
extension (PFE) across the slab was mapped.
-------
RESULTS
Three schools and one office building, each located in Maine,
were selected for this study. The major characteristics of these
buildings are summarized in Table 1, and each building is discussed
separately.
TABLE 1. SUMMARY OF BUILDING CHARACTERISTICS
Building
School A
School B
School C
Building D
Subs lab Aggregate
25-60% pass 0.25
inch* sieve
25-60% pass 0.25
inch sieve
0.50 to 0.75 inch
in diameter
25-60% pass 0.25
inch sieve
Radon Control System
1 perforated pipe,
1 suction point
perforated pipe
perforated pipe
perforated pipe;
footing bridge
SCHOOL A
System Description
This is a two story slab-on-grade building. The floor plan
consists of a large rectangle and three pods. The school was
constructed with two separate passive subslab depressurization
systems. System 1 has four risers exiting through the roof, and is
connected to an interior footing drain loop. This system covers
the rectangle and two of the pods. System 2 is a single point
passive depressurization system in the third pod with a gravel-
filled well connected to a roof exhaust. An interior footing
between the two systems effectively separates the subslab area into
two zones. Figure 1 schematically illustrates the layout of the
school and the two radon control systems.
The perimeter loop system is connected to a storm drain in two
places. If air is drawn from the storm drain into the perimeter
loop, it would be difficult to produce a low pressure field beneath
the slab. The plans do not indicate use of water traps or reverse
flow values to separate the storm drain from the perimeter loop.
Subslab aggregate specifications for the school required a 12
inch layer of base material, with 95% compaction. Size
specifications for the subslab aggregate are in Table 2.
* 1 inch =2.54 centimeters
-------
TABLE 2. SCHOOL A SUBSLAB AGGREGATE SPECIFICATIONS
Sieve Size
3 inch
1 inch
0.5 inch
0.25 inch
# 40
# 200
Percent Passing
100
80-100
35-75
25-60
0-25
0-5
The subslab aggregate specifications allow for a large percentage
of stones less that 0.25 inch in diameter. Observation of the
subslab material during the investigation revealed that a large
percentage of the aggregate was, in fact 0.25 inch or smaller.
This fine-grained material can impede subslab PFE (1).
Procedures
The System 1 perimeter loop was constructed of 4 inch poly
vinyl chloride (PVC) perforated pipe wrapped in geotextile wrap and
buried in the subslab aggregate. This loop was connected to solid,
Schedule 20 PVC pipe risers. This system was designed to provide
passive depressurization to a subslab area of about 48,000 square
feet.* PFE measurements were made to determine the strength and
distance of the pressure field, from the perimeter loop. The
objectives of investigation of this system were to determine if the
perimeter loop was helping to extend the subslab pressure field and
if the interior footing was restricting PFE.
The researchers searched for the perimeter loop/scorm drain
connection to determine if there were large air leaks into the
subslab drainage layer. Unfortunately, the connection was not near
any access holes. Building maintenance personnel said the storm
drain emptied into a pond located on the school grounds, not to
daylight. Subsequent airflow measurements did not indicate any air
leaks into the system.
According to available building plans, System 2 began beneath
the slab in a pit filled with 0.5 cubic yard** of crushed stone.
Schedule 20 PVC pipe was used for the exhaust riser in this system.
The plans did not specify the size of the stone used in the pit,
but the builders' standard specifications for stone used for this
purpose is angular crushed natural stone, free from shale, organic
matter and debris; and 0.375 to 0.75 inch in diameter. This single
point system was intended to provide depressurization beneath an
area of approximately 8800 square feet.
* 1 square foot = 0.093 square meter
** 1 cubic yard = 0.765 cubic meter
-------
The locations for the subslab PFE measurements were selected
based on opportunity rather than systematically. Nearly all floor
surfaces in the school were covered with vinyl tile, and school
officials were hesitant to allow many slab perforations. As a
result, a compromise between scientific evaluation and suitable
drilling locations was made. Locations chosen for the slab
penetrations included under floor-mounted doorstops and beneath
carpets. Figure 1 illustrates approximate positions of all test
points.
Measurements on the two systems included exhaust pipe airflows
and subslab-to-indoor pressure differences. The measurements were
made in the passive mode (without a fan, as designed) and with an
in-line centrifugal fan at two airflow rates. The other stacks
were capped during the active tests to prevent short-circuiting.
The fan used is rated at 80 cfm* at 0 inch water column (WC).**
Results
Figure 2 shows subslab pressure versus distance measurements
for System i — the perimeter loop system — using the fan for
depressurization. The subslab pressure field did not reach test
point 3, which was 28 feet from the suction point. It was not
possible to drill holes in the immediate vicinity of the footing
(which was approximately 18 feet from the suction point) to
determine if the footing was restricting the pressure field. As an
alternative, a test hole was drilled in another location, at the
same distance from the perimeter loop as the interior footing
(approximately 18 feet). No pressure field was detected at that
test point. These measurements indicate that the subslab aggregate
was restricting PFE, so that it was not possible to determine the
effects of the footing.
Figure 3 shows the results of the subslab pressure versus
distance measurements for System 2 — the pit system. These
results also confirm that there is fine-grained material under the
slab. A static pressure of -1 inch WC in the suction pit does not
extend a pressure field beyond 10 feet. It is suspected, but not
confirmed, that the exhaust riser did not rest in a pit as
specified, but terminated in the subslab aggregate. This would
result in a suction's being applied to only 12.6 square inches* of
* 1 cubic foot per minute (cfm) =0.47 liter per second
** 1 inch water column = 249 pascals
-------
aggregate, rather than the surface area of a pit. The importance of
the area of aggregate exposed to the suction has been documented in
previous research (2) . This situation illustrates that inspection
of the system installation by a qualified radon professional is
essential during building construction.
We had hoped to test whether the interior footing interferes
with PFE. Unfortunately, a direct measurement of this was not
possible since the pressure field did not extend to the footing.
Both systems were also tested in the passive mode; however, no
airflow or negative pressure was measured under the slab. As a
result, if this system is to be used to consistently exert a
negative subslab pressure, it must be operated in an active mode.
SCHOOL B
System Description
This school, still under construction during this
investigation, has a rather extensive subslab pipe network
connected to a single exhaust riser. The pipe network is
constructed of 4 inch diameter perforated PVC pipe. The diameter
and number of holes per linear foot** of pipe was not documented by
the builder. Figure 4 illustrates the pipe network in a portion of
this school.
At the time of the investigation, the riser was terminated 2
feet above the slab surface, conversations with the site manager
revealed that there was no intention to run the riser outdoors
unless indoor radon measurements indicated that a radon control
system was required. At the time of the investigation these
measurements had not been made; however, subslab measurements
indicated a radon concentration of 1,200 pCi/L*** beneath the slab.
The building architect was informed of these levels and was
considering running the riser through the roof during construction.
Subslab aggregate specifications were not available, but
conversations with the building project manager indicated that
crushed stone similar to the aggregate in School A was used. A
polyethylene barrier was placed on top of the crushed stone, a 2
inch layer of sand placed above the aggregate, and the slab poured
on the sand. Sand was probably used to reduce the chance of the
slab's cupping as the result of unequal drying between the top and
bottom. However, the sand interfered dramatically with the subslab
* 1 square inch = 6.45 square centimeters
** 1 foot = 0.305 meter
*** 1 picocurie per liter (pCi/L) = 37 becquerels per cubic meter
(Bq/m3)
-------
PFE measurements. In many cases, the sand would fill in around the
hole as soon as the drill bit was removed. It took several tries
at most holes to take a pressure reading in the subslab aggregate
rather than in the sand layer.
Procedures
An in-line centrifugal fan (rated at 80 cfm at 0 inch WC) was
placed on the end of the riser. Airflow out of the system and
pressure differences were measured at the riser and at various
distances from the riser.
Results
Subslab-to-indoor pressure differences are shown in Figure 5.
It was expected that the strongest negative pressure readings would
be found near the subslab pipes. The predicted pressure pattern is
not clear in Figure 5. The strength of the subslab pressure rises
and falls in a cyclic way, but the pipe pattern would have to be
different than was shown on the building plans. There is a low
pressure field at almost all tested points, with a mean of -0.1
inch WC. This would be expected because the pipe network should be
exerting a near uniform pressure at all points. Some of the
variation in the measurements may be an artifact of the sand's
interfering with the pressure taps. For example, the relatively
low negative pressure at 30 feet appears like an outlier for both
airflows.
SCHOOL C
System Description
The radon control system at this school was being installed in
an addition to the original building. The foundation, slab, and
walls were in at the time of the measurements, but the roof was not
on. While this might have some effect on the test measurements,
the subslab PFE was strong enough and extensive enough to be
conclusive. The slab is a 6700 square foot rectangle. A 6 inch
perforated pipe runs under the center of the slab through the
drainage layer. The drainage layer is composed of stone pebbles
that are 0.5 to 0.75 inch in diameter. The frost walls are poured
concrete, and there are very few slab penetrations. Figure 6 shows
a floor plan of the building and the location of the test points.
-------
Procedures
Conversations with the site manager revealed that the
perforated pipe ran straight across the width of the slab, exiting
through a footing. It then snaked its way underground, beneath an
adjacent slab, and finally exited to daylight. Assured by the site
manager that we had correctly identified both ends of the pipe, an
in-line centrifugal fan was attached to one end of the pipe, and
the other end was carefully sealed with duct tape. Test holes were
drilled through the slab, and subslab pressures were measured and
recorded.
Results
Figure 7 shows the subslab pressure profile for this school.
Pressures at test points beyond 10 feet measured -0.012 (±0.005)
inch WC beneath the slab. Test Point 1, located directly above the
perforated pipe, measured -0.016 (±0.002) inch WC. Subslab
differential pressure measurements were not made across the width
of the building because building materials were piled on one side
of the slab. The subslab flow and pressure characteristics
indicate a fairly tight foundation, low leaks, and highly permeable
fill. The pressure field does not drop even within 2 feet of the
slab edge, and there is only a 25% decrease in pressure field
strength at all other test holes as compared to the pressure at the
pipe. This indicates that a layer of clean, coarse gravel beneath
the slab provides a more uniform pressure field, compared to the
fields found in the other buildings described in this paper.
This type of system, with a perforated pipe in the middle of
a layer of aggregate, behaves like an air plenum. It is expected
that pressure down the length of the pipe would be fairly uniform
because the resistance to airflow of the perforations is large
compared to the resistance to airflow in the pipe. This
distributes the low pressure pattern along a line down the center.
Airflow through the aggregate bed will come largely from the slab
edges that are parallel to the pipe. The air will move
perpendicular to the direction of the pipe. If the stem wall and
slab are fairly tight (as in this case), the resistance of the
cracks and holes through them will be much greater than through the
aggregate bed.
This can be modeled using design aids for thermal storage
stone pebble beds. Applying the design aid to this problem, it is
expected that drawing 196 cfm from the pipe would result in a
differential pressure that is 0.002 inch WC less 30 feet from the
center pipe than 1 foot from the pipe.
On the other hand, if 196 cfm was drawn through a system with
a single suction point, this type of uniform pressure field would
not be expected unless the leaks were so small that very low air-
flows were drawn from the system. This is because the resistance
to airflow through the aggregate increases as the speed of air
8
-------
moving through the aggregate increases. As air moves through the
aggregate bed to the central suction point, the cross-sectional
area of bed it travels through becomes increasingly smaller. The
air speed must increae in order to conserve mass so the resulting
pressure drop with distance from a central suction point is
expected to be nonlinear and greater than for the plenum-type
system achieved with the central pipe system.
BUILDING D
System Description
This building was designed to facilitate radon reduction, if
needed, using a subslab depressurization system. Subslab aggregate
specifications are the same as those for the School A. In addition
to the subslab aggregate, a perimeter drain loop pipe network with
two passive risers was installed. In order to facilitate subslab
PFE, interior footings were not used except in a small area
surrounding a vault. To extend the pressure field across the
footing, the top of the footing was recessed 4 inches in the
doorway, and a 4 inch layer of gravel was placed in this recess.
The floor slab was then poured. The building floor plan is shown
in Figure 8. The purpose of testing in Building D was to determine
the effect of the recessed footing on PFE. Measurements were not
made to assess the effectiveness of the drain tile system.
Procedures
The PFE measurements were performed using two suction points,
as shown in Figures 9 and 10. Suction was applied first to point
SP1 (Figure 9) and then to SP2 (Figure 10) . Pressures were
measured under the slab on both sides of the opening through the
footing in order to determine the effect on PFE.
Results
The baseline differential pressure under the slab was neutral,
indicating that the passive system was not exerting a negative
pressure under the slab. However, these measurements were made
during the summer with little temperature differential between
indoors and outdoors.
The opening in the footing does appear to facilitate subslab
PFE. A pressure of - 32 inch WC at SP1 produced a pressure of -
0.240 inch at TP7 which is 6 feet away and on the same side of the
footing (Figure 9). Pressure at TP3, 5 feet from SP1 but on the
other side of the footing, was -0.02 inch WC. Although this is not
a very strong negative pressure, it is apparent that the negative
pressure does reach across the opening in the footing. When
suction is applied at SP2 (Figure 10), a slight negative subslab
pressure is also observed across the footing.
-------
We suspect that there is some obstruction between TP3 and TP4
which is causing the quick drop in the pressure field between these
two points that are only 1 foot apart. This drop in pressure would
not be caused by the footing because a large drop in pressure also
occurs when suction is applied at SP2 which is on the same side of
the footing (Figure 10).
Maintaining a continuous layer of aggregate under a door
opening (which is otherwise surrounded by footings) should be
researched in a building with high permeability aggregate. The
aggregate produces a very uniform subslab pressure (as seen in
School C results), and any pressure drop across a footing could
then be attributed to the footing rather than the subslab
aggregate.
CONCLUSIONS
The PFE measurements in these four Maine buildings support
AEERL recommendations for radon prevention in the construction of
schools and other large buildings (1) . The recommendations,
together with the supporting data from this research, are
summarized as follows:
1) install a 4 to 6 inch layer of subslab aggregate (ASTM #5) - In
the four buildings researched for this project, the building with
the best PFE was the building with clean, coarse subslab aggregate
(School C). The three other building had subslab material with
more fines (see Table l); thus, the pressure field did not extend
as far or was as strong.
2) Avoid barriers to subslab communication - Although openings in
the footings helped to bridge subslab barriers, the PFE was
somewhat reduced (Building D). Research of this construction
technique should be pursued in a building with clean, coarse
subslab aggregate.
3) Install a suction pit and radon vent pipe - Previous research
(2) has shown that a suction pit will improve PFE by increasing the
surface area of the applied suction. The data from School A help
to confirm the importance of a suction pit.
4) Attach a suction fan to the vent pipe to activate if needed - In
the two buildings that were designed to operate passively (School
A and Building D), measurements showed that subslab PFE was not
achieved without use of a suction fan. However, these two
buildings did not have clean, coarse subslab aggregate, and
measurements were made during the summer with minimal temperature
differential.
This study demonstrated the need for observing installation of
radon prevention systems before the slab is poured in order to
ascertain system design parameters. This study also indicated
that, in areas where gravel is not readily available or expensive,
use of a subslab pipe network will help to extend the subslab
10
-------
pressure field (Schools A and B) . However, the PFE developed by a
pipe network is not as evenly distributed as PFE through a coarse,
crushed aggregate bed. The builder must also consider the
additional expense of installing the pipe network.
REFERENCES
1. Leovic, K.W., A.B. Craig, and D.B. Harris, Radon Prevention in
the Design and Construction of Schools and Other Large
Buildings, Architecture/JResearch, October 1991, 1:1, pp. 32-
33.
2. Craig, A.B., K.W. Leovic, and D. B. Harris, Design of New
Schools and Other Large Buildings Which Are Radon Resistant
and Easy to Mitigate. Presented at the Fifth International
Symposium on The Natural Radiation Environment, Salzburg,
Austria, September 1991.
11
-------
Connect to
storm drain
System 1
Passive perimeter
drainage loop
System 2
single point
passive system
Library with
load-bearing
walls
Figure 1. School A subslab depressurization system
layout (not to scale). SP denotes suction
points; TP denotes test points.
12
-------
TP2
TP3
Perimeter loop
pipe
LU
E
_
LU
a.
v.
:'
LU
DC
-
System 1 at 61 cfm
System 1 at 31 cfm
-0.5
10 15 20
DISTANCE (ft)
25
Figure 2. School A System 1 subslab pressure
characteristics. TP denotes test points.
13
-------
Shz
_
E
-0.2
-0.4
-0.6
-0.8
TP5 TP6
10 15 20
DISTANCE (ft)
TP4
30
Figure 3. School A System 2 subslab pressure
characteristics. SP denotes suction
point; TP denotes test points.
14
-------
10ft-
6
\
Suction
point
All test points
along this line
\
Figure 4. School B subslab pipe network layout.
15
-------
_
_
~-
.
_
r
-0.2
-0.4
-0.6
-0.8
-1
It
Airflow a\ 50 cfm
Airflow at 26 cfm
, i
15 20
DISTANCE (ft)
:-
-
Figure 5. School B subslab pressure characteristics,
16
-------
42ft'
T
2ft
TP4
20ft
TP3 X
10ft
10ft
T
This end capped
4 in. drainage pipe
X
TP1 X
Fan installed on this
end
10ft
i
TP5 X
Figure 6. School C test point locations.
TP denotes test points.
17
-------
TP5 TP1 TP2 TP3
TP4
_
f
* '
I I 6.
o ~
-0.004 ~
-0.008 ~
-0.012 "
-0.016 ~
-
•
•
\
\
t
Subslab drainage pipe
\
-0.02 H
20 10
/
•
*
•
•
0 10 20 30 40 50
DISTANCE (ft)
Figure 7. School C subslab pressure characteristics,
TP denotes test points.
18
-------
Vault room with
footings. See Figure
9 for details.
Figure 8. Building D floor plan (not to scale)
19
-------
1 1
1 1
0 1
SP2
I I 1 1 1
1 1 1 1 1
23456
TP4 TP3
-0.004 -0.02
V V
SCALE (ft)
III 1 1 1 1 1 1
7 8 9 10 11 12 13 14 15 16 17
TP2 TP1 SP1 TP7
-0.348 -1.23 -32 -0.240
S^ NOTE: All SP and TP
Vault
room
measurements are in
inches WC.
Footing
TP6
-0.167
Figure 9. Building D subslab pressure field
characteristics with suction at SPl.
SP denotes suction points; TP denotes
test points.
20
-------
SCALE (ft)
•1 1 h
8 9 10 11 12 13 14 15 16 17
SP2
-49
TP4 TP3
-0.254 -0.026
TP2
TP1
-0.005
SP1
TP7
-0.003
-x—1
TP8 TP9
-2.57 -0.362
Vault
room
Footing
x
TP5
-0.046
NOTE: All SP and TP
measurements are in
inches WC. No measurement
at TP2 and TP6.
TP6
Figure 10.
Building D subslab pressure field
characteristics with suction at SP2. SP
denotes suction points; TP denotes test
points.
21
-------
XP-5
Title: Radon Measurements in the Workplace
Author: David Grumm, U. S. EPA, Office of Radiation Programs
This paper was not received in time to be included in the preprints, and
the abstract was not available. Please check your registration packet for a
complete copy of the paper.
-------
XP-6
RADON SURVEY OF OREGON PUBLIC SCHOOLS
by: George L. Toombs & Ray D. Paris
Oregon Health Division
Suite 705
800 NE Oregon #21
Portland, Oregon 97232
ABSTRACT
To assist in addressing the concerns of potential elevated
radon levels in public schools nationwide, the Oregon Health
Division conducted a limited study to determine the average
radon levels in the schools in Oregon.
Thirty-one schools were selected at random out of a
population of 1,190 statewide to participate in this study
during the 1990-1991 school year. Long-term alpha track
detectors were placed in each of the ground-floor and basement
classrooms to obtain the average radon levels this entire nine-
month school year.
The results of this study showed that the mean radon con-
centrations in Oregon public schools statewide was 1.1 pCi/lit-
er. This compares to a mean of 1.2 pCi/liter for indoor radon
in homes in the state. Two of the schools surveyed had many
rooms above the 4 pCi/liter EPA guidelines.
Follow-up with these schools and their options for lower-
ing the levels in the elevated rooms are discussed.
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.
-------
OREGON HEALTH DIVISION PUBLIC SCHOOL RADON PROJECT
by: George L. Toombs & Ray D. Paris
Oregon Health Division
Suite 705
800 NE Oregon #21
Portland, Oregon 97232
HISTORY
Indoor radon surveillance in Oregon was initiated in 1984
by the Bonneville Power Administration (BPA) as part of their
regional Residential Conservation Program covering the five
Northwestern States. Their measurements were essentially
conducted over a 3-month period, mainly during the winter
months. It should be noted that their service districts were
the only areas covered in this study. As of January 1, 1991, a
total of 12,871 homes had been tested. The indoor radon value
for these homes had an arithmetic mean of 1.2 picocuries per
liter (pCi/1) with about 4 % of the homes exceeding the
Environmental Protection Agency's (EPA) guideline value of 4
pCi/1.
The Division initiated the Oregon Radon Project (ORP) in
March of 1988 to obtain needed information covering a 12-month
period. The results of this 12-month project gave an
arithmetic mean for indoor radon at 1.4 pCi/1. This compared
very closely with BPA's value of 1.2 pCi/1. The conclusion was
that indoor radon problems in Oregon are small when compared to
other regions of the country.
Although in Oregon, indoor radon problems are relatively
small, areas would have to be defined wh^re it could pose
problems to large numbers of people. This prompted the
Division to develop a project to evaluate potential radon
concerns in Oregon public schools.
PUBLIC SCHOOL PROJECT
Elevated radon concentrations in schools have been a
growing public health concern nationally. This is because high
levels have been found in a large number of schools around the
country and children are likely to be more vulnerable to the
effects of radon. Therefore, many states, including Oregon,
are involved in projects to assess the indoor radon situation
in schools.
Oregon has 1190 public schools statewide. It was beyond
the scope of this project to test every school. Therefore, the
-------
goal of the Division's Public School Radon Study was to conduct
a limited study to provide an estimate of the average radon
level in Oregon school buildings. Thirty-one schools were
randomly selected to be tested.
The schools selected to participate in this study were
chosen from the Oregon School Directory 1989-1990 edition.
Each was given a number from 1 to 1190. A random number table
was then used to determine which schools would be contacted and
asked to participate in the study. The 31 schools that chose
to participate were locatqd statewide in 20 of the 36 counties
and are listed in Table 1. The distribution of these
participating schools followed the population density very
closely. The sizes ranged from a three-room country school in
the Willamette valley to a 78-room complex in the southern part
of the state. A monitoring device was placed in every ground-
floor and basement classroom in each school.
Alpha-track detectors were used exclusively in this study.
The design of the device, durability, ease of installation,
reporting format options of the vendor and cost were the
factors considered in selecting the vendor by using a
competitive-bid process. The selected vendor also had to be
listed in the National Radon Measurement Proficiency Report of
EPA dated January 1990. Radiation Safety Systems, Inc. (RSSI)
in Morton Grove, Illinois, was awarded the bid.
The alpha-track monitoring devices received from RSSI were
delivered to each of the schools by Health Division staff. The
purpose of the study was explained to the personnel along with
instructions on how to install the devices. They returned to
the Division the data sheets, the numbered shipping foil
packets, and the floor plan showing the rooms where the devices
were installed. This provided for accountability and the
assurance that the detectors were actually installed. The
Division also inspected the installations in about 20% of the
schools after the devices were installed as part of the overall
quality assurance plan.
Each of the 31 schools was notified, as a reminder, in
March 1991 that the devices would be picked up in May. The
Division also provided additional informational brochures about
radon in the home and schools at this time.
The monitoring devices were picked up by the Health
Division during the last week of May and first week of June
1991. Any device that had the security seal broken or had
visible damage was not processed.
Quality assurance is a vital part of any study. The
objective is to insure the data obtained is valid. The guality
-------
assurance plan for this study consisted of administrative
controls, placement techniques, documentation, supervision and
followup. It included sending the vendor "spiked" monitoring
devices exposed to known levels of radon. Blank and duplicate
devices were also returned to the vendor for processing.
Fictitious names and addresses of schools were used for these
controls so the vendor could not distinguish them from the real
ones. A total of 789 devices were used in this study. Only 30
of the installed devices were unable to be processed because
they were either damaged or missing. This is about a 96%
success rate, which is very high when compared to school survey
projects of other states.
RESULTS
Figure 1 shows the radon concentration in pCi/1 in
relation to the percentage of classrooms tested. The
arithmetic mean for the indoor radon levels in all rooms of the
31 schools tested was 1.1 pCi/1. The summary of results per
pCi/1 for each school tested, are listed in Table 2. This
compares very favorably with indoor radon levels found in homes
as shown by the BPA and Oregon Radon Project data listed below.
Mean Radon Measurement Number of
Value in pCi/1 Period Measurements
Bonneville Power data 1.2 3 months 12 871 (homes)
Oregon Radon Project 1.4 12 months l' 562 (homes)
Oregon Public Schools l.l 9 ^nths 31 schools (689 rooms)
Only two schools out of the 31 tested had any classrooms
greater than the EPA guideline of 4 pCi/1. Not all rooms in
those schools exceeded the 4 pCi/1 value. One school had 14
rooms out of 39 greater than that value, and the other school
had 12 out of 35 rooms exceeding the guideline. The highest
value observed in a classroom was 9.6 p^i/1. When the
concentrations observed in all rooms were taken into
consideration in each of these two schools, neither had an
average indoor radon value greater than the 4 pCi/1 guideline.
The Health Division subsequently worked with these two
schools which had classrooms with radon levels greater than the
EPA guideline of 4 pCi/1. Specifically, we wanted to define
the radon sources and the options available to reduce the
levels. This required additional short-term testing as well as
continuous monitoring in one of the schools to evaluate the
source and to observe how the heating/ventilation/air
conditioning system (HVAC) affected the radon levels in the
classrooms.
-------
Elevated levels were observed in the utility tunnels under one
of these schools. The rooms directly above these tunnels
generally had higher levels. Continuous monitoring of the
classroom with the highest concentration (9.6 pCi/liter) showed
a strong correlation between radon levels and the mode of
operation of the HVAC system. Figure 2 shows the radon
concentrations in relation to the various operation modes of
the HVAC system.
This school, using the above data, is planning on making
changes in ventilation and HVAC operation to minimize radon
levels in the classrooms. Also, they are considering
appropriate mitigation techniques to lower the radon
concentrations in the tunnels to further reduce the levels in
the classrooms.
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.
-------
TABLE 1. PARTICIPATING SCHOOLS
NAME OF SCHOOL
SURVEYED
Applegate Elem.
Ardenwald Elem.
Awbrey Park Elem.
Burns Union High
Butte Falls High
Colton Elem.
Crooked River Elem.
Dickey Prairie Elem.
Elgin Jr. High
Fairplay Elem.
Ferndale Elem.
Franklin High
Glide Jr. High
Grant Union High
Hanby Jr. High
Irish Bend Elem.
Judson Middle School
Lapine Elem.
Mabel Rush Elem.
Madras Jr. High
Molalla Primary
Mt.Vernon Elem.
North Medford High
Park Place Elem.
Parkdale Elem.
Prospect Elem.
Riley Creek Elem.
Rocky Heights Elem.
Sand Ridge Elem.
Scappoose High
Territorial Elem.
COUNTY
Lane
Clackamas
Lane
Harney
Jackson
Clackamas
Crook
Clackamas
Union
Benton
Umatilla
Multnomah
Douglas
Grant
Jackson
Benton
Marion
Deschutes
Y.amhill
Jefferson
Clackamas
Grant
Jackson
Clackamas
HoodRiver
Jackson
Curry
Umatilla
Linn
Columbia
Lane
GRADES
K-5
K-6
K-5
9-12
9-12
K-6
K-5
K-8
7-8
K-5
K-6
9-12
7-8
9-12
7-8
K-8
7-8
K-6
K-4
7-8
K-5
K-6
9-12
K-6
K-5
K-8
K-8
K-6
K-8
9-12
K-4
NO. OF
STUDENTS
126
295
397
340
63
489
595
93
180
186
241
1,325
146
235
224
23
862
479
559
373
446
109
1,374
334
278
139
304
481
72
535
127
NO. OF
ROOMS
SURVEYED
17
16
31
30
7
28
28
7
35
10
20
28
13
14
6
3
42
27
26
23
24
11
74
10
18
19
21
28
7
39
6
Total
Total Oregon Public School
Enrollment 1990-1991 (K-12):
11,430
668
495,922
-------
TABLE 2. PUBLIC SCHOOL RADON PROJECT
SUMMARY OF RESULTS IN Picocuries per liter (pCi/1)
School
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
No. of Rooms
Tested
17
16
31
30
7
28
28
7
35
10
20
28
13
14
6
3
42
27
26
23
24
11
74
10
18
19
21
28
7
39
6
No. of Rooms
Over 4 . 0
0
0
0
0
0
0
0
0
12
0
0
0
0
0
0
1
0
0
0
0
0
0
0
0
0
0
0
0
0
14
0
MAXIMUM
3. 1
2.3
0.9
2.3
1.4
3.5
0.8
1.3
8.7
2. 1
1.1
0.8
1.8
2.8
2.6
5.0
2. 1
1.3
2.7
0.7
0.5
0.8
1.3
0.4
1.9
2.9
4.0
1.2
1.2
9.6
3.0
MINIMUM
0.8
0.5
0.2
0.2
0.7
1.2
0.2
0.2
0.4
0.4
0.3
0.2
0.8
0.4
0.5
0.4
0.3
0.1
0.1
0.2
0.2
0.2
0.2
0.2
0.3
0.5
0.4
0.4
0.3
1.2
0.3
Mean ± S.D.
1.6±0.7
1.6±0.6
0.7±0.2
0.910.2
1.010.3
1.910.7
0.510.2
0.910.4
3.411.7
1.110.8
0.710.2
0.410.2
1.210.3
1.010.7
1.210.9
2.012.6
0.710.4
0.410.2
1.110.6
0.4+0.1
0.310. 1
0.610.2
0.410. 1
0.310.1
1.010.4
1.510.7
0.810.8
0.610.2
0.710.3
4.012.0
1.511.1
TOTALS 668 27 MEAN 2.412.1 0.410.3 1.110.8
-------
FIGURE 1.
OREGON HEALTH DIVISION
PUBLIC SCHOOL RADON PROJECT
30
Q
LU
CO 25~
LU
CO
S 20-
o
O
cc
o
LL 10-
O
h-
LU 5-
0
CC
LU
Q.
n —
^>
Percent of classrooms at various concentrations based
upon 9 consecutive month measurements (Sep '90 - Jun '91)
n = 668
MEAN ± S.D. = 1.1 ± 0.8
MAX = 9.6
4.2% > 4pCI/l
1.9% > 5pCI/l
0% > 10pCI/l
12345678
RADON CONCENTRATION IN pCi/L
10
-------
FIGURE 2.
Effect of various modes of HVAC system operation on
radon concentrations in school classroom
ROOM D1
Off
Minimum
Outside
Air
Exhauct Off
Supply On
Exhaust On |
Supply Off •
4 1
VENTILATION CYCLES
-------
Session XI Posters
Radon Prevention in New Construction
-------
XIP-1
MODEL STANDARDS AND TECHNIQUES FOR
CQMTROL OF RADOM IM NEW BDILPIMflfl
by
David N. Murane
USEPA Radon Division
ABSTRACT
It is anticipated that EPA's "Model standards and Techniques for
Control of Radon in Nev Buildings" vill be published in final
fora in the Spring of 1992. This tvo part paper will provide:
(1) An update on ths key provisions of the final "Standard", and
(2) Details of an outreach program designed to inform, educate
and gain support of State and local building officials for
adoption of the Standard in National and local building codes.
-------
XIP-2
COMBINED VENTILATION AND ASD SYSTEM
by David Saum
Infiltec
Falls Church, VA 22041
Fred Sickels
NJ Department of Environmental Protection
Princeton, NJ 08625
ABSTRACT
Radon mitigation in new construction is generally accomplished
by installing an active subslab depressurization (ASD) system, but
most U.S. houses have no mechanical ventilation system that will
assist them in meeting the ventilation recommendations of the
ASHRAE 62-1989 standard. This paper describes the performance of
a single fan system for houses that was designed to provide both
radon mitigation and increased ventilation. Systems were installed
in three Maryland hous,es with radon levels between 4 and 20 pCi/L.
Performance monitoring included continuous radon and carbon dioxide
measurements.
-------
XIP-3
EVALUATION OF PASSIVE STACK MITIGATION IN 40 NEW HOUSES
By: Michael Nuess
Washington State Energy Office
N. 1212 Washington St., #106
Spokane, WA 99210-2401
ABSTRACT
The Passive Stack Study is a study of the performance of passive stacks (PS) in 40 new
residential homes in Washington State. It is anticipated that the study period will be about 18
months. The total study cost will be about $200,000. It is anticipated that the Washington State
Department of Health will manage the project with support from the Washington State Energy
Office-Energy Extension Service, U.S. EPA, and the Bonneville Power Administration. The
study homes will be located in the same geographical/climatic area (Spokane, Washington) in
order to:
• reduce weather-driven variables
• maximize the potential for appropriate candidate houses:
• houses with indoor radon levels encompassing the range of levels occurring in the
state and the region
•houses with code-required PS installed
• houses in both low-medium and high permeability soils
• houses with both winter and summer climate conditions
• minimize study COSt
The study will:
• measure and compare the effect of PS — installed according to the new Washington State
Ventilation and Indoor Air Quality Code (V&IAQ Code) -- on reduction of indoor
radon during both heating season and summer conditions in the study homes
• measure and compare the effect on indoor radon of the code-allowed partial substructure
sealing of the houses versus complete below grade sealing
• sufficientiy characterize the study houses in order to:
•enhance interpretation of the data obtained
•enable design of the detailed follow-up investigations
Homes will be selected into a four component matrix of 40 houses. This matrix structure allows
division of the study houses into two principal categories of soil type, and two principal
categories of substructure type. Each of the four components will consist of 10 houses:
High Perm Soil Med Perm Soil High Perm Soil Med Perm Soil
Basement Slab on Grade Basement Slab on Grade
-------
Each study house will be evaluated in order to obtain information about other important
variables:
• site factors (terrain, shielding, large surface pavements, etc.)
• building factors (area, volume, height, geometry)
• substructure details (gravel, pressure field extension characteristics, geometry)
• degree of substructure sealing (visual inspection)
• house depressurization potentials (characterization of mechanical systems
pressure impacts and use patterns)
• passive stack installation details
The follow-up investigations will intensively investigate selected study houses and subsets of the
study houses in order to:
• explain and substantiate the results obtained
• investigate the effects of several parameters on the PS performance:
• desirable fan capacity for activated stacks
• pressure-difference impacts of different residential mechanical systems
• sub slab membrane location and effectiveness
• alterations of pipe/gravel interface
• etc.
The study will hopefully recommend enhanced system design parameters and modifications to
code requirements related to PS systems.
-------
XIP-4
RADON REMEDIATION AND LIFE SAFETY CODE*
by: Lyle Sheneman, Architect
Chem-Nuclear Geotech, Inc.
Grand Junction Projects Office
Grand Junction, Colorado 81503
ABSTRACT
The preface of the Uniform Building Code states that the code "is dedicated to the development of better
building construction and greater safety to the public..." It accomplishes its mission through enactment of
uniform building laws that are based on the performance and use of building materials and systems.
Professionals in the building trades are schooled in the proper use of the code and are committed to
its principles.
The purpose of this paper is to address the concerns that radon mitigation systems are, at times, designed
and installed without apparent regard to life safety systems, as required by the Uniform Building Code.
Remedial action contractors need a heightened awareness of the necessity to either be aware of the code
or to enlist the assistance of professionals in the design and installation of these systems.
SCHOOL INSTALLATIONS
One of the major areas of focus for radon mitigation has been the nation's public school buildings.
Because schools are occupied by children, building code officials have subjected these buildings to
vigorous analysis of life safety concerns. Occupants of a school have the expectation that if the building
is threatened by fire or smoke, corridors are available as safe exits from the building. The corridors are
designed to resist involvement in the fire or entrainment of smoke for a period of 1 hour to allow the
occupants ample time to exit the building. All walls, ceilings, floors, and doors are constructed to be part
of this 1-hour system of protection.
*Work performed under the auspices of the U.S. Department of Energy, DOE Contract
No. DE-AC04-86ID12584.
-------
Because large unobstructed areas are required for instruction, a school is often constructed with long-
span, open-web steel joists that commonly bear on corridor walls. This type of construction results in a
grade beam being installed under those walls, dividing the sub-slab area into small "noncommunicating"
areas. Consequently, a radon mitigator attempting to depressurize the area below the floor slab in the
corridor must place the point of entry through the slab in the corridor and extend a vent stack through the
roof deck to evacuate the radon (see Figure 1). As this vent stack is extended through the ceiling finish, it
penetrates the protective barrier and violates its integrity, leaving a point of entry for both fire and smoke.
A chase is placed around the vent stack to finish and protect it, but this chase must be constructed with
materials that equal the construction of the corridor as outlined by the code. The radon vent stack and
chase must also be placed in the corridor in such a way to not obstruct the egress of persons in the
corridor (see Figure 2). The corridor width has been established by the occupant load of the building and
must not be reduced by the width of the chase.
ROOFING
INSULATION
ROOF DECK
VENT STACK
WITH CAP
FLASHING
SOIL
Figure 1. Radon vent stack section.
-------
CORRIDOR
CLASSROOM
PREFERRED CHASE
LOCATION
CLASSROOM
CHASE NOT PERMITTED
IN THIS LOCATION
1'-0" +- CHASE WIDTH
REQUIRED CORRIDOR
WIDTH BASED UPON
OCCUPANT LOADS
Figure 2. Radon vent stack location.
GENERAL BUILDING CONSTRUCTION
Material selection of exterior walls of all buildings, including schools, is governed by the code and
depends on the building's proximity to the property line, the proximity of adjacent buildings, and the
building usage. Openings in or penetrations through an exterior wall are not permitted when that wall is
within a specified distance to a property line or within the allowed distance to an adjacent building. If a
penetration is required, it must be "protected" as defined by the code. In many cases, a radon mitigator
will place a penetration through the wall, extend the radon vent stack through the opening, and terminate
the stack at rooftop height to vent the radon at a safe distance from the building's occupants. While this
seems to be an acceptable solution, the penetration through the wall must be accomplished in accordance
with the code.
-------
In the initial building survey for radon mitigation, all systems need to be considered. Underlying the
programmatic design criteria of a building are its systems requirements, such as ingress/egress, life
safety, structural, electrical, plumbing, and heating, ventilating, and air conditioning (HVAC).
Once these systems are understood, an evaluation is possible of the relationship between them and the
proposed radon vent system.
In many instances, the space above the ceiling finish materials and below the structural members is
commonly used as a chase for mechanical, electrical, and plumbing systems. It may also be used as a
return air plenum for the HVAC system. In such a case, all material placed in this space must meet code-
specified criteria. Exposed polyvinyl chloride (PVC) pipe, which is normally used in a radon vent
system, may not meet those criteria because of its flame spread and smoke emission characteristics. As a
result, the radon vent chase may need to be extended through this space to provide protection to both the
space and the vent system without altering the proper function of any system.
Commercial roofing systems are normally covered by a bond or warranty. A qualified technician must
alter such a roofing system to not void the bond or warranty. Although this is not a code-covered item, it
is an important issue to the property owner.
Many localities have adopted energy codes that govern the specifications and installation of a building's
insulation system. The workmanship employed in the installation of the radon vent system can be
critical to the proper performance of the insulation system. Disruption of the thermal barrier could
result in increased operating costs and heating/cooling loads on the HVAC system and could compromise
the comfort level of the building's occupants.
RESIDENTIAL INSTALLATIONS
Noise and vibration from the operation of a fan are ongoing problems in all active radon ventilation
systems. Many manufacturers have developed acceptable components that, when installed correctly,
either eliminate or reduce this annoyance to acceptable levels. However, some mitigators have developed
unique and creative ways of dealing with this problem.
One such solution is to cut a hole in the bottom of a 5-gallon plastic bucket and place it in the attic of the
subject property. The radon vent stack is then extended through the bucket and the bucket is filled with
concrete. While this would no doubt stop any vibration from extending down the vent stack, it introduces
a few problems of its own. A 5-gallon bucket with a 4-inch-diameter pipe extended through it will
contain slightly more than 0.5 cubic foot of concrete. Concrete weighs 155 pounds per cubic foot If this
bucket of concrete were to sit directly on the gypsum board ceiling, the result could be immediately
dramatic. If the ceiling did not fail at the time the concrete is placed in the bucket, it would, over a period
of time, begin to show signs of deformation because of the imposed load. While roof trusses are
designed to be primarily top chord load-bearing members, they are capable of carrying short-term point
loads on their bottom chords. However, similar to the gypsum board ceiling, they will show signs of
deformation when subjected to loads imposed on their bottom chords for long periods of time.
RECOMMENDATIONS
While it is idealistic to think that all individuals involved in the alteration of buildings are well versed in
the life safety codes, it is imperative that all alterations be done in strict accordance with the codes. To
accomplish this, appropriate building officials should review the documents outlining the system
installation. In addition, a professional architect or engineer should review and concur with the proposed
-------
system design and components. Each system should be documented by a detailed set of working
drawings showing the system components and techniques employed in their installation. This set of
working drawings would be necessary for obtaining any required building permits, would provide the
vehicle for the professional review, and would serve as a valuable resource for any future warranty work
that the system may require.
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.
-------
XIP-5
A PASSIVE STACK SYSTEM STUDY
By: Geoffrey Hughes and Katherine Coleman
State of Washington Department of Health
Office of Toxic Substances LD-11
Airdustrial Center Bldg. #4
P. 0. Box 47825
Olympia, WA 98504-7825
ABSTRACT
Washington's interim radon-resistive construction standards
require the installation of passive stack systems in new homes in
eight counties. The Washington State Department of Health and the
Washington State Energy Office are collaborating on a study on the
effectiveness of these passive stack systems. The research
entails the selection and characterization of 40 study houses,
followed by a series of radon tests. Each house will be tested in
both summer and winter in a series of four two-week periods of
alternating open and closed stack conditions. A subset of houses
will also be tested before and after the sealing of below-grade
penetrations. This paper summarizes the passive stack system
research methodology and,provides an up-to-date progress report of
the study.
-------
Session XII Posters
Radon in Water
-------
XIIP-1
RADON IN WATER MEASUREMENTS USING A COLLECTOR - BUBBLER
Robert E. Dansereau and Joseph A. Hutchinson*
New York State Department of Health
Laboratory of Inorganic and Nuclear Chemistry
P.O. Box 509, Albany, N.Y. 12201-0509.
*Current address
Lockheed Engineering and Science Co.
975 Kelly Johnson Drive
Las Vegas, NV 89119
ABSTRACT
A Collector-Bubbler (C-B) device was developed to allow precise and
sensitive field or laboratory measurements of 222Radon (Rn) in water. The
device, which is used in conjunction with Lucas cell (LC) counting, is rugged,
inexpensive and easy to use. After a measured quantity of water is collected
directly in the C-B, no liquid transfer is required. Rn is immediately purged
into a LC, which can be measured in the field. Rn decay and loss during sample
transport and transfer are eliminated.
INTRODUCTION
In recent years a great deal of concern has focused on indoor radon as a
serious health problem. Although the major source of indoor Rn is usually the
underlying soil and rock, drinking water is also a source, and in some cases, the
predominant one. In order to evaluate water supplies as a Rn source, studies
measuring waterborne Rn concentrations have been conducted by the U.S.
Environmental Protection Agency (1,2,3) and States (4). These measurements will
increase dramatically if the EPA's recently proposed Drinking Water Regulations
(5) are implemented. The Country's 68,000 public water supplies will have to be
measured quarterly during the first year and until they comply with the 300 pCi/L
maximum contaminant level (MCL) and annually thereafter. Analytical techniques
must be capable of measuring this level within a ±30% uncertainty. Although
these regulations do not apply to the estimated 13 million private wells, its
expected that the new regulation may result in many of them being tested.
The two techniques commonly used and proposed in the new regulations for
measuring Rn in water are the Lucas cell (LC) and liquid scintillation (LS)
methods (6). Both measurements are usually performed in the laboratory since a
liquid scintillation spectrometer is required for the LS method and the LC method
requires transfers using fragile equipment. Sample transport and preparation for
counting allows Rn decay and possible release from the sample. This reduces the
measurement sensitivity and could give inaccurately low concentrations.
DESCRIPTION
The C-B device (Fig 1) utilizes a 20-cc graduated nylon syringe with a
hollow plunger. A one-cm diameter medium-porosity fritted disc is fitted into
a machined hole at the inner end of the plunger. Luer stopcocks are attached to
both the tip of the syringe (outlet) and to the outside end of the plunger
(inlet). A tube containing 2 g of activated charcoal is fitted inside the hollow
plunger, and is connected to the inlet stopcock.
-------
PROCEDURES
COLLECTION
The sampling procedures used in the EPA recommended methods (6) are used
for sample collections with the C-B. For laboratory measurement a bottle is
filled in the field following the LC collection procedure and is shipped to the
laboratory. For field measurements follow the LS collection procedure. With the
outlet stopcock open and the inlet stopcock closed hold the C-B vertically,
immerse the tip (outlet) below the water surface and draw a measured 15-ml
aliquot of water into the C-B. Close the outlet stopcock.
RADON TRANSFER
Position the C-B with the outlet stopcock pointing upward. An initial purge of
Rn is accomplished by opening the inlet stopcock and withdrawing the plunger 5
cc, causing air to pass through the charcoal tube and bubble through the sample.
(The function of the charcoal tube is discussed in the testing section). This
also creates the necessary air space above the water to allow for sample bubbling
(Rn transfer) . The inlet stopcock is closed and the clamp is secured to prevent
plunger travel. The C-B is connected to a transfer system (Fig 2) which consists
of a drying column attached to a fine metering valve. An evacuated Lucas cell
is fitted above the valve. (A hand-operated vacuum pump is used for field
measurements) . Open both the Lucas cell and the C-B outlet stopcocks. Start the
bubbling action (Rn transfer) by opening the metering valve two turns. The C-B
inlet stopcock is then opened and the metering valve is adjusted to maintain the
desired bubbling rate. The Lucas cell is closed immediately after the bubbling
ceases.
MEASUREMENT AND CALCULATION
Allow four hours for secular equilibrium to be established with the short
lived Rn daughters. Following the manufactures directions count the Lucas cell
for 20-100 minutes. Divide the net sample count-rate by the counting efficiency,
decay factor and sample volume to obtain a value for the Rn concentration.
SENSITIVITY
The sensitivity of measurements using the C-B is dependent upon the
performance of the counting equipment. LC counters usually have an absolute
efficiency of 75% and relative efficiency of 225% (2.25 cpm/dpm for 222Rn in
equilibrium with 218Po and 214Po) . Using a cell with a background of 0.5 cpm, a
lower limit of detection of 0.16 pCi can be achieved for a 20 minute count. The
minimum detectable concentration for a 15-mL sample would be 10 pCi/L. The
measurement uncertainty at 300 pCi/L, the proposed MCL for drinking water, is ±
10%. This is well below the 30% uncertainty limit in the proposed drinking water
regulations. This sensitivity is determined for field measurements where the
time for Rn decay is the four-hour daughter ingrowth period.
-------
S TANDARCIZATION.
226 standardizati°n of this LC/C-B method is performed by purging the Rn from
a Radium solution obtained from the National Institute of Standards and
Technology. The solution is transferred to a 40-mL vial, which has a Teflon
lined cap. The vial is completely filled (no voids). After a Rn ingrowth
period, a 15 mL aliquot of the standard is drawn into the C-B and measured
following the measurement procedures section. Since the Rn is quantitatively
transferred from the C-B to the LC, the counting efficiency for this measurement
is given by:
E = S-B
AID
where:
E - efficiency (cpm/dpm)
S = standard count rate (cpm)
B = background count rate (cpm)
A = activity of standard (dpm)
I = Rn ingrowth factor
D = Rn decay factor
TESTING
REMOVAL OF Rn FROM PURGE AIR
Depending on field conditions, the ambient air used to purge Rn from the
sample into the Lucas cell could contain an appreciable quantity of radon, thus
resulting in an erroneously high Rn in water measurement.
The effectiveness of the C-B's internal charcoal in removing Rn from the
purge air was tested by drawing air containing 270 pCi/L of Rn through the device
and into a Lucas Cell. The C-B did not contain water. Four consecutive times
the C-B was tested with no (<0.6 pCi/L) detectable Rn passing into the cell. Rn
release from the charcoal was checked by passing ambient air through the system
after each run, then again one and six days later. The immediate checks
contained no measurable Rn, while the later checks contained less then 1% of the
Rn collected on the charcoal. The presence of Rn in the later checks is due to
degassing. Therefore, compressed gas (helium or aged air), as used in the EPA's
LC method, is not required for C-B/LC measurements.
EFFECT OF INITIAL LC PRESSURE
The effect of partially evacuated Lucas cells on the Rn transfer efficiency
from the sample to the cell was evaluated. Aliquots of water containing ~ 8,000
pCi/L of Rn were measured using the C-B with Lucas cells having varying reduced
initial pressures. The transfer was quantitative when the initial pressure was
less than 360-mm Hg. Therefore, a hand-operated portable vacuum pump, capable
of reducing pressure to 125-mm Hg, can be used for field measurements.
FIELD MEASUREMENTS
Field measurements were made at several public water supplies using the
C-B, a portable Rn counter 165-cc Lucas cells. Samples were also measured by the
LS method to provide reference values. The C-B results agreed very well with the
-------
reference values. Additional measurements will be performed at selected sites,
identified in the NYS study (4), to evaluate the C-B/LC field method performance
fr\v a tj^ /^c* var^/r^ rt-F T>n s*r*+- 4 *r4 ^ *» 1 AYTA! o
for a wide range of Rn activity levels.
SUMMARY
Results of our initial evaluation of the Collector - Bubbler device for
waterborne Rn measurement indicate that its use improves the accuracy of the
EPA's LC method. Additionally, its use facilitates field measurements. The need
for fragile equipment (glass radon bubblers), compressed gas, a mechanical vacuum
pump and/or an elaborate water degassing system is eliminated. Individuals
measuring indoor radon with a portable Lucas cell counting equipment can easily
and accurately measure radon in water.
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. U.S. Environmental Protection Agency. Environmental Radiation Data:
Report 43, Office of Radiation Programs, Report EPA/520/5-86-007
(Washington, DC: USEPA), 1985a.
2. U.S. Environmental Protection Agency. Nationwide Occurrence of Radon and
Other Natural Radioactivity in Public Water Supplies. EPA/520/5-85-008
(Montgomery, AL:USEPA), 1985b.
3. U.S. Environmental Protection Agency. Distribution Tables for the
National Inorganic and Radionuclides Survey. Results Memorandum to Arthur
Perler, STB from Jon Longten, WSTB, 1988.
4. New York State Dept. of Health. Report of Statewide Surveillance for Radon
in Selected Community Water Systems, Bureau of Public Water Supply
Protection (Albany, NY), 1990.
5. U.S. Environmental Protection Agency. National Primary Drinking Water
Regulations; Radionuclides. Federal Register 56:33050; 40CFR Part a 141,
142; 1991.
6. Whittaker, E.L.; Akridge, J.D.; Giovano, J. Two Test procedures for radon
in drinking water: Interlaboratory collaborative study. Las Vegas, NV:
U.S. EPA Environmental Monitoring Systems Laboratory, Office of Research
and Development; EPA 600/2-87/082; 1989.
-------
TO LUCAS CELL
1
FRITTED DISC .
r
FIRST
GRADUATION
Dl 1 IM/^CD i
rLUIMvatn , '
CLAMP *-C
•
OTnornri/ D
k
u
>
i 1
/
t
Hr
—^b. (V/t^v
^ RTOPrOCK A
^ O 1 \Ji \j\J\Ji\ f\
- RAFtRH
* DrMiiitL
SECOND
GRADUATION
1
THRFADFD AREA
•
Figure 1. Diagrzun of the Collector-Bubbler (charcoal tube not shown)
-------
LUCAS CELL
—STOPCOCK
12/2 BALL JOINTv
12/2 SOCKET^ I LKOVAR TO PYREX SEAL
FINE METERING VALVE
SILICONE SEPTA-
MAGNESIUM PERCHLORATE<
-16 GAUGE HYPODERMIC
NEEDLE
-TUBING
-ASCARITE
COLLECTOR - BUBBLER
Figure 2. Rn transfer system (C-B, not shown, attaches to the drying column)
-------
XIIP-2
MEASUREMENTS OF RADON IN WJ
via
SODIUM IODIDE DETECTORS
Paul N. Houle, Ph.D.
Dept. of Physics
E. Stroudsburg University
E. Stroudsburg, Pa. 18301
&
David Scholtz, M.S.
Prosser Laboratories
Rt. 115 & State Rd.
Effort, Pa. 18330
ABSTRACT
The measurement of radon in water is typically accomplished
with a liquid scintillation system or a lucas cell system.
This paper describes a technique vhich uses Marinelli
Beakers along with the typical radon in air measurement
equipment to determine radon in water concentrations.
Counting times, LLD's, and corresponding analyses are
included.
Using this technique, laboratories that already perform
radon in air measurements will be able to add radon in water
measurements with little added expense.
-------
XMP-3
CONTINUOUS MEASUREMENT OP THK RATYW mm RNTRRTTOW TW WATER
USING ELECTRET ION CHAMBER METHOD
P.K.Hopke Ph.D.,
Clarkson Dniversity
Potsdam, NT 13699-6670
and
P.Kotrappa Ph.D.,
5310 H Rad Elec Inc
Spectrum Drive
Frederick, MD 21701
ABSTRACT
A radon concentration of 300 pCi/L has been proposed by U.S.
Environmental Protection Agency as a limit for dissolved radon
in water for Municipal water supplies. There is a need for a low
cost continuous monitor to ensure that a daily average does not
exceed this limit. A system has been designed that uses the
principle of electret ion chamber. Water flows through a
container at a predetermined low flow rate of about 0.2 liter per
minute. Radon exhaled into the container during the flow is
monitored continuously with a specially designed electret ion
chamber. Electret is removed once a day for measurement and
inserted back after measurement. Calibration factors
derived in comparison with liquid scintillation system are used
to convert the data into an average radon concentration in water.
Sensitivity and error analysis presented in the paper
demonstrated that this relatively low cost instrument can give
satisfactory results at 300 pCi/L.
-------
XIIP-4
PERFORMANCE TESTING THE WD200 RADON IN WATER MEASUREMENT SYSTEM
G. VANDRISH, L DAVIDSON
INSTRUSCIENCE LTD
OTTAWA, CANADA
MAY 20/1992
ABSTRACT
THE SUITABILITY OF THE WD200 SCINTILLATION CELL FOR THE RAPID AND
SIMPLE MEASUREMENT OF RADON IN WATER AND IN AIR HAS BEEN
EXHAUSTIVELY STUDIED BY INTERCOMPARISON WITH NIST REFERENCE
MATERIALS.
THE MINIMUM DETECTABLE LEVEL, SENSITIVITY AND REPRODUCIBILITY,
HAVE BEEN EXAMINED FOR SEVERAL WATER SAMPLING PROCEDURES TO
ESTABLISH THE MOST ACCURATE, SENSITIVE AND RELIABLE METHOD.
PROCEDURES INCLUDE SAMPLING DIRECTLY FROM THE WATER SOURCE OR
DECANTING A BOTTLED SAMPLE LOCATED WITHIN THE UNIT.
STUDIES WERE MADE OF BACKGROUND BUILDUP, MEMBRANE DETERIORATION,
RESPONSE TIME AND STABILITY.
BACKGROUND
THE TWO PRINCIPAL METHODS FOR RADON IN WATER; THE DEGASSING
METHOD AND THE LIQUID SCINTILLATION METHOD HAVE LIMITED PRACTICAL
UTILITY IN THE FIELD.
DEGASSING METHODS ARE LABOUR INTENSIVE, TIME CONSUMING AND
REQUIRE SKILL TO DEPLOY, WHILE LIQUID SCINTILLATION METHODS USE
CHEMICALS, CAN BE EXPENSIVE AND PROVIDE RESULTS AT A LATER TIME.
THE SYSTEM DESCRIBED IN THIS PAPER SATISFIES THE NEED FOR A
QUICK AND ACCURATE FIELD METHOD FOR RADON IN WATER.
THE SUITABILITY OF THE WD200 FOR MEASURING RADON IN AIR AND
RADIUM IN WATER WAS ALSO INVESTIGATED AND PRELIMINARY RESULTS ARE
REPORTED IN THIS RESEARCH.
-------
DESCRIPTION OF WD200 SYSTEM 2
THE WD200 RADON IN WATER MEASUREMENT SYSTEM IS A WATER
SAMPLER/EQUILIBRATION CHAMBER AND SCINTILLATION CELL IN ONE
UNIT. THE SYSTEM IS DESIGNED TO MEASURE RADON OVER A WIDE RANGE
OF ACTIVITIES (2 PCI/L AND HIGHER) QUICKLY AND RELIABLY WITH A
MINIMUM OF OPERATOR INVOLVEMENT OR SKILL. ALTHOUGH THE SYSTEM IS
DESIGNED TO OPERATE WITH SAMPLE VOLUMES-OF 0.1 TO 0.15L, MUCH
LARGER VOLUMES CAN BE EMPLOYED IN SPECIALIZED APPLICATIONS (SUCH
AS RADIUM IN WATER).
TO OBTAIN A MEASUREMENT THE OPERATOR SIMPLY TAKES A WATER SAMPLE
WITH THE EQUILIBRATION CHAMBER (OR INSERTS A FILLED SAMPLE
BOTTLE), WAITS FOUR HOURS FOR THE RADON TO DEGAS PASSIVELY FROM
THE WATER SAMPLE INTO THE AIRSPACE AND DIFFUSE THROUGH THE
MEMBRANE INTO THE SCINTILLATION CELL (WHERE IT COMES TO SECULAR
EQUILIBRIUM WITH ITS DAUGHTERS). A TEN TO THIRTY MINUTE COUNT IS
THEN OBTAINED WITH A SUITABLE SCALAR (RN2000) AND THE RADON IN
WATER CONCENTRATION OBTAINED BY CALCULATION WITH A SIMPLE
FORMULA.
FOR THE 0.1 TO 0.15L WATER SAMPLE VOLUMES EMPLOYED A SENSITIVITY
OF 0.2 CPM/PCI/L AND A MINIMUM DETECTABLE LEVEL OF 2 PCI/1 OF
RADON IN WATER IS POSSIBLE AT 20C.
ALTERNATIVELY, WITH THE EQUILIBRATION CHAMBER REMOVED, THE WD200
CAN BE EMPLOYED AS A PASSIVE RADON GAS SCINTILLATION CELL FOR
RADON IN AIR MEASUREMENTS WITH A SENSITIVITY OF 0.5 CPM/PCI/L.
THIS CALIBRATION VALUE IS REFERENCED TO THE USEPA.
THE RESULTS DESCRIBED HERE ALSO DEMONSTRATE THE POSSIBILITY OF
MEASURING RADIUM IN WATER WITH THE SAME SCINTILLATION CHAMBER
AND A ONE LITER SAMPLE SIZE.
-------
FORMULA RELATING RADON IN WATER CONCENTRATION TO MEASURED COUNT
RATE WITH RN-2000 RADON DETECTOR
Rn(W)=(0.26+Va/Vw)*(C/T/S/D(t))
Rn(W) IS RADON IN WATER CONCENTRATION IN PCI/L
Va IS VOLUME (L) OF AIR (SCINTILLATION CELL +
AIRSPACE)
Vw IS VOLUME (L) OF WATER SAMPLE
C IS TOTAL COUNT (MIN)
T IS COUNT PERIOD (MIN)
S IS SENSITIVITY OF CELL
0.26 IS OSWALD COEFFICIENT (20C)
D(t) CORRECTION FOR DECAY OF RADON FROM TIME OF
SAMPLING
-------
EXAMPLE
TYPICAL PARAMETERS ARE
Va = 0.320 L (VOLUME CELL=0.17L;VOLUME EQUILIBRATION
CHAMBER=0.279L;VOLUME WATER SAMPLE=0.13Lj
Vw = 0.13 L (BOTTLE SIZE)
C = 880 COUNTS IN 10 MINUTES
T = 10 MINUTES (COUNT LONGER FOR BETTER STATISTICS)
S = 0.52 CPM/PCI/L (EPA AND NIST TRACEABLE)
D(t) = 0.97 (4 HOURS DELAY SAMPLING TO COUNTING)
0.26 = OSWALD COEFFICIENT FOR 20C ( +5C=0.42; +10C=0.35;
4-30 = 0.20) TEMPERATURE MUST BE HELD CONSTANT FOR BEST
ACCURACY
THUS:
Rn(W) = (0.26 + 0.32/0.13)*880/10/0.52/0. 97
478 PCI/L
BY TAKING LARGER SAMPLE VOLUMES, OR COUNTING FOR LONGER PERIODS
BETTER SENSITIVITY CAN BE ACHIEVED.
FOR THE SAMPLE SIZE CHOSEN (0.13L), AN APPROXIMATE SENSITIVITY OF
0.2 CPM/PCI/L IS POSSIBLE. THE BACKGROUND OF THE SCINTILLATION
CELL AND THE DETECTOR DETERMINE THE MINIMUM DETECTABLE LEVEL.
THUS WITH A BACKGROUND OF 0 . 2 CPM (NO CONTAMINATION); A
CONSERVATIVE ESTIMATE OF THE MINIMUM DETECTABLE LEVEL WOULD BE 2
PCI/L FOR THE 0.13L SAMPLE SIZE.
-------
SUMMARY OF TESTING
RADON DIFFUSION AND SOLUTION IN THE PLASTIC CONSTRUCTION
MATERIALS, LEAKAGE RATES AND LOSSES, DIFFUSION TIMES FROM
SOLUTION AND THROUGH THE DIFFUSION MEMBRANE WHICH AFFECT
MEASUREMENT TIMING AND SYSTEM UTILITY, BACKGROUND BUILDUP,
CONTAMINATION OR MEMBRANE DETERIORATION, SENSITIVITY, AND
REPRODUCIBILITY WERE ALL ADDRESSED DURING THIS INVESTIGATION.
THERE IS NO EVIDENCE OF LEAKAGE OF RADON FROM THE WD200 DURING
THE CALIBRATION EXPERIMENTS. LEAKAGE WOULD HAVE RESULTED IN LOWER
THAN EXPECTED RADON BUILDUP AND HAVE BEEN ESPECIALLY APPARENT
OVER THE LENGTHY TEN DAY BUILDUP PERIODS EMPLOYED.
RADON AND RADIUM BACKGROUND BUILDUP IN WD200
CONSTRUCTION JjATERIALS
SINCE THE WD200 IS CONSTRUCTED OF ABS AND PVC PLASTIC THERE IS
THE POSSIBILITY OF SOME RADON PENETRATION AND/OR A RESIDUAL
BUILDUP DURING PROLONGED HIGH RADON LEVELS WHICH COULD BE
RELEASED DURING LATER EXPERIMENTS AS A "RADON BACKGROUND".
THIS POTENTIAL PROBLEM WAS INVESTIGATED AFTER EACH CALIBRATION
RUN IN THE FOLLOWING MANNER. THE SOURCE WAS REMOVED, THE
EQUILIBRATION CHAMBER WAS FLUSHED AND THE WD200 WAS RECLOSED
AFTER A PERIOD OF ONE HOUR (*JO PERMIT GAS PHASE RADON TO ESCAPE),
THE SYSTEM WAS THEN FOLLOWED FOR UP TO 10 DAYS TO ALLOW THE
BUILDUP OF RADON FROM ADSORBED RADIUM OR FROM RADON DISSOLVED IN
THE PLASTICS.
NO MEASURABLE RESIDUAL WAS NOTED IN ALL EXPERIMENTS. RADON
PENETRATION AND/OR SOLUTION WOULD HAVE ALSO RESULTED IN LOWER
THAN EXPECTED RADON LEVELS, AN EFFECT WHICH WAS NOT OBSERVED IN
THESE STUDIES.
NO DETERIORATION NOR CONTAMINATION OF THE MEMBRANE WAS NOTED IN
THESE EXPERIMENTS. THE DIFFUSION MEMBRANE CONTACTED THE WATER
SAMPLE ONLY BRIEFLY IN MOST MEASUREMENTS.
-------
DIFFUSION RATES
THE WD200 CELL WAS PLACED OPEN IN A CALIBRATION CHAMBER AND THE
INGROWTH CURVES MONITORED. SECULAR EQUILIBRIUM WAS ACHIEVED IN
THE 3-4 HOUR PERIOD TYPICAL OF RADON DAUGHTERS, INDICATING THAT
THE DIFFUSION RATE THROUGH THE MEMBRANE WAS NEGLIGIBLE IN
COMPARISON TO THE TIME TO ACHIEVE SECULAR EQUILIBRIUM (4HRS).
THE REVERSE DIFFUSION RATE WAS ALSO STUDIED BY REMOVING THE UNIT
FROM THE CHAMBER AND LETTING THE RADON ESCAPE. NO DIFFUSION
CONTROLLED DELAY DUE TO MEMBRANE EFFECTS OR SOLUTION IN THE
PLASTIC WAS OBSERVED IN THE INGROWTH AND DECAY CURVES.
THESE EXERIMENTS WERE REPEATED WITH AN 0.12L WATER SAMPLE. THE
TIME REQUIRED TO ACHIEVE PARTITION EQUILIBRIUM WAS SMALL IN
COMPARISON TO THE TIME TO ACHIEVE SECULAR EQUILIBRIUM.
-------
SAMPLING PROCEDURES
THE ACCURACY OF SAMPLING DIRECTLY WITH THE WD200 EQUILIBRATION
CHAMBER OR WITH A O.13L SAMPLE BOTTLE WAS EXAMINED BY MEANS OF AN
NIST REFERENCE STANDARD. THE STANDARD WAS PLACED IN THE SAMPLE
BOTTLE, SEALED FOR ABOUT ONE DAY, THEN TRANSFERRED BY ONE OF TWO
METHODS TO THE WD200 SAMPLING CHAMBER
IN THE FIRST METHOD THE BOTTLE WAS INSERTED GENTLY INTO THE
SAMPLER, UNCAPPED, AND THE SAMPLER IMMEDIATELY SEALED. THE UNIT
WAS THEN TILTED TO DECANT THE SAMPLE FROM THE BOTTLE.
*TYPICAL RESULTS ARE 988, 970, AND 990 PCI/L FOR A 1000
PCI/L NIST TRACEABLE SAMPLE. IN GENERAL, RESULTS WERE WITHIN
5% OF THE NIST TRACEABLE STANDARD.
IN THE SECOND METHOD THE SAMPLE WAS POURED FROM THE BOTTLE INTO
THE SAMPLER WHICH WAS THEN SEALED.
*TYPICAL RESULTS ARE 733, 888, AND 920 PCI/L FOR A 1000
PCI/L REFERENCE SYSTEM. IN GENERAL, GOOD ACCURACY AND
REPRODUCIBILITY WERE DIFFICULT TO ACHIEVE.
EVIDENTLY THE FIRST PROCEDURE PROVIDES ACCEPTABLE RESULTS WHILE
THERE IS SOME LOSS OF RADON DUE TO POURING IN THE SECOND
PROCEDURE.
THE FEASIBILITY OF USING THE WD200 TO SAMPLE DIRECTLY FROM A
WATER RESERVOIR WITHOUT THE USE OF A SAMPLER BOTTLE IS CURRENTLY
BEING STUDIED.
-------
CALIBRATION
THE WD200 SYSTEM WAS CALIBRATED AGAINST A 1000 PCI/L NIST
TRACEABLE REFERENCE STANDARD. 0.120 LITERS OF THE STANDARD WAS
PLACED IN THE EQUILIBRIUM CHAMBER AND ALLOWED TO DEGAS FOR FIVE
HOURS BEFORE SEALING. THE INGROWTH OF RADON INTO THE DETECTOR WAS
THEN MONITORED CONTINUOUSLY FOR SIX TO TEN DAYS. THE OSWALD
FACTOR EMPLOYED IN THESE CALCULATIONS WAS 0.27 (18C).
TABLE
INGROWTH OF RADON FROM NIST STANDARD/CONTINUOUS MEASUREMENTS
TIME COUNT CALCULATED
SOURCE ACTIVITY
(HR) (MIN) (PCI/L)
67 72.4 1056
96 88.7 996
116 93.9 987
120 104.9 1019
144 113 987
IN ALL CASES, THE CALCULATED AND MEASURED SOURCE ACTIVITIES ARE
WITHIN 5% OF THE NIST TRACEABLE STANDARD.
THE RESULTS DEMONSTRATE THAT AFTER A FIVE DAY INGROWTH PERIOD A
0 12L RADIUM IN WATER SAMPLE OF 1000 PCI/L WILL GENERATE ABOUT
1000 COUNTS/10 MINUTES. THUS BY INCREASING THE SAMPLE SIZE TO
ABOUT A LITER (10X), A RADIUM IN WATER LEVEL OF ABOUT A PCI/L CAN
BE MEASURED AFTER A ONE DAY INGROWTH PERIOD (PROVIDED THAT THE
SYSTEM IS NOT CONTAMINATED). THE ONE LITER SAMPLE SIZE IS EASILY
ACHIEVED WITH SLIGHT MODIFICATION OF THE SYSTEM.
-------
MEASUREMENT OF RADIUM WITH THE WD200
THIS TOPIC WAS INDIRECTLY EXPLORED WITH THE 1000PCI/L NIST
TRACEABLE STANDARD DURING THE CALIBRATION STUDIES. RESULTS
DEMONSTRATE THAT WITH A 130 ML SAMPLE VOLUME A COUNT RATE OF ONE
COUNT PER MINUTE PER PCI/L RESULTS AFTER A FIVE DAY INGROWTH
PERIOD. IF THE EQUILIBRATION CHAMBER IS INCREASED IN SIZE TO
SEVERAL LITERS, A PRACTICAL MEASUREMENT COULD BE MADE AFTER A ONE
DAY INGROWTH PERIOD WITH A MINIMUM DETECTABLE LEVEL IN THE
SUBPICOCURIE RADIUM RANGE.
MEASUREMENT OF RADON IN AIR WITH THE WD200 SYSTEM
THE WD200 SAMPLER SYSTEM OPERATES AS A REGULAR DIFFUSION CELL
WITH A SENSITIVITY OF 0.5 CPM/PCI/L IN AIR.
WD200 RADON IN WATER SCINTILLATION CELL
-------
WD200 RADON IN WATER SCINTILLATION CELL
WINDOW
SCINTILLATION
CELL
DIFFUSION
MEMBRANE
SEALING GASKET
SCREW THREAD
FASTENER
EQUILIBRATION
CHAMBER
-------
^-»•"-.-; • ;_.
• •' '^.fcfc,
•-*r.
-------
XIIP-5
TEMPORAL VARIATIONS IN BEDROCK WELL WATER RADON AND RADIUM. AND
WATER RADON'S EFFECT ON INDOOR ATR RADON
by: Nancy W. McHone, Geologist
Margaret A. Thomas, Environmental Analyst III
CT Dept. of Environmental Protection
Natural Resources Center, 165 Capitol Ave.
Hartford, CT 06106 (203) 566-3540
Alan Siniscalchi, Radon Program Coordinator
CT Department of Health Services Radon Program
150 Washington St.
Hartford, CT 06106 (203) 566-3122
ABSTRACT
Monthly sampling during 1989-1990 of a high radon bedrock well
in Connecticut revealed radon and radium variations of more
than 94% over an 18 month period (1). This research involves
systematic sampling of granitic bedrock radon and radium in
well water and provides data for evaluating single
'representative' well water radon analyses. Five wells are
being sampled hourly, daily, and weekly to document temporal
variations in well water radon and radium. These data are
bSing compared with water use, water chemistry, meteorological
data? and seasonal climatic effects A 'high ' radon w ell
(662 000 pCi/L), three 'moderate- radon wells (1 5,000 - 25 000
pCi/L), and a relatively 'low' radon well (3,500 pCi/L) are
included in the study. Water radon data is compared with
mSnihly aSha track indoor air radon measurements to examine
its derivative and integrated effects on indoor air radon.
Generally, hourly radon measurements rose quickly during the
earty hours of the day with 13% to 48% variation overall.
Daily radon levels varied from 24% to 41%. Radon and radium
analyses" are part of weekly testing which will continue through
Radon variability potentially caused by a difference in
Although the work described in this paper was partially
funded by the U.S. Environmental Protection Agency the
contents do not necessarily reflect the views of the
Agency and no official endorsement should be inferred.
-------
INTRODUCTION
From 1985 to 1989 about 600 private and public bedrock
wells in Connecticut were tested for radon (1, 2, 3). Ninety-
five percent of the wells tested were over the EPA proposed MCL
of 300 pCi/L for public water supplies. It was found that
radon values varied statewide, and by rock type. In
particular, granites and granitic gneisses contained the
largest percentage of wells with elevated radon values (2).
Repeated radon testing was conducted on a well in the
Nonewaug granite by the CT Department of Health Services and
the U.S. Geological Survey. Radon levels in this well ranged
from a high of 660,000 pCi/L to a low of approximately 50,000
pCi/L and these values sometimes changed from the high to the
low value in just one month. Analyses of preliminary data from
one and a half years of nearly continuous monthly water radon
sampling indicated a possible seasonal variation. Radon
values, from another study testing one well frequently over a
period of up to six hours, varied within that short time period
(3). To learn more about radon variation in well water, and to
determine if such change occurs in other wells, a study was
designed to look at temporal variations of radon and radium in
well water. Because both short term and long term variations
were to be examined, the study was designed in three phases.
Wells were to be tested hourly during phase one, daily during
phase two, and weekly during phase three.
STUDY DESIGN
For the purposes of this study, a "low" well has water
radon levels around 3000 to 4000 pCi/L, a "moderate" well is
one in the range of 15,000 to 25,000 pCi/1, and a "high" well
is over 50,000 pCi/L. Five wells are included in the study.
One is the previously mentioned high well in the Nonewaug
granite, three are moderate level wells in the Eastford phase
of the Canterbury gneiss, and one is a low radon level well in
the Hope Valley Alaskite gneiss.
The Nonewaug granite is a two-mica granite containing many
pegmatite bodies (4). Compositionally similar to Nonewaug is
the Eastford gneiss, described as a muscovite-biotite-
microcline-oligoclase-quartz gneiss (5). The Hope Valley
gneiss ranges from albite alaskite to oligoclase quartz
monzonite in composition (6, 7).
PHASE ONE (HOURLY TESTS)
Table 1 illustrates the various sampling parameters of the
three phases of the study.
-------
TABLE 1. TESTING PARAMETERS AND WELL IDENTIFICATION
TESTING PARAMETERS
Water testing interval
Hourly Daily Weekly
Water radon
Water temperature
Barometric pressure
Alkalinity
x
PH
Eh
x
Conductance
x
Dissolved oxygen
x
Water use
Rainfall
Air radon (measured monthly)
Radium
WELLS
Well number Radon range (pCi/L)
Low radon well
Intermediate wells
High radon well
MS-48
ED-10
ED- 15
ED-18
BM-3
3,315
13,325
12,055
15,400
33,560
- 5,953
- 24,725
- 56,825
- 31,330
- 597,815
For the purposes of this study, the terms "low",
"intermediate" and "high" are used merely as relative terms
to distinguish groups of radon values, and are not intended
to be qualitive descriptions for these wells.
-------
Phase one consisted of two days of hourly tests on each of
the five wells. Its purpose was to measure possible radon
variation during a typical "water use day". The testing began
before the home owners arose in the morning and continued until
after they went to bed at night, yielding sixteen to eighteen
tests per day. Each hour a water sample was collected for
radon analysis and the temperature, pH, Eh, alkalinity,
conductance, and dissolved oxygen of the water were determined.
In addition, a record was kept of the barometric pressure and
the amount of time the well pump ran since the last sample,
representing "normal" household water use. The two days of
testing on each well were separated by five to twenty days.
PHASE TWO (DAILY TESTS)
Radon results from the first phase of testing were
examined to determine the time of day at which radon levels
appeared to be most stable (Figure 1). Daily tests were then
conducted during these times of radon value stability. The
same tests for water chemistry and physical characteristics
were performed during this phase as in phase one. Each well
was tested at the same time of day for fourteen consecutive
days. In order to examine the effects of rainfall on radon
levels, rain gauges were installed at each well location and
checked daily.
PHASE THREE (WEEKLY TESTS)
This phase consists of weekly testing of the five wells
for one year. Phase two results were used to determine the day
of the week on which there seemed to be the least variation in
water radon levels. Some consideration was also given to the
laboratory schedule for analysis, resulting in a Wednesday
sampling day. During this phase the same water analyses were
performed as in phase one. and water samples were also
collected for radium-226 and radium-228 analyses. Also alpha
track detectors were installed to monitor the air radon in the
kitchen and basement of each home, to examine if air radon
fluctuations mimic water radon changes. These detectors were
not placed in the high radon home (BM-3), where air and water
are mitigated. The detectors were changed each month and sent
to the manufacturer for analysis. In addition, one year alpha
track detectors were left in each kitchen and basement.
SAMPLING METHODS
Water samples for radon analysis were collected from an
inverted funnel according to the EPA method (8) after running
the water for ten minutes (except as noted below for the daily
samples). Ten ml samples were transferred immediately to 25 ml
glass scintillation vials containing 10 ml of scintillation
-------
Hourly Water Radon
500000
BM-3
250000
8 12 16 20 24
Time of Sample
ED-15
28000 -
o 24000
O.
c 20000
o
16000 -
4200
> 4000
4 8 12 16 20 24
Time of Sample
MS-48
a
o_
3800
3600
ED-10
O
a_
c
o
-o
o
a:
18000 -
16000
8 12 16 20 24
Time of Sample
ED-18
8 12 16 20
Time of Sample
Legend
O Day One
• Day Two
8 12 16 20 24
Time of Sample
Figure 1. Water radon levels for samples taken hourly
on each of two days. Note changes in scales.
-------
fluid. Samples were delivered to the Department of Health
Services Laboratory the next morning for same day analysis on a
UTC Packard Tri-Carb 4530 Liquid Scintillation Spectrometer.
Water was collected from an outside tap at each house except at
BM-3, where water was collected directly from the pressure tank
outlet via a garden hose. During the winter, one well was
tested by collecting water from the kitchen faucet. For the
water chemistry tests, probes of the various instruments were
inserted through holes in the styrofoam lid of an acrylic box.
The box was connected via rubber tubing to a garden hose
attached to the outside tap. Water was allowed to run through
the box and thus around the probes. The water temperature was
also recorded in the box. The probes measured dissolved oxygen
(YSI Model 57 Oxygen Meter), Eh (Markson Model 93 Eh Meter), pH
(Orion Research Model 211 pH Meter) and conductance (Markson
Conductance/TDS Meter). Alkalinity was measured by titration
with a LaMotte Chemical Alkalinity Test Kit Model WAT-MP-DR. A
digital time totalizer from GFA Engineering measured, by
induction, current traveling to the well pump each time the
pump operated.
RESULTS
Nine previous radon tests of the Eastford phase of the
Canterbury gneiss yielded a geometric mean of 18,000 pCi/L with
a range from 3,000 to 35,000 pCi/L (1, 2). The ten
exploratory tests done for this study on the same geologic unit
have a geometric mean of 28,000 pCi/L with a range from 17,000
to 74,000 pCi/L.
RADON
Hourly tests
Radon levels in the high well (BM-3) rose at the beginning
of the first day from a low, at 6 AM, of 435,460 pCi/L to a
high, at 9 AM, of 488,390 pCi/L (Figure l). This high was
maintained until about 3 PM, at which time values began a
decline. This decline continued to the end of the testing day,
when the radon was back at the morning low. The second day of
testing, twenty days after the first, began with a 6 AM test
value of 297,860 pCi/L and rose to 318,900 pCi/L at 9 AM.
After an hour at that level there was a steady decline to a low
of 251,040 pCi/L at the end of the day at 10:45 PM, a
difference of 27% in 12 3/4 hours.
Changes in the moderate wells during a day of testing were
less dramatic. Although total radon values were lower, they
followed a similar trend of rising swiftly at the beginning of
the water use day. ED-10 rose from a low of 16,145 pCi/L at 5
AM to a high of 21,335 PCi/L at 9 AM (Figure 1), an increase
-------
of 32%, before dropping back to around 19,000 pCi/L for six
hours, then increasing to around 21,000 pCi/L for the rest of
the day. The second day of testing, eight days later, began
with a low of 16,560 pCi/L at 5 AM, climbing to a high of
21,680 pCi/L at 10 AM, before a small decline occurred followed
by a gradual rise for most of the rest of the day. Maximum
levels reached 21,675 pCi/L, a total difference of 5120 pCi/L
for the day, a 31% increase.
The second moderate well, ED-15, showed a pattern similar
to ED-10's on its second day of testing, with an overall
increase of 8,435 pCi/L, or 42% (Figure 1). The first test
day, however, produced much less variation. At 5 AM the radon
level was 21,375 pCi/L, rising to 22,815 pCi/L at 7 AM. It
stayed around 22,000 pCi/L until 6 PM, when it dropped suddenly
to 15,030 pCi/L. At 9 PM the radon level rose to 23,260
pCi/L, to 25,175 pCi/L at midnight, then dropped to 20,960
pCi/L at 1 AM.
Moderate well ED-18 basically rose quickly in the morning
of both days, to highs at 11 AM and noon, then gradually
declined (Figure 1) . On day one the second sample was lower
than the first, but radon values then increased, and there was
a sharp anomalous decrease at 5 PM. Excluding this anomalous
decrease, the variations were 43% and 58% over the two days.
The radon pattern of the low well, MS-48, was different
(Figure 1), as there was no morning increase. Instead, on both
days, the radon levels decreased gradually all day from morning
highs. Variations during each day of testing were 16% and 5%.
Attempts to correlate radon values with water use resulted
in statistically significant (Student's t test, t = 1.33 or
higher at 10% significance level) correlations on only four of
the ten days of testing (Table 2) . Three of these were
positive correlations, the other negative. The low well (MS-
48) had positive correlations on both days, the high (BM-3) and
one moderate well (ED-15) had correlations on one day each, and
the other two moderate wells showed no correlations. All wells
had significant correlations between radon and conductance, but
MS-48 and ED-15 each correlated on only one day. Two of the
other wells had a negative correlation one day, positive the
other. Correlations with other parameters were more variable
(Table 2) .
Daily tests
During the daily tests an additional question was
addressed — is there a difference when collecting samples
between running the water two minutes — as recommended by EPA
— and running it ten minutes — as is commonly practiced. As
-------
TABLE 2. T-TEST RESULTS FOR RADON AND OTHER PARAMETERS
HOURLY TESTS
BM-3-1 BH-3-2 ED-10-1 ED-10-2 EO-15-1 ED-15-2 ED-18-1 ED-18-2 MS-48-1 MS-48-2
Water temperature
Barometric pressure
Alkalinity
PH
Eh
Conductance
Dissolved oxygen
Water use
DAILY TESTS
Water temperature
Barometric pressure
Alkalinity
P«
Eh
Conductance
Dissolved oxygen
Water use
Rainfall
WEEKLY TESTS
Water temperature
Barometric pressure
Alkalinity
P«
Eh
Conductance
Dissolved oxygen
Water use
Rainfall
Ra-226
COMBINED TESTS
Water temperature
Barometric pressure
Alkalinity
PH
Eh
Conductance
Dissolved oxygen
t
1.87
4.39
-1.43
-1.13
1.63
-1.43
-1.62
-0.57
Btt-3
t
-3.54
2.34
-2.03
-3.36
1.81
-0.56
1.76
-2.79
0.53
2.42
1.68
6.80
-0.15
-0.36
3.23
-6.71
0.39
-0.91
2.45
4.29
3.22
6.27
6.29
-3.72
3.1*
-4.06
t
-2.19
-1.03
2.80
1.25
2.64
1.47
0.16
2.12
ED- 10
t
-0.45
-2.60
-1.19
-2.26
5.34
-2.03
0.82
4.49
0.24
-0.80
-1.08
0.55
-0.03
-0.83
1.97
2.36
1.36
-0.67
0.43
-0.32
-0.81
-1.26
3.09
0.57
-6.35
t
--
5.32
-1.08
0.85
3.85
-1.43
1.56
-0.08
ED- 15
t
1.61
-0.04
-1.32
2.32
0.27
-2.37
3.46
0.61
-1.21
-0.04
1.95
0.11
0.60
-1.20
0.22
-0.30
2.98
-0.29
4.83
0.26
5.70
-0.15
-1.92
0.12
0.23
t
1.72
0.18
-4.38
-5.04
2.57
-2.13
1.26
-0.40
ED- 18
t
0.20
-0.50
0.01
-1.36
-0.43
-1.67
0.68
1.24
0.03
0.31
-0.30
-1.07
-0.76
0.38
-1.55
-0.64
2.29
-0.69
2.18
-2.78
5.25
0.80
-4.02
-2.98
-0.29
t t
-0.67 3.04
0.59 2.81
0.44 1.65
0.56 -3.80
0.74 -2.85
1.14 -6.20
0.06 2.32
0.66 -2.22
MS-48
t
-0.21
0.34
-1.35
-0.86
-0.17
0.49
0.83
-1.45
-0.35
-0.49
-1.03
1.55
-1.52
-3.03
1.03
-2.44
0.72
-6.60
2.50
-0.83
2.39
1.05
-1.62
0.53
-2.97
t t t t
-0.63 -0.39 -1.40 1.51
4.30 -2.51 1.36 1.91
-0.98 -0.85 1.76 -2.11
0.16 -0.40 1.71 0.13
-0.11 -0.36 2.22 0.04
1.54 -2.31 1.76 0.02
-1.45 -1.07 2.87 -0.16
-0.63 0.65 2.90 1.50
Note: t = 1.33 at 10%
significance level
for hourly tests,
t = 1.34 for daily
tests, and t = 1.30
for weekly and for
combined tests
-------
can be seen from Figure 2, ten minute radon levels are
generally slightly higher than two minute levels, an average of
about 4.3% higher. Over the fourteen days of tests, variation
within the five wells ranged from 16% to 57% for the two minute
samples, and from 24% to 41% for the ten minute samples. The
overall higher radon values of the ten minute samples may be
partially attributed to water sampled directly from the well,
rather than a mixture of well and pressure tank water, as may
occur when the water is run for only two minutes. If pressure
in the tank was low when the water was turned on, fresh water
would begin entering the tank sooner than if the initial
pressure was higher, thus delaying the onset of pumping and the
flow of fresh water into the tank.
The different water use patterns of the well users (Figure
3) were expected to influence the radon levels we would find
during the daily water sampling. The well pump electrical
current, as measured by the time totalizer, was used as a
determination of water use. The correlations of radon with
water use are variable. The high and low wells both had
negative correlations (t = -2.79 and -1.45, respectively, t =
1.34 at 10% significance level). One of the moderate wells had
a positive correlation (t = 4.49), while the other two did not
show significant correlations (t = 0.61 and 1.24). The
coefficients of variation for radon during these two weeks of
daily tests ranged from 0.08 for BM-3, through 0.09 for MS-48,
0.10 for ED-10 and ED-18, to 0.25 for ED-15.
Significant correlations of radon with water chemistry
indicators is variable (Table 2). The best correlation is
between radon and pH - three wells had a negative correlation,
one a positive correlation, and only MS-48 had no correlation.
This compares with only three days of significant correlations
during the ten days of hourly testing.
Weekly tests
The weekly tests began on 11 September 1991 and will
continue for a year. As of this writing forty-two weeks of
testing are complete. Except for the high well, radon levels
declined gradually to mid-winter (Figure 4). ED-15 then
continued to decline, while ED-10, ED-18 and MS-48 began a slow
increase. Coefficients of variation over this time have ranged
from 0.06 for MS-48 through 0.21 for ED-15 (which was 0.25
during daily testing).
The greatest variation, 0.73, occurred in the radon values
from the high well - from a high of 597,820 pCi/L to a low of
33 560 PCi/L . in mid-November, 1991, radon in BM-3 dropped
from 482 380 pCi/L to 33,560 pCi/L in one week, stayed near
this level for four weeks, then increased again. Another drop
-------
Daily Water Radon
BM-3
ED-10
^ 420000 ,r
< *
O 390000 -
Q.
C
o
T>
D
o:
360000 -
330000 -
300000
o
a.
o
TJ
O
SAT MON WED FRI SUN TUE THU
Day of Sample
ED-15
c 30000 -
o
TJ
. 20000
SAT MON WED FRI SUN TUE THU
Day of Sample
MS-48
4000
O
a.
c
o
T3
O
Ct
O
Q.
C
O
T>
O
o:
SAT MON WED FRI SUN TUE THU
Day of Sample
ED-18
28000 -
20000 -
SAT MON WED FRI SUN TUE THU
Day of Sample
Radon Samples
o 2 minute
• 10 minute
SAT MON WED FRI SUN TUE THU
Day of Sample
Figure 2. Daily radon levels for all wells, showing water radon values
for samples collected after running water 2 minutes and 10
minutes. Note changes in vertical scales.
-------
BM-3
ED-1.0
o
CL
c
o
T>
D
ac.
450000
400000
350000
300000
0.4
SUN TUE THU SAT MONWED FRI
Day of week
ED-15
O
CL
c
o
D
cr
3
o
0)
v>
3
l_
a>
15
0.4
SUN TUE THU SAT MONWED FRI
Day of week
MS-48
0.8
0.6
3
O
a>
"5
3500
SUN TUE THU SAT MONWED FRI
Day of week
o
a_
c
o
-o
o
a:
O
_c
0.6 >-
0.4
SUN TUE THU SAT MONWEO FRI
Day of week
ED-18
28000 -
a>
o
SUN TUE THU SAT MONWED FRI
Day of week
Legend
• Radon
O Water use
Figure 3. Daily radon and water use in all weils.
-------
Weekly Water Radon
BM-3
ED-10
o
Q.
C
O
-D
D
CC
600000 h
500000
400000
300000
200000
100000
10/2712/262/24 4/24 6/23
Date of Sample
ED-15
o
CL
c
o
TJ
o
ct:
26000
24000
o 22000
a.
20000
8000 I-
6000
10/27 12/26 2/24 4/24 6/23
Date of sample
ED-18
10/2712/26 2/24 4/24 6/23
Date of sample
10/2712/262/24 4/24 6/23
Date of sample
MS-48
o
Q.
D
ce
10/2712/26 2/24 4/24 6/23
Date of sample
Figure 4. Weekly radon levels for all wells. Solid line is second
order regression. Note changes in vertical scales.
-------
occurred in early March, from 440,670 pCi/L to 62,420 pCi/L in
one week. This time radon stayed at this level for an
additional eight weeks before rising again to its higher level.
Then in June, over two weeks, the radon values went from
486,872 pCi/L to 45,638 and back to 439,335 pCi/L. Because of
the huge variation found in this well, duplicate samples are
routinely collected each week.
During phase three, correlations of radon with water
parameters were best for conductance, dissolved oxygen and
water use, all of which had three significant correlations for
the five wells (Table 2). However, only with water use were
the correlations all of the same sign.
RADIUM
Except for BM-3, radium-226 ranged from 0 to 4.32 pCi/L
with arithmetic means from 0.6 to 1.4 pCi/L for the individual
wells. With error bars of approximately 0.7 pCi/L, there is
little perceived variation in the radium-226 values. In BM-3
radium-226 ranged from 0.1 to 70.8 pCi/L, with an arithmetic
mean of 38.9 pCi/L (Figure 5). There is a good positive
correlation between radium-226 and radon in this well (Table 2
and Figure 5). Other researchers have found groundwater levels
of radium-226 to be generally lower than average in the
northeastern United States (10) and that radium does not
necessarily correlate with radon (9, 11).
Radium-228 values for all wells yield arithmetic means of
between 0.5 and 0.7 pCi/L. However, with error bars greater
than the reported values in most cases, it is difficult to
ascertain variation in the radium-228.
AIR RADON
Air radon results are shown in Figure 6. Alpha track
detectors from the Terradex Corporation were placed in the
kitchen on the wall near the sink and on a wall in the basement
of each test home. Homes ED-15 and ED-18 are less than five
years old, BM-3 was built in 1970, ED-10 about 1935, and MS-48
in 1928 (but on a foundation built around 1900). Except in
home ED-15, there is a washer and dryer in each basement. All
basement walls are concrete except ED-10, which is built of
unmortared fieldstones.
Coefficients of variation for monthly alpha track readings
are large, ranging from 0.34 to 1.27 since September. Basement
air radon at home ED-10 has an arithmetic mean of 10.8 pCi/L.
From a value of 7.5 pCi/L for September it rose to 18.5 pCi/L
for November, then declined. Kitchen air radon began with its
highest value - 3.1 pCi/L - then declined. Its arithmetic mean
-------
BM-3 The "high" well
u
Q.
c
o
°
600000
500000 -
400000 -
300000 h
200000 -
100000 -
10/27 12/26 2/24
1991 Date of test
4/24 6/23
1992
Figure 5. Radium-226 and radon in well BM-3.
-------
Air radon
o
16
o 12
Q.
4
0
ED-10
OCT DEC FEB APR
Month of test
o
c
"D
o
12 h
8
4
0
ED-15
OCT DEC FEB APR
Month of test
ED-18
MS-48
o
CL
c
o
a:
OCT DEC FEB APR
Month of test
c
0
T3
O
OCT DEC FEB APR
Month of test
Legend
O Kitchen air radon
• Basement air radon
Figure 6. Air radon from four week alpha tracks. Note
changes in vertical scales. BM-3 is not represented here, as
its air radon has been mitigated.
-------
is 1.9 pCi/L. Water radon for this well has increased overall,
although it declined from mid-December through late February
(Figure 4).
The large increase of water radon in ED-15 during
December, 1991, occurred during a month in which kitchen air
radon declined slightly. The ten-fold increase in basement air
radon in Ed-15 in March 1992 is an anomaly, as there were no
significant changes in water radon or in kitchen air radon.
Home ED-18 had a steady increase in kitchen air radon from
October through March, while the water radon decreased from
September through December, then increased. The basement air
radon generally decreased during this time.
The low well, MS-48, began with the highest kitchen radon
(6.3 pCi/L) at a time the water radon was declining from a
high. Except for the first month, the kitchen and basement air
radon are very similar (Figure 6).
DISCUSSION
The solubility of uranium and radium in water is partially
limited by the Eh and pH of the ground water (9). Radon,
however, being an inert gas, is not so affected. Comparisons
of radon levels in well water with Eh and pH of the water gave
mixed results. When the pH and radon results from all phases
of testing are combined for each well, only the high well, BM-
3, shows a statistically significant correlation between pH and
radon (Table 2). As an examination of Table 2 shows,
correlations between pH and radon for the individual phases are
quite variable.
Statistically significant correlations between Eh and
radon occur for all welli, when all of the data is combined,
although ED-10 has a positive correlation while the other four
are negative. As with pH, the individual correlations for Eh
within the three phases of the study are variable (Table 2).
For the combined results, when the relationship between
dissolved oxygen and radon is examined for each well, there are
three significant negative correlations, in wells BM-3, ED-10,
and MS-48. Within the individual phases, dissolved oxygen -
radon correlations occur in ten of the twenty tests (Table 2).
The alkalinity tests showed statistically significant
positive correlations between alkalinity and radon in all wells
except ED-10, for the combined data.
For the individual testing phases, the largest number of
correlations between radon and water chemistry occurred with
-------
conductance, a measure of total dissolved solids (Table 2).
These were both positive and negative correlations. When the
combined test results are examined, statistically significant
correlations between conductance and radon occur only in BM-3
and ED-18; for BM-3 the correlation is positive, for ED-18 it
is negative.
There seem to be no consistent significant correlations of pH,
Eh, dissolved oxygen, alkalinity or conductance with radon in
all five wells. Szabo and Zapacza (9) reached this same
conclusion for groundwaters in the Newark Basin.
Statistically there was correlation between water use and
radon on only four of the ten days of hourly testing, three
positive and one negative. However, water use seemed to have a
dramatic effect on early morning radon levels in all except the
low well, MS-48, as radon levels rose quickly during the first
few hours of the water use day (Figure 1) . The low well, MS-
48, however, had positive correlations between radon and water
use for hourly testing on both days. BM-3 had a statistically
significant positive correlation between radon and water use on
one day of the hourly tests and ED-15 had a negative
correlation on one day. The other six days of tests produced
no significant correlations.
On the daily tests MS-48 and BM-3 had negative
correlations between radon and water use, while ED-10 had a
positive correlation (Table 2). On the weekly tests there are
significant positive correlations between water use and radon
for wells ED-10, ED-15 and ED-18. This variation seems to
indicate there is no relationship between radon and water use
other than the early morning depression of water radon levels.
The large variation in water radon levels found in the
high well, BM-3, is not reflected in any of the water
parameters for which testing was done, including water use and
rainfall. The variation may be caused by seasonal changes in
groundwater levels which bring in radon-rich water at some
water levels and radon-poor water at others. The well
penetrates 235 feet of granite which is known to be highly
pegmatitic (4) . Some of the water moving into the well must
pass through a pegmatite rich in uranium and/or radium bearing
minerals. However, comparison of rainfall with radon does not
reveal any recurring pattern of rainfall amounts before or
during the times of sudden low radon levels. Measurements of
groundwater levels in the area or water levels in the well
itself may prove useful in testing this theory, as may
examination of the bedrock mineralogy to look for radon
sources and transport paths.
-------
AIR RADON
Air radon results are summarized in Figure 6. Both
basement and kitchen air radon levels have generally decreased
in home ED-10, while the water radon levels have gradually
increased over the September to June weekly testing period.
The nearby ED-15 home also has an overall decrease in air
radon, while the water radon has also decreased. Home ED-18
has mixed results for air radon. After an initial drop,
kitchen radon steadily increased while basement radon decreased
overall. The water radon decreased from September through
December, then increased (Figure 4).
Kitchen and basement air radon are remarkably similar to
each other at the home of the low well (MS-48). The kitchen
radon declined rapidly at the beginning of the study, then
increased slightly in December. Basement radon has increased
slightly throughout the study. At the same time, water radon
decreased through September and October, then increased (Figure
4).
These air data are being evaluated for the water borne
radon contribution. Of the four homes in this study in which
air radon was monitored along with water radon, it appears that
in the kitchens of homes ED-15, ED-18 and MS-48 a significant
amount of the air radon could be coming from the water. In
those homes kitchen air radon value trends seem to follow the
trends of the water radon values. Basement air values also
follow the water trends in homes ED-15 and MS-48, possibly due
to air flow between'the two levels of the homes. A full
diagnosis of the homes is required in order to identify
dominant indoor air radon sources originating from water, soil
or bedrock sources.
EARLY IMPLICATIONS OF THE STUDY
This study has demonstrated that variation occurs in the
radon levels found in water wells. Early morning well samples
are likely to give abnormally low radon results for that well.
Wells tested later the same day may yield radon results as much
as 58% higher than the early morning test. Seasonal variations
were also found in well water radon. Wells ED-18 and MS-48 had
lower water radon in early winter, followed by increases
beginning about mid to late December. Water radon in well ED-
10 has increased overall throughout phase three of the study,
although with some variability. The high well, BM-3, had three
sudden large decreases followed by more gradual increases in
its water radon levels (Figure 4).
These results indicate a need for multiple testing of
wells being considered for mitigation, to determine the maximum
-------
range of radon levels likely to occur in that well. The high
well, BM-3, is an excellent example of the need for multiple
testing. If it had been tested while at one of its low values,
mitigation equipment might have been designed for a threshold
of 40,000 pCi/L of radon from the water. When the levels rose
to 600,000 pCi/L the equipment would likely be inadequate.
Because this well's second low period lasted nine weeks,
seasonal tests may avoid multiple tests being done in one of
these lows.
Variations in radium-226 levels in BM-3 were also
significant enough to cause the radium problem to be overlooked
if initial testing yielded a low value (Figure 5).
The results of this study to date indicate that a single
well water radon analysis is not necessarily representative of
the radon character of the well. Multiple water radon tests
over several seasons are likely to be required to assess the
range of radon values. These values can vary by as much as 94%
for an individual well. Radon values have been found to be
most stable from mid-morning to mid-afternoon, possibly due in
part to water use and well recharge dynamics.
ACKNOWLEDGMENTS
The authors especially wish to thank Denis Healy of the
U.S. Geological Survey for his invaluable help and
encouragement, without which this study would not have been
possible. And we also wish to thank the owners of the wells
used in the study for their gracious participation and
cooperation.
-------
REFERENCES
1. Dupuy,C.J., Healy, D., Thomas, M.A., Brown, D.R.,
Siniscalchi, A.J. and Dembek, Z.F. A survey of naturally
occurring radionucleides in groundwater in selected bedrock
aquifers in Connecticut and implications for public health
policy. In: Gilbert, C.E. and Calabrese, E.J., (eds.)/
Regulating drinking water quality. Lewis Publishers,
Chelsea, Michigan, 1992. p. 95.
2. Rothney, L.M. A survey of radon-222 occurrence in
Connecticut private well water; assessing geologic and
hydrologic parameters. M.S. Thesis, Yale University, 1987.
160 pp.
3. Torgersen, T., Mackie, D., and Benoit, J. Geologic control
of Rn-222 concentrations in groundwater and its impact on
the health risks of indoor Rn-222. Unpublished report,
1988. 29 pp.
4. Gates, R.M. The bedrock geology of the Woodbury
Quadrangle. State Geological and Natural History Survey of
Connecticut, Quadrangle Report 3, 1954. 23 pp.
5. Dixon, H.R. and Pessl, F., Jr. Geologic map of the Hampton
Quadrangle, Windham County, Connecticut. U.S. Geological
Survey, Map GQ-468, 1966.
6. Snyder, G.L. Bedrock geology of the Willimantic
Quadrangle, Connecticut. U.S. Geological Survey, Map
GQ-355, 1964.
7. Rodgers, J. Bedrock geologic map of Connecticut.
Connecticut Geological and Natural History Survey, 1985.
2 sheets.
8. Radon in water sampling program. EPA/EERF-MANUAL 78-1,
U.S. Environmental Protection Agency, Montgomery, Alabama.
11 pp, 1978.
9. Szabo, Z. and Zapecza, O.S. Geologic and geochemical
factors controlling uranium, radium-226, and radon-222 in
groundwater, Newark Basin, New Jersey. In: Gundersen,
L.C.S. and Wanty, R.B., (eds.), Field studies of radon in
rocks, soils and waters. U.S. Geological Survey Bulletin
1971: 243, 1991.
-------
10. Paulsen, R.T. Radionuclides in ground water, rock and
soil, and indoor air of the northeastern United States and
southeastern Canada — a literature review and summary of
data. In: Gundersen, L.C.S. and Wanty, R.B. (ed.), Field
studies of radon in rocks, soils and waters. U.S.
Geological Survey Bulletin 1971: 195, 1991.
11. Szabo, Z. and Zapecza, O.S. Relation between natural
radionuclide activities and chemical constituents in ground
water in the Newark Basin, New Jersey. In; Graves, B.
(ed.), Radon, radium, and other radioactivity in ground
water: hydrogeologic impact and application to indoor
airborne contamination. Proceedings of the NWWA Conference.
Somerset, NJ, 1987. p. 283.
• U.S COVERNMEVTPWNTINCOFFICI:! 992 -6<.e -00 3/60 021
------- |