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
x°/EPA The 1992 International
Symposium on Radon
and Radon Reduction
Technology:
Volume 4. Preprints
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 I-7
-------
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 II-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 Hoornbeek, 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.
Willhoit, Orange County, NC, Board of Commissioners III-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. Poppell, 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. Bernardin, University of
Wisconsin-Milwaukee VI-2
Data and Models for Radon Transport Through Concrete
Vern C. Rogers and K. Nielson, 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, ACUREX Corp 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 VIM
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 VIII-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
Bj0rn 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. Fisk,
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. B0hmer, 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 IXP-5
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-NuclearGeotech, 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, Inc XllP-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 Health Services XIIP-5
XVI
-------
Session XI
Radon Prevention in New Construction
-------
THE EFFECT OF RADON-RESISTANT CONSTRUCTION TECHNIQUES IN A
CRAWLSPACE HOUSE
by: C. S. Dudney and D. L. Wilson
Health and Safety Research Division
Oak Ridge National Laboratory
Oak Ridge, TN 37831-6113
and
T. M. Oyess
Air and Energy Engineering Research Laboratory
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
ABSTRACT
An extensive battery of radon diagnostic tests were performed
on two different houses built on the SAME foundation. A house was
studied as part of a multi-year evaluation of radon mitigation
measures in the Tennessee Valley. The original house was destroyed
in a fire caused by a faulty space heater. The second house was
built on the same foundation and incorporated passive techniques to
reduce radon transport via the crawlspace to the living area. The
rebuilt house achieved a 60% reduction in the volumetric transport
of air from crawlspace to living area when the heating and cooling
system was not operating. An even greater reduction was found when
the heating and air conditioning system was operating, reflecting
the benefits of new, well-sealed ductwork. In spite of these
enhancements, the passive measures failed to reduce radon
concentrations to below 4 pCi/L. The primary reason is believed
to be the enclosure of the open carport to form a garage at the
time of reconstruction.
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.
Research sponsored by U.S. Environmental Protection Agency
under Interagency Agreement DW89935096-01-0 with the U.S.
Department of Energy.
Managed by Martin Marietta Energy Systems, Inc., for the U.S.
Department of Energy under contract DE-AC05-840R21400.
*** T
1 pCi/L « 37 Bq/m
-------
INTRODUCTION
According to recent state and Federal surveys, radon levels in
excess of the Environmental Protection Agency (USEPA) guideline of
4 pCi/L are widely distributed throughout the U.S. As a result, the
USEPA has supported on a regional basis research programs to
quickly develop mitigation technologies required for variant macro-
geologic and house-construction characteristics. Also of major
national importance is the development of low cost radon resistive
construction techniques for new construction. For existing homes,
mitigation options are generally limited to active or mechanical
systems. These systems, although effective initially, are not
totally fail safe. Occupant interaction, performance degradations,
and other problems may eventually lead to elevated radon once
again. The ability to build into a home a permanent, non-
mechanical, maintenance free radon mitigation system is important
to the nation.
When a fire unexpectedly destroyed part of House 11 (Hll) in
Huntsville, AL, a rare opportunity became available to test several
theories about radon entry and transport. The home, a single-story
ranch with crawlspace, was one of four in Huntsville that in 1987
had been fully instrumented with monitors, sensors, and a complete
weather station as part of a USEPA radon mitigation research
program. For 3 years, detailed data were gathered from Hll. Then,
in November 1990 a fire destroyed over 30% of the home
superstructure. Because the superstructure suffered considerable
damage, the entire home with the exception of the crawlspace walls
had to be rebuilt.
Based on the data collected in Hll before the fire, three
major radon entry problems were identified: excessive floor
penetrations, leaky forced air return ductwork, and the location of
the heating and air conditioning (HAG) blower in the crawlspace.
Working with the contractor, these problems were either eliminated
or minimized during reconstruction. The following paper summarizes
the data collected before the fire and the effectiveness of the
passive design after the fire.
STUDY DESIGN
Before the fire of November 1990, Hll was a single-story
house, built around 1967, with a full crawlspace below the living
area. The vapor barrier in the crawlspace had noticeable punctures.
There was an attached carport with slab-on-grade construction. The
house was heated and cooled by an electric heat pump with the
blower unit and ductwork located in the crawlspace. The building
site is on flat land with many trees in the backyard. During the
time of measurements before the fire, occupants of the home
included two adults and two children, of whom two or three were
home most of the time.
-------
The initial (i.e., before fire) study design included a
variety of radiological, environmental, and house dynamic
measurements with near-continuous and episodic data acquisition
schedules. Continuous measurements of radon, temperature, and
relative humidity were performed in substructural (i.e.,
crawlspace) and superstructural (i.e., living area) zones.
Differential pressures were monitored between substructural and
superstructural zones and between substructural and outdoor zones.
These monitors were also used on an episodic basis to evaluate the
impact of the clothes dryer and MAC system operation on
differential pressures across various building membranes.
Episodic measurements included (1) building leakage area, and
(2) selected infiltration, exfiltration, and interzonal transport
rates, performed using both passive/integrated and active/real-time
test devices. Most episodic experiments were performed as radon
diagnostic measurements to support the selection, implementation,
evaluation, and optimization of radon mitigation techniques. Blower
door tests were used to characterize the overall leakiness of the
houses and substructural compartments. Integrated measurements of
interfloor transport and air infiltration rates were evaluated for
approximate 1 month periods with passive perfluorocarbon tracers
(PFTs). Substructure to superstructure transport rates were also
measured in brief, real-time, tracer gas experiments to evaluate
the impact of forced-air HAC systems. Both the integrated and
real-time measurements were performed prior to and following the
installation of selected weatherization and/or radon mitigation
systems to evaluate their impact on various interzonal air flows.
Diagnostic tests emphasized infiltration and exfiltration of
the crawlspace and interzonal transport between the crawlspace and
upstairs living compartments. In addition to the PFT and
substructure to superstructure transport experiments, single
compartment measurements of air exchange rate were performed in the
crawlspace as a function of vent area and the operation of a
submembrane depressurization system. These air exchange
measurements were used to quantify the effectiveness of selected
passive and active mitigation systems when strong fluctuations in
radon availability hampered systematic analysis of the radon
concentration data.
In addition to the extensive radon diagnostic measurements,
several factors in the study design strongly influenced the
selection and implementation of radon control measures in the study
house. A progressive series of weatherization, HAC duct sealing,
passive crawlspace venting, and active source control measures were
employed. Quantitative estimates of their relative effectiveness
were limited by large, seasonal variations in radon source
strength.
-------
MEASUREMENT METHODS
Measurements performed in support of this radon mitigation
study were collected on continuous and episodic bases. In the study
house, continuous measurements were automatically acquired every 6
seconds, accumulated or averaged every 30 minutes, and subsequently
recorded in electronic data loggers. Episodic measurements were
manually performed at selected intervals in the study to enhance
comparisons between pre- and post-mitigation periods, pre- and
post-weatherization periods, and on/off periods of HAC operation.
The experimental designs, equipment, and basic methodology are
described below.
RADIOLOGICAL MONITORING
Continuous radon monitors were constructed at Oak Ridge
National Laboratory based on techniques developed by Wrenn (1) and
modified by Perdue et al. (2) . Room air entered the internal volume
of the instrument by diffusion through two layers of foam that
excluded short-lived decay products. Ionic species resulting from
the decay of Rn then accelerated through an approximate 900 V
electrostatic field to an aluminized plastic sheet. Subsequent
alpha decay from these ions excited a thin layer of scintillator
(i.e., zinc sulfide) located beneath the sheet. The scintillator
emitted a burst of photons that were focussed through a pipe to a
photon counting photomultiplier. The electronic pulse from the
photomultiplier was then broadened, amplified, and counted with a
data logger.
Wrenn chamber zero response factors and radon sensitivity
factors were periodically measured to track the performance of the
instruments. Sensitivity factors averaged typically
0.8-1.5 cpm/(pCi/L). Zero response factors averaged typically
0.7-1.5 cpm with a few notably higher exceptions. The lower limits
of detection for these instruments were approximately 0.5-1.0
pCi/L, at which concentration the signal-to-background ratios are
approximately unity.
Episodic measurements of radon gas from soil tubes and
naturally occurring openings in the backyard (i.e., holes in
surface soils and rocks) were performed on a periodic basis to
track seasonal variation in radon source strength. Grab sampling
was performed with Lucas cells, whose responses were calibrated
with laboratory generated atmospheres of known concentration.
TRACER GAS: PASSIVE MEASUREMENTS
Integrated measurements of air infiltration, exfiltration, and
intercompartment transport were performed using PFT sensors
purchased from the National Association of Home Builders (NAHB).
Two PFTs were used in these studies to differentiate between
substructural and superstructural zones. PFT sources and receivers
-------
were placed in accordance with NAHB guidelines. One PFT source was
typically placed for every 28-37 m of floor area in each substructural
and super structural zone. A single PFT receiver was placed at two
separate locations in each substructural and super structural zone. One
additional replicate and one additional blank (i.e., unexposed)
receiver were placed alongside the PFT receivers located in the
superstructure. Receivers were exposed for 4 to 8 week periods in which
the house was undisturbed by research or mitigation activities that
would significantly impact the measured air flows within the house.
CRAWLSPACE TO LIVING AREA TRANSPORT RATE
Near-real time measurements of gaseous transport between
crawlspace and living-area zones were performed in the study house to
investigate the impact of HAC operation. Refrigerant gas (R12) was
briefly injected and mixed in the crawlspace and then monitored over
time using independent infrared spectrometers in the living area and
crawlspace. To determine the crawlspace to living area (i.e.,
interfloor) volumetric transport rates, a simple two compartment model
was used. The model assumed uniform mixing in both the crawlspace and
living area, and zero concentration outside of these two compartments.
The volume of the living area times the change in refrigerant
concentration in the living area (i.e., A[Refrigerant]LA) divided by At
was modeled as the volumetric transport rate from crawlspace to living
area times the refrigerant concentration in the crawlspace minus the
volumetric transport rate from the living area to all other zones times
the refrigerant concentration in the living area. A similar model (see
equations below) was used for the changes in the refrigerant
concentration in the crawlspace. The equations used in these
computations were:
(1)
where :
VLA is the volume of the living area,
At is the time interval for data acquisition,
[Refr]LA is the concentration of Refrigerant 12 in the living area,
TRCS->LA is the transport rate from the crawlspace to the living
area,
[Refr]cs is the concentration of Refrigerant 12 in the crawlspace,
TRLA->other is tne transport rate from the living area to all other
areas,
Vcs is the volume of the crawlspace,
TRLA->CS is tne transport rate from the living area to the
crawlspace, and
TRcs->other is tne transport rate from the crawlspace to all other
areas
-------
Initial estimates of the transport rates were obtained by linear
regression of the computed change in concentration onto the
concentrations observed in the crawlspace and in the living area.
The final estimates of transport rates were determined by trial and
error with graphical comparison of modelled and observed
concentrations.
RESULTS BEFORE FIRE
Several potential methods for radon mitigation were evaluated
during the study. Table I summarizes the baseline conditions and
the sequence of mitigation efforts that were undertaken at Hll
during late 1987 and 1988. For each period, the average radon
concentration recorded at the monitoring sites in the crawlspace
and living area are indicated. Figures 1 and 2 indicate the daily
average radon concentrations recorded in the crawlspace and living
area during the course of the study. Interpretation of these radon
results is complicated by the large seasonal fluctuations in the
concentration of radon in the soil gas (3).
A major consideration in the design of a mitigation system is
the leakiness of the structural shell. Using a blower door to
depressurize the house, it was determined that neither the living
area nor the crawlspace was tight enough to allow pressurization
without unacceptable energy penalties (see Table 2).
Measurements showed that the major route for radon entry into
this house is through the crawlspace. Two different measurement
techniques were used to evaluate processes causing gas transport
from crawlspace to living area. Tables 3 and 4 summarize results
from passive, PFT-based measurements and active, refrigerant-based
measurements, respectively. Figures 3 and 4 illustrate the data and
modelling results for the refrigerant-based measurements in March
1988. Between the Jan/Feb (1988) and Feb/Mar (1988) measurement
periods, the HAG ductwork was sealed, resulting in a 35% reduction
in the crawlspace-to-living-area transport rate measured by the
PFT. An even bigger improvement (~50%) was seen in the refrigerant
measurements when the HAC fan was operating continuously, as
opposed to intermittently, which was the case for the PFT
measurements.
Increased air exchange between the outdoors and the crawlspace
was investigated as a potentially valuable passive method of
mitigation for a crawlspace house. The crawlspace air exchange rate
was evaluated under three different conditions affecting apparent
ventilation of the crawlspace: (1) with all crawlspace vents closed
and taped, (2) with all vents closed but not taped, and (3) with 10
vents fully opened (see Table 5) . The air exchange rate was
moderately increased, relative to closed and taped, when no tape
was used (about 60% increase). When the vents were fully opened
there was a seven-fold increase (about 600%) in the rate of air
exchange.
-------
The final mitigation system selected, after consideration of
the diagnostic data and of the radon changes summarized in Table 1,
consisted of: (1) a plastic membrane attached to the crawlspace
walls, and (2) a submembrane depressurization system. The
depressurization system consisted of a standard mitigation fan
connected to a set of polyvinyl chloride (PVC) pipes to distribute
the suction. The system was designed to run continuously. Table 6
shows the air exchange rates observed in the crawlspace depending
on the status of the crawlspace vents and the mitigation blower.
The increase in infiltration, as measured by the air exchange rate,
is about equal to the flow through the mitigation system.
POST-FIRE RESULTS
HOUSE FIRE AND RADON RESISTIVE REBUILD AT Hll
On November 11, 1990, between 11:00 and 11:30 am, a fire
erupted in the master bedroom. The fire, started by a faulty space
heater, spread quickly throughout the house, consuming
approximately 35% of the living area. The remainder of the
superstructure suffered considerable heat and water damage. The
crawlspace, however, was virtually undamaged. After a review of the
damage, the insurance company declared the house a 75% loss and
contracted with a local builder to rebuild the superstructure.
The primary premise of the second phase of the project was to
advise the contractor on radon resistive construction techniques
for the reconstruction. These building modifications and techniques
were to have a minimal cost impact, be effective, and be readily
adaptable by most home builders.
Based on the previous house diagnostic data, several major
design problems were noted that in theory would facilitate radon
transport into the living area. The most noteworthy problem was the
location of the mechanical blower of the HAC system in the
crawlspace. All attempts to seal the mechanical unit and the
return pan joist were not very successful. PFT measurements
indicated rapid transport of crawlspace air into the living area
while the system was in operation. Also observed were numerous
floor penetrations for water, electrical, and sanitary service.
Utilizing the pre-fire diagnostic data, a simple approach was
devised for the reconstruction: 1) minimize the number of
interfloor penetrations and 2) relocate the mechanical system from
the crawlspace to the living area. Installation of a Mylar barrier
on the floor was considered; however, it would have resulted in a
dramatic cost increase in the reconstruction and probably would
have not be easily adapted by most home builders.
After briefing the contractor on the basics of radon entry,
the following radon resistive guidelines for the reconstruction
were jointly agreed upon:
-------
Reroute all HAC duct work, the house water supply, and
electrical wiring through the attic.
Seal penetrations for sanitary drains with expansion foam
or caulk.
Replace old flooring or cover it with new building cpde
plywood (3/4 in.)-
Seal all cracks and gaps in the new floor with caulk or
expansion foam.
Relocate the mechanical system from the crawlspace to the
living area or attic.
Due to cost and design considerations, the following contractor
modifications were made:
In areas where the original flooring was intact (< 30%),
the old floor was left in place. All holes were sealed
with caulk.
The HAC return could not be located in the attic since
changing the filters would require the use of a ladder.
Instead, the furnace was installed in the hall closet and
a sidewall return was located at floor level.
The other task that the contractor was requested to perform
was to keep track of any additional expenses for added material or
labor cost associated with the suggested modifications. Although he
failed to keep accurate records, he estimated the total
modifications resulted in an increase of less than $100.
OWNER MODIFICATION OF THE HOUSE
Shortly after reconstruction began, the owner of the house
decided to enclose the slab-on-grade carport and convert it into an
attached, single-car garage. Radon grab sample measurements
performed inside the garage were greater than 20 pCi/L. Left
unmitigated, the garage radon would complicate the planned study
since the exact source of the living area radon would be unknown.
A temporary, -but effective pressurization mitigation system was
installed by inserting a 20 in. window fan, in the garage window.
The system gave an impressive 0.02 in. WC differential pressure
(relative to the outdoors) . Figure 5 illustrates the effectiveness
of the temporary mitigation system. For the duration of the study,
the window fan was left on.
1 in. = 2.54 cm
1 in. WC = 249 Pa
-------
PRE-DIAGNOSTIC REPAIRS TO THE HOUSE
The pre-diagnostic inspection of the reconstructed house
revealed six minor and two major floor penetrations which had n^
been sealed by the contractor. The minor penetrations (< 12 in )
were mostly due to drain pipes or old HAC vent holes 2an.d were
easily sealed with expansion foam. Two major holes ("6 ft ) , each
located under the bathtubs, required building a sealed containment
box between the floor studs.
Another sizable hole (6 in.2) was located in the return duct.
Although the return duct was located upstairs, the contractor
installed a condensation drain pipe for the mechanical system
through the floor inside the return. By sealing the hole, the
pressure difference inside the return increased by over 100%.
Although the fire did not affect the Submembrane
Depressurization System (SMD) in the crawlspace, reconstruction of
the house did. Approximately 25% of the barrier, all near the
entrance of the crawlspace, was damaged during the reconstruction
and required replacement. The damaged area was removed and replaced
with new liner allowing a 3 ft overlap. As in the previous
installation, double strips of 1 x 1/8 in. double-sided foam tape
were used to secure the liner. The remaining 75% of the barrier,
although intact, was pressed very tightly against the soil from
extensive abuse during reconstruction. Despite the abuse,
differential pressure measurements performed in several locations
beneath the barrier indicated sufficient pressure extension for
mitigation.
DIAGNOSTIC TESTING OF Hll
For comparison purposes, similar diagnostics were performed on
the rebuilt house. These tests were:
Continuous radon measurements (Table 7 and Figure 5)
Solution pipe grab samples (Table 8)
Interzonal transport of tracer gas (Table 9 and Figures
6 and 7)
Air exchange (Table 10)
Shell blower door diagnostics (Table 11)
1 in.2 = 0.0006 m2
1 ft2 = 0.09 m2
***
1 ft = 0.3 m
-------
RADON CONCENTRATION AS A FUNCTION OF HAC CYCLE
Due to cost and time restraints, continuous radon data were
collected for only 4 days. The data set, although small, indicates
several important points: 1) the passive design did not completely
prevent radon from entering the living area and 2) operation of the
relocated HAC blower failed to increase the radon levels in the
living area (Figure 5 and Table 7).
EFFECT OF RADON IN SOLUTION CAVITIES
In the backyard of Hll, there are small (~4-in. diameter)
holes, called solution cavities, that appear to communicate with
subterranean voids. The increase in the radon concentration within
the house during the week (Table 7) correlates well with the
increasing radon concentration within the solution cavities
(Table 8) . Between 1987 and 1991, we have observed a strong
correlation between the radon concentration in the backyard
solution cavities and Hll (3). If the radon concentration increases
within the cavities, the levels within the house increase as well.
Similar several order-of-magnitude, rapid, increases of radon
concentration in the solution cavities have been recorded in the
past, usually with coincident increases of radon concentration in
the house. Obviously, the fire or reconstruction had no effect on
the activity of the solution cavities.
EPISODIC TRACER STUDIES
To evaluate the impact of floor penetration minimization,
interzonal refrigerant transport experiments were performed. Based
on the data (Table 9 and Figures 6 and 7) , HAC-indep^nden't
transport was reduced more than two-fold (44 versus 18 ft /min)
compared to pre-fire results. The HAC system still appears to
increase the amount of air transported from the crawlspace.
However, the volume of HAC-dependent transport was reduced more
than five-fold (138 versus 25 ft /min) compared to pre-fire
results.
SHELL DIAGNOSTICS
Comparing shell tightness between the old and new house is
important for interpreting the radon results. If, for example, the
rebuilt house is exceptionally tighter, then in theory, higher
living area radon levels would be expected for the same source
potential. To answer this question, episodic air exchange
measurements (Table 10) and blower door measurements (Table 11)
were performed on the house. These measurements indicate that the
shells of both the crawlspace and the living area have about the
same degree of airtightness after reconstruction as before. Air
exchange rates and equivalent leakage areas are not significantly
different.
ft3/min =1.7 m3/hr
-------
CONCLUSIONS AND RECOMMENDATIONS
The rebuilt house achieved a 60% reduction in the volumetric
transport from crawlspace to living area when the heating and
cooling system was not operating compared to pre-fire results.
There was an even greater reduction (>80%) in the
crawlspace-to-living-area transport that is attributable to
operation of the HAC blower unit. These results serve to illustrate
the potential for radon resistant construction techniques to reduce
radon entry rates in houses with crawlspaces. In principle, a 60%
reduction in crawlspace-to-living-area transport should lead to
better air quality in the living area due to reduced levels of: (1)
radon and other pollutants from soil gas, (2) spores and other
biogenic species that arise from molds and mildews that grow in the
crawlspace, and (3) energy expenditure due to less unconditioned
air flowing into the living area. The added cost for installing the
measures that led to these benefits was estimated to be about $100.
The HAC relocation and the reduction in floor penetration did
not completely prevent radon entry into the living area. Radon
levels measured in the rebuilt house are not significantly lower
than those in the original house. It is possible that enclosing the
carport to make a garage may have led to new sources of
depressurization relative to the soil gas beneath the slab or to
new radon entry pathways. Because of the karst terrain in this
area, there may be subterranean conduits underneath the slab
capable of delivering large amounts of radon to the house. In the
backyard of this house we have observed a 3-in. diameter hole to
release radon at a rate of more than 200 pCi/s for 1 week. The rate
of release from this hole ranged from 0 to 900 pCi/s during a 3
month period of observation. Another possible explanation is the
failure of the materials used in construction to adequately prevent
radon permeation. It is recommended that EPA perform radon
permeability measurements on plywood to determine if it has
sufficient radon resistance to act as a passive barrier alone.
ACKNOWLEDGEMENTS
The before-fire results were obtained in a study supported by
the Tennessee Valley Authority (J. P. Harper and R. B. Maxwell,
project officers), the U.S. Department of Energy (S. L. Rose,
project officer), and the U.S. Environmental Protection Agency (R.
B. Mosley, project officer). R. J. Saultz assisted us with the
before-fire studies and J. J. Walls assisted us with the post-fire
studies.
-------
REFERENCES
1. Wrenn, M. E., Spitz, H., and Cohen, N. Design of a continuous
digital-output environmental radon monitor. IEEE Transactions
in Nuclear Science. 22:645, 1975.
2. Perdue, P. T. , Dickson, H. W., and Haywood, F. F. Radon
monitoring instrumentation. Health Physics. 39:85, 1984.
3. Wilson, D. L., Gammage, R. JB. , Dudney, C. S., and Saultz, R.
J. Summertime Elevation of Rn Levels in Huntsville, Alabama.
Health Physics. 60:189, 1990.
-------
TABLE 1. CHRONOLOGICAL HISTORY OF MITIGATION AT HI 1 BEFORE THE FIRE.
Julian
rvif-po
1987 &
1988
263-293
295-333
338-012
20-40
44-76
86-97
100-132
132-145
145-153
166-177
180-202
209-215
215-222
222-228
229-236
246-276
287-307
310-342
Radon (pCi/L)
Crawl-
space
9.8
3.1
4.4
4.2
__
4.6
7.9
3.0
11.2
9.5
1.6
2.8
4.6
6.4
1.0
0.5
0.5
Living
Area
_-
5.0
2.4
3.6
2.2
1.6
1.0
0.9
1.5
6.9
5.8
1.9
2.0
5.8
7.7
0.8
0.6
0.5
Status of Mitigation Systems
HAC
Sealed
No
No
No
No
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Mixing
None
None
None
None
None
None
None
Fans
Fans
None
None
None
None
None
None
None
None
None
CS Vents
Original
Original
Original
Original
Original
Original
New
New
Exhaust
New
New
New
Closed
Closed
New
New
New
New
Barrier
Original
Original
Original
Original
Original
Original
Original
Original
Original
Original
Original
Sealed
Sealed
Sealed
Sealed
Sealed
Sealed
Sealed
Barrier
Flow, %
100
100
75
0
100
100
100
Comments
Upgraded energy
conservation
HAC Sealed
About 1 ,000 f t3/min
-------
TABLE 2. BLOWER DOOR DEPRESSURIZATION: ORIGINAL HOUSE.
Location
Living Area
Crawlspace with All Vents Closed
Blower Door Depressurization Results
4-Pa Equivalent
Leakage Area (ft )
7.33
6.89
Air Exchange at 50-Pa
(0.02 in. WC)
(fa'1)
69
150
TABLE 3. PFT TRANSPORT MEASUREMENTS: ORIGINAL HOUSE.
Type and Direction of Flow
Infiltration
Exfiltration
Interzonal
Outdoors to Crawlspace
Outdoors to Living Area
Total
Air Exchange Rate (h"1)
Crawlspace to Outdoors
Living Area to Outdoors
Crawlspace to Living Area
Living Area to Crawlspace
Average Volumetric Air Flows (ft3/min)
Time Period
Jan/Feb
(1988)
138
5
143
0.48
82
61
66
9
Feb/Mar
(1988)
147
18
164
0.56
115
50
40
8
Sep
(1988)
312
90
402
1.36
309
130
44
42
TABLE 4. REFRIGERANT TRANSPORT MEASUREMENTS: ORIGINAL HOUSE.
HAC
Status
Off
On
Crawlspace to Living Area Transport Rates (ft3/min)
Original
Oct (1987)
14
361 ± 42
Post-Weatherization
Mar (1988)
44
182
Post-Mitigation
Oct (1988)
51± 5
194 ± 4
TABLE 5. CRAWLSPACE AIR EXCHANGE MEASUREMENTS: ORIGINAL HOUSE.
Impact of Crawlspace Vents on Infiltration with HAC Turned Off
Vent Status
Crawlspace Air Exchange (h"1)
Closed & Taped
0.49
Closed
0.82
Open (N = 10)
3.35
-------
TABLE 6. EFFECT OF MITIGATION SYSTEM ON AIR EXCHANGE IN ORIGINAL HOUSE,
Status of Crawlspace
Vents
Open
Closed and Taped
Air Exchange (If1) with HAC Off
Mitigation Blower Status
Off
1.62
0.29
On
2.00
0.87
Theoretically expected change based on HAC blower unit
causing an air flow equalling 51 ft3/min (30 m3/hr)
Change
0.38
0.58
0.54
TABLE 7. RADON CONCENTRATION AS A FUNCTION OF HAC CYCLE: REBUILT HOUSE.
Julian Start
Time
176.4
177.4
178.4
178.9
Julian Stop
Time
177.4
178.4
178.9
179.4
Time in
Hours
25
24
12
12
HAC Cycle
Off
On
Off
On
Living Area
Average
pCi/L
5.8
5.7
6.4
9.9
Crawl-
space
Average
PCi/L
6.9
5.9
13.1
16.8
Garage
Average
PCi/L
16.4
2.4'
3.2
4.1
Garage mitigation system activated on Julian Day 177.
TABLE 8. SOLUTION CAVITY GRAB SAMPLE DATA (pCi/L).
Hole Number
1
2
3
4
Julian 176
17
194
912
24
Julian 177
208
687
2,258
582
Julian 178
3,049
5,047
5,108
4,235
Julian 179
4,270
2,558
4,996
5,606
-------
TABLE 9. EPISODIC CRAWLSPACE-TO-LIVING-AREA TRANSPORT RATES (ft3/min) FROM
INTERZONAL TRACER EXPERIMENTS: REBUILT HOUSE.
HAC Status
Off
On
Post-Weatherization
March, 1988
44
182
Post-Mitigation
October, 1988
5U5
194.+4
Post-Fire
June, 1990
18
42
TABLE 10. EPISODIC AIR EXCHANGE (If1) COMPARISON.
HAC Status
Living Area
Crawlspace
Before Fire
Off
0.22
0.82
After Fire
Off
0.26
0.84
Before Fire
Continuously On
0.37
0.86
After Fire
Continuously On
0.50
0.71
TABLE 11. BLOWER DOOR DEPRESSURIZATION COMPARISON.
Location
Living Area
Crawlspace with All Vents
Closed
4-Pa Equivalent
Leakage Area (ft2)
Before
Fire
7.33
6.89
After
Fire
4.44
7.89
Air Exchange at 50-Pa
(0.20 in. WC)
(h-1)
Before
Fire
69
150
After
Fire
71
110
-------
Crawlspace
40
O 30
c
o
I
0)
en
f
074 147
Julian Date (1988)
221
^
001 074
Julian Date (1989)
147
Figure 1. Radon in crawlspace prior to fire.
-------
10
O
o.
I6
oc
£5 4
I
Q 2
10
O
CL
I8
DC
o>
I
Q 2
Living Area
001 074 147
Julian Date (1988)
220
001 074
Julian Date (1989)
147
Figure 2, Radon in living area prior to fire,
-------
-^ £000
I
03
£ 1,500
n
,
HAC Operating Continuously
Data from Crawlspace n
Data from Living Area o
Modelling Coefficients (ft^S/min)
Transport from living area to crawlspace: 50
Transport from crawlspace to elsewhere: 194
Transport from crawlspace to living area: 182
Transport from living area to elsewhere: 141
1 2 3
Hours Since Start of Experiment
Figure 3. Interzonal transport while HAC is operating before reconstruction.
asoo
2.000
1.500 -
1.000 -
500 -
HAC Not Operating
Data from Crawlspace
Data from Living Area
o
Modelling Coefficients (ft~3/min)
Transport from living area to crawlspace:
Transport from crawlspace to elsewhere:
Transport from crawlspace to living area:
Transport from living area to elsewhere:
11
94
44
44
1 2 3
Hours Since Start of Experiment
Figure 4. Interzonal transport while HAC is off before reconstruction.
-------
b
MAC On MAC Off
176.2 176.6 177 177.4 177.8 178.2 178.6 179 179.4
Julian Day
Upstairs + Crawlspace o Garage
Figure 5. Radon at three sites in rebuilt house,
-------
HAC Not Operating
Data from Crawlspace
Data from Living Area
Modelling Coefficients (ft
Transport from living area to crawlspace:
Transport from crawlspac© to elsewhere:
Transport from crawlspace to living area:
Transport from living area to elsewhere:
Figure
asoo
2 3
Hours Since Start of Experiment
6. Interzonal transport while HAC is off after reconstruction.
I
§
IB
I
c
O
8
2,000 -
1,500 -
1.000 -
£0 500
02
O)
HAC Operating Continuously
Data from Crawlspace
Data from Living Area
Modelling Coefficients (ft 3/min)
Transport from living area to crawlspace:
Transport from crawlspace to elsewhere: 60
Transport from crawlspace to living area: 42
Transport from living area to elsewhere:
Hours Since Start of Experiment
Figure 7. Interzonal transport while HAC is operating after reconstruction.
-------
PERFORMANCE OF SLABS AS BARRIERS TO RADON
IN 13 NEW FLORIDA HOMES
by: James L. Tyson Charles R. Withers
Florida Solar Energy Center Florida Solar Energy Center
Cape Canaveral, FL Cape Canaveral, FL
ABSTRACT
The Florida Solar Energy Center has been involved with the Florida Radon Research
Project in the past year demonstrating radon resistant construction techniques.
FSEC has installed mitigation systems in 13 new Florida homes.
Testing has been conducted in three major areas:
Soil - Native and fill soils have been tested for radon and radium content, and
permeability has been measured. Sub-slab radon is measured for the
finished house. Findings will be presented and compared to indoor
radon levels.
Slab - Cracks on each slab have been mapped out and measured for size, air
flow, and radon levels. Pressure field extensions for the activated
mitigation systems are measured for each house. Selected crack and
pressure field extension maps will be presented illustrating general and
unique cases.
House - House testing includes blower door, infiltration, duct leak, and radon
stress testing. General results of each test will be presented along with
illustrations of special cases.
-------
XI-3
HVAC CONTROL OF RADON IN A NEWLY CONSTRUCTED
RESIDENCE WITH EXHAUST ONLY VENTILATION
By: Mike Clarkin and Terry Brennan
Camroden Associates, Inc.
RD #1, Box 222
Oriskany, NY 13424
Timothy Dyess
U. S. Environmental Protection Agency
MD54, AEERL
Research Triangle Park, NC 27711
ABSTRACT
A test house in Pennsylvania has been modified so that the
air handler pressurizes the basement and prevents radon entry.
This system has shown effectiveness at keeping radon levels low.
A central exhaust system that meets minimum ASHRAE 62-89
recommendations has been installed. Radon levels were monitored
to assess the impact of a continuously operating exhaust fan on
the effectiveness of the radon control technique. Power
consumption of the HVAC air handler, changes in infiltration with
the air handler on and off, and the moisture content of the
framing in several locations in the building were monitored.
-------
A SIMPLIFIED ANALYSIS OF PASSIVE STACK FLOW RATE
Pah I. Chen
Mechanical Engineering Department
Portland State University
Portland, OR 97207
ABSTRACT
Passive stack is a potential indoor radon mitigation technique
in which sub-slab ventilation is driven by free convection caused
by the heating of the air in the stack without external mechanical
power input.
An equation governing steady state free convection induced by
the heating of a vertical stack wall will be expressed first. For
field application, a simplified computation can achieve the flow
rate through a stack of known diameter and height, as subject to a
given house temperature between inlet and outlet of the stack.
This equation provides significant insight for one to pursue
ways to enhance the passive stack flow rate. Enhancement ideas
including the improvement of sub-slab pressure field extension,
reduction of pressure differential, use of stack material of higher
heat conductivity, and addition of a low wattage heating coil for
additional needed flow are discussed.
-------
INTRODUCTION
The passive stack technique has been considered a potential
method for use in indoor radon mitigation. While the effectiveness
of this technique is still being evaluated, its potential benefits
are numerous. The aspect of energy conservation is high on the
list, followed by the lower operating cost and simplicity in
construction and maintenance of the system.
A field study on passive stacks in Maryland was reported by
Saum and Osborne (1). Their measurement provided interesting
evidence that the radon levels in the selected houses drop
substantially when the passive stack is open versus the condition
when the stack is closed. Other than their paper, there seems to
be a lack of other reports on the same subject matter. And, up to
this point, analytical work related to passive stacks is not known
to be available.
This paper presents a simplified computation of flow rate
through a passive stack of known diameter subject to house
temperature and pressure differentials between inlet and outlet of
the stack. The governing equation provides much insight for
improving the flow rate and thus the potential of the passive stack
technique.
GOVERNING EQUATION
The passive stack is normally a circular duct. Three typical
laminar velocity profiles inside the stack are shown in Figure 1.
The lower profile is near the base of the stack where cold soil gas
enters into the warm stack. A slight convection effect is visible
as the air near the stack wall is heated and moves a little faster
than the main stream. The middle profile is for a section where
the stack is well-heated inside the house. It is shown that the
air velocity near the wall picks up as the entire stream moves
faster upward. The top profile is for flow above the roof where
the stack is cold in winter. The profile shows a down draft of air
current near the stack wall while the warmer main stream is still
heading upward to the stack exit.
The flow inside the stack is driven mainly by pressure and/or
temperature difference while resisted by viscosity along a fluid
column. The governing equation of a laminar, steady-state, one-
dimensional free convective flow inside the stack is
-------
dz
-------
Since the fluid located away from the wall is in hydrostatic
equilibrium,
._p/,
where Pf is the density of the fluid located away from the stack
wall. Substituting this into equation (2), we get
For an ideal gas, this equation becomes,
- *< - i>
Assuming the temperature difference (T - T f) is constant and
integrating equation (6) , we get,
" Tf)z (7)
w =
(8)
where p is the coefficient of volume expansion of air or l/Tf.
This equation basically states that the velocity in the stack
is a function of fluid buoyancy and the height of the stack. The
velocity can be enhanced simply by promoting greater heat transfer
to the stack.
This equation implies that the difference between the average
air temperature and the temperature of the main stream, T - Tf, is
dominating the stack velocity. Two temperature profiles are shown
in Figure 2. Both temperature inside the house and the air inside
the stack are increasing vertically along the stack as shown in
Figure 2(a). Above the roof, air temperature drops towards ambient
temperature at the exit. In Figure 2(b), the difference between
-------
the average air temperature and the temperature of the main stream
is plotted along the stack height. For the portion inside the
house, T - Tf is nearly a constant. For the portion above the
roof, the temperature difference drops rapidly. There, average
upward movement of the air slows down. The stack may be totally
"plugged" when T - Tf = 0, where air ceases to move upward. This
is shown as the dotted line intersecting a hypothetical stack
height. This phenomenon is caused by the drop of air temperature
in the stack due to the cold intruding ambient air.
Roofline
Stack
height
Slab
\
Average temperature
L of air inside the stack
Roofline
Temperature
inside the house
Slab
I I
! I
|\ Plugged effect
AT = T - Tf
(a) The temperature profile of air in stack
(b) Temperature difference AT = T-Tf
Figure 2. Temperature Profiles
The velocity obtained from equation (8) is the maximum
velocity achievable in the passive stack. From this velocity, one
can evaluate the maximum suction pressure achievable at the
entrance of the passive stack. The following is an example
illustrating how this equation can be used in the passive stack
flow rate calculation.
-------
Example
Air at 40 F enters a 3" diameter vertical stack of 30 ft long
beneath the slab. The air is heated from a constant stack wall
temperature. The heated air rises in the stack and exits at the
top of the stack. This creates a vacuum effect at the stack
entrance which draws air in from beneath the slab.
If the temperature of the exit air is 50 F, what is the
volumetric flow rate of air in the stack?
Since the air velocity is basically induced by buoyancy, the
flow is assumed laminar. The temperature difference between the
wall and the mainstream is assumed to be 6 F through the stack, so
P = (9)
H 460*46 l '
From equation (8), we get,
w2 = 2 x 32.2 x x 30
460+46
w = 4.79 ft/S
The volumetric flow rate, V, in a 3" pipe, is
n(&*
V = Aw = i^ x 4.79 = 0.235 ft3/S = 14.11 JftVmin
This flow rate does not seem to be much, but for a sub-slab
with good communication it may be sufficient to reduce the indoor
radon level, especially if the radon concentration level below the
slab is low (e.g., less than 10 pCi/L) to begin with.
The reduction of radon level, however, does not come without
an energy penalty. The penalty is due to heat loss through the
passive stack. But if the passive stack is proven useful, there is
an energy savings when compared to an operational fan. The heat
loss can be computed using heat convection equation.
PLOW RATE ENHANCEMENTS
US EPA has taken a cautious approach (3) in recommending
passive soil ventilation as an indoor radon reduction technique.
-------
The concern seems to be whether the soil depressurization can be
sufficiently sustained by employing the passive technique without
an active fan. But an EPA publication suggests that "if a fairly
substantial piping network is already in place, the ventilation
system that is installed connecting to these pipes might initially
be designed and operated in a passive mode to determine if passive
operation is sufficient"(4). This is a trial-and-error approach in
determining the effectiveness of the passive technique. If the
system is proven insufficient, it should be modified to operate in
an active mode. This approach seems to work better for new
constructions as laying a pipe network beneath an existing slab is
not likely to be cost effective.
For both new and existing houses, the following methods may be
considered to enhance the passive stack flow rate and thus sub-slab
depressurization. Some of the methods have been mentioned in radon
literature.
(1) Sub-slab Communication
As mentioned, buoyant force can only generate a slight amount
of suction at the slab. So, anything one does to improve the
sub-slab communication would improve the passive stack
performance. An EPA publication (4) suggests the following
sub-slab preparations to improve the sub-slab communication:
(a) use crushed aggregate with minimum of 80% of the aggregate
at least 3/4" in diameter (or DOT #2 gravel) 4" minimum depth,
and (b) lay perforated PVC pipe in the gravel before the slab
is poured and connect the pipe to the exhaust of the system.
Good sub-slab communication will provide uninterrupted soil
gas ventilation and without much effort. The passive system
will benefit more with good sub-slab communication than an
active system.
(2) Pressure Differentials
Although the pressure term is not directly shown in equation
(8), it is manifested in temperature, Tf, based on the ideal
gas relation. Depressurization implies that Tf is lower and
requires more heat input to raise the soil gas temperature as
it flows through the stack.
There is not much one can do to minimize the influence of
atmospheric movement. But the reduction of sub-slab
depressurization is needed and can be achieved, for example,
-------
by sealing cracks on slab and floor joints. Also, reduction
of stack pressure loss can be achieved by avoiding elbows and
tees in the piping system.
Using flue gas from the furnace to supplement the lift of air
in the stack has been referred to in radon literature.
However, the operation of a furnace is intermittent and it is
for colder period of the year. It cannot be relied upon for
continuous radon mitigation.
For existing houses installed with a pressurization system in
conjuction with a passive system, one can take advantage of
the higher pressure at the stack inlet to force the diluted
radon to ventilate through the passive stack.
(3) Stack Materials
Material selection for passive stack implementation is a
dilemma between heat transfer characteristic and energy
conservation. One can improve the heat transfer from the
house to the air in the stack by using aluminum duct because
its heat conductivity is hundred times higher than that of PVC
pipe. Un-heated portions should stay with plastic pipe. Of
course, one should question whether this is energy efficient
as the heat loss through the aluminum duct is far greater than
that through the plastic pipe
(4) Heated Stack
Portion of the stack if heated either by an electric heating
coil or by electricity generated from photovoltaic cells would
increase the air temperature and thus air buoyancy. This
applies to houses in mild climate as well as to houses with
air-conditioning.
For existing systems, using an inserted section of heated coil
inside the PVC pipe would be more economical than replacing
the entire pipe by a metallic pipe. For new systems, use
aluminum pipe in the heated area, PVC pipe in unheated area
and in areas above the roof. Of course, one should compare
the initial cost as well as the operating cost of both passive
and active systems before making a final decision.
-------
CONCLUSIONS
Passive stack technique utilizes the fluid buoyancy induced
flow from warmer indoor temperature to move the air inside the
stack. The upward air movement causes the sub-slab to ventilate.
The magnitude of sub-slab depressurization created, however, is
small compared to that caused by an active fan. Because of this,
intuitively, only sub-slab of small area with good sub-slab
communication can benefit by using this technique. And, passive
stack technique can be more beneficial for new constructions (4).
In this case, one can pre-lay a radial pattern of perforated pipes
beneath the slab to improve the flow rate through the stack. For
both new and existing residential houses, using one (heated stack
in plastic duct seems to be the best) or a combination of
enhancement methods suggested would increase the stack flow rate.
Other suggestion to include a mini fan also available in literature
(5). But, mini fan installation changes the passive nature of the
technique. In any case, a rain cap on top of the stack is
definitely needed to prevent the moisture from affecting the heat
transfer capability of the stack wall. Sealing of cracks on slab
and floor joints, and the prevention of house depressurization due
to mechanical appliances is imperative.
The simplified equation as presented is intended to provide a
quick estimation of the flow rate that can be generated by the
passive stack technique. At this time, experimental or field data
on passive stack flow rate are not known to be publicly available.
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 author would like to thank Kevin M. Chen for his
assistance in the final preparation of the paper, and Chuck
Eastwood of Bonneville Power Administration for his valuable
comments.
REFERENCES
1. Saum, D.W., and Osborne, M.C. Radon Mitigation Performances
of Passive Stacks in Residential New Construction", 1991
International Symposium on Radon and Radon Reduction
-------
Technology, Atlanta, February 19-23 1990, vol. V. Preprints,
14 pages
2. Chapman, A.J. Heat Transfer, 2 nd ed. Macmillan Co., 1967
3. Mosley, R.B., and Henschel, D.B. Application of Radon
Reduction Methods. EPA/625/5-88/024 August 1988
4. Clarkin, M., and Brennan, T. Radon-resistant Construction
Techniques for New Residential Construction. EPA/625/2-
91/032 February 1991
5. Saum, D.W., Mini Fan for SSD Radon Mitigation in New
Construction. 1991 International Symposium on Radon and
Radon Reduction Technology, Philadelphia, April 2-5 1991,
vol 4, 11 pages
6. Grisham, C.M., Radon Prevention in Residential New
Construction: Passive Designs That Work. 1991 International
Symposium on Radon and Radon Reduction Technology,
Philadelphia, April 2-5 1991, vol 4, 13 pages
-------
FACTORS THAT INFLUENCE PRESSURE FIELD EXTENSION IN NEW RESIDENTIAL
CONSTRUCTION: EXPERIMENTAL RESULTS
by: Prill, R., Washington State Energy Office - Energy Extension Service
Fisk, W., Indoor Environment Program, Lawrence Berkeley Laboratory
Gadgil, A., Indoor Environment Program, Lawrence Berkeley Laboratory
ABSTRACT
Field experiments were conducted in houses to determine the influences of soil and subslab
aggregate permeability, slab sealing, a subslab pit, and a sub-aggregate membrane on the subslab
pressures resulting from radon mitigation by subslab ventilation (SSV). Eight new, single-family
detached houses with concrete basements were studied. The combination of soil and subslab
aggregate permeability varied between houses as a consequence of the large range in soil
permeability and the use of three different types of aggregate. In two of the houses with high
permeability soil, a plastic membrane was placed between the soil and the subslab aggregate to
evaluate potential enhancement of the subslab pressure field due to the membrane. A test
apparatus was used to produce typical SSV operating conditions, and to measure the associated
air flow rates and pressures. Multiple small test holes were drilled through the slab to permit
measurement of the magnitude of the pressure differences between the basement and the subslab
aggregate.
Large improvements in subslab pressure field extension were demonstrated by creating a 10-
inch radius pit immediately below the SSV pipe penetrations through the slabs. Pressure field
extension was further enhanced substantially when visible cracks and joints in the slabs were
sealed to the extent practical.
These experiments have increased our understanding of the relationships between SSV
system performance and soil and aggregate permeabilities, SSV system characteristics, and the
sealing of slabs. Optimization of SSV systems in new residential construction is essential to the
development of low pressure radon control that is efficient, effective, and reliable.
BACKGROUND
The Environmental Protection Agency (EPA) has determined that indoor radon is one of the
most significant indoor air pollutants in terms of health risk to the population of the U.S. In some
areas of the U.S. building codes have responded by requiring new structures to be built with
features intended to allow effective and efficient radon control to be achieved, should it be
necessary. The most common features prescribed in the new codes relate to the optimization of
subslab ventilation (SSV) techniques that reduce radon entry.
Subslab ventilation systems consist of a pipe (or multiple pipes) installed through the
concrete slab floor. An in-line fan is typically attached to the pipe(s) to create a pressure within
the pipe and thus produce a pressure zone beneath the slab. SSV systems installed and operated
to depressurize the soil and materials beneath slab floors (subslab depressurization or SSD) are
-------
currently the most common radon mitigation technique used to reduce radon entry into
residential buildings in the U.S.
Subslab pressurization (SSP) systems, are configured and operated to force outside air
through the pipe and beneath the slab. The resulting positive pressure field in the subslab region
may still allow some radon entry, however the concentration of radon in the entering soil gas will
usually be reduced. Field studies have shown that SSP systems are sometimes more effective
than SSD systems in reducing indoor radon concentrations in structures built in high permeable
soils and where soil radon concentrations are low to moderate (Turk 1991).
The spatial extent and magnitude of the negatively pressurized area established by SSD
systems beneath the slab is an important performance factor. The pressure difference that
normally draws radon-laden air into structures must be reversed so that air flows from the
structure into the soil. The extent of the depressurized (or pressurized) areas under buildings is
known as the pressure field extension (PFE). Extension of the pressure field to include all areas
of the slab normally assures significant and reliable reductions in radon entry rates. The term
normalized pressure field extension (PFE) is used for the ratio of sublsab pressure to pressure at
the location where the SS V pipepenetrates the slab.
Field studies and experience have shown that slab sealing, subslab aggregate, SSV pipe
location(s) and pressures, and the existence of a pit at the SSV pipe penetration through the slab,
are important parameters related to radon mitigation. Experience in existing structures has also
shown that when an uninterrupted (unbounded) layer of high permeable gravel or other media
exists between the soil and the slab, pressure field extension can often be easily established
beneath the entire slab. Additionally, the existence of a pit in the aggregate and soil at the SSV
pipe location through the slab generally enhances PFE (Pyle 1988, Craig 1991).
Openings through the slab such as cracks, joints, and holes can act to "short-circuit" the
pressure field, reducing the PFE. Thorough substructure sealing generally enhances PFE and
increases the overall energy efficiency of SSD systems by reducing the volume of conditioned
indoor air that is drawn out of structures by the SSD system, and also by allowing the SSD fan to
be down-sized. In homes that are thoroughly air-sealed above grade, but lack complete slab and
substructure sealing, powerful SSD system fans have the potential to backdraft atmospherically
vented combustion appliances, exposing occupants to hazardous combustion gases.
OBJECTIVES
The principal objective of this study was to determine the influence of subslab aggregate
type on the performance of subslab ventilation (SSV) systems, both SSD and SSP, in new single
family detached structures. Other variables that influence SSV, as indicated by PFE, were also
investigated.
A secondary objective was to develop a more complete numerical model to evaluate SSV
performance as a function of building substructure, soil characteristics, and selected SSV system
parameters (such as the magnitude of the depressurization at the suction point, the size and
location of cracks and other openings in the slab, and the permeability of the gravel layer). This
objective is addressed in a separate document (Gadgil 1991).
MOTIVATION
A primary motivation for conducting the study was to establish a greater understanding of
the influence of various grades of aggregate relative to PFE in order to guide the development of
-------
building codes and specifications. At the time of the study, many building code jurisdictions
within the state of Washington, and in other parts of the U.S. require a minimum grade of
aggregate beneath residential concrete floor slabs (WSBCC 1990, Nuess 1989). Justification for
requiring specific grades of aggregate was difficult, due to the lack of field studies and numerical
modeling comparing PFE to aggregate type. Findings relating aggregate type to PFE would
allow code jurisdictions, engineers, architects, and builders to incorporate the most appropriate
subslab aggregate type into new structures.
Another building code related variable investigated in this study was the influence of slab
sealing on PFE. A Washington State building code in effect at the time of the study required
sealing of only those slab openings that would be inaccessible to the occupants (i.e. behind
finished walls, beneath shower stalls, etc.).
SITE SELECTION
Eight new homes located in and near Spokane, Washington were studied. The building
departments in Spokane and Stevens counties actively cooperated in the study. They notified
researchers immediately before footing inspections of new single family structures were to be
conducted and supplied builder names and addresses. Researchers visited the sites and visually
assessed the relative homogeneity and permeability of the soil, depth of footings below grade,
and complexity of the structure geometry. When particular sites appeared suitable for inclusion
in this study, builders were contacted (builders were rarely present at time of footing inspection
visit), the research project was described and the builder's cooperation solicited. After the
builder agreed to participate in the research, the building sites were revisited and soil
permeability measured within the excavation and in the undisturbed soil immediately
surrounding the building site using methods described by Garbesi (1988).
Final selection of sites/houses was based on the homogeneity of the soil, soil permeability
(relatively high or low), specifications for a basement of 1 meter minimum depth below grade,
and simplicity of the substructure geometry, and a signed agreement with the builder.
STRUCTURE CHARACTERIZATION
All phases of the substructure construction were monitored and documented: interior and
exterior footings and obstructions, plumbing and utility trenches, backfill, aggregate type and
installed thickness, and slab thickness. Thorough characterization of the structures and soil was
necessary in order to provide the level of detail necessary for development of the numerical
model.
SOIL
Three of the study houses were built in high permeability soil and the remaining five houses
in low permeability soil. The permeability of the low permeability soils was below the detection
limits of the measurement equipment (approximately 10" ^ m2). The high permeability soils
ranged from 9.3 x 10'11 to 15 x 10"12 m2.
AGGREGATE
The homes were built using one of three common grades of aggregate beneath the concrete
slab floors. Each grade had a desired permeability range: relatively high, medium, and low
-------
permeability. Samples of aggregate at each house were collected during placement and shipped
to Lawrence Berkeley Laboratory (LBL) for laboratory measurement of air permeability (Gadgil
1991). The installed thickness of the aggregate throughout the slab area was documented.
The medium permeability aggregate corresponded to minimum grading specifications
(ASTM 67) allowed by the Spokane County Building Department (at the time of this study) for
use under residential concrete slab floors. The minimum grade aggregate, and larger, was
intended to allow effective and efficient radon mitigation by SSV, if mitigation was necessary.
This minimum grade material was identified by the supplier as 3/4" Round and known locally in
the construction industry as "radon rock". Three study houses were built using this material.
The low permeability aggregate (ASTM 8), was used in two study houses located outside
Spokane County, since the material did not meet Spokane County minimum grading
requirements. One local supplier identified the aggregate as 3/8" Exposed, while another supplier
identified it as #8 Pea Gravel.
The high permeability aggregate (ASTM 4) was identified by the supplier as 1-1/2" Round
or "Drain Field Rock" and was used in three of the houses.
PLASTIC MEMBRANE OVER SOIL
A 6 mil plastic membrane was placed between the gravel layer and high permeability soil in
two houses. Researchers installed the plastic membrane in a manner similar to that what might be
expected from a building contractor or subcontractor. The plastic was overlapped at least 0.6 m
at the seams but not sealed. The plastic sheets were placed loosely (not stretched) and the edges
trimmed such that they contacted the walls of the perimeter and interior footings. The plastic was
not sealed at the edges. When the gravel was placed over the membrane it was observed that the
edges of the membrane were pulled back from the footings due to irregularities in the soil
surface, in spite of the extra membrane material that had been allowed by loose placement. This
resulted in numerous gaps along the footings. The gap widths ranged from less than 1 cm to
more than 8 cm.
The combinations of soil, aggregate and membranes for each of the study houses are shown
in Table 1.
FIELD EXPERIMENTS
TEMPORARY SSV SYSTEM AND TEST HOLES
After the house was framed, but before the finish work was completed, a temporary SSV
system was installed in each structure in order to conduct PFE measurements. A 10cm (4in) dia.
hole was drilled through the slab in a central location to simulate a typical SSV penetration. An
SSV system was also installed near the perimeter of the slab in two of the houses. Smaller test
holes of 0.9 cm (3/8 in) diameter were drilled through the slab floor at various locations. Test
holes along the perimeter were located at least 30 cm (1 ft) from the walls to avoid footings. The
test holes were used to assess the PFE and magnitude of pressure in the aggregate layer relative
to the pressure applied at the SSV penetration (See Figure 1.). The number of test holes at each
house ranged from 21 to 35, while the area of the slabs tested ranged from 45.7 m2 to 190.3 m2
(492 ft2) to 2,049 ft2). The density of holes across the slabs ranged from one test hole per 1.8 m2
to one test hole per 9.1 m2 (19.4 ft2 to 98 ft2).
-------
TABLE 1.
MATRIX OF FIELD TESTS OF SS V PERFORMANCE
Total Slab
Area(m2)
No. of Slab
Bays*
Soil Perm.
Range (m2)
001
181.1
3
4.6-6.8xlO-n
002
45.7
2t
1.1-3.7X10-10
003
48.8
2
7.5-9.3X10-H
House Code
004
163.4
2*
BDL
005
109.0
2$
BDL
006
119.7
3*
BDL
007
140.3
2
6.3-7.5xlO-10
008
190.3
3~
8.7-15xlO-l2
Backfill Perm.
Range (m2) 2.8-lOxlQ-H
1.9-3.6xlO-!0 0.3-1.1x10-12 3.9-1.6x10-10 BDL-1.4xlO-9 8.2-250xlO-12 0.4-9.6x10-1° 4.1-7.7x10-12
Vendor's
Aggregate
Name
Localion(s) of
Slab
Penetration
forSSV
SSVPit
(Y and/or N)
SSD Pressures
(Pa)
SSP Pressures
(Pa)
Extra Tests
With Sealed
Wall-Floor
Joint (Y or N)
Perimeter
SSV Location
YorN
3/4" Round
Center of
Central Bay
Y
-125
-375
+125
+375
N
N
Plastic Membrane N
1 1/2" Round
East Bay
YandN
-125
-375
+125
+375
N
Y
N
3/8" Exposed
Center of
North Bay
YandN
-125
-375
+125
+375
Y
N
N
1 3/4" Round
Center of
West Bay
Y
-125
-375
+125
+375
N
N
N
3/4" Round
Center of
West Bay,
Perimeter of
West Bay
YandN
-125
-375
+125
+375
N
Y
N
#8 Pea Gravel
Center of
East Bay
YandN
-125
-375
+125
+375
Y
N
N
1 3/4" Round
Center of
North Bay
Y
-125
-375
+125
+375
N
N
Y
3/4" Round
Center of
North Bay
Y
-125
-375
+125
+375
N
N
Y
BDL = below detection limit of approximately 10*13 m2
Number of subslab regions bounded by footings
TFooter between bays terminates approximately 0.75 m from one end of slab permitting pressure extension between bays.
TSoil saturated approximately 18 cm below slab
Footer between bays terminates approximately 4.3 m from one end of slab permitting pressure extension between bays
Three subslab regions; however, interior footings terminated approximately 1 m from perimeter footings permitting pressure extension between
bays.
"North bay footer terminates approximately 4 m from end of slab, south bay footer end approximately 8 m from end of slab permitting pressure
extension between all bays
PRESSURE AND AIR FLOW MEASUREMENTS
Pressures at the SSV penetration and air flow rates were measured in the temporary SSV
apparatus constructed of 4" diameter PVC pipe. Air flow rates were measured in series using a
pitot tube, a sharp edged orifice flow plate, and a hotwire anemometer. Pitot tube and orifice
plate pressures were measured using a calibrated electronic digital manometer (measurement
uncertainty +/-1 Pascal). Fans typical for radon mitigation were used to pressure and
depressurize beneath the slab. Air flow rates and thus pressures were adjusted using a
mechanical damper located in the pipe near the fans.
-------
E
CO
CO
Figure 1. SSV and test hole locations at slab in house 002 (typical
of all 8 houses). Backfill pressure measurement locations dictated
by driveways, sidewalks, etc.
Note: Not to scale
The digital manometer was also used to measure differential pressures, relative to the
pressure in the basement, at the SSV penetration and at the test holes distributed throughout the
slab. Small tubes were temporarily sealed into the test holes. These tubes remained capped
during the tests with one hole at a time accessed for the individual test hole pressure
measurements. All test hole pressures were measured during each test condition.
Pressures were also measured in the backfill region at 2 to 4 locations around each house.
Probes were installed 0.2 m from the exterior concrete substructure walls and to a depth of 1.5 m
in full basements, and to a depth of 0.2 m from the top of the footing in shallow basements. The
probes were also used to measure soil permeabilities and remained in place during testing for
each house.
RESULTS
LIMITATIONS OF DATA
Due to the differences in the sizes and geometry of the structures, soil conditions, aggregate
and soil permeabilities, interior footing details, differing substructure crack-widths, and other
variables, it is not possible to directly compare the test results between houses. This discussion
will be limited to findings for individual houses only.
A numerical model developed by Lawrence Berkeley Laboratory, validated in part using the
data from this study, does allow comparison of variables among different idealized structures and
configurations (Gadgil 1991).
-------
EFFECTS OF INDIVIDUAL VARIABLES ON PFE
SSVpit
Initially, the experimental methodology called for a series of PFE tests to be conducted
through a carefully cut hole in each slab, followed by an additional series of PFE tests after a
small 25 cm (10 in) radius pit was excavated in the aggregate and soil immediately below the
SSV pipe. However, after these tests were completed in the first house, it was clear that the pit
had a significant impact on PFE. Considering the minimal cost and basic nature of this measure,
it was decided to eliminate some of the non-pit tests from this research effort with the
assumption that this measure will be generally incorporated in SSV systems.
When a 25 cm (10 in) radius pit was excavated in the gravel and soil at the SSV pipe
location in house 006, the PFE increased substantially (Figure 2).
Addition of a pit at house 002 (high permeability soil and large grade aggregate) allowed a
reduction in SSV pressure by a factor of 3 while producing a significant increase in PFE
(Figure 3).
These results suggest that adding pits at SSV systems can increase SSV performance and
simultaneously allow down-sizing of SSV fans.
23456789 10
DISTANCE FROM SSV (METERS)
-125P* NbHt-NoSedlo» A -I25P* Ht-KoSetBn* X -US Pi Hi-Slab Scaled
Figure 2. Effect of SSV pit and slab-wall joint sealing on pressure
field extension in house 006 (low permeability soil, small subslab
aggregate). Lines connecting data points on this figure arc included
for visual guidance only.
-------
HI
cc
a:
a.
350-
30fr
25a
200-
150-
100-
50-
-375 Pa NO PIT
° -125 Pa PIT
2345
DISTANCE FROM SSV (METERS)
Figure 3. Effect of SSV pit on pressure field extension in house
002 without slab-wall joint sealing (high permeability soil, large
subslab aggregate). Lines connecting data points on this figure are
included for visual guidance only.
Slab sealing
All intentional openings through the slab (i.e. plumbing and utility penetrations), and visible
surface cracks were sealed prior to all PFE tests. In some homes the wall-floor joint and cold
joints in the slab were intentionally left unsealed to allow before and after sealing tests to be
conducted. Sealing was performed by the researchers using methods that might be expected fron
a building contractor: the slab-wall joint area was scraped to remove loose concrete, vacuumed,
brushed with a solvent, and then sealed with a one-part urethane caulking material.
The effect of slab sealing on PFE in house 006 was substantial. Slab sealing at the wall-flooi
joints (approximate joint width 1 to 2 mm) served to establish pressures throughout the entire
subslab region in the range of 85% to 95% of SSV pressure (Figure 2).
Unfortunately it was not possible to thoroughly seal the slabs in houses 003, 004, and 005,
due to the unscheduled erection of frame walls over the wall-floor joints.
Subslab aggregate
As stated previously, one of the primary objectives of the study was to investigate the
influence of soils and subslab aggregates of various permeabilities on PFE. Since the subslab
aggregate layer could not be tested individually across the different houses and soil types, it is
impossible to attribute differences in PFE between separate houses to the aggregate type. See
Gadgil (1991) for results of numerical modeling of aggregate and soil permeability effects on
PFE.
When the ratio of aggregate permeability to soil permeability is large (greater than 2 orders
of magnitude) excellent PFE should be produced by typical SSV systems. However, as the
-------
aggregate permeability approaches the permeability of the soil, PFE should decrease since the
resistance to soil-air flow through the aggregate is no longer significantly less than the resistance
to flow through the soil beneath.
Plastic membrane
It was hypothesized that a membrane placed between the high permeability soil and the
gravel layer might function to contain the pressure field within the subslab gravel layer (increase
the ratio of aggregate permeability to soil permeability). Since the pressure field might then be
"shaped" horizontally from the SSV location, instead of dissipating radially into the high
permeable soil, the PFE would be increased.
Since the sub-aggregate membrane could not be tested individually across the different
houses, aggregate and soil types, it is impossible to attribute differences in PFE at the individual
houses to the membrane. Numerical modeling, using this and similar experimental data, is
required to assess PFE, radon entry, and other potential effects of sub-gravel membranes under
various conditions.
SSDvsSSP
In all the houses, and under all configurations except one, subslab pressurization produced
greater PFE than did subslab depressuration (see Table 2). We have no theoretical explanation
for this trend. Additionally, SSV flow rates were generally lower in the SSP mode compared to
SSD at identical pressures in the SSV. It must be cautioned that problems associated with long
term effectiveness of some SSP systems have been reported (Prill 1988).
TABLE 2. SUBSLAB PRESSURIZATION VS SUBSLAB DEPRESSURIZATION
Ratio of Pressure Field Extension (Subslab Pressure Divided by SSV System Pressure) for SSD
to Pressure Field Extension for SSP, Based on Data from Test Holes in Each House
SSV Pressures (Pa) House Code
001 002 003 004 005 006 007 008
No Pit/No Seals
125/-125 1.14 1.0 1.0
3757-375 1.0 1.12 1.0
No Pit/Seals
125/-125 1.03
375/-37S 1.0
Pit/No Seals
1257-125 1.05 1.17 1.02 1.07 1.0
3757-375 1.01 1.09 1.05 .98
Pit/Seals
1257-125 1.03 1.04 1.0
3757-375 1.02 .99 1.0
-------
SSV pressures
As SSV pressures were increased, normalized PFE is decreased as shown in Table 3 for all
houses and all configurations except one. The normalized PFE does not increase in direct
proportion to increases in SSV pressure. Therefore, even though increased SSV pressure does
increase the overall PFE, efficiency is compromised, as discussed by Gadgil (1991). An increase
in SSV pressure causes air velocities in the aggregate layer also to increase, resulting in inertial
losses. This trend is also evident from the laboratory measurements of pressure gradient vs
velocity in the gravel (Gadgil 1991). This effect can be seen in the PFE test results for house 003
(Figure 4).
TABLE 3. EFFECT OF INCREASED PRESSURES AT SSV SYSTEMS ON
PRESSURE FIELD EXTENSION
Averaged Normalized Pressures from Multiple Test Holes
SSV Pressures (Pa)
001
002
House Code
003 004 005
006
007
008
No Pit/No Seals
-125
-375
Ratio
.29
.25
1.16
.31
.25
1.24
No Pit/Seals
-125
'375*
Ratio*
Pit/No Seals
-125
-375 .
Ratio*
Pit/Seals
-125
-375
Ratio
.33
.27
1.22
.34 .84 .43 .60
.29 .78 .40 .58
1.17 1.08 1.07 1.03
.58 .93
.52 .96
1.11 .97
.25
.21
1.19
* Normalized subslab pressures (averaged) at -125 Pa SSV pressure divided by normalized
subslab pressures (averaged) at -375 Pa SSV pressure.
-------
234
DISTANCE FROM SSV (METERS)
Test holes located in secondary aggregate bay
(PFE obstructed by continuous footing)
-125 Pa No Pit-Slab Sealed * -125 Pa Pit-Slab Sealed
Test holes located in SSV aggregate bay
-125 Pa No Pit-Slab Sealed A -125 Pa Pit-Slab Sealed
-380 Pa No Pit - Slab Sealed a -375 Pa Pit - Slab Sealed
Figure 4. Effects of SSV pit and slab-wall joint sealing on pressure
field extension in house 003 (high permeability soil, small subslab
aggregate). Footing was continuous - no openings between
seperate (bounded) aggregate areas. Lines connecting data points
on this figure are included for visual guidance only.
SSV air flow rates
Changes in the air flow rate in the SSV systems, as a result of slab wall joint sealing,
excavating a pit, and excavating a pit and sealing slabs varied between the houses (Table 4).
-------
TABLE 4. AIR FLOWS IN SSV SYSTEMS OPERATING AT -125 (Pa)
Configuration House Code
001 002 003 004 005 006 007 008
No Pit/No Seals
1/s 12 8
(cfm) (26) (18)
No Pit/Seals
1/s 10
(cfm) (21)
Pit/No Seals
1/s 20 45 16 17 16 63
(cfm) (43) (95) (34) (36) (34) (133)*
Pit/Seals
1/s 16 1 63 18
(cfm) (34) (2) (133)t (39)
*-97 Pa SSV pressure
t@-110 Pa SSV pressure
Pressures in backfill regions
Measurement of pressures in the backfill region of the houses were conducted for each test
configuration and with the SSV fan(s) on and off the range of values, for all probes at each
house, are summarized in Table 5. Those data with greater than +/- .5 Pa fluctuation, and where
fan-off pressures exceeded 3 Pa, are not reported.
The SSV systems induced some pressure at the backfill region in every house and under most
test configurations. The induced pressures in the backfill region tended to increase as slabs were
sealed and pits added at the SSV.
-------
TABLE 5.
SSV INDUCED PRESSURES IN BACKFILL REGION
Range of Pressure with SSV Fan-On Minus Pressure with SSV Fan-Off (Pa)
House
Code
001
002
003
004
005
006
007
008
SSV
Pressures
(Pascals)
-125
-375
-125
-375
-125
-375
-125
-375
-125
-375
-125
-375
-125
-375
-125
-375
No Pit No Pit Pit
No Slab Sealing Slab Sealing No Slab Sealing
-3 to -6
-8 to -14
-2 to -7
Oto-3
Oto-9
Oto-1
OtoO
Noisy Data Noisy Data
Due to Wind Gusts Due to Wind Gusts
Otol 0 to -10t
-5 to -15 -9 to -32
1 to -5*
Pit
Slab Sealed
-6 to -8
-5 to -16
-8 to -22
-4 to -39
Oto-7
-1 to -3
Oto-1
tl29Pa
t-97 Pa
Interior footings
Interior footings that create a bounded area around the gravel in the SSV bay can significantly
restrict PFE from other subslab regions. PFE measurements were made on both sides of interior
footings, at test holes of equal distance from the SSV penetration, allowing comparison of
normalized pressures in the SSV bay relative to other bays in the structure (Table 6). The PFE
test results, illustrated in Figure 4, show the effect of a continuous footing bisecting the
aggregate bays in house 003 (high permeability soil and small aggregate). The PFE test results
for house 004 (low permeability soil and large aggregate) indicate that only 15% to 40% of the
SSV pressure is transferred to the secondary aggregate bay (Figure 5).
-------
TABLE 6.
FOOTING EFFECTS ON PRESSURE FIELD EXTENSION
House
Code
001
002
003
004
005
006
007
008
Percent of Normalized SSV Pressures Measured at Secondary Bay Test Holes
Compared to Primary Test Holes of Equal Distance From SSV
Continuous Not Continuous Size of
Footings Footings Footing Openings (m)
50%
10% to 20%
25% to 40%
15% to 20%
50% to 75%
25% to 75%
90%
25% to 50%
0.75
4.3
1
4, 8
Note: All values approximate
°I
eg 0.8-
oc 0.7-
D
$0.6-
cc 0.5-
S0.4-
S 0.3-
1 0>2"
go.v
M
M
x xx xx
X X
x xx
X
ft
2 4
DISTANCE FROM PfT (METERS)
| x SeOONDAHYBW - PRIMARY BAY |
-125 Pa SSV Pressure Pit-No Slab Sealing
Figure 5. Effect of continuous interior footing on pressure field
extension between two aggregate bays in house 004 (low
permeability soil, large subslab aggregate). Footing was
continuous - no openings between seperate (bounded) aggregate
areas.
-------
SUMMARY AND CONCLUSIONS
New home construction provides the opportunity to optimize construction features that
enable effective and efficient radon mitigation to be employed by SSV, if necessary. A variety of
measures are currently recommended by the U.S. Environmental Protection Agency (1991), the
National Association of Homebuilders (1989), and others, while some local and regional codes
require and enforce specific measures in new structures (Nuess 1989, WSBCC 1991). In addition
to providing data useful for the development of guidance for the construction of new structures,
these results increase the understanding of radon reduction techniques in existing structures.
These results show SSV systems should have a pit, or sump, at the pipe location through the
slab. SSV pits significantly increase pressure field extension, resulting in systems that are
potentially more effective and reliable. The experimental results show that smaller SSV
pressures can be used to produce equal or improved PFE when a pit is present. This allows SSV
system fans to be down-sized accordingly.
This study also demonstrates that sealing of openings in the substructure can significantly
improve PFE and reduce the required fan size and flow rate.
For effective pressure field extension, footings that bisect subslab areas should have
openings to allow SSV pressures to extend throughout the aggregate layer. Continuous interior
footings reduced PFE between the bounded aggregate areas by at least 50% in all houses.
As pressures are increased in SSV systems, PFE is generally improved. However, these
increases in pressure result in a decrease in the normalized PFE. Results from this study show
that the normalized PFE was decreased by as much as 24%. For purposes of efficiency, creation
of SSV pits and thorough sealing of slabs should be accomplished before increases in SSV
pressures are considered.
ACKNOWLEDGEMENTS
This research was supported at Washington State Energy Office - Energy Extension Service
by Bonneville Power Administration (BPA) Coop Agreement DE-FC79-82BO34623
Modification A027 (WSE) 90-04. This research was also supported at the Indoor Air Program of
Lawrence Berkeley Laboratory by BPA through interagency agreement DE-AI79-90BP06649,
and by the Assistant Secretary for Conservation and Renewable Energy, Office of Building
Technologies, Building Systems and Materials Division of the U.S. Department of Energy under
Contract DE-AC03-76SF00098.
We gratefully acknowledge the efforts of WSEO-EES staff members Amy Mickelson and
Mike Nuess who assisted in the field measurements, and Dena Ackerman for data entry and
document preparation. We thank Chuck Eastwood of BPA for technical management of this
project.
-------
REFERENCES
I.Turk, B.H., Prill, R.J., Grimsrud, D.T., and Sextro, R.G. Effectiveness of radon control
techniques in fifteen homes. In: JAQMA, Vol. 41, No.5, pp. 723-734, 1991.
2.Pyle, B., Williamson, A., Fowler, G, Belzer, F., Osborne, M., and Brennan, T. Radon
mitigation techniques in crawl space, basement and combination houses in Nashville, Tennessee.
In: the 1988 Symposium on Radon and Radon Reduction Technology. USEPA, Denver,
Colorado 1988.
S.Craig, A.B., Lepvic, K.W., and Harris D.B. Design of radon resistant and easy-to-mitigate
new school buildings. In: Proceedings of the 1991 International Symposium on Radon and
Radon Reduction Technology. Philadelphia, Pennsylvania, 1991.
4.Gadgil, A.J., Bonnefous, Y.C., Fisk, W.J., Prill, R.J., and Nematollahi, A. Influence of subslab
aggregate permeability on SSV performance. LBL-31160 Preprint, Lawrence Berkeley
Laboratory, Berkeley, California, 1991.
5.WSBCC (Washington State Building Code Council). Radon resistive contraction standards.
In: Washington State Ventilation and Indoor Air Quality Code. State of Washington, 1990.
6.Nuess, M. Northwest Residential Radon Standard. Vol.1. Project Report. Bonneville Power
Administration, Portland, Oregon, 1989.
7.Garbesi, K. Experiments and modeling of the soil-gas transport of volatile organic componds
into a residential basement. LBL-25519 Rev. Lawrence Berkeley Laboratory, Berkeley,
California, 1988.
S.Prill, R. I, Fisk, W.J., and Turk, B.H. Evaluation of radon mitigation systems in 14 houses
over a two-year period. LBL-25909. Lawrence Berkeley Laboratory, Berkeley, California,
1988.
9.NAHB (National Association of Homebuilders) National Research Center. Radon handbook
for the building industry. NAHB National Research Center, Upper Marlboro, Maryland, 1989.
10.U.S. EPA. Radon resistant construction techniques for new residential construction, technical
guidance. EPA/625/2-91/032. U.S. Environmental Protection Agency, 1991.
-------
EVALUATING RADON-RESISTANT CONSTRUCTION PRACTICES IN FLORIDA
By: John W. Spears, Harry E. Rector, David P. Wentling
GEOMET Technologies, Inc.
20251 Century Boulevard
Gennantown, MD 20874
ABSTRACT
This paper discusses the results of an evaluation of the
radon resistance of 20 new homes built in accordance with the
Florida draft radon construction code provisions. This project
is part of the ongoing Florida Radon Research Program.
Preconstruction soil measurements included soil permeability,
soil gas radon, radium content and other physical properties of
the soil. Postconstruction measurements included indoor radon,
subslab radon, pressure differentials, pressure field extension,
duct leakage, air infiltration and air leakage. Radon entry rate
or soil gas entry rate was estimated by using perfluorocarbon
tracers (PFTs) buried in the soil around the building and
measuring the soil-based PFT concentrations in the home.
Recommendations on improving the draft radon construction
code are given.
This work was supported by the Florida Department of
Community Affairs under Contract No. 91 RD-41-14-00-22-009, as
part of the Florida Radon Research Program under Interagency
Agreement No. RWFL 933783-03.
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
One of the primary goals of the Florida Radon Research
Program (FRRP) is to develop a set of radon-resistant new home
construction standards that can be adopted by the state building
code. Within the FRRP, a family of new house evaluation projects
(NHEPs) was launched to evaluate the performance and
effectiveness of radon-resistant building design and construction
criteria and to develop recommendations based on the evaluation
for code refinements. To accomplish this, 20 new homes were
constructed to code specifications in the Gainesville area as an
integral component of the workstream of established builders and
then submitted to a systematic series of tests to evaluate the
efficiency of improved slab construction, subslab suction
systems, and other facets of the code in reducing indoor radon.
The fundamental questions addressed by the project relate to
issues of radon entry into homes and the beneficial impact of
various construction characteristics on reducing radon entry.
Once the cause and effect relationships linking building
characteristics and radon entry are better understood, better
advice can be offered to builders, through standards or codes, on
how to build radon-resistant homes.
Most of the work to characterize radon entry into homes has
looked at radon entry in an existing home and then measured the
change in radon entry due to changes in the building (e.g.,
caulking, sealing, subslab depressurization). This approach has
offered a direct measure of change in building performance and
allows for the quantification of the effectiveness of the change.
In new homes, however, the problem is quite different. The
precondition with which to compare performance does not exist.
To date, the EPA new house evaluation program has merely measured
the indoor radon concentration in new radon-resistant homes and
attempted to correlate it to radon soil measurements with little
success (see, for example, NAHB 1991). Empirical relationships
have been proposed to broadly correlate soil-based measurements
with indoor radon for populations of houses on a regional basis
(e.g., Kunz 1989), but predictions for individual houses are
bracketed by large uncertainties. This is not so much a problem
with basic understanding, the theoretical framework is largely in
place (Nazaroff, Hoed, and Sextro 1988), but rather one of
measurement.
-------
STUDY DESIGN AND PROTOCOL
Marion and Alachua County builders were recruited in
cooperation with the local home builder associations. The
building sites offered by the builders were evaluated in the
light of radon data from the statewide radiation survey, geologi-
cal maps from the University of Florida and other information
about the area. Each home was built in accordance with the
provisions of the Florida draft radon construction code (Dixon
1991). The code provisions address critical areas such as soil
covers, slab design, space conditioning systems and subslab
depressurization systems. Details of the draft code are
available through the Florida Department of Community Affairs.
During construction inspections were made of the footing
placement, slab mix, slab cracking, control joints and the active
subslab depressurization (ASD) system installation.
Each home received a common core set of measurements
(Table 1) to evaluate radon potential of the soil, the building's
overall resistance to radon and soil gas entry as well as
detailed data to evaluate dynamic forces that influence building
performance and to evaluate subsystems performance.
The measurement goals of this project focused on soil
conditions at each building site (the primary radon source), as
well as physical and dynamic conditions after completion. Soil
measurements included radium content, soil gas radon, soil
permeability, moisture content and physical characteristics.
Building measurements include leakage rate, soil gas entry rate,
radon concentrations and crack dimensions. Building dynamics
tests include pressure effects of the heating and air-
conditioning (HAC) system performance and the ASD fan on the
indoor and subslab environment.
DATA ANALYSIS
Basic characteristics of the study homes are summarized in
Table 2. These are of modest size (three bedrooms, multiple
baths) averaging about 1,600 ft floor space.
All of the study homes are of slab-on-grade construction;
the principal differences among the study homes lie in the slab
edge detail (monolithic versus poured-into-stem wall) and subslab
features of the ASD system (manufactured ventilation mat versus
suction pit). Three of the study homes (sites 6, 7, 8) are of
townhouse design. There is one two-story home (site 5) in the
study (see Table 2).
-------
TABLE 1. MEASUREMENT PACKAGE
Measurement
Preconstruction
Soil permeability
and soil gas radon
Postconstruction
Soil gas entry
efficiency
Long-term radon in house
Long-term radon in the
sub slab area
Pressure field
extension
Subslab
depressurization
system performance
Air infiltration and
leakage
Radon entry rate
Ins trumentat ion
MK-II permeameter
Type GP soil gas sampling probes
Pylon AB5 sealer or equivalent
Calibrated scintillation cells
Perfluorocarbon tracer (PFT)
placed in the soil
Capillary adsorption tube (CAT)
in the house
Alpha track detector (ATD) in the
house
ATD below the slab
Plastic tubing network installed
under slab before slab is poured,
Micromanometer
ATD, Scintillation cells, Anemo-
meter, Micromanometer
Blower door
PFT
Micromanometer
Anemometer
Digital temperature
PFT in house and in soil
CAT in house
ATD in house and in subslab
Measurement
Protocol
Williamson
and Finkel
(1990)
Derived from
D'Ottavio
et al. (1987)
USEPA (1989)
USEPA (1989)
Tyson and
Withers
(1991);
Spears and
Rector (1991)
Williamson
and Finkel
(1990)
Williamson
and Finkel
(1990); Dietz
and Cote
(1982)
Derived from
D'Ottavio
et al. (1987)
-------
TABLE 2. PHYSICAL CHARACTERISTICS OF STUDY HOMES
Site
ID
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
Area
(ftr)
2,289
1,141
1,769
2,795
1,200
1,106
1,106
1,106
1,600
1,771
1,402
1,402
1,607
1,484
1,600
1,607
1,607
1,950
2,369
2,112
Slab Edge
Detail
Monolithic
Monolithic
Stem Wall
Stem Wall
Stem Wall
Monolithic
Monolithic
Monolithic
Monolithic
Monolithic
Monolithic
Monolithic
Monolithic
Monolithic
Monolithic
Monolithic
Monolithic
Stem Wall
Monolithic
Stem Wall
Subs lab
Detail
Ventilation Mat
Suction Pit
Ventilation Mat
Ventilation Mat
Ventilation Mat
Ventilation Mat
Ventilation Mat
Ventilation Mat
Suction Pit
Suction Pit
Suction Pit
Suction Pit
Suction Pit
Suction Pit
Suction Pit
Suction Pit
Suction Pit
Ventilation Mat
Ventilation Mat
Ventilation Mat
Other
Two Story
Triplex- Center
Triplex- Left
Triplex- Right
-------
PRECONSTRUCTION MEASUREMENTS
The building sites are on primarily sandy soils with varying
minor proportions of silt and clay. Measured permeabilities
generally ranged between 10" and 10* m with occasional
occurrences of thin lenses with permeabilities at or below
10* m . Soil gas radon concentrations (at 1 m depth) varied
from a few hundred pCi/L to just over 20,000 pCi/L, generally
exceeding levels indicated by the radium data (<1 pCi/g), and
indicating the presence of a radon source at depth.
Preconstruction and postconstruction radon measurements are
summarized in Table 3.
POSTCONSTRUCTION MEASUREMENTS
Radon
Indoor radon levels were measured for 3 days using charcoal
and for 30 days using alpha track detectors (ATDs). Subslab
radon was measured during the 3-day indoor test with a grab
sample and with a 30-day ATD during the 30-day indoor test.
Short-term radon indoors ranged from 0 to 8.5 pCi/L and 30-day
radon ranged from 0.9 to 3.8 pCi/L. Subslab radon ranged from a
low of 54 pCi/L to a high of 3,061 pCi/L. These measurements
were made with the ASD system capped.
Only one of the study homes exhibited indoor radon levels
that could cause concern. At site 2, the 2-day (charcoal
canister) screening level was at 8.5 pCi/L and the 1-month (ATD)
confirmatory level was at 3.8 pCi/L. These data were collected
in November 1991. The ASD fan was installed and activated in
early January 1992. A follow-up test (1-month ATD) beginning in
mid-January 1992 with the ASD system activated showed that indoor
levels were reduced to 2.9 pCi/L.
The main feature distinguishing house 2 from the others in
this study is the measured radium content of the soil. At this
site the soil radium content was found to be over 5 pCi/g,
largely supporting the nearly 2,700 pCi/L observed in the
preconstruction soil gas sample. As noted earlier, all other
homes enrolled in the study were in soils of relatively lower
radium content (commonly <1 pCi/g) in the surface layers.
Slab Integrity
The simplest measures of slab integrity for the project
arise from simple inspections. The vapor barrier inspection was
ordinarily coordinated with placement of the subslab
-------
TABLE 3. SUMMARY RADON DATA FOR STUDY HOMES
Site
ID
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
Preconstruction
Grab
Soil
Gas
pCi/L
1,163
2,669
826
1,457
272
704
704
704
944
311
1,340
139
N/A
865
1,487
961
837
20,194
2,875
473
Grab
Fill
Radium
pCi/R
0.1
0.2
0.4
0.3
0.4
1.0
1.0
1.0
N/A
0.3
N/A
0.2
0.2
0.3
0.3
0.4
N/A
3.2
0.3
0.4
Grab
Soil
Radium
pCi/R
0.4
5.5
0.6
0.5
1.0
0.5
0.5
0.5
0.6
0.4
0.8
0.5
0.4
0.6
0.6
0.6
0.4
0.6
1.0
0.3
Postconstruction
AID
Subslab
pCi/L
361
N/A
946
220
1,004
842
1,226
1,144
120
768
54
154
303
177
249
N/A
1,143
3,061
311
304
ATD
Indoor
pCi/L
0.9
3.8
2.0
1.1
2.3
2.2
1.4
1.1
2.2
1.2
N/A
1.7
1.4
0.8
1.5
N/A
2.3
2.7
1.4
0.9
Grab
Subslab
pCi/L
175
1,586
2.896
369
2,130
2,144
565
1,231
282
714
272
639
355
166
292
263
459
1,404
1,294
554
Charcoal
Indoor
pCi/L
1.1
8.5
2.3
1.0
0.8
3.4
1.8
2.2
0.8
1.1
1.1
1.3
1.6
1.1
0.9
1.7
1.4
1.5
1.9
0.0
-------
pressure/sample lines prior to the concrete "pour. Inspection of
the finished slab for shrinkage and settlement cracks was
accomplished prior to installation of the floor coverings. Two
of the study homes (sites 6 and 19) exhibited cracks in the
hairline (visible) and fine (<0.04 cm) categories. These homes
have the distinction of presenting the longest dimension for
rectangular slabs. The slab for site 6 (shared by sites 7 and 8)
spans 32 m along the main axis. The slab poured for site
19 spans 22 m. The next longest span (21 m) occurred at site 18.
The slabs of the other 18 homes were free of cracks at the time
of final inspection.
Pressure and Air Leakage
Air exchange rates (from PFT testing) ranged from just under
0.2 air changes per hour (ACH) to just under 0.7 ACH. The
relative tightness of the homes is evidenced by the blower door
results in most cases. Airflows at 50 Pa are largely correlated
with the PFT results, ranging from just over 5 ACH to just over
15 ACH. Duct leakage is extremely small in these houses,
generally amounting to less than 1 ACH at 50 Pa.
Although craftsmanship of the HAC system largely precludes
general depressurization of the homes from duct leakage effects,
the physical layout of the system nonetheless depressurizes
individual rooms when the doors are closed. Such patterns arise
because individual rooms ordinarily have supply registers while a
single return grille serves the entire indoor airspace. As
shown in Figure 1, at least one room becomes depressurized
(relative to outdoors) when the interior doorways are closed,
reducing the supply/return balance. In one case, the
depressurization reached nearly 10 Pa. In some cases, these
localized depressurization effects were transferred across the
slab.
Houses constructed with the ventilation mat option for
extension of the ASD-induced subslab pressure field seem to also
transmit the HAC-induced effect better than the houses served by
the ASD pit. This effect is overwhelmed, however, when the ASD
system is operating. With the ASD fan running, the reference
pressure at the stack ranged from 350 to 500 Pa. In ventilation
mat-equipped houses the pressure field was readily extended to
all of our subslab measurement points because the largest
transfer distance from the nearest mat edge through the fill was
never more than a few meters.
-------
rJ
a
a
o
a
Ed
CO
CO
Ed
K
a.
HAC ON
ROOM-TO-ROOM
20
10 -
-10
r
I
iii
iii
GRm
BRs
12 3456789 10 11 12 1J 14 15 16 17 18 19 2 0
BITE ID
FIGURE 1. ROOM-TO-ROOM PRESSURE PATTERNS DUE TO HAC OPERATION
-------
Soil Gas Entry
The PFT methodology is a straightforward adaptation of the
constant injection technique for measuring air exchange and
interzonal airflows in buildings (Dietz and Cote 1982; Dietz et
al. 1986). The soil-based PFTs provided additional measures of
radon entry potential. Further, by placing separate PFTs along
each building edge, we were able to examine spatial differences
in soil gas entry. Basic mass-balance relationships, illustrated
in Figure 2, are employed to calculate the effective emission
rate (Se) for a given soil-based PFT from measured indoor
concentrations of the soil-based PFT (Ce), and values (calculated
previously from the indoor PFT data) for the total airflow
exiting the building (Qt) composed of infiltration from the
atmosphere (Q,) plus radon entry airflow coming from the
atmosphere through the soil and into the building (Qe) .
SOIL-BASED PP
FIGURE 2. MASS-BALANCE RELATIONSHIPS FOR SOIL-BASED PFTS
For the case at hand, the total airflow exiting the building
consists of direct flow from the local atmosphere plus flow from
the atmosphere, through the soil, and into the building. Indexed
-------
against standardized emissions for each soil-based PFT, the
fraction of PFT emissions that make it into the building provides
a qualitative measure of soil gas entry. The full mathematical
development is presented in a recent report (GEOMET 1992).
Soil gas entry for these homes generally transported no more
than 10 percent of the PFT emissions to the indoor airspace. The
highest PFT entry rate observed was about 50 percent along one
side of house 4. None of the houses exhibited homogeneous
results; PFT entry was usually highest (by a factor of two or
more) along one edge. This dominance does not appear to be an
artifact associated with a particular PFT: the dominant face was
found to change from house to house. Earlier studies in New York
State conducted by researchers from Brookhaven National
Laboratory showed nearly complete PFT entry in homes with
basements (D'Ottavio et al. 1987). The lower PFT entry rates
observed here are most likely due to a combination of relatively
low pressure forces allied to the higher flow resistance offered
by the soil and subslab fill.
While the PFT entry data provide numerical values, further
refinement is needed to develop quantitative estimates of soil
gas entry. Given measurements of the PFT concentrations in the
soil gas, we could treat the soil regime as a chamber that is
flow-coupled to the indoor airspace and formulate equations where
the entry flow is the only unknown.
Although we have no direct measurements of soil gas PFT
concentrations, we can nonetheless estimate the concentration
using a series of assumptions. The molecular weights of the PFT
compounds used are on par with the atomic weight of radon. If we
assume that the PFT will diffuse through the soil similarly to
radon, then we can hypothesize a length scale using the concept
of 1-hour diffusion distance. This is similar to the lifetime
diffusion distance concept applied to radon, but with the
"lifetime" established at 1 hour instead of the inverse of the
radioactive decay constant. Armed with this concept, we can
dimension a sphere that contains 1 hour's emissions and, from our
earlier estimates of the fraction of emissions entering the
building, assume that this also corresponds to the fraction of
the volume that is swept into the building.
Assuming a bulk diffusion constant of 4xlo"6 m2s"1, we get a
1-hour diffusion length of about 12 cm, forming a 7.2 L sphere at
each soil-based PFT source. We can then estimate the entry flows
by extending this concept along each building edge. If the
effective entry rate of a soil-based PFT were found to be
20 percent, for example, this would correspond to 1.4 L/h passing
through every 24 cm section along a particular side (4.2 volume
-------
elements per meter of length); if the side were 10 m long, the
total entry flow would be about 60 L/h. Soil gas entry rates
estimated through these assumptions ranged from a few liters per
hour to a few hundred liters per hour. Houses tested under cold
season conditions generally showed higher soil gas entry flows
than houses tested during the warm season. While this is
consistent with general expectations, warm season testing was
largely confined to one subdivision. Return tests would be
needed to establish seasonal differences in individual houses to
confirm this apparent trend.
The second area of uncertainty involves the migration of
PFTs in the soil. Selection of the diffusion constant has been
based on the broad assumption that the PFTs would diffuse through
the soil at a rate similar to that of radon. While this
assumption may be reasonable, confirmatory measurements (e.g.,
benchtop tests, possibly field tests) should be considered.
Further, broad concerns could be raised over possible
sorption/desorption of the PFTs in the soil matrix. Such effects
could lead to underestimations of soil gas transport; no
information was found, however, to establish or deny such
effects, except that the experiments of D'Ottavio et al. (1987)
observed complete entry of the soil-based PFTs. Measurements of
PFT diffusion in the soil could also address this issue. Also,
the summary approach utilized here also implicitly assigns soil
conditions (e.g., porosity, moisture content) that may not
prevail in detail from site to site.
The PFT-based estimates of soil gas entry allow us to
evaluate radon entry from a more dynamic perspective. Using the
field measurements to approximate the soil gas concentration in
the swept zone of the soil near the footing, we can evaluate
indoor levels attributable to soil gas entry along the building
perimeter. To keep things simple, we have taken the data at face
value; our only adjustment is to assume an outdoor radon level of
0.3 pCi/L (consistent with annual average conditions observed in
the national ambient survey reported by Hopper et al. 1991).
Figure 3 plots the fraction of observed indoor radon
attributable to soil gas entry along the building perimeter. At
flow rates above 50 to 100 L/h, indoor radon levels are fully
accounted for by the perimeter entry flows. At lesser flows, the
entry rates can account for only relatively small percentages of
the observed radon levels. For study houses 1, 4, 18, and 19,
the estimated soil gas entry flows from the perimeter account for
the majority of the radon entry needed to match the observed
indoor concentrations. Although house 2 required the second
-------
8
w
B
g o.i-=
H
t<
0.01
PFT-BASED PREDICTIONS OF IHDOOR RADON
OUTDOOR RADOV AT 0.3 pCi/L
100
200 300
SOIL GAS ENTRY, L/h
400
500
SOU GRAB H SLAB GRAB SLAB ATD
FIGURE 3. RADON ENTRY ESTIMATE FROM PFT DATA
-------
highest radon entry rate at 1.3 pCin s (house 18 required
1.4 pCim" s* ) , estimated soil gas entry flows of 50 L/h would
support no more than 0.3 pCim" s" , accounting for only about
20 percent of the radon entry needed to accommodate the observed
indoor radon. This residual of just over 1 pCim" s" was the
largest observed across the study houses.
Possible explanations for this include diffusion across the
slab as well as pressure-driven flow that recirculates air
between the indoor airspace and the subslab. We did not attempt
to calculate the flux attenuation factor for the slab that would
need to be invoked to explain such transport by diffusion alone;
given the general absence of cracks prior to installation of
floor coverings, however, we expect the diffusive transport to be
too small. The prospects for closed-circuit transport caused by
HAC-induced depressurization of individual rooms relative to the
zone containing the return, however, is more plausible (even
though the entry/re-entry pathways are not clearly evident).
Estimates of soil gas entry were used to evaluate the soil
gas concentration needed to produce indoor concentrations above
4 pCi/L. To provide some feeling of the possible range,
calculations were performed with temperature corrections for soil
emissions spanning the 13* to 28 *C range of average monthly
temperature (USDOC 1962). As shown in Figure 4, the four houses
exhibiting dominant perimeter soil gas flows (sites 1, 4, 18, 19)
would appear to be sufficiently resistant for sustained soil gas
concentrations well above 2,000 pCi/L. The results for house 18
are particularly encouraging because the indoor concentration
measured with the ATD was at 2.7 pCi/L, making for a smaller
contribution from outdoor radon concentrations. The 20,000 pCi/L
lower bound corresponds well with the 27,000 pCi/L soil gas
concentration observed in the preconstruction test.
Costs
The builders recruited to the study installed either the
ventilation pit or the ventilation mat below the slab. The stack
design specifications were the same for all houses. Eight of the
stacks were terminated in the roof cavity and capped, the
remaining 12 were carried through the roof and capped. Following
the project testing, stacks that terminated above the roofline
were left uncapped to provide for passive venting. Stacks
terminating in the roof cavity were left in the capped condition
at the end of testing. The average cost charged to the project
by the builders for the passive system was $112 for the pit
system and $490 for the ventilation mat. One house (site 2),
equipped for the ventilation pit system, required activation to
-------
1, 000,000:j
8
1
«
M.
O
EH
w
<
100,OOOi
10,000 -
m
RAIXW ENTB.Y EFFICIENCY
OTTTOQQR. RADON - 0.3 pCi/1
m
m
**<
m
m m
*
m
m
m
m
m
m
X , w V V I I I I I f I I I I I I t I I I I I I I I I I i I I I I I I I I I I I I I I I | | |
123456789 1011121314151617181920
SITE ID
FIGURE 4. ESTIMATES OF SOIL GAS RESISTANCE
-------
reduce indoor radon levels. The high suction fan (Radonaway
model HS300B with muffler and couplings) was purchased at retail
for $676.45; the alarm system (checkpoint with visual and audible
warning) added an additional $100. Additional labor and
materials ($233.65) brought the total cost to just over $1010.
Although none of the ventilation mat systems required
activation, we can reliably estimate that the added cost of a
suitable fan and couplings would be $144.50. Assuming labor and
additional materials equivalent to our single case for the pit
design, the additional cost for an activated ventilation mat
system would be just over $478, giving a total cost at just over
$968. The ventilation mat system, if it needs to be activated,
offers installation costs that are 14 percent less than the
ventilation pit system.
CONCLUSIONS
Summary analysis of the data indicates that the code
elements incorporated into the study houses offer fairly good
resistance to radon entry. Even though the majority of the
builders hedged on the concrete mix, visible cracks were absent
in all but two houses, and did not appear to contribute to radon
entry. Similarly, the system of pressure forces measured in
these houses was not consistent with mechanisms intuitively
linked to radon entry. Houses with the stronger measured
pressure-related effects did not seem to attach the pressure
effect to radon entry from the perimeter. Nonetheless, various
field measures that relate to radon potential, when coupled to
PFT-based estimates of soil gas transport, seem to match well
enough to "predict" indoor concentration to well within a factor
of two in houses exhibiting soil gas flows above 50 L/h.
Over the near term, it would be desirable to develop a
simplified model to portray the radon resistance; this would be
of direct value in implementing the various code options for new
construction and could also be applied to existing structures.
Although such a model could be as straightforward as a simple
"figure of merit," to depict the greatest indoor-to-subslab
difference that a particular configuration can resist,
formulations need to be developed to incorporate the radon source
term, effective permeability of the construction ensemble (e.g.,
fill plus vapor barrier plus slab), as well as the pressure
forces and flow resistance along entry pathways.
While we can recognize correspondence among various data
elements, devising realistic systems of forces is best realized
within a model-guided context. Of particular concern is the
radon transport that apparently bypasses perimeter flows in the
-------
swept zone near the footing. Our concern lies not so much with
the occurrence: physical mechanisms can be invoked to explain the
observations. Rather, we are concerned over the lack of clues
from construction details, design and craftsmanship of the HAC
system, and other features that should indicate how well the
driving forces are attached to the entry flows. Tracer studies
to separately label the subslab fill and the building perimeter
soils could provide quantitative guidance.
Concerns also arise from environmental modulation of the
radon source term. From the limited opportunities to repeat some
of the field measurements in this study, it is clear that
seasonal influences need to be taken into account. Further,
longer term follow-up would be indicated to establish a basis for
accepting the first year performance as a reasonable measure of
the life cycle for the system. Items that fall into this
category include the formation of cracks, and moderating
influences of the building in sheltering the immediate subslab
zone from climate cycling.
CODE RECOMMENDATIONS
Working with the builders on the project gave us some
insights into the practical implications of the draft radon code.
Each builder was thoroughly briefed on each element of the draft
code and a commitment was made by the builder before construction
began.
During our inspection and construction monitoring process we
observed several areas where there was resistance to following
the draft code provisions and other areas where the builder
focused his attention.
Soil Cover
All builders in the study paid a great deal of attention to
the soil cover or vapor barrier. Most builders actually went
beyond the code and sealed all joints and penetrations with
mastic. The vapor barrier as a radon barrier has great intuitive
appeal to builders and therefore they are not reluctant to focus
attention on this element of the construction process. To build
on this inclination it may be advisable to focus more attention
on the performance of the polyethylene vapor barrier as a radon
barrier. In particular the code should address sealing the
intentional penetrations in the barrier.
-------
Slab Design
Most of the builders in this study chose to ignore the 4 in.
slump code specification and poured slabs with a 5 in. slump.
The builders encountered great resistance from the concrete
subcontractors to the 4 in. slump specification. Additionally,
we were unsuccessful in preventing water from being added on
site. Subcontractors cite problems with "workability" of 4 in.
slump concrete. As noted earlier, slabs inspected prior to
finished floor covering exhibited minimum cracking. We do not
know, however, about crack formation over the long term. Also,
we observed no significant difference in slab cracking between
4 in. slump and 5 in. slump. It would be desirable to conduct
side-by-side long-term slab testing of 4 in. and 5 in. slump
concrete to determine if 4 in. slump significantly reduces
cracking over 5 in. slump.
The current draft code emphasizes the performance of the
concrete as the radon barrier. If the emphasis were also placed
on the soil cover or vapor barrier as the radon barrier, perhaps
the concrete specification would not be as critical. As stated
earlier this would follow the builder's natural inclination to
improve the soil cover and pour 5 in. slump concrete. This
formula seemed to work for the 20 homes in this study.
Space Conditioning System
The requirements of the Florida Energy Efficiency Code for
Building Construction 1991 were well enforced in the Gainesville
study area; therefore, most of the homes had tight ducts and
relatively minor "HAC-ON" depressurization of the indoor
airspace. The builders used mechanical contractors who were
already experienced in duct sealing techniques as a result of the
enforcement of the new energy code provisions. The use of a
pressure differential test with HAC on and HAC off could be a
useful inspection tool for code enforcement.
Subslab Depressurization System
The pit design and the ventilation mat-design were evenly
dispensed among the study houses. Builders chose the pit over
the mat primarily because they were unfamiliar with the mat
material. While the pit showed to be lower first cost, the mat
had 14 percent lower cost if it is was activated because it could
function with a less expensive fan. Additionally, the fan
required for the pit design is much noisier than the mat fan;
therefore, it is less likely to be preferred by the homeowner.
Further the mat system showed a greater pressure field extension
than the pit.
-------
General Comments on the Code
The draft code in its current form is complex and many
elements are readily overlooked by builders, because it requires
so many changes to their work flow. Simplifying the code to
contain only the additions to the existing code should make it
more usable.
REFERENCES
D'Ottavio, T.W., R.N. Dietz, C. Kunz, and B. Kothai. 1987.
Radon Source Rate Measurements Using Multiple Perfluorocarbon
Tracers. Indoor Air '87, Proceedings of the 4th International
Conference on Indoor Air Quality and Climate, Vol. 2, pp.
388-392.
Dietz, R.N., and E.A. Cote. 1982. An Inexpensive
Perfluorocarbon Tracer Technique for Wide-Scale Infiltration
Measurements in Homes. Environment International. Vol. 8, pp.
419-433.
Dietz, R.N., R.W. Goodrich, E.A. Cote, and R.F. Wieser. 1986.
Detailed Description and Performance of a Passive Perfluorocarbon
Tracer System for Building Ventilation and Air Exchange
Measurements. Measured Air Leakage of Buildings. ASTM STP 904.
H.R. Trechsel and P.L. Lagus, Eds., American Society for Testing
and Materials, Philadelphia, PA, pp. 203-264.
Dixon, R.W. 1991. Florida's New Construction Standards.
Presented at the 1991 AARST National Fall Conference, Rockville,
MD.
GEOMET. 1992. Evaluation of Radon-Resistant Construction
Practices in New Homes in Florida. Report No. IE-2588, GEOMET
Technologies, Inc., Germantown, MD. DCA Contract No. 91 RD-41-
14-00-009.
Hopper, R.D., R.A. Levy, R.C. Rankin, and M.A. Boyd. 1991.
National Ambient Radon Study. In Proceedings: The 1991
International Symposium on Radon and Radon Reduction Technology,
Volume 4, EPA-600/9-91-037d (NTIS PB92-115385), p. P9-79.
Kunz, C. 1989. Influence of Superficial Soil and Bedrock on
Indoor Radon in New York State Homes. Report No. 89-14, New York
State Energy Research and Development Authority, Albany, NY.
-------
NAHB. 1991. Field Test and Demonstration of Radon Prevention
Techniques in New Homes. (Prepared for USEPA, New Jersey
Department of Community Affairs, and Jersey Central Power and
Light.) NAHB Research Center, Upper Marlboro, MD.
Nazaroff, W.W., B.A. Moed, and R.G. Sextro. 1988. Soil As A
Source of Indoor Radon: Generation, Migration, and Entry. Radon
and Its Decay Products In Indoor Air (Nazaroff, W.W. and A.V.
Nero, Eds.), pp. 57-112.
Spears, J.W., and H.E. Rector. 1991. Evaluation of Radon-
Resistant Construction Techniques in the Florida New House
Evaluation Program. Presented at the 1991 AARST National Fall
Conference, Rockville, MD. 'A,
Tyson, J.L., and C.R. Withers. 1991. Installation and
Evaluation Techniques Used To Measure Pressure Field Extensions
From Sub-Slab Depressurization Systems Installed In New Florida
Homes. Presented at the 1991 meeting of the American Association
of Radon Scientists and Technologists, Rockville, MD.
USDOC. 1962. Decennial Census of United States Climate -
Monthly Normals of Temperature, Precipitation, and Heating Degree
Days. Climatography of the United States No. 81-6, Weather
Bureau, U.S. Department of Commerce, Washington, DC.
USEPA. 1989. Indoor Radon and Radon Decay Product Measurement
Protocols, EPA-520/1-89-009 (NTIS PB89-224273), Office of
Radiation Programs, U.S. Environmental Protection Agency,
Washington, DC.
Williamson, A.D., and J.M. Finkel. 1990. Standard Measurement
Protocols: Florida Radon Research Program. EPA-600/8-91-212
(NTIS PB92-115294), Office of Research and Development, U.S.
Environmental Protection Agency, Washington, DC.
-------
LABORATORY INVESTIGATIONS FOR THE SEARCH
OF A RADON REDUCING MATERIAL
By: Lakhwant Singh, Jaspal Singh, Surinder Singh,
and H. S. Virk
Department of Physics
Guru Nanak Dev University
Amritsar-143005, India
ABSTRACT
The majority of common building materials contain small
amounts of the naturally-occurring radioisotopes of the uranium
and thorium decay series. The gaseous members of the series Rn-
222 and Rn-220 are of particular importance from the health hazard
point of view due to inhalation exposure. Because of a short
half-life, Rn-220 emanates from building materials only in small
quantities. Rn-222, being a noble gas with a long half-life, can
more through crevices, material pores, and structural failures.
So, in the areas where high levels of radon activity are found,
the choice of materials that act as barriers to radon diffusion
becomes important.
Radon diffusion can e decreased in two ways: by covering the
floor and walls with sheets or foils of radon-resistant materials,
or by coating with suitable paints. The choice of materials
depends on the feasibility of application, availability, and cost.
Both types of experiments are in progress. The percentage
reduction of radon activity due to coating of soil, cement,
whitewash, paint, etc., from radon sources is under study. On the
other hand, different foils of varying thickness are being
investigated as a radon reduction technology. The results from
both experiments will be presented in the paper.
-------
Session XII
Radon in Water
-------
XII-1
RISK ASSESSMENT IMPLICATIONS OF TEMPORAL VARIATION OF RADON AND RADIUM WELL
WATER CONCENTRATIONS
Submitted by:
Alan J. Siniscalchi, M.S., M.P.H. (1); Zygmunt F. Dembek, M.S. (2); David R.
Brown, Sc.D. (1); Carolyn J. Dupuy, M.S., S.M. (1); Margaret A. Thomas,
M.S. (2) Nancy W. McHone, M.L.S. (2); Barbara S. Weiss, B.A. (1); and Maria
C. van der Werff, B.S. (3).
(1) State of Connecticut Department of Health Services, Hartford, CT 06106
(2) State of Connecticut Department of Environmental Protection, Hartford,
CT 06106
(3) U.S. Environmental Protection Agency Region I Office, Boston, MA 02203
ABSTRACT
The accurate assessment of the risk of radon exposure from various household
sources relies on the careful documentation of both dose and length of
exposure. The importance of accurate exposure assessment for both general
population risk assessment and the individual household evaluation cannot be
overemphasized. Current EPA protocol suggests a single water sample be
analyzed for determining the contribution of the drinking water source to
the total radon exposure in homes. The State of Connecticut Department of
Health Services has collected periodic samples of well water that reveal
temporal variations in both the radon and radium concentrations. Radon
concentrations in one well ranged from 34,000 pCi/L to 661,000 pCi/L. The
results of two analytical methods are compared and recommendations for
changes in EPA sampling, analysis and risk assessment procedures are
discussed.
-------
INTRODUCTION
Exposure to radon is causally associated with lung cancer (1,2). The
discovery of elevated radon in residential structures prompted the U.S.
Environmental Protection Agency (EPA) and many state health departments to
conduct statewide surveys to determine the magnitude of residential radon
exposure throughout the country. The EPA provided initial guidance to
states and the general public (2,3,4). The EPA, many state agencies and
others recognized that home drinking well water can be a significant source
of radon-222 (5,6). It was also hypothesized that this source of radon can
pose a risk for gastrointestinal tract cancers through ingestion and
additional risk for lung cancer from off-gassing and subsequent inhalation
(7). Other radionuclides such as radium-226 and uranium-238, have been
found in Connecticut and other states (8,9). Since radium-226 has been
associated with increased risk for development of both bone and head
carcinomas (9), its presence in water must also be acknowledged in
conducting risk assessment evaluations of radon-contaminated wells. The
maximum contaminant levels proposed by the EPA for radon-222, radium-226
and other radionuclides have also drawn increasing attention to radon in
wells (9,10).
RECOGNITION OF TEMPORAL VARIATION ISSUES
The temporal variation of radon concentrations in the air of homes has
been recognized. Both diurnal and seasonal fluctuations were recognized in
the early guidance literature (2,3) prepared by the U.S. Environmental
Protection Agency. A major premise of the original air testing protocol
was designed to compensate for these fluctuations by screening the highest
potential areas of homes. For example, the recommendations for conducting
the screening test call for exposing the radon testing device in the
"lowest livable area" under closed house conditions. These closed house
conditions are "necessary to keep the radon level relatively constant
throughout the testing period (3)."
EPA guidance for further (follow-up) measurements also recognized these
variations. These recommendations specifically mention that "radon levels
can vary greatly from season to season..." Many other references
emphasizing the need to obtain the "average radon concentration" can be
found in the Citizen's Guide and other documents (2,3,4).
In an earlier pamphlet, the EPA provided guidance on the evaluation of
radon in water of homes (5). This pamphlet did mention conducting well
water follow-up tests but only to verify results of initial tests that
"indicate that you may have a radon problem." The possibility of seasonal
variations in water radon concentrations was not mentioned.
-------
Subsequent literature being prepared by the EPA also has not discussed
seasonal variation of radon in water or recommended obtaining an annual
average radon concentration (11,12,13).
DISCOVERY OF TEMPORAL VARIATION IN CONNECTICUT
In February 1989, the Connecticut Radon Program was contacted by a
Connecticut resident for advice on the mitigation of the radon in his home.
He had conducted multiple radon tests and found variations of both the air
and water radon levels. In reviewing the test data obtained by the
homeowner, the Radon Program discovered high levels of radon in both the air
(up to 70 pCi/L) and water (up to 549,316 pCi/L) (15). The Program, in
conjunction with the Department of Environmental Protection Natural
Resources Center (DEP NRC) and the U.S. Geologic Survey (US6S) conducted
extensive testing of the home and well (BM-3) to document what appeared to
be large temporal variations of radon levels (15,16,17).
The initial evaluation of this home and well (BM-3) is described in
"Part I" of our temporal variation studies. A most important finding was
the confirmation of large temporal variations of the radon levels in this
well which ranged from 36,000 to over 661,000 pCi/L.
In order to determine whether this was a unique well representing
unusual groundwater dynamics or whether many wells undergo this type of
fluctuation, the Departments of Health Services and Environmental Protection
conducted a more detailed study. This study, entitled "Temporal Variations
in Private Well Water Radon and Radium Levels in Specific Geologic
Formations and Effects on Indoor Radon," was funded as the first-year
Innovative Project as part of the EPA State Indoor Radon Grants (SIRG)
Program (17). Preliminary findings, considered in Part II of our temporal
variation studies, are reported in a second paper in this symposium (19).
This paper investigates the risk assessment implications of these
findings and suggests recommendations for modifications in EPA sampling,
analysis and risk assessment procedures.
METHODS
A number of studies were conducted in order to gain a better
understanding of the variations of radon and other radionuclides in water
and to examine more closely the relationship between the air and water radon
concentrations.
-------
PART I EARLY STUDIES OF A HIGH RADON CONCENTRATION WELL (BM-3)
Radon in Air Sampling & Analysis
Assessment of the radon in air concentration was obtained using both
active and passive radon monitors. Long-term radon levels were monitored
using alpha-track (AT) devices. Short-term radon levels were obtained using
activated charcoal (AC) passive devices. Additional detailed short-term
exposure data were obtained using a continuous radon monitor. All devices
were obtained from companies listed with the EPA's Radon Measurement
Proficiency (RMP) Program. Quality control/quality assurance (QA/QC)
procedures were used during all tests (20).
Radon in Water Sampling and Analysis
The EPA-approved sampling method for collecting water into a 10
milliliter (mL) sample transferred from the tap stream with a syringe and
placement under a 10 mL mineral oil-based scintillation fluid contained in a
scintillation vial was used. The beta activity of the radon-222 progeny was
then counted using the liquid scintillation method (20,21). Radon-222
concentrations were then calculated (20). Samples were collected on a
monthly to bimonthly basis during two separate six month periods.
A portion of samples in the earlier studies were split with the U.S.G.S.
and analyzed by the alpha scintillation Lucas cell method (emanation)
(16,21).
PART II INNOVATIVE PROJECT STUDIES
The detailed methodology used in this portion of our temporal variation
studies are reported in a companion paper (19). The study is designed to
determine both long and short-term well water radon and radium levels
(18,19,22). A total of five wells with radon concentrations ranging from
3,000 to 662,000 pCi/L are included. These five wells can be divided into
three categories consisting of one "high" (36,000 to 661,000 pCi/L) well
(BM-3), three "moderate" (15,000-25,000 pCi/L) wells (ED-10, ED-15, & ED-18)
and one -"low" (3,OOO pCi/L) well (MS-48). Three phases of the study were
organized. Phase I is designed to identify variations in radon
concentrations that may result from water use. In this phase, five wells
were sampled on an hourly basis for two non-consecutive days. Sampling was
initiated before the "water use day" began and sampling continued for 18 hra
(18,22).
Phase II studies consisted of same-time daily sampling of the same wells
for 14 consecutive days. This phase was designed to identify changes in the
radon concentration resulting from changing weather conditions and other
short-term factors (18,22).
-------
The Phase III studies consisted of weekly testing of the five wells for
a period of one year. This phase was designed to identify long-term trends
in the radon-222 and radium-226 concentrations caused by factors such as
seasonal variations. The data is being compared to information obtained
from the Connecticut Hydrologic Monitoring Network and the National Weather
Service Stations located throughout Connecticut (18,22).
Studies of an additional well (RQ-1) were conducted as part of a second
Innovative Project. As with the other wells, bimonthly samples were
collected using the EPA-approved sampling method. Radon analysis was
performed using a liquid scintilation method.
RESULTS
PART I STUDIES OF A HIGH RADON CONCENTRATION WELL
Samples of radon-222 obtained during the early radon studies of the
"high" study well are shown on Table 1. These findings reflect the average
of two radon-222 sampling results analyzed using the liquid scintillation
method. A portion of the samples were split and analyzed using the
emanation method. The values reported in Tables 2, 4 and the liquid
scintillation analysis for Table 1 are the average of two samples obtained
at the well.
TABLE 1. Radon-222 pre-treatment biweekly samples from a "high" radon
concentration study well (BM-3)
Radon (pCi/L)
Date Liquid Scintillation Emanation
03/28/89 35,699 29,224
06/07/89 208,615 81,483
06/23/89 40,565 35,548
07/05/89 261,830 76,595
07/19/89 570,335 193,672
08/02/89 503,740 145,856
08/09/89 437,235 250,773
08/16/89 40,010 53,846
08/23/89 200,555 157,151
07/18/90 661,219 251,406
08/22/90 620,610 224,299
09/20/90 568,815 201,153
10/18/90 560,155 517,688
11/19/90 57,968 64,820
12/20/90 46,823 36,815
-------
The well was also sampled for radium-226 and radium-228. Radium-228
values were not distinguishable from zero. Table 2 contains values for
Radium-226 determined from samples taken from well BM-3.
TABLE 2. Radium-226 pre-treatment samples from a "high" radon concentration
study well (BM-3).
Date Radium-226 (pCi/L)
06/23/89 4.54
07/05/89 7.71
07/19/89 10.99
07/26/89 11.35
08/09/89 12.22
08/16/89 8.23
08/23/89 6.61
07/18/90 35.85
08/22/90 30.87
09/20/90 24.76
10/18/90 36.24
11/19/90 10.43
12/20/90 4.81
Significant variance in Radon-222 and Radium-226 concentration during
the six month sampling period was found. A variance of 206,147 pCi/L with
liquid scintillation and 76,931 pCi/L with emanation were identified. Table
2 reveals the Radium-226 data to have a variance of 11.32 pCi/L.
Lower values in the radon-222 concentrations were obtained by the alpha
scintillation {Lucas cell) emanation analysis as compared with the radon-222
values obtained using the liquid scintillation method. The accuracy of the
values using the emanation method is questionable when greater than 20,000
pCi/L because of the lack of comparable standards (16).
Radium-226 concentrations also revealed a temporal variation during the
study period, with a variance of 11.32 pCi/L. The absolute concentration
change illustrated in Table 2 was similar to the change in the radon-222
concentration.
PART II INNOVATIVE PROJECT STUDIES
The results of the preliminary analysis of this component of Connecticut
temporal variations are discussed in the companion to this paper (19).
These studies show temporal variations for radon within all three phases
and with all five wells. Weekly radon sampling revealed that the greatest
variance for radon in weekly radon levels was observed in the high well,
BM-3. The least amount of variance (216 pCi/L) was noted in the low well,
designated as MS-48.
-------
TABLE 3. Recent (Phase III) radon-222 and radium-226 sampling results of the
high radon well (BM-3)
Date
09/11/91
09/18/91
09/25/91
10/02/91
10/10/91
10/16/91
10/23/91
10/30/91
11/06/91
11/13/91
11/20/91
11/26/91
12/04/91
12/11/91
12/18/91
12/24/91
12/31/91
01/08/92
01/15/92
01/21/92
01/28/92
02/04/92
02/11/92
02/18/92
02/25/92
03/03/92
03/05/92
03/10/92
03/17/92
03/24/92
03/31/92
04/07/92
04/14/92
04/21/92
04/28/92
05/05/92
05/12/92
05/20/92
05/27/92
Radon-222 (pCi/D*
382,100
433,680
493,200
327,830
450,930
435,060
252,360
503,180
459,090
471,730
482,380
33,560
43,265
40,630
39,208
105,170
263,260
294,730
147,650
363,190
405,780
354,910
293,980
597,820
409,130
440,670
201,950
63,032
41,382
61,105
36,168
34,200
63,032
55,610
52,432
53,845
352,070
205,710
459,870
Radium-226 (pCi/Ll*
34.0
0.86
37.92
36.07
42.6
37.12
42.58
33.69
44.54
70.8
46.08
58.08
0.11
33.16
29.8
0.1
49.25
45.2
44.62
47.2
-
40.53
47.65
66.53
58.0
64.71
31.58
47.16
26.11
35.2
27.56
29.67
26.21
26.78
32.00
15.00
34.22
29.68
37.8
*liquid scintillation method
-------
Preliminary analysis of the recent Phase III values of the highest well
(BM-3) reveal a similar fluctuation in well radon levels as observed in the
earlier studies. The cause of this fluctuation is unknown, but water table
level may be a contributing factor (19). Table 3 illustrates the additional
weekly sampling data observed in well BM-3. With the exception of well BM-3,
radium-226 levels in the wells, although exhibiting variance, often were so
small as to be indistinguishable from zero.
The DHS has examined an additional well (RQ-1) from 11/91 to 4/92 in which
the radon level has had a range of 36,000 to 252,000 pCi/L with a variance of
80,513 pCi/L. Radium-226 had a range of 0 to 19.2 with a variance of 6.5
pCi/L. The data from RQ-1 is seen in Table 4. Radon analysis was performed
using a liquid scintillation method (20,21).
TABLE 4. Radon-222 and Radium-226 Pre-Treatment Samples From a High Radon Well
(RQ-1) in Geologic Strata Different From Well BM-3.
DATE
11/12/91
12/30/91
01/14/92
01/29/92
02/10/92
02/26/92
03/11/92
03/17/92
03/24/92
03/31/92
04/23/92
04/10/92
RADON (pCi/Ll
240,613
222,613
133,788
50,298
35,570
47,163
230,160
74,070
252,463
116,434
211,093
88,718
RADIUM-226 (pCi/Lj
0
13.05
1.37
7.11
8.91
-
14.37
16.34
16.45
16.69
19.32
17.16
DISCUSSION AND CONCLUSIONS
The'results of all studies illustrate temporal variation of radon-222
concentrations in Connecticut groundwater. While these studies were limited
to six wells, they indicate that such variation is not unique.
When the wells with the greatest variation are examined by their
geographic location other trends can be seen. Previous reports by this
department have shown that good correlation can be seen between radon
concentration and the underlying bedrock (8,23,24,25). Additional studies
are needed to determine whether the temporal variations are greater within
certain bedrock units.
The difference in the absolute value of radon-222 concentration between
the two analytical methods suggest that the liquid scintillation method may
be the most reproducible for radon levels greater than 20,000 pCi/L.
-------
Figure 1 is a log plot of the variance versus the mean for each well
sampled. Although the sample size differs for the wells (BM3 = 39, RQ1 =
12, ED 18 = 38, ED10 = 37, ED15 = 38, MS48 = 36), it can readily be seen
that the variance increases with an increase in mean well radon-222 level.
If all wells containing radon-222 were similar in variance, then a straight
line perpendicular to the x-axis could be expected. This does not occur,
and the greatest variance is seen in wells having the highest levels of
radon-222.
The temporal variations in radon-222 and radium-226 concentrations of
these wells have many implications. It illustrates the need to conduct
follow-up sampling of radon-222 water analysis. The weekly radon levels
with the largest variance were in the highest wells (BM-3 and RQ-1). The
amount of variance in radon (and of measurable radium-226) for RQ-1 places
it in a category similar to BM-3.
Cancer risk assessment of radon exposure is related to the cumulative
exposure of radon progeny over time. Therefore, it is important to identify
and quantitate the factors that vary over time. As an important contributor
to indoor radon levels, water radon variation should influence sampling
strategy. These findings suggest that at high levels of water radon,
variation in indoor air radon contributed by household water may be an
important source of uncertainty in the individual risk assessment.
Therefore it appears the greatest health risk from radon in water may
occur from wells not properly assessed, especially if the lower limit of
radon is greater than 30,000 pCi/L. A well in this category has the
potential for radon levels in excess of 200,000 pCi/L and measurable
radium-226 levels. A recommendation for conducting multiple samples over at
least 4 seasons is advisable.
The limitations of radon water treatment require knowledge of the upper
range of radon concentration during a given year. This problem would be
especially critical if the water radon peak occurred during the heating
season when soil gas levels are also at their peak.
Temporal variations have implications when applied on a national scale.
For example, estimates of the radon-222 concentrations from household water
may be underestimated if seasonal water radon peaks occur during the heating
season. Assessments of the air radon in the homes with the wells will help
to address these issues.
Table 5 lists the suggested guidelines for EPA and states to consider in
providing advice for well characterization. These suggested guidelines are
offered to avoid the aforementioned problems inherent in the assessment of
temporally varying radon-222 and radium-226 household well water sources.
-------
TABLE 5. SUGGESTED GUIDELINES FOR EVALUATING A WELL AS A SOURCE OF RADON-222
AND RADIUM-226.
1. Collect all samples using the EPA method. This method is preferred to
avoid loss of radon-222.
2. Analyze all samples using the liquid scintillation method to avoid
underestimating the radon-222 concentration in wells greater than 20,000
pCi/L.
3. Conduct a minimum of 4 samples over a minimum of 12 months prior to
especially if initial samples reveal radon-222 concentrations of 30,000
pCi/L or greater. At least two samples should be analyzed on all wells
4. Identify maximum peak radon-222 concentration in order to select proper
degree of treatment.
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.
ACKNOWLEDGEMENTS
The. authors especially wish to thank Denis Healy of the U.S. Geological
Survey for his invaluable help in the early sampling and analysis. We also
wish to thank the owners of the wells used in the study for their gracious
cooperation.
-------
1000000
Variance
100000
10000=
1000
100
1000
o
i r
10000 100000
Mean Value (pCi/L)
i i i M
100000
BM-3
ED-10
RQ-l -*- ED-IS
ED-18 -e- MS-48
RE 10 *-°9 ^'°* °* Variance vs Mean Value of
Radon-222 levels (pCi/L) for each well
-------
REFERENCES
1. Committee on the Biological Effects of Ionizing Radiations. Health
Risks of Radon and Other Internally Deposited Alpha-emitters. BEIR IV.
National Academy Press, Washington, D.C., 1988.
2. Radon Reference Manual. EPA 520/1-87-20. U.S. Environmental Protection
Agency, Washington, D.C. 1987.
3. A Citizen's Guide to Radon. OPA-86-004. U.S. Environmental Protection
Agency, Washington, D.C., 1986.
4. Radon Reduction Methods. OPA-87-010. U.S. Environmental Protection
Agency, Washington, D.C., 1987.
5. Removal of Radon from Household Water. OPA-87-011. U.S. Environmental
Protection Agency, Washington, D.C., 1987.
6. Becker, A.P. and Lachajzyk. Evaluation of watesrborne radon impact on
indoor air quality and assessment of control options. U.S.
Environmental Protection Agency contract #68-02-3178, February 1984.
7. Cross. F.T., Harley, N.H., and Hofman. Health effects and risks from
Rn-222 in drinking water. Health Phys. 48:649, 1985.
8. Dupuy, C.J., Healy, D., Thomas, M.A., Brown, D.R., Siniscalchi, A.J.,
and Dembek, Z.F. A survey of naturally occurring radionucludes in
groundwater in selected bedrock acquifers in Connecticut and
implications for public health policy. IN; C.E. Gilbert & E.F. Calabrese
(eds) Regulatory Drinking Water Quality. Lewis Publishers, Boca Raton,
FL, 1992. p. 95.
9. U.S. Environmental Protection Agency. Water pollution control; national
primary drinking water regulations; radionuclides; advance notice of
proposed rulemaking. Federal Register. 51(189):34836. September 30,
1988.
10. U.S. EPA. Drinking water regulations under the safe drinking water
act. EPA Fact Sheet, May 1990.
11. Radon and You, Testing and Reducing Radon Levels in Connecticut Homes.
Radon in Real Estate Transaction. Department of Health Services,
Hartford, CT p. 12, 1992.
12. Radon and You, Testing and Reducing Radon Levels in Connecticut Homes.
Diagnosing and Correcting High Radon Levels in Air. Department of
Health Services, Hartford, CT, p. 6, 1992.
-------
13. Radon and You, Testing and Reducing Radon Levels in Connecticut Homes.
Radon in Water. Department of Health Services, Hartford, CT, p. 9, 1992.
14. New Citizen's Guide (draft) U.S. Environmental Protection Agency,
Washington, D.C., 1992.
15. Memorandum to Alan Siniscalchi from Maria van der Werff, April 14, 1989.
16. Letter to Alan Siniscalchi from Denis Healy, May 9, 1989.
17. Letter to homeowner from Z.F. Dembek, December 20, 1989.
18. Innovative Project (proposal). Temporal variations of well water radon
and evaluation of radon sources from bedrock and surficial materials.
Submitted to: U.S. Environmental Protection Agency, Boston, MA, 1991.
19. McHone, N., Thomas, M. and Siniscalchi, A.J. Temporal variations in
bedrock well water radon and radium and water radon's effect on indoor
air radon. Paper at International Symposium on Radon and Radon
Reduction Technology, Minneapolis, Minnesota, September, 1992.
20. Connecticut Department of Health Services Quality Assurance Project Plan
to U.S. EPA Region I, December, 1990.
21. Radon in water sampling. EPA/EERF-78-1. U.S. EPA, 1978.
22. McHone, N. Temporal variations in bedrock well water radon and
radium...Progress report for EPA annual report, March 16, 1992.
23. Toal, B.F. et al. Radon exposure assessment - Connecticut. M.M.W.R.
Vol. 28, No. 42, pp. 713-715.
24. Siniscalchi, A.J. et al. Radon exposure in Connecticut: analysis of
three statewide surveys of nearly one percent of single family homes.
Paper presented at 1990 International Radon Symposium, Atlanta Georgia,
February 19-23, 1990.
25. Siniscalchi, A.J. Radon: what you don't know could hurt you. Vol. 5,
No. 3. Connecticut Academy of Science and Engineering (CASE Reports),
Hartford, CT, 1990.
7910F
-------
XI1-2
SEASONAL VARIABILITY OF RADON-222. RADIUM-226. AND RADIUM-228 IN GROUND
WATER IN A WATER-TABLE AQUIFER. SOUTHEASTERN PENNSYLVANIA
by : Lisa A. Senior
U.S. Geological Survey,
111 Great Valley Parkway, Malvern, Pennsylvania 19355
ABSTRACT
Monthly sampling of a well drilled into the Chickies Quartzite, a water-table aquifer in southeastern
Pennsylvania, indicates that concentrations of radon-222, radium-226, and radium-228 fluctuate seasonally at that
location. The 150-foot -deep well was sampled monthly for 3 years. Water samples were analyzed for concentrations
of radon-222, and dissolved radium-226, radium-228, sulfate, chloride, nitrate, nitrite, and ammonia; dissolved
oxygen concentration, pH, and specific conductance were measured in the field. Maximum concentrations of radon-
222, radium-226, and radium-228 were measured in autumn or early winter when annual depth to water was greatest,
and minimum concentrations of radon-222 and the radium isotopes were measured in the spring when annual depth to
water was least. The relatively low concentrations of radon-222, radium-226, and radium-228 in the spring may
represent dilution by recharge. For the period of record, radon-222 concentrations ranged from 2,800 to 7,900 pCi/L
(picoCuries per liter), radium-226 concentrations ranged from 4.4 to 15.3 pCi/L, and radium-228 concentrations
ranged from 62.8 to 155 pCi/L. Radon-222 concentrations in the ground-water samples correlate directly with
concentrations of radium-226, radium-228, and ammonia and inversely with pH and the concentration of sulfate. The
seasonal fluctuation of radon-222 concentrations in a water-table aquifer suggests that seasonal sampling of ground
water may be necessary to characterize radon-222 concentrations in ground water.
-------
XII-3
RADON IN TAP WATER FROM DRILLED WELLS IN NORWAY
by: Bj0rn Lind and Terje Strand
National Institute of Radiation Hygiene
P.O.Box 55, N-1345 0steras, Norway
ABSTRACT
The results of measurements of radon concentrations in samples
of tap water from deep drilled wells (70-100 m) from different
parts of Norway are reported. The measurements were performed by
liquid scintillation counting. The radon concen-trations were
generally found to be low (less than about 500 kBq/m3) and for
most of the houses the most important radon source is the bedrock
and subsoil. The highest concentrations in tap water (up to about
7 MBq/m3) were found in typical granite areas in the south-
eastern part of the country. In one of these areas, more
extensive measurements were undertaken with the aim to study the
time variations in the source term and influencing factors.
INTRODUCTION
Degassing of radon from tap water has been found to lead to
elevated indoor concentrations of radon (222Rn) in several
countries (1-4). High concentrations of radon in tap water are
usually associated with deep drilled wells in radium rich
granites. In Finland, concentrations of radon in tap water of 77
MBq/m3 have been reported (1) . In Denmark and Sweden,
concentrations of radon up to 1.1 MBq/m3 (5) and 9 MBq/m3 (6),
respectively, have been found. In this paper concentrations up
to 7 MBq/m3 are are reported.
There are, however, a number of other naturally occurring
radionuclides in ground water, i.e. radionuclides of uranium and
radium, but they are dosimetrically of much less importance than
radon (7). Except for radium (226Ra), these nuclides will not be
discussed in this paper.
Many variables must be considered when assessing the
proportion of airborne radon that is derived from a water source.
-------
In addition to the concentration of radon in the tap water, these
variables include: water-usage rates, household air volume, air
exchange rate, type of water use and degassing efficiency. Models
have been developed (8,9) to estimate an average value for air-
to-water concentration ratio defined by:
f = - (1)
where Ca is the average radon concentration in indoor air
attributable to water-use and Cw is the radon concentration of
the tap water. Most published values of this ratio are between
10'5 to 10'4 (8,9) .
Most of the households in Norway have their water supplies
from surface sources with low concentrations of radon. It is
estimated that less than 10% of the Norwegian population have
their supply of household water from drilled wells. However,
because of polution of surface water (lakes and rivers) it is an
increasing trend to use ground water as a source of tap water.
This increase of the number of households using ground water,
could bring water to be a more significant source of indoor radon
in the future than at present.
MATERIALS AND METHODS
The measurements of radon were performed by liquid
scintillation counting. The water samples were prepared at the
site by drawing 10 ml of water from the tap into a glass syringe
and then transfered to a standard 20 ml glass vial prefilled with
10 ml of Lumagel; a toluene based scintillant solution. The vials
were then immediately sealed. In order to get fresh water from
the well, the tap was turned on for about 10 minutes before the
water samples were taken. The vials were counted in a LKB Wallac
1215 scintillation counter at the National Institute of Radiation
Hygiene 'in Norway. Calibration of the method were carried out
using a standard solution of radium. The overall uncertainty in
the measurements is estimated to be about 20% at the 95% confi-
dence level, and the lower limit of detection to be 1 kBq/m3.
The concentration of radon in the water samples was
measured by a method described by Cooper et.al. (10). From a one
liter sample the radium was concentrated and precipitated with
lead nitrate and barium sulphate, then redissolved in a standard
liquid scintillation vial in 7 ml of alkaline EDTA, and then
finaly, 15 ml of a toluene based scintillant solution (Lumax) was
added. The vials were then sealed and set aside (3-4 weeks) to
allow for radon ingrowth from radium prior to counting. The
-------
method was calibrated using a standard solution of radium. The
lower limit of detection was estimated to be about 10 Bq/m3 and
the overall uncertainty, including the chemical procedures, to
be less than 20% at the 95% confidence level.
The continuous measurements of radon in indoor air were
performed by a portable pulse-counting ionization chamber
developed by Balzer et.al. (11) and manufactured by Gammadata
MStteknik AB in Sweden. The instrumet was calibrated both at the
Swedish Radiation Protection Institute in Stockholm and at our
calibration facility in Norway.
RESULTS AND DISCUSSION
In our study, measurements of radon concentrations in tap
water were made on samples from a total of 229 deep drilled
wells. The results of the measurements are shown in Table 1. The
wells have been separated into three categories according to
geology of the areas - granite, alum-shale or various geological
formations. The table shows that the concentrations of radon in
tap water are more than an order of magnitude higher in wells
from granitic areas than alum-shale areas. Alum-shales are
generally known to have higher concentrations of radium and
higher emantation ratious of radon than most granites (12) .
However, alum-shales are usually located in the upper layers of
the ground while the drilled wells are very deep, usually between
70 and 100 m, and therefore they are most probably into limestone
with generally low concentrations of radium. The average concen-
trations of radon in the category "various geological formations"
is assumed to be a representative average for drilled wells in
Norway because these wells are located in different geological
formations and they are from different places of the country.
Table 1. Concentrations of 222Rn in tap water.
Geology
Number of
wells
222Rn concentration in
tap water (Bq/m3)
Average Range
Granitic areas 52
Alum shale areas 92
Various geological
formations 85
1070 130-7000
70 <5 - 470
250
<5 -1250
-------
In Table 2, the results of the radium measurements are
shown. The highest concentrations were found in samples from
drilled wells in typical granitic areas. In one of these wells
a radium concentration of 3.6 kBq/m3 were found. Assuming an
annual intake of 300 1 of water, and using published conversion
factors (13) , it is estimated that the annual effective dose is
about 0.3 mSv. In our study, only 48 samples were analysed.
However, it is assumed that the average concentration of radium
in tap water from drilled wells in Norway is close to the very
average for the category "varies geological formations". The
annual effective dose is thereby estimated to be less than 20
/uSv. This is almost negligible compared with other sources of
naturally occuring radiation.
Table 2. Concentrations of 226Ra in samples of tap water.
226Ra concentration in
Geology Number of tap water (kBq/m3)
wells Average Range
Granitic areas 9 0.84 0.17-3.6
Various geological
formations 39 0.16 <0.1 -1.6
In two of the houses, more extensive measurements were
made. These two houses, designated A and B in the further
discussion, are assumed to be very typical norwegian single
family houses. The number of persons living in each of the houses
are 3 and 2, respectively.
The radon concentration of the tap water for house A and
B was found to 4.3 and 1.7 MBq/m3, respectively. The water-use,
and the type of water-use (shower, dishwasher, washing machine,
etc.), was recorded as accurately as possible during the measure-
ment period. Both continuous and passive measurements of radon
in air were performed in different rooms. The results of the
measurements are shown i Figure 1 and 2. As illustrated the
variation of the radon concentration are very closely related to
the water use. Based on the continuous measurements the average
radon concentrations in air for the two houses were found to be
250 Bq/m3 and 200 Bq/m3, respectively. From passive measurements
by nuclear track detectors (3 months integration time), the
annual average radon concentrations for the two houses was esti-
mated to 300 and 190 Bq/m3, respectively. The contribution from
other sources of radon (mainly the buidling ground) were
-------
1200-
LJ
L.J-
*l
-too 5
03-29 03-30 03-31 04-Ot 04-02 04-03
Tim*
Figure 1. Radon concentration in air and water-use versus
time for house A.
03-24
03-23 03-2S
Tme
100
50
Figure 2. Radon concentration in air and water use versus
time for house B.
-------
estimated the be 100 Bq/m3 and 65 Bq/m3, respectively. These
estimates were based on continuous measurements in the two houses
during a three day period when no water were released into the
air.
The average air-to-water ratios for house A and B were
experimentally found to 4.7 10"5 and 7.3 10"5, respectively.
These air-to-water ratios are in accordance with most other
published values (8,9).
According to the families living in the two houses, the
water-use was fairly normal during the measurement period. Based
on the data in figure 1 and 2, the annual water-use for house A
and B was estimated to be 98 000 liters and 65 000 liters,
respectively. This corresponds to 33 000 liters per persons pr.
year, and this is about half the water-use estimated by Nazaroff
et.al.(8) for people in the USA.
The effective house volumes of the two houses were calcu-
lated to 160 m3 (A) and 130 m3 (B). The effective house volume
is calculated based on information on the occupancy in different
rooms and transport of air from rooms where water are released.
Details about this calculations will not be discussed in this
paper.
The transfer efficiency from water to air depends on type
of source and the temperature of the water. However, according
to Nazaroff et.al. (8), the variability of the transfer
efficiency is likely to be a small factor in the overall
variability of the air-to-water ratio. Most published values are
in the range 0.44 - 0.68. According to the recorded data on water
usage for the two houses, dishwashers, showers and washing
mashines are the most important sources. Owing to the very high
transfer efficiency of these sources (8) , an average value of
0.65 were adopted in our discussion.
From decay measurements of radon and tracer gass
measurements using SF6 the average air exchange rates for house
A and B were estimated to 0.56 h"1 and 0.75 h" , respectively.
Based on the singel-cell modell (8) the air-to-water ratio
were then calculated by:
W e
f = (2)
V Xv
where W is the water-use pr. hour (m3 h"1) , e is the use-weighted
average transfer efficiency of radon from water to air, V is
the effective volume of the residence (m3), and Av is the average
air exchange rate of the residence (h"1)
-------
From these calculations the air-to-water ratios for house
A and B were found to 4.0 10"5 and 6.3 10"5, respectively. These
calculated values are in very good agreement with the
experimental results of 4.7 10"5 and 7.3 10"5.
CONCLUSIONS
From the results and discussion in this paper the following
conclusions may be drawn:
1. Radon in household water is not a very important source of
indoor radon in Norway. Most norwegians (about 90%) have
their water supply from surface water with very low radon
concentrations (< 2 kBq/m3). The concentration of radon in
tap water from deep drilled wells, on the other hand, are two
to three order of magnitude higher. The average radon
concentration tap water from drilled wells in Norway is
estimated to 250 kBq/m3. Using the air-to-water ratios found
in our study, this would give an annual contribution to the
average radon concentration in domestic air of 10 - 20 Bq/m3.
The contribution to household air, if surface water is the
main source, is estimated to be less than 0.1 Bq/m3.
2. The air-to-water ratios for the two houses in our study was
found to be in accordance with other published values (8).
Air-to-water ratios were experimentally found to be 4.7 10"5
and 7.3 10"5 for the two houses. These values are very close
to the calculated ratios of 4.0 10"5 and 6.3 10"5 from
equation 2.
3. The concentrations of radium (226Ra) in tap water are very low
and the average annual effective dose is estimated to be less
than 20 /iSv.
REFERENCES
1. Asikainen, M. and Kahlos, H. Anomalously high concentrations
of uranium, radium in water from drilled wells in the Helsini
region. Geochim. Cosmochim. Acta.43: 1681-1686, 1979.
2. Asikainen, M. and Kahlos, M. Natural radioactivity of
drinking water in Finland. Health Phys.39: 77-83, 1980.
3. McGregor, R.G. and Gourgon, L.A., Radon and radon daughters
in homes utilizing deep well water supplies, Halifax County,
Nova Scotia, J.Environ.Sci.Health A15: 25-35, 1980.
-------
4. Hess, C.T., Casparius, R.E., Norton, S.A. and Brutsaert, W.F.
Investigation of natural levels of Rn-222 in groundwater in
Maine for assessment of related health effects. Natural
Radiation Environment III, Houston, Texas, U.S.DOE Conf.
780422, 1:529-547, 1978.
5. Ulbak, K. and Klinder, O. Radium and radon in danish
drinking water. Radiat.Prot.Dosim.7 (1-4): 87-89, 1984.
6. Kulich, J., M0re, H. and Swedjemark, G.A., Radon og radium
i hushallsvatten. Swedish Radiation Protection Institute,
Report 88-11, 1988.
7. Salonen, L. Natural radionuclides in ground water in
Finland. Radiat.Prot.Dosim.24 (1-4): 163-166, 1988
8. Nazaroff, W.W., Doyle, S.M., Nero, A.V., Jr., and Sextro,
R.G. Radon entry via portable water. In: Nazaroff, W.W. and
Nero, A.V.,Jr., eds. Radon and its decay products in indoor
air. Wiley Interscience, New York, USA, 1988: 131-157.
9. Hess, C.T., Vietti, M.A., Lachapelle, E.B., and Guillemette,
J.F. Radon trensfered from drinking water into house air.
In: Cothern, R.C. and Rebers, P.A., eds. Radon, radium and
uranium in drinking water. Lewis Publishers, Michigan, USA,
1990: 51-67.
10. Cooper, M.B. and Wilks, M.J. An analytical method for radium
(Ra-226) in environmental samples by the use of liquid
scintillation counting. Australian Radiation Laboratory,
Report of August 1991, ISSN 0157-400.
11. Balzer, P., G0rsten, K.G. and Backlin, A. A pulse-counting
ionization chamber for measuring the radon concentration in
air. Gammadata Sweden, February 1990.
12. Stranden, E. and Strand, T. Radon in an alum shale rich
norwegian area. Radiat.Prot.Dosim.24 (1-4): 367-370, 1988.
13. Kendall, G.M., Kennedy, B.W., Adams, N. and Fell, T.P.
Effective dose per unit of radionuclides by adults and young
people. Radiat.Prot.Dosim.16 (4): 307-312, 1986.
-------
XII-4
A RAPID ON-SITE DETECTOR OF RADON IN WATER
L. Grodzins, Professor of Physics
Massachusetts Institute of Technology
Founder and Chairman of the Board,
NITON Corporation, Bedford MA 01730
Stephen Shefsky, Scientist
NITON Corporation
ABSTRACT
The RAD-H2Ois a portable instrument that has been developed to measure radon in
water at the site of collection. The RAD-H2O combines a special water aeration unit with
the RAD7, NITON's electronic radon-gas detector. The device takes full advantage of
the features of the RAD7, including its computer-controlled strong air pump, on-line
humidity meter, and its ability to measure radon gas on the basis of the 3-minute 218po
alone. The measuring procedure is simple, fast and accurate over a range of radon concen-
trations from under 100 picoCuries of radon per liter of water to over 1,000,000 pCi/L,
providing readings in less than 30 minutes after a sample has been obtained, including a
paper record output complete with the alpha particle energy spectrum. The instrument is
described and results from testing private wells and public water systems are presented to
show the accuracy (with respect to EPA standards), precision, rapidity of measurement and
speed of recovery after measuring a high concentration of radon in water.
-------
INTRODUCTION
Radon is found in measurable concentrations in the water from all wells, with variations
that range over a factor of 100,000; that in public wells are typically between 50 and 1,000
picoCuries per liter of water, (pCi/L) while those in private wells are typically in the range
of 500 to 20,000 pCi/L. The highest concentration measured by NITON Corporation,
using standard liquid scintillation techniques, was 2,800,000 pCi/L from a well in New
Hampshire.
Elevated concentrations of radon in the water are considered a health hazard by the
Environmental Protection Agency which is proposing to set the maximum permissible
concentration at 300 pCi/L for public water systems serving at least 25 people.(1). When
that rule change takes effect in 1993, the 300 pCi/L is expected to be the de facto standard
for private wells as well.
Most private and small public wells in the United States are expected to have radon
concentrations in excess of that limit. In a survey of 2,000 wells in Maine, for example,
the mean radon concentration was 10,000 pCi/L, with the highest value being 1 million
pCi/L, and 64 public wells had a mean concentration of 2,000 pCi/L( 2 ). A survey in
North Carolina of 225 wells found a mean concentration of 1,400 pCi/L (3). Some areas
have much higher radon levels: a survey of 366 wells in Finland found a mean concen-
tration of 55,000 pCi/L (4).
NITON's results for 2,147 private wells, obtained mainly in 1991, are shown in the
related Figures 1 and 2. Figure 1 shows the number of wells versus radon concentration.
Figure 2 shows the percentage of wells versus radon concentration.
-------
Radon in Water from Private Wells
800
700
600
=
"a!
« 500
CO
O 400
41
.0
s
2 300
200
100
< 300
> 300
1,000 2 5,000 ± 10,000 S 25,000 £ 50,000 2 100,000 2 250,000
Radon in Water, pCi/L
Figure 1. Number of wells versus radon concentration.
-------
Radon in Water from 2,147 Private Wells
OJD
100
1
SO
i
I
;:
< 300 > 300 > 1,000
5,000
10,000 > 25,000 2 50,000 2 100,000 > 250,000
Radon Concentration in Water, pCi/L
Figure 2. Percentage of wells versus radon concentration.
The New England states -- which account for the preponderance of the data -- have
similar distributions to those in Figures 1 and 2. 90% of the wells in New England have
radon concentrations exceeding 300 pCi/L, 70% arc above 1,000 pCi/L and 20% arc
above 10,000 pCi/L. Of the private wells in New Hampshire, tested by NITON, 1% had
concentrations above 250,000 pCi/L. Even states with much lower radon-in-watcr
concentrations, such as New Jersey, have a majority of wells exceeding the proposed EPA
guidelines. There is a strong need for an inexpensive device for accurately measuring the
radon in water at concentrations from below 100 pCi/L to above 1 million pCi/L.
There are two laboratory methods approved by the EPA for measuring radon in
water (5). The most common method uses a liquid scintillation (LS) counter technique in
which the radon-bearing water is mixed with an appropriate liquid scintillation cocktail that
contains a scintillant and a desorbant, such as xylenc, to pull the radon from the water into
-------
the scintillant. The vial, containing the water and the cocktail, is then placed in the liquid
scintillation counter where the radon concentration is determined from the counting rates
of the alpha and beta particles emitted by the radon and its daughters.
In the second approved technique, the radon from the water is stripped by an appro-
priate aeration method and the radon-bearing air is pulled into an evacuated Lucas cell. The
alpha rays from the radon and its polonium decay products are detected by the light that is
emitted when the rays strike the scintillation phosphor that covers the inner surface of the
Lucas cell. The analysis in both methods is done in a laboratory where the necessary
equipment exists. (The latter method has been adapted for a field instrument by Pylon
Corporation; the Lucas cell is evacuated by a hand pump.) Both methods are sensitive over
the full range of radon concentrations encountered in public and private wells.
The LS method, which NITON Corporation uses to measure the radon in water
samples sent to its laboratory, is particularly well suited to commercial processing since it
uses sophisticated liquid scintillation counters that run in computer-controlled mode to
automatically process many samples a day. The disadvantage of the LS method is mainly in
the total turn-around time from collecting the sample to receipt of the laboratory results, a
process that typically takes from 2 to 5 days, depending on the mail or other means of
returning the samples to the lab. Professional testers and mitigators often want the results
sooner. Indeed, many would like to be able to measure the radon in the water at the site
itself. We here describe a portable instrument, which we call the RAD-fyO for which
patents have been applied, that yields rapid, inexpensive measurements of radon in water at
the collection site.
-------
DESCRIPTION OF THE RAD-H2O METHOD
The essential idea of the RAD-fyO method is to desorb the radon from the sample
of water into the air of a closed-loop, gas-tight circulation line that contains the desorbing
mechanism, the water sample, an air pump and a detector of radon gas. The desorption
results in a partition of the radon concentration between the fixed quantity of water and
the fixed quantity of air, with more than 95% of the radon in the air of the closed loop.
After partition equilibration, the measurement of the radon in the air gives a direct
measurement of the radon in the water.
The RAD-H2O makes use of the RAD7, a detector of radon-in-air manufactured by
NITON Corporation, that has an internal, computer-controlled pump. It has the special
capability of determining the radon concentration from the count rate of 21^Po, the
daughter of radon that decays with a 3-minute half-life with the emission of a 6.0 MeV
alpha particle.
Figures 3 and 4 show the principal elements of the RAD-H2O . The radon-bearing
water is collected in the approved manner in a standard container shown in the left half of
Figure 3. The nominal collector is the 40 cc "EPA collection vial" with a cap sealed by a
Teflon/rubber disc, but 10 cc and 20 cc vials are more practical for concentrations above
about 50,000 pCi/L, while 250 cc bottles are especially useful for radon concentrations
below about 300 pCi/L. The collection vial is filled to the brim in the appropriate manner
so as to minimize the loss of radon from the water. The tester can place the full vial directly
into the RAD-fyO for an immediate measurement, or can seal the vial with the standard
cap for measurement at a later time.
The aeration unit is shown schematically in the right panel of Figure 3. The aeration
unit is designed so that standard collection vials can be attached with a gas tight seal. When
the aeration unit is attached to the collection vial, the fritted diffuser at the end of the insert
tube extends several inches into the collection vial.
-------
From Radon
Gas Detector
J
To Radon Gas
Detector
Standard collection vial
with teflon seal
Aeration unit for a standard
collection vial.
Figure 3. Schematic views of collection vial and aeration unit.
Figure 4 shows the aeration unit connected to the RAD7. The fritted-end tube of the
aeration unit is connected to the output (push) port of the HAD/V while the air chamber
above the water sample is connected to the input (pull) port of the RAD7. A 25 cc
desiccant tube is inserted into the input line to maintain a reasonable low relative humidity
in the test chamber of the RAD7. The entire air circulation line is gas tight.
-------
Aeration
Unit
Figure 4. The connection of the aeration unit to the RAD7
The coupling of the collection vial to the aeration unit results in some water being
displaced but that does not affect the concentration of radon in the water that remains. If
the coupling is done in less than a minute and it generally takes about 10 seconds
there is negligible loss of radon concentration from the water sample.
The .RAD7is turned on and the measurement of the concentration of the radon in
water begins. The air pump forces the air in the closed loop through the submerged
aeration tube, through the radon-bearing water and into the air chamber, from which it is
pulled through the desiccant tube into the .RAD/test chamber and then back to the
aeration tube. The air is continuously circulated until the radon in the water volume and
the radon in all parts of the air line, including the air chambers in the RAD7, are in
partition equilibrium.
The time for equilibration depends on the aeration configuration, pump capacity,
collection volume, etc. The 40 cc vial reaches partition equilibration in less than 2 minutes;
the 250 cc vial reaches partition equilibrium in about 6 minutes using the same pump and
aerator. Using the known volume partition function of 4:1 for radon partitioning between
air and water, calculations predict and experiments confirm that, at partition equilibration,
99% of the total radon is in the air and only 1% is left in the water of a 40 cc vial; if a 250 cc
collection vial is used, the air in the closed loop would still contain about 95% of the total
-------
radon. The pump can be kept on after equilibration has been reached but that uses up the
desiccant and puts an unnecessary burden on the power supply of the portable system.
The radon has a half-life of 3.8 days. When the radon atoms decay in the test chamber
of the RAD7, emitting a 5.49 MeV alpha particle, the resulting 218Po atom is typically a
positively charged ion. A positive high voltage on the test chamber wall pushes the 218Po
ions to the surface of the alpha particle detector where they stick. The efficiency of the
surface barrier detector is about 50% for counting the 6 MeV alpha particles that decay with
a half-life of 3 minutes. Since the energy resolution of the RAD7 detection system easily
distinguishes the 218Po decay from other alpha emitters, the RAD-fyO system has an
effective resolving time of about 3 minutes.
This short time constant is important for two independent reasons. First, the count rate
on the detector of the RAD-H2O grows with a 3-minute half-life, reaching 94% of its
radioactive equilibrium in just 12 minutes. Second, after completing the measurement, the
RAD-HzO can be purged with a half-life of 3 minutes. In both cases, one need not
contend with the longer-lived decay products that can incapacitate a detector for many
hours.
The time for a measurement of the radon in the water sample depends on the concen-
tration of the radon in the water and the efficiency of the radon-in-air detector. The
RAD/has a sensitivity of 0.25 counts of 6 MeV alpha particles per minute, per pCi per
liter of radon in air. The sensitivity per pCi of radon in the water sample is also 0.25 counts
of 6 MeV alpha particle per minute, since the transfer of radon from the water to the air is
at least 95% efficient, and the volume of the air in the closed loop of the RAD7h about
one liter.
The factor to convert counts per minute to pCi/L of water depends on the collection
volume. One count per minute corresponds to 100 pCi/L in a 40 cc collection vial and to
16 pCi/L in a 250 cc vial. We emphasize that a fully purged RAD7, with its excellent
energy resolution and spectrum analysis, has negligible background; much less than one
count per hour in the 6 MeV window.
Example 1: 300 pCi/L of radon in water. This is the proposed EPA maximum for
public drinking water. A 40 cc collection vial will contain (40/1000) x 300 = 12 pCi,
which will lead to a count rate of 12 x 0.25 « 3 counts per minute. In 34 minutes of
counting, the radon concentration in the water vial will be known to within a standard
deviation of 10%. If a 250 cc collection vial is used, the count rate will be 18 cpm, and the
same accuracy can be reached in just 6 minutes of counting time. Thus a full measurement
can be made in 27 minutes: 5 minutes to collect the sample, 6 minutes for partition
equilibrium, 10 minutes for radioactive equilibrium, and 6 minutes for counting. We know
of no system, even under laboratory conditions, that can measure such a low concentration
in a shorter time.
Example 2: 100 pCi/L of radon in water. This is the standard that one would like to
meet to ensure a reliable measurement at 300 pCi/L. To measure a concentration of 100
-------
pCi/L with a 40 cc collection vial requires close to 2 hours of counting. However, if one
uses a 250 cc vial, the count rate at equilibrium will be about 6 counts per minute. This
means that in less than 18 minutes, one will have a reliable measure of even the lowest
concentration of practical interest.
Example 3: 20,000 pCi/L. Private wells typically have radon concentrations exceeding
1,000 pCi/L. The State of Maine surveys (2), after concluding that a significant percentage
of the wells in the State had concentrations exceeding 20,000 pCi/L, set a guide line for
mitigation at 20,000 pCi/L, suggesting that wells with concentrations below that value
need not be mitigated. A concentration of 20,000 pCi/L, collected in a 40 cc vial, results
in 200 counts per minute in the J!lAD7after equilibrium has been reached. The time
needed for the measurement itself is small compared to the time needed for equilibration.
We note that, in principle, one can reduce that measurement time by taking advantage of
the reproducible time-dependence of the equilibration process. In practice, however, there
is probably no need to do so.
Example 4: 1,000,000 pCi/L. The highest concentration of radon in water that we
can confirm was about 3,000,000 pCi/L, measured from a well in New Hampshire in the
laboratory of NITON Corporation, using liquid scintillation techniques. Concentrations
above 1 million pCi/L are rare but, when they occur, they can present problems since
radon-gas detectors are not generally designed to handle the high count rates. Further,
once high levels of radon are in the detector chamber, they are not easily purged for the
next test.
Obvious ways of handing the problem are: use a much smaller collection vial; let the
radon decay before measuring (a solution that could take weeks of waiting); dilute the
concentration with radon-free public water or distilled water; or make use of the partition
function to reduce the concentration of the radon in the water.
The partition function for radon in air to water is 4:1. Thus if we seal a collection vial
half full of the radon-bearing water, the radon concentration in the water at partition
equilibrium will be one-fifth the original concentration. The radon in the air volume of the
container will quickly escape when the container is uncapped; the radon in the water,
however, takes many minutes to diffuse out of the water if the container is not shaken.
If the 40 cc collection vial is filled with only 5 cc of water then, after equilibration with
the air of the vial, the radon concentration in the 5 cc will be l/29th the original
concentration. Thus a concentration of 1 million pCi/L will produce a count rate of
35,000 pCi/L. We note that if this method is to be used, one must know the temperature
of the container prior at equilibration since the partition function is temperature
dependent.
-------
The sequence of steps to using the .RAD-H^O are as follows:
1. Take a 5-minute background reading and verify that the count rate
is <0.5 per minute.
2. Collect samples of the water in the approved manner to minimize
the escape of radon. In the absence of any knowledge about the
concentration of the radon in the water, we recommend that a
10 cc, 40 cc and 250 cc vial be filled.
3. Connect the aeration unit via hoses and desiccant to the input and
output of the RAD7.
4. Attach the 10 cc water sample to the aeration unit and seal the top
to make an air-tight circulation loop.
5. Press start on the RAD7, which is the .RAD-H-jO protocol mode.
The internal computer will then automatically carry out the
next set of steps.
6. The aeration pump will be on for 3 minutes and will then shut off.
7. The count rates of the 6 MeV alpha particle will be determined for
successive intervals of 6 minutes, 6 minutes and T, where T is
10 minutes or the time to obtain 100 counts, whichever is
longer. The radon concentration in the water will be inferred
from the last count rate. The data can be ported to a computer
via an RS232 port or can be outputted directly to the attached
graphic printer to give all relevant information, including a
copy of the alpha particle spectrum.
8. If the count rate is too low for an accurate measurement in a
20-minute count, the computer will notify you by a steady
beeping, advising you to change to the appropriately larger
water sample for the remainder of the test. The .RAD/will be
restarted with a 3 minute aeration.
9. The RAD7 should be purged of radon after the measurement with
a programmed procedure lasting at least 10 minutes.
10. A new measurement can begin as soon as the background level
has been reduced sufficiently to make an accurate measurement
of the next sample. That time is typically 10 to 30 minutes.
-------
RESULTS OF MEASUREMENTS OF PRIVATE WELLS
In this section we present some of the results from the continuing development of the
RAD-H2O to optimize the procedures so as to obtain the optimum accuracy, precision,
convenience and speed.
Most of the tests have been done following a simple procedure in which comparisons
are made, on the same sample of water, between the RAD-^O and NITON's standard
liquid scintillation methods that are used daily to process tests of water samples obtained
from public and private well systems. The LS results are taken as the benchmark; NITON
is part of the EPA Radon-in-Water Collaborative Study conducted by the Environmental
Monitoring System Laboratory in Las Vegas, so that our LS methods are periodically
checked against EPA standards and are accurate to 10%.
The comparison procedure is illustrated with the example of a 40 cc VOC collection vial
with a Teflon/rubber seal. The NITON laboratory first measures this sample by its
standard procedure of drawing 10 cc of the water from the vial into a syringe containing
5 cc of scintillation/eluting cocktail. The 15 cc mixture is slowly emptied into bottom of
a glass LS vial that already contains 5 cc of scintillation/eluting cocktail. The LS vial is
capped and the 20 cc shaken prior to placement in the LS counter.
Counting begins about 4 hours later, after radioactive equilibrium has been obtained.
Within a few seconds after the 10 cc sample has been drawn from the 40 cc collection vial,
10 cc of distilled water is added to that vial to fill it completely. Tests indicate that the
40 cc collection vial now contains 75% of the radon strength that it had originally;
essentially no radon gets lost in the extraction-refilling process.
The diluted collection vial is then treated according to the protocol described above.
We have verified that -99% of the radon is extracted from the water sample by the 3-minute
aeration procedure by comparing the radon concentration in the vial before and after
aeration.
Tests have been made over the range from 50 pCi/L to 110,000 pCi/L using the 40 cc
collection vial. Figure 5 shows the correlation between the counting rates obtained with
the LS method to that obtained with the RAD-HzO. Decay and background corrections
have been made and the -RAD-fljO data have been adjusted for the dilution factor. There
is considerable scatter to the data but the linear relation has an acceptable correlation.
-------
1000
g 800
<*.
a
e
f
O «OQ
I
200
Comparison of Count Rates in LS and RADH2O
-48075 0.091972X R- 0.9913")
2000
4000
6000
8000
1.000 10'
Counts/min in LS counter (10 ml of water)
Figure 5. Correlation between LS and RAD-H2O counting rates
Figure 6 shows the ratio R of LS counts to RAD-HzO counts, plotted versus the radon
concentration in the water as determined by the LS counter. There is a substantial scatter in
the ratio, especially at low radon concentrations where the uncertainties of both techniques
are large.
-------
.2
'>
a
I
Q
C/J
-------
SUMMARY AND CONCLUSION
A new technique has been developed for measuring radon in water at the test site. The
method makes use of the fact that aeration of a water sample into a closed loop of air that
contains a radon gas meter can transfer more than 95% of the radon in the water to the air
where the meter can measure the concentration by well-established techniques. The
RAD-H2O method is simple and reliable. It is also rapid and partially automated. And it
is accurate over the full range of encountered radon-in-water concentrations.
The work described in this paper was not funded by thre
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) EPA Office of Water Fact Sheet, WH-550, National Primary Drinking Water
Regulations for Radionuclides, June, 1991
2) C.T. Hess, et al. 1979. "Radon-222 in Potable Water Supplies in Maine" Land and
Water Resources Center, University of Maine at Orono.
3) M.K. Sasser and J.E. Watson, "An evaluation of the Radon Concentration in North
Carolina Ground Water Supplies", Health Physics 34, 667 (1978)
4) M. Asikainen and H. Kahlos, "Anomalously High Concentrations of Uranium, Radium,
and Radon in Water from Drilled Wells in the Helsinki Region", Geochimica et
Cosmochimica Acta, 43, 1981 (1979)
5) EPA/600/2-87/082, March, 1989. Two Test Procedures for Radon in
Drinking Water.
'< U.S. GOVERNMENT PRINTING OFFICE: 1992-648-003/60,022
------- |