EPA-600/9-91-037b
November 1991
PROCEEDINGS: THE 1991 INTERNATIONAL SYMPOSIUM ON RADON
AND RADON REDUCTION TECHNOLOGY
Volume 2: Symposium Oral Papers
Technical Sessions VI through X
Symposium Co-chairpersons:
Timothy M. Dyess
U.S. EPA
Air and Energy Engineering Research Laboratory
Susan M. Conrath
U.S. EPA
Office of Radiation Programs
Charles M. Hardin
Director
Conference of Radiation Control Program Directors
(CRCPD), Inc.
S. Cohen and Associates
1311 Dolley Madison Blvd.
McLean, VA 22101
EPA Contract 68-D0-0097
Work Assignment 1
EPA Task Officer:
Timothy M. Dyess
U.S. Environmental Protection Agency
Air and Energy Engineering Research Laboratory
Research Triangle Park, NC 27711
Prepared for:
U.S. Environmental Protection Agency
U.S. Environmental Protection Agency Office of Air and Radiation
Office of Research and Development Washington, DC 20460
Washington, DC 20460
-------
TECHNICAL REPORT DATA
(Please read Inurucliam on the reverse before compt''
1. REPORT NO.
EPA-600/9-91-037b
4. Tt f LE AND SUBT 11 LE
Proceedings: The 1991 International Symposium on
Radon and Radon Reduction Technology; Vol.2. Sym-
posium Cral Papers, Sessions VI Through X
6. PERFORMING ORGANIZATION CODF
7. AUTHOR(S)
Miscellaneous
8. HEH FORMING ORGANIZATION RfcPOHl NO.
9. PERFORMING ORGANIZATION NAME AND ADDRESS
S. Cohen and Associates
1311 Dolley Madison Boulevard
Mcl.ean, Virginia 22101
12. SPONSORING AGFNCY NAMF AND ADDRFSS
EPA, Cffi.ce of Research and Development
Air and Energy Engineering Research Laboratory
Research Triangle Park, North Carolina 27711
. REPORT DATE
November 1991
10. PROGRAM ELEMENT NO.
11. CONTRACT/GRANT NO.
68-D0-0097, Task 1
13. TYPF Of- RFPORT AND PFRIOD COVFRED
Proceedings; 4-7/91
14. SPONSORING AGFNCY CODE
EPA/600/13
15.supplementary notes AEERL task officer is Timothy M. Dyess, Mail Drop 54, 919/541-
2802. Cosponsor EPA/CRP's project officer is Susan M. Conrath, Mail Code ANR-
464, Washington, DC 20460, 202/245-6736.
i6. ABsiHAci pr0ceedings, in four volumes, document the 1991 International Sympo-
sium on Radon and Radon Reduction Technology, held in Philadelphia, PA, April
2~5, 1991. In all, 65 oral papers (including the welcome address, the lead address,
and the keynote address), 14 panel session papers, and 40 poster papers were pre-
sented. The papers addressed a wide range of radon-related topiesi government pro-
grams and policies, health studies, health risk communication, measurement me-
thods, radon reduction methods in existing houses, radon transport and entry dyna-
mics, survey results, geological data, radon-resistant new construction methods,
and radon measurement and mitigation in schools and other large buildings. The
symposium speakers included EPA personnel, representatives from federal and
state environmental/health agencies, research and development groups, academic
and medical personnel, manufacturers of testing equipment, and those in the con-
struction and real estate industries. Attendees represented 14 countries other than
the U. S. The international papers provided updates on government policies, results
of surveys, and technological developments in radon md radon reduction technology.
17.
a.
DESCRIPTORS
KEY WORDS ANO DOCUMENT ANAL YSfS
b. IDENTIFI ERS/OPE N ENDED TERMS
Pollution
Radon
Measurement
Residential Buildings
Schools
Geology
Construction
Environmental Biology
Pollution Control
Stationary Sources
Health Effects
COSATI Field/Group
13B
07 B
14 G
13 M
051
08G
06F
18. DISTRIBUTION.' STATEMENT
Rclori.se to Public
10. SECURITY CLASS {This Report)
Unclassified
?1 NO. OF PAGES
429
20. SECURITY CLASS (This page)
Unclassified
27. PRICE
EPA Foitr. 2220-1 (9-73)
-------
NOTICE
This document has been reviewed in accordance with
U.S. Environmental Protection Agency policy and
approved for publication. Mention of trade names
or commercial products does not constitute endorse-
ment or recommendation for use.
-------
ABSTRACT
The 1991 International Symposium on Radon and Radon Reduction Technology
was jointly sponsored by the U.S. Environmental Protection Agency (EPA) Air and Energy
Engineering Research Laboratory (AEERL), EPA Office of Radiation Programs (ORP), and
the Conference of Radiation Control Program Directors (CRCPD), Inc. The symposium
was held in Philadelphia, Pennsylvania on April 2-5, 1991. The objective of the
symposium was to provide an international forum for the exchange of technical
information on radon and radon reduction technology in the indoor environment. Oral
papers and poster presentations conveyed recent advances in radon research and radon
reduction methods in the following fields: the assessment of radon derived health
impacts, government programs and policy, the measurement of radon and radon
progeny, soil/geology and radon source potential, and diagnostics and application of
radon reduction and radon resistant construction techniques. The Symposium
Proceedings, published in four volumes, include 65 oral papers (including the welcome
address, the lead address, and the keynote address), 14 panel session papers, and 40
poster session presentations.
i i i
-------
PREFACE
The 1991 International Symposium on Radon and Radon Reduction Technology
was held on April 2-5, 1991, in Philadelphia, Pennsylvania. Sponsored jointly by the U.S.
Environmental Protection Agency (EPA) Air and Energy Engineering Research Laboratory
(AEERL), EPA Office of Radiation Programs (ORP), and the Conference of Radiation
Control Program Directors (CRCPD), Inc., the symposium was devoted to the exchange
of current research developments in radon and radon reduction technology in the indoor
environment. Education in radon research and the dissemination of technical information
were the primary objectives of the meeting.
The co-chairpersons of the symposium were Timothy M. Dyess of EPA-AEERL,
Susan M. Conrath of EPA-ORP, and Charles M. Hardin of CRCPD. The opening address
was given by Charles M. Hardin, Executive Director of CRCPD, who briefed the audience
on the objectives of that organization. The welcome address was given by Edwin B.
Erickson, Administrator of EPA Region 3. The lead address, entitled "Comparative
Dosimetry of Radon in Mines and Homes: An overview of the NAS Report", was
presented by Jonathan M. Samet of the New Mexico Tumor Registry at the University of
New Mexico. This paper addressed the results of recent risk estimate calculations. The
keynote address, given by John R. Garrison, Managing Director of the American Lung
Association, focused on ways to promote public respect for the health hazard which
indoor radon presents.
In all, 62 oral papers, 14 panel session papers, and 40 poster papers were
presented. The papers addressed a wide range of radon topics: government programs
and policies, health studies, health risk communication, measurement methods, radon
reduction methods in existing houses, radon transport and entry dynamics, survey results,
geological data, radon resistant new construction methods, and radon measurement and
mitigation in schools and large buildings.
The symposium speakers included EPA personnel, as well as representatives from
federal and state environmental/health agencies, research and development groups
conducting radon testing, academic and medical personnel, manufacturers of testing
equipment, and those in the construction and real estate industries. Attendees
represented 14 countries (other than the U.S.), most of which were European. The
international papers provided updates on government policies, results of surveys, and
technological developments in radon and radon reduction technology issues.
i v
-------
The symposium has been published in four volumes:
Volume 1: Symposium Oral Papers
Opening Session and Technical Sessions I through V
-- Session Opening: Lead and Keynote papers
- Session
- Session
-- Session
-- Session
-- Session
III
IV
V
Government Programs and Policies
Relating to Radon
Radon-Related Health Studies
Measurement Methods
Radon Reduction Methods
Radon Entry Dynamics
Volume 2: Symposium Oral Papers
Technical Sessions VI through X
-- Session
-- Session
-- Session
-- Session
-- Session
VI: Radon Surveys
VII: State Programs and Policies
Relating to Radon
VIII: Radon Prevention in New
Construction
IX: Radon Occurrence in the Natural
Environment
X: Radon in Schools and Large
Buildings
Volume 3: Panel Papers and Poster Papers
Technical Sessions I through V
-- Panel Session I: Risk Communication
-- Panel Session
II: Detection of Radon Measurement
Tampering
-- Panel Session III: Short-Term / Long-Term
Measurement
- Poster Papers for Sessions I, II, III, IV, and V
Volume 4: Poster Papers
Technical Sessions VI through X
-- Poster Papers for Sessions VI, VII, VIII, IX, and X
-------
The 1991 International Symposium on Radon
and Radon Reduction Technology
Volume 1: Oral Papers
Opening Session and Technical Sessions I through V
Table of Contents
Opening Session
The Conference of Radiation Control Program Directors, Inc.
Its Beginning, Role, and Operation
Charles M. Hardin, Executive Director, CRCPD 0-3
Comparative Dosimetry of Radon in Mines and Homes:
An Overview of the NAS Report
Jonathan M. Samet, New Mexico Tumor Registry, University of New Mexico 0-9
Keynote Address
John R. Garrison, Managing Director, American Lung Association 0-19
Session I: Government Programs and Policies Relating to Radon
The Need for a Coordinated International Assessment of the Radon Problem
Friedrich Steinhausler, International Atomic Energy Agency, Austria 1-3
European Radon Research Sponsored by the Commission of European Communities
Martial Olast, Jaak Sinnaeve, and Augustin Janssens, CEC, Belgium 1-21
The UK Radon Programme
Michael O'Riordan, National Radiological Protection Board, UK 1-29
The U.S. DOE Radon Research Program: A Different Viewpoint
Susan L. Rose, Office of Energy Research, U. S. DOE 1-41
Policies and Progress of EPA's Radon Action Program (Abstract)
Margo Oge, EPA - Office of Radiation Programs 1-53
Session II: Radon-Related Health Studies
Residential Radon Exposure and Lung Cancer in Women (Paper withdrawn)
Goran Pershagen, Karolinska Institute, Sweden 2-3
An Evaluation of Ecological Studies of Indoor Radon and Lung Cancer
Christine Stidley and Jonathan M. Samet, New Mexico Tumor Registry,
University of New Mexico 2-5
Review of Radon and Lung Cancer Risk (Abstract)
Jonathan M. Samet, University of New Mexico; and
Richard W. Hornung, NIOSH; 2-21
Lung Cancer in Rats Exposed to Radon/Radon Progeny (Abstract)
F. T. Cross and G. E. Dagle, Pacific Northwest Laboratory 2-23
vi i
-------
Startling Radon Risk Comparisons
JoAnne D. Martin, DMA-RADTECH, Inc 2-25
Estimating Radon Levels from Polonium-210 in Glass
J. Cornells, State University of Gent; Hans Vanmarcke, Nuclear Research Center, Belgium;
and C. Landsheere and A. Poffijn, State University of Gent, Belgium 2-41
Expanded and Upgraded Tests of the Linear-No Threshold Theory
for Radon-Induced Lung Cancer
Bernard L Cohen, University of Pittsburgh 2-49
Session III: Measurement Methods
Current Status of Glass as a Retrospective Radon Monitor
Richard Lively, Minnesota Geological Survey; and Daniel Steck,
St. John's University 3-3
Soil Gas Measurement Technologies
Harry E. Rector, GEOMET Technologies, Inc 3-13
Results From a Pilot Study to Compare Residential Radon Concentrations
with Occupant Exposures Using Personal Monitoring
B. R. Litt, New Jersey Department of Environmental Protection;
J. M. Waldman, UMDNJ; and N. H. Harley and P. Chittapom,
New York University Medical Center 3-29
Rapid Determination of the Radon Profile in a Structure by Measuring
Ions In the Ambient Atmosphere
W. G. Buckman and H. B. Steen III, Western Kentucky University 3-41
Intercomparison of Activity Size Distribution Measurements with Manual and
Automated Diffusion Batteries - Field Test
P. K. Hopke and P. Wasiolek, Clarkson University; E. 0. Knutson,
K W. Tu, and C. Gogolak, U. S. DOE; A. Cavallo and K. Gadsby,
Princeton University; and D. Van Cleef, EPA-NAREL 3-51
Influence of Radon Concentrations on the Relationship Among Radon
Measurements Within Dwellings (Abstract)
Judith B. Klotz, Janet B. Schoenberg, and Homer B. Wilcox,
NJ State Department of Health 3-69
The Use of Indoor Radon Measurements and Geological Data in Assessing
the Radon Risk of Soil and Rock in Construction Sites in Tampere
Anne Voutilainen and llona Makelainen, Finnish Centre for
Radiation and Nuclear Safety 3-71
Session IV: Radon Reduction Methods
Causes of Elevated Post-Mitigation Radon Concentrations in Basement Houses
Having Extremely High Pre-Mitigation Levels
D. Bruce Henschel, EPA - Office of Research and Development; and
Arthur G. Scott, AMERICAN ATCON, Inc. 4-3
v i i i
-------
A Measurement arid Visual Inspection Critique to Evaluate the Quality of
Sub-Slab Ventilation Systems
Richard W. Tucker, Gemini Research, Inc.; and Keith S. Fimian, Radonics, Inc 4-21
Pressure Field Extension Using a Pressure Washer
William P. Brodhead, WPB Enterprises 4-35
A Variable and Discontinuous Subslab Ventilation System and Its
Impact on Radon Mitigation
Willy V. Abeele, New Mexico Environmental Improvement Division 4-43
Natural Basement Ventilation as a Radon Mitigation Technique
A. Cavallo, K. Gadsby, and T.A. Reddy, Princeton University 4-61
Session V: Radon Entry Dynamics
A Modeling Examination of Parameters Affecting Radon and Soil Gas
Entry Into Florida-Style Slab-on-Grade Houses
R. G. Sextro, K. L. Revzan, and W. J. Fisk, Lawrence Berkeley Laboratory 5-3
Effect of Winds in Reducing Sub-Slab Radon Concentrations Under
Houses Laid Over Gravel Beds
P. C. Owczarski, D. J. Holford, K. W. Burk, H. D. Freeman, and
G. W. Gee, Pacific Northwest Laboratory 5-19
Radon Entry Into Dwellings Through Concrete Floors
K. K. Nielson and V. C. Rogers, Rogers and Associates
Engineering Corporation 5-29
Radon Dynamics in Swedish Dwellings: A Status Report
Lynn M. Hubbard, Nils Hagberg, Anita Enflo, and Gun Astri Swedjemark,
Swedish Radiation Protection Institute 5-41
Soil Gas and Radon Entry Potentials for Slab-on-Grade Houses
Bradley H. Turk, New Mexico; David Grumm, Yanxia Li, and Stephen
D. Schery, New Mexico Institute of Mining and Technology; and
D. Bruce Henschel, EPA - Office of Research and Development 5-53
Direct Measurement of the Dependence of Radon Flux Through
Structure Boundaries on Differential Pressure
D. T. Kendrick and G. Harold Langner, Jr., U.S. DOE/Chem-Nuclear
Geotech, Inc 5-69
Radon Resistance Under Pressure
William F. McKelvey, Versar, Inc.; and Jay W. Davis, Versar A/E, Inc 5-83
ixs
-------
The 1991 International Symposium on Radon
and Radon Reduction Technology
Volume 2: Oral Papers
Technical Sessions VI through X
Table of Contents
Session VI: Radon Surveys
Factors Associated with Home Radon Concentrations in Illinois
Thomas J. Bierma and Jennifer O'Neill, Illinois State University 6-3
Radon in Federal Buildings
Michael Boyd, EPA - Office of Radiation Programs; and
Terry Inge, S. Cohen & Associates 6-23
Radon in Switzerland
H. Surbeck and H. Volkle, University Perolles; and W. Zeller, Federal
Office of Public Health, Switzerland 6-31
A Cross-Sectional Survey of Indoor Radon Concentrations in 966 Housing
Units at the Canadian Forces Base in Winnipeg, Manitoba
D. A. Figley and J. T. Makohon, Saskatchewan Research Council 6-47
Radon Studies in British Columbia, Canada
D. R. Morley and B. G. Phillips, Ministry of Health; M. M. Ghomshei,
Orchard Geothermal Inc.; and C. Van Netten, The University of British Columbia 6-61
The State of Maine Schools Radon Project: Results
Lee Grodzins, NITON Corporation; T. Bradstreet,
Division of Safety and Environmental Services, Maine; and
E. Moreau, Department of Human Services, Maine 6-71
Radon in Belgium: The Current Situation and Plans for the Future
A. Poffijn, State University of Gent; J. M. Charlet, Polytechnical
Faculty; E. Cottens and S. Hallez, Ministry of Public Health;
H. Vanmarcke, Nuclear Research Center; and P. Wouters, BBRI, Belgium 6-87
A Radiological Study of the Greek Radon Spas
P. Kritidis, Institute of Nuclear Technology - Radiation Protection, Greece 6-95
Session VII: State Programs and Policies Relating to Radon
Washington State's Innovative Grant: Community Support Radon Action Team for Schools
Patricia A. McLachlan, Department of Health, Washington 7-3
Kentucky Innovative Grant: Radon in Schools Telecommunication Project
M. Jeana Phelps, Kentucky Cabinet for Human Resources; and
Carolyn Rude-Parkins, University of Louisville 7-13
x
-------
Regulation of Radon Professionals by States: the Connecticut Experience and Policy Issues
Alan J. Slniscalchi, Zygmunt F. Dembek, Nicholas Macelletti, Laurie
Gokey, and Paul Schur, Connecticut Department of Health Services;
Susan Nichols, Connecticut Department of Consumer Protection; and
Jessie Stratton, State Representative, Connecticut General Assembly 7-25
New Jersey's Radon Program - 1991
Jill A. Lapoti, New Jersey Department of Environmental Protection 7-37
Session VIII: Radon Prevention in New Construction
A Comparison of Indoor Radon Concentrations Between Preconstruction and
Post-Construction Mitigated Single Family Dwellings
James F. Burkhart, University of Colorado at Colorado Springs; and
Douglas L. Kladder, Residential Service Network, Inc 8-3
Radon Reduction in New Construction: Double-Barrier Approach
C. Kunz, New York State Department of Health 8-19
Radon Control - Towards a Systems Approach
R. M. Nuess and R. J. Prill, Washington State Energy Office 8-29
Mini Fan for SSD Radon Mitigation in New Construction
David W, Saum, INFILTEC 8-45
Building Radon Mitigation into Inaccessible Crawlspace - New Residential Construction
D. Bruce Harris and A. B. Craig, EPA - Office of Research and Development; and
Jerry Haynes, Hunt Building Corporation 8-57
The Effect of Subslab Aggregate Size on Pressure Field Extension
K. J. Gadsby, T. Agami Reddy, D. F. Anderson, and R. Gafgen,
Princeton University; and A. B. Craig, EPA - Office of Research and Development 8-65
Session IX: Radon Occurrence in the Natural Environment
Combining Mitigation and Geology: Indoor Radon Reduction by Accessing the Source
Stephen T. Hall, Radon Control Professionals, Inc. 9-3
Preliminary Radon Potential Map of the United States
L.C.S. Gundersen, R. R. Schumann, J. K. Otton, R. F. Dubiel, D. E. Owen,
and K. A. Dickenson, U. S. Geological Survey; and R. T. Peake and S. J. Wirth,
EPA - Office of Radiation Programs 9-13
Technological Enhancement of Radon Daughter Exposures Due to Non-nuclear Energy Activities
Jadranka Kovac, D. Cesar, and A. Bauman,
University of Zagreb, Yugoslavia 9-33
A Site Study of Soil Characteristics and Soil Gas Radon
Richard Lively, Minnesota Geological Survey; and Daniel Steck,
St. John's University 9-43
Geological Parameters in Radon Risk Assessment -
A Case History of Deliberate Exploration
Donald Carlisle and Haydar Azzouz, University of California at Los Angeles 9-59
xi
-------
Session X: Radon in Schools and Large Buildings
Seasonal Variation in Short-term and Long-term Radon Measurements in Schools
Anita L Schmidt, EPA - Office of Radiation Programs; and
John T. MacWaters and Harry Chmelynski, S. Cohen & Associates 10-3
Diagnostic Evaluations of Twenty-six U. S. Schools - EPA's School Evaluation Program
Gene Fisher and Bob Thompson, EPA - Office of Radiation Programs; Terry Brennan,
Camroden Associates; and William Turner, H. L. Turner Group 10-25
Extended Heating, Ventilating and Air Conditioning Diagnostics in Schools in Maine
Terry Brennan, Camroden Associates; Gene Fisher and
Robert Thompson, EPA - Office of Radiation Programs; and
William Turner, H. L. Turner Group 10-37
Mitigation Diagnostics: The Need for Understanding Both HVAC and Geologic Effects in Schools
Stephen T. Hall, Radon Control Professionals, Inc. 10-57
A Comparison of Radon Mitigation Options for Crawl Space School Buildings
Bobby E. Pyle, Southern Research Institute; and Kelly W. Leovic,
EPA - Office of Research and Development 10-73
HVAC System Complications and Controls for Radon Reduction in School Buildings
Kelly W. Leovic, D. Bruce Harris, and Timothy M. Dyess, EPA - Office of
Research and Development; Bobby E. Pyle, Southern Research Institute;
Tom Borak, Western Radon Regional Training Center; and David W. Saum, INFILTEC 10-85
Radon Diagnosis in a Large Commercial Office Building
David Saum and Marc Messing, INFILTEC 10-105
Design of Radon-Resistant and Easy-to-Mitigate New School Buildings
Alfred B. Craig, Kelly W. Leovic, and D. Bruce Harris,
EPA - Office of Research and Development 10-117
xi i
-------
The 1991 International Symposium on Radon
and Radon Reduction Technology
Volume 3: Panel Session Papers
and
Poster Papers
Technical Sessions I through V
Table of Contents
Panel Session I: Risk Communication
Apathy vs. Hysteria, Science vs. Drama: What Works in Radon Risk Communication (Withdrawn)
Peter Sandman, Rutgers University PNL1-3
American Lung Association's Radon Public Information Program
Leyla Erk McCurdy, American Lung Association PNL1-5
Radon Media Campaign
Dennis Wagner and Mark Dickson, EPA - Office of Radiation Programs PNL1-13
Developing a Community Radon Outreach Program: A Model for Statewide Implementation
M. Jeana Phelps, Kentucky Cabinet for Human Resources PNL1-29
Panel Session II: Detection of Radon Measurement Tampering
Policy and Technical Considerations for the Development of EPA Guidance on
Radon and Real Estate
Lawrence Pratt, EPA - Office of Radiation Programs PNL2-3
State Property Transfer Laws Now Include Radon Gas Disclosure
Michael A. Nardi, The Nardi Group PNL2-13
Guidelines for Radon/Radon Decay Product Testing in Real Estate
Transactions of Residential Dwellings
William P. Brodhead, AARST PNL2-17
How to Determine if Radon Measurement Firms are Providing Accurate Readings
Herbert C. Roy and Mohammed Rahman, New Jersey Department
of Environmental Protection PNL2-37
What Happens When You Do 477 Radon Inspections Preceded by Grab Samples? (Abstract)
Marvin Goldstein, Building Inspection Service, Inc. PNL2-51
Exploring Software Device Management Routines that Ensure the Overall Quality
of Continuous Working Level and Continuous Radon Monitor Performance
in a Field Environment
Richard Tucker, Gemini Research; and Rick Holland, Radonics, Inc PNL2-55
Use of Grab Samples as a Quality Assurance Tool to Enhance Overall Radon
Measurement Accuracy and Reproducibility
Brian Fimian, Radonics, Inc.; and Richard Tucker, Gemini Research PNL2-75
xi i i
-------
Panel Session III: Short-Term/Long-Term Measurement
Predicting Long-Term Indoor Radon Concentrations from Short-Term Measurements:
Evaluation of a Method Involving Temperature Correction
T. Agami Reddy, A. Cavallo, K. Gadsby, and R. Socolow,
Princeton University PNL3-3
Correlation Between Short- and Long-Term Indoor Radon Concentrations in Florida Houses
Susan E. McDonough and Ashley Williamson, Southern
Research Institute; and David C. Sanchez, EPA - Office of Research and Development PNL3-21
Relationship Between Two-day Screening Measurements of Radon-222 and
Annual Living Area Averages in Basement and Nonbasement Houses
S. B. White, N. F. Rodman, and B. V. Alexander, Research Triangle
Institute; and J. Phillips and F. Marcinowski,
EPA - Office of Radiation Programs PNL3-33
POSTER PRESENTATIONS
Session I: Government Programs and Policies Relating to Radon
The State Indoor Radon Grants Program: Analysis of Results After the First Year of Funding
(Withdrawn)
Laurie Amaro, EPA - Office of Radiation Programs P1 -3
EPA Radon Policy and Its Effects on the Radon Industry
David W. Saum. INFILTEC P1-5
EPA's National Radon Contractor Proficiency Program
G. Lee Salmon, John MacKinney, John Hoornbeek, and
Jed Harrison, EPA - Office of Radiation Programs P1-13
Draft Guidance to States on Radon Certification Programs
John Hoornbeek, EPA - Office of Radiation Programs; and
Barbara Zakheim, S. Cohen & Associates P1-23
National Radon Measurement Proficiency (RMP) Program
Philip P. Jalbert, John Hoornbeek, and Jed Harrison,
EPA - Office of Radiation Programs P1-45
Session II: Radon-Related Health Studies
Occupational Safety During Radon Mitigation: Field Experience and Survey Monitoring Results
Jean-Claude F. Dehmel, S. Cohen & Associates; Peter Nowlan,
R. F. Simon Company; and Eugene Fisher, EPA - Office of Radiation Programs P2-3
Cost Benefit Analysis of Radon Mitigation Systems in 157 Iowa and Nebraska Homes
Kenneth D. Wiggers and Tom D. Bullers, American Radon Services, Ltd; and
J. Peter Mattila and Laurent Hodges, Iowa State University P2-17
The Effect of Passive Cigarette Smoke on Working Level Exposures in Homes
Raymond H. Johnson, Jr. and Randolph S. Kline, Key Technology, Inc.; and
Eric Geiger and Augustine Rosario, Jr., Radon QC P2-25
xi v
-------
Session III: Measurement Methods
Characterization of Structures Using Simultaneous Single Source
Continuous Working Level and Continuous Radon Gas Measurements
Brian Fimlan and John E. McGreevy, Radonics, Inc P3-3
Radon and Water Vapor Co-adsorption on Solid Adsorbents
Neguib M. Hassan, Tushar K. Ghosh, Sudarshan K. Loyalka, and
Anthony L. Hines, University of Missouri-Columbia P3-13
Calibration of Modified Electret Ion Chamber for Passive Measurement
of Radon-222 (Thoron) in Air
P. Kotrappa and J. C. Dempsey, Rad Elec, Inc P3-35
Unit Ventilator Operation and Radon Concentrations in a Pennsylvania School
Norm Grant, Quoin Partnership; and William P. Brodhead, WPB Enterprises P3-43
Session IV: Radon Reduction Methods
Radon Mitigation Failure Modes
William M. Yeager, Research Triangle Institute; D. Bruce Harris,
EPA - Office of Research and Development; and
Terry Brennan and Mike Clarkin, Camroden Associates, Inc P4-3
Mitigation by Sub-Slab Depressurization Under Structures Founded on
Relatively Impermeable Sand
Donald A. Crawshaw and Geoffrey K. Crawshaw, Pelican
Environmental Corporation P4-15
A Laboratory Test of the Effects of Various Rain Caps on Sub-Slab
Depressurization Systems
Mike Clarkin, Terry Brennan, and David Fazikas,
Camroden Associates, Inc. P4-31
Analysis of the Performance of a Radon Mitigation System Based on Charcoal Beds
P. Wasiolek, N. Montassier, and P. K. Hopke, Clarkson University; and
R. Abrams, RAd Systems, Inc P4-43
Control of Radon Releases in Indoor Commercial Water Treatment
D. Bruce Harris and A. B. Craig, EPA - Office of Research and Development P4-63
Session V: Radon Entry Dynamics
Model Calculations of the Interaction of a Soil Depressurization System With the
Radon Entry Process
Ronald B. Mosley, EPA - Office of Research and Development P5-3
Effects of Humidity and Rainfall on Radon Levels in a Residential Dwelling
Albert Montague and William E. Belanger, EPA - Region 3; and
Francis J. Haughey, Rutgers University P5-23
xv
-------
The 1991 International Symposium on Radon
and Radon Reduction Technology
Volume 4: Poster Presentations
Technical Sessions VI through X
Table ot Contents
Session VI: Radon Surveys
A Cumulative Examination of the State/EPA Radon Survey
Jeffrey L. Phillips, EPA - Office of Radiation Programs;
and Jane W. Bergsten and S. B. White, Research Triangle Institute P6-3
Seasonal Variation in Two-Day Screening Measurements of Radon-222
Nat F. Rodman, Barbara V. Alexander, and S. B. White, Research
Triangle Institute; and Jeffrey Phillips and Frank Marcinowski, EPA -
Office of Radiations Programs P6-21
The State of Maine School Radon Project: Protocols and Procedures
of the Testing Program
Lee Grodzins and Ethel G. Romm, NITON Corporation; and
Henry E. Warren, Bureau of Public Improvement, Maine P6-31
Results of the Nationwide Screening for Radon in DOE Buildings
Mark D. Pearson, D. T. Kendrick, and G. H. Langner, Jr.,
DOE/Chem-Nuclear Geotech, Inc. P6-37
Session VII: State Programs and Policies Relating to Radon
Quality Assurance - The Key to Successful Radon Programs in the 1990's
Raymond H. Johnson, Jr., Key Technology, Inc. P7-3
Radon In Illinois: A Status Report
Richard Allen and Melanie Hamel-Caspary,
Illinois Department of Nuclear Safety P7-15
Session VIII: Radon Prevention in New Construction
Radon Prevention in Residential New Construction: Passive Designs That Work
C. Martin Grisham, National Radon Consulting Group P8-3
Preliminary Results of HVAC System Modifications to Control
Indoor Radon Concentrations
Terry Brennan and Michael Clarkin, Camroden Associates;
Timothy M. Dyess, EPA - Office of Research and Development; and
William Brodhead, Buffalo Homes P8-17
Correlation of Soil Radon Availability Number with Indoor Radon and
Geology in Virginia and Maryland (Visuals only)
Stephen T. Hall, Radon Control Professionals, Inc. P8-33
xvi
-------
Session IX: Radon Occurrence in the Natural Environment
Geologic Evaluation of Radon Availability in New Mexico: A Progress Report
Virginia T. McLemore and John W. Hawley,
New Mexico Bureau of Mines and Mineral Resources; and
Ralph A. Manchego, New Mexico Environmental Improvement Division P9-3
Paleozoic Granites In the Southeastern United States as Sources of Indoor Radon (Visuals only)
Stephen T. Hall, Radon Control Professionals, Inc. P9-23
Comparison of Long-Term Radon Detectors and Their Correlations with
Bedrock Sources and Fracturing
Darioush T. Ghahremani, Radon Survey Systems, Inc P9-33
Geologic Assessment of Radon-222 in McLennan County, Texas
Mary L. Podsednik, Law Engineering, Inc. P9-45
Radon Emanation from Fractal Surfaces
Thomas M. Semkow, Pravin P. Parekh, and Charles 0. Kunz,
New York State Department of Health and State University of
New York at Albany; and Charles D. Schwenker, New York State
Department of Health P9-59
National Ambient Radon Study
Richard D. Hopper, Richard A. Levy, and Rhonda C. Rankin, EPA-Office of Research and
Development; and Michael A. Boyd, EPA - Office of Radiation
Programs P9-79
Session X: Radon in Schools and Large Buildings
Design and Application of Active Soil Depressurization (ASD) Systems in
School Buildings
Kelly W. Leovic, A. B. Craig, and D. Bruce Harris, EPA -
Office of Research and Development; Bobby E. Pyle, Southern Research Institute; and
Kenneth Webb, Bowling Green, KY, Public Schools P10-3
Radon in Large Buildings: Pre-Construction Soil Radon Surveys
Ralph A. Llewellyn, University of Central Florida P10-15
Radon Measurements in North Dakota Schools
Thomas H. Morth, Arlen L. Jacobson, James E. Killingbeck,
Terry D. Lindsey, and Allen L. Johnson, North Dakota State
Department of Health and Consolidated Laboratories P10-27
Major Renovation of Public Schools that Includes Radon Prevention:
A Case Study of Approach, System Design and Installation, and
Problems Encountered
Thomas Meehan, TFM/SAF AIR P10-41
The State of Maine School Radon Project: The Design Study
Henry E. Warren, Maine Bureau of Public Improvement; and
Ethel G. Romm, NITON Corporation P10-47
Planning and Implementing the National School Radon Survey
Lisa Ratcliff, EPA - Office of Radiation Programs; and Jane W. Bergsten,
Ronaldo lachan, Harvey Zelon, and Lisa Williams, Research Triangle
Institute P10-53
xvi i
-------
Session VI
Oral Presentations
Radon Surveys
6- 1
-------
FACTORS ASSOCIATED WITH HOME RADON" CONCENTRATIONS IN ILLINOIS
Thomas J. Bierma and Jennifer O'Neill
Environmental Health Program
Department of Health Sciences
Illinois State University
Normal, Illinois 61761
ABSTRACT
The Illinois Department of Nuclear Safety has performed short-term
alpha-track radon testing in over 3000 homes throughout Illinois. Data were
also collected on a wide array of household characteristics and test: condi-
tions. Analysis of these data revealed a number of interesting patterns.
Contrary to many investigations, the highest average concentrations were
not obtained in homes monitored in winter. Though a number of models were
explored in the analysis, few could explain more than 10% of the variation in
radon concentrations and no model could explain more than 20%. This suggests
thaL other factors, such as local geology, may be largely responsible for
inter-house variations.
Among houses with a crawl space elevated radon was associated with having
an indoor entrance and not ventilating or insulating the crawlspace. In
houses with basements, the foundation construction materials were important
explanatory variables. Surprisingly, common entry routes, such as sump pits,
cracks, drains, and exposed earth, were not associated with radon concentra-
tion. Also contrary to other studies, energy efficiency was positively asso-
ciated with radon concentration.
Results suggest that factors governing radon concentrations in the
Midwest are poorly understood.
6- 3
-------
INTRODUCTION
The primary factors governing the entrance of radon into the home are
well known: 1) the soil radon gas concentration and soil permeability, 2) the
existence of entry routes between soil and home interior, and 3) a pressure
difference between home and soil to provide a driving force for radon entry.
However, specific conditions that influence these factors are not well under-
stood, and other factors, such as the distribution of radon in the home and
the infiltration of outside air, can influence the concentration of radon in
various parts of a home.
Though a number of studies have been performed to examine the
relationships between local conditions and indoor radon concentrations,
considerable uncertainty remains. In addition, relatively little work has
been conducted on factors influencing home radon concentration in the Midwest.
This study was performed using data collected on approximately 3,000
homes in Illinois that were tested for radon as a part of the Illinois
Department of Nuclear Safety's Radon Program. In addition to radon testing,
data were collected on a number of household and monitoring conditions. The
purpose of this analysis is to build a predictive model for indoor radon
concentrations in Illinois. Such a model could not only help identify homes
with a higher risk of elevated radon, but could also assist in directing
mitigation efforts.
A PRIORI MODEL SPECIFICATION
Data on a number of household and monitoring variables were collected
during monitoring (see the following section for details). An a priori model
was developed to guide analysis of these variables and their relationships to
indoor radon. The model is also useful in identifying gaps in the array of
variables included on the questionnaire.
Table 1 presents the household and monitoring variables evaluated in
this analysis. Indicated in the table expected relationships between each
variable and monitored radon concentration. These expectations are based upon
relationships reported by others in the literature and on the authors' own
field experience. Table 1 lists the basic references used in creating the
table. A detailed discussion of the current literature and a priori model is
not presented here due to space limitations. The reader is referred to re-
views such as references (1-3). However, a few comments should be made to
clarify subsequent modeling.
Air infiltration may b associated with an increase or decrease home
radon concentrations. Increases may occur if infiltration occurs due to house
depressurization, which can increase infiltration of soil gas. Similarly, the
mixing of air within a house can increase or decrease monitored radon depend-
ing upon the location of the monitor. For example, basement radon levels may
decrease and first floor radon levels may increase as basement air is distrib-
6- 4
-------
uted through a forced-air heating system. Therefore, variables related to air
infiltration and air distribution may demonstrate either a positive or nega-
tive relationship to radon in modeling results.
No information on local geology, weather conditions, or other radon
sources (other than basement wall construction material and exterior brick)
was collected. Thus, the contribution of these factors to radon
concentrations cannot be assessed with this data.
METHODS
In the Fall of 1986, the Illinois Department of Nuclear Safety began a
systematic radon testing program for houses in Illinois. Testing was
performed on a county-by-county basis using local assistance from county
health departments, Cooperative Extension Service personnel, or other organi-
zations. The program continued through the Spring of 1990, though no testing
was performed from the Summer of 1988 to the Fall of 1989.
An alpha-track detector was typically placed in the lowest livable area
of the home. After approximately one month of exposure, detectors were
returned to the contract laboratory (Tech/Ops Landaur, Inc.) for evaluation.
In addition to placement of detectors, surveyors completed a question-
naire with the assistance of a head-of-household. The questionnaire included
a number of questions on house characteristics and on the household occupants.
Though houses were not selected randomly, selection procedures, typical-
ly involving identification of houses from county highway maps, were intended
to avoid any systematic bias. Due to the complex interaction of factors
influencing radon concentrations, such selection processes may be relatively
free of bias (1). At least 20 houses were tested in each county. The number
of houses tested in a county increased with increasing population. The 3,021
monitored houses analyzed in this study were drawn from 73 of 102 Illinois
counties. These 73 counties contain approximately 88% of the Illinois popula-
tion ,
All data were evaluated for distributional characteristics and coding
errors or ambiguities. For many quantitative variables demonstrating non-
normal distributions, transformations were performed to enhance normality. In
most, cases, natural logarithm transformations were sufficient. For other
non-normal quantitative variables, however, no useful distributions could be
derived from transformation. Many variables demonstrated bi- or tri-modal
distributions. Such variables were typically transformed to artificial cate-
gorical variables using either theoretically or empirically based cut-points.
All final variables and their coding schemes are presented in Table 1. For
quantitative variables, scatter plots were assessed for evidence of hereroski-
dasticity and non-linearity. No problems were identified.
Multiple linear regression was used to explore models of monitored radon
concentration. A series of theory and non-theory based models were explored.
These are explained in more detail in the following section.
6- 5
-------
RESULTS
OVERALL
Radon concentrations were approximately log-normally distributed with a
geometric mean of 2.84 pCi/1 and a standard geometric deviation of 2.30. All
subsequent modeling and statistical analysis used the natural log transform of
radon concentration.
It was anticipated, prior to analysis, that substructure type, moni Lot-
location, and season of monitoring would be the primary explanatory variables
across all types of homes, but. that. Lhe effects of monitor location and season
of monitoring could vary by substructure type. Table 2 presents the ANOVA
results for radon concentration across the four basic substructure types
encountered. Results confirm that substructure is a critical explanatory
variable, with crawl space/no basement, producing the lowest concentrations and
basement/crawl space combination producing the highest concentration.
Some of the difference in means, however, may be due to monitor location
rather than a direct effect of substructure type. A small percentage of
houses with basements were monitored on the first floor, presumably because
the basements were not considered "livable". Table 2 presents ANOVA results
using first floor measurements only. Though significant differences are still
noted, slab-on-grade, rather than basement/crawlspace combination represents
the substructure associated with the highest average monitored concentrations
No analysis was conducted on the joint effect of monitor location and
substructure type since interaction is an artifact of the survey method: a
monitor was placed in the basement only if the home had a basement. The joint
effect of substructure and season are presented in Table 3. Both independent
variables are significantly associated with the dependent variable, and there
appears to be an interaction effect. In basement homes without a crawlspace,
fall monitoring produced the highest concentrations. For slab-on-grade homes,
fall monitoring produced the lowest concentrations.
Due to the apparent interaction between substructure type and monitoring
location and season, as well as probable interactions with other explanatory
variables, subsequent analyses were conducted separately for each of the four
substructure types.
Before leaving Table 3, however, it is important to note that, conLrary
to many research findings, winter was not the season of highest radon concen-
trations. Summer produced Lhe highest or second highest concentrations in
each substructure type.
A number of regression models v/ere explored for each substructure type.
These are presented in Tables h through 7. Model A in all tables represents
the results from a series of simple linear regressions using the variable
listed as the only independent variable. Interpretation of statistical sig-
nificance under multiple comparisons is an obvious problem. Consistency with
theory and the findings of other studies, as well as consistency across sub-
structure types, should be used in evaluating the results. This point will be
explored further in the discussion sec Lion.
6- 6
-------
Models B and C reflect models built upon theory and the results of other
models. In Model B, only the monitor location variables were used. In Model
C included the addition of monitoring location.
Models D through F represent non-theoretical modeling. Model D includes
all variables simultaneously in the model. Because many of the independent
variables are correlated, multicolinearity is a significant problem in Model
E. However, by comparing the regression coefficients and p-values to those
obtained in other models, one can use Model D to identify those variable that
are relatively sensitive or relatively insensitive to the inclusion of other
variables in the model . Model E uses Model C plus a forward selection proce-
dure, and Model F used a stepwise selection procedure without any forced
variables. Variables included through such non-theoretic methods should only
be considered curious possibilities as true explanatory factors for indoor
radon. Considerable support from other studies would be needed to enhance the
reliability of such models.
Model G represents the best attempt at a complete, theory-based model.
Beginning with a base of Model C, additional variables were added for which a
good theoretic foundation exists, and which demonstrated results consistent
with theory in Models A and D. Model G should be considered the best model
for prediction in Illinois houses.
The most immediate observation from the regression analyses is the
relatively low explanatory power of the models. The percentage of variation
in radon explained by the models was generally under 10%. Only for homes with
both a basement, and crawlspace does theoretical modeling achieve an r-square
greater than 0.2. (Note: although Model D can produce a greater R-square,
non-theoretic models are likely to have far less predictive power than explan-
atory power.)
Specific variables of interest include monitoring location and season.
For homes with basements, knowing whether the monitor had been located in the
basement or not was an important predictor of monitored radon in all models.
Knowing whether the monitor was located in a first-floor bedroom or elsewhere
on the first was not, generally, a good explanatory variable. This changed in
the non-theoretic models for homes with basement/ no crawlspace, possibly due
to the addition of the energy efficiency or central air conditioning varia-
bles. For homes without basements, bedroom location was important only for
homes with a crawlspace, and then, only when monitoring season was included in
the model.
Monitoring season was important in nearly all models though the effects
were not consistent across substructure types (supporting the previous ANOVA
results). Only for slab-on-grade homes, where sample size was small, did the
season variables generally have p-values greater than 0.1 and not produce a
significant r-square change from Model B to Model C.
BASEMENT/CRAWLSPACE SUBSTRUCTURE
For homes with a basement/crawl space combination, an entrance from the
interior of the home (most likely the basement) to the crawlspace produced a
consistent increase in monitored radon. Because the dependent variable is
logaritlmiic, coefficients represent the multiplicative effects of an independ-
ent variable on the radon concentration. The coefficient of 0.26 in Model G
6- 7
-------
indicates that a crawispace entrance increases home radon about 30%. Similar-
ly, venting a crawispace demonstrated a consistent decrease in radon, with
Model G indicating approximately 20% higher radon levels for homes with un-
vented crawl spaces. Coefficients for crawispace insulation and the presence
of exposed earth in the crawispace were both in the correct direction but did
not demonstrate consistent statistical significance.
Basement foundation construction material also appeared to be important.
Construction materials, from lowest to highest in their apparent contribution
to home radon levels are poured concrete, block, "other material", stone and
mortar, and brick. From Model G, brick is associated with an average increase
in radon levels of greater than 50% over poured concrete. (It is interesting
to note that brick reverses signs in Model D. This is apparently an artifact
due to multicollinearity.)
Energy efficiency demonstrated a consistent positive relationship with
radon. The greater the occupant-assessed energy efficiency, the greater the
radon. Perhaps for similar reasons, the number of people in the house was
negatively related to radon. Both variables may be related to the amount of
air exchange and the size of the house.
Among the most significant findings is that the standard entry routes in
a basement (cracks, sump pit, exposed earth, etc.) were not important explana-
tory variables. Some even had coefficients with a sign opposite that predict-
ed by theory. Also contrary to theory, the presence of a woodsl.ove or fire-
place had a negative coefficient and large p-value.
There are a number of interesting curiosities in the regressions of
houses with basement/crawlspace combinations. In Model A having room air
conditioning or an electric space heater were associated with lower radon
concentrations. These did not retain their significance in Model D and may
reflect the effects of energy efficiency. Another interesting point is the
importance of a brick exterior appearing only in the non-theoretic models.
Having an interior entrance to the basement was associated with higher radon
levels in Model A, though this association lessened dramatically in Model D.
A test for an interaction effect between basement monitor location and
basement entry was negative (results not presented here).
BASEMENT/NO CRAWLSPACE SUBSTRUCTURE
For basement homes without a crawispace, a similar pattern appears.
Basement construction materials, in the order of their apparent contribution
to radon, are: poured concrete, block, brick, and stone and mortar. This
order, and magnitude of effect, are roughly consistent with the findings of
basement/crawlspace homes.
Standard basement entry routes were, again, not significant, though the
presence of cracks had a marginal p-value and correctly signed coefficient in
Model A. The presence of a woodstove or fireplace also was not an important
explanatory factor.
An interaction term between basement monitor location and B-FINISH
(using the basement as a bedroom or living area) was found to be significant
in a separate test. The inclusion of this term in Model G resulted in a
dramatic change in the effect of B_FINISH. Together, these variables would
6- 8
-------
appear to have a substantial effect on radon. For a home with a basement used
as a bedroom of living area, basement levels averaged nearly b times higher
than in homes without finished basements.
Energy efficiency demonstrated the same relationship with radon as in
basement/crawl space homes.
Among the curiosities, the presence of central air conditioning or
central forced-air heat., which may be inversely related to energy efficiency,
demonstrated significance in Model A. Brick exterior was negatively related
to radon, the opposite effect of that in basement/erawlspace homes.
CRAWLSPACF./NO BASEMENT SUBSTRUCTURE
In homes with a crawlspace but. no basement., having an entrance to the
crawl space from the home interior was an important predictor of radon in all
models. Insulation and ventilation of crawlspace were directionally
consistent with theory, Lhrough p-values were marginal.
A number of curious associations were found. Having an attic fan was
consistently an important explanatory variable, though in the opposite direc-
tion generally expected from theory. Other variables, such as having hot
water heat, gravity feed furnace, and "other" fuel type, had low or marginally
low p-values in at least one model. Having a gravity feed furnace, for exam-
ple, produced an average 75% increase in home radon level. Explanations for
such associations are unclear.
SLAB-ON-GRADE SUBSTRUCTURE
For slab-on-grade homes, age of house and energy efficiency (which may be
related to age of house) were consistently important explanatory factors,
indicating that the older and less energy-efficient, the home, the lower the
radon.
A curiosity was the very strong inverse relationship between central air
conditioning and radon. having central air was associated with an average
decrease in radon of about 50%. Having a fireplace or woodstove was associat-
ed with a decrease in radon, though p-values were relatively large.
Because of the relatively small sample size for slab-on-grade homes,
caution should be used in assessing the importance of variables based on p-
value alone.
DISCUSSION
A number of limitations should be recognized when drawing conclusions
from this study. The methods used to select homes for testing produce a
greater likelihood of bias than more rigorous, pseudo-random selection meth-
ods. Though reasonable steps were taken to assure quality and consistency in
data collection, the use of local personnel and home occupants is likely to
introduce some error.
Because sampling was conducted by independent local agencies as time
permitted, a correlation between local factors (such as housing or geology)
6- 9
-------
and Lime-variant. factors (such as weather) may have been introduced. In
addition, lack of local geological data means that correlations between geolo-
gy and household conditions cannot be identified. Thus, some results may be
due to the confounding effects of geology and time-variant factors.
Finally, due to the high degree of correlation between variables, model-
ing is prone to error. Model results should be considered suggestive, not
confirmatory, evidence of Lhe acLual underlying relationships.
ft appears that the vast majority of radon variation is due to factors
other than the common household factors considered in this analysis. Such
factors may include differences in regional or local geology, housing con-
struction types, weather conditions at the time of monitoring, or household
factors not considered (such as the existence of thermal bypasses or vented
appliances). This does not indicate, however, that the factors considered in
this survey are not important determinants of radon (for example having a
brick foundation may increase radon readings an average of 50%) but only that
other factors appear to be more important, in explaining the variation between
houses. Household factors may not be consistent in their effects. This
supports policy recommendations that all homes be tested for radon, despite
the apparent presence or absence of known risk factors.
Location of the monitor in the basement produced, as expected, a signif-
icant increase in radon concentration. Location of the monitor in the bed-
room, as opposed to elsewhere on the first floor, was generally unimportant.
For research primarily intended to evaluate determinants of home radon concen-
trations, first floor monitoring in all homes is desirable to allow direct
comparisons across all house types.
Season of monitoring was an important explanatory factor. However, the
relationship between season and home radon concentration was not consistent
with other investigations nor was it consistent across housing substructures.
The finding that summer radon concentrations could be as high or higher than
other seasons suggests that current guidelines for winter monitoring be re-
evaluated. However, since these data reflect Lhe measurement of different
homes in different, seasons, they should be interpreted with caution. Addi-
tional research on seasonal effects in the Midwest and elsewhere in the coun-
try are needed.
Basement, foundation construction material was a relatively consistent
predictor of radon. Brick and stone foundations were consistently higher than
poured concrete, even after adjustment, for age of house. Block foundations
were slightly elevated compared to poured concrete. This suggests opportuni-
ties for low cost mitigative strategies if foundation walls are accessible,
and if a suitable radon barrier can be found.
An entrance to a crawlspace was consistently associated with increased
radon. Insulation, exposed earth, and lack of ventilation in the crawlspace
were also associated with increased radon, though less consistently. Tliese
findings suggest that the common weatherization practices of sealing crawl-
space entries, insulating floors, and installing vapor barriers not only save
energy, but may reduce radon (through vapor barriers may need to be sealed and
vented). Limiting crawlspace ventilation as an energy-saving measure, howev-
er, does not appear advisable.
6-10
-------
The work described in this paper was not funded by the U.S. Environmen-
tal Protection agency and therefore the contents do not necessarily reflect
the views of the Agency and no official endorsement should be inferred.
REFERENCES
I. Nero, A. Radon and its decay products in indoor air: an overview. In:
W.W. Nazaroff and A.V. Nero (ed.), Radon and its Decay Products in
Indoor Air. John Wiley and Sons, New York, 1988. p. 1.
?. Nazaroff, W., Moed, B., and Sextro, R. Soil as a source of indoor radon:
generation, migration, and entry. In: W.W. Nazaroff and A.V. Nero
(ed.), Radon and its Decay Products in Indoor Air. John Wiley and Sons,
New York, 1988. p. 57.
3. Sextro, R. Understanding the origin of radon indoors- building a pre-
dictive capability. Atmospheric Environment. 21:431, 1987.
4. Turk, B., Prill, R., Grimsrud, D., Moed, B., and Sextro, R. Character-
izing the occurrence, sources, and variability of radon in Pacific
Northwest homes. J. Air Waste Manage. Assoc. 40:498, 1990.
5. Fisk, W. and Mowris, R. The impacts of balanced and exhaust mechanical
ventilation on indoor radon. In: B. Seifert, H.Esdorn, M. Fischer, 11.
Ruden, J.Wagner (ed.) Indoor Air '87. IiistiLute for Water, Soil and Air
Hygiene, Berlin, 1987. p. 316.
6. Cohen, B., and Gromicko, N. Variation of radon levels in U.S. homes
with various factors. J. Air Poll. Cont. Assoc. 38:129, 1988.
7. Bierma, T., and Toohey, R. Correlation of lung dose with Rn concentra-
tion, potential alpha-energy concentration and daughter surface deposi-
tion: a Monte Carlo analysis. HealLh Physics, 57:429, 1989.
8. Sextro, R., Moed, B., Nazaroff, K., Revzan, K., and Nero, A. Investiga-
tions of soil as a source of indoor radon. In: P. Hopke (ed.). Radon and
Its Decay Products - Occurrence, Properties, and Health Effects. Ameri-
can Chemical Society, N.Y., 198/. p. 10.
9. Borak, T., Woodruff, B., and Toohey, R. A survey of winter, summer and
annual average Rn-222 concentrations in family dwellings. Health Phys-
ics, 5/:465, 1989.
10. Arnold, L. A scale model study of the effects of meteorological, soil,
and house parameters on soil gas pressures. Health Physics, 58:559,
1990.
II. Mosely, R., and Henschel, D.B. Application of radon reduction methods.
EPA/625/5-88/024, U.S. Environmental Protection Agency, Research Trian-
gle Park, NC, 1988.
6- I I
-------
TABLE 1. VARIABLE NAMES, EXPLANATIONS, CODING, AND THEORETICAL RELATIONSHIP TO INDOOR
RADON CONCENTRATION
Relationship to
Variable naae Description Coding3 indoor radon*5'0
Substructure
BMTCRL
BKTNCC
CRLNOB
SLAB
Season
WINTER
SPRING
SUMMER
FALL
Monitor
LOCBMT
LOCBDS
OTHER
N_PEOPLE
LNAGEHS
NSMOKER
ADu_^ijAB
PC_B3ICK
EN2R EFF
d
:ent AC
ROOK AC
FAN AT'IC
Basenent and crawlspace
Base«ent but no crawlspace
Crawlspace but r.o basenent
Slab-on-grade only
Monitoring done during the winter
Monitoring done during the spring
Monitoring done during the suGier
Monitoring done during the fall
Monitor located in the basenent
Monitor located in the bedroce
Monitor located soaewhere else on
first floor
Bow 3any people will be living in the
house for the next month?
What is the aae of the house?
Is there at least one SBOker in the house?
Does the house have any attached asphalt or
concrete slabs(attached garage, carport slab,
patio, driveway, etc.)?
What percent of the outside of the house is
covered with brick?
Sun of occupant ratings on five energy
efficiency characteristic. Each
characteristic rated on scale of 1 to 5.
Co you use central air-conditioning during
wars weather?
Do you use a rooi air_conditioner during
warn weather?
Do you use a whole house and/or attic fan
durinq warn weather?
count
natural log
of years
01=0, >01=1
A scale of
5=ieast to
25=!iost
energy
efficient
++
+-
V-
V-
t/ "
6- 12
-------
TABLE 1. VARIABLE NAMES, EXPLANATIONS, CODING, AND THEORETICAL RELATIONSHIP TO INDOOR
RADON CONCENTRATION {CONTINUED)
Variable naie
Descriotion
Coding3
Relationship to
indoor rador.b,c
CENT FA
8 BASE
CENT GF
ELEC SPA
Do you use a forced air central heating systen
during cold weather?
Do you use a hot water baseboard or radiator
systec during cold weather?
Do you use a gravity flow central heating systen
during cold weather?
Dc you use an electric space heater during
cold weather?
+/-
+/-
ROOD Do you use a fireplace or wood burning stove
during cold weather?
KEROHEAT Do you use a kerosene space heater during cold
weather?
GAS_STCV Do you use a gas stove during cold weather?
EIAT_OTHER Do you use soae other kind of heating systei
during cold weather that was not aentionea on
this questionnaire? If sc, specify.
3_FINISH Is all or a portion of the baseient freguently
used as a bedrooi or living area? (If so, speci
foundation The outside baseaent walls are priiarily coaposed
BLOCK Concrete or cinder block
CRET Poured concrete
NORT Stone and sortar
BRK Brick
OTHER Other and tile
-/-
2NT BASE
Can you enter the baseaent fron inside
the house?
+/-
CRACKS
Does the baseaent floor)or sub-surface
floor in a split-level hoie) have large
cracks or holes?
DRAINS
SUMP
Does the baseaent fioor(or sub_surface
floor in a split-level hone) have drains?
Does the baseient floor(or sub-surface
floor in 3 split-level hose) have suap
pumps?
6-13
-------
TABLE 1. VARIABLE NAMES, EXPLANATIONS, CODING, AND THEORETICAL RELATIONSHIP TO INDOOR
RADON CONCENTRATION (CONTINUED)
Variable naie
Eescriotion
Coding"
Relationship to
indoor radcnfc'c
EARTH
Does the baseaeni fiocrfor sub-surface
floor in a split-level hoie) have exposed
earth?
+
CSP ENTR
Can you enter the crawl space froa inside the
house(froa baseient, for example)?
+
CSP INSL
Is the floor above the crawl space insulated?
CSP VENT
Will the crawl space be vented to the outside
curing the lonitoring period?
CSP EXP
Does the crawl space have exposed earth?
a. All Yes/Nc variables codes as C=nc, l=yes
b. ++ = strongly and positively related
+ = positively related
- = negatively related
— = strongly and negatively related
+/- = both positive and negative relationships nay be expected
? = toe little infornation tc predict a relationship
c. Based largely upon references I through 11
d. These variables were used to define datasets and do not appear explicitly in later tables.
6-14
-------
TABLE 2: ANALYSIS OF VARIANCE RESULTS FOR RADON CONCENTRATION 3Y SUBSTRUCTURE TYPE.
Monitor
Geouetric lean
Geouetric
location
Substructure type
concentration (pCi/1)
Standard Dev.
N
All(a)
All types
2.84
2.30
37021
Baseient and crawlspace
3.56
2.21
752
Crawl space, no basement
1.67
2.11
572
Baseient, no crawlspace
3.16
1 o 1
6 « 61
1,548
Slab-or-grade only
2.32
2.33
134
First floor(b)
All types
1.79
2.14
853
Baseient and crawlspace
1.95
1.88
78
Crawlspace, no fcaseaent
1.67
2.11
572
Baseient, no crawlspace
1.81
2.07
69
Slab-on-grade only
2.32
2.33
134
(a) p-value = <.001 for at least two aeans differing by substructure type.
!b) p-value = <.001 for at least *wo neans differing by substructure type.
TABLE 3: GEOMETRIC MEAN CONCENTRATION IN pCi/1 (AND SAMPLE SIZE)
BY SEASON OF MONITORING AND SUBSTRUCTURE TYPEa
BASEMENT BASEMENT NO BASEMENT NO BASEMENT
CRAWLSPACE NO CRAWLS PACE CRAWLSPACE NO CRAWLSPACE
SEASON
WINTER
3.74
(509)
3.22
(1137)
1 .68
(354)
2.27
(97)
SPR TNG
2.34
(122)
2.51
(218)
1.39
(103)
2.83
(17)
SUMMER
4. /6
(65)
3.74
(118)
1.67
(45)
2.34
(14)
FALL
3.94
(56)
3.78
( 75)
2.03
(70)
1.70
(6)
a. p-values ~ <.001 for both main effects and interaction
6-15
-------
TABLE 4: MULTIPLE REGRESSION RESULTS FOR MODELS OF HOME RADON CONCENTRATION:
HOMES BUILT OVER BASEMENT AND CRAHLSPACE
VARIABLE
MODEL A
MODEL B
MODEL C
MODEL D
MODEL E
MODEL F
MODEL G
INTERCEPT
-
.52447
.90441
.53720
.13636
-.06271
.42381
-
;<.01)
(<.C1)
(<.01)
(<.01)
(.79)
(.20)
SINTER
-.13391
-.11985
-.13353
-.19473
-.17511
(.34)
(.38)
; .34)
(.14)
(.19)
SPRING
-. 6C418
-.54158
-.62871
-.65254
-.48379
-.64939
(<.01)
(<•01)
(<-oi)
(<.01)
(<.01)
(<•01)
SUMMER
.15321
.10257
.05199
-.03172
-.08679
(.74)
1.82)
(.91)
(.94)
(.84)
LOCBMT
.66483
.66463
.57401
.55822
.58134
.63876
.57936
(C.01)
;<.oij
(<.01)
Ml)
(<.C1)
(C.01)
(<.01)
LOCBDR
-!02230
-.02230
-.08867
-.22949
-.11436
-.11555
: .91)
;.91)
(.65)
(•24)
(.55)
(.54)
NPEOPLS
-.05905
(.03)
-.05277
(.05)
-.04495
(.07)
ADJJLAB
. 1C375
.08877
(.17)
(.26)
NSMOKES
.09009
(.21)
.10630
(-13)
LNAGEHS
-.04625
(.18)
-.05477
(.20)
-.06157
(.13)
ENERJFF
.02528
.01958
.02655
.02516
.02513
: .01)
(.04)
(<.Ci)
(<•01)
(<.oi)
PC BRICK
-.07020
.50320
.34598
(.45) (<.0i; (<.01)
CENT AC .09944 .CS401
(.17) (.33)
ROOMJ.C -.14919 -. 06130
(.06) (.49)
FAN_ATIC .05429 .1C389
(.50) (.30)
CENT FA .05843 -.C3392
(.52) (.81)
M BASE -.05531 .C1585
(.52) (.92)
CENT GF -.30802 -.27069
(.19) (.26)
ELECJPA -.22500 -.02960
(.06) (.80)
FPJJSJE -.03622 -.10073
(.53) (.15)
KEROHEAT -.05649 .00570
(.47) (.96)
GAS STOV -.31293 -.24238
(.14) (.24)
HEAT OTHER .12606 .31080
(.35) (.03)
6-16
-------
TABLE 4: MULTIPLE REGRESSION RESULTS FOR MODELS OF HONE RADON CONCENTRATION;
HOMES BUILT OVER BASEMENT AND CRASLSPACZ {COiTIH'JE)
VARIABLE
MODEL A
MODEL B
MODEL C
MODEL D
MODEL E
MODEL F
MODEL G
BLOCK
.09408
.10501
.13430
(.27)
(.23)
(.10)
KCRT
.13457
.41139
.39434
(.32)
;<.01l
(<.01
BRK
.33532
-.13263
.34835
.50465
(<•01;
(¦15)
(<.fll)
(<.G1)
0:HESBKT
.18767
.32299
.34548
(•21)
(.03)
(.02)
BJINISH
•C1927
.02686
(.81)
(.75)
ENTBASE
.26429
.05568
(.02)
(.61)
CRACKS
-.05275
-.03644
(.47)
(.61)
DRAINS
-.04133
-.11459
(.56)
(.20)
SUMP
-.03177
-.00855
(.56)
(.26)
EARTH
.06191
.08333
(.45)
(.32)
CSP_ENIR
.30078
.21091
.23927
.22255
.25269
(<•01)
1.03)
(.31)
(•01)
(<•01)
CSPJNSL
-.03825
-.06228
(.62)
(.40)
CSP_VENT
-.13419
-.20053
-.18782
-.18901
-.19639
(.07)
(<-ai)
(.01)
(.01)
(.01)
CSF_EXP
.06678
.09223
.19421
.2C2C0
.14370
(.47)
(.34)
(.03)
(.02)
(•13)
Multiple R?
.07481
712360"
725315"
7l9353~"
""718892""
.21718
(p-vaiue)
(<•01)
(<•01)
(<•01)
•<•01)
(<•01)
;<.oi)
R^-Change
.04879
.12955
(p-value)
(<.01)
(<.91;
6- 17
-------
TABLE 5: MULTIPLE REGRESSION RESULTS FOR MODELS OF HOME RADON CONCENTRATION:
HOMES BUILT OVER EASEMENT BUT NO CRAWLSPACE
VARIABLE MCCEL A MCCEL B MODEL C MODEL D MODEL E MODEL F MODEL G
INTERCEPT
-
.68446
.96252
.73018
.84825
.50784
,78389
-
(<.01)
(<•31)
(.02)
(<•31)
(.02)
(<•01)
WINTER
-.18722
-.16864
-.20245
-.17015
-.17590
(.13)
(.17)
(.10)
(.16)
(.14)
SPRING
-.46281
-.41092
-.42518
-.40259
-. 23534
-.41147
(<.01)
(<.Q1)
(<•01)
(<.01)
(<•01)
(<.01)
SUMMER
-.09921
-.09921
-.11711
-.07166
-.09674
(.76)
(.76)
(.71)
(•32)
(.76)
LOCBMT
.47646
.47646
.39752
.40905
.39413
.57138
.39319
(<.C1)
(<.01)
(.01)
(<.01)
(.01)
(<•01)
(<.01)
LQCBDR
-.29852
-.29852
-.33525
-.47366
-.43520
-.47317
-.36979
(.18)
(.19)
(.14)
(.04)
(.05)
(-04)
(.11)
NJEOPLE
.32780
.03166
(.12)
(.08)
ADJSLAB
.07140
.11599
(.16)
(.04)
NJMOK5R
.02038
-.00685
(.67)
(.89)
LNA3EHS
-.02615
-.08970
-.08629
-.09404
(.29)
(<.C1)
«.G1)
(<.01)
ENERJF?
.01519
.32134
.01692
.01666
.01699
(.02)
(<.01)
(.01)
(.02)
(.01)
PC3RICK
-.11062
' fu'
-.07839
I "in
CESTAC
-.11208
-.12129
-.12048
-.12346
(.02:
(.05)
(.02)
(.02)
ROOM_AC
.04593
-.08926
(.43)
(.21)
FANATIC
-.11001
-.05317
(.13)
(.20)
CENTJA
-.11838
-.13650
-.11611
(.05)
(.24)
(.04)
WJASE
.10301
-.03680
/ .15*
I ~ LJ ,
(.77)
CENT_GF
.01615
-.14266
(.92)
(.47)
ELEC_SPA
-.02207
-.02826
(.83)
(.75)
FPJfSJE
.02871
.01815
(.37)
(.72)
KEROBEAT
.11347
.12184
(.29)
(.25)
GASJTCV
-.01028
.02939
(.94)
(.83)
HEAT_OTHER
.06506
-.03535
(.52)
(.77)
BLOCK
.21535
.28573
.27904
.26975
.27563
(<•01)
(<.01)
(<•01)
(<•01)
(<.01}
6-18
-------
TA3LH 5:
MULTIPLE
REGRESSION RESULTS FOR MODELS O?
HOMES BUILT OVER BASEMENT BUT
HOME RADON CONCENTRATION:
0 C3AWLSPACE (CONTINUED)
VARIABLE
MODEL A
MODEL B MODEL C
MODEL 3
MODEL E
MODEL F
MODEL G
MORT
.35438
.46092
.48845
.48387
|<.01)
f <¦31)
(<.01)
(<.01j
BRK
.26990
.38522
.40233
.40290
.49618
(<.01)
«.01)
(<•01!
(<•011
(<•01)
OTHERBMT
-.04626
.07880
.39319
(.72)
(.55)
(<•01!
B_F"HI3H
-.0693?
-.10759
-.11224
.C8526
(.15)
(.03)
(.02)
(.51)
LOCFIN (a)
.64436
(.12)
ENTJASE
.02046
-.91569
-.74116
(.85)
(.89)
(•38)
CRACKS
.08293
.05840
(.09)
(.25)
DRAINS
.34636
.06675
• .52)
(.35)
SDMP
-.32555
-.04301
(.59)
; .39)
EARTH
.09408
.13934
.15703
(.31)
(.15)
(.10)
Multiple R?
.02426 .03894
.10581
.08507
.08036
.09169
(c-value)
(<•01) (<•01)
(<.01)
(<.01)
(<.01}
Ml)
R^-Change
.01468
.G6687
(p-value)
(<•01)
(<.ci)
6- 19
-------
TABLE 5: MULTIPLE REGRESSION RESULTS FOR MODELS OF HOME RADON CONCENTRATION:
HOMES BJIIiT OVER CRAHLSPACE BUT NO BASEMENT
VARIABLE
MODEL A MODEL B
MODEL C
MODEL D
MODEL E
MODEL F
MODEL 3
Intercept"
. 743901
""763841"'
128924""
"""76085"
750219
174458
{<•01)
Ml)
(.42)
Ml)
(<•01)
(<•01)
WIHTER
-.16455
-.19250
-.23470
-.26874
-.29359
(.14)
(.09)
(.07)
(.02)
(.01)
SPRING
-.38879
-.42165
-.40197
-.44044
-.19195
-.50885
Ml)
(<•01)
(.01)
Ml)
(.03)
Ml)
SUHHER
-.21478
-.27490
-.32321
-.41428
-.39123
(.24)
(.14)
(.11)
(.03)
(.04)
LCCBMT
LOCBDR
.11316 .11316
.14463
.15581
.17344
.15547
(.13) (.13)
(.05)
(.04)
(.02)
(.34)
NPSOPLE
.00389
.01128
(.89)
(.70)
ADJ_SLAB
.04341
.10461
(.58)
(.20)
.Nl SMOKER
.11543
.06467
(.11)
(.38)
LNAGEHS
.04566
.03790
(.24)
(•41)
ENERJFF
.00355
.01355
(.72)
(.21)
PCJRICK
.02862
.03722
(.77)
(.71)
CENT_AC
-.02534
.06893
(.72)
(•51)
ROOK_AC
.01281
.02986
(.87)
(.77)
:AN_AIIC
-.29695
-.25446
-.24573
-.26775
(<,01)
(.02)
(.02)
(•ci)
CENTJA
-.05005
-.17104
(.52)
(.22)
WBASE
.30619
.22883
.36088
.32856
(.03)
(.21!
(.01)
(.02)
CENI_GF
-.57371
-.56661
[•07;
(.09;
ELEC_SPA
-.16027
-.11494
(.181
(.35)
FP_8S_HS
-.07352
-.06947
(.34)
(.40)
KEROHEAT
-.06352
-.00899
(.71;
(-96)
GAS_STOV
.19141
.07700
(.17]
(.54)
HEAT_OTHER
-.19467
-.24163
(.09)
(.12)
CSPENTR
.22661
.28046
.25974
.21505
.23106
(<•91)
(Ol)
(<•01]
(.01)
:<.oi)
6-20
-------
TABLE 5: MULTIPLE REGRESSION RESULTS FOR MODELS OF HOME RADON CONCENTRATION:
HOMES BUILT OVER CRASLSPACE BUT NO BASEMENT (CONTINUED)
VARIABLE
MODEL A
MODEL B
MODEL C
MODEL D
MCCEL E MODEL F
MODEL 3
CSP IH3L
-.12095
-.11359
-.10899
(.12)
(.16)
(.14)
CSP VENT
-.08019
-.13048
-.12354
(.26)
(.08)
(.09)
CSF EXP
.03655
.01987
(.65)
(.83)
Multiple R*
.00492
.02936
.11306
.07685 .05571
.06398
(c-value)
(.13)
(.01)
(<•01)
(<.31) (.oa;
(<.01)
R^-Change
.02444
.08370
[p-value)
(.01)
(.01)
6-21
-------
TABLE 7: MULTIPLE REGRESSION RESULTS FOR MODELS CF HOME RADON CONCENTRATION:
HOMES BUILT OVER SLAB-ON-GRADE ONLY
VARIABLE
MODEL A
MODEL B
MODEL C
MODEL D
MCDEL E
MODEL F
HCCEL G
INTERCEPT
-
.72577
.30640
-1.88722
.47208
1.07722
-1.47057
-
(c.01)
(.44)
(.08)
(.23)
(<.C1)
(.05)
WINTER
.44310
.43118
.55750
.49918
. 43731
(•27)
'.28)
(.18)
(21)
(.30)
SPRING
.70936
.63447
.79930
.63153
.63846
(•11)
{•16;
(.09)
; .15)
(.15)
SUMMER
.45747
.38257
.59130
.56041
.36715
(.32)
(.41;
(.22)
(.23)
(.42)
LOCBDR
.24511
.24511
.21398
.11732
.19420
.19473
(.14)
(.14)
(•21)
(.52)
(.25)
(.24)
N_PEOPLE
-.06949
-.04225
(.31)
(.59)
N_SKOKES
.25388
.26591
(.13)
(.13)
LNAGEHS
.13552
.23415
.23172
(.14)
(.03)
: .02)
ENERJFF
.03203
.08430
.05745
(.17)
(<.01j
(.03)
PCJRICX
-.09390
-.23546
(.65)
(.28)
CESTAC
-.39525
-.54079
-.39441
-.39525
(.02)
(.01)
(.32)
(.02)
ROOM_AC
.10428
-.41429
(.60)
(•13)
FANATIC
.26593
.57891
(.43)
(.10)
CEUTJA
-.16207
.20092
(.39)
(.54)
W_BASE
.13815
.38537
(.63)
(.31)
CENT_GF
.11750
.32170
(.82)
(.58)
ELECJPA
-.03449
.13368
(.92)
(.71)
F?_WS_HE
-.21622
-.25058
(.24)
(.19)
XEROHEAT
.73459
.91013
(.24)
' .17)
GAS_STOV
-.20941
.20125
(.64)
(.70)
HEATOTHER
.09399
.01911
(.72)
(.95)
Multiple
.03009
.03912
.24870
.08740
.05156
.10304
(p-value)
(.14)
(.37)
(.10)
(.08)
(.02)
(.07)
R^-Change
.01904
.20958
(p-vaiue)
(.55)
(.09)
6-22
-------
RADON IN FEDERAL BUILDINGS
Michael Boyd
U.S. Environmental Protection Agency
Office of Radiation Programs
Washington, DC 204 60
Terry Inge
Sandy Cohen and Associates
McLean, VA 22101
ABSTRACT
The Environmental Protection Agency (EPA) has provided
guidance to Federal agencies in response to requirements of the
Indoor Radon Abatement Act (IRAA) for testing Federal buildings
for radon. Twenty-two agencies have reported complete or partial
results to EPA. These data are included in the first report to
Congress on radon in Federal buildings which is now in
preparation. Initial analysis indicates that the percentage of
Federal buildings tested with screening levels above 4 picocuries
per liter (pCi/L) may be somewhat lower than that routinely cited
for houses. However, given the large number of such buildings in
the U.S., this percentage indicates there are many Federal
buildings with elevated radon levels. Except for the Armed
Services, most of the Federal studies have focused on office
buildings. The data collected thus far does clearly demonstrate
that office buildings are not immune to radon problems. Because
of the large number of people that work in a typical multistory
office building, and the ease with which radon can be transported
through a building's ventilation system, the Federal Building
Survey results have already demonstrated the prudence of testing
the workplace for radon.
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.
6-23
-------
INTRODUCTION
The Indoor Radon Abatement Act (IRAA) was passed by Congress
in October, 1988. Among its provisions was a requirement in
Section 3 09 for all Federal departments and agencies that own
buildings to test a representative sample of those buildings for
radon. The Environmental Protection Agency (EPA) reviewed the
radon study designs submitted by each affected agency and
provided information available on high risk areas, testing
protocols, and other technical guidance. EPA also provided
direct support (measurement devices, device analysis, quality
control activities) to interested agencies through interagency
agreements (IAGs).
EPA established contact with 48 Federal departments and
agencies and notified them of the provisions of IRAA. It was
determined that 23 of these agencies owned at least some of their
own buildings and were responsible for conducting radon surveys.
Many of the remaining agencies had their buildings tested by the
General Services Administration (GSA).
The first report to Congress on the results of radon testing
has been drafted and includes at least some results from 22 of
the 23 affected agencies. The reporting agencies primarily used
passive detectors (typically alpha track detectors or activated
charcoal devices) to make screening measurements of radon in
ground contact rooms, stairwells, and on upper floors near
elevator shafts and ventilation ducts. The duration of the radon
tests ranged from 2-7 days for the charcoal devices to 3 months
for the alpha track detectors. Some agencies, such as GSA, chose
to sample all of their buildings, while others sampled only a
portion. The EPA shared with other Federal agencies existing
information on surface uranium deposits in the U.S. as well as
the results of State radon surveys. Some agencies used this
information to select regions to include in their sampling
strategy.
The testing plans of the agencies differed in certain
aspects including the type of detector used, the sampling
strategy, the length of time the detector was deployed, and the
season during which the test was conducted. Nevertheless, there
is remarkable similarity between the surveys. Initial analysis
indicates that the percentage of Federal buildings tested with
screening levels above 4 picocuries per liter (pCi/L) may be
somewhat lower than that routinely cited for houses (see Tables 1
& 2). Because many buildings operate under steady-state
conditions and because seasonal differences may not be as
important a factor in determining radon entry in big buildings as
it is for houses, the results reported in this paper may not
differ greatly from the annual averages. Most agencies have
already begun performing long term tests in areas that screened
above 4 pCi/L. Results from these tests will be useful in
determining how much variability in radon levels occurs as a
result of seasonal changes.
6-24
-------
The types of buildings owned by a particular agency have a
significant effect on the measurement results. For example,
laboratories with one-pass air handling systems and large office
buildings with significant outside air intake tend to have fewer
radon problems than smaller buildings or houses. In addition to
larger buildings, the Army, Navy, and Air Force also conducted
residential testing. Because of the large number of test results
reported by the Army, the distribution of radon levels by Army
building type is presented later in this paper.
MATERIALS AND METHOD
EPA reviewed and ultimately approved the radon survey
designs submitted by Federal agencies. Approval of the plans was
often contingent on an agency agreeing to modifications suggested
by EPA. Many of the agencies chose to perform 3-month
measurements using alpha track detectors. The next most popular
device was the diffusion barrier charcoal canister which was
typically deployed for 5 to 7 days. Electret ion chambers and
continuous monitors were also used in some cases. EPA's policy
is that any measurement device listed in the Indoor Radon and
Radon Decay Product Measurement Protocols (EPA 520/1-89-033,
March 1989) is acceptable for making short-terra measurements,
except that grab sampling may not be used alone. Grab sampling
can be a useful confirmatory measurement made in conjunction with
a measurement from some other approved device.
In responding to the survey designs, EPA recommended that
all occupiable rooms in ground contact be tested. Open areas in
ground contact were recommended to be tested at a density of one
detector every 2000 square feet. In addition, it was recommended
to place at least one detector on every floor. Placement of
detectors in stairwells, near elevator shafts, and in the
vicinity of each vertical service shaft was encouraged.
Agencies were advised to test during the winter heating
season when possible. The reporting constraints imposed by IRAA
did result in extensive testing during other seasons, however.
EPA is currently investigating the effect of climate on indoor
radon concentrations in big buildings. Preliminary indications
are that seasonal differences do not have as marked an effect in
big buildings as they do in houses.
IRAA also required private water supplies to be tested for
radon. Since most Federal buildings are on public water
supplies, very little water testing was performed.
6-25
-------
RESULTS
Many of the Federal agencies are still conducting radon
surveys. This is particularly true for the Armed Services where
world-wide testing and large property holdings are involved. The
results reported to EPA as of February, 1991 are given in Table 1
for the total number of reported measurements and Table 2 for the
total number of measured buildings. Each building represented in
Table 2 is characterized by its highest single reading. It is
not statistically valid to lump together surveys that have
different sample selection strategies, differing measurement
devices and techniques, and differing periods of measurement.
Nevertheless, the results do provide a qualitative indication of
the extent of radon contamination in big buildings.
Table 1. Preliminary Federal Agency Radon Survey Data
Summary Data for 84,642 Measurements
Radon Concentration
Number of Measurements
Percent of Total
0-2 pCi/L
71,346
84.3
2-4 pCi/L
8,851
10.5
4-10 pCi/L
3,511
4.1
10-20 pCi/L
812
1.0
20-200 pCi/L
115
0.1
> 200 pCi/L
7
<0.1
Table 2. Preliminary Federal Agency Radon Survey Data
Summary Data for 52,031 Buildings
(Includes 29,671 Army Housing Units)
Radon Concentration
Number of Buildings
Percent of Total
0-2 pCi/L
42,864
82.4
2-4 pCi/L
5,948
11.4
4-10 pCi/L
2,469
4.7
10-20 pCi/L
649
1.2
20-200 pCi/L
94
0.2
> 200 pCi/L
7
H
•
O
V
6-26
-------
The effect of combining measurement data from houses,
schools, daycare centers, and hospitals with office building data
has implications for the extent to which these surveys can be
used to indicate radon contamination in big buildings.
Fortunately, the data from the Army was coded so that building
types could be analyzed separately. The Army survey uses alpha
track detectors deployed for 90 days. Using the Army as an
example, Figures 1 & 2 represent the format used for reporting
data in the report to Congress. Figure 1 represents the total
number of measurements by concentration range and Figure 2
represents the number of affected buildings, using the highest
reading in each building to characterize it. Figures 3-5
represent the measurement data separated by building type. As
these figures indicate, there is not a great difference between
the radon distribution for these 3 categories. With the
exception of the Army, Navy, and Air Force, most of the remaining
agencies reported primarily the results of testing in the
workplace.
Most agencies reported that their buildings were served by
municipal water systems and so they were not required to test
their water for radon. Radon concentrations in surface water and
water that has been aerated are typically very low. A few
agencies did report results for buildings served by private water
supplies. Of the 163 water supplies tested, 46 were above 300
pCi/L.
Some agencies have offices that are located on the upper
floors of multistory office buildings. If the Federal government
does not occupy the entire building, then generally the Federal
tenant may only legally test for radon on the floors that it
occupies. Sharing space in large office buildings is common for
EPA, with its 10 regional locations. EPA also has a significant
number of field stations and laboratories that have one-pass air
handling systems. For these reasons, EPA did not expect to find
many instances of elevated radon levels in its buildings. This
expectation proved correct as can be seen from the graphs of
EPA's data (Figures 6 & 7). EPA's measurements were made with
diffusion barrier charcoal canisters which were, in most cases,
deployed from Monday morning until Friday afternoon. The
canisters were then returned for analysis to one of the EPA
Office of Radiation Programs (ORP) laboratories located in
Montgomery, Alabama and Las Vegas, Nevada.
In general terms, distance from the source (usually the
soil) and extent of dilution flow (degree of outside air intake
and air exchange rate) have a major effect on radon
concentrations on upper floors of buildings. EPA has not yet
found an example where radon levels were above 4 pCi/L on upper
floors of a building and below 4 pCi/L in the basement or ground-
contact floor (the recommended screening location). For the
usual case, where the source of radon is soil gas, simple
diffusion will result in radon concentrations decreasing on upper
floors of the building. Depending on the ventilation pattern,
6-27
-------
radon levels may not decrease in a linear fashion, floor by
floor, as predicted by diffusion alone. In fact, the
configuration of the air handling system may result in the second
highest radon level in a building being found on the top floor.
This case has been observed where the supply air was being pumped
from the basement directly to the top floor to assist in cooling
the building during the summer. In some cases, radon may enter a
building through a conduit that penetrates the foundation. For
this reason, telephone rooms and electrical cabinets are likely
places to find radon on upper floors.
EPA's recommendations to Federal agencies to sample
extensively in ground contact areas during periods of minimal
outside air intake appears to be sound. The importance of
testing big buildings during the winter is being examined in
tests that will be completed this year. Based on these findings,
an interim protocol for testing big buildings for radon will be
issued. The Federal building surveys that have been completed to
date indicate the need for increased testing of the workplace. A
big building protocol will be especially useful for the expected
increase in radon testing in this area.
For most Federal landlords, the minor costs associated with
testing will be followed by the good news that their building
does not have a radon problem. For those buildings that indicate
elevated radon concentrations, a significant health threat will
have been identified which can then be corrected. EPA is
actively investigating diagnostic and mitigation tools for
application in big buildings. It is anticipated that EPA will
continue to provide support to agencies in addressing the
mitigation of problem buildings and that the lessons learned and
technologies developed from these efforts will be transferred to
the private sector.
CONCLUSIONS
Since the individual agency results have yet to be
transmitted to the Congress, it would be premature to list them
here. The results reported for EPA and the Army are by
permission. Several conclusions from the Federal building study
may be drawn.
1. Soil gas is the primary source of radon in Federal
buildings, as it is for other building types.
2. The effect of seasonal differences on radon levels in
big buildings has not yet been confirmed, but appears
to be less significant than for houses and other small
buildings. EPA still advises testing during the winter
or during periods of minimal outside air intake.
3. The highest radon levels are most likely to be found in
ground contact areas. Radon levels in other parts of a
6-28
-------
building will be determined by the distribution pattern
of air handling systems, by diffusion, and by the
availability of direct paths to the sub-slab
environment such as telephone and electrical conduits.
4. From the data received thus far, the percentage of
Federal buildings with radon screening measurements
above 4 pCi/L appears to be somewhat less than the
corresponding percentage of houses. Because of the
steady state conditions maintained in many big
buildings, the difference between screening and annual
measurements in big buildings may be less than for
houses.
5. The sampling strategy presented by EPA, namely to test
extensively in ground contact areas, in stairwells,
outside elevator shafts, and near vertical service
shafts, appears to be effective for locating buildings
with potential radon problems.
6. Most Federal buildings are connected to municipal water
supplies and are not expected to have significant radon
in their water. For buildings that do use private
water supplies, a significant percentage are likely to
be above 300 pCi/L.
7. The prudence of testing workplaces for radon is
confirmed by the results of this study.
6-29
-------
RADON IN SWITZERLAND
H.Surbeck and H.Volkle
Federal Office of Public Health
c/o Physics Institute, University
Perolles
CH-1700 Fribourg, Switzerland
W.Zeller
Federal Office of Public Health
Bollwerk 27
CH-3001 Bern, Switzerland
ABSTRACT
Based on measurements in nearly 1600 homes, representing 0.15% of the
housing stock, we estimate that the Swiss live on the average in rooms
with a radon concentration of 80 Bq/m3 and that 5% of them are exposed to
concentrations exceeding 200 Bq/m3.
Radon research in Switzerland started nearly a decade ago and shows
that building materials and household water use present no serious radon
problems, the soil being the main radon source. The highest values are
found in homes on highly permeable building grounds (Karst terrains,
rockslides).
We discuss the results of the radon surveys and explain how we try to
get a representative exposure estimate from biased data. We also present
geological aspects of the radon situation in our country and outline the
policy for the new decade that will see surveys concentrated on the search
for hot spots.
Several mitigation techniques have been tested successfully but few
homeowners are interested to take remedial actions. There is no great
public concern on radon in Switzerland; radon is natural.
6- 31
-------
INTRODUCTION
As early as in 1908 Gockel (1) reported on radon ("Radiumemanation")
measurements in Switzerland. He already knew that the radon concen-
tration in the soil gas depends on various geological factors, meteorolo-
gical conditions like wind speed and on the soil moisture content !
Well seven decades later one started to realize that exposure to radon
may present a serious health problem and small scale radon surveys were
carried out in Switzerland in the early 1980s.
Alarmed by high values (up to 5 kBq/ra3 in living rooms) found in homes
in a city in the Western Swiss Jura Mountains (2,3) a task force was set
up to study the radon situation in Switzerland. This eventually has led to
a nationwide 5-year research program (RAPROS) that started in 1987.
It took some time to correct the then widely accepted but unproven
"facts" like : "high radon concentrations are mainly due to building
materials", "granitic bedrock shows a high uranium concentration and
therefore homes in the Alps have high radon levels", "there can't be high
radon concentrations in homes on Jurassic limestone".
Building materials and domestic water use showed to be a negligible
radon source in Switzerland (4,5), the main source being the soil.
Enhanced 22GRa have been found in various soils not of granitic origin and
the highest activity (880 Bq/kg dry weight) has been measured in a soil
covering Jurassic limestone.
We show the general radon situation in Switzerland and how we try to
gain representative exposure estimates from biased data. Geological
aspects of the radon problem are discussed. Mitigation techniques tested
in Swiss homes are presented and the policy for the new decade is
outlined. This policy is characterized by a concentrated search for radon
hot-spots.
GENERAL RADON SITUATION
FREQUENCY DISTRIBUTION AND AVERAGE RADON EXPOSURE
The frequency distribution of the radon concentrations in about 5000
rooms, corresponding to nearly 1600 buildings, representing 0.15% of the
6-32
-------
Swiss residential housing stock, is shown in figure 1. The radon levels
have been determined by exposing passive (etched-track) detectors for at
least two months. In general two detectors are placed per building, one in
the basement and one in an inhabited room at or above ground floor. We ask
people to use the rooms as usual in order to get radon concentrations
under realistic conditions. About 80% of the measurements have been
carried out during the winter. Few homes are represented by both summer-
and winter-values. These measurements show that summer levels are on the
average 1/3 lower than the winter levels.
The raw data in figure 1 are not representative for the radon exposure
of the Swiss for the following reasons :
1) Mainly in early surveys single family homes are overrepresented.
2) Certain regions are overrepresented due to particular
research programs like the search for radon sources in
the Jura Mountains (6) or because of the initiative of local
authorities.
3) Most measurements have been carried out during the winter
and thus don't give the annual mean.
To correct for bias 1) we sort the room data into building classes like
single family homes, blocks of flats, farms and "others". For multistory
buildings different stories (up to the forth floor) form separate classes.
For every class the number of radon values falling into a concentration
interval (subclass) is then multiplied by the percentage of the population
living in the respective class (1980 census data). Summing up the weighted
subclass contents over all classes leads to the new frequency
distribution. This first weighting is carried out for every canton (State
of the Swiss Confederation) or in the case of small cantons for a group of
cantons.
To correct for bias 2) the numbers in the subclasses of each canton
are multiplied by the percentage of the Swiss population living in this
canton. This frequency distribution having an arithmetic mean of 80 Bq/m3
is more representative for the radon concentration to which the Swiss are
exposed in their homes than the arithmetic mean of 140 Bq/m3 from the raw
data in figure 1.
The mean of 80 Bq/m3 still lacks the correction for bias 3) and for the
fact that few people stay at home 24 hours a day. We estimate that these
two factors lower the above 80 Bq/m3 to an annual mean of about 70 Bq/m3.
The Swiss map in figure 2 shows the geographical terms used.
REGIONAL DISTRIBUTION
In figure 3 we show the regional distribution of the 1540 buildings
that have at least one inhabited room at or above ground floor measured.
6-33
-------
99, %
99
O unirtfvatoited poo* at or below
prourta floor
144# V«lu«f
9)
U
e
3 ?0
•I
»•
•«
91
I
* »
v ? •«
%:
i«
40
30
se
10
• inhabited room at
or above around floor
3957 values
1 r
10 2«
-i 1 r
!• 1M 100
n 1 1 r
500 1000 2000 S000 10000
Rn-222 concentration [ Bq/n3 ]
Figure 1. Frequency distribution of radon neasurenen* results
CI TV OF C.
WESTERN
SWISS JURA
MOUNTAINS
SWISS
n'
COL DU
MARCHAIRUZ
¦*
PLATEAU
A
UPPER RHINE UALLEV
ALPS
iee kn
Lakes
Figure 2. Swiss nap. The dashed lines roughly represent the axis of
the respective geographic unit.
6-34
-------
©
#
©
0
©
GO
#
#
o
< s
O
S ~ 1®
o
10 - 20
o
20 - S0
o
50 - 100
&
100 - 200
/
Number of but Idi mgs
Measured per 840 Km^ region
fraction of
butldinfli
mith cone•
> 20O Bq^« 3
©
©
©
Figure 3. Regional distribution of radon measurement results
6- 35
-------
z
< 1
1 -5
fraction
of
buildmgj
5- 10
C per mile 1
Number of buildings
per 840 k«2
€ in thous-an
Figurt 4. Fraction of buildings measured
6-36
-------
We consider radon levels exceeding 200 Bq/m3 in this type of rooms as an
indicator for a possible radon problem. The fraction of homes with at
least one room exceeding this level is shown in this figure for each of
the regions. The division into regions is the one used for the 1:50000-
scale maps. Each rectangle measures 24 km times 35 km (840 km2).
There are at least two regions in Switzerland with clearly enhanced
radon concentrations : the Jura Mountains in the west and the Upper Rhine
Valley in the east. Geological aspects of the radon problem in these two
regions are discussed below. Enhanced levels are present in the south-
eastern part of Switzerland too. The Swiss Plateau where most of the Swiss
live is essentially free from radon problems.
The more than 1500 homes measured so far represent 0.15 % of the
residential housing stock in Switzerland. This may be sufficient to
calculate a Swiss average but as can be seen from figure 4 many regions
are not well represented. We don't really know what "well represented"
means. What percentage of homes has to be measured per 840 km2 unit until
one can declare it as "affected" or "safe" ? A hint comes from a recent
survey in the southern part of Switzerland (Ticino) that nearly doubled
the number of homes measured in this region. From a comparison of the
frequency distributions of the radon values before and after this survey
we conclude that a representative sample has to contain at least 1 % of
the residential buildings. Another hint comes from the now best covered
region ( 3.5 % of the 6000 homes are sampled) where we have been measuring
for more than 8 years. The frequency distribution changed slightly over
the years and now has become quite stable. We don't expect any surprise
from further measurements. We therefore recommend to sample 1 to 3 % of
the housing stock before any region can be declared as safe or affected.
GEOLOGICAL ASPECTS OF THE RADON SITUATION IN SWITZERLAND
There are mainly three factors that determine the radon risk of a
building ground : 1) 226Ra activity concentration in the soil, 2) Fraction
of the 222Rn produced that is available for transport (emanation) and 3)
Gas permeability of the soil.
We will show the range of values found in Switzerland for these three
factors and discuss geological aspects of two high risk sites.
RA-226 ACTIVITIES IN ROCKS AND SOILS
Uranium data for Swiss rocks, taken from a recent compilation by
Scharli (7) are shown in figure 5. The term "Uranium" used by Scharli is
somewhat misleading for the quantity measured has been the 222Rn daughter
concentration. He neglects any disequilibrium in the 23eu series down to
the 222Rn daughters. We therefore call his "Uranium" values 22fiRa taking a
conversion factor of 12.3 Bq/kg per ppm U. From this figure it is obvious
6-37
-------
fU IU IU - ITI
U 111
111
111
lll
u
u
111
o
in
U
u
U
U
111
y
u
w u
111
VI
111
111
HI
111
111
111
111
t- H
H
H
1-
»-
»-
K
»-
h
»-
h
h
H
>-
h 1-
1-
w
K
1-
K
1-
1-
H
M M
M
M
M
M
H
CO
M
M
M
M
>
M
M
H H
M
M
M
M
M
M
M
M
z z
J
z
z
*
ffi
z
o
o
»-
I
J
u
U h
_}
111
N
u
_l
z
o
z
ft!
c «r
o
c
111
o
o
o
N
M
o
X
111
M
Ł
c
O
>
o
O -1
a.
13
UJ
Ł
CD
H
_l
Ul
X
o o
o
o
w
z
o
o
o
u
13
U
X
o
O
w
¦J
I
U
w
K
Ą
N
z
Ł
N
>
z
(L
(L
X
O
Cf
0.
CL
CL
z
c
z
CL
ft!
c
a
C
Ł
tf
o
tt
o
o
_l
C
III
Ł
o
Ł
z
CL
W
IGNEOUS
-J METAMORPHIC
ROCKS
1000
500
100
S0
NUMBER OF VALUES
$ "
I I "] CONTAI
= =A= J MEAN I 68 X 0
I I J THE VA
NS
OF
VALUES
U|I||1||1||1|_I>K
ZI-l-ZZKC-f
OmmOOC->C
t-ZKh-t-XUW
in o ~ in in *
in -i > ~ k u
Z O I Z -I o
M o z C M tt
-i i ii in
SEDIMENTARY
ROCKS
is
2S
t nU ° u
t §8
u ez 5
SOILS —
a; c
u y
a.
(L UJ
D Z
Figure 5. Ra-226 activities in Swiss rocks and soils
6-38
-------
that "granite" is not synonymous with "high activity".
Activities in Swiss soils are shown in the figures 5 and 6. The 226Ra
concentrations in figure 5 have been determined by high resolution gamma
spectrometry on dried soil samples. The data given in figure 6 are from in
situ gamma spectrometry measurements (8). Contrary to the laboratory
measurements the poor statistics for in situ measurements exclude a
precise determination of the 23=U or the 234Th concentrations. The 23SU
contribution to the 186 keV 22SRa line has thus to be calculated assuming
perfect equilibrium down the 23Bu series. This leads to an underestimation
of the 226Ra activity in soils with a 23°Th ( and thus 22SRa) excess. A
23°Th excess is present in Jura Mountains soils. When comparing activities
in figures 5 and 6 one has also to take into account that the laboratory
data are for dried samples whereas the in situ values refer to the
undisturbed wet soil.
The complex nature of the Swiss geology and the important impact that
Quaternary had on our country makes it very difficult to find any
correlation between the regional activity distribution in figure 6 and
geological or tectonic maps. Soils in many parts of Switzerland are not
derived from the underlying bedrock. The most striking example is found in
the Western Swiss Jura Mountains where 226Ra activities of up to 880 Bq/kg
dry weight are present in soils covering Jurassic or Cretaceous limestone
having only about 20 Bq/kg of 226Ra.
A peculiarity of these soils is that 23°Th and 22GRa are largely in
excess of 23BU (determined quantity is 33*U), the latter being present in
"normal" quantities (30-50 Bq/kg dry weight). There is still no
explanation for this widespread anomaly. The watch industry, being very
prominent in the Jura Mountains has used large quantities of radiura-
activated luminous paint but we can hardly blame them for this
"contamination". The 226Ra and it's natural precursor 23°Th are nearly at
equilibrium even in soil samples taken close to a former radium processing
workshop. In samples of luminous paint from this workshop the 23°Th
activity is orders of magnitude lower than the 226Ra activity.
A hint for the origin of the enhanced activities may come from the
regional distribution of the 226Ra activity and it's dependence on the
altitude. There is a general trend for higher activities towards the
southwest (the main wind direction). Enhanced ( > 100 Bq/kg dry weight)
2zeRa activities are abundant at high altitudes (figure 7) and no Ra
anomaly has been found so far below about 900 m above sea level. This
altitude roughly corresponds to the upper ice margin of the Rhone glacier
during the latest glacial period. These two observation are consistent
with the idea by Pochon (9) of an aeolian origin of an important part of
the Jura Mountains' soils.
RADON EMANATION
The few emanation measurements on Swiss Plateau soils (mainly glacial
till) show that for these soils about 30% of the radon produced can escape
6-39
-------
fx
11
F i sure 6. Ra-22G and Ac—228 in Swiss soils.
Detprninpd by in situ ganni sp»ctronetrs
1BBB
800
O eastern slope
4- western slope
+ {
eee
408
200 —
0
_r
T
~r
+
+
T"
T
888 980 1888 1188 1208 1380 I486 1500
Altitude of sanpling point C neters above sea level J
Figure 7, Soil sanples Col du Marchairuz
6-40
-------
to the pore space and is thus available for transport. This fraction is
far higher (about 70%) for "high radium" soil samples from the Jura
Mountains. An example is shown in figure 8. The 2210~1om2 i.e. variations of at least four orders of magnitude have been
found in Swiss soils. The highest values show badly consolidated
rockslides.
High radon concentrations in the soil gas only present a radon risk if
the permeability of the soil is sufficiently large to allow for an
efficient radon transport to the foundation of a house. Even "normal"
radon levels in the soil gas may lead to a considerable radon risk if the
gas permeability of the soil is very high.
Therefore the product of the radon concentration in the soil gas times
the soil's gas permeability may be a better measure of the radon risk than
the bare radon concentration. This "radon availability" is plotted in
figure 10 for several regions in Switzerland. The envelopes have been
generously drawn around the respective data sets. Individual data points
are not shown in this figure. In the regions TI and FR we could not find
homes with high indoor radon concentrations whereas the regions RA and SI
are characterized by high indoor levels. Measurements in the region SI
include gas samples taken in unconsolidated rocks. Despite the large
scatter of the data points there is some evidence that building grounds
with radon availabilities larger than lO^Bq/m to 10~6Bq/m present a radon
risk.
GEOLOGICAL ASPECTS OF TWO HIGH RISK REGIONS
Western Jura Mountains, a Limestone Karst Region
As can be seen from the figures 5 and 6 22B Ra concentrations in the
soils of the Western Jura Mountains are on the average well above the
values for soils from the Swiss Plateau. At first sight this seems to
correspond well to the high percentage of increased indoor radon levels
found in this region (figure 3). But a closer look at the houses with high
concentrations shows that they are built directly onto the bare limestone
bedrock. The contact with the soil is limited to a less than 30 cm high
zone round the walls. More important than the soil 22
-------
500
i t 1 i
» t » i
t t i t
Ł
n
• w*
4i
3
71
L
T3
\
a
CO
400
300
200
71
~>
100
I I I I
0
L
I
O Ra-226
• Rn-222 daughter!
sanple sealed with epoxy
T
~r
T
10 28
Tine t days 3
"i r
"l r
30
40
sanple dried, stored in leaky container
Figure 8. Radon buildup in a Jura Mountains soil sanple
DRILLED HOLE
DIA. 7 CM
INFLATABLE
PACKER
PUMP
1000 ccn
SOIL
AEROSOL
FILTER
AEROSOL
FILTER
UALUE
TO LUCAS-
CELL
C188 ccn}
>,» V V ¦
K\' •
-OH:
GH
GAS ENTRANCE HOLES
PG
CYLINDRICAL PROTECTION GRID
BRASS TUBE
12 nn
Figure 9. Soil sas sanding eguipnent. To deternine soil perneability
we neasure the tine it takes to punp 1000 ccn at constant
pressure into the borehole. Constant pressure is produced
by the punp piston's own weight.
6-42
-------
I j I 111 II1
e. s
Paranittr >
Radon-availability I Bq/nJ
0.1
t—| i | i 111 | | 111 1—j i | i 111 | | ) 11 r—| i | i |11 | j 11
5 IB 58 188 588 1888 2880
Gas perneabilitu of -the soil C 18"^^ ]
Figure 18. Radon availability in Swiss soils
PERCOLATING HATER
SOIL
BEDROCK
CKARST)
HOUSE
RADON DEGASSING FROM PERCOLATING HATER
Figure 11. Proposed radon transport in • ktrit systtn
6- 43
-------
of the buildings there is even a visible connection to the karst system.
The radon concentrations in caves below this city are very high (up to
40 kBq/ra3 (10,11)). In combination with the high gas permeability this
karst system represents a very powerful radon source. Even small
connections to this source are sufficient to supply large quantities of
radon to the basement of a house.
There remains to explain the high radon levels in the air of the karst
system. The Jurassic limestone contains only about 20 Bq 226Ra/kg, by far
not enough to sustain 30 to 40 kBq/ra3 222Rn in the cave air. We have
therefore proposed (12) that percolating water is transporting large
quantities of radon from the (high 226Ra) soil to the caves (figure 11).
Rockslides in the Alps
There are many villages in the Swiss Alps built onto badly consolida-
ted rockslide debris. In the Upper Rhine Valley these rockslides contain
"Verrucano". In Switzerland the term "Verrucano" means an old clastic
sediment frequently showing enhanced 22SRa concentrations. 226Ra values in
soils from the Upper Rhine Valley are shown in figure 5. The combination
of relatively high 226Ra activities with the extremely high gas permeabi-
lities in these rockslides seems to be the reason for high indoor radon
concentrations. Contrary to the situation in the Jura Mountains the homes
are built onto the "high 22SRa" material.
MITIGATION
Remedial actions are still in a test phase. Homeowners willing to
participate in pilot projects have been offered a substantial financial
support by the Federal Office of Public Health that also plans and
supervises the work.
Pilot projects carried out so far have shown that passive methods like
sealing floors are insufficient. Combining sealing with subfloor suction
has lead to the successful mitigation of several homes at still reasonable
costs.
The most dramatic reduction (to nearly outdoor radon levels) has been
achieved by an air conditioning system that allows for the control of air
flow and pressure in the basement. A heat exchanger keeps the energy
consumption low. This installation is for research only. It is too
expensive for a general use but a scaled down version may give comparable
results at reasonable costs.
In a high risk area a future homeowner could be convinced to install a
subslab suction system. We hope that he will have considerably lower radon
levels in his new home than his neighbour living with the highest radon
6-44
-------
concentration ever measured in Switzerland (45 kBq/m3 in winter).
In Switzerland the mitigation technique has not yet passed to the
private sector. There is no "radon business" in our country for there is
no real public concern about radon. This may be partly due to the lack of
limits or recommendations for indoor radon levels but distinctly more
important is the general feeling that something natural like radon can't
be harmful.
THE NEW DECADE
in the 1980s we have gathered enough data to make a reasonable
estimate for the average radon exposure of the Swiss. This average will
change only slightly even if we could double the number of buildings
measured. What we need more is to find the homes with extreme values,
homes really worth remedial actions. Therefore any new survey will
concentrate on the search for high risk regions. This search will be
guided by the knowledge gained on the correlation between geology and
radon concentration. A survey started in November 1990 in the eastern part
of Switzerland has already been planned according to this new concept.
Etched-track detectors are placed in villages on high permeability grounds
(rockslides, karst, clean coarse gravel with low lying water table,
important fault zones) and/or close to known or suspected uranium
mineralizations.
The new decade will also see recommendations or even limits for safe
indoor radon concentrations and an increased engagement and responsibility
of local authorities. The federal government will concentrate on research,
the search for high concentrations, scientific support and quality
control.
We will plead for sensible radon concentration limits. Radon is only
one of the many carcinogenic substances present in our environment.
The work described in this paper was not funded
by the U.S. Environmental Protection Agency and
therefore the contents do not necessarily reflect
the views of the Agency and no official endorsement
should be inferred.
REFERENCES
1. Gockel, A., Ueber den Gehalt der Bodenluft an radioaktiver
Emanation, Phys.Zeitschr., 9 (1908) 304
6-45
-------
2. Lauffenburger, T. and Auf der Maur, A., The Concentration
of Radon in a Town where Radium-Activated Paints were
used. Jfo: Proceedings of the 6 th Int. Congress of IRPA,
Berlin (West), May 1984
3. Surbeck, H. The Search for Radon Sources, a Multi-
disciplinary Task.
Rad.Prot.Dosim.,24,1/4 (1988) 431-434.
4. Buchli, R., Burkart, W., Correlation among the terrestrial
gamma radiation, the indoor air Rn-222 and the tap water
Rn-222 in Switzerland. Health Phys. 57, (1989) 753-759.
5. Schuler, Ch., Crameri, R., Burkart, W., Assessment of the
indoor Rn contribution of Swiss building materials.
Health Phys. (1991), in press.
6. Surbeck, H. and Piller, G. A Closer Look at the Natural
Radioactivity in Soils. In : Proceedings of The 1988
Symposium on Radon and Radon Reduction Technology,
Denver CO, October 17-21, 1980, Environmental Protection
Agency Report EPA-600/9-89-006, Research Triangle Park, 1989
7. Scharli, U., Geothermische Detailkartierung (1:100000) in der
zentralen Nordschweiz, mit besonderer Beriicksichtigung
petrophysikalischer Parameter,
Thesis, ETH-Diss.Nr.8941, Zurich 1989
8. Murith, C., Volkle, H., Surbeck, H., Ribordy, L., In situ
gamma spectrometry in Switzerland using a portable gamma
ray spectrometer. In : Feldt, W. (ed.) Proc. XVth Regional
Cong, of IRPA, Visby, Gotland,Sweden,10-14 Sept. 1989, 389-394
Fachverband fur Strahlenschutz, ISSN 1013-4506
9. Pochon, M. Origine et evolution des sols du Haut-Jura
suisse. Memoires de la Societe Helvetique des Sciences
Naturelles. Vol.XC. 1976
10. Piller, G. and Surbeck, H., Radon and Karst,
in : Feldt, W. (ed.) Proc. XVth Regional Cong, of IRPA, visby,
Gotland,Sweden,10-14 Sept. 1989,p.15-20
Fachverband fur Strahlenschutz, ISSN 1013-4506
11. Rybach, L., Medici, F., Surbeck, H., Geological aspects
of radon exposure in Switzerland, In: Proc. Colloque Int. sur
la Geochimie des Gazes, Mons, Belgium, October 6-13, 1990,
in press
12. Surbeck, H., Medici, F., Rn-222 transport from soil to karst
caves by percolating water. In : Proc. of the 22nd Congress
of the IAH, Lausanne, Switzerland, August 27 - September 1,
1990, in press
6-46
-------
A CROSS-SECTIONAL SURVEY OF INDOOR RADON CONCENTRATIONS
IN 966 HOUSING UNITS AT THE
CANADIAN FORCES BASE IN WINNIPEG. MANITOBA
D.A. Figley & J.T. Makohon
Saskatchewan Research Council
Building Science Division
Saskatoon, SK S7N 0W9
CANADA
Acknowledgements: The authors would like to acknowledge the assistance of
1). Harvey-Smith, Project Manager, Department of National
Defence for his valued assistance and comments during the
course of this study.
ABSTRACT
This paper summarizes the results of a cross-sectional survey of indoor radon
concentrations in a total group of 966 housing units at the Canadian Forces Base (CFB)
in Winnipeg, Manitoba. The major objective of the study was to characterize the
distribution of indoor radon levels in the housing group as the first step in
developing a radon control strategy. Subsequent investigations on sub-groups of these
houses (not reported here) were conducted to examine the building factors associated
with the indoor concentrations and the efficacy of post-construction control measures.
Measurements were obtained from 670 of the 966 housing units (69% participation) .
The study group was composed of large numbers of nominally identical housing units of
several different building styles. The two-day average measurements were taken using
charcoal canisters during extremely cold weather, -28°C to -35°C. A short
questionnaire administered to the occupants by the field workers who installed and
removed the canisters recorded basic data on occupant activities and building factors.
For the entire group, the geometric mean concentration was 112 Bq/m3 (3.0 pCi/L),
approximately twice as high as the geometric mean obtained by an earlier summertime
study of 563 Winnipeg houses. Data was subgrouped based on geographic location within
the city, and the subgroup geometric mean concentrations varied between 25 and 206
Bq/m3 (0.7 and 5.6 pCi/L). Individual house measurements ranged from <10 Bq/m3 to
>5400 Bq/m3 (<0.3 pCi/L to >146.0 pCi/L). No building or occupant factors were
initially identified as being associated with the variation in levels.
6-47
-------
INTRODUCTION
In the fall of 1989, the Department of National Defence (DND) retained the authors
to design and conduct a study to investigate the indoor radon levels in the residences
occupied by DND personnel at the Canadian Forces Base (CFB) Winnipeg, MB. These
residences included housing units (PMQ's) owned by DND, bulk leased (BL) housing units
rented by DND, and barracks units (BU). Radon levels were also surveyed in 46 areas
of the officers' and non-commissioned members' messes and other occupied areas on both
north and south areas of CFB Winnipeg.
Radon has been identified as a naturally occurring pollutant that is broadly
distributed throughout Canada. In 1977 to 1980, Health and Welfare Canada conducted
a study to survey indoor radon levels in 14000 homes in 19 cities across Canada (1).
These data are frequently referred to in discussions regarding the radon situation in
Canada and are used to rank cities with respect to their radon risk potential. In this
study, Winnipeg was identified as the Canadian city having the highest geometric mean
indoor radon level (57 Bq/m3) based on a sample of 563 houses. For the purposes of
this paper, the conversion 37 Bq/m3 = 1 pCi/L can be used.
Many studies of indoor radon levels have been conducted and while a more complete
understanding of the factors that influence indoor levels is emerging, at present, the
only reliable method of estimating the radon concentration in a specific building is
to measure it (2).
The study design included three parts to be conducted consecutively:
Part 1. Cross-Sectional Survey of Indoor Radon Concentrations.
The focus of this part of the project was to provide an overview screening of Lhe
radon concentrations occurring in the homes. The data would also provide a statistical
database for future studies. DND requested that all of the occupants of both the owned
and leased housing units be given the opportunity to participate in the study. In an
attempt to obtain the highest indoor readings, measurements were taken in the lowest
levels of the houses during calm, cold weather.
Part 2. Detailed Engineering Study of a Selected Sub-group of Houses.
This work focused on identifying the building factors that influence indoor radon
concentrations and provided information for the development of mitigation techniques.
It included a more intensive study on a sub-group of approximately 40 houses identified
in the part 1 work as having the highest and lowest indoor radon concentrations.
Part 3. Mitigation Study on a Small Group of Houses.
The focus of this work was to select five houses with high Indoor radon
concentrations, make building modifications and evaluate the impact of the
modifications on the indoor radon concentrations.
All houses having part 1 screening levels >150 Bq/m3 had alpha track monitors
installed for the period from October 90 to March 91.
6-48
-------
This paper deals with the results of the cross-sectional survey (part 1) of the
study.
For the purposes of data presentation and discussion, the current Manitoba
Government interim guidelines (3) are referred to in the report. In summary, these
are:
1) If the screening measurement is about 150 Bq/m3 or lower and was taken during
cold weather with the house closed up, there is little chance that the home
will have an annual average concentration greater than 150 Bq/m3. Follow-up
measurements are probably not required.
2) If the screening measurement is about 150 - 800 Bq/m3, consider performing
follow-up measurements.
3) If the screening measurement is about 800 Bq/m3 or greater, perform long term
(minimum three months) measurement as soon as possible.
OBJECTIVES
Part 1 of the study had two major objectives:
1) Measurement of indoor radon levels in all CFB Winnipeg residences (PMQ, BL
and BU) and selected other buildings to give DND an accurate assessment of
the current indoor radon levels. These data would be used to determine if
additional measurements or mitigation work were required to ensure indoor
radon levels were maintained below levels established by DND.
2) Characterization of the distribution of radon levels and analysis of the
levels in conjunction with selected building and occupant factors. The
analysis would identify factors that are statistically associated with the
radon concentration and will be used in subsequent phases of work.
STUDY DESIGN
The initial phase of the study was a cross-sectional survey to measure the two day
average concentration in (nominally) all of the 966 residential units potentially
inhabited by base personnel. CFB Winnipeg engineering staff also prepared a list of
17 buildings to be monitored. A total of 46 monitors were placed in various locations
in the lowest levels of these buildings.
Prior to conducting the study, thti base command prepared an information package
containing basic information about radon and a brief overview of the proposed study
which was mailed to all occupants of homes in the study. Only homes that were occupied
during the test period were monitored since gaining access to homes where the occupants
were not present was not permitted. Participation in the study was at the discretion
of the occupant.
6-49
-------
The study consisted of an Initial home visit to install the radon monitor and a
foliow-up visit 48 hours later to remove the monitor and complete (with the homeowner's
assistance) a short questionnaire concerning the building construction and occupant
activities. If the homeowner Indicated that they would not be home when the monitor
was to be removed, they were carefully instructed as to the protocol for repackaging
the monitor. The homeowner would leave the monitor in the mailbox for pick-up.
Twenty-three temporary contract employees were used over a four day period to
install and remove the monitors. They were all paid on an hourly basis and instructed
to take as much time as necessary to complete each house visit (average 15 minutes).
On the day before the monitoring began, a two hour seminar was held to train all of
the personnel assisting with the study. A phone-in help line was manned at all times
so that workers could phone in for assistance. Only five calls requesting minor
information were received during the study.
The field work was conducted from 10-14 December, 1989. During the test period,
the weather was clear and relatively calm with the outdoor air temperature varying
between -28°G to -35°C.
Additional monitors and questionnaires were available at the base engineering
office for persons who phoned in to say they were in the city but would not be home
when the visits were being made. These people were invited to come to the engineering
office to pick up the materials for self administration. These data are not included
in this report.
For this survey, the sample population and the target population were identical
since all residences were included in the survey. Considerations as to sample size,
representativeness of sample and estimation of the distribution of indoor radon levels
are eliminated in a total sampling program. This is an important point in designing
radon research projects since the nature of radon concentration distributions varies
widely depending upon local circumstances.
The following potential biases may affect the study, however, they are not
considered significant in the analysis.
Although all of the residences occupied by base personnel were included in the
survey population, some houses were not monitored. For most cases, the reason for not
being included was that the occupants were not home at the time the house was initially
visited (between the hours 8:00 to 21:00 Monday or Tuesday). Several attempts were
made at various times of the day over the two day period.
The non-participants may bias the selection of the data group towards residences
where a co-operative individual was home, however, there does not appear to be any
systematic reason why this would affect the validity or interpretation of the study
results. The demographic and building data obtained from the questionnaire would
correctly account for these occupant differences. Of the 966 potential residences,
670 measurements were obtained.
The questionnaire contained primarily descriptive and quantitative questions
concerning the building and the occupant activities during the two day monitoring
period. A section for general homeowner comments was also included.
6-50
-------
METHODOLOGY
The indoor radon concentration was measured using RADPAC TM activated charcoal
canisters. The nominal exposure time suggested by the supplier was two days, however,
as long as the exposure time was accurately recorded, exposures in the range of 45 to
72 hours were acceptable.
The canisters were installed approximately 0.6 m above the floor in the lowest
level of the residence, centrally located away from drafts and in accordance with the
manufacturer's instructions.
All of the canisters were received by the supplier for analysis within 48 hours
of removal from the house.
The monitor supplier was listed as a registered participant in both the US EPA and
Health and Welfare Canada quality assurance programs. In 10 locations, duplicate
canisters were installed as an internal check of the measurements.
QUESTIONNAIRE DESIGN
The questionnaire was designed to be either self-administered or filled in by the
survey employee with assistance from the homeowner. It consisted of 36 questions
requiring;
1) a yes or no response about building characteristics.
2) basic physical information about the building such as the number of windows,
main floor area or type of space heating system.
3) selection of a ranked descriptor to rate the condition of the foundation walls
and floor.
4) estimation of hours spent doing specific household activities or frequency
of door/window openings.
The purpose of the questionnaire was to obtain information on building and
occupant factors that would influence the indoor radon concentration either directly
or indirectly.
ANALYSIS
The survey yield for the entire housing group was 670 measurements from a total
population of 966 (69%). Since the geographic location was considered to be an
important factor associated with the indoor radon concentration, the data were sub-
grouped 1 to 8 (somewhat arbitrarily) based on location.
6- 51
-------
The sub-groups used in the initial analysis were combinations of streets based on
a common geographic area. For the sub-groupings used, the smallest yield was 62% of
all possible houses so all of the areas were considered to be adequately sampled.
Information from the questionnaire can be used to group the houses into different
categories. Most of the questionnaire information can not be directly assigned a
quantitative value that could be used in a mathematical analysis, but will be useful
in identifying houses that can be grouped together on the basis of some common
characteristics and compared with respect to other factors.
RESULTS
The authors or project manager should be contacted for information on the detailed
survey results.
A frequency distribution of all of the indoor concentration data is presented in
Figure 1 and replotted with the logarithm of the concentration in Figure la. All
logarithms are taken to the base 10. The distribution in Figure la follows the log-
normal distribution and therefore, the geometric mean (GM) rather than the arithmetic
mean is used to describe the central tendency of the data. For the entire group, the
geometric mean indoor radon concentration was 111.8 Bq/m3 (123.1 Bq/m3 for the housing
units only) which was well above the 57 Bq/m3 geometric mean for Winnipeg given by the
Cross-Canada study.
The class intervals for the histograms were selected to allow group frequencies
corresponding to the Manitoba guideline values to be calculated.
A second set of frequency distributions were prepared by breaking the population
into six geographic areas (somewhat arbitrarily) based on the physical proximity of
groups of streets Figure 2. Also included are the separate data from the north and
south base building areas. The data are presented in Table 1.
The geometric means, arithmetic means and standard deviations for the geographic
sub-group of data are given in Table 1. These values show the wide variation in mean
radon concentrations both within the groups and based on location. It is important
to note that the grouping based on location is, in some cases, a surrogate grouping
based on other building factors such as house style or ventilation system type. The
sub-groups also include buildings owned and maintained by DND and bulk leased housing
that is owned and maintained by the leasing company.
In future analysis (beyond the scope of this report) the data for the individual
geographic locations (where applicable) may be further sub-grouped based on the general
retrofit status of the detached and semi-detached houses. Over the past years,
different levels of improvements have been made to the DND housing stock. The present
housing stock falls into one of the following categories:
1) original construction as built in the 1940's.
2) replacement of windows and doors with more modern units.
6- 52
-------
3) replacement of windows and doors along with reinsulation and siding of the
exterior walls above grade.
Other possible analysis include examination of the effect of house style,
heating/ventilating system type, foundation type and foundation condition.
DISCUSSION
The initial screening survey indicated that there is a wide variation in radon
levels in the buildings occupied by CFB Winnipeg personnel.
Overall, the indoor radon concentrations are much higher than the 57 Bq/m3
geometric mean level obtained in the Cross-Canada study by Health and Welfare Canada.
To a large extent, this may be related to the test conditions under which the
measurements were taken. The Cross-Canada study used short term (<10 minute)
measurements conducted in the summer. For this study, two day averages during sealed
house conditions in very cold weather were taken. While detailed modelling is beyond
the scope of this study, several building science principles support these results:
1) The cold outdoor temperatures would result in high sustained negative
pressures at the lower level of the buildings. This would maximize the
pressure potential driving radon into the buildings.
2) Although the high negative pressures should result in an increase in the air
exchange rate for the houses, a concerted effort on the part of the homeowners
to keep all windows and doors closed (as compared to summer when children are
home from school and window/door opening may provide the only cooling
ventilation) may have offset the pressure effect and resulted in lower overall
outdoor air exchange rates. Many of the homeowners reported taking special
care to keep their homes "sealed up" during the winter to minimize drafts and
reduce heating costs.
The groups with the lowest geometric mean indoor radon levels (25-30 Bq/m3) were
the south and north base buildings - groups 7 & 8 and the two storey six/eight family
units in group 5 located adjacent to the north base. All of these buildings had hot
water heating systems and no mechanical ventilation systems.
All of the other groups were single or double family residences. While a detailed
analysis is not provided, there is a general tendency for the north base areas to have
higher geometric mean indoor radon levels (group 1 - 206.5 Bq/m3, group 2 - 173.4
Bq/m3, group 4 - 147.0 Bq/m3) than the south base areas (group 3 - 110.2 Bq/m3, group
6 - 156 Bq/m3) .
Table 2 lists the values for the replicate measurement tests. For the ten
locations, two charcoal monitors were placed side by side and exposed for the same time
period. In nine cases, the agreement was within a maximum range of 16.7% and typically
much smaller. For the test at Location E, the monitor results varied by a factor of
six. There are no procedural differences that would account for this anomalous result.
Using a paired t-test analysis, the differences between the measured values (excluding
Location E) were not significant at the 5% level of significance.
6-53
-------
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.
Although this project was funded by the Canadian Department of National Defense,
DND does not endorse the products or techniques used by the authors.
REFERENCES
1. Letourneau, E.G., McGregor, R.G., Walker, W.B., "Design and Interpretation of Large
Surveys for Indoor Exposure to Radon Daughters", Radiation Protection Dosimetry,
Vol.7, No.1-4.
2. "Radon Reduction Techniques for Detached Houses - Technical Guidance, Second
Edition", United States Environmental Protection Agency, Washington, DC., January,
1988.
3. "RADON - An Interim Guide for Manitoba Homeowners", Manitoba Energy and Mines
Information Center, Winnipeg, MB., 1989.
4. Health and Welfare Canada, "A Radon Guideline for the Department of National Health
and Welfare".
6-54
-------
TABLE 1. INDOOR RADON CONCENTRATION DATA FOR GEOGRAPHICAL SUB-GROUPS
GROUP
NO. OF
HOMES
NO. OF
MEAS.
GEO. MEAN
Rn (Bq/m3)
ARITHMETIC
MEAN (Bq/m3)
NO. >
150 (Bq/m3)
NO. >
800 (Bq/m3)
1
214
145
206.5
319.1
88
10
2
180
111
173.4
206.3
68
1
3
243
181
108.2
183.0
47
6
4
106
65
147.0
200.8
34
1
5
105
77
25.7
29.2
0
0
6
118
91
152.1
266.7
40
4
7
-
35
27.2
47.0
2
0
8
-
11
27.8
35.0
0
0
6- 55
-------
TABLE 2. COMPARISON OF REPLICATE MEASUREMENTS
r— 1 i i
| Location
1
RADPAC
Bq/m3
% Diff.
|Location A
125.8
125.8
0
|Location B
J
88.8
88.8
0
|Location C
103.6
99.9
3.6
|Location D
55.5
66.6
16.7
|Location E
140.6
806.6
82.6
|Location F
388.5
388.5
0
|Location G
299.7
299.7
0
|Location H
629.0
629.0
0
|Location I
510.6
503.2
1.4
|Location J
92.5
88.8
4.0
6-56
-------
1 70
1 60
1 50
1 40
1 30
120
1 1 0
100
90
80
70
60
50
40
30
20
10
0
/¦
rir
Y'V
y a'a
A
A
one measurement off scale
5487 Bq/m^
vV/\
A A
I
-1
tn
n
innn.
; I 11 1 i 1 1 "r".' i 1 ' | ¦ . ¦ | . . ¦¦¦ j ' i 1 1 > i 111 i 1 1 ' y ' i ' | i i i j
400 600 800 1000 1200 1400 1600 1800 2000 2200 2400 2500 2800 30C
200
RADON CONICLN TRA1 ION, Bq/m3
FIGURE 1. FREQUENCY DISTRIBUTION OF ALL INDOOR MEASUREMENTS
-------
190
180
170
1 60
150
1 40
1 30
1 20
1 10
1 00
90
80
70
60
50
40
30
20
10
0
0 0.25 0.5 0.75 1 1.25 1.5 1.75 2 2.25 2.5 2.75 3 3.25 3.5 3.75 4 4.25 4.5 4.75 5
LOG RADON CONCENTRATION, Bq/-m3
FIGURE 1A. FREQUENCY DISTRIBUTION OF ALL INDOOR MEASUREMENTS (LOGARITHM)
-------
City limits —
•o
D.
Highway #1
MILES
City limits
FIGURE 2. MAP OF WINNIPEG SHOWING EIGHT SAMPLING GROUP
-------
RAJJQN STUDIES IX BRITISH COLUMBIA, CANADA
D.R. Morley1, M.M. Ghomshei2, C. Van Netten3 and B.G. Phillips1
1 - Province of British Columbia. Ministry of Health, Radiation Protection
Service, 200 - 307 West Broadway, Vancouver, B.C. V5Y 1P9, Canada.
2 - Orchard Geothermal Inc. 500 - 342 Water Street, Vancouver, B.C. V6B 1B6,
Canada.
3 Faculty of Medicine. The University of British Columbia, 5804 Fairview
Avenue, Vancouver, B.C. V6T 1V5, Canada.
ABSTRACT
Three radon studies, involving 150 background gamma measurements and
long-term alpha track tests in a total of 400 homes have been conducted in
three geologically distinct areas of the province of British Columbia.. A
positive correlation between the background gamma radiation and the measured
radon level can be depicted only at the regional scale.
In the Coastal Area where the terrestrial gamma radiation is low, no
homes were found to exceed 4 pCi/1 on the main floor. In the Kootenays, where
the background gamma is relatively high, considerably higher radon levels were
encountered. The highest radon levels were encountered in the West Kootenay
where the terrestrial gamma level is comparatively higher than East Kootenay.
In West Kootenay about 45% of the homes demonstrate radon levels above 4 pCi/1
and '/% above 20 pCi/1 on the main floor. In the same area more than 60% of
the homes demonstrate radon levels above 4 pCi/1 in the basement. In the East
Kootenay, where the radon levels were found to be lower than in West Kootenay,
the terrestrial gamma levels are generally lower.
In some areas, the age of the house and the combustion air supply system
seem to correlate with the radon level. In West Kootenay, where the highest
radon levels were encountered, a positive correlation was observed between the
radon level in the basement, and the age of the house. The correlation was
negative for the main floor (possibly due to air circulation between main
floor and the basement in the new houses). In both Kootenay areas, the homes
with combustion air supply from outside demonstrated reduced radon levels on
the main floor (roost probably due to increased inside pressure!.
6-61
-------
INTRODUCTION
Surveys of radon gas in homes have been carried out. in Canada since
1977(1). These early studies, although they covered over 14.000 homes, were
baaed upon single grab measurements normally in the basement of the house
during summer and therefore gave poor indications of the annual exposure(2)
and the correlation between annual exposure and construction parameters. The
British Columbia (B.C.) Ministry of Health began making terrestrial gamma ray
measurements in 1980 using both Thermoluminescent Dosimeters (T.L.D.s) in 25
locations(3.), and a portable high pressure ionization chamber (Reuter Stoke
RSS-111) in 150 locations. The areas of higher gamma activity generally
corresponded with rock structures where uranium is likely to occur(4). We
found the province could be divided into three gamma background areas. The
coastal region of the province has very low gamma background, a moderate or
normal gamma background regions that is located in the interior of the
province: and an elevated gamma radiation area of the province which is
scattered about the interior and associated with areas favourable for uranium
deposits. Figure 1 shows the province of British Columbia and the three areas
where our, long-term radon surveys were carried out during 1988 - 89.
t
YUKON
ALASKA
t
1
t
t
BRITISH
ALBERTA
1
\
\
1
i
f
COLUMBIA
\
I
«
\
I
J
200 km
Vancouver * Castle^ ^Crajibrook
-T~
t
WASHINGTON
MONTANA
1 IDAHO (
Figure 1 - British Columbia with the locations of 3 radon surveys
6- 62
-------
The first long -term radon study was carried out in the town of
Cactlegar. the West Rootenay region of the province. A previous radon grab
aampIs study indicated elevated radon levels in many basements there(5). The
region has an elevated gamma background and there has been uranium exploration
in the area. The second study was carried out in the East Kootenav
(Cranbrook) region of the province. Moderate gamma radiation levels had been
detected in the region. However .just south of the region, moderately elevated
radon levels had been measured in Montana(6). The third study was conducted
in the Greater Vancouver Region of the coastal British Columbia low gamma
background region. In a previous grab sample study(l) by Health and Welfare
Canada in this area, only low radon concentrations were found.
This paper will compare the data obtained in these three long-term r=adon
surveys, the terrestrial gamma surveys, other geological and construction data
available. This is the first step in developing a good potential radon risk
model for homes in this province.
MKTHODOIDGY
RADON SURVEYS
The first survey was conducted in Castlegar, in the West Koot.enay area
of B.C. The homes are located on glacial terraces created by the Columbia
River. The soil is dry gravelly and permeable. The monitors were installed
in July of 1987 .and removed in March of 1988. This period was representative
of the observed annual weather pattern. 74 homes were monitored (73 homes
returned monitors). All but one home had an upstairs and a downstairs (or
basement) monitor. The monitors were mounted 4-7 feet above the ground,
away from drafts and placed in an area where the family commonly resided.
Measurement of the terrestrial gamma were made at each house, usually outside
on the front yard.
The second survey was conducted in the East Kootenays where 157 of the
160 monitors were recovered. The monitors were placed one per house. In this
study, the owner decided if the monitor should be placed in the basement or
upstairs. The monitors were again placed 4-7 feet above the ground, away
from draft and in a central living area. They were installed in January 1988
•and removed in July 1988. It was our observation that this period represented
an average annual weather conditions. Mo3t of the monitors were placed in
Cranbrook, the principal community in the area but some were placed in the
nearby communities of Fernie, Invermere, Kimberley, Creston and Golden. Those
communities are located in valley bottoms at the foot of the Rocky Mountains.
Hie third set of 140 monitors was distributed in Greater Vancouver area.
135 monitors were recovered. Although the terrestrial gamma background is
consistently low the geology varies from the Eraser River delta (rich farming
land) to the North Shore mountains and includes bed rock and glacial out-wash.
One monitor was placed in the main living area of each home 4-7 feet above
the floor and away from drafts and corners. No terrestrial gamma measurements
were made in Greater Vancouver in this study since previous studies had
detected no significant difference from one home to another in this area. The
radon monitors were placed in January 1988 and removed in August 1988. This
period was observed as representative of the average annual weather pattern.
Similar weather patterns have been observed in previous years by Ghomshei et
6- 63
-------
al(6) during their radon stadies.
At the time of monitor placement, a questionnaire was completed by the
surveyor. Information was collected on the location of the monitor is;. house
construction, age of home, home occupancy, basement or slab construction,
possible radon pathways, heating and ventilating systems and the geological
environment. Surveyors were instructed on required procedures prior to going
out so that survey consistency would be maintained.
ANALYTICAL PROCEDURES
Terradex alpha tract radon detectors were obtained from Landauer Inc.
They were type DRN and had a detection limit of about (.4 pCi/1) - month.
Terrestrial gamma measurements were made using a Reuter-Stokes RSS-111
Environmental Radiation Monitor. The monitor's gamma ray response extends
from .060 MeV to above 8 MeV. Correction for cosmic rays was made by
recording the barometric pressure and subtracting the corresponding cosmic ray
component, as specified in the operators manual for the Reuter Stokes
iustrument.
RESULTS AND DISCUSSION
TERRESTRIAL GAiflA MEASUREMENT
TerreatriftI gamma radiation level3 were determined in 150 areas of the
province (2 to 100 measurements/area). The province (Figure 1) can be divided
into 3 regions of terrestional radiation intensity. The first or low
background area is the coastal strip composed of the two tectonic belts which
were most recently rafted into North America to build the province. The
second, moderate terrestrial radiation area, composes much of the interior
area of the province. Within this interior area are large areas of high
terrestrial background. These areas correspond to the areas identified by the
British Columbia Ministry of Mines as being favourable environments for
uranium deposits(7).
There is a good correlation between average terrestrial gamma background
and average radon concentration (see Figure 2). This however does not- follow
through to the individual homes. There was no correlation between the
individual homes terrestrial gamma intensity and the radon levels found in the
basement or upstairs. The elevated terrestrial gamma was not the only
indicator of potential radon problems in communities. In the Caatlegar, West
Kootenay area the soil was dry, gravelly, and permeable while in the East
Kootenay area a number of communities (Cranbrook and Creston) were underlain
with clay which appears to retard radon migration.
6- 64
-------
Radon Cone, in pCi/l
10
Castlegar
Cranbrook
Vancouver
80
40 60
Gamma ray intensity in mR/y
20
Figure 2 - Main Floor Radon Levels as a Function of
Terrestrial Gamma Radiation
CASTLEGAR, WEST KOOTENAY ARM
All but one of the homes surveyed in Castlegar had two levels. The land
was strongly sloped and the lower levels were sunk into the ground on at least
3 sides of the home. Most of the second level was located above ground. The
average main radon level was 6.5 pCi/l and the average basement level was 10.6
pCi/l. A low pass filter was applied to smooth out some of the fluctuation in
the data (figure 3). As can be seen from figure 3, about 45% of the homes
demonstrated radon levels above 4 pCi/l and 7% were above 20 pCi/l on the main
floor. More than 60% of the homes had radon levels above 4 pCi/l in the
basement. Approximately 15% of the homes had higher radon concentrations
upstairs. In these homes fresh air entering at the basement level was
probably diluting the radon in the vicinity of the monitor. The age of the
home had a marked influence on the radon concentration (Figure 4j. Older
homes can be characterized as having poorly constructed basement foundations
with a doorway sealing them off from a leaky upstairs. New homes, although
they have better constructed basements, have open stairways, an occupied
basement, a central heating system, and a better sealed housing envelop.
Although les3 radon enters the newer building, it gets distributed over both
floors and is retained there. There was no direct correlation between
upstairs and downstairs radon levels. If a home was supplied with make up
combustion air the average radon level was downstairs 10.8 +./- 10.5 and 5.4
+/- 6.0 pCi/l upstairs, if no combustion air was supplied the average radon
level downstairs was 10.2 +/-- 12.0 pCi/l and upstairs was 8.2 +/- 10.5 pCi/l.
Although the evidence is not strong it appears that the combustion air supply
may reduce the negative pressure in the home reducing radon infiltration.
6-65
-------
00
90
80
70
60
50
40
30
20
10
0
30 35
40 45
50
10 15
20 25
5
Radon in pCi/1
Figure 3 - Distribution of Main Floor and Basement Radon Values (Castlegsr, B.C.)
RADON CONC. IN pCl/l
14
0-5 YRS 6-10 YRS 11-20 YRS 21-30 YRS 30+ YRS
AGE OF HOME
¦¦ MAIN FLOOR K? BASEMENT
Figure 4 - Radon in Castlegar Homes as Function of Age
6- 66
-------
KAST KOOTKNAY DATA (CRANBROOK AND VICINITY)
The East Kootenay area had, unlike the West Kootenay, a wide variety of
housing types. Table 1 gives the average radon values found in the region.
Terrestrial gamma levels are relatively constant throughout the region. It is
interesting to note that Creston and Cranbrook are underlain with clay. This
clay is not permeable .and presents problems when septic fields are
constructed. The other near-by communities are located on rocky and coarse
soil which is much more permeable. These communities have higher radon
levels. No clear trend was detected when comparing the age of the houses with
radon concentration. This may be because both upstairs and downstairs
measurements were not conducted in each home. Combustion air intake however
does appear to have some impact on radon levels. If combustion air is
supplied the average radon concentration was 1.56 +/- 2.90 pCi/1 on the main
floor and 1.6V +/- 1.82 pCi/"l in the basement. If combustion air was not
supplied the average radon concentration was 2.29 +/-2.75 pCi/1 on the main
floor and 2.29 +/- 3.36 pCi/1 in the basement. 10% of the main floor radon
levels exceeded 4 pCi/1 and 1% exceeded 20 pCi/1.
6-67
-------
Table 1 Radon Concentrations in the Kaat Kootenay Area
(Cranbrook and Vicinity)
Area
Number
of Homea
Average Cone,
in pCi/1.
Standard
Deviation
All Survey Sites
Main Floor
Basement
90
67
1.7
2.0
2.8
1.9
Fernie
Main Floor
Basement
3
7
3.2
1.7
2.9
.6
Cranbrook
Main Floor
Basement
47
41
q
1.7
. 7
1 7
Kimberley
Main Floor
Basement
15
10
2.1
3.4
2.3
2.8
Invermere
Main Floor-
Basement
6
4
7.1
1.6
8.5
.9
Golden
Main Floor
Basement
4
5
1.7
3.1
1.8
1.7
Creston
Main Floor-
Basement
14
1
1.4
.9
1.0
GREATER VANCOUVER AREA
Radon concentrations in the Vancouver region were very low (average of
0.49 +/- 0.23 pCi/1 and ranged between .2 and 1.6 pCi/1 in the 135 homes
measured. There is no significant difference from one area of the city to
another despite a large variation in geological environments. The three
factors that are common to the city which may explain the low radon levels are
the low terrestrial gamma, an abundance of "hard pan" clay that deters radon
infiltration and higher than average rainfall.
CONCLUDING REMARKS
Terrestrial gamma measurements can be used in British Columbia to
predict, -a community radon potential. They however cannot be used to predict
an individual home's concentration of radon. Soil structure particularly
permeability appears to have a marked impact on radon potential (9). The
province can be divided into three radon risk areas. There is a wide coastal
6-68
-------
strip where the risk of radon exposure is very low compared to other areas of
North America. However the majority of the province lies in .an interior belt
where the radon risk is typical of other areas of North America. There are
patches within that belt where both the terrestrial gamma .and radon risk are
relatively high. Further study is required to delineate these areas and all
homeowners in these areas should make it a priority to test their homes.
Modern changes to house construction have been reported to increase radon
concentration in the living area(lU). Although we have also seen this in the
Castlegar study, some construction techniques such as supplying combustion
area and well built basements are tending to mitigate this trend. There is
some concern that annual variations in household radon levels may also have to
be investigatedt10). At this time, fifteen additional regional radon surveys
are being carried out. These additional surveys should allow us to more
accurately delineate the geological, meterological, and construction
parameters that impact on radon levels in the province of British Columbia.
ACKNOWLEDGEMENTS
Tliis work was supported by the East Kootenay and We3t Kootenay Health
Units and the British Columbia Centre for Disease Control.
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 tie-
inferred.
6-69
-------
REFKIiKNCKS
LeTourneau, E.G. et al. Design and interpretation of large surveys foi~
indoor exposure to radon daughters, Radiation Protection Dosimetry.
Vol.7, Mo.1 4, p.303 - 308 Nuclear Technology Publishing.
Oswald, Richard A. The need for Jong-term radon measurements,
Environmental Radon Update, p.1-4, August 1986.
Morley and Green. Background radiation levels in British Columbia 1980
-1983. province of British Columbia Ministry of Health publication.
1984.
Bates et ai. Royal Commission of Inquiry into Uranium Mining, Province
of British Columbia, October 30, 1980.
McGregor, ft.G. Background levels of radon and radon daughters in homes
in the Castlegar-Trail area of British Columbia, Radiation Protection
Bureau, Health and Welfare Canada. Ottawa, November 1978.
Report No.10. Radon monitoring results from B.P.A's residential
weatherization program. U.S. Department of Energy, January 1989.
British Columbia. Ministry of Energy, Mines and Petroleum Resources.
Submission to the Royal Commission on Uranium Mining. Pages 16, 4b,
1979.
Ghomshei, M.M. and Slawson, W.P. Secular variations of radon in
metropolitan Vancouver, British Columbia, Canada. Paper presented at
1990 international Symposium on Radon -and Radon Reduction Technology at
Atlanta. OA. February, .1990.
Nielson, K.K. and Rogers, v.C. Radon transport properties of soil
classes for estimating indoor radon entry. Paper presented at
Conference on Indoor Radon and Lung Cancer, October 15-19, 1990.
Richland, Washington, U.S.A.
Ergent, G.W., Kathren, R.L., Cross, F.T. The effect of home
weatherization on indoor radon concentration. Paper presented at
Conference on Indoor Radon and Lung Cancer. October' 15-19. 1990.
Richland, Washington, U.S.A.
6-70
-------
The State of Maine Schools Radon Project: Results
L. Grodzins, Professor of Physics
Massachusetts Institute of Technology
and Founder & Chairman of the Board of
NITON Corporation, Bedford, MA 01730
T. Bradstreet, Director of Information and Education
Division of Safety and Environmental Services
Augusta, ME 04333
E. Moreau, Manager of Indoor Air Quality Program
Division of Health Engineering
Department of Human Services
Augusta, ME 04333
ABSTRACT
A comprehensive study has been made of the radon concentrations in every frequently occupied
room on or below grade in every public school in the State of Maine. 32% of the 653 school
buildings covered in this report had at least one room with a radon level exceeding the EPA
guideline of 4 picoCuries of radon per Liter of air (4 pCi/L). 8.7% of the 13,353 rooms had a
radon level > 4 pCi/L; 1.9% of the rooms had radon concentrations > 10 pCi/L; 0.7% of the
rooms had radon concentrations
> 20 pCi/L. The radon concentrations were not distributed uniformly among the schools; a
building tended to have a radon problem or it was essentially free of radon. The radon concen-
trations were not uniformly distributed throughout the state. The schools in the counties contiguous
to New Hampshire were far more likely to have a serious radon problem than were schools in the
central part of the state, especially along the coast. And we note a strong correlation between the
geographical results of this state-wide school survey and the previous state-wide results of radon in
homes.
6- 71
-------
INTRODUCTION
This is a report on the findings of the comprehensive survey of radon carried out in 1990 in
more than 13,000 classrooms in more than 650 school buildings in the state of Maine. Overall,
33% of the school buildings had radon levels exceeding the EPA action level of 4 pCi/L. Schools
with elevated radon values were not, however, uniformly located in the state. The western counties
tended to have considerably higher radon levels than elsewhere. Some schools in these counties
had mean concentrations exceeding 4 pCi/L and several had mean concentrations greater than 20
pCi/L. Most striking is the strong correlation between the radon levels in the schools of a county
with the radon levels in the homes in that county.
This study is unique in several ways. It is, to our knowledge, the first complete state-wide
school survey in which every regularly occupied room on or below grade in every school was
measured; nearly 50% of all school rooms in Maine were tested. It is the first to survey all the
schools using a single radon detection method analyzed by a single company. Without this unified
approach, the present study would not have been practical. It is the first survey in which the
placement and retrieval of the radon detectors were carried out by the school custodians, with all
scientific and technical decisions handled in advance by the testing firm, NITON Corporation; the
98.5% success rale of this procedure has important economic implications for future surveys.
The design of the study is described elsewhere in this meeting (1). So, too, are the protocols
and procedures of the testing program (2). For completeness, a summary follows. The next
section presents the results, first in overview, then in greater detail. The last section correlates the
school results with other information, particularly the radon survey of homes in the state and
summarizes our conclusions.
PROCEDURES
The radon tests were carried out using NITON'S patented liquid scintillation charcoal detectors.
These small, 1" diameter by 2" long, detectors contain a cartridge holding about 1.5 grams of
charcoal mixed with desiccant. For each school, NITON made up individual packages containing
the test vials, data sheets, and a copy of the school floor plans marked with locations for placing
the test vials. Most important, the package included a set of simple, comprehensive, step-by-step,
check-off instructions.
Every regularly occupied room on or below grade was tested over a week-end under closed
building conditions. The air-handling systems were generally operated continuously. School
custodians set out and retrieved the tests and returned them to the NITON Laboratory in
Massachusetts, using next-day UPS service. This protocol worked exceptionally well even for
remote one-room school houses, including those on islands off the coast and those in Indian
reservations; only 1.5% of the rooms had to be resurveyed because of faulty procedures.
The NITON LS vials were set out on Friday afternoons, retrieved Monday morning, generally
arrived at the laboratory in Massachusetts and were counted in the automated LS counters on
Tuesday. The NITON "2-day" diffusion barrier is most sensitive to the last 48 hours of testing so
the first evening of the test effectively established the base line of closed conditions.
All test vials were counted for 5 minutes each. Prompt return of the test vials meant that the
radon decayed by only 25% to 35% between the time the vials were closed and the time they were
6-72
-------
counted. As a consequence, the 5-minute count in the automated scintillation counter resulted in a
standard deviation a <5 % at a concentration level of 4 pCi/L; o -10% at 1 pCi/L; and O -20%
at 0.4 pCi/L. All vials with radon concentrations exceeding 3 pCi/L were rerun for 20 minutes
each ( a < 2%). Results, quoted to the nearest 0.1 pCi/L, were sent to the State of Maine the
following day.
RESULTS
QUALITY OF THE DATA
The results from NITON vials were compared with themselves and with independent tests.
Over the course of the study, side-by-side tests were run in a total of 33 buildings. The results
were excellent. The mean of the absolute differences between the side-by-sides was 0.2 ±0.16
pCi/L. The mean of the absolute differences for results exceeding 2 pCi/L was 0.2510.17 pCi/L.
Only 2% of the absolute differences were as great as 0.6 pCi/L; none were higher. The precision
of the results was <5% at 4 pG/L, about the same as the statistical uncertainty of the initial liquid
scintillation test.
Quality control of the NITON vials is carried out routinely with in-house radon standards, and
checked periodically using independent radon quality control laboratories. We followed three
additional procedures to establish the quality control for the Maine survey: 1) The NITON liquid
scintillation vials were specially tested at the Environmental Measurements Laboratory of the
Department of Energy; 2) 100 NTTON LS detectors, in pairs, were compared with 50 Charcoal
Canisters; i.e. 3 detectors (2 NITON LS and 1 CC) were run side-by-side. The Charcoal Canisters
were tested by the State of Maine Indoor Air Quality group. Most of the compared results were
within 0.1 pCi/L. Two comparison tests differed widely: In one test, the LS values were 3.3 and
3.0 pCi/L, the CC result was 1.2 pCi/L; in another, the LS values were 5.5 and 5.3 pCi/L while
the CC result was 3.0 pCi/L.
3) 30 NITON LS detectors were compared by the State of Maine with continuous monitors.
Half the tests lasted 8 hours, half lasted 16 hours; NITON detectors are calibrated from 8 to 72
hours. The mean radon concentrations ranged from 4.4 pCi/L to 67 pCi/L, with short-term
variations ranging from 0.6 pCi/L to 74.5 pCi/L, according to the continuous monitor. The mean
of the 30 NITON results was 10% higher than the mean of the means of the 30 results from the
continuous monitor. These comparisons give considerable confidence in the results for the
individual schools and for the overall survey.
STATE-WIDE RESULTS
School Rooms
The results for 13,353 school rooms are presented in Table I and Figures 1 and 2. The
frequency distribution of radon in the school rooms of Maine had a most probable value of
<1 pCi/L, a median value of 1.1 pCi/L and a geometric mean of 1.05 pCi/L. These values are
not much different from those obtained by NITON in a survey of 5,000 school rooms in
Massachusetts. 8.7% of the school rooms in Maine had radon values of 4 pCi/L and above (the
corresponding number in Massachusetts was 6%); 1.9% of the Maine school rooms had radon
values of 10 pCi/L and above (Massachusetts was 1.1%); 0.7% of the Maine school rooms had
radon values of 20 pCi/L and above (Massachusetts also had 0.7%).
6-73
-------
TABLE 1. THE FREQUENCY DISTRIBUTION OF RADON IN SCHOOL ROOMS
AND SCHOOL BUILDINGS IN MAINE
Concentration, pCi/L
Rooms with >
Concentration
Column 2 as % of
13,353 Rooms
Buildings with 1 or
more rooms >
Concentration
Column 4 as % of
653 Buildings
0.4
11,550
86.5
644
98.6
1.0
7,322
54.8
587
89.9
1.5
4,884
36.6
520
79.6
2.0
3,365
25.2
445
68.1
2.5
2,429
18.2
373
57.1
3.0
1,828
13.7
308
47.2
3.5
1,420
10.6
248
38.0
4.0
1,164
8.7
213
32.6
5.0
808
6.1
167
25.6
6.0
601
4.5
126
19.3
7.0
454
3.4
93
14.2
8.0
370
2.8
77
11.8
9.0
312
2.3
70
10.7
10.0
255
1.9
54
8.3
11.0
222-
1.7
49
7.5
12.0
197
1.5
44
6.7
13.0
172
1.3
38
5.8
14.0
155
1.2
35
5.4
15.0
144
1.1
33
5.1
16.0
131
.9
31
4.7
17.0
118
.8
28
4.3
18.0
109
.8
24
3.7
19.0
102
.7
21
3.2
20.0
98
.7
21
3.2
25.0
68
.5
14
2.1
30.0
48
.35
11
1.7
40.0
16
.11
7
1.1
50.0
2
.01
2
.3
6-74
-------
The distribution is plotted in Figure 2 as a lognormal probability graph. The data exhibit the
fact, now familiar in most radon surveys, that the radon distributions follow a normal probability
distribution up to about 4 to 5 pCi/L. The probability of observing elevated radon concentrations is
higher than would be predicted on the basis of a normal distribution. This study found twice as
many school rooms with radon concentrations above 10 pCi/L, and ten times as many above 20
pCi/L, as would be inferred from a normal distribution.
S.7% of School Rooms
1.9% of School Rooms
0.7% of S ihor»l Rooms
0-1 1-2 2-3J.4 4.5 6.7 7.8 8.9M0 10-15 15-20 >20
Radon Concentration, pCi/L
Figure 1. The distribution of radon in the school rooms of Maine
N.B. The uncertainties in these statistical values, and those given later in the paper, are due
almost entirely to the uncertainty in the accuracy of the test results, which we assume to be -10%
on the basis of NITON's overall accuracy in several EPA Quality Assurance rounds. A 10%
uncertainty in the absolute accuracy results in a corresponding 10% uncertainty to the median,
arithmetic and geometric means, as well as to the percentage of school rooms exceeding 10 pCi/L
and 20 pCi/L. The percentage of school rooms above 4 pCi/L depends more strongly on the
absolute accuracy because of the steepness of the distribution at 4 pCi/L. For example, if
NITON's absolute calibrations have a systematic error such that all results are 10% too high (and
recall that the EPA accepts a 25% absolute uncertainty) then only 7% of the school rooms are
6-75
-------
above 4 pCi/L. If NTTON's values are 10% low then 10.5 % of the school rooms are above 4
pCi/L. The sensitivity of the results to the absolute accuracy in the tests is a compelling reason
why surveys should be carried out using a single method and, wherever possible, by a single
group using the same calibration standards. In practice it is very difficult to accurately compare
surveys carried out by different methods or by different laboratories.
The distribution is plotted in Figure 2 as a lognormal probability graph. The data exhibit the
fact, now familiar in most radon surveys, that the radon distributions follow a normal probability
distribution up to about 4 to 5 pCi/L. The probability of observing elevated radon concentrations is
higher than would be predicted on the basis of a normal distribution. This study found twice as
many school rooms with radon concentrations above 10 pCi/L, and ten times as many above 20
pCi/L, as would be inferred from a normal distribution.
Frequency Distribution of Ration jConcefttratiojis
in School Rooms of M^ine
Geometric; Me^n: 1.95 pCi/L
10
i
0.3
.01
5 10
20 30
1
1
SI
70 80
99 95
99.9
99.99
Percentage of school rooms having less than given concentration {%)
Figure 2. Lognormal plot of the distribution of radon in the school rooms of Maine.
6-76
-------
School Buildings
The data for the 653 school buildings examined in this survey are presented in Table I.
Columns 4 and 5 give, as a function of radon concentration shown in column 1, the number and
percentage of school buildings that have at least one room with a concentration greater than that
level. 32.6% of the buildings had at least one room with a radon level o> 4 pCi/L. Stated the
other way, 67.4% of the buildings had no room with a radon concentration exceeding the EPA
action level. 8.3% of the buildings had at least one room with a concentration of 10 pCi/L or
greater, and 3.2% of the buildings had at least one room with a radon concentration of 20 pCi/L.
These state-wide percentages tell an incomplete story since both the geographic location and the
size of the school are critical variables.
School Size
Table II shows the distribution of the number of rooms per school building. The typical school
in Maine has fewer than 20 rooms, 151 schools have fewer than 10 rooms. The third row of Table
II gives the number of school buildings that have at least one room with > 4 pCi/L of radon. The
bottom row of the table gives the percentages of buildings that have at least one elevated radon
reading. The percentages vary from 21% to 50% but, within statistical uncertainties, the
percentages are essentially constant. This is a most surprising finding since one would expect, a
priori, that the larger the school the greater the probability of finding an elevated radon level.
TABLE II: SOME DISTRIBUTIONS IIS
SCHOOL BUILDINGS IN MATN1
7
Number of Rooms per Building
<10
10-19
20-29
30-39
40-49
>50
Total number of buildings
151
184
141
90
40
42
Total number of buildings with at
least one room > 4 pCi/L
40
65
41
36
20
9
% of "High-Radon" Buildings
27%
35%
29%
40%
50%
21%
The explanation is that radon is not randomly distributed in the schools. School rooms, in a
building that has a common architecture and air-handling system, show very similar radon
concentrations. To emphasize the lack of randomness, consider the larger buildings with more
than 50 rooms. If radon were distributed randomly we would find 8.7% of the rooms of each
school with elevated radon concentrations. The probability that a school with 50 rooms would
have no elevated radon room is (0.913)^ = 1.1%; the actual probability is 79%.
The "one-room" school houses do not follow a random pattern either, though one expects, on
the basis of the state-wide data, to find about 30% of the buildings with a radon problem. The lack
of randomness is demonstrated in Table III, which breaks out the data of the third row of Table II
to give, as a function of the size of the school, the number of buildings that have a given
percentage of the school rooms with > 4 pCi/L. Thus, of the schools that have at least one radon
problem, there were 5 large school buildings (>50 rooms) in which fewer than 10% of the rooms
had elevated radon. There was, however, one large school building in which more than 80% of
the rooms had an elevated radon problem.
6-77
-------
['ABLE ITT: THE NUMBER OF BUILDINGS AS A FUNCTION OF TIIE NUMBER OF
ROOMS AND THE PERCENTAGE OF ROOMS WITH ELEVATED RADON LEVEIES
Number of Rooms per Building
<10
10-19
20-29
30-39
40-49
>50
<10% of the rooms
0
14
20
19
9
5
10% - 39% of the rooms
22
37
12
11
7
1
40% - 59% of the rooms
1
5
5
3
0
1
60% - 79% of the rooms
5
6
3
2
3
1
>80% of the rooms
12
3
1
1
1
1
Total number of buildings
40
65
41
36
20
9
A number of school buildings (bottom row of Table III) were saturated with radon. There is
negligible probability that any of these saturations could occur by chance. Every one of the 48
rooms in one school building was far above the EPA guidelines; the median value was 25 pCi/L.
All of the buildings in the lower part of the table support the general observation that high radon
levels lend to cluster, there are relatively few buildings that have isolated rooms with elevated radon
levels.
The school buildings in the second row of Table III have, typically, only 1 or 2 high-radon
level rooms. Unfortunately, the odd high radon value can be very high indeed. For example, in
one school of 23 rooms, having a median radon level of 1 pCi/L, there was one classroom with
27 pCi/L; in a 4-room school house where 3 of the rooms were under 4 pCi/L, there was one
room with 38 pCi/L. In the next section we examine a few of the radon distributions in individual
school buildings.
Results of Individual Schools
The present study involved more than 650 school buildings and NITON has surveyed more
than 500 other school buildings during the past two years. The buildings have different ages,
architectures, geological sites, air-handling systems, etc. From the mitigator's point of view, each
building is unique. From the radon surveyor's point of view, there are definite patterns of radon
distributions that can be useful guides to understanding the origins of the radon problem.
Figure 3 shows three radon distributions found in schools in Maine. Distribution A is a "radon-
free" building; the median value is well below 1 pCi/L and no concentration is greater than 2 pCi/L.
There is no correlation between the radon levels and the room location. The most probable value is
similar to the radon level found outdoors.
Distribution C is a "radon-infested" building in which the radon levels are about the same in
every room. The distribution is nearly Gaussian; the mean of 24.9 pCi/L is same as the median
value.
Distribution B is very similar to C, though the mean and the median are both below 4 pCi/L.
Every room in the school has a potential radon problem as the levels fluctuate with changes in the
weather and the air-handling.
6- 78
-------
I ptliL
M
l \
i i...
i..
i ,c.in. &» pciit i
: iSi
s
n ; in :• SI i
^ S \
N i N ! ^ ¦
^ ^ S • N 1
s ^ ; s 1
JUL
¦
1
* r » *5 * •
I i I i i i i
t i i 5 s : s s * s i
Radon Concentration, pCi/L
Figure 3. Radon distributions in three schools in Maine.
Figure 4 shows three distributions observed in school buildings of Maine. These are rather
typical of the broad distributions that almost always show a correlation between the radon
concentration and the room location. Contiguous rooms show similar radon values; changes in
concentration take place over several to many rooms..
6-79
-------
Radon Concentration, pCi/L
Figure 4. Radon distributions in three schools in Maine
The most probable radon concentration of distribution D is below 1 pCi/L and most of the
rooms are radon-free. Nevertheless, a few rooms have values well above the EPA guidelines; it is
our experience that these rooms are generally localized to the same area of the school.
Distribution 0 is also sharply peaked at a low radon value, but the median value is close to
2 pCi/L, indicating a radon problem. We again anticipate a strong correlation between the radon
concentration and the geography of the room.
Distribution F has no reading above 3.5 pCi/L. But the median value of 2.5 is high. This
building should be carefully monitored over time since it is likely that there will be periods during
the year when the radon levels will rise by at least a factor of two and most of the rooms will have
concentrations exceeding EPA guidelines.
Only distributions similar to those of A in Figure 3 can give us reasonable assurance of a school
without a real or potential radon problem. The assurance is not, however, a guarantee. We have
several examples of schools in this survey in which there is one or at most two elevated radon
concentrations in an otherwise radon-free school. To Find such rooms, one must survey every
room on or below grade.
6- 80
-------
Survey Results bv Maine County
Table 4 presents the results by county. The last 3 columns give the results of the Maine survey
of radon in homes.
TABLE 4: RADON IN TIIE SCHOOLS AND HOMES OF MAINE, BY COUNTY
County
Total
School
Rooms
Median
pCi/L
Maximum
pCi/L
% >4
pCi/L
Total
Houses
Maximum
pCi/L
% >4
pCi/L
Androscoggin
1,065
1.3
21.8
5.34
39
17.7
18
Aroostock
1,493
1.5
18.7
11.25
95
25.2
43
Cumberland
2,222
1.4
59.2
16.42
120
82.7
41
Franklin
345
.5
15.8
6.66
19
9.7
16
Hancock
654
1.2
21.4
5.5
49
19.4
27
Kennebec
1,062
1
43.5
5.83
57
19.4
26
Knox
151
.6
4.5
.66
24
9.7
29
Lincoln
116
.6
13.8
4.31
12
5.9
8
Oxford
580
1.5
37.7
14.47
37
30.2
51
Penobscot
1,783
.7
17.9
3.64
72
5.7
17
Piscataquis
384
.6
5.7
0.26
37
22.5
32
Sagadahoc
499
.9
5.3
1.4
32
8
19
Somerset
746
.8
26.4
4.28
27
5.8
19
Waldo
223
1
7.6
6.27
26
13
19
Washington
539
1
12.6
1.66
36
12.2
14
York
1,491
1.4
41.6
15.75
73
33
38
Total Tests
13,353
755
The counties vary widely in population and, therefore, in the number of schools and school
rooms. There are large variances in the measures of radon concentration in the homes of several of
the counties, particularly, Lincoln, Knox and Waldo. The school measurements are much more
secure. Even the smallest county had more than 100 school tests and the median value is measured
to an uncertainty of less than 0.2 pCi/L.
Table 4 gives three indicators of the radon concentration in the schools of the different counties.
The maximum radon value, column 4, can be a statistical outrider and is not a useful measure of
the radon problem in the county. The percentage of rooms that exceed 4 pCi/L, column 5, is a
much more useful indicator since it focuses on that part of the distribution which demands action.
The median radon values, column 3, while not giving the full description that would be obtained
from the geometrical moments of each distribution, does give an easily understood measure of the
6- 81
-------
radon problem and is, in our view, the best single number to quote. If half the rooms have a radon
concentration below 0.7 pCi/L, which is the case in 5 counties, one can be quite sure that fewer
than 10% of the school rooms will have elevated values. On the other hand, if half the rooms have
a radon concentration greater than 1.4 pCi/L, which is the case in 4 counties, one can be quite sure
that more than 10% of the rooms will have concentrations greater than 4 pCi/L.
The four counties with the highest radon concentrations in the schools also have the four
highest radon concentrations found in homes. The correlation between the percentage of school
rooms in a county that are > 4 pCi/L (column 5) and the median radon concentration found in the
county (column 3) is r2 = 0.58. The correlation between the values in column 5 and the highest
radon value found in any school room in the county (column 4) is r2 = 0.65. There is little
correlation between the median and the maximum values in a county.
Figure 5 shows, by county, the percentage of school rooms and homes that exceed the EPA
action level. The values for homes have been divided by a factor of 3. (These home readings are
generally basement readings. The first floor level in New England is, as a rule of thumb, about
one-third of the basement level.)
There is a striking correspondence between the results for schools and those for homes in the
same counties. There are several exceptions to the obvious correlation, but most have large
uncertainties in the individual values due to the small sample sizes.
'5a
V
0.
E3 Percent of School Buildings >= 4 pCI/L
¦ Percent of Home* >* 4 pCi/Li {« 3)
Counties in Maine
Figure 5. Radon Concentrations in School Rooms and Homes by County in Maine
6-82
-------
SAGADAHOC
ANDROSCOGIN
Median Levels of Radon
in Counties in Maine. In pCi/L
6- 83
-------
The final figure shows a map of Maine divided into its 16 counties. The median radon
concentration for each county is given in pCi/L. The map has been shaded to show the strengths
of the radon values; the darker the shading the higher the median value. Radon is obviously
correlated with the geography of Maine. The four southwestern counties, York, Cumberland,
Androscogin and Oxford have uniformly high mean radon values. There is then a band of
moderate radon concentrations extending from Kennebec through Waldo, Hancock and
Washington. The central coastal counties of Sagadahoc, Lincoln and Knox have low radon
concentrations, as do the Maine-woods counties of Franklin, Somerset and Piscataquis.
Aroostook county, with towns and schools bordering New Brunswick, Canada, is another area of
high radon concentration. As we noted above, the areas bordering New Hampshire and New
Brunswick have the highest radon concentrations in both school rooms and homes.
SUMMARY
The principal aim of this comprehensive survey of radon in the schools of Maine was to find
those schools that should be mitigated immediately, as well as those schools that have potential
problems that must be monitored over time. That aim has been well met. A second aim was to
obtain a data base of the radon concentrations in every school, which would serve as the bench
mark and guide for spot checks and sample surveys that might be conducted in future years either
as part of a general "due diligence" program or because of changes in construction or air handling.
That aim, too, has been well met. A third aim was to find generalities and correlations that might
aid in understanding the radon problems in the state. This paper presents the sum of those
findings.
The frequency distribution of radon in school rooms in Maine is similar to that found in surveys
of school rooms elsewhere; for example, in Massachusetts. The distribution follows a lognormal
curve up to radon concentrations of about 5 pCi/L. The occurrence rate at the higher concentra-
tions is greater than would be predicted by the normal curve.
8.7% of the school rooms tested over weekends under closed building conditions were found to
have radon concentrations greater than the EPA action level of 4 pCi/L; 1.9% of the rooms were
above 10 pCi/L; 0.7% were > 20 pCi/L. The elevated radon concentrations were not randomly
distributed. A school that had one room with > 4 pCi/L, had, on the average, 5 such rooms. The
odds that a school building had at least one room with > 4 pCi/L, was about one in three. The
odds were almost independent of the size of school.
The patterns of radon distributions in the schools can be conveniently divided into three broad
groups: 1) those radon-free schools with median (or geometric mean) values well below 1 pCi/L,
no concentration greater than 2 pCi/L, and no correlation between the radon values and the position
of the school room; 2) those radon-infested schools with median and mean values above ~3
pCi/L, and with a majority of the values exceeding the EPA guidelines; 3) schools that have
median (or geometric mean) values in the 1-2 pCi/L range, have broad distributions (large standard
deviations of the geometric means), and generally a strong correlation between the position of the
room in the building and its radon concentration. The first group is the only one with a strong
probability of being radon-free under all circumstances of weather and air handling; the second
group must be mitigated early; the last group encompasses a wide variety of situations with few
common denominators other than the need for close examination and monitoring.
The radon concentrations showed a strong correlation with geography. The median radon
concentrations in the western counties — Oxford, York, Cumberland and Androscogin — and the
6-84
-------
northernmost county, Aroostock, were more than twice the values in the coastal counties of Knox
and Lincoln and the Maine woods counties of Franklin, Somerset, Piscataquis and Penobscot.
The counties with highest (lowest) radon concentration in the schools generally had the highest
(lowest) average concentration in the homes. This strong correlation between radon levels in
school rooms and homes in the same geographical area implies that the underlying geological
factors are the determinants for the average or median radon concentrations in county-size areas.
We anticipate that this is a general conclusion that will be observed throughout the country.
The correlation between radon levels in homes and in schools is strengthened by comparing the
Maine results with those obtained by NITON Corporation tests of more than 5,000 school rooms
in Massachusetts. The EPA state-wide studies of homes found that 25% of the "lowest-livable"
rooms in Massachusetts had values > 4 pCi/L, compared to 30% for Maine. Massachusetts has
only 6% of its school rooms exceeding the EPA guideline, compared to 8.7% for Maine. Thus
there are fewer school rooms with high readings in Massachusetts to about the same degree that
there are fewer homes with high radon readings.
We also consider it worth noting that the percentage of school rooms with > 4 pCi/L of radon is,
for both Maine and Massachusetts, about one-third the percentage of homes with basements with
radon concentrations > 4 pCi/L. This correlation implies that the average radon values in schools
is not much different from the radon concentrations found on the first floors (the living areas) of
homes in the same geographical area.
ACKNOWLEDGEMENTS
L.G. would like to take this opportunity to thank Anne McGuineas and Christopher Collins of
the staff of NITON Corporation for their indispensable help with the paper generally, and the
generation and understanding of the data, specifically.
The work described in this paper was not funded by the U.S.
Environmental Protection Agency and therefore the contents do
not necessarily reflect the views of the Agency and no official
endorsement should be inferred.
REFERENCES
1. Warren, H.E. and Romm E.G. The State of Maine School Radon Project: The Design
Study. EPA 1991 International Symposium on Radon and Radon Reduction Technology,
Philadelphia, PA. April 2-5, 1991.
2. Grodzins, L., Warren, H.E., and Romm, E.G. The State of Maine School Radon Project:
Protocols and Procedures. EPA 1991 International Symposium on Radon and Radon
Reduction Technology, Philadelphia, PA. April 2-5, 1991.
3. Grodzins, L. Radon in Schools in Massachusetts. EPA 1990 International Symposium on
Radon and Radon Reduction Technology, Atlanta, GA. February 19-23, 1990.
6-85
-------
RADON IN BELGIUM : THE CURRENT SITUATION AND PLANS FOR THE FUTURE
A. Poffijn1, J.M. Charletj;, e. Cottens3, S. Hallez3,
H. Vanmarcke4, P. Wouters5
^ Nuclear Physics Lab, RUG, B-9000 Gent, Belgium
~ Polytechnical Faculty Hons, B-7000 Mons, Belgium
3 Ministry of Public Health, B-1000 Brussels, Belgium
Ł Radiation Protection, SCK/CEN, B-2400 Mol Belgium
5 BBRI, B-1342 Limelette, Belgium
ABSTRACT
An overview is given of the current knowledge about radon
in Belgium. The results and experience obtained from
regional and local surveys are presented .
The indications about the risk of indoor radon are
discussed.
The general outline and main purposes of the national
action program, as prepared for presentation to the
competent authorities, are also dealt with.
INTRODUCTION
In Belgium, the yearly average population dose from
natural ionising radiation is estimated to be of the order of
3.5 mSv (1). This value is about three times higher than what
was generally accepted some 15 years ago. This increase is a
major consequence of the introduction of the notion effective
dose-equivalent and of a better knowledge of the respiratory
track dosimetry.
As a matter of fact, radon progeny alone is responsible for
some 50% of the total annual effective dose.
From the point of view of policy the range of variation of ex-
posures is an even much more important quantity. Radon concen-
trations are found to vary over more than, three decades, being
at least one order of magnitude greater than the range for oth-
er natural sources of ionising radiation.
Based-upon the avaiable radon data, some hundred of houses are
expected to exist, where the inhabitants are exposed to levels
giving rise to doses higher than the limits for radiation work-
ers.
Infiltration of radon bearing soil gas is the major source of
radon in dwellings. As a lot of uranium anomalies occur in Bel-
gium, the study of the geological parameters is an essential
tool for the evaluation and reduction of this infiltration. The
specific construction characteristics of Belgian dwellings
should of course also be taken into account.
6-87
-------
RISK ESTIMATIONS
According to the results of
average indoor radon concentration
a net distinction between the
wherein the country may roughly be
AND REGULATIONS
a national survey (l) the
in Belgium is 48 Bq/m3, with
two major geological zones
divided (Fig.l).
The Netherlands
North Sea
Antwerp
I i I M U
zone I
Ghen t
mean value 33 Be
the Scheldt Brussels
Liege
he neuse
zone 13
//dG m
Lu >;ei7;b
O
Oi
3
•<
urg
Figure l : Mean radon concentration in the two major
geological zones of Belgium
6- 88
-------
Using a dose conversion factor of 50jxSv per Bq/m3 and a mean
time spent indoors of about 0.8, an average effective dose
equivalent of 1.9 mSv per year may be put forward as reference
value for a member of the Belgian population, corresponding to
a collective dose of 19000 Sv.
Up to now, near all risk estimations for radon are extrapola-
tions of the results of epidemiological studies among workers
(2). From these data, an overall (absolute) risk of the order
(3 ± 1.5)*10 per Sv can be expected to hold for the general
population (3). This value agrees well with the ICRP risk fac-
tor of 1.65/Sv for radiation of artificial origin. Using this
risk factor, some 250 up to 750 of the total number of lung
cases per year in Belgium (6000) may be associated with radon.
The corresponding life-time risk is of the order of 0.3% and is
of the same order as the risk for a mortal accident at work.
Due to the increasing awareness of the significance of indoor
radon, many national and international authorities have drawn
up dose-control policies. The proposed reference levels vary
between 150 and 800 Bq/m3 for existing buildings and between
140 and 250 Bq/m3 for future constructions (Table 1).
TABLE 1 : RECOMMENDATIONS FOR RADON LIMITATION
Radiological
Authority
Existing Homes
(Bq/m3)
Future Homes
(Bq/m3)
ICRP
CEC
WHO
Germany
Sweden
UK
USA
400
400
200
250
800
200
150
200
200
200
250
140
200
150
6-89
-------
In Belgium the highest values have been reported in the prov-
ince of Luxemburg (Fig.2). Based upon the avaiable data some
10000 dwellings are expected to have radon concentrations of
more than 400 Bq/ra , with an exceptional high proportion being
situated in the province of Luxemburg (Table 2).
L'lJ3tiŁ V.
IfAIKAtlT
-
Ed o-so
Q 50-100
~ 1.00-200
"• • 'I
'w, y
MBM Hyfflilite
rianUR
70
I-
' ^
X ¦ i
r
l
I
'J ^
,, J**
k* llu=ni:
ISlif
/
i
> i ,'Hy '1
Figure 2 : Indoor radon concentrations in Belgium
6-90
-------
TABLE 2 : DISTRIBUTION OF PROBLEM DWELLINGS IN BELGIUM
Province Number(percent) of dwellings >400 Bq/m3
Antwerp
Brabant
Hainault
Limburg
Liege
Luxemburg
Namur
East-Flanders
West-Flanders
TOTAL
680 (0.08)
1500 (0.4)
7400 (9.0)
800 (0.5)
10380 (0.3)
RADON SOURCES
For most house the soil and the building materials form
the major sources of radon indoors. As part of a general study
on building materials, the exhalation rate of some 120 commonly
used materials in Belgium was measured. From the obtained re-
sults ranging from 10 ° Bq/kg*s for bricks up to 10 x Bq/kg*s
for materials made of phosphogypsum, it can be concluded that
building materials in Belgium contribute only for a minor part
to indoor radon (max. 20 to 30 Bq/m3) .
On the other hand the infiltration of radon bearing soil gas
can easily give rise to very high indoor levels.
Evidence of this statement was obtained by means of a case-
control study in two neighbouring villages.
Referring to the results of a survey for uranium exploration
in the Belgian Paleozoic (4), a case-village was selected,
where the uranium content of the soil reaches locally 70 up to
80 ppm. In some 50 houses built-on on close to this anomaly the
radon concentration was monitored for one year through 4
seasonal measuring campaigns. The same was done in a comparable
neighbouring control-village.
6- 91
-------
From the best log-normal fits (Fig.3) it can be concluded that
the radon concentrations in the case-villaae (with an average
of 117 Bq/m3 and a median value of 77 Bq/m3) are significantly
higher than the values obtained in the control-village (average
and median resp. 80 and 54 Bq/m3). In about 3% of the houses
in the case-village the radon concentration exceeds 400 Bq/m3.
As the results of the national survey indicate that in only
some 0.3% of the total house stock this value is expected to
be exceeded, the zone around the anomaly should be classified
as high risk area. Special prevention and inigitation actions
should be taken there without any delay.
CONTROL
'3
O!
CASE
Ł
ao
o
o
3
y
O
I
1
3 A DON CONCENTRATION ( Bq/rr,3 )
Figure 3 : Cumulative radon distributions in the case &
control village
6-92
-------
RADON PLAN(S) FOR THE FUTURE
The growing uneasiness among the population about radon
and the lack of any guideline or strategy nessecitated a
coordinated action plan covering topics as mapping, prevention,
mitigation and information of the public.
Therefore, in close collaboration with governmental responsi-
bles and different concerned research groups a national radon
research project was worked-out and agreed upon.
It primarely focusses on the province of Luxemburg and on two
smaller areas - coinciding more or less with the territory of
the town of Vise and the village of Bifevre - where alarming
high radon levels (>4000 Bq/m3) were registrated.
In this project attention is given to the following points :
- Databank for all radon measurements in Belgium
- Detailed radon risk map of the province of Luxemburg
and of the area around Vis§ and Bievre
- Detailed inventory of a selected number of risk areas
- Effectiveness of mitigation techniques
- Quality control exercices
- Information of the public.
ACKOWLEGMENT
This work is partly funded by the Commission of the
European Communities.
The work described in the paper was not funded by the U.S. En-
vironmental Protection Agency and therefore the contents do not
necessarily reflect the views of the Agency and no official
endorsement should be inferred.
REFERENCES
1. Poffijn,A. The indoor radon problem in Belgium. Xu :Indoor
air quality and ventilation,Selper Ltd, London, 1990, p.339.
2. Marks,S. Radon epidemiology, a guide to the literature,
DOE/ER-0399, U.S. Department of Energy, Springfield,
Virginia, 1988, 136pp.
3. Jacobi,W. Possible lung cancer risk from indoor exposure to
radon daughters, Radiat. Prot. Dosim. 7: 395, 1984.
4. Charlet,J.M.et al. Reconnaissance survey for uranium in the
Belgian Paleozoic, Professional Paper 196, Brussels, 1983.
6- 93
-------
A RADIOLOGICAL STUDY OF THE GREEK RADON SPAS
P. Kritidis
Environmental Radioactivity Laboratory
Institute of Nuclear Technology - Radiation Protection
15310 Aghia Paraskevi, Athens, Greece
ABSTRACT
A number of balneological units located in four regions of Greece and
using thermal spring waters of high 222Rn concentrations (0.1-6 MBq m-3) have
been investigated. The concentrations of 222Rn and 226Ra in the water used, as
well as those of the short-lived decay products of 222Rn (RnD) in the indoor
air have been determined. The annual doses to the personnel and the patients
have been evaluated. The results are discussed in the frame of the
Justification and ALARA principles. The main problem concerns the absence of
scientific statements relating the benefit from the procedures applied (or
certain part of it) to the exposure to 222Rn and RnD.
Keywords: natural radioactivity, radon spas, balneotherapy, inhalation doses
The Work described in this paper was not funded by the U.S. Environmental Protection Agency and there fore the contents do
not necessarily reflect the views of the Agency and no official endorsement should be inferred.
INTRODUCTION
The water is known to be, generally, a minor source of indoor radon. This
is riot always the case in the balneological (water-physiotherapy) units using
thermal spring waters. The balneotherapy is an ancient (Greek, Roman) practice
which has survived until our days in many parts of Europe. After the discovery
of radioactivity, it was found that certain spa waters are characterised by
high concentrations of 222Rn. In Greece, the first studies of this kind were
carried out by Pertesis during the 20s and 30s (1-3). In four regions of
thermal springs the concentrations of 222Rn in the water exceeded 100 kBq m-3
(originally expressed in the old Mache units, where 1 Mache = 13.5 kBq m~3).
These regions are shown in Fig.l. The maximum radon concentrations were
measured in the Artemis and Apollon springs of the Ikaria island and equaled
10 and 7.5 MBq m-3 respectively. The major radon spas have been investigated
later by other authors (4,5) and the early values of Pertesis have been,
generally, confirmed. Nevertheless, we could not. find, during 1983, any study
with radiation protection goals, i.e. study considering the critical pathway
of man's exposure - the inhalation of RnD related to the radon released from
the thermal waters. During the period 1984-88 we have visited repeatedly the
balneological premises in the 4 spa regions mentioned, measuring the
concentrations of 222Rn and 226Ra in the waters used, the concentrations of
RnD in the air of the main and auxiliary indoor spaces, collecting information
about the existing practices, occupancy factors etc. The results of these
studies, together with the related dose estimations, are given in the present
paper.
6- 95
-------
THE GREEK RADON SPAS: GENERAL DATA
The major Greek radon spas are located in four regions (Fig.l): Ikaria
island in the Eastern Aegean Sea (9 springs, 3 units), Karaena Vourla (1 common
source and 4 units) and Edypsos (5 springs and 5 units) in the Northern
Evoikos gulf and Loutraki in the Eastern Korinthiakos gulf (5 springs and 5
units).
In all cases the spas are located right on the coast and, also, in the
mild climate central latitudes of Greece. This allowed their establishing,
throughout the centuries, as attractive places of the so-called "therapeutical
tourism", where certain physiotherapy practices are combined with usual by-sea
vacation, swimming, fishing etc. In most cases, the high temperature of the
water allows its use for bathing procedures, typically in separate individual
baths, but also in common pools (Kamena Vourla). In one case the water has to
be warmed before use. The water is collected, usually, in some basic
reservoirs and directed further to the points of use. This results in certain
losses of radon gas, of the order of 10-30 %.
During the typical individual bath procedure the patient enters a tub
filled with warm spa water (or mixture with cold spa water) and spends about
20 min in it. The total time spent in the bath, including undressing and
dressing, is about 30 min. The personnel has to prepare the bath (clean and
fill the tub with new water) and often, also, to help the patients before
and/or during the procedure. In most cases there are at least 5 baths per
member of the personnel and their in-the-bath occupancy exceeds 50%.
There are various architecture designs of bath premises. In some cases
the baths are isolated, from the ventilation point of view, from the common
spaces (corridors, halls), but communicate one with another through big holes
(0.5 to 4 m) in the top parts of the common walls. This construction leads to
higher ventilation of the bath, but also to dependence of the average radon
concentration in the bath air on the current working occupancy of the whole
premise. In other cases the baths are totally isolated one from another, but
there is some air exchange with the common spaces. This leads to higher
average concentrations of radon in the bath air, but also in the air of the
common spaces, especially under "stack effect" conditions in the building.
The significant dispersion of the water during the tub filling, its
enhanced temperature and the duration of the procedure lead to the release of
practically 100% of the radon in the air. The use of 1 m3 spa water per
procedure in a bath of 25 m3 air volume will result in a transfer coefficient
of 0.04 between the water and air concentrations of radon under the absence of
ventilation. In the extreme case of the 6 MBq m-3 Apollon spring water it
would correspond to average concentration of 222Rn in the bath air equal to
240 kBq m~3.
The drinking therapy is another use of certain spa waters. In this case
the patients exposure to inhaled radon is significantly lower, with the
exception of cases like LU3, where certain patients combine the drinking with
a sort of inhalation therapy, by staying 10-15 min inside the hall built
around the twin thermal spring. In all cases, the intake of 226Rq is a pathway
to be considered as well. The water consumed per season is about 20 1.
6-96
-------
INSTRUMENTS AND METHODS
The concentrations of 222Rn and 226Ra in the spa waters have been
determined by use of the total alpha-counting method (Lucas cell instrument)
in a closed loop air circulation system (6). The glass sampler is designed to
be used for radon extraction as well, which avoids any sample transfer and the
related losses. The concentration of 2 2 2Rn is measured soon after sampling to
ensure maximum counting statistics. Temperature and volume corrections are
applied. After the "in situ" determination of 222Rn, the 0.25 1 water sample
is fully de-emanated and sealed for about 30 days, to allow the radioactive
equilibrium between 226Ra and 222Rn. Then a second measurement of 222Rn is
made to determine 226Ra. The LLD (2o of bokg) of the method is 25 Bq m-3,
which is quite adequate for 222Rn, and acceptable for 226Ra (where it
corresponds to 6.3 (jSv per year for a standard 0.8 m3 water consumption and to
0.16 ^Sv for the 20 1 average patient consumption of spa water per season.
The concentrations of RnD have been determined separately by use of an
express variant of the 3-interval total alpha-counting filter method (6). The
time of aii- sampling is 1 min arid the counting intervals - (1-5), (6-10) and
(11-15) min after the end of sampling. The sampling flow rate is 100 1 min-1
and the active filter area - 7 cm2. The LLD (2cr of bckg) is 30 Bq m 3
equilibrium equivalent concentration (EEC) of 222Rn - a value adequate for the
levels observed in the radon spas studied. It was also possible to determine a
'quasi-equilibrium factor", as the ratio of EEC to the concentration of 2i8p0.
RESULTS AND DISCUSSION
The concentrations of 222R11 and 226Ra in the waters examined, as well as
the range of concentrations of RnD measured in the air of the premises using
these waters (if any) are given in Table 1. Note that the water sampling has
been done at the point of use (if any) and not at the source spring.
Therefore, in most cases the values for 222Rn are slightly lower than those in
the source point. The waters of the Ikaria springs 13-17 are not used for any
curative procedures indoors. The concentrations of RnD in the outdoor air
close to these springs are measurable, but insignificant from the radiological
point of view, so they are not given in Table 1. Note also, that only the
waters of the units L2 and J A are used for drinking therapy.
The balneological premises in Kamena Vourla use a common water reservoir,
while the other premises have separate reservoirs, in certain cases - even
two, one for hot water and a second, where the water is cooled before use.
The maximum concentrations of RnD in air have been measured, in all
cases, at the place of the water use (bath, pool etc.). The minimum values
have been measured at the auxiliary spaces used by the personnel and/or the
patients (during waiting). The occupancy factors of the various spaces for the
patients and the personnel have been estimated both by observations and from
the information provided by the staff.
The "maximum transfer factor" is the ratio of the maximum RnD
concentration measured in air to the 222Rn concentration in the water used. It
can be seen that these factors vary significantly from premise to premise by
6-97
-------
more than two orders of magnitude: minimum value 0.00032, maximum value 0.085
and average value 0.012. One must note that both extreme values have been
measured in the air of personal bathrooms. It is also interesting, that the
highest values of RnD in air (K3, K4) are related to water concentrations of
222Rn significantly lower that the maximum observed (II, 12). We do not
consider here the case I1R, which is a reservoir of radon water, not visited
by patients. It is interesting, nevertheless, to note that certain
individuals, attracted by the idea of the "primary curative source", have been
seen to apply a sort of "private inhalation therapy" in the entry of the
reservoir, inhaling several minutes its 120 kBq m-3 RnD air!
TABLE 1. CONCENTRATIONS OF NATURAL RADIONUCLIDES MEASURED IN THE WATER
AND THE AIR OF VARIOUS GREEK RADON SPAS (Bq m-3).
Region,
premise
222Rn in
226Ra in
RnD in
Maximum 1
water
water'
air
transfer factor |
(
LI
450000
18-400
1
0.0009 j
L2
175000
103
20-200
0.0011 i
Loutraki
L3
170000
150-3500
0.021 i
L4
140000
55
180-440
0.0031 j
L5
90000
15-40
0.0004 j
Kamena
K1
850000
1500
20-3700
0.0044 i
Vourla
K2
it
..
90-1400
0.0016 |
K3
ti
110-15000
0.018 i
K4
"
It
11-18000
0.021 |
Ikaria
11
5700000
950
1300-6900
0.0012 i
I1R
5700000
950
120000
0.021 1
12
5700000
1200
500-1800
0.00032 i
13
3000000
200
-
;
14
625000
360
-
-
15
480000
200
-
-
16
270000
3400
-
- i
17
200000
4700
-
i
18
160000
3500
70
0.00044 i
19
<100
3400
<20
!
Edypsos
El
200000
3400
150-1750
0.0088 j
E2
72000
1600
200
0.0028
E3
10000
1100
400-850
0.085 i
E4
9300
2700
30-130
0.014 j
E5
1000
5000
<20
i
The differences in the transfer factors observed reflect, mainly, the
variability of ventilation rates, but also the differences in the water-
temperatures and the "typical" dispersion of the water during its use. This
variability leads to insignificant correlation between the water and air
concentrations of radon (CC=0.2). It is interesting to note, that an weak
negative correlation (CC=-0.3) is observed between the concentrations of 226Ra
6-98
-------
and 222Rn in the waters examined. The discussion of this point is beyond the
scopes of the present study.
The estimations of the annual effective dose equivalents (EDE) for the
personnel and the patients of the radon therapy centers examined are given in
Table 2. The occupancy of the personnel is 5 months per season and, in most
cases, 50% of the working time is supposed to be spent in the areas of highest
radon concentrations in air. The patients are advised, typically, for 20 bath
procedures of 30 min each per season. The intake of radon water for the
drinking therapy patients is, typically, 20 1 per season.
TABLE 2.
ESTIMATIONS OF ANNUAL EFFECTIVE DOSE EQUIVALENTS FOR THE
PERSONNEL AND THE PATIENTS OF VARIOUS GREEK RADON SPAS (mSv)
Region, premise
Personnel
inhalation
Patients
inhalation ingestion
of RnD
of RnD
of 226Ra
LI
1.9
0.05
—
L2
0.9
0.01
0.001
Loutraki
L3
16
0.4
-
L4
4.0
0.03
0.0005
L5
0.2
0.005
-
Kamena
K1
17
0.4
(0.01)+
Vourla
K2
6.6
0.15
..
K3
70
1.7
..
K4
85
2.0
Ikaria
11
32
0.8
(0.006)
I1R
-
(1.3)*
-
12
8.5
0.2
(0.008)
18
0.35
0.01
(0.02)
19
0.1
0.002
(0.02)
Edypsos
El
8.2
0.2
(0.02)
E2
1.0
0.02
(0.01)
E3
4.0
0.09
(0.01)
E4
0.6
0.015
(0.015)
E5
0.1
0.002
(0.03)
* Based on 3 min per day "private inhalation therapy" (see text).
+ The values in parentheses are hypothetical. These waters are not used for
drinking therapy.
It can be seen, that the estimated yearly EDE for the personnel exceed,
in 2 cases, the current 50 mSv a-i dose limit, in 5 cases - 30% of this limit
(controlled area conditions) and in other 4 cases lay between 10% and 30%, of
this limit (supervised area conditions). If we apply the new 0.04 per Sv fatal
cancer risk coefficient recommended by ICRP (7), then risks of the order of
10% could be attributed to the personnel of the units K3 and K4 after 30 years
of work and risks between 1% and 5% - in other 6 cases.
6-99
-------
The dose estimations for the patients vary 3 orders of magnitude, with
maximum values not exceeding 2 mSv a-i. These doses are mainly due to the
inhalation of RnD, while the ingestion of water (where applied) has an
insignificant contribution.
CONCLUSIONS
Two different radiation protection problems arise from the results
presented above.
1. The members of the personnel of certain radon therapy premises are
exposed to doses which not only exceed the 10% and/or 30% of the current dose
limit, but, in 2 cases, the limit itself. Nevertheless, these working areas
are not classified as supervised or controlled and no radiation protection
measures are applied. We have, in all cases, a violation of the Justification
and ALARA principles and in 2 cases - also of the Dose Limitation principle.
2. The patients are exposed to doses similar to those reported for other
cases of medical use of ionising radiations. Nevertheless, we could not find
any scientific material dealing with the dose-benefit relation of the radon
therapy procedures and, therefor, no risk-to-benefit analysis and no
optimisation can be applied.
It seems necessary to draw further the attention of the international
radiation protection community to the problem of the radon balneology
practice, in order to achieve, gradually, the conformity of this practice
(whatever it could mean) with the basic radiation protection principles.
REFERENCES
1. Pertesis, M. About the radioactivity of the Kamena Vourla spas. Geological
Service of Greece: No 16, Athens, 1926.
2. Pertesis, M. The Greek mineral springs. Geological Service of Greece: No 24
Athens, 1937.
3. Pertesis, M. About the radioactive hot springs of Ikaria island.
In: Proceedings of the Academy of Athens No 14, 1939.
4. Ioakimoglou, G., Georgalas, G. and Fokas, E. Analyses of the therapeutical
spas of Kamena Vourla. In: Radiosprings Kamena Vourla, Athens, 1940.
5. Gioni-Stavropoulou, G. Hydrological and hydrogeological studies.
Publication of the Institute of Geological and Mining Research No 39, 1983.
6. Kritidis, P. and Angelou, P. Investigation of the concentration of 222Rn
and its daughter products in the Loutraki spas. Greek Atomic Energy
Commission report DEMO 84/7G, 1984.
7. ICRP/90/G-01, Recommeridatioris of the Commission - 1990. Draft, Feb. 1990.
6-100
-------
BULBAR A
JUGOSLAVIA
ALBANIA
. \ \\\\
gr.ee &e ^
c^a>
m
Eduosos
TURKEV
A J * vW« i ¦ ¦¦ in i i i i ¦ ¦ i ¦' i
^ K Mourla
Ionian xŁs
3td
v\\wi\
i^\i ^''iVi'NS
Loutraki
*
& Aegean MjŁ——
^ Sea Ikaria
$, % o
9
_ i
0 • y'
cJ^
f
"3
Cretian Sea
H,
V cL
Mediterranean Sea
Fig.l. Locations of the regions of the major Greek radon spas.
6-101
-------
Session VII
Oral Presentations
State Programs and Policies Relating to Radon
-------
WASHINGTON STATE'S INNOVATIVE GRANT:
COMMUNITY SUPPORT RADON ACTION TEAM FOR SCHOOLS
Patricia A. McLachlan
Department of Health
Olympia, Washington 98504
ABSTRACT
In February, 1990, the Environmental Protection Agency awarded
the Washington State Department of Health $100,000 from the State
Indoor Radon Grants Program to fund an innovative project titled,
"Community Support Radon Action Team for Schools." The Department
of Health contributed an additional $34,000 to the project and
organized a team of public and private sector experts. The goal of
the team was to write a manual of cooperative and cost-effective
approaches school administrators could use to assess and mitigate
radon exposure in schools.
The team of federal, state and local experts from the fields
of health, education, energy, building science and codes, safety,
administration, communication, and radon testing, diagnostics and
mitigation, chose to write and evaluate the manual with the
cooperation of a school district in northeastern Washington.
The manual includes chapters on administrator's overview,
radon facts, radon awareness, radon and liability, strategic
planning, public informational materials, school radon testing,
building inspection and radon diagnostics, radon mitigation, long-
term radon management, and case studies. These chapters and the
experience gained in their application in the school district will
be discussed.
7-3
-------
INTRODUCTION
Despite the identification of elevated radon exposure levels
(100 pCi/L) in some buildings in Washington State's five
northeastern counties, very few residences, schools or public and
commercial buildings have been tested or mitigated. Within the
Core Radon Program, the Department of Health (DOH), the State's
lead agency responsible for a radon program, has insufficient
resources to help school districts that want to tackle their radon
problems but lack the funds and technical expertise.
DOH with its State Radon Task Force encourages state agencies,
local governing bodies and other organizations to work
cooperatively to reduce the public health risk from radon. In
discussions with personnel in the school community, government
agencies and the private radon industry, DOH found a manual was
needed to help schools resolve radon issues. Since a variety of
relevant expertise was present in Washington State, a team approach
to developing the manual had merit. EPA agreed and awarded funding
for DOH's innovative project, "Community Support Radon Action
Team."
THE RADON ACTION TEAM
The Community Support Radon Action Team, a group of public and
private sector experts, met ten times from March, 1990 to February,
1991 and numerous times in small working groups to develop the
School Radon Action Manual. The team was composed of health, radon
and building science experts from DOH, Region 10 EPA, the
Washington Energy Extension Service (WEES), the Spokane County
Health District (SCHD) and the City of Spokane Building Services
Department (SBSD). Also, Faytek, Inc., Quality Conservation and
Thomas J. Gerard & Associates, Inc. from the private sector
provided expertise on radon testing, diagnostics and mitigation,
and HVAC (heating, ventilation and air conditioning) systems,
respectively.
In addition, the school community was represented on the team
by a manager of state school facilities from the Office of the
State Superintendent of Public Instruction (OSPI); a writer and a
safety coordinator from the Education Service District 101 (ESD
101, one of nine regional agencies in Washington providing
administrative and instructional support to local school
districts); an administrator experienced in school radon testing
from Spokane School District 81 (SSD); and an administrator, a
public information officer, a supervisor of school maintenance and
an HVAC specialist from Central Valley School District 356 (CVSD).
The Northwest Regional Foundation (NRF, a private, non-profit
corporation committed to facilitating change in communities)
7-4
-------
provided a facilitator to help this group of people work as a team
to write the manual and evaluate its application in the CVSD.
Finally, as a legal consultant, a Washington State Assistant
Attorney General (OAG) contributed his expertise about liability.
THE TEAM PROCESS
In initial meetings, team members discussed the project goal,
participants' self-interests, the process of radon problem solving
in a school community, manual contents, working group assignments
and site selections for the case studies in CVSD. The major goal of
the team was to compile educational, problem solving and
organizational resources in the School Radon Action Manual.
The manual was designed to help school personnel communicate with
their school community about radon and use internal resources as
well as the private sector to cost-effectively assess and remediate
for radon in their schools.
Radon Action Team members represented a wide variety of
expertise and self-interests. Health professionals focussed on the
need to communicate the health effects of radon exposure accurately
and effectively to the public. School administrators desired to
test, diagnose and mitigate for radon in a cost-effective manner
while informing and involving their communities. Radon testing
professionals demanded a scientific approach that complied with EPA
interim protocols. Building science professionals viewed each
building as an integrated system that demanded careful, logical
problem solving techniques to tackle radon and other indoor air
quality problems. Team members decided to pool their knowledge and
concerns in the manual development realizing they had differences
in perspectives and opinions which would be debated during the
writing process.
In fact, many vigorous discussions did take place over the
course of the year. One often debated question was: How much
testing, diagnostics and mitigation work should school personnel
attempt before they call in the private sector? A second question
was: How should a school district communicate about radon to its
community which often wants problems solved immediately? It has to
be pointed out that the district needs to train its staff, hire
consultants, request bids, raise funding and plan to remediate
radon problems as they are discovered over several years of testing
and diagnostics A third question was: How can schools achieve
cost-effectiveness? Team members decided they could provide
accurate and concise guidance on such things as radon testing
options, building inspection and radon diagnostics, and public
informational materials. They concluded, however, that ultimately
it would be school administrators with an intimate knowledge of
their communities and resources who with the help of the manual
would answer these questions.
7-5
-------
SCHOOL RADON ACTION MANUAL
The School Radon Action Manual contains sections designed for
school district administrators, public information officers,
building managers and maintenance personnel. The manual
organization follows a school radon action process (illustrated in
Figure 1) recommended by the team. A list of the sections with a
summary of their contents follows:
The "Administrators' Overview" includes what radon is, where
radon is found, what the health risks are and when radon was
recognized as a health hazard. Other topics are how radon enters
a building, how it is measured, how radon concentrations are
reduced and who can perform the radon testing and mitigation. In
this section, the team recommends that school district staff
involved in radon testing, building diagnostics or mitigation
attend an EPA endorsed training course. Also, it is recommended
that school district consultants show that they have successfully
participated in EPA's Radon Measurement Proficiency Program, or
employ individuals who have passed EPA's National Radon Contractor
Proficiency Program.
"Radon Facts" gives greater detail on radon discovery, radon
and radon progeny, the health effects of radon, the health risk to
children and comparisons of school to home exposures.
"Risk Awareness" deals with assessing risk, the nature of
radon risk, 4 pCi/L as an action level, the Indoor Radon Abatement
Act, the health risk to children and smokers, challenges to EPA's
risk estimates, getting to ALARA (As-Low-As-Reasonably-Achievable) ,
managing and communicating radon risk and the risk awareness
process.
"Radon and Liability" concludes that health hazards presented
by radon and indoor pollution in schools and public buildings may
be substantially reduced by technical analysis of the problem and
a careful administrative response from management. Failure to
initiate the analysis and respond to the problem presents the risk
of liability for any school or public institution. Suggestions for
a program that schools can develop to deal with radon and other
indoor air quality problems are given.
"Strategic Planning" deals with how a school district may
develop a plan for dealing with radon in its buildings. Topics
covered include prerequisites for planning, action steps, timelines
and financing. Formulation of a radon action team is recommended.
"Public Informational Materials" includes internal and
external communications strategies utilized by the Central Valley
School District as it dealt with radon in its schools. It includes
7-6
-------
ADMINISTRATOR
OVERVIEW
RISK
AWARENESS
RAOOH
ANO
LIABILITY
RAOON
FACTS
TEST
YES
NO
STRATEGIC
PLANNING
SCHOOL RADON
ACTION PROCESS
SCHOOL
RAOON
TESTINO
PUBLIC
INFORMATIONAL
MATERIALS ,
BUILDING
INSPECTION
ANO RAOON
DIAGNOSTICS
YES
SYSTEM
SELECTION
RAOON
MITIGATION
ELEVATEO
TESTS
NO
EVALUATE
LONG-TERM
RAOON
MANAGEMENT
CASE
STUDIES
FIGURE I. School Radon Action Process
RECONSIDER
AT LATER
DATE
DOCUMENT
ANO REPORT
TO PUBLIC
7- 7
-------
strategies for communicating with staff, administrators, students,
parents and the news media. Sample press releases and letters are
included in this section.
"School Radon Testing" deals with testing school buildings for
radon. It provides information on qualifications necessary to
perform testing, school district requirements, testing procedures
and forms, testing methods with advantages and disadvantages, and
evaluation of testing results.
"Building Inspection and Radon Diagnostics" describes how to
inspect buildings and perform or oversee diagnostic testing for
radon entry locations. It includes checklists for review of
testing data and for mechanical and structural inspections.
"Radon Mitigation" provides strategies for mitigation if
elevated levels of radon are found in a building. Topics include
radon entry, causes of pressure differentials, variations in radon
concentrations between rooms, mitigation techniques and dealing
with contractors.
"Long-term Radon Management" provides the basics for continual
monitoring of indoor air quality, including radon, for the
district. Topics include team design, program design, public
policy guidelines and documentation.
"Case Studies" documents the application of the manual in six
schools in the Central Valley School District (CVSD). This
section describes the radon action process that CVSD employed with
the manual and team members' expertise. Based on this limited
application of the manual in one school district, the team offers
suggestions to other school districts.
The "Glossary" defines key terms school district personnel
must understand to communicate meaningfully about radon as a public
health issue. The "Bibliography," a "Team Members' List" and
"Federal and State Contacts" complete the manual.
CASE STUDIES
During this project, part of the manual's school radon action
process (see Figure 1) was evaluated using six buildings in the
Central Valley School District of Spokane County, Washington. The
school community (including school board, faculty and staff,
parents and students) was informed of the project and the radon
action process by Radon Action Team members through the use of the
sections: "Radon Facts," "Risk Awareness," and "Public
Informational Materials." School personnel were trained by team
professionals using parts of the sections: "School Radon Testing"
and "Building Inspection and Radon Diagnostics." The manual
7-8
-------
sections, "Administrator's Overview," "Radon and Liability,"
"Strategic Planning," "Radon Mitigation," and "Long-Term Radon
Management" were still being developed during this time so they
were not evaluated in this school district.
Selected for evaluation of the educational, communications,
testing and diagnostics processes were three elementary schools, a
junior high school, a high school and an administration building.
Team members made presentations about the project and the school
radon action process at meetings of the school board,
administrators, faculty, staff, Parent Teacher Association (PTA)
and press. Four junior high science instructors wrote model radon
awareness curriculum which they taught and are refining for
distribution next summer. Literature on radon was displayed and
made available to staff and the public in the building reception
areas. Letters were sent home to parents, and articles published
in the newspapers. A spirit of openness and cooperation was
encouraged by the radon action team and the school administrators.
The team decided to employ charcoal canisters to test the
administration building and the high school and electrets to test
the other four schools. In both cases, school maintenance
personnel were given training from the manual in placing the
detectors, retrieving them, and keeping records. Charcoal devices
were sent to the manufacturer's lab for analysis while electrets
were read by school personnel. Faytek, a private EPA proficient
testing company, provided training and oversight throughout the
whole testing process.
Elevated radon levels were found in the administration
building, high school, and three elementary schools. Building
inspection and radon diagnostics were performed by both team
members and school personnel. School personnel provided
information about building histories and basic building operation.
They completed some initial mitigation involving sealing cracks and
adjusting HVAC systems. Most of the detailed diagnostics was
performed by radon professionals from Quality Conservation, EPA
proficient contractors, and a mechanical engineer from Gerard and
Associates, all team members. Quality Conservation developed
remediation plans for two elementary schools and the high school.
Due to time and funding limitations, the team's efforts ended after
the three remediation plans were given to the Central Valley School
District.
SUGGESTIONS FOR SCHOOL PERSONNEL
Some suggestions emerged from the Radon Action Team's work on
the Central Valley School District case studies. This school
district is to be applauded for its progressive approach to radon
problem solving and its offer to share lessons learned from this
7-9
-------
project with other school districts.
Suggestions are as follows:
As a part of strategic planning, the team recommends that
schools assemble a Radon Action Team which incorporates relevant
expertise from both the public and private sectors. The team
organized for this project provides a model for team member
selection (although smaller teams are appropriate for individual
school districts). It is important that regular meetings of this
team be held and that progress reports be shared with and decisions
supported by upper level administrators (school board members,
superintendents, district level administrators, and principals) in
the school district. As part of the school district's operations
strategy, it is recommended that EPA testing protocols should be
followed. Decision points and procedures for immediate risk
interventions should be developed. Thought should be given to
scheduling, and minimizing class disruptions and loss of detectors.
During the public informational process, we suggest that a
public information officer or a superintendent be the primary
contact for all information requests. This contact person should
be a team member, well-informed about radon issues, activities in
the schools and the district's strategic plan. Requests for
information should be answered accurately, openly and quickly, with
a timeline given for the radon action process (eg. when test
results will be reported, when buildings will be fixed). A good
relationship should be established with the press at the outset of
the process. The contact person needs to be flexible, calm and
ready to handle "incidents" with concerned individuals and groups.
Staff in buildings with preexisting indoor air quality problems may
show a heightened interest or sensitivity to radon testing. More
communication may be needed. PTA meetings work well to inform
parents and faculty. The public information officer should be
accompanied by other team members who have expertise in radon
health effects, testing, diagnostics and mitigation, to gain public
credibility through answering a broad range of questions.
Before a school district begins the radon testing process,
school personnel should evaluate the various options for testing,
considering cost-effectiveness, available internal and external
resources, liability issues and time constraints. These options
include use of private testing firms, use of school personnel or a
combination of the two. If school personnel will be involved in
performing school radon testing, the team suggests that school
personnel (perhaps one or two maintenance personnel) receive EPA
approved training. Also, a private, EPA proficient testing company
should be employed as a consultant to oversee placement, retrieval,
and recording of test results. Quality control procedures must be
performed for both charcoal canisters and electrets. If school
personnel read electrets, they must have training in appropriate
7-10
-------
analytical techniques. Clear procedures should be used from the
outset to ensure accurate and complete record keeping of data.
Building maps should show radon levels for each room and be color
coded to make radon hotspots evident. Originals of all data should
be kept in one file.
During testing, building inspection and radon diagnostics,
plans and maps are required. Team members found that both
architectural drawings and AHERA (asbestos) plans were sometimes
inaccurate, inadequate or missing altogether. Also, the district
public information officer needs to be prepared to answer concerns
from school community members who want immediate action to lower
radon levels. Radon diagnostics and mitigation take time and money
to perform. Scheduling testing before a holiday such as winter
break may give the district more time to perform this process.
SECOND YEAR PLANS FOR THE SCHOOL RADON ACTION MANUAL
The Washington State Department of Health (DOH) was awarded
second year EPA funding to complete the review process for the
draft School Radon Action Manual and publish the Manual. Also, the
Radon Action Team will design training curriculum and materials for
state and national presentations on the School Radon Action Manual.
Trained instructors from the Radon Action Team will be available
for presentations to regional and national conferences of school
administrators, teachers, facilities maintenance personnel, public
relations staff, public health officials, and radon program and
industry representatives. Finally, evaluation of the Manual's use
in one Demonstration School District will result in a revised
edition, if appropriate.
ACKNOWLEDGEMENTS
The author wishes to acknowledge the contributions made to the
School Radon Action Manual by the following Radon Action Team
members: Jerry Leitch and Misha Vakoc (Region 10 EPA); Mike Nuess
and Rich Prill (WEES); Michael F. LaScuola (SCHD); Robert L.
Stilger (NRF); Bob Eugene and Steve Belzak (SBSD); Mike Roberts
(OSPI) ; Jim Kerns and Dick Moody (ESD 101) ; Dave Jackman, Karl
Speltz and Skip Bonnucelli (CVSD); Jody Schmitz (SSD); Ray Tekverk
and Jan Fay (Faytek); John Anderson and Jack Bartholomew (Quality
Conservation); and Tom Gerard (Gerard and Associates). Other
important contributions were made by T.R. Strong, Robert R. Mooney,
Kate Coleman and Ed Scherieble (DOH); Dick Sovde (CVSD); and Larry
Watters (OAG).
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.
7-11
-------
KENTUCKY INNOVATIVE GRANT
RADON IN SCHOOLS' TELECOMMUNICATION PROJECT
M. Jeana Phelps
Radon Program, Radiation Control Branch
Division of Community Safety
Kentucky Cabinet for Human Resources
Frankfort. Kentucky
and
Carolyn Rude-Parkins
Occupational Training and Development
University of Louisville
Louisville, Kentucky
The Work described in this paper was not funded by the U.S. Environmental Protection Agency and there fore the contents do
not necessarily reflect the views of the Agency and no official endonement should be inferred.
ABSTRACT
One of the many challenges facing the U.S. Environmental
Protection Agency and individual state radiation control programs
is to provide decision-making information and technical support
to school administrators about radon testing and mitigation in
school buildings.
This paper provides information on the development, delivery
and overall implementation of a model radon telecommunication
outreach program to school administrators in Kentucky and to the
following states: Alabama, Arkansas, Georgia, Louisiana,
Michigan. Missouri, Mississippi. North Dakota, Nebraska, New
Jersey. Ohio, Pennsylvania, Texas, South Carolina. Wisconsin, and
West Virginia, Virginia, and Florida.(Figure 1)
The two hour interactive broadcast will be delivered by
satellite to all locations through the Kentucky Educational
Television Network (KET).
Kentucky Educational Television
• Alabama
• Arkansas
• Georgia
• Louisiana
• Michigan
• Missouri
• Missisippl
• North Dakota
> Nebraska
• New Jersey
• Ohio
• Pennsylvania
• Texas
• South Carolina
• Wisconsin
• West Virginia
• Virginia
• Florida
Figure 1: KET Star Channel Network
State Contacts
Star Channel Network
7-13
-------
In schools around the world, students are "talking" to other
students they may never see; teachers are instructing students
who are not in the room; professionals are meeting with far away
colleagues without leaving their desks. Distance learning
concepts and technologies make these interactions possible.
Distance learning is defined as the application of
telecommunications and electronic devices which enable students
to receive instruction that originates from some distant
location. It involves four major concepts:
1. students are seperated from one another and from the teacher;
2. interactive two-way technologies are used to unite them;
3. the learning is planned, delivered, and evaluated by an
institution;
4. students organize themselves and collaborate around a goal
which may be theirs or the teacher's (1,2).
SEPARATED
Virtually any subject can be taught via distance learning.
During 1989-1990, the National University Teleconference Network
(NUTU) offered 94 programs, both fee-based and free. The
greatest percentage (41%) of these programs were directed to
Engineers with the remaining being delivered to: Education
(38.6%), Medical and Allied Health (27.3%), and general interest
(34.1%) (3). Distance learning provides a connecting network to
learners who may be in the next building, another city or in a
different country. Distance learning also allows the learners to
"tune-in" at a time and location convenient to their schedule.
UNITED
The key to distance learning is two-way interactive
technology. In the past, print technology and the U.S. Postal
Service delivered correspondence courses to students in dispersed
locations. Radio and telephone added an audio component and the
possibility for two-way interaction. Later, education television
broadcasts, cable programs, and video cassettes offered other
ways of distributing course materials. Interactivity was added
via telephone and then computer.
Today, computers, satellites and telecommunications
technologies such as fiber optics expand the power'of older
distance learning technologies. Computers make it possible to
access electronic mail, bulletin boards, interactive computer
conferences, data bases of information and computer assisted
instructional lessons and courses. Microwave dishes and
satellite links enable the computer to reach across the street
and around the world. Telephones, fiber optics and satellites
permit one-way and two-way interactive video conferencing.
7-14
-------
Distance learning technologies fall into the categories of
video, voice and data communications. Video is the primary means
of delivery. It is the most attractive, complex and expensive.
Audio telephone delivery provides one-on-one interaction or
handles groups of three or more in different locations via an
audio bridge. Audiographic teleconferencing permits the
transmission of still images and audio signals over telephone
lines. Electronic blackboards and tablets. FAX, slow scan, and
compressed video also use telecommunications technologies.
Video, voice, and data communications can reach learners at
any time and in any place. But the real power of distance
learning lies in two-way interactive technologies (2). The
reason for this is the capability for interpersonal
communication. One-way systems leave the learners passive,
uninvolved and isolated. Two-way systems let them actively
exchange ideas, information and feelings. It puts high touch
into high technology. They can reach out and touch a real
someone.(2)
KENTUCKY EDUCATIONAL NETWORK (KET)
Kentucky is the first state in the nation to fund and
construct a statewide telecommunications delivery system. KET
was established by the Legislature in 1968 to provide
instructional services to teachers. The Satellite uplink costs
$500,000 and $2,000 for each school downlink. It is the largest
system in the nation. The KET Star channel system consists of a
transmitting and receiving site at the KET Network Center in
Lexington, Kentucky and has more than 1,300 receiving sites (down
links) located throughout Kentucky at public schools. The KET
satellite can uplink programs not only to Kentucky schools but
also to schools throughout the nation (Figure 1 and Figure 2).
Statewide Telecommunications Service To:
Elementary/Secondary Schools
Vocational Schools
Colleges and Universities
Figure 2: KET Reaches all Public Schools in the State
7-15
-------
This potential nationwide KET linkage will allow the Kentucky
radon program to offer the "Radon in Schools" broadcasts to
radiation control offices and education departments in the
following states.
Alabama
Georgia
Michigan
Mississippi
Nebraska
Ohio
Texas
Wisconsin
Virginia
Kansas. Oklahoma. Tennesse
currently KET STAR Channel par
states can be arranged.
Arkansas
Louisiana
Missour i
North Dakota
New Jersey
Pennsylvania
South Carolina
West Virginia
Florida
, and North Carolina are not
icipants, but linkage to these
TARGET AUDIENCE
The radon program, funded through the U.S. Environmental
Protection Agency. State Indoor Radon Innovative Grant, will
apply KET's high technology approach to providing radon
information to school administrators, school building managers,
daycare operators, and others engaged in managing buildings where
citizens learn and work.(Figure 3)
Once the broadcast dates have been scheduled and
pre-broadcast materials developed, all state radiation control
offices, radon programs, state departments' of education and
other target participants will be invited to link up and take
part in this project.
School
Facility
Managers
STAR
School
Administrators
Private/Public
KET
State
Licensed
Daycare
Centers
Radon Testing
and
Mitigation Info
To
SATELLITE
Colleges
Universities
Technical School
Admhiistrators
Figure 3: KET/Radon Target Audience
7-16
-------
For Kentucky public school officials, the radon in schools
broadcast will arrive soon after they receive confirmatory radon
measurements for their building(s). For others such as daycare
and nursing home operators, the information will arrive as they
begin to consider how to go about testing and mitigating their
buildings.
By using the Star Channel satellite and distance learning
concepts, the radon program staff can reduce the need to travel
statewide to disseminate radon information to these groups. An
additional benefit of this communication media is that the
audience is exposed to a consistent message with minimal
presenter bias.
KET-RADON PROJECT GOALS
The second KET/Radon Innovative grant goal will go beyond the
initial broadcasts to research the potential for using
telecommunications to deliver information about radon. To do
this, the KET project staff will investigate existing
telecommunication networks and apply this knowledge to the
delivery of radon information and technical training. Another
aspect to be investigated will be the possibility of linking the
USEPA Regional Radon Training Centers, through a central training
delivery system with on-site training center faculty in
designated locations (Table 1).
TABLE 1: KET/RADON INNOVATIVE GRANT GOALS
KET - Radon Project Goals
• Design and Deliver two Radon Programs
utilizing the Star Satellite and
telecommunications.
• Investigate the potential for using KET Network
to deliver Radon information, Training, and
Continuing Education.
7-17
-------
SCHEDULED BROADCASTS
The Spring broadcast date is expected to be scheduled prior
to the time school officials finalize their 1991-1992 budget
requests. Hopefully, the timing of the broadcast and the
technical information provided, will be sufficient to allow
school officials to request funding for radon testing and
mitigation projects (Table 2).
TABLE 2: BROADCAST SCHEDULE
Radon Broadcasts
• Spring 1991 - Testing
•Fall 1991 - Mitigating
The Spring broadcast will emphasize radon testing protocols
for schools and illustrate ways to communicate radon risk
information to parents, staff, children and the public (Table 3).
TABLE 3: SPRING BROADCAST DESIGN
Spring 1991
Broadcast Design
• Testing for Radon in Schools
• Decision steps based on
• Initial Screening Measurements
• Confirmatory Measurements
• Risk Communication Suggestions
• Resources Available
• Live Panel of Radon Experts for Teleconference
7-18
-------
Kentucky school representatives, experienced in radon testing
and mitigation in their own school building(s). will share
insights and lessons learned. These segments will be pre-taped
in the actual school setting. Also, the radon school based
outreach program being implemented by the Jefferson County Public
School District and Jefferson County Parent-Teacher's Association
will be highlighted (Table 4).
TABLE 4: FALL BROADCAST DESIGN
Fall 1991
Broadcast Design
• Mitigating for Radon in Schools
• Decision Steps
• Risk communication suggestions
• Resources Available
• Live Panel of Radon Experts for
Teleconference panel
The Fall Radon in Schools broadcast will provide school
mitigation information. Again, Kentucky school representatives
will share their experiences in an attempt to communicate the
message that radon issues are manageable within a school setting
and that there are others who have survived the process (Table 4).
Each two hour broadcast will feature a "live" interactive
question and answer period for the last 30-40 minutes of the
program. A panel consisting of radon experts from U.S. EPA and
other agencies will respond "on air" to participants' questions.
The audience will be provided with an 800 telephone number to the
KET studio and encouraged to call in with their questions. The
800 call-in number will remain active for one hour after the
broadcast. Trained radon staff will answer the incoming calls,
write out the questions and route them to a panel member.
A panel member will read the question aloud and provide an
answer or refer the caller to additional resources. Any caller
who does not have their question addressed on live broadcast will
receive a written answer by mail following the program. Callers
will not be identified by name or school and their voices will
not be aired.
7-19
-------
The entire broadcast plus the live panel call-in segment will
be taped for future broadcast schedules. Individual tapes will
be made available through a loan program (Figure 4).
BROADCAST PRODUCTION
io
1. Instruction it
presented before a
earners In the KET
aludio In Lexington.
Television transmission
b beamed to satellite
ia earth orbit.
7. Expert panel
monitors audience
quextkxu (voice &
data) aad provides
advice.
Distance Learning
Process
6. Information from
telephone la delivered to
the studio by telephone
lines.
ir
3, Relayed to a
school's
sateHite dish.
The audience watches
the lesson on
television.
5. Onslght raclllltation.
Figure 4: Distance Learning Process
Pre-broadcast activities can be divided into two major steps;
audience preparation and broadcast production processes. In
regard to the participants, the State Department of Education,
Human Resources Licensing and Regulation Department, and KET will
assist by providing mailing labels for the target audiences. The
project staff person, in coordination, with the USEPA and KET and
others, will prepare the broadcast design. KET staff will take
the lead in taping and editing broadcast visuals. Field taping
of actual school settings is planned as part of a pre-production
activity. Existing video productions may also be integrated.
Other pre-broadcast activities include sending a radon
resource literature package along with each invitation. In each
school district, a KET receiving school will be identified as a
radon resource broadcast center. These resource centers will
provide on-site facilitation during and after the broadcast by a
"radon trained" resource person. County Extension Agents, school
representatives, local health officials, and others with
demonstrated knowledge of radon testing and mitigation will be
designated to serve as on-site facilitators. The facilitators
will receive a pre-broadcast orientation to ensure statewide
consistency of information. For security reasons, all non-school
related participants (daycare operators, etc.) will be assigned
to a resource center. Closest to their home/work.
7-20
-------
EVALUATION
The nature of this project requires an extensive system of
checks and balances at each phase of development. This will be
provided by the KET staff assigned to the project, as well as the
project designated radon staff person. USEPA Region IV staff,
and Regional Radon Training Center staff will provide ongoing
assistance and review. An in-state work group will also be
assigned to on-going evaluation. Members of this group will
include, staff from the radon program. State Department of
Education; Jefferson, Fayette, and Warren County School
Districts; University of Kentucky Cooperative Extension Service.
Kentucky Parent-Teachers Association, and others.
An immediate post-broadcast evaluation questionnaire and a
six week follow-up will be administered to all participants. The
results from the immediate and delayed evaluation will serve to
measure the degree of change in participant attitude regarding
radon.
ACKNOWLEDGEMENTS
Special thanks to Tim Tassie, KET Acting Deputy. Executive
Director of Broadcasting, for giving flight to the radon distance
learning ideas outlined in this paper.
Also, to Paul Wagner/Patsy Brooks, USEPA Region IV, James
Vaughn, Jefferson County Public School District, Sharon Solomon,
and Virginia Maurer, Parent Teachers' Association of Kentucky,
Jim Judge, Ky Department of Education, and to Chick Craig and his
team who have provided assistance to Kentucky schools.
References
1. Carl, D.R. (1989). A response to Greenville Rumble's "On
defining distance education". American Journal of Distance
Education. 3(3), 65-67.
2. Rude-Parkins, Ph.D., Carolyn, (1990) Distance Learning
Technology and Education; Key Ideas for Kentucky Educators.
Kentucky Department of Education.
3. National University Teleconference Network (NUTN). NUTN News,
Vol. 8, No. 3, Winter 1990.
7-21
-------
Appendix A
JEFFERSON COUNTY SCHOOL DISTRICT
AND
RADON PROGRAM
1. Work cooperatively to establish a radon comunication outreach program
through the school to parents, staff, and students.
2. Promote the Jefferson County School District radon testing project and the
communication outreach program as a model for other school districts in the
state and the nation.
MODEL COMMUNICATION OUTREACH
Audience
School Administrators Building Maintenance Personnel
Teachers Ancillary School Staff
Parents Parent-Teachers Association
Students
Message
Long-term exposure to elevated levels of rdon gas is associated with increased
risk of developing lung cancer.
Testing for radon gas is easy and mitigation methods are effective.
All homeowners should test their homes. Schools, daycares, public and commerical
buildings should also be tested.
If elevated levels are discovered, action should be taken to reduce the levels.
7-22
-------
RADON COMMUNICATION OUTREACH PROGRAM THROUGH SCHOOLS
Audience
Type of Communication
Menage
School Administrators
Building Maintenance
Personnel
Technical/Support
and Motivational
• Testing Protocols
• Decision process
after testing
• Mitigation strategies
• Technical assistance
- Public disclosure
- Encourage them to
test their homes
Teachers Informational and -Levels of radon in
Ancillary School Staff Motivational school, by room
• Mitigation strategy
• Encourage them to
test their homes
Parent-Teacher's
Association
Informational and
Motivational
- Levels of radon In
school, by room
- Mitigation strategy
- PTA can help school
administrators to
reach parents with
radon information
- Encourage them to
test their homes
Parents Informational and -Levels of radon in
Motivational school, by room
•Mitigation strategy
- Encourage them to
test their homes
Students Informational and • Facts about radon
Motivational and indoor air quality
- How to improve air
quality
*Risk Communicatioh information
Provided for each Group
Method
Kentucky Educational
Television-Radon in
Schools Broadcast
Spring/Fail 1991
• Dissemination of
informational literature
-Presentations through the
Kentucky Education
Association
- Presentation at Stat*
and District Meetings
and Workshops
- Assistance to individual
schools/districts
- PTA host Radon Awareness
Program
•PTA distribute radon info
to parents
• Host testing campaigns
- Dlstrlct/School/PTA
sponsored radon awareness
programs
• Distribution of radon
literature through PTA/
District office
- American Lung Association
lesson on radon in
"Growing Healthy"
curriculum
-Weekly Reader Poster
contest
-------
REGULATION OF RADON PROFESSIONALS BY STATES: THE CONNECTICUT
EXPERIENCE AND POLICY ISSUES
Alan J. Siniscalchi, M.S., M.P.H. (1); Zygmunt F. Dembek, M.S. (1),
Nicholas Macelletti, M.S. (1); Laurie Gokey, M.P.H. (1); Paul Schur,
M.P.H. (1); Susan Nichols, B.S. (2); and Jessie Stratton, B.S. (3)
(1) State of Connecticut Department of Health Services
(2) State of Connecticut Department of Consumer Protection
(3) State Representative, Connecticut General Assembly
ABSTRACT
The desire of state governments to provide information on proficient
radon professionals to their citizens has resulted in a variety of
informational and regulatory programs. The State of Connecticut, has
developed a new registration program for radon professionals, pursuant
to Connecticut Public Act 90-321, that combines requirements for
successful completion of both state registration and federal proficiency
programs in order to be listed. Under this program, radon testers and
testing companies are required to successfully participate, in the
current round of the EPA Radon Measurement Proficiency (RMP) Program.
Radon mitigation contractors must successfully participate in the Radon
Contractor Proficiency (RCP) Program and register with the Department of
Consumer Protection. Radon Diagnosticians are required to successfully
participate in both the RMP and RCP Programs. The process by which
Connecticut developed this program and recommendations for EPA policy
changes are discussed.
7-25
-------
INTRODUCTION
The State of Connecticut Department of Health Services (DHS) began
receiving inquiries on radon and requests for information on radon
professionals soon after high levels of radon were discovered in a home
in Pennsylvania. Many of the callers requested our recommendations for
radon testing companies and mitigation contractors. Most callers
wanted information on what types of testing services were available,
what prices were reasonable, which companies had the most experience
and whether they could conduct the test themselves. Callers who had
already tested asked for information on mitigation systems and lists of
"approved" mitigation contractors. Some of the better informed callers
also asked about whether it is appropriate for the same company that
conducted the radon test to also conduct the mitigation operation. In
time, other consumers also reported problems with the radon testing and
mitigation companies.
The DHS response to these requests has evolved through a number of
changes. These changes have included both the information provided on
radon assessment and control and the means by which we regulate radon
professionals. Our ultimate goal is to provide sufficient information
to Connecticut consumers to enable them to make informed choices on
testing and mitigation services.
This paper will outline the evolution of radon professional
regulation in Connecticut, provide recommendations for other state
programs, and propose changes for the EPA proficiency programs.
EARLY GUIDANCE ON RADON PROFESSIONALS
The DHS first attempt at providing guidance to Connecticut
consumers on radon professionals was the development, in 1985, of a
list of radon testing companies. The information was obtained from
results of an early round of the EPA's Radon Measurement Proficiency
(RMP) Program. The DHS determined that additional information on both
the types of testing services offered by each company and the price of
each testing service would also be beneficial. Information on DHS
recommendations regarding screening devices and procedures was also
included.
During 1986 the DHS provided the public with a list of mitigation
contractors used by the EPA in the northeast since little information
was available on Connecticut mitigation companies. As mitigation
companies became established in Connecticut, the DHS expanded the list
to include information on mitigation company services. Since this list
of contractors was not derived from a state or federal proficiency
program, disclaimers were added to warn the public that the DHS could
not be responsible for a company's performance.
7-26
-------
During 1986-88 the DHS prepared additional informational material
on radon exposure in Connecticut that was mailed in response to
telephone inquiries on radon testing. This material included fact,
sheets summarizing the results of the various statewide radon surveys,
and the two EPA radon pamphlets (1,2).
In August 1987, the DHS held a news conference to announce the
results of our second statewide survey. At this conference the
Department recommended that all Connecticut citizens test their homes
for radon regardless of the geographic area of their residence, since
our surveys to date had revealed the potential for high levels of radon
in all areas of the state. The DHS formally organized a Radon Program
in December 1987 to publicize these recommendations and conduct
additional surveys. A third survey revealed consistent results in the
percentage of homes (20%) with radon levels in excess of the U.S. EPA
guideline of 4 picocuries (pCi/L) per liter (3,4,5). These results
further emphasized our recommendation that all Connecticut residents
test their homes for radon (3,4,5).
During 1989, a second state agency became increasingly involved in
assisting Connecticut consumers in evaluating radon professionals.
This agency, the Department of Consumer Protection (DCP), recommended
that individuals offering mitigation services register as "home
improvement contractors." This existing DCP registration program
provided consumers with additional protection in the form of the "Home
Improvement Guaranty Fund" with monies that can be used to correct
poorly installed or unfinished mitigation systems.
During Lhis time, the Radon Program greatly revised the format of
how information on radon is provided to Connecticut residents. This
new approach included the development of two information packets.
Information packet "A" included a list of testing companies, a fact
sheet and the EPA pamphlet "A Citizen's Guide to Radon" (1).
Information packet "B" included the EPA pamphlet "Radon Reduction
Methods A Homeowners Guide" (2), and the list of diagnostic services
and mitigation contractors. Although the DCP registration procedures
for mitigation contractors was not mandatory at this time, the DHS
added a notice to the contractor list advising consumers of the
benefits of this program. Table 1 summarizes the other registration
requirements and recommendations. It should be noted that while the
DHS Bureau of Laboratory Services had existing regulatory authority to
require registration of testing laboratories, the DHS could not require
mitigation contractor registration. During early 1990 the Radon
Program also recommended that consumers select mitigation contractors
that had successfully participated in the EPA Radon Contractor
Proficiency (RCP) examination.
REGISTRATION REQUIREMENTS AFTER OCTOBER 1990
Many consumers were not satisfied with the status of our radon
professional lists since these lists did not require contractors to
demonstrate competency by completing a proficiency examination or
program.
7-27
-------
A number of Connecticut State Representatives and agency staff began to
independently suggest alternative approaches toward a more formal
regulation of radon professionals. Many of these alternative
approaches called for mandatory certification or registration
programs. A proposed bill was submitted by the Conecticut General
Assembly's General Law Committee that asked for a mandatory licensure
program for all radon professionals. This proposed program would have
required individuals offering radon testing, diagnostic or mitigation
services to obtain a license from the DHS, that would be renewable on
an annual basis. While the bill would have offered consumers better
protection against fraudulant. radon professionals than a registration
program, it would have required additional agency staff to implement.
The DHS testified against the bill, pointing out that due to fiscal
constraints and a lack of staff it would not be possible to implement
the legislation if it was enacted.
TABLE 1. REQUIREMENTS FOR LISTING OF RADON PROFESSIONALS
IN CONNECTICUT PRIOR TO OCTOBER 1990
Professional Class
Requi reraent Testing Diagnostics Mitigation
primary and
secondary
Department of; Health
Services (DHS) Registration Yes Yes Yes
EPA Radon Measurement
Proficiency (RMP) Program Yes No No
F.PA Radon Contractor
Proficiency (RCP) Program No Rec.* Rec.
Education Rec. Rec. Rec.
Department of Consumer
Protection (DCP) Registration No No Rec.
^recommended
One of the authors (Representative Jessie Stratton) began work on
an alternative proposal that would increase the Department's ability to
regulate radon professionals at little cost to the agency. After
meeting with the representatives of appropriate state agencies
(including the other authors) a bill was proposed to accomplish these
goals using a more formal registration program. This proposed bill
survived various committee meetings and hearings and was signed into
7-28
-------
law by Governor William A. O'Neill on May 3, 1990. The requirements of
Public Act 90-321 "AN ACT CONCERNING PERSONS WHO TEST FOR OR MAKE HOME
REPAIRS TO ELIMINATE THE PRESENCE OF RADON GAS AND DIRECTING THE
DEPARTMENT OF HEALTH SERVICES TO ADOPT REGULATIONS ESTABLISHING SAFE
LEVELS OF RADON IN POTABLE WATER AND PAYMENTS FROM THE HOME IMPROVEMENT
GUARANTY FUND" became law on October 1, 1990. This bill stated that
"the Department of Health Services shall publish a list from time to
time of: companies that perform radon mitigation or diagnosis; primary
testing companies and secondary testing companies."
Table 2 summarizes these new requirements.
TABLE 2. REQUIREMENTS* FOR LISTING OF RADON PROFESSIONALS
IN CONNECTICUT AFTER OCTOBER 1990
Professional Class
Requirement Testing Diagnostics Mitigation
primary and
secondary
Department of Health Services Yes Yes Yes
(DHS) Registration
EPA Radon Measurement
Proficiency (RMP) Program Yes Yes No
EPA Radon Contractor
Proficiency (RCP) Program No Yes Yes
Education EPA "EPA "EPA
Measurement** Approved" Approved"
Department of Consumer
Protection (DCP) Registration No No Yes
* Under Connecticut Public Act 90-321
** Required by the DHS Bureau of Laboratory Services only
One should note the most significant changes relate to the
registration process for radon professionals. Diagnosticians and
mitigation contractors are now required to fulfill federal proficiency
requirements by successfully participating in the RCP Program in order
to be listed by DHS. This requirement of successful participation in
the RCP Program established a minimum level of proficiency for radon
contractors.
7-29
-------
Individuals offering both radon testing and diagnostic services are
specifically required to successfully participate in the federal RMP
Program to be listed by the DHS under PA 90-321.
Both diagnosticians and mitigation contractors are required to
fulfill an educational requirement specified in the bill. This
requirement states that diagnostic specialists and the on-site
supervisor must have "attended a program approved by the United States
Environmental Protection Agency."
Finally, mitigation contractors are required to register with the
DCP as "home improvement contractors." This requirement was added to
the legislation to ensure that Connecticut residents retaining the
services of radon mitigation contractors will be afforded the same
protection available to consumers using the services of any other home
improvement contractor. This protection can include receiving funding
to complete unfinished installations or correct problem systems.
PROBLEMS ENCOUNTERED WITH THE IMPLEMENTATION OF PUBLIC ACT 99-321
The authors have documented the following problems in the
implementation of PA 90-321. The first problem relates to the bill's
reference to "EPA-approved" courses in outlining the educational
requirements for the three radon professional groups. This term,
suggested by EPA staff and others, referred to courses offered by EPA
contractors including the newly organized Regional Radon Training
Centers. It was chosen to avoid the resource intensive problem of state
agency approval of source providers.
A number of radon professionals who wished to be listed by the state
asked if courses they had taken from private vendors were considered
"EPA- approved". Inquiries to the EPA revealed that they did not
"approve" any radon courses at that time, although they did endorse the
courses offered by the Regional Radon Training Centers.
A second problem relates to the infrequent review periods or
"rounds" offered by the EPA Radon Measurement Proficiency (RMP)
Program. Public Act 90-321 now requires both primary and secondary
testing companies and diagnosticians to have "successfully completed"
the RMP program in order to be listed. Therefore, newly organized
companies must wait up to a year or more for the next test round prior
to being listed.
A third problem relates to the requirement of individuals who wish
to be listed as offering services as a radon diagnostician participate
in both the EPA RMP Program and the Radon Contractor Proficiency (RCP)
Program. The final language of the bill allowed any RMP participant
including secondary companies to be listed as a diagnostician. The
authors have found that only those companies with instruments capable of
performing real-time radon measurements and successfully participating
in the RMP Program with these instruments can conduct accurate
diagnostic evaluations.
7-30
-------
A forth major problem relates to the program's design as a listing
rather than a registration program. While the program places little
demand on the Department, it allows radon professionals to conduct
business without requiring proof of proficiency and training.
SUGGESTED CHANGES IN EPA POLICY TO ASSIST STATES IN THE REGULATION OF
RADON PROFESSIONALS
The Department's preliminary experience with PA 90-321 has already
modified the author's thoughts on the definition of the model radon
professional registration program. The following changes are suggested
to aid states in the development of their radon programs. These
suggestions include both changes in EPA policy and recommendations for
states in their interpretation of EPA policy.
Most changes are related to the educational requirements. Table 3
lists recommended educational requirements which call for "EPA-
equivalent" radon training courses specifically designed for radon
testers and mitigators. EPA-equivalent being defined as having a course
content based on courses offered by the EPA or by the EPA Regional Radon
Training Centers.
The change in language from "EPA-approved" to EPA-equivalent will
more accurately reflect the current EPA policy on approval of training
providers. At this time the EPA only approves providers of the
"hands-on" mitigation training course (see below).
TABLE 3. RECOMMENDED REGISTRATION REQUIREMENTS OF RADON PROFESSIONALS
BY STATE AND LOCAL GOVERNMENTS
Professional Class
Requirement Testing Diagnostics Mitigation
primary and
secondary
Health Agency
Registration
EPA Radon Measurement
Proficiency (RMP) Program
EPA Radon Contractor
Proficiency (RCP) Program
Education
Yes
Yes
No
Yes
Yes
Yes
Yes
No
Yes
EPA-Equivalent EPA-Equivalent EPA
Measurement Measurement & Equivalent
Mitigation Mitigation
Consumer Agency Registration
No
No
Yes
7-31
-------
Ideally the EPA would determine if a course provider is offering an
EPA-equivalent course. EPA staff would review new course offerings
submitted by the providers and make a determination of "equivalency".
The EPA could also develop a model accreditation program for radon
course providers. State agencies could determine if a course is
considered EPA-equivalent by comparing the course outline to the EPA
accreditation model. States with limited resources devoted to radon
issues can make this determination by simple comparison of the course
outline.
The measurement course should consist of a one day program while
the mitigation course would include a "hands on" training component. A
list of approved courses that include this "hands on" component will be
maintained by the EPA. This course will include actual practice in
assembling radon mitigation systems. Diagnosticians would be required
to complete both courses, since they must be both proficient, with
testing methods and knowledgeable about mitigation systems.
A second proposed change would require diagnosticians to
participate in the RMP Program as a primary company using a radon
testing device capable of obtaining real-time pinpoint radon
measurements. This type of device is needed to conduct accurate
diagnostic evaluations of homes and other buildings. This requirement
would ensure that diagnosticians will not try to conduct diagnostic
radon measurements with passive radon detection devices.
Perhaps the most significant changes would have these requirements
mandated not only for professionals who wish to be listed, but for all
professionals conducting business in a state.
SUGGESTIONS FOR THE EPA RADON MEASUREMENT PROFICIENCY (RMP) PROGRAM
The EPA's Radon Measurement Proficiency Program has proven
extremely useful for states such as Connecticut in providing
information on competent radon professionals. The authors have
identified a few minor changes that would improve the RMP's Program's
ability to provide useful information to states. These changes are
summarized on Table 4.
The most significant changes again relate to educational
requirements. The authors recommend both prerequisite and continuing
education requirements for all radon testing company personnel involved
with placement and retrieval of radon measurement devices. This
requirement will help to ensure that accurate testing procedures are
followed by professional radon testers. In addition, laboratory
directors will be required to complete a course on techniques and
procedures used in radon analysis. This requirement will ensure that
accurate analysis will be conducted by using appropriate quality
assurance techniques. Both a continuous RMP Program application
process and a continuous randomly selected blind testing are also
recommended. A continuous testing round policy will allow newly-formed
companies to immediately participate in the RMP program. The current
system of periodic "test rounds" has resulted in waiting periods of up
to one year. The continuous blind testing plan will ensure that
laboratories maintain a high level of proficiency throughout the year.
7-32
-------
TABLE 4. CURRENT AND RECOMMENDED REQUIREMENTS FOR RADON TESTING
COMPANIES PARTICIPATING IN THE U.S. ENVIRONMENTAL PROTECTION AGENCY
(EPA) RADON MEASUREMENT PROFICIENCY (RMP) PROGRAM
Company Type & Requirements
Current
Recommended
PRIMARY COMPANY:
EPA Lab. Analysis*,
Education
None
EPA Measurement** or
Equivalent
QA Program (Analysis)
Yes
Yes
Chamber Testing (Submitted)
Yes
Yes
Chamber Testing (Blind)
Selected
Random (year-long)
(test round
period only)
SECONDARY COMPANY:
Documentation of Approval
Listing Only
Yes
by the Primary Co.
Education
None
EPA Measurement**
or Equivalent
QA Program (Sampling)
No
Yes
Chamber Testing (Blind)
No
Random (year-long)
* Prerequisite and continuing educational requirement for laboratory
director only
** Educational requirement for all staff involved with testing device
placement
and retrieval
RECOMMENDATIONS FOR THE EPA RADON CONTRACTOR (RCP) PROFICIENCY PROGRAM
The authors have found the EPA Radon Contractor Proficiency Program
(RCP) to be a useful addition to the RMP Program. We have some minor
recommendations for improving this program's value in providing
information to consumers who wish to contract for installation of
mitigation systems. Table 5 summarizes these changes which again
emphasize prerequisite and continuous educational requirements. The
authors have recommended that the EPA emphasize the separation of the
radon diagnostician and mitigation contractor services. This
separation should be used in both the registration requirements of the
RCP Program, where the Radon Program would recommend a separate
listing, and within the text of EPA literature on radon reduction
techniques (6,7).
7-33
-------
An example of a document where a text change is needed can be seen
in Section 4.1 of the EPA booklet "Application of Radon Reduction
Methods" (7). This section, entitled "Choice of Diagnostician/Miti-
gator," describes the diagnostician but does not emphasize the
distinction:
"The person primarily responsible for the diagnosis of the problem
is called the "diagnostician." The person who will be primarily
responsible for the design, installation, and post-installation
evaluation of the radon reduction system is referred to here as the
"mitigator." These may or may not be the same person (7)."
The authors also recommend that participants in the RCP Program
must agree to follow the "RCP Program Guidelines" in all their
mitigation activities. These guidelines are designed to insure that
only well-designed, efficient systems are installed in homes and only
when needed.
CONCLUSIONS
The authors are of the opinion that adoption of these changes would
increase the value of EPA's Proficiency Programs by providing useful
information which would enable our nation's consumers to make informed
decisions regarding the selection of radon professionals. By requiring
participating in the EPA proficiency programs and following the
recommendations listed on Table 3, states can offer better protection
to their residents. Consumers in those states will know that all
listed professionals possess a minimum level of knowledge and training
in proper testing diagnostic and mitigation procedures. It is our
belief that such information and consumer protection safeguards
contribute to reducing radon exposure and decreased mortality among our
nation's citizens.
The work described in this paper was not funded by the U.S.
Environmental Protection Agency and therefore the contents do not
necessarily reflect the views of the Agency and no official
endorsement should be inferred.
ACKNOWLEDGEMENT
The authors wish to thank Ms. Suzanne Lessard for her expert
assistance and Dr. David R. Brown for his comments in the preparation
of this paper.
7-34
-------
TABLE 5. CURRENT AND RECOMMENDED REQUIREMENTS FOR RADON DIAGNOSTICIAN
SPECIALISTS AND MITIGATION CONTRACTORS PARTICIPATING IN THE U.S.
ENVIRONMENTAL PROTECTION AGENCY (EPA) RADON CONTRACTOR PROFICIENCY
(RCP) PROGRAM
Professional Class and
Current
Recommended
Requirements
Diagnostician:
Prerequisite Education
No*
EPA Measurement and
Mitigation Courses
Participation in the
or Equivalent
EPA Radon Measurement
No*
Yes
Proficiency (RMP) Program
Participation In the EPA
No*
Yes
Radon Contractor Proficiency
(RCP) Program
Continuing Education
No*
Yes
M i t: i p.at: i on Con trar. tor:
Prerequisite Education
Rec.**
EPA Mitigation
Course*** or
Equivalent
Participation in the F,PA
Yes
yes
Radon Contractor Proficiency
(RCP) Program including use of
the RCP Guidelines (see text)
Rec.
Yes
Continuing Education
Rec.**
Yes
* Current EPA proficiency programs do not recognize diagnostic
specialists as a separate professional class
** Will be required after July 1991
*** Required for on-site supervisor only, recommended for all staff
Rec. — Recommended
7-35
-------
REFERENCES
1. U.S. Environmental Protection Agency (EPA). A citizen's guide to
radon. What it is and what to do about it. EPA-86-004 U.S.
Environmental Protection Agency, Cincinnati, Ohio, August 1986. 14
pp.
2. U.S. EPA. Radon reduction methods A homeowners guide. RD-681 U.S.
Environmental Protection Agency, Cincinnati, Ohio, July 1989, 24 pp.
3. Toal, B.F., Dupuv, C.J., Rothney, L.M., Siniscalchi, A.J., Brown,
D.R. and Thomas, M.A. Radon exposure assessment - Connecticut.
Morbidity and Mortality Weekly Report (MMWR) 38:713, 1989.
4. Siniscalchi, A.J., Rothney, L.M., Toal, B.F., Thomas, M.A., Brown,
D.R., van der Werff, M.C. and Dupuy, C.J. Radon exposure in
Connecticut: Analysis of three statewide surveys of nearly one
percent of single family homes. Proceedings of the 1990
International Symposium of Radon and Radon Reduction Technology.
U.S. EPA, 1990.
5. Siniscalchi, A.J. Radon: What you don't know could hurt you.
Connecticut Academy of Science and Engineering (CASE) Reports
5(3):1, 1990.
6. Henschel, D.B. Radon reduction techniques for detached houses.
Technical Guidance (2nd Ed.) EPA/625/5-87/019. U.S. Environmental
Protection Agency, Cincinnati, Ohio, January 1988. 192 pp.
/. Mosley, R.B., and Henschel, D.B. Application of radon reduction
methods. EPA/625/5-88/024. U.S. Environmental Protection Agency,
Cincinnati, Ohio, August 1988. 92 pp.
4871F
7-36
-------
NEW JERSEY'S RADON PROGRAM - 1991
Jill A. Lapoti
New Jersey Department of Environmental Protection
7-37
-------
NEW JERSEY RADON PROGRAM, 1991
Background
Early in 198 5, the Pennsylvania Department of Environmental
Resources contacted the New Jersey Department of Environmental
Protection (NJDEP) and described finding high indoor radon levels
in homes along the geologic formation known as the Reading Prong.
Since the Reading Prong extends from Pennsylvania, through northern
New Jersey, and into southern New York State, it was likely that
a similar hazard existed in homes in New Jersey. A few months
after this initial notification, a greater sense of urgency was
added to the situation as a result of an article about radon and
the Reading Prong which appeared in the New York Times. As a
result of the article, the State received a large number of phone
calls from concerned citizens.
Early on the NJDEP identified two major issues: 1) there was
a potential indoor radon exposure problem in the State which
required testing and remediation whenever necessary, and 2) the
extent of the problem needed to be identified. It would not have
been enough to assume that only the Reading Prong area was
affected, but that was the natural starting place to begin studying
and testing.
A review of available geologic data showed that uranium, of
which radon is a natural decay product, was commonly present in a
greater geographic area of the State than the Reading Prong. Based
on this data, the NJDEP estimated that 1.6 million homes could
potentially be affected. That meant as many as 4 million people
or more might be affected, greater than one half of New Jersey's
population. Two facts were apparent; indoor radon posed an
extremely large potential environmental hazard and no single state
agency had the resources to deal with a problem of such magnitude.
In late 1985, planning began on what actions to take and how to
involve all levels of government, as well as the private sector
wherever possible.
The New Jersey Legislature also recognized the magnitude of
the situation and enacted two separate pieces of legislation
providing $4.2 million and mandating specific activities. The
NJDEP was designated the lead agency and required to develop an
information outreach program to educate New Jersey residents about
the problem and methods of testing and remediation. Additionally
the NJDEP was to institute a program of confirmatory monitoring for
residents whose initial radon tests showed 4 picoCuries per liter
(4 pCi/1) or higher and to also conduct a statewide scientific
study to identify areas at risk for residential exposure to high
levels of radon. Finally the legislation required the NJDEP to
develop a certification program for companies offering radon
testing and mitigation services. The New Jersey Department of
Health (NJDOH) was required to conduct an epidemiologic study to
identify potential risk of lung cancer associated with residential
exposure and also to develop a voluntary registry of residents with
7-38
-------
a radon exposure history.
Activities
The information outreach program that the NJDEP developed,
centers around a toll-free "800" number that is open to callers
every work day from 8:00 a.m. to 5:00 p.m. Since July of 1985 when
the information phone line was first set up, more than 125,000
calls have been logged. Many callers want information about
testing and remediation, so brochures were prepared and a standard
information packet is sent to callers upon request. To date over
60,000 of these packets have been sent out. More than 350
presentations by Radon Program staff have been made to audiences
including homeowners and local officials, realtors, health
professionals, educators and students, testers and mitigators, and
a number of professional groups at conferences convened for the
purpose of information exchange. Other public awareness and
education outreach activities include production of a radon slide
show, which was also converted to a video. Three billboards were
put up along roadways in high exposure areas in an attempt to
generate more awareness about radon testing. Radon Program staff
worked with representatives from New Jersey Transit on a project
to put placards in buses, so as people rode to work or went
shopping they would repeatedly see the radon testing message. A
mass mailing to almost a half million households in the Tier 1
area, resulted in about 40,000 inquiries about radon, its health
effects, and testing and mitigation programs. More recently, an
insert was included in energy bills, which the participating
utility company estimates goes to about 2,000,000 customers. It
generated over 1,000 telephone inquiries, which is a small
percentage, but calls are still coming in and the mailing was at
no cost to the NJDEP. An article about radon, its identification,
hazards, and control was prepared by Radon Program staff and is
scheduled to appear in a real estate magazine and also in a New
Jersey Transit publication which is available to commuters.
As important as it was and is to promote public awareness
about the hazard of radon and the importance of testing, the NJDEP
knew it could not offer every potentially affected homeowner a free
test kit. Some communities, where an initial few high readings
were found, did make radon test canisters available for free or at
greatly reduced prices. Instead, the NJDEP established a program
offering free confirmatory testing to any homeowner who requests
it because their initial test results are equal to or above 8
pCi/1. Up to and including October 1988, the confirming test was
offered if the initial result was equal to or above 4 pCi/1. This
program has now been expanded to include followup measurements on
homes which have been mitigated. The confirmatory and followup
programs were an effective means to monitor the growing industry
providing radon testing services and home mitigation services.
From October 1985 through October 1988, when confirmatory
testing was offered for a test result equal to or greater than 4
7-39
-------
pCi/1, 7,223 tests were conducted. Since the level was raised to
8 pCi/1 in November 1988, an additional 1,909 tests have been
performed, making a total of 9,132 confirmatory tests conducted
through December 1990. From October 1985 through December 1990,
2,389 followup remediation tests have been conducted.
Perhaps the most significant undertaking in the beginning of
the New Jersey Radon Program was determining the extent of the
potential radon exposure problem.
To start, the NJDEP delineated the geographic area of the
Reading Prong that ran through the State in order to make an
initial evaluation of the number of potentially affected homes.
The number exceeded 250,000. Then a review of available geologic
data for the State was conducted. It showed uranium deposits
extended beyond the Reading Prong formation. Additionally, an
examination of a New Jersey Geological Survey literature review
showed that "radioactive mineralizations" were present throughout
northern New Jersey. This meant the potential geographic area was
any part of the State north of Trenton, and that approximately 1.6
million homes were affected. Further the number of homes was
increasing in that area as more people were building in the
northwestern portion of the State during the 1980's. This initial
review of available geologic data gave New Jersey officials a sense
of the magnitude of the radon problem in the State. However
officials were aware that an extensive statewide radon study needed
to be conducted to determine where elevated radon levels were most
likely to be found and to better understand how environmental and
structural factors contribute to radon entry in homes.
Work on the legislatively mandated Statewide Study of Radon
was begun in the summer of 1986 when a contractor for the project
was selected. The study was to prepare a risk assessment of
contracting lung cancer as a result of exposure to indoor radon and
radon progeny. Almost 6,000 homes were tested in different
geologic areas of the State over the course of the study. In order
to estimate an annual exposure rate, the contractor took the
average of radon readings based on a six month heating season and
a six month non-heating season. Residency periods and smoking
history were major factors in the risk assessment. Statistics
showing risk of contracting lung cancer were compiled on both
county and selected municipal levels. The findings confirmed, and
further defined, the initial areas of concern identified by the
State.
In the autumn of 1987, using information from both the initial
NJDEP geologic data review and data already collected during the
statewide study, the voluntary certification program, and the
Cluster Study Program, the NJDEP released the first "Tier" map
entitled, "Preliminary Recommendations for Radon Testing". The map
outlined three tiers: Tier 1 was "test as soon as practical", Tier
2 was "test within one year", and Tier 3 was "test if concerned".
Municipalities were categorized as Tier 1, 2, or 3 based on the
percentage of homes measured with radon levels greater than or
7-40
-------
equal to 4 pCi/1. Data on 25 homes was required to classify a
municipality into a particular tier. If there was insufficient
data, then classification of the municipality was based on the
geological province data in which the municipality was located.
The tiers are drawn on municipal boundaries because these were
considered the smallest workable political and geographic
subdivisions on which to identify radon potential.
Both a press release and a direct mailing to every homeowner
in Tier 1 were done in conjunction with the map release. The
mailing was sent to almost a half million home and resulted in
approximately 40,000 inquiries about the radon issue and testing
recommendations.
The Tier map continues to be periodically updated based on
data submitted to the NJDEP by radon testing firms currently
participating in the "Interim Voluntary Certification Program".
Over the past four years the Tier boundaries have altered. The
reported test results have shown that although the initial
designated areas were on track further identification and
definition are possible and necessary. Recently the tiers ceased
to be defined as recommendations for testing. Instead, they are
defined as radon potential. The current criteria used to classify
municipalities into a particular Tier are outlined in Table 1.
TABLE 1
Criteria for Tier Designation
Tier Municipality* Geologic
Province**
Tier 1 - High Radon
Potential
Tier 2 - Moderate Radon
Potential
Tier 3 - Low Radon
Potential
>25% of homes
tested have
radon levels
>4.0 pCi/1
5-24% of homes
tested have
radon levels
>4.0 pCi/1
0-4% of homes
tested have
radon levels
>4.0 pCi/1
>2 5% of homes
tested have
radon levels
>4.0 pCi/1
5-24% of homes
tested have
radon levels
>4.0 pCi/1
0-4% of homes
tested have
radon levels
>4.0 pCi/1
* Criteria used if there are at least 2 5 homes that have been
tested in the municipality.
** Criteria used only when municipality data is insufficient
(less than 25 homes tested for radon) and at least 100 homes have
been tested in the province.
7-41
-------
The New Jersey Legislature had also mandated requirements for
the NJDOH. An epidemiological study of radon and lung cancer based
on actual radon measurements in homes and detailed smoking
histories for individual subjects was conducted by the NJDOH. It
was an extension of a previously conducted lung cancer study among
New Jersey women. Residence criterion was established and both
year-long alpha track detector measurements for estimating subject
exposure as well as four-day canister quick screening for current
residents were done. The entire study group, cases and controls
combined, was 835 women. Detailed smoking histories were taken for
the subjects. The findings reported by the NJDOH suggested "the
trend for increasing risk with increasing radon exposure was
statistically significant". Consequently, "the study suggests that
the findings of radon-related lung cancer in miners can be applied
to the residential setting. Excess radon exposures typical of
homes may increase risk of lung cancer; extremely high residential
exposures would be associated with very serious lung cancer risks."
The NJDOH reported that the study findings supported the State's
initiatives for technical information and services, citizen
education, and research studies, and that smoking avoidance
education for the public should also be included and emphasized in
any radon reduction program.
The NJDOH was also charged with establishing and operating a
Voluntary Radon Exposure Registry. Residents who were found to
have high indoor radon levels which they had been exposed to for
some time, could be listed on the registry. They are to receive
follow-up information about hazard reduction, risk, and medical
treatments. The registry is also a source for background
information about exposures and exposure areas.
Current Program Activities
Two major programs are currently underway which should improve
radon protection efforts in New Jersey. The first is the
legislatively required certification program for testers and
mitigators. The second is the federal State Indoor Radon Grant
program.
The New Jersey Legislature enacted a law requiring that the
NJDEP develop a mandatory certification program for all radon
testers and mitigators who want to operate in the State.
Initially, the NJDEP established a voluntary program in which
testers and mitigators voluntarily submitted proof to the NJDEP
that they met certain requirements. These companies were included
on an "Interim Voluntary Certification" list. These companies have
been the major source of information about home testing done in the
State. To date they have supplied data for more than 140,000 tests
conducted statewide.
Final regulations have been adopted, and as of May 13, 1991,
no tester or mitigator may continue to operate in New Jersey, if
he or she has not applied for and met the State's certification
7-42
-------
requirements. The certification process begins with a tester or
mitigator taking a training course that is given by the NJDEP or
that is NJDEP-approved. Then the applicant must take an
examination. There are four exams, each given for a particular
title, and they are Radon Measurement Specialist, Radon Measurement
Technician, Radon Mitigation Specialist, and Radon Mitigation
Technician. Finally, there is an application form on which the
applicant reports his or her qualifications and experience, and
this form must be submitted to and reviewed by Radon Program staff.
Applicants may choose to submit their certification forms for
review prior to taking the examination.
However, it is not sufficient to simply await data that is
supplied by testers and mitigators. There remain large portions
of the homeowning population who know about radon and its
associated risk but still do not test. And there is also a large
population group that may be unaware of radon problems although
they might very well be at risk. With funds from the United States
Environmental Protection Agency's State Indoor Radon Grant program,
the NJDEP is working to increase awareness and educate the public
about radon issues.
One project is the development of school activities to teach
children about radon and also about the concept of risk, using
radon as an example. The intent is that these children will grow
up being more aware of potential hazards in life and how to make
rational risk based decisions regarding them. It is also hoped
that the children will carry the message home to their parents.
Somehow, adults find it hard to ignore information that is
presented to them by a child who has just learned a new and
interesting lesson in school. Especially, when that lesson has
direct bearing on all their lives.
Another project that received funding is training local health
officials to evaluate elevated radon areas. This creates a
valuable working resource, lessening the burden on Radon Program
staff in conducting labor intensive radon evaluation studies.
Evaluations of elevated radon areas are needed when a home test
result is at or above 200 pCi/1 because it has been found in a
number of communities that as many as three quarters of the
surrounding homes will have readings exceeding 4pCi/l. A protocol
was developed for State employees to conduct area evaluations and
recently, with grant funding, local health officers are being
trained in the protocol. It consists of confirmatory testing,
public meetings to explain the situation and plan of action,
selection of candidate homes for radon testing based on geologic
data, house structure and a gamma survey, radon canister placement
and pickup by evaluating staff, and a public report of findings and
recommendations. In the first year of the project, 28 local health
officers were trained and others have expressed interest.
Contacting and communicating with low income residents and
residents of metropolitan areas (urban environments) about radon
presents a unique challenge. Currently, two grant projects are
7-43
-------
being funded to identify and assess the radon exposure, testing,
and remediation needs of low income and disabled persons, and also
urban populations especially focusing on multifamily dwellings.
Many of the standard means for informing and educating the public
are not applicable to these population groups. Additionally,
questions such as testing and remediation expenses and building
owner responsibility and liability must be dealt with.
A fifth grant project is to survey real estate transactions
in New Jersey. This project has four objectives: 1) to assess the
current radon knowledge and information needs of buyers, realtors,
bankers, and real estate attorneys; 2) to assess the assistance and
notification that current home buyers are receiving about radon;
3) to develop additional information pieces for all of these
groups; and 4) to develop guidelines and policies on radon testing
and real estate transactions.
Since the New Jersey Radon Program began work in the spring
of 198 5, the direction of the program has been identifying the
extent of the radon problem in the State, educating the public
about radon, and assuring that the latest and most effective means
for control and mitigation are available. The NJDEP believes that
residential exposure to radon is the most serious environmental
health threat facing New Jersey citizens today. The NJDEP has
taken steps to make each State resident aware of the hazards of
radon exposure by providing information about potential radon
occurrence in local areas via the Tier map, advertising the toll-
free Radon hotline number, and preparing and distributing
informational materials.
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.
7-44
-------
Session VIII
Oral Presentations
Radon Prevention in INIew Construction
8-1
-------
A COMPARISON OF INDOOR RADON CONCENTRATIONS BETWEEN
FRECONSTRUCTION AND POST-CONSTRUCTION MITIGATED
SINGLE FAMILY DWELLINGS
James F. Burkhart, Physics Department,
University of Colorado at Colorado Springs
Colorado Springs, CO 80933
Douglas L. Kladder, Residential Service
Network, Inc.,
525 E. Fountain
Colorado Springs, CO 80903
ABSTRACT
We have done a detailed study comparing indoor radon
concentrations among single family dwellings in Colorado Springs
that were mitigated prior to the completion of construction and
similar buildings that were mitigated after construction. There
appears to be evidence which indicates that "preconstruction"
mitigation is more effective at lowering indoor radon
concentrations than "post-construction" mitigation.
A total of 102 owners of single family dwellings, in two
different areas within the city, agreed to participate in the
study. Thirty-nine homes formed the preconstruction mitigation
category (with 14 of these homes having only passive systems), 24
had been mitigated after construction and the final 39, chosen as
a control group, had never been mitigated but shared similar soil
and surficial geological features with the mitigated homes
(including distance to nearby faults). Eighty nine homeowners
successfully completed the test. All of theses houses were
tested over the same 48-hour period, under closed-houso
conditions, thereby controlling the variables of weather and, to
some extent, occupants' usage.
By analyzing the data obtained, we can conclude that there is
a statistically significant difference in post-mitigation indoor
radon concentrations (as measured by simultaneous charcoal
screening tests) between the preconstruction and the
post-construction mitigated homes. The preconstruction category
exhibited the lower radon average, although both mitigation
categories had averages below 4.0 pCi/L. Such a conclusion could
have an impact on current mitigation practices, especially as
they pertain to new housing construction.
Esthetics, installation costs and operating costs of the two
mitigation techniques (pre and post-construction) are also
discussed herein.
8-3
-------
INTRODUCTION
The purpose of this study is to assess the relative
effectiveness of radon reduction methods in residential
structures when they are utilized after the home is constructed
as opposed to when the home is mitigated prior to the completion
of construction. It is hoped that the results discussed herein
will provide information for the building industry and those
agencies which assist it in developing approaches to mitigating
new and existing homes.
This study was conceived by the authors when it was noted that
data collected from post-mitigation testing over the last three
years were giving the indication that post-construction
mitigation provided similar results to mitigations performed
prior to the completion of construction. However, such a
conclusion was difficult to make due to varying environmental
conditions which affected test results. Consequently, this study
was designed to remove many of the typical testing variables by
testing all subject homes simultaneously and on the same floor.
As will be seen later, the hypothesis that active mitigation,
whether performed during or after construction, had essentially
the same results proved to be incorrect based upon the total data
obtained.
The study was conducted concurrently within two different
areas of Colorado Springs, Colorado, which we refer to as Area 1
and Area 2. The two study areas offer a unique opportunity for
comparison since they are both infill subdivisions where a
significant number of homes have no radon mitigation system at
all (Category 1). These unmitigated homes serve as a basis for
reference as to what a mitigated home might have been if no radon
reduction techniques had been used. Furthermore, these same
areas had a relatively large number of homes that had been
mitigated with active systems {i.e.; operating fans installed)
after construction (Category 2) and prior to the completion of
construction (Category 3). A fourth category was necessary to
distinguish between these homes mitigated during construction
using active systems and homes using only caulking, membranes or
sub-slab ventilation without fans. In this region, these latter
homes are called "radon ready" by the authors. We designated
these radon ready houses as category 4.
Homeowner participation was voluntary and solicited on a
neighborhood-wide basis through the two appropriate homeowner's
associations, therefore no preselection of mitigation techniques
occurred. However, subsequent interviews with participants
indicated that all mitigated homes with active systems
(Categories 2 and 3) employed sub-slab or sub-membrane
depressurization techniques as the primary mitigation method. No
attempt has been made to determine relative ventilation rates
within test homes.
Homes in Area 1 were all within a half mile radius while homes
8-4
-------
in Area 2 were within a one-quarter mile radius. The homes in
both areas were custom homes, ranging in size from 3,000 to 4,000
square feet of livable area. Most homes had finished walk-out
basements.
The number of homes initially participating in this study fell into
the four categories as noted in Table 1 below. The numbers in the
brackets, on this same chart, show the number of participants who
conducted the charcoal canister test correctly and who were
subsequently used as our data base.
TABLE 1. NUMBER OF HOMES PARTICIPATING IN THE STUDY
Category Area 1 Area 2 Total
1 Homes never mitigated 26 (22) 13 (13) 39 (35)
2 Homes mitigated after construction 12 (12) 12 (11) 24 (23)
3 Homes mitigated during construction 19 (15) 6(4) 25 (19)
4 Homes made "radon-ready" for future 10 ( 8) 4(4) 14 (12)
mitigation
GEOLOGY OF THE TEST AREAS
A previous study (1) had already shown correlations between
certain characteristics of the soils and geology of these two
areas and the indoor radon concentrations as measured by
screening tests. Specifically, elevated radon concentrations are
predicted for these two areas because of low shrink-swell
potential (indicating very little clays) and relatively high
permeability of the soil as determined from the Soil Conservation
Service County Soil Surveys (2). The surficial geology of both
areas is made up of rock derived from the Pikes Peak batholith
(3) which is known to contain 5.0 ppm of uranium (4). Finally,
Area 2 is known to be relatively close to a major fault system.
This fact is believed to contribute to enhanced radon transport.
A more precise breakdown of the above characteristics for each
of the two areas is as follows:
Area 1 soil has a low shrink-swell potential with a
permeability of 2 to 6 inches of water per hour. The surficial
geology is a Dawson Arkose with some Verdos alluvium (both
dervived from the Pikes Peak granite). The average distance of
these homes to a major fault is 2.8 km.
8-5
-------
Area 2 soil has a low shrink-swell potential, also, with a
permeability of 6 to 20 inches of water per hour. The surficial
geology is Rocky Flats alluvium (which is also derived from the
Pikes Peak granite). The average distance of these houses from a
major fault is .75 km.
Ignoring house construction details completely, the above
characteristics would lead one to predict elevated radon in homes
in both areas and the higher permeability and closer distance to
a fault in Area 2 would suggest even higher radon levels in those
homes. These predictions will be seen to be verified when the
actual measurements are discussed in the Statistics section,
below.
TESTING METHODOLOGY
Radon Measurements Laboratory, housed at the University of
Colorado-Colorado Springs, is a primary lab for the evaluation of
radon concentrations using the 48 hour, four-inch, open faced
charcoal canister. These canisters are of typical design with
approximately 70 grams of 8 X 16 mesh Calgon charcoal encased in
a four-inch diameter canister, one-and-five-sixteenths inches
high, covered with a 30-50 % open-mesh retainer screen. The
laboratory has analyzed over 8,000 canisters over the last three
years.
Canisters are read using a three inch by three inch Nal(Tl)
crystal housed within a commercial lead shield. A 1,024 channel
MCA is used to look at the three most intense lead-214 and one
bismuth-214 photopeak lying between 220 and 692 KeV. The minimum
detectable activity (MDA) at the 3 O level was calculated to be
0.13 pCi/1 for canisters measured 3 hours after closing and
slightly higher for the balance of the canisters.
The usual quality assurance procedures were in place during
this testing period with 100 % of the blanks being identified and
duplicates above 4.0 pCi/1 all within the 10 % precision
expected. The 2 a error was 0.17 pCi/1 at 1.0 pCi/1 and 0.4 pCi/1
at 30 pCi/1. This low error was maintained by measuring all the
canisters (after equilibrating) the same day the test concluded.
The canisters were delivered to the participants by the
authors along with a detailed instruction sheet. The instruction
sheet augmented prior phone conversations and further oral
instructions at the time the canisters were delivered. The tests
were all to begin on the morning of December 17th and conclude on
the morning of December 19th, 1990. The canisters were placed in
an open area in the basement (in most cases, the family room), 30
inches off of the floor in the center of the room. The canisters
were sealed by the homeowner and placed outside for pick-up by
8-6
-------
the authors. Non-compliance with the instructions, or failure to
perform the test, led to 13 of the original 102 participants
being dropped from the subsequent data base. This gave us an 87 %
compliance with the fairly stringent test requirements.
THE WEATHER DURING THE TESTING PERIOD
Since all of the homes were tested during the same time period
and the distance between the two test areas is only a few
kilometers, the weather was identical for all houses. It is
probably safe to assume, therefore, that pressure differentials
brought on by outside temperatures, wind, surface conditions
(i.e.; frozen soils) and atmospheric disturbances were also
similar.
Nonetheless, it is instructive to review the climatological
data for that 48 hour period because the weather conditions were
clearly such as to promote an honest screening test by
discouraging surreptitious ventilation. Table 2 below shows the
weather data from the morning of December 17th through the
morning of December 19th. Not shown on this table is the fact
that the winds were gusty for a short time on the morning of the
18th, with a peak gust of 48 mph from the northwest.
TABLE 2. CLIMATALOGICAL DATA FOR THE TEST PERIOD
Date temp (high and low) pressure winds precipitation
Dec 17 30°F 17°F 2 9.78 4- 8.2mph light snow
Dec 18 4 9°F 17°F 2 9.62 —> 10.8rnph none
Dec 19 27°F 2l°F 2 9.60 —» 8.0mph light snow
STATISTICS
This section is in two parts. First, the raw data will be presented
in histogram form for each area separately and then both areas
combined. Second, the results of the t-tests (testing the means of
two populations to see if the populations are the same or different)
will be given after each histogram.
RAW DATA IN HISTOGRAM FORM
Figure 1 below compares the indoor radon concentrations as measured
8-7
-------
during the testing period in Area 1 with the number of houses having
a particular radon concentration. The black bars refer to those
houses which were never mitigated (Category 1) and the bars with hash
marks within them refer to houses which have passive systems only
(Category 4), the so-called "radon ready" homes.
(O 1
| Area 1-Categories 1 and 4 |
i 13 Category 4 ¦ Category 1
2 4 6 9 1012141618
pCi/l
Figure 1. Radon in homes in Area 1, Categories 1 and 4
Area 1-Categories 2 and 3
category 2
category 3
%
W/
p
%
25
1.
75
2.25 2.
75
pCi/l
in
hoir.es in Area
I I
I
Figure 2. Radon in homes in Area 1, Categories 2 and 3
8-8
-------
Figure 2 above makes the same comparison between number of houses
and radon concentrations in Area 1 only using houses mitigated after
construction {Category 2) and houses mitigated during construction
(Category 3).
Comparing Category 1 and Category 4, in Area 1, and using "he null
hypothesis that the two categories represented the same population, a
t-test was performed. The t-test, with a t value of .017, te^ls us
that the two categories are indistinguishable. It would appear that
"radon ready" houses have the same radon as unmitigated houses. The
statistics are given in Table 3.
Comparing Category 2 and Category 3, in Area 1, and using the null
hypothesis that the two categories represented the same population, a
single tailed t-test, with a t value of 2.416 indicates that the two
populations are indeed different at the 95% confidence level with the
houses mitigated during construction (category 3) having the lower-
radon mean. The statistics are summarized in Table 3
Figure 3 below compares the indoor radon concentrations as measured
during the testing period in Area 2 with the number of houses having
a particular radon concentration. The black bars refer to those
houses which were never mitigated (Category 1) while the bars with
hash marks within them refer to houses which have passive systems
only (Category 4).
4
Area 2-Categories 1 and 4
31
category 1
category 4
cn
LU
2-
2 4 6 8 10 12 14 16 18 2022242628
pCi/1
Figure 2. Radon in homes in Area 2, Categories 1 and 4
8-9
-------
Figure 4 compares the indoor radon concentrations in Area 2 with
the number of homes at a particular radon concentration. Here, the
black bars refer to homes mitigated after construction (Category 2)
while the hash mark bars refer to homes mitigated during construction
(category 3).
3
Area 2-Categories 2 and 3
¦ Category 2
H Category 3
0.25 0.751.251.75 2.25 2.75 3.25 3.75 4.25
pCl/l
Figure 4. Radon in homes in Area 2, Categories 2 and 3
Comparing Category 1 and Category 4, in Area 2, and using the null
hypothesis that the two categories represented the same population, a
one-tail t-distribution, with a t value of 1.304, seems to confirm
the null hypothesis. That is, as in Area 1, "radon ready" homes have
the same average radon as do unmitigated homes. The statistics are
shown later in Table 4.
Comparing Category 2 and Category 3, in Area 2, and using the null
hypothesis that the two categories represented the same population, a
one-tail t-test, with a t value of .091, seems to confirm the null
hypothesis. That is, homes mitigated during construction have the
same average radon as do homes mitigated after construction. It
should be mentioned that the small number of homes (only 4) in
category 3 make this conclusion far from certain, although
statistically justified. The statistics are §hown later in Table 4.
Finally, the data from the two areas is combined, thereby making
any conclusions more general and, because of the larger numbers
involved, more convincing. We begin by showing a histogram of the
combined data, Categories 1 and 4 in figure 5.
8-10
-------
Combined Data-Categories 1 and 4
Category 1 (Unmitigated)
Category 4 (Radon Ready)
4 6 9 101214161820 22 2426
pCi/!
.gure 5. Radon in ail the homes combined, Categories 1 and 4
When we combine all the data from both areas, we can also compare
radon levels in homes which were mitigated during construction
(Category 3) and homes mitigated after construction (Category 2).
This comparison is given below in figure 6.
Combined Data-Categories 2 and 3
Category 2 (Existing Houses)
Category 3 (New Construction)
0.25 0.75 1.25 1.75 2.25 2.75 3.25 3.75 4.25 4.75
pCi/I
.gure 6. Radon in all the homes combined, Categories 2 and
8-1J
-------
Comparing unmitigated homes (Category 1) with "radon ready" homes
(Category 4) in the combined data, and using the null hypothesis that
the two categories really represent the same population, a single
tailed t-test with a t value of .987 seems to confirm the null
hypothesis. At this point, it seems safe to say that "radon ready"
homes are no better at reducing radon concentrations than are
unmitigated homes. The statistics are shown in Table 5.
A last comparison is now made. This is comparing houses mitigated
during construction (Category 3) with houses mitigated after
construction (Category 2) with all data combined. Again, the null
hypothesis is that the two categories will represent populations with
similar averages and standard deviations, i.e.; that it makes no
difference in indoor radon levels if a house is mitigated during or
after construction. This time, it is probably safe to reject the null
hypothesis because a single tailed t-test indicates that the two arc-
separate populations at the 98% confidence level, with a t value cf
2.059. The statistics are shown in Table 5.
To show the effectiveness of the radon prevention measures in the
three mitigation categories, a final histogram is presented. Figure 7
compares the average of each of the categories when all cf the data
is combined.
15
Combined Data Averages
0
1
2 3
Categories
4
Figure 7. Average radon in all homes combined, broken down by
category
8-12
-------
DATA REVIEW
After receiving the questionnaires and the exposed canisters, i:he
authors found several conflicting comments regarding descriptions of
the type of system installed. Consequently, a combination of
participant interviews, site visits and construction files were
reviewed to verify which category each house really belonged within.
All mitigated houses were reviewed in this manner which yielded some
additional insights for this study:
1) Several new home owners were under the impression that adequate
systems had been installed in their homes by the builders. Some of
these systems turned out to be only barrier techniques (sealing or
sub-concrete polyethylene). Perhaps more notable were homes that had
sub-slab perforate piping systems that were stubbed up in the
basement (most were sealed and one was open into the home). As an
interesting note, this survey was tne first time some of the homes
were tested after occupation. For the purpose of the study, these
homes were moved into Category 4 with Category 3 retaining only
active sub-structure depressurization systems.
2) Two homes had utilized a sub-concrete mesh system where all of
the rest of the survey utilized foundation drains or a combination of
foundation drain and interior piping approaches for negative field
propagation. These two homes were more than twice the mean of the
other existing homes. Inspection of these homes indicated that the
problem was not necessarily with the membrane, but rather with the
installation. Fans were installed inside with extensive positive
side piping. Non-standard fittings were utilized, which discharged
beneath windows and near dryer vent openings. As the purpose of ~he
study was to distinguish between during- and post-construction
techniques as they are actually being installed, these two houses
were maintained in the Area 2 data pool. The balance of the
mitigated properties were carried out by the same RCPP listed
contractor. Although it is not the purpose of this paper to
distinguish between installers, it reinforces the need for proper
training of those involved in radon mitigation.
3) Some homes which had active mitigation systems installed, after
construction, had inoperable fans. These homes were moved to
Category 4 since the authors felt that they represented a passively
vented system as in a "radon ready" approach. At this time, no
attempt has been made to distinguish between barrier versus passive
systems. As an interesting side light, one homeowner insisted that
her system was operating because it was not unplugged. She was only
convinced when she inspected the fan. This system was installed
three years ago before the present EPA mitigation guidelines
requiring certain operating indicators for the homeowners were
developed (5).
RESULTS OF THE STUDY
What follows is a discussion of each area separately, culminating
8-13
-------
in a discussion of both areas combined. However, it should be kept
in mind that because of the smaller data base of Area 2, conclusions
based upon this smaller data base may prove to be less convincing.
RESULTS FROM AREA 1
A comparison of the mean radon levels listed in Table 3 clearly
indicates that mitigation during or after construction had beneficial
effects. In fact, the means of both Categories 2 and 3 were well
below the current EPA guideline of 4.0 pCi/L. As these were
screening measurements taken at the lowest living area, current
approaches would recommend no further action by the homeowner (6).
TABLE 3. RADON LEVEL MEANS AND STANDARD DEVIATIONS FROM AREA 1
Category
Descript ion
Number
Mean
Standard
Deviation
1
Unmitigated
22
9.8
5.25
2
Post-construction mitigation
12
1. 94
1.7 2
3
During construction mitigation
15
0 .78
0 . 64
4
Radon ready
8
9.77
6.63
Homes that were mitigated during construction with active
sub-slab systems (Category 3) outperformed those active systems
that were installed after construction (Category 2). This
conclusion is based on a one-tail t-distribution at the 95%
confidence level.
Homes that were built with radon ready systems or had
passively vented systems showed statistically no benefit over
homes that had no mitigation work done.
RESULTS FROM AREA 2
As was seen in Area 1 using unmitigated houses as reference
(Category 1), mitigation which occurred during or after construction
showed significant beneficial reductions. Additionally, both the
mean of Categories 2 and 3 were well below the current screen action
level of 4.0 pCi/L (Table 4).
8-14
-------
TABLE 4. RADON LEVEL MEANS AND STANDARD DEVIATIONS FROM AREA 2
Category Description
Number Mean
Standard
Deviat io
1
2
3
4
Unmitigated
Post-construction mitigation
During construction mitigation
Radon ready
13
11
4
4
16.57 ±8.39
1.43 0.86
1.49 1.74
10.27 8.71
Homes that were mitigated during construction with active
systems (Category 3) did not show a statistical difference from
those homes that were mitigated after construction (Category 2).
This result is certainly different from that obtained in Area 1.
This may be due to the smaller sample volume and the effect of
the non-mitigation guideline homes. One might also speculate
that the higher soil porosity in Area 2 allows equal propagation
of a sub-slab negative pressure field regardless of the use of a
perimeter drain system (Category 2) or a perimeter drain system
plus a sub-slab pipe network (Category 3).
Although the mean of radon ready homes (Category 4) in Area 2
was lower than non-mitigated homes (Category 1), no statistical
difference can be demonstrated. Therefore, the conclusion for
Area 2 is the same as for Area 1 in that no reduction benefit was
seen on radon ready installations.
RESULTS FROM BOTH AREAS COMBINED
In order to better answer the question that served as the
hypothesis for this paper, both data sets were combined. This
approach can be justified due to similarity of home construction,
unmitigated levels and soil type. The only difference noted,
however, was slightly different soil porosity. The comments made
above regarding unmitigated homes (Category 1) with respect to
mitigated homes (Categories 2, 3 and 4) remain the same when the data
is combined. That is, any active mitigation system is beneficial and
no benefit was derived from radon ready homes (See Table 5 below).
TABLE 5. RADON LEVEL MEANS AND STANDARD DEVIATIONS FROM BOTH AREAS
Category Description Number Mean Standard
Deviat: i on
1
2
3
4
Unmit igated
Post-construction mitigation
During construction mitigation
Radon ready
35
23
19
12
12.32 ± 7.28
1.70 1.37
0.93 0.96
S.94 5.98
8-15
-------
When all data is combined, including the anomalies mentioned
earlier, one can determine statistically that systems installed
during construction (Category 3) outperformed systems installed
after construction (Category 2). Categories 2 and 3 are two
distinctly different populations as verified by the one-tail
t-test at the 98% confidence level.
IMPLICATIONS OF RESULTS
It is interesting to note that the existing homes that were
mitigated after construction (Category 2) had a mean screening
result of 1.70 pCi/L ± 1.40. Although this is at a level below
the current EPA action level of 4.0 pCi/L, it is right at
contemplated values for the new proposed guideline of 2.C pCi/L.
(Ref 7). Although it is reasonable to assume upper floors of
these homes would be at lower concentrations of radon, it should
be noted that due to terrain and architectural plans, many of
these lower level floors contain family rooms and bedrooms. The
adoption of 2.0 pCi/L guideline for living areas may be difficult
to consistently achieve with mitigation techniques observed in
this study.
Similarly, the homes that had active mitigation system
installed during construction exhibited a mean result of 0.93
pCi/L ± 0.96. Within one standard deviation all of these
Category 3 homes would exhibit screening levels beneath both the
existing guideline of 4.0 pCi/L and the proposed guideline of 2.0
pCi/L.
The overall mean of new homes constructed with active systems
(Category 3, mean 0.93) would lend partial credence to the
(Option 1) prescriptive approach proposed in the draft model
standards for new buildings. (Ref 8). However, the approach of
not requiring, or not emphasizing post-occupancy testing may
result in not identifying improper installations, as this study
did. This may, on the other hand, speak to proper education of
installers and the extension of the RCPP program to home builders
as well as specialty radon mitigation sub-constractors.
The inability to distinguish between "radon ready" systems
(Category 4) and non-mitigated homes reinforces the need for
testing within 30 days of occupancy for a non-activated radon
ready home. This is referred to as Option 2 of the Draft Model
Standards for New Buildings. Furthermore, the results of
Category 3 indicate the ability to reduce levels to below 2.0
pCi/L once the radon ready system is made active by addition of a
fan. It would be prudent to emphasize testing after actuation of
the system fan for the same reasons as indicated above.
Homeowners' understanding of proper system operation was
inadequate in some cases. Interviews with participants indicated
little information was passed on from previous homeowners or
building contractors. This comment is more pertinent wi~h
8-16
-------
respect to homes which were constructed with radon ready systems.
In this case, some homeowners felt that a complete system had
been installed. This car. be dealt with either in a regulatory
manner or perhaps a greater emphasis can be placed on the present
Radon Contractor's Proficiency Program and particularly the
Mitigation Guidelines (5).
The data made available from this study will, with further
evaluation, offer opportunities to assess differences between
finer points of mitigation installations. A more detailed review
of homes in Categories 2 and 3 that fell outside the standard
deviation of the mean can be made to assess these installation
differences. A comparison of individual results to soil porosity
and soil gas measurements can also be made in order to assist in
developing a predictive model, at least for this geological area.
Furthermore, a more detailed review of Category 4 homes needs
to be made to determine which radon ready approaches may offer
the most cost effective benefit.
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.
8-17
-------
REFERENCES
1. Burkhart, J.F. and Huber, T.P., Surficial materials influence
on radon production and transport: A Colorado example. In:
Proceedings of Technical Exchange Meeting on Assessing Indoor
Radon Health Risks, DOE, Grand Junction, Colorado, September,
1989, Section N.
2. U.S. Soil Conservation Service, 1981. Soil Survey of El Paso
County Area, Colorado, U.S. Department of Agriculture,
Washington, D.C.
3. Scott, Glenn R. and Wobus, Reinhard A., Reconnaissance
Geological Map of Colorado Springs and Vicinity, Colorado,
pub. by U.S. Geological Survey, Reston, VA, 1973.
4. Phair, G. and Gottfried, D., The Colorado front range,
Colorado, U.S.A., as a uranium and thorium province, in J.A.S.
Adams and W.M. Lowder, eds., The Natural Radiation
Environment. University of Chicago Press, Chicago, o. 7-38,
1964.
5. Environmental Protection Agency, Radon Contractor Proficiency
Program, Radon mitigation guidelines, October, 1989.
6. Environmental Protection Agency Citizen's Guide, August., 1986.
7. Environmental Protection Agency, Technical Support Document of
the 1990 Citizen's Guide to Radon, August 16, 1990.
8. Environmental Protection Agency, Proposed model standards and
techniques for control of radon in new buildings, November 2,
1989 .
8-18
-------
RADON REDUCTION IN NEW CONSTRUCTION: DOUBLE-BARRIER APPROACH
C. Kunz
Wadsworth Center for Laboratories and Research
New York State Department of Health
Albany, New York 12201-0509
ABSTRACT
A double-barrier design with the space between the barriers having
little resistance to gas flow is described for those parts of homes and
buildings that interface with the soil or surficial rock to reduce soil-gas
(radon) entry into structures. The outside or soil-side barrier interfaces
with the soil. A barrier placed on the soil under the subslab aggregate is
an important element in this design. This forms the outer barrier for the
floor. The subslab aggregate forms a permeable layer, while a plastic
membrane above the aggregate, the slab, and caulking form the inner barrier.
If hollow block are used, barrier coatings can be placed on both the soil
side and interior wall of the blocks, while the hollow space in the blocks
forms the permeable space. The hollow-block walls are connected to the
subslab aggregate to form a small interconnected permeable volume that can be
managed in the following ways to reduce soil-gas entry into the structure.
1. Sealed.
2. Passively vented to outdoor air.
3. Passively depressurized using an internal stack.
4. Actively depressurized.
5. Actively pressurized.
In addition to basements with hollow-block walls, the double-barrier
technique can be adapted to solid wall, crawl space and slab-on-grade
construction including various combinations.
8-19
-------
INTRODUCTION
In the long terra, substantial reduction in radon exposure can result
from improved new home and building construction techniques that reduce radon
entry. In addressing this approach to reducing radon exposure, the EPA has
published a report "Radon-Resistant Residential New Construction" (1) in
which construction techniques to minimize radon entry in new structures and
to facilitate its removal after construction are described. This report is
the first edition of technical guidance for constructing radon-resistant
structures to be issued by the EPA, and they anticipate future editions as
additional experience and approaches become available. The EPA report
includes a section on barriers to reduce radon entry including wall coatings,
sub-slab membranes, caulking, sealing and prevention of slab cracking.
Another section discusses designs for post-construction active or passive
sub-slab ventilation. A primary element in these designs is a minimum of 4
in. of aggregate under the slab. The preferred material is crushed aggregate
with a minimum of 80% of the aggregate at least 3/4 in. in diameter. This
highly permeable bed under the slab is necessary for good communication in
the event that sub-slab ventilation is needed. The aggregate is placed
directly on the soil and represents a large permeable volume into which radon
can diffuse or flow from the soil and rock under and around the foundation.
The radon that accumulates in the permeable aggregate can then flow with
little resistance to any penetrations in the barriers above the aggregate.
These barriers include the membrane placed over the aggregate, the slab and
any caulking and sealing of the wall floor joint, cracks and penetrations.
Having a permeable volume between the soil and the barriers reduces the
effectiveness of the barriers. Barriers are most effective when interfacing
with the soil. A similar situation occurs when hollow blocks are used to
construct the foundation walls. Radon that infiltrates through the outer
wall and into the hollow cavity of the block walls can then flow with little
resistance to any penetrations of the inside wall barriers. Again, barriers
to radon entry are most effective on the outside or soil-side of the wall.
An indication that aggregate under the slab increases radon entry into
structures was obtained in a survey of over 6,000 homes in New Jersey (2).
The data collected in this study show a definite relationship between age and
radon concentration. On average, houses built since World War II tend to
have higher indoor radon concentrations than houses built between 1900 and
about 1945. Initially, it was suspected that newer houses had higher indoor
radon concentrations because newer houses tend to be tighter and have lower
air exchange rates. However, closer examination of the data indicated that
the differences in radon concentrations associated with tightness did not
fully account for the decline in radon concentration with increasing age in
20th-century houses. The authors speculated that the use of sub-slab
aggregate, which increased in the post-World War II era, could also
contribute to the higher indoor radon observed in newer homes.
It is difficult to determine the effectiveness of the barriers to radon
entry suggested by the EPA, when used in the passive mode, since it is not
possible to know what the indoor radon concentrations would be for a house if
the radon-resistant techniques were not employed. The initial results,
however, have led the EPA to conclude "that in the presence of a moderate-to-
high radon source, radon prevention techniques that are passive only may not
produce indoor radon levels consistently below 4 pCi/1." In a study of 15
8-20
-------
full-basement homes in New York State which were built employing radon-
resistant techniques in an area with above-average levels of indoor radon,
most of the homes required active sub-slab ventilation systems (3). The
results from the Hew Jersey survey and the initial results of the homes built
with radon-resistant construction indicate that sub-slab aggregate
interfacing directly with the soil or rock under a home can increase radon
entry into the home and decrease the effectiveness of barriers placed above
the aggregate.
DOUBLE-BARRIER CONSTRUCTION
It is the purpose of this paper to suggest a design for new home
construction that is more effective in reducing radon entry in the passive
mode but one that can be readily adapted to active mitigation systems if
needed. The design proposes to reduce soil-gas entry by using double-barrier
construction for the sub-grade structure of homes and buildings. A primary
element in this approach is to have a radon barrier under the subslab
aggregate at the soil interface.
The double-barrier approach is illustrated in Figure 1 for a basement
with block walls and a sump. The hollow space in the block walls is
connected to the subslab aggregate via weeping holes or some other low
resistance pathway for air flow, to form an interconnected permeable space
that surrounds the entire subgrade structure. Barriers to radon transport
such as membranes, coatings, caulking, sealing, etc., are placed on both the
soil side and inside of the permeable space. Since radon barriers are most
effective at the soil interface, most of the barrier effort should be
concentrated on the sub-aggregate and outside wall barriers. The barrier
below the aggregate may be a composite of materials such as cement, tar,
plastic film, fine sand, and clay. Barriers at the soil interface should be
resistant to both diffusive and convective flow. A special effort should be
made to seal the outside wall barrier at the wall-footing joint and the
barrier below the aggregate at the footing-aggregate and aggregate-sump
joints.
The double-barrier subgrade construction creates a reasonably small
volume between the inside and outside barriers that can be managed in several
ways to reduce radon entry. Without a barrier below the aggregate, the soil
and rock under and around the house will be directly connected to any
mitigation system used to reduce radon entry. The double-barrier approach
works toward decoupling this direct connection. For the double-barrier
system shown in Figure 1, passively venting the hollow-block walls to outdoor
air will allow outdoor air to flow with little resistance into the permeable
space. As gas from the permeable space is drawn through any penetrations in
the interior or upper barriers into the basement by indoor-outdoor pressure
differentials, outdoor air can flow into the permeable space with little
resistance. The outside air flow reduces the draw on soil-gas at any
penetration in the outer or below barriers and thereby reduces the flow of
soil-gas radon into the permeable space. Alternatively, the permeable space
could be treated by depressurization (passive or active) or pressurization
(active). For these approaches it would be best to not vent the block walls
to outside air. Radon entry reduction can then be accomplished by creating
either a reduced pressure or increased pressure in the permeable space.
Having created a reasonably small interconnected permeable space with sealing
8-21
-------
VENT TO
OUTDOOR AIR
OUTSIDE WALL
BARRIER AT —
SOIL INTERFACE
CHANNELS
IN FOOTING
SOLID ROCK
INSIDE WALL
BARRIER
DEPRESSURIZATION
OR
PRESSURIZATION
-SUMP
DISCHARGE
SUBSLAB
AGGREGATE
SEALS
SUMP
DRAIN
TILE
BELOW SUBSLAB AGGREGATE
BARRIER AT SOIL INTERFACE
Figure 1. Double-barrier construction for a basement with sump.
8-22
-------
on both the soil side and inside, it is expected that, if passive venting
(using a stack through the house interior), active suction, or positive
pressure flow is necessary to reduce indoor radon to acceptable
concentrations, then relatively low flow rates would be successful.
An example for an active pressurization system would be to draw air from
ceiling vents in the highest level of the house and blow this air into the
permeable space between the double barriers (Figure 2). The fan could be
located in the basement and relatively low flow rates ("20 cfm) should
suffice. In this manner, heated air from the highest interior level of the
house would be used to pressurize the double-barrier system heating the floor
and walls of the basement while reducing heat loss via exfiltration from the
higher levels of the house.
It is of primary importance to ensure that water effectively drains from
the permeable substructure space between the double barriers, This can be
accomplished with a sump as shown in Figure 1. It may be necessary to grade
the soil forming the base of the subslab aggregate toward the drain tiles and
the sump to aid in preventing the accumulation of water in the subslab
aggregate. If it is possible to drain the subslab aggregate to grade or to a
sewer, then this drainage option could be used instead of or with a sump.
Solid pipe should be used and it should be sealed at the outside or soil-side
barrier.
Exterior footing drainage of gravel and/or perforated piping is used by
many builders and presents a problem to the double-barrier design approach.
The gravel and/or perforated piping of the exterior drainage system runs
around the outside perimeter of the wall-footing joint. It represents a
permeable volume in which radon can accumulate and flow to any penetrations
in the wall and wall-footing joint. To minimize radon entry, the exterior
drainage system should be drained to daylight or to a sewer and not connected
to the subslab aggregate and sump via weeping holes or other methods.
Connecting the exterior drainage system to the subslab aggregate would
provide a pathway for soil-gas radon to enter the permeable zone of the
double-barrier system. Exterior perimeter drainage systems increase the need
for careful sealing at the exterior wall-footing joint.
The double-barrier approach is illustrated for slab-on-grade and crawl
space construction in Figure 3. Drainage of water that might accumulate in
the sub-slab-on-grade aggregate can be accomplished using a sealed sump as
shown in Figure 1 or by drainage to grade or a sewer using solid pipe. If
the double-barrier system is not effective in the passive mode (sealed,
vented to outdoor air, or passively depressurized using stack ventilation),
then active pressurization or depressurization can be employed. When a
barrier is placed directly on the soil of a crawl space and the floor of the
house is sealed, one obtains a double-barrier system with the space between
the soil barrier and the floor being the permeable space. The crawl space
can then be vented to outdoor air or the crawl space can be sealed and
passively depressurized, or actively pressurized or depressurized. To reduce
the volume of air to be pressurized or depressurized, a permeable layer of
aggregate or other construction to form a permeable space with barriers on
both the soil side and house side can be used as shown in Figure 3. Sealing
the floor and using a double barrier at the soil surface results in a triple-
barrier system where the two permeable spaces could be treated independently.
8-23
-------
SUMP
BARRIER AT
SOIL INTERFACE
Double-barrier pressurization -using interior air.
3-24
-------
VENT TO
OUTDOOR AIR'
DEPRESSURIZATION
OR
PRESSURIZATION
fTTZTl
tm
III#
BARRIER AT SOIL
INTERFACE
A fed Ł.
4 4 A UII4 4i )
SLAB-ON-GRADE
VENT TO
OUTSIDE AIR
DEPRESSURIZATION
OR
PRESSURIZATION
\
BARRIER AT SOIL
-1 INTERFACE
CRAWL SPACE
Figure 3. Double-barrier systems for slab-on-grade and crawl space
construction.
8-25
-------
For example, the aggregate could be passively depressurized and the space
below the floor could be vented to outdoor air.
SUMMARY
Radon-resistant construction designed to decouple houses from the soil
has been suggested and used in various forms. The EPA refers to constructing
a pressure break between the foundation and the soil. Brennan and Osborne
(A) suggested that a drainage mat be used to form an air curtain around the
foundation. A Denver builder excavates to a depth of 10 ft. and constructs a
crawl space under a wood basement floor (1). The crawl space is then
actively ventilated. Walkinshaw (5) constructs a shell inside the basement
and then ventilates the space between the interior shell and the basement
floor and walls.
The double-barrier approach described in this paper attempts to modify
normal building practices to be more radon-resistant at moderate cost.
Barriers under the aggregate and on the outside of hollow-block walls
interfacing with the soil and rock will be the most effective barriers in
reducing radon entry. The double-barrier construction creates a relatively
small permeable volume between the inside and outside barriers that can be
managed in several ways, either passively or actively, to reduce radon entry.
A key element in this design is to maintain water drainage from the permeable
space between the barriers and from around the foundation. There are many
types and variations of house and foundation construction. Very often these
variations are dictated by the local and regional surficial geology. It is
not possible to describe a radon-resistant design readily applicable to all
types of construction and water drainage conditions. However, a better
understanding of how water drainage systems around foundations can increase
the potential for radon entry will enable builders to make water drainage and
radon-resistant construction more compatible. Double-barrier construction is
such an attempt to make water drainage and radon resistance work together.
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.
8-26
-------
REFERENCES
1. Osborne, M.C., Radon-Resistant Residential New Construction, United
States Environmental Protection Agency, EPA/600/8-88/087, July 1988.
2. Task 3 Final Report, Sampling Design, Data Collection and Analysis,
Statewide Scientific Study of Radon, Prepared for the New Jersey
Department of Environmental Protection, Prepared by Camp Dresser and
McKee Inc., April 1989.
3. Demonstration of Mitigative Techniques in Existing Houses and New House
Construction Techniques to Reduce Indoor Radon, Prepared for New York
State Energy Research and Development Authority and the U.S.
Environmental Protection Agency, Prepared by W.S. Fleming and
Associates, Inc. (NYSERDA report in publication).
4. Brennan, T. and Osborne, H.C., Overview of Radon-Resistant New
Construction, In Proceedings of the EPA 1988 Symposium on Radon and
Radon Reduction Technology, Denver, CO, Oct. 1988.
5. Walkinshaw, D.S., The Enclosure Conditioned Housing (ECHO) System: A New
Approach to Basement Design, In Proceedings of The Fifth International
Conference on Indoor Air Quality and Climate, Toronto, Canada, Aug.
1990, Vol 3, p. 257.
8-27
-------
RADON CONTROL - TOWARDS A SYSTEMS APPROACH
Nuess, R. M. and Prill R. J.
Washington State Energy Office
Energy Extension Service
N 1212 Washington, Suite 106
Spokane WA 99201
The Work described in this paper was not funded by the U.S. Environmental Protection Agency and there fore the contents do
not necessarily reflect the views of the Agency and no official endorsement should be inferred.
ABSTRACT
The normal operation of a continuous mechanical ventilation system, incorporated
into a relatively airtight house, and designed to control pressure-differences, has
been demonstrated to provide sufficient control of radon entry in a two story
residential building.
This was accomplished via a "two-cell barrier-enhanced pressure-difference control
system."
Ventilation rates, energy usage, moisture levels, pressure-differences, and radon
concentrations were monitored. Changes in radon concentrations in several
building locations, as a function of distinct pressure-difference configurations, have
been measured.
Indications are that this design offers the new residential construction industry an
opportunity to realize affordable control of radon entry, while simultaneously
optimizing potentials for moisture control, energy efficiency, and control of other
indoor air pollutants.
INTRODUCTION
This project explores the use of an airtight building envelope (and separately
isolated airtight crawlspace) integrated with a continuously operating mechanical
ventilation system, to enhance pressure-difference control strategies for
minimizing soil-air entry into the indoor air.
8-29
-------
This approach seeks to obtain robust control of radon entry, while concurrently
optimizing potentials for several building design goals including: moisture control,
energy efficiency, control of other indoor air pollutants.
CONTEXTING ASSUMPTIONS
There are several primary design goals for the environment control system
"house:" safety, comfort, durability, healthy indoor air, and energy efficiency. These
goals are not only increasingly achievable, but can be mutually advantaged in a
manner that can reduce net system cost.
A systems approach that seeks to optimize a building's performance with regard to
several desirable performance qualities might well include several very successful
radon solutions.
The radon source of concern is soil-air. The primary goal with regard to control of
indoor radon is the prevention of soil-air entry into the indoor air. Radon is a
given component of soil-air though its concentration both varies from one site to
another and is not readily predictable. In one case, measurements of soil radon
within a distance of 9 meters varied by a factor of 250 (1). Hence, the degree of
soil-air entry control required is neither constant nor predictable.
Two conditions are necessary for soil-air entry:
• There must be openings in the building envelope that couple the soil-air to the
indoor air.
• There must be a driving force, a pressure-difference that results in a flow from
the soil-air zone into the indoor air zone.
While significant reduction of all coupling pathways from the soil zone is
reasonably achievable, elimination of them is not. It has been observed that even
very small openings are sufficient to allow unacceptable radon levels (2). Control
of pressure-differences may be the practical key to adequately limiting the entry of
soil-air pollutants, including radon. Envelope tightness may be most important for
its role in enabling and enhancing pressure-difference control.
The tight building envelope, coupled with a properly designed mechanical
ventilation system, can play a central role in a systems approach that incorporates
pressure-difference control to limit soil gas entry. The tighter the air barriers of the
system, the more effective the pressure-difference control for a given amount of fan
power. This dovetails nicely with the desirable advantages of a tight building
envelope for several other building performance purposes, including:
• Comfort - fewer drafts; minimal temperature stratification; reduced noise, dust,
pollen, insects.
• Energy efficiency - large net reduction in heating/cooling loads.
8-30
-------
• Moisture control - structural durability and reduced maintenance costs.
• Enhanced dilution/removal of pollutants generated indoors - via improved
ventilation effectiveness, control and capability.
Several of the design elements serve to advantage multiple design goals. This
should be recognized when attempting to allocate costs. For example, soil-air may
also contain other pollutants of concern, such as garbage gasses (methane),
herbicides, fungicides, pesticides, spores of soil fungi, etc. (3). The cost of preventing
soil-air entry should be life-cycled against the delivery of several health benefits.
Also, in this particular project, the cost of mechanical ventilation and the tight
envelope must be apportioned to comfort, energy performance, radon control,
moisture control, and control of other pollutants.
SPECIFIC HYPOTHESIS
The normal operation of a commercially available continuous mechanical
ventilation system incorporated into a tight house, and designed to control
pressure-differences, can provide sufficient day-to-day control of those
pressure-differences (induced by weather, internal household activities, and
mechanical systems) to prevent entry of radon and other soil-air pollutants. This
can be reasonably accomplished by developing a "two-cell, barrier-enhanced
pressure-difference control system." (4).
BUILDING DESCRIPTION
In 1988, a tightly sealed and energy efficient two-story residential building was
constructed with the intent to exceed any energy performance standards currently in
place in the U.S. The building was among those instrumented and continuously
monitored for one year as part of the Residential Construction Demonstration
Program (RCDP), a multipurpose research and development effort of the
Bonneville Power Administration and the Washington State Energy Office. As an
RCDP Cycle II Future House, the expected energy performance of the building was
designed to exceed that required by the Northwest Power Planning Council's Model
Conservation Standards by 30%.
The building was constructed in Spokane, WA. Spokane has a winter outdoor
design temperature of 4°F (-15°C), 6882 normal heating degree days, and 411 normal
cooling degree days. Spokane weather has the characteristics of a mild arid climate
in the summer and a cold coastal climate in winter. Winter solar potentials are
limited by both the climate and the site. The building was calculated to have an
annual need of 2.5 kWh/ft2 (97 MJ/m2) for space heating.
8-31
-------
25001
2000'
16 200 megajouies per year 97 MJ/(m • y r)
kWh 1500-
month
1000-
500-
total thermal load
useful solar + appliance gains ~
useful appliance gains
9000
7200
AS,.SVMWWM,i%W.VM,.'iVMViV.,iV.V,V.V.'j,i':'lll"
Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec
5400 MJ
month
3600
1800
Figure 1. Predicted space heating profile.
The building is of double-wall construction and the wall thermal resistance is
approximately R45 (.126 W/[m2*K] ). This insulation level extends from the ceiling
to the concrete footing, except as interrupted by doors and glazing (Figure 2). The
glazing area is 328 ft2 (30.5 m2) and is 18% of the conditioned floor area 1780 ft2 (166
m2). Fifty-five percent of the glazing faces south. The glazing thermal resistance is
approximately R4 (1.5 W/[m2*K] ). The ceiling is insulated to R60
(0.095 W/[m2*K] ). The continuous thermal envelope is completed by R25
( 0.227 W[(m2*K]) fiberglass batt insulation laid directly upon the ground (over a
gravel capillary break).
A continuous air barrier was established with the interior drywall by gasketing the
drywall to the wood framing and sealing any penetrations through the drywall.
Upon completion of construction, the building had a tested air leakage rate of 1.2 air
changes per hour (ACH) at an induced indoor/outdoor pressure-difference of 50
pascals. One year later it was tested at 1.4 ACH at 50 pascals. The measured Pacific
Northwest average is 9.3 ACH at 50 pascals (5). The vapor retarder was established
on the interior surface of the drywall with a rated paint. The glue in the laminated
subflooring provided the floor vapor retarder.
The building is divided into two distinct cells, that are atmospherically decoupled
from both each other and the outdoor air (Figure 2). The tightness and isolation of
these two "cells" enables pressure-difference control with the mechanical
ventilation system (and prevents contamination of air in cell 1 by air in cell 2. Cell 1
contains all occupied space, so that the breathable indoor air is contained in cell 1.
The volume of cell 1 is 16,500 ft^ (467 m^). Cell 2 is a plenum by which stale air
from cell 1 is removed. Though atmospherically decoupled, it is thermally coupled
to cell 1, so it provides warm floors. Cell 2 adds another 3000 ft^ (85 m^) to the
conditioned volume.
8-32
-------
The first floor subfloor was the selected air barrier between cell 1 and cell 2. All
joints in the tongue and groove exterior grade plywood were sealed with urethane
sealant during installation. Special care was taken to identify and seal any holes
created in this barrier by the construction process (eg; temporary nailing for wall
bracing, sawhorses, measuring and cutting tables). A tracer gas was injected into cell
2 prior to carpet installation and two small air leaks were located using a detection
instrument.
HVAC SYSTEM
A small (5000 to 7000 btuh) commercially available integrated residential heat
recovery ventilation system (HPV) provides continuous ventilation, partial space
heating, space cooling, water heating, as well as the desired pressure-differences (6).
The unit consists of a water heating tank and a space conditioning module (SCM).
The SCM contains 2 constant speed fans, 2 coils, and utilizes a reversible
refrigeration cycle to provide heating or cooling via the same ductwork. During the
winter heating cycle, heat is extracted from stale exhaust air and delivered to either
the domentic hot water tank or the mixed air supply. In summer, heat is extracted
from the mixed air supply and either exhausted outside or used to heat domestic
water.
recirc
Figure 2. Exhaust Air Side. Figure 3. Supply Air Side.
8-33
-------
On the supply air side, two 60 ft (18 m) long by 4 in (10 cm) diameter PVC earth tubes
provide filtered outdoor air which is mixed with recirculating air before it passes
over the supply-side coil and is distributed to individual living areas by fan F3
(Figure 3). The earth tubes are buried approximately 4 ft (1.2 m) below grade and
serve to temper both winter and summer air.
On the exhaust air side, stale indoor air is removed from kitchen and bathrooms by
fan F1 and ducted to cell 2. From this point it remains isolated from cell 1. The stale
air travels across cell 2, then exits via a sealed duct which leads to the SCM.
Continuous operation of fan F2 (SCM exhaust fan) is necessary to maintain a lower
pressure inside the SCM than in the mechanical room, so that no leakage back into
the indoor air occurs. After passing through the SCM the air is exhausted above the
roof line.
TWO-CELL, BARRIER ENHANCED PRESSURE-DIFFERENCE CONTROL
Cells 1 and 2 are isolated from each other, from the outdoor air, and from the indoor
air; by accessible and maintainable air barriers. Sealed ducts allow controlled air
passage. Continuous mechanical ventilation removes stale air from cell 1 and
delivers it outside via cell 2. Depending on which fans are selected to operate, cell 2
can be either pressurized or depressurized relative to cell 1 and/or the soil-air. This
project incorporated four fans in the ventilation system in order to enable
comparison between these two approaches, as well as other possible ventilation and
pressure-difference configurations:
• Continuously pressurize and flush cell 2: This is the baseline operating
condition. Fan Fl removes stale indoor air from cell 1, depressurizing cell 1
relative to outside and pressurizing cell 2. Fans F2 and F3 are part of the
commercial unit and operate at a constant speed. Fan F1 must produce a greater
flow than fan F2 in order to maintain cell 2 at a greater pressure than cell 1. A solid
state speed control allows adjustment of fan Fl. Dampers allow adjustment of the
flows through the SCM, but adjustment is limited to the range of flows required by
the SCM.
• Continuously depressurize and flush cell 2: Fan Fl does not operate, so fan F2
depressurizes both cell 1 and cell 2. Cell 2 is at a lower pressure than cell 1.
• Continuously pressurize cell 2: Fan Fl operates but no flushing of cell 2 occurs.
Cell 2 is decoupled from the exhaust loop, and stale air is removed directly from cell
1.
• Mimic typical housing leakage and ventilate: Increase the envelope equivalent
leakage area of cell 1 to typical levels by introducing deliberate openings in floor and
ceiling (The northwest average is 125 in^ (806 cm~). The measured cell 1 leakage
8-34
-------
area is 16-20 in2 (100-130 cm2). The ventilation system operates (decoupled from
cell 2). Leakage distribution can also be adjusted. Typical outside vents can be
installed in the erawlspaee,
• Mimic typical housing leakage and do not ventilate: Increase the envelope
equivalent leakage area of cell 1 to typical levels by introducing deliberate openings
in floor and ceiling. Typical outside vents can be installed in the erawlspaee.
ENERGY PERFORMANCE
Limited data is available at this time and results should be considered preliminary.
The building was completed and occupied during January of 1989. Shortly thereafter
extensive energy performance monitoring was begun by the Washington State
Energy Office, via a subcontract with W.S. Fleming Inc. Selected air and water
temperatures, air and water flows, relative humidities, and electrical energy usage
have been monitored and recorded (six second averages) on a multi-channel
datalogger. Data collection for the first year has been completed. Once the data are
analysed a more complete energy performance profile will become available.
Zoned electric resistance heaters were separately submetered. The integrated heat
recovery ventilation system, which provides continuous ventilation, partial space
heating, space cooling and water heating was also submetered. Electrical main and
submeter data were recorded by the author. For the one year period between March
4,1989 and March 3, 1990, electric resistance heating used 1.4 kWh/ft2 (54 MJ/m2).
The HPV unit used 3.5 kWh/ft2 (135 Mj, ..t2) for continuous ventilation, space
heating and cooling, water heating, and pressure-difference control. The HPV
system is estimated to provide 44% of the space heating load. This brings the total
cost of space heating to $216/year. The measured average Kwh consumption for
heating conventional electrically heated homes in the Pacific Northwest is 12,420
Kwh, which amounts to $596 (5).
MOISTURE PERFORMANCE
Humidity sensors (7) were calibrated and placed inside structural wood framing in
six locations prior to completion of construction. Two sensors were placed in the
attic, two in the walls, and two in the floor of cell 1. An attempt was made to select
locations with the greatest moisture potential; generally downwind from the
prevailing wind direction, shaded areas on the north side, and (for walls) high in
the building:
• Attic top chord - north side near center of building.
• Attic bottom chord - north side near center of building.
• East wall exterior framing stud - north side on upper level.
8-35
-------
30
Cell 2 Rim
February 1989 - March 1990
Attic Top Chord
N Wall E
N Wall W
Cell 2 Interior
Attic Lower Chord —5
0
FFFMJ J AONNNDDDJ JF F M
Month
Figure 4. Percent wood moisture content in six locations.
• West wall exterior framing stud - north side on upper level and near electrical
outlet.
• Warm joist in cell 2 - center of building.
• Cold joist cell 2 - next to north rim joist and on cold side of air-vapor barrier and
insulation.
Thirty-seven intermittent readings were recorded (approximately weekly during the
heating season) and corrected for temperature. Monitored moisture levels in all
locations dropped by the end of the summer following completion of construction,
and remained approximately constant through the following winter (Figure 4).
Moisture levels remained constant during the second winter as well.
The clock timer on the HPV unit is set to provide continuous exhaust ventilation,
so the unit's exhaust fan (F2) drawing stale air from cell 2 is always activated. If
there is a demand for water heating the compressor also operates. If there is a
demand for space heat or cooling the supply fan (F3) also activates. Both fans
operate at a constant speed and flows must be adjusted by dampers.
The baseline mode of operation has been to adjust fan F1 to maintain a slightly
lower pressure at the ceiling of cell 1 than that outside (thus also pressurizing cell 2).
This typically resulted in a 2-13 Pa lower pressure at the ceiling of cell 1 relative to
outdoors during space heating. The neutral pressure plane was maintained above
the ceiling of cell 1, there was no exfiltration, and therefore all air exchange was
induced by the HPV unit. The resultant pressure in cell 2 was generally 3 to 7
pascals greater than the pressure in cell 1, during space heating mode of operation.
The supply air fan (F3) operates during space heating and cooling, and tends to
pressuize the building, by increasing the flow of outside air through the earth tubes.
VENTILATION PERFORMANCE DYNAMICS
8-36
-------
However, when it is off (during periods of ventilation and water heating), fans F1
and F2 remain on, so the cell 1/cell 2 pressure-difference increases (10 to 20 Pa).
Additionally, a manual timer switch in the bathrooms enables short pulses of
greatly increased ventilation by boosting fan F1 to full power, and the cell 1/cell 2
pressure-differences become even larger (45 to 60 Pa).
Intermittent measurements by the author indicate that the mechanically induced air
exchange rate for the first year has been roughly .6 ACH, or equivalent outdoor air
supply for 11 persons at 15 cfm (7 L/s) per person. Since the pressure in cell 1 was
lower than the pressure outside (therefore no exfiltration), all the air leaving cell 1
had to pass through fan Fl. A Kurz Model 435 Linear Air Velocity Transducer was
used to measure the mass flow of air in the duct downstream of fan Fl.
The purpose of fan F4 is to pull outdoor air through the earth tubes and provide
adequate outdoor air supply. It was found to be unecessary and was not operated.
The negative pressure of cell 1 induced sufficient flow in the earth tubes. When the
unit's supply fan (F3) did not operate (ventilation and water heating modes) the
earth tube flow averaged 27 cfm (13 L/s). When the supply fan operated the average
earth tube flow was 57 cfm (27 L/s). Approximately one third of the outside air
supply was via the earth tube. Envelope infiltration, due to the induced negative
pressure of cell 1, provided the remaining outside air.
The pressure-difference control under these conditions appears to have been very
robust. Though pressure-differences were not continuously monitored, they were
frequently observed during cold and windy periods. No reversals of the desired
pressure-difference directions were observed.
CONTROL OF RADON
RADON PHASE ONE
Five continuous radon monitors (CRMs) were placed in the same location for five
days to establish a comparison baseline. The monitors were then placed in five
different locations in the building for a thirteen day period between November 11,
1989 and November 23,1989. Fan Fl was off during the first part of of this period, so
that fan F2 depressurized both cell 1 and cell 2 relative to outside.
After the first 112 hours, Fan Fl was activated and adjusted to maintain a slightly
lower pressure at the ceiling of cell 1 relative to outside during the space heating
mode. This resulted in a 10 to 60 Pa greater pressure in cell 2 (depending on the
HPV system's operating mode at time of read).
Cumulative CRM data were recorded intermittently and averaged over each elapsed
time period. Thirty-two readings were recorded. Average radon levels in cell 2
decreased dramatically when fan F2 was activated; average radon levels in cell 1 also
8-37
-------
showed a tendency to decrease (Figure 5).
Pressure-Difference Between Cells 1 and 2: jplj;!
AP: cell 2 < cell 1
AP: cell 2 > cell 1
Radon Levels In Cell 2: Y//j
Radon Levels in Cell 1 (4 locations):
- 925
-¦ 740
- 555
- 370
A' . house < outsidQ
Cumulative Hours - October 1989
Figure 5. Radon levels in five locations.
RADON PHASE TWO
Two CRMs were placed in separate locations within cell 1 (CRMs la and lb) and two
CRMs were placed in separate locations within cell 2 (CRMs 2a and 2b). Hourly
radon averages were recorded for the three month period between 17 February and
21 May 1990. During this period, the baseline system configuration (continuously
pressurize and flush cell 2 while depressurizing cell 1) was held constant for the
first 23 days. Then 9 deliberate alterations in system configuration were made. After
each alteration the system was returned to the baseline configuration, before the
next alteration was initiated. Upon completion of the 9 alterations the system was
returned to the baseline configuration for 22 days.
The alterations had powerful effects upon Cell 2 radon levels, and clear effects upon
radon levels in Cell 1. Effects of wind, rainfall, and temperature did not appear to
have noticeable influence on radon levels, except that Cell 2 radon levels may have
showed some indication of response to temperature. Nonetheless the effect was
subtle relative to the effects of system operation.
The baseline system configuration and the alterations to it are discussed below, and
referenced in the following graphs.
8-38
-------
140
120
100-
o
Q.
80-
60-
Radon Responses to System Alterations
February 17 to March 20,1990
Beginning Baseline
Alteration 1
•d
^
02/17/90 02/27/30 03/09/90 03/19/90
02/22/90 03,04/90 03/14/90
Hourly Data
Figure 6. Baseline and Alteration 1.
Baseline. The baseline operating condition was established for 23 days, from 2/17 to
3/11. During this period the average radon concentration in cell 1 was on the order
of 1 pCi/1. The average radon concentration in cell 2 was 7 pCi/1. Intermittently
recorded cell 2 pressures ranged from 1 to 7 pascals greater than those of cell 1. The
average was 3 pascals. Cell 1 was 7 to 15 pascals lower in pressure than outside. The
average was 9 pascals.
Alteration 1. Fan F1 was turned off on 3/11 at 10:20 am. The pressure in cell 2 had
been greater than the pressure in cell 1, but now shifted to about 30 pascals lower
than cell 1, since F2 now pulled air through the cell 2 plenum. In two hours radon
levels in cell 2 had increased by a factor of three. Radon levels in cell 2 averaged 29
pCi/1. Radon levels in cell 1 increased also to an average of about 2.5 pCi/1. On
3/17, six days later, fan F1 was turned back on. Pressure in cell 2 returned to 3-4
pascals greater than cell 1. Radon levels in cell 2 dropped by a factor of three in
three hours and returned to baseline levels.
8-39
-------
140-
Radon Responses to System Alterations
March 21 to April 23,1990
120
100-
Alteration 3
Alteration 4.
Alteration 2
-^l! iU-
80-
o
Q.
cell 2
60-
•~r
cell 2
cell 2
cell 1
vT * vv
ohr~"
03/21/90
mi
03/26/90 04/05/90 04/15/90
Hourly Data
Figure 7. Alterations 2,3, and 4.
Alteration 2. On 3/29 fan F1 was turned off again for approximately 34 hours. F2
continued to operate, but the duct connection from F2 to cell 2 was disconnected
and exhaust air taken from cell 1 instead. The ductwork joining the two cells
remained open. Cell 2 was atmospherically coupled to cell 1, but was decoupled
from the ventilation loop. During this condition the cell 1 pressure was 10 pascals
lower than the outdoor pressure and cell 2 was about 7 pascal lower than than the
outdoor pressure. Hence, both cells sucked on the ground, but only cell 1 recieved
ventilation. Ceil 1 radon levels increased by a factor of 12 to an average peak of 12
pCi/1. Cell 2 radon levels increased by roughly a factor of 20 to an average peak of
120 pCi/1. When the system configuration was returned to the baseline, radon
levels quickly returned to baseline levels.
Alteration 3. On 4/4 fan F1 was turned off for 62 hours. Also F2 and F3 were turned
off. There was no ventilation. The ductwork joining the two cells was sealed to
atmospherically decouple the cells from each other. Radon levels in cell 2 rose
gradually (whereas in the previous alterations they had risen abruptly) to a peak of
8-40
-------
66 pCi/1. Then Fl, F2, and F3 were reactivated and radon levels in both cells
returned abruptly to baseline levels. The pattern of a more gradual rise in radon
levels also seemed to occur in the cell 1 radon levels which rose to over 5 pCi/1. The
gradual rise is assumed to be attributed to soil recharging and the slower response
related to the lesser stack pressures.
The three fans were activated 62 hours after the initial alteration. However, cell 2
remained atmospherically decoupled from the ventilation cycle (F2 drew exhaust air
from cell 1), as well as isolated from cell 1 by the sealed ductwork. The condition
was that cell 2 was pressurized by fan F1 but there was no flushing of the air in cell 2.
The resultant pressure in cell 2 was 45 pascals greater than that in cell 1. Radon
levels returned to baseline in about 6 hours, hence were already at baseline levels
when the ductwork was reconnected and the system configuration returned to
baseline.
140
Radon Responses to System Alterations
February 17 to March 20,1990
120-
Ending Baseline
..Alteration 4
80-
Q.
60-
40-
ceil 2
20-
/V v
04/24/90
04/29/90
Hourly Data
Figure 8. Alteration 4 and Baseline.
Alteration 4.. Fan F1 was turned off. The system was altered as described in
alteration #2 for roughly 24 to 30 hours, then returned to the baseline configuration.
This process was repeated six times. In each case cell 2 radon levels responded as
they had in alteration #2.
8-41
-------
Baseline. The baseline operating condition was restablished for 22 days, from 4/29
to 5/21. During this period the average radon concentration in cell 1 was less than
one pCi/1. The average radon concentration in cell 2 was about 2.5 pCi/1.
There is some uncertainty associated with the radon measurements. The CRMs
were research instruments that were not calibrated immediately prior to these
measurements. However, they were compared to each other by operating them in
the same location for six days at the beginning of this project. All 4 CRMs tracked
radon levels consistently (Figure 9).
a
a
s
s
9
9
to
n
16 to 22 September 1999
Figure 9. Radon instrument comparisons.
Several months after the project they were again compared to each other, and it was
discovered that only CRM la had a correctly operating air pump. This CRM was
then compared for about 30 hours to a Pylon AB5 with a PRD, which had been
recently calibrated with a Pylon Calibration Standard. The two monitors tracked
radon fluctuations consistently. When the data from this project was later
reviewed*, it was discovered that CRMs lb and 2b had diverged widely (radon
levels declined) from their respective matched pairs and never recovered. These
were suspected to be the points of pump failure. Data from these units was
removed from the data set for the time periods after their responses suggested pump
failures.
CRMs la and 2a did respond in a consistent manner throughout the entire
measurement period. Fortunately one was located in cell 1 and the other in cell 2. It
was also fortunate that CRM la ~ which was recording the lowest and least variable
radon levels ~ was also the CRM that continued to have a correctly operating
pump and was compared to the calibrated Pylon instrument after the study. Both
the consistent response of these two remaining CRMs to the repetitive nature of
alteration #4, and the similarity of their ending baseline responses to their starting
baseline responses, suggest that these two CRMs responded with acceptable
accuracy to radon fluctuations throughout the study.
* This study was not funded and was conducted as time allowed.
8-42
-------
It is not known when the the pump on CRM 2a began to fail. The data set suggests
that CRM 2a was operating correctly during the study period and likely failed after
the study period.
COST
It is very difficult to assign costs to the radon prevention and mitigation features of
this building, since virtually all the features that control radon also enhance energy
performance, durability, comfort, and the control of other pollutants. The simple
energy-only payback for these features is 10 to 20 years. The building's useful life has
also been extended due to such features as the vented rain screen designed to extend
siding life, and the elimination of air transported moisture into the exterior walls.
Since these features primarily address energy, and there is a clear energy payback for
them, it can be argued that there is no incremental cost for the control of radon
entry. The building cost $80,824, approximately $45 to $48 per square foot.
INDICATIONS
1. Control of soil-air entry with pressure-differences using a continuous mechanical
ventilation system incorporated into a tight house is readily achievable.
2. The initial radon experiment indicates that the radon source at this building
location may be sufficient to allow the demonstration of radon control, as well as
the comparison and evaluation of the impact of different pressure-difference
configurations on radon entry.
3. Airflows through the HPVAC-80 can be reduced by installation and adjustment of
dampers so that the flows necessary to maintain required pressure-differences can be
reduced (within the limits of the range of flows required by the HPV). The target
goal of maintaining soil gas entry control with a mechanical system operating at .35
ACH may be achievable at this level of envelope tightness.
4. Careful attention to air-vapor barrier installation can enable sufficient control of
moisture levels in the Spokane climate, even under conditions of constant and
relatively large pressurization.
FUTURE DIRECTIONS
• Evaluate this concept at sites where the known soil radon source is high.
• Evaluate the degree of pressure-difference necessary to control radon and
determine the associated air exchange rates, climatic conditions, and energy costs.
• Compare different pressure-difference scenarios and their impact on radon levels.
8-43
-------
U.S. EPA & New York State Energy Office, Reducing Radon in
Structures/Training Manual, unit 2, page 4.
Brennan T (1990) Evaluation of Radon Resistant New Construction Techniques.
Preprints of The 1990 International Symposium on Radon and Radon
Reduction Technology, Vol. 5, p.l.
White J H (1988) Radon—Tust Another Soil Gas Pollutant? Presented at 81st
Annual Meeting of the Air Pollution Control Authority, Dallas TX, USA, June
1988, p.5.
This concept was built upon conversations with Bo Adamson, Lund Institute of
Technology, Sweden. Also discussions by Sven-Olov Ericson and Hannes
Schmied (1987) Modified Design in New Construction Prevents Infiltration of
Soil Gas That Carries Radon, American Chemical Society,
0097-6156/87/0331-0526.
Sensible Living for the 90's, Bonneville Power Administration, BP A
DOE / BP-23821-4 10/90.
HPVAC-80 Envirovent, DEC International Inc., Therma-Stor Products Group,
Madison, WI.
HS-1 Humidity Sensors and Model J-3 Moisture Meter, Delmhorst Instrument
Co., Boonton, New Jersey, U.S.A.
8-44
-------
MINI FAN FOR SSD RADON MITIGATION IN NEW CONSTRUCTION
David W. Saum
Infiltec
Falls Church, VA 22041
ABSTRACT
Subslab depressurization (SSD) systems in new houses constructed
with well sealed slabs and good aggregate beds will probably achieve
excellent radon mitigation performance with fans that are considerably
smaller than the 80 Watt fans that are currently recommended. This
paper describes the development, testing, and evaluation of a low
power radon mitigation fan for installation in new houses. This "mini
SSD fan" uses only 10 Watts of power, and its radon mitigation
performance is shown to be almost as good as the larger fans. Since
the EPA plans to recommend the installation of SSD systems in hundreds
of thousands of houses that are constructed each year in high radon
areas, the long term energy savings involved in reducing fan power
could involve billions of dollars. In addition, the mini SSD fan
lowers the installation cost of the radon mitigation system, and this
might encourage more builders to follow with the EPA recommendations.
8-45
-------
BACKGROUND
One strategy for radon mitigation in new construction is for
builders in areas with high radon levels to install SSD systems in all
the houses that they build. Experience with SSD in new construction
suggests that when houses are constructed with SSD combined with a
well sealed foundation and a porous aggregate layer, then the
performance of SSD radon mitigation has been shown to be excellent.
Although many houses with these systems would not have had a radon
problem above the EPA action level of 4 pCi/L, even the lower level
houses would probably experience some radon mitigation. Since most
radon exposure occurs in these lower level houses (due to their large
numbers) , the net result of installing SSD in all houses in high radon
areas is a substantial decrease in radon exposure for occupants of
these houses.
A primary objection to installing SSD systems in all houses would
be the costs, both for initial installation and for energy and
eventual replacement. This objection could be reduced by using the
lowest cost components consistent with good radon mitigation
performance, long life, and low energy costs. This paper describes
the development, test, and life cycle cost evaluation of a low power
SSD radon mitigation fan for installation in new houses. This "mini
SSD fan" uses 10 Watts of electric power compared to the 80 Watts of
the standard fan, and its performance is shown to be almost as good as
the larger fans. Since the EPA plans to recommend the installation of
SSD systems in hundreds of thousands of houses that are constructed
each year in high radon areas, the long term energy savings involved
in reducing fan power could involve billions of dollars, and these
energy savings might provide some assistance in solutions to problems
such as global warming and U.S. energy independence. In addition, the
mini SSD fan lowers the installation cost of the radon mitigation
system, and this might encourage more builders to follow the EPA
recommendations.
PREVIOUS EXPERIENCE
Although SSD is by far the most common radon mitigation
technique, the details of its operation are not entirely clear and the
size of the fan that is necessary for effective mitigation is not well
understood. As a result, most mitigators use 80 Watt fans for most of
their mitigation jobs, and most of the industry experience is based on
the use of these fans. For new house construction, it seems that
smaller fans might be successful if the builder provides a good site
preparation by installing at least a 4" depth of large diameter
aggregate under the slab, and by sealing all slab penetrations.
Several documents have been written about construction details for
radon resistant new construction, including: a new ASTM standard1, a
Bonneville Power Administration report , and the EPA new construction
guide . Unfortunately, these documents do not contain much discussion
of fan performance versus size or of life cycle costs. The EPA will
soon issue recommendations on model code language for radon resistant
new construction, and there will be a technical support document with
life cycle cost calculations.
However, there is one study that suggests that very small fans
might provide good performance: the February 1990 EPA Symposium Paper
8-46
-------
Radon Mitigation Performance of Passive Stacks in Residential New
Construction4 by Saum and Osborne. This research showed that one
builder's passive stacks (SSD systems without any fans) offered
significant radon mitigation performance in both summer and winter.
Table 1 shows a summary of the radon mitigation results for this
study. The passive stacks reduced the radon levels by about 66%, and
45 Watt SSD fans reduced radon levels by an average of 98% of the pre
mitigation levels. Most performance reductions in these passive stack
houses are thought to be due poor installation of the stacks, or to
depressurization of the basement by leaky forced air return ducts
which reversed the passive stack pressures. This suggests that a
small SSD fan would boost the passive stack pressures enough to
overcome most of these residual mitigation problems.
MINI FAN DESIGN
The first step in this project was to design a low power and low
cost fan that could be used in a conventional new home SSD system. It
was assumed that the builder would install a 4 inches of coarse
aggregate under the slab, seal all slab penetrations, and run a stack
pipe (3" or 4" PVC) up through the slab and exiting though the roof.
In order to take advantage of the passive stack effect, the stack
should run through the heated part of the house. The desirable fan
characteristics were considered to be low power, long life, and low
cost. Conventional radon fans use about 80 Watts, have an estimated
100,000 hour life, require 2 pipe couplings to connect to the stack
pipe, and the total cost of fan and couplings is about $150.
The final mini SSD fan design is shown in Figure 1. The
conventional radon fan system consists of a fan motor, fan housing,
and two pipe couplings. To reduce noise, the conventional 45 or 90
Watt radon fans use a backward inclined blade, but 10 Watt fans are so
quiet that a conventional low-cost axial fan blade can be used. For
simplicity and lower cost, the mini fan is built into one pipe
coupling which serves as a combined fan housing and pipe coupling.
The final mini fan design consists of a high quality 10 Watt, 3"
diameter, axial fan mounted in a 3" diameter PVC ring, and enclosed in
a 3" flexible pipe coupling. When a 3" stack pipe is used, the fan
housing serves as the pipe coupling. If a 4" stack pipe is used, then
two 4" to 3" pipe couplings can be used to couple the fan to the 4"
stack. The use of 3" stacks would be recommended because of the low
air flows, the reduced costs, and the consistency with the use of 3"
plumbing stacks already installed in houses. To complete the radon
control system, a pressure gauge capable of monitoring the low
expected stack pressures was developed from the commercially available
"Fancheck" type pressure indicator. The Fancheck is a modified Dwyer
air flow meter that consists of a small ball in a tapered clear-
plastic tube. The conventional Fancheck indicates pressures greater
than about 0.2" wc, but this device was modified for use with the mini
SSD fan by using a much lighter ball so that it indicates pressures of
only a few hundredths ™ wc.
8-47
-------
COST ESTIMATES
FAN PARTS COSTS
The parts cost of the mini SSD fan is less than $20, and the
modified pressure gauge is about the same cost (about $10) as the
Fancheck gauge. Therefore the mini SSD fan could be sold to builders
for considerably less than a standard radon mitigation fan system
consists of an 80 Watt fan with 2 pipe couplings (about $150), and a
pressure indicator ($10) . It is anticipated that the mini SSD fan
system could be sold to builders for about $75, half the cost of the
standard 80 Watt fan system.
LIFE CYCLE COSTS
The largest cost savings are in the life-cycle costs, not the
initial installation costs. Table 2 shows a comparison of life cycle
costs between a 10 Watt SSD fan and a standard 80 Watt SSD fan. Three
types of recurring costs are assumed: electric cost for running the
fan continuously, wasted heat costs for warming house air that is
exhausted through the fan, and fan replacement costs. These
calculations show that the mini SSD fan would cost about $29 per
average year, while the standard 80 Watt fan would cost about $135 per
year to operate continuously.
If builders follow the forthcoming EPA recommendations that all
houses in radon prone areas have radon resistant features built into
them, then it is reasonable to assume that at least 100,000 of the
1,000,000 new home built every year will have SSD fans installed.
Under these assumptions, the estimated savings for installing a 10
Watt fan, rather than an 80 Watt fan, are shown to be $11 million in
the first year, $476 million in 10 years, and $4.6 billion in 30
years.
RADON MITIGATION PERFORMANCE
Ideally we would like to know the performance trade-off between
fan power and radon mitigation performance under a wide variety of
conditions: geology, climate zones, building practices,
heating/cooling system variations, contractor variables, failure
modes, etc. With this type of data, we could begin to make a
calculation of the cost per life saved with different SSD fan systems.
Unfortunately, this type of study is far beyond the scope of this
project.
PASSIVE STACK PERFORMANCE
The passive stack experiment data shown in Table 1 suggests that
the mini SSD fan radon reductions in new houses should be somewhere
between the performance of the passive stack systems (about 66%
reduction) and the performance of the 45 Watt fan systems (about 98%
reductions). It seems likely that if the performance of the passive
stacks is based on the extremely weak forces of stack effect, then the
performance of 10 Watt fan systems will be much closer to the 4 5 Watt
fan systems than to the passive stack systems. We believe that is not
unreasonable to assume that the mini SSD fan will give at least a 90%
average reduction of elevated radon levels in new construction,
8-48
-------
assuming the recommendation on the subslab aggregate layer and sealing
is followed.
MINI FAN PERFORMANCE IN ONE HOUSE
Only one prototype fan was available for experiments until last
month when an additional half dozen prototype units were received from
the fan manufacturer. The original prototype has been tested for
several months under worst case conditions: an older house with no
slab sealing and a poor subslab aggregate bed. These conditions were
expected to be much worse than would be found in new houses built for
radon resistance. The data from these tests is shown in Figures 2-5.
These Figures show that even in an older house with no sealing of
cracks and an uneven subslab aggregate bed, the 10 Watt mini SSD fan
lowers the radon level from 10 pCi/L to 2.1 pCi/L (a 79% reduction),
versus a reduction to 0.8 pCi/L (a 92% reduction) for a 45 Watt fan.
Figures 2 and 3 show the performance of the two fan3 over a month as
the fans are being cycled off and on every 3.5 days. Figures 4 and 5
show the averages of 4 on/off cycles so that the variations are
smoothed out. Note that both fans reduce radon levels within a few
hours after the fans are turned on, but the larger fan seems to have
depleted the radon under the slab more extensively than the smaller
fan, and it takes longer for the radon levels to build up in the house
after the larger fan is turned off.
PERFORMANCE LIMITATIONS
Some general limitations of SSD fans that may apply to the mini
SSD fan have not been fully investigated yet: 1) problems with large
slabs, 2) problems when sand or other low porosity aggregates are used
below slabs, and 3) problems when the soil below the slab is very
porous. These situations are not well understood for even the
standard SSD radon mitigation systems. It seems likely that this type
of problem could be addressed by written guidance that would be
included with the mini SSD fan systems.
FUTURE DEVELOPMENT PLANS
The future plans for development of the mini SSD fan call for more
field tests of the prototype, refinement of the design, field tests
with cooperative builders, certification (UL or equivalent), volume
purchase agreements, and eventual sale to builders and mitigators.
DISCLAIMER
The work described in this paper was not funded by the U.S.
Environmental Protection Agency and therefore the contents do not
necessarily reflect the views of the Agency and no official
endorsement should be inferred.
REFERENCES
1. Radon Control Options for the Design and Construction of New Low-
Rise Residential Buildings, Standard Guide 1990, American Society for
Testing and Materials, Philadelphia, PA, 1990.
8-49
-------
2. Nuess, M., Northwest Residential Radon Standard Project Report,
DOE/BP-1273,Bonneville Power Administration, Portland OR, October
1989.
3. Osborne, M.C., Radon-Resistant Residential New Construction,
EPA/600/8-88/087, July 1988.
4. Saum, D.W., Osborne, M.C., Radon Mitigation Performance of Passive
Stacks In Residential New Construction, Presented at the 1990
International Symposium on Radon and Radon Reduction Technology,
Atlanta, 1990.
8-50
-------
Table 1 Radon Mitigation Performance Data from Passive Stack Study
BOUSES WITB PASSIVE STACKS, NO FANS
Test
Stack
Stack
Radon
House
Open
Sealed Reduction
No.
(pCi/L)
(pCi/L>
(%>
Comment Comment
126
0.3
6.1
95%
summer data
126
0.1
13.6
99%
winter data
162
4.7
8.5
45%
duct leaks, poor communication
40
8.8
12 .8
31%
duct leaks, poor communication
53
1.1
2.7
59%
209
1.2
6.5
82%
105
0.6
1.8
67%
42
1.9
9.4
80%
duct leakage
84
4.9
5.8
16%
duct leaks, poor communication
206
2.9
19.9
85%
winter - duct leakage
206
0.6
2.4
75%
summer - duct leakaqe fixed
AVERAGE:
2.5
8.1
70%
BOUSES f/ITB
FANS IN
PASSIVE STACKS,
FANS OFF
Test
Stack
Stack
Radon
House
Open
Sealed
Reduction
Ho.
(oCi/L)
(pCi/L)
(%)
Comment
383
1.8
13.0
86%
winter
308
4.5 na
na
na
summer, duct leakage
308
8.0
33.5
76%
winter, duct leakage
221
1.5
12.0
88%
winter, duct leakage
181
6.7
7.4
9%
stack in ur.heated garage
233
na
18 na
1.9 na
not used, fan inside basement
237
7.4
13.7
46%
stack in unheated garage
184
12.7
26.4
52%
duct leakaqe
AVERAGE:
6.3
17.7
64%
BOUSES WITH
FANS IN
PASSIVE STACKS,
FANS ON
Test
Fan
Stack
Radon
House
On
Sealed
Reduction
No.
(pCi/L)
(PCi/L)
(%)
Comment
383
0.1
13.0
99%
winter
308
0.4
na
na
summer, duct leakage
308
0.2
33.5
99%
winter, duct leakage
221
0.3
12.0
98%
winter, duct leakage
181
0.1
7.4
99%
stack in unheated garage
233
na
18 na
na
not used, fan inside basement
237
0.6
] 3.7
96%
stack in unheated garage
184
0 . 8
26.4
97%
duct leakaqe
AVERAGE:
0.4
15.1
98%
COMMENTS AND CONCLUSIONS
1. This data was collected by an EPA Office of Research & Development
funded study conducted in 1989-90 and performed by Infiltec and Ryan Homes.
2. All radon measurements are averages of one or more weeks of hourly
continuous radon data collected with Pylon or Femtotech monitors.
3. Passive stacks lowered radon by about 1/3, fan systems by about 1/40
4. Passive stacks provided mitigation in summer as well as winter.
5. Passive stack performace appeared to be reduced by duct leaks and
poor subslab communication causing blocked pipes.
6. Passive stacks provided some mitigation in all cases.
7. A low power fan to assist the passive stack might overcome many of the
the problems caused by house pressures or poor communication.
8. Limitations of study: one builder (Ryan Homes), one region (D.C. Metro),
small number of houses (16), one HVAC type (heat pump).
9. All radon control systems were installed by the builder without
supervision by a radon mitigation expert.
10. Houses with summer and winter data are included twice in averages.
8-51
-------
Table 2 Life Cycle Costs of the Mini SSD Fan and Standard SSD Fan
CALCULATION ASSUMPTIONS:
GENERAL ASSUMPTIONS:
Electric rate
Gas rate
Oil rate
Fuel co3t escalation
Inflation rate
New houses built
House lifetime
New houses with SSD
New houses with SSD
HEATING ASSUMPTIONS:
Heating degree day
Gas efficiency
Gas heat for 1 cfra
Oil efficiency
Oil heat for 1 cfm
Elec. efficiency
Elec. heat for 1 cfm
Heat pump efficiency
Heat pump for 1 cfm
Avg fuel for 1 cfm
FAN ASSUMPTION'S:
Fan power
Exhausted house air
Exhaust heat cost
Fan life
Fan & gauge cost
Fan replacement cost
VALUE
$0.08 per kwh
$0.60 per therm
$1.00 per gallon
0.00%
0.00%
1 million/yr
30 years
10.00% of total
0.1 million/yr
VALUE
COMMENT
5000 deg
F days
70%
$1.12 per
year
70%
$1.36 per
year
90%
$4.60 per
year
200%
$2.08 per
year
$2.29 per
yN /
approximate U.S. electric rate
approximate U.S. gas rate
approximate U.S. oil rate
to simplify long term calc.
to simplify long term calc.
approximate U.S. average
used as long term limit
SSD — SubSlab Depres. system
this is a guess
COMMENT
approximate U.S. degree days
gas furnace and ducts
gas cost/yr/cfm of exhaust
oil furnace and distribution
oil cost/yr/cfm of exhaust
elec. furnace Ł ducts
electric cost/yr/cfm exhaust
heat pump and distribution
heat pump cost/yr/cfm exhaust
average cost/yr/cfm exhaust
STANDARD FAN
""80 watts
25 cfm
$57 /yr
11.42 yrs
$150
$250
MINI FAN
10 watts
3.125 cfm
$7
11.42 yrs
$75
$175
COMMENT
continuous electric
this is a guess
avg oil, gas & elec.
rated @ 100,000 hrs
cost to builder
fan plus install
ECONOMICS FOR SINGLE BOUSE:
STANDARD FAN
electricity (/yr)
heat loss (/yr)
replacement (/yr)
$56
$57
$22
MINT FAN
57
$7
$15
$ SAVING
—~m
$50
$7
% SAVING
8l%
86%
30%
Cost (/yr) $135
Cost (/house life) $4,056
$29
$885
$106
$3,172
78%
78%
ECONOMICS FO.R U.S.:
Costs (1st year)
Costs (10 years)
Costs (30 years)
STANDARD FAN
$14 million
$608 million
$5,882 million
MINI FAN
million
$133 million
$1,283 milllion
$ SAVING
$11 million
$476 million
$4,599 million
NOTE:
1. Costs of pipe installation and slab sealing ignored since they are conur.on
to both SSD fan systems.
2. Exhaust air leakage of 25 cfm for 80 Watt fan is a guess.
3. The estimate of 10% of new home builders installing SSD systems assumes
the EPA will recommend new home SSD in radon prone areas.
8-52
-------
Power Cord
Exhaust
Fan Motor
3" Pipe Coupling 3" Pipe Section
Figure 1 Vertical Cross-Section Schematic of Mini Fan
8-53
-------
20
\
b
Q.
C
o
"D
cd
CH
j- Waynesboro, VA house, 20 yeArs old, finished beserrent
'r Wafe-out basement, some aggregate under slab, no slab sealing J
Singles Do'nt SSD system draws 0.06" wc wilh 10 Walt fan
4 weekly far. cycles, fan on for 3.5 days eacn week
Far tutred eff
fAj \v
+—»—I—I—H
189
an turned on
1990 Julian Day
2C9
— Estimated lower mitiqation limit of 2.1 oCi/L
— Radon (pCi/L)
Figure 2 - Mitigation Performance of 10 Watt Fan
\
b
CL
c
o
"C
03
OtL
20
'5
Waynesboro, VA house, 20 years c'd, 'inished basemen:
Wa'k-out basement, some aggregate urder slab, no slat
Single point SSD syst
wesk:y fan cycles!
:n draws 0.7" wc with 45 Wat',
ar off for 3.5 days each week
5
an turned on
Fan shut off
sealing
an
w.
25
Uay
— Estimated lower mitigation limit of 0.8 pC'/L
— Radon (pCi/L)
Figure 3 - Mitigation Performance of 45 Watt Fan
8-54
-------
20
..J
\
b
CL
C
o
"D
CO
CŁ
Waynesboro, VA ho'-se, 20 years oa, finished basement
Walk-out basenient, Sonne aggregate under slab, no slab seating
Sing's point SSC system draws 0.06" wc w:th '0 Watt fan
Average o' 4 week cycles, fan en for 3.5 days each week
Fan '.urnec on
Fan turned off
205 206
208
Estimated lower litigation limit of 2.1 pCi/L
Hourly 'p-dor, -dvcaged aver wceklorv; 'rn on/of cycles
Figure 4 Average Mitigation Performance of 10 Watt Fan
\
b
Cl
c
c
"a
C3
cr
'5
Waynesboro, VA house, 20 years o'd. finished basement
Walk-cut basement, sime aggregate under slab, re s ab sealing
Single point SSD system draws 0.7" wc with 45 Wat: fa-
Average ctf f- week cycles, fan off for 3.5 days eacn week
ran turned on
t-
*0 -
Fnr, turner off
rJ
V\
V
Af
i /
25
Sunday 8/5
26
-| ..
27 28
29
-I-
Average Days
— E-oUmalftd icwnr m:liqatio~ liTil of 0.8 pCi/L
— Average of 4 wceVlong fan cr/off eye'es
Figure 5 Average Mitigation Performance of 45 Watt Fan
8-55
-------
BUILDING RADON MITIGATION INTO INACCESSIBLE CRAWLSPACE
NEW RESIDENTIAL CONSTRUCTION
D. Bruce Harris and A. B. Craig
U. S. EPA
Air and Energy Engineering Research Laboratory
Research Triangle Park, NC 27711
and
Jerry Haynes
Hunt Building Corporation
El Paso, TX 79984
ABSTRACT Specifications for new residential housing
units for base personnel at Ellsworth Air Force Base, Rapid
City, SD, called for demonstrated radon levels below 4
picocuries per liter (pCi/L) before they would be accepted
by the Air Force. Hunt Building Corporation decided that it
would be cheaper to build radon control systems into all the
units than to have to retrofit some later. The Radon
Mitigation Branch of EPA's Air and Energy Engineering
Reasearch Laboratory assisted during the design and
installation of the active soil depressurization (ASD)
systems and followup measurements. The buildings utilized
below grade wood floor construction over an inaccessible
crawlspace because of the highly expansive soils. The
initial installations demonstrated the need for complete
sealing of the floor system. An effective quality control
scheme was instituted which tested the negative pressure
field established under every building and required
additional sealing until each corner of the floor was under
at least 2.5 pascals (Pa). Early data indicate that moderate
levels of radon (100 pCi/L) exist in the crawlspace when the
mitigation fan is off for several days and virtually none
when it is on. Results from several buildings are presented.
This paper has been reviewed in accordance with the U.
S. EPA's peer and administrative review policies and
approved for presentation and publication.
BACKGROUND
A major housing project which will be leased by the Air
Force at Ellsworth Air Force Base, Rapid City, SD, is being built
by the Hunt Building Corporation. The housing project has 828
residential units in 251 buildings, consisting of singles,
8-57
-------
duplexes, and quadruplexes. The area where the houses are being
built is known to be radon-prone. The Air Force had tested 30 of
their 2500 houses and found about 60% of them above 4 pCi/L. As
a result of this high radon incidence, the Air Force's initial
Request for Proposal (RFp) contained a radon performance clause
requiring the builder to test the houses before occupancy to
guarantee that they were below 4 pCi/L. A unit testing below 4
pCi/L is accepted by the Air Force for occupancy and a 1-year
alpha track test is comraehced. If radon levels in the living
area test below 4 pCi/L for the first year, then Hunt has met its
performance requirement and no longer has any responsibility. If
the house initially tests above 4 pCi/L, then Hunt must bring the
level below 4 pCi/L before occupancy. When the 1 year test
increases above 4 pCi/L, then the Air Force will stop payment
until the level has been brought below 4 pci/L.
With limited radon control experience, Hunt Building
Corporation contacted EPA to seek advice on the best way to
construct these raultifamily housing units to ensure radon levels
of below 4 pCi/L or to mitigate them to this level if they are
found to contain higher levels when tested. The decision was made
by Hunt to install a radon mitigation system in all units since
retrofit into an inaccessible crawlspace would be very difficult
and potentially expensive.
MITIGATION SYSTEM DESIGN
Most of the units are two-story quadruplexes with the lower
level built approximately 3 ft (1 m) below grade (Figure 1). The
8-58
-------
individual units are separated by a double wall for sound
deadening, but no firewall. Because the units are built on
expansive soils the Air Force is requiring that the lower floors
be treated plywood over joists with a crawlspace below. The
units are built with a 6 in. (15 cm) crawlspace between the
bottom of the joists and the clay under the units. The building
site is being excavated to a depth of 5 ft (1.5 m) and backfilled
with compacted glacial aggregate in an effort to stabilize the
ground and minimize movement. This is moraine till which is
quarried on site consisting of a moraine stone with a great deal
of fine sand and some soil in it.
After reviewing the various techniques which AEERL has
tested on the mitigation of radon in crawlspace houses, it was
recommended that the most cost-effective way to mitigate the
house with a high level of assurance of lowering levels to below
4 pCi/L was to use either suction under a polyethylene sheet in
the crawlspace or suction on the crawlspace itself. It was
decided that there was an excellent chance of making suction on
the crawlspace satisfactory by doing a thorough job of sealing
the plywood subfloor. The plywood is tongue and groove along the
8 ft (2.4 m) edge, and all 4 ft (1.2 m) edges are on joists. The
plywood is screwed to the joists. No outside vents are in the
crawlspace, and every effort is made to make the crawlspace as
airtight as possible. Moisture should be controlled by the active
mitigation system. The joint between the floor and the concrete
wall is sealed with polyurethane caulk. Any cuts through the
polyethylene and plywood for pipes are carefully sealed around
the pipe with polyurethane foam. Hunt has built a box in the
8-59
-------
joists under the bathtub so the bathtub trap does not penetrate
into the crawlspace and provide a possible radon entry route.
One suction pipe per unit is used. This 6 in. pipe, in the
wall between the two middle units of each quadruplex, extends
straight up to the fan in the attic and exits the roof
immediately above.
TESTING FIRST UNITS
Since the building season is short in South Dakota, testing
began as soon as the first crawlspace floor was installed. AEERL
sent a team to Ellsworth Air Force Base to install a temporary
fan on the system and to measure pressure reduction at the
various points of the crawlspace below the plywood floor. When
negative pressures can be achieved throughout the crawlspace as
measured in all four of the corners, the greatest distances from
the suction point, then no soil gas should be sucked into the
house. A 6 in. pipe was installed through the deck of a
quadruplex at the intersection of the central "party" wall and
the central beam which ran the full length of the unit. An axial
aligned centrifugal fan was mounted on the pipe with power
provided through a speed controller. An electronic manometer
measured the pressure under the deck at several locations. Once a
few small leaks had been sealed, at least -0.010 in. WC (-2.5 Pa)
was obtained at each corner of the building and more than 1.0 in.
WC (248.8 Pa) fan suction was recorded.
A duplex was the second unit tested. No perimeter or
penetration sealing had been done so this provided an excellent
opportunity to test the effect the extra sealing had on the
suction obtained. No suction was detected at the corners and only
8-60
-------
0.2 in. WC (49.8 Pa) of fan suction was measured which was the
pressure drop through the floor opening. The fan suction was
monitored while the sealing crew proceeded to close the openings.
Little change was noticed until most of the wall joint and the
pipe penetrations were closed. As the final openings were closed,
dramatic increases in the fan suction were seen. The installation
crew understood the need for carefully sealing all openings as a
result of this test. This effort also provided the basis for the
strict quality control and quality assurance program instituted
by Hunt.
Every deck was tested when it was sealed and all plumbing
activity was finished. A separate crew performed the tests and
repaired breaks in the seal. A temporary fan depressurized the
crawlspace and the sub-floor suction was measured. Any unit that
failed to draw at least 0.010 in. WC in the corners was checked
for leaks and resealed before additional construction activities
were allowed to proceed. All buildings completed so far have
exceeded the requirements when the seal was completed and a 6 in.
fan installed. It may be possible to reduce the power consumption
further by adding a speed control to the fans.
RESULTS
As the first units were finished, an AB-5 Pylon radon
monitor was used to check for radon in three units: a finished
one with an operating mitigation system, a unit with a finished
floor which had been sealed for 1 week, and a third with a just-
completed floor without final sealing. No radon was found under
the floor in the finished unit nor in the mitigation system duct.
8-61
-------
Likewise, the completed but not sealed unit had no measurable
radon in the crawlspace. The crawlspace under the sealed floor
did yield levels of 100 pCi/L. This is a moderate source
strength, but could be enough to elevate unmitigated units above
the EPA action level of 4 pCi/L because the shell of each house
is very tight and low dilution could be expected.
Acceptance testing of the first 130 unit section yielded
levels between 0.8 and 2.4 pCi/L except for a single unit which
had a carbon cannister reading of 16.0 pCi/L. A check of the
system in this house found that the circuit breaker had been
turned off? consequently, the reading was really an indication of
the level that could have been expected if no mitigation system
had been installed. A retest with the fan operating showed the
levels reduced to 2.5 pCi/L. This fan inoperation showed that the
effort and commitment of Hunt to install the mitigation systems
and insist on an effective quality assurance program was well
worth the investment. Retrofitting mitigation systems into these
units would have been much more costly than doing it during the
building phase.
8-62
-------
00
unit 1
unit 2
unit 3
6-*- suction hole
unit 4
crawlspace
\
'' garage slabs
Figure 1. Quadruplex foundation plan
-------
THE EFFECT OF SUBSLAB AGGREGATE SIZE ON PRESSURE FIELD EXTENSION
K.J. Gadsby, T. Agami Reddy,
D.F. Anderson, and R. Gafgen
Center for Energy and Environmental Studies
Princeton University
Princeton, NJ 08544
Alfred B. Craig
Air and Energy Engineering Research Laboratory
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
ABSTRACT
Four sizes of commercially available crushed blue gravel
aggregate (3/8, 1/2, 3/4, and 1 in. nominal diameter) were tested
in a laboratory apparatus designed to experimentally determine the
aerodynamic pressure drop coefficients of porous material such as
soil and gravel. Permeability values for crushed stone of 1/2 and
3/4 in. nominal diameter were found to be 10-20 times higher than
those reported in a previous study for river-run gravel of the same
nominal diameter. Pressure field extension of this aggregate, when
suction is applied to a single central suction hole (a practice
widely used for mitigating buildings with elevated radon levels by
the subslab depressurization technique), is also generated based on
a disc flow model previously studied by the authors. Application
of the disc flow model to residences with both basements and slab-
on-grade construction is also described. Theoretical computations
indicate that the permeability of soil around the foundation walls
and periphery of the residence is a more crucial parameter
affecting the pressure field extension than the permeability of the
subslab gravel bed. The laboratory studies will be field verified
in new construction of schools and houses in the future.
This paper has been reviewed in accordance with the U.S.
Environmental Protection Agency's peer and administrative review
policies and approved for presentation and publication.1
'This work was funded by the U.S. Environmental Protection
Agency under Cooperative Agreement No. CR-817013.
8-65
-------
PROBLEM STATEMENT
Subslab air flow dynamics provide important diagnostic
information for designing optimal radon mitigation systems based on
the subslab depressurization (SSD) technique. An earlier study
(Refs. 1 and 2) showed that subslab air flow induced by a central
suction point can be mathematically treated as radial air flow
through a porous bed contained between two impermeable discs.
Subsequently, it was suggested that subslab material commonly found
under residential buildings be categorized and tested in the
laboratory in order to deduce their aerodynamic pressure drop
coefficients. This would then permit the pressure field extensions
to be inferred which would be of practical and realistic importance
to radon mitigators. To this end, a laboratory apparatus was
designed and built which is described fully in Ref. 3.
The scope of this study was limited to four aggregate mixes —
No. 600, 601, 602, and 603 — provided by Vulcan Materials from
their Manassas Quarry. The physical specifications provided by
Vulcan Materials are given in Table 1. The aggregate is crushed
blue gravel, a material commonly used as subslab material for
residential and commercial housing.
This paper first reports aerodynamic flow coefficients of the
four aggregate mixes obtained from EPA's laboratory apparatus.
Subsequently, it describes how the disc flow model can be applied
to residences with both basement and slab-on-grade construction.
Finally, it shows generated pressure field extension plots for each
aggregate in the framework of the above model and discusses the
practical implications of these plots in the design of SSD systems.
EXPERIMENTAL RESULTS
The laboratory apparatus, described fully in Ref. (3), is
shown in Fig. 1. It consists of a straight length of 8 in." PVC
pipe approximately 5 ft long to the bottom of which a 1/4 in.
aluminum removable screen is fitted. The screen, which slides in
and out along a groove machined into the PVC pipe, is perforated
with a pattern of 1/4 in. diameter holes to permit air flow. The
porous material is loaded into the test column from the top after
which another sleeve, containing flow straighteners and the air
supply inlet tube, is placed on top of the column. Air is supplied
to the system from the top and several pressure taps provide
information on the pressure drop in the porous bed as air flows
down through the porous bed in the test column. The total air flow
rate is accurately measured. The experimental procedure involves
* For readers more familiar with metric units, 1 in. = 2.54 cm and
1 ft = 30.48 cm.
8-66
-------
measuring the pressure drop (AP) through a known and predetermined
bed length (AL) for a specific value of total flow rate (q).
Several such tests are carried out over a range of q values for
each sample of aggregate and for different samples of the same
aggregate. A least-square regression finally provides estimates of
the permeability (k) and the flow exponent (b) of the aggregate.
Table 2 assembles results of the porosity experiments, three
different samples of each aggregate tested at least twice. This
involved choosing a volume of the aggregate material and then
finding the porosity by measuring the volume of water needed to
completely saturate each sample (Ref. 4). The values of porosity
( 0.9) .
This is an indirect indication that the experimental design was
satisfactory. The relatively lower R2 values for mix No. 603, which
is the largest gravel, could be a result of the fact that, in order
to keep Reynolds numbers low, the pressure drop was measured close
to the sensitivity of the instruments (which was 0.25 Pa or 0.001
in. WC).
8-67
-------
The mean and the coefficient of variation (CV) in percentage
values for k and b are also given in Table 3. Permeability values
of mix No. 600 are around 10"7 m2, gradually increasing to about 7
x 107 m2 for mix No. 603. It is worth pointing out that mixes No.
601 and 602 (1/2 and 3/4 in., respectively) have k values which are
10-20 times larger than those of river-run gravel of the same
nominal diameter, as reported in a previous study (Refs. 1 and 2) .
A possible explanation for this is that river-run gravel beds have
lower porosity; i.e., they tend to pack more closely (Refs. 1 and
2) . This important finding suggests that crushed aggregate is more
suitable as a subslab material fill than river-run gravel for
buildings to be mitigated using the SSD system since the former is
likely to have a larger pressure field extension from the suction
hole. The flow exponent b does not vary too much with mix. The
relatively higher value for mix No. 602 does seem surprising since
both No. 601 and 603 have lower values for b. A possible reason
for the low b value for mix No. 603 is that it had relatively more
fines in the aggregate. CV values for k are low (less than 3%) ,
while those for b are high, as much as 12% for mix No. 603.
PRESSURE FIELD EXTENSION
This section gives results of computing the pressure field
extension under the slab when the four crushed gravel sizes are
used as subslab fill material, using the disc flow model described
in Ref. 1. Consequently, the following conditions are assumed:
only one suction hole is used,
the suction hole is located at the center of the slab,
the slab is circular, and
the edges of the slab communicate uniformly with the
ambient air.
Figure 2 illustrates how flow conditions in an actual house
with a basement can be visualized in the framework of the disc
model. The square basement is approximated as a circle of radius
r0 while the extra flow path H through the soil around the house is
accounted for by effectively increasing the circle radius to R.
Thus the flow is assumed to occur between two impermeable discs,
the upper end being the underside of the basement slab and the
lower end being the soil beneath the gravel bed. Since the
material under the slab is gravel while the extra flow path around
the sides of the house is through soil of different permeability
than that of the gravel, this model was modified to represent a
disc made up of two materials of different permeabilities: soil
between radii r0 and R, and gravel between r0 and the central
mitigation pipe. As pointed out in Ref. 1, the flow regime would
likely be turbulent through the gravel bed and laminar through the
soil.
8-68
-------
Consequently the total pressure drop AP (in head of water)
across the two-material bed is equal to the pressure drop through
the gravel plus the pressure drop through the soil. From equations
derived in Refs. 1 and 2:
A P =
_1_ . _^_a . P a ./ Q \b . 1 l-jb _
kg g p„ \2nh) (l-b) * ° * >
w a
9
P,
2izh
•In
R
where kg - permeability of the gravel bed,
v - kinematic viscosity,
g - gravity
p - density,
q - total air flow rate,
h - thickness of the gravel bed,
b - flow exponent,
rc - radius of the basement
r, - radius of the suction pipe,
k, - permeability of the soil surrounding the house,
R - total radius of flow (equal to that of the
basement and of the extra flow path through
the soil).
Note that the pressure drop given by Equation (1) is not the
entire pressure drop to be overcome by the mitigation fan. The
pressure drop due to entrance effects into the mitigation pipe and
that due to the straight pipe, fittings, and bends could account
for as much as 50% of the entire pressure drop in the mitigation
system.
Two types of construction were studied: slab-on-grade and
basement houses. In the framework of the disc model, the
difference between the two is solely in terms of the extra flow
path through the soil. Basement houses will tend to have higher H
values (see Fig. 1) than slab-on-grade houses; consequently the (R-
r0) value will be correspondingly different. The following values,
deemed representative, have been chosen in all calculations that
follow:
slab-on-grade: R = r„ + 1 m
basement house: R = r„ + 3 m
Also selected were r, = 5 cm (i.e., a 4 in. diameter suction
8-69
-------
pipe for the mitigation system) and a value of h = 0.05 m (which is
based on experience and supported by tests in an actual house
described in Ref. 1) . Equation 1 cannot be directly used to
compute the required suction pressure for different values of
basement radius r0, since the total air flow rate q is not an
independent variable. It is determined by the practical criterion,
which is that the pressure below the actual slab (i.e., up to a
radius r0) should be lower than the ambient pressure P, by an amount
larger than the depressurization of the house arising from natural
causes (e.g., wind, stack effect, heating, ventilation, and air
conditioning system (HVAC) depressurization). A realistic range
for this depressurization is 3-10 Pa (Ref. 5). Consequently, the
flow rate q should be such that a pressure drop equal to this
depressurization occurs in the outer portion of the disc containing
soil; i.e., between R and rc. Since the flow is laminar in this
region, q is computed from:
where AP is the prespecified minimum depressurization below
all points of the slab expressed in head of water.
The value of q is then used in Equation (1) to calculate the
required suction pressure at different values of slab radius r0.
Note that the numerical value of k,, soil permeability, greatly
influences q and consequently the pressure field extension. Hence
the two following cases were chosen in order to study the
sensitivity of the pressure field extension on the type of boundary
fill material (Ref. 6):
(i) k, = 10"8 m2, corresponding to sand and gravel mixtures,
(ii) k, = 10"10 m2, corresponding to fine sands.
Additionally, a value of 10 Pa has been selected as the
required pressure drop through the soil between radii R and rQ due
to reasons discussed above.
Figures 3 and 4 present the computed suction pressures for
cases (i) and (ii), respectively, for the four gravel sizes tested
and for both basement and slab-on-grade type construction. Note
that differences in field extension between gravel sizes for the
subslab bed material, especially mixes No. 602 and 603, are much
less important than those resulting from soil type selection.
Figures 3 and 4 show that there is an order of magnitude difference
in pressure requirements between the two construction types; it is
lower for case (ii), lower permeability soil. This fact highlights
the importance of having a ring of low permeability soil around the
foundation walls and the periphery of the building. There is an
(2)
<7 = (AP)
Va Pa
8-70
-------
important difference in suction pressures between basement and
slab-on-grade construction for case (i) , but this is so for the
tighter soil (Fig. 4). This is because most of the pressure drop
occurs across the outer ring of soil, with practically no drop
through the gravel bed itself. Thus theoretically, for fine soils
(case ii), the pressure field extension is very large (radius
greater than 50 m). However, in practice, looseness in packing of
the soil, short-circuiting of flow paths, or small holes or cracks
in the slab will drastically decrease this value.
Finally, Fig. 5 presents the total air flow rate for case (i)
for slab-on-grade and basement construction. These flows are
independent of the gravel size since they have been computed from
Equation (2) following the condition that the pressure drop in the
outer ring consisting of soil is equal to a prespecified minimum
amount taken as 10 Pa in this study. Note that the air flow rates
between the slab-on-grade and basement cases differs by almost a
factor of three.
FUTURE WORK
This study was undertaken to experimentally determine the
aerodynamic flow coefficients (permeability and flow exponent) of
four crushed aggregate mixes of commercially available stone. The
sizes, ranging from 3/8 to 1 in. nominal diameter, are gravel sizes
often used as subslab fill for residences and large buildings. An
important finding of this study is that permeability values of 1/2
and 3/4 in. crushed gravel were 10-20 times higher that those
reported in a previous study for river-run gravel of the same
nominal diameter. Field validation of the computed pressure field
extension using laboratory test results is important and such tests
are currently being planned in newly constructed schools,
commerical buildings, and houses. The results presented here
should be used with caution until such time that the apparatus and
experimental design are more fully validated.
ACKNOWLEDGEMENTS
D. Harrje and R. de Silva contributed generously during the
design and construction of the laboratory column. Useful
discussions with A. Cavallo and R. Sextro are also acknowledged.
8-71
-------
NOMENCLATURE
b flow exponent
dv equivalent diameter of gravel
CV coefficient of variation
g acceleration due to gravity
h thickness of the porous bed
k permeability of gravel
AL length of porous bed in flow direction
AP pressure drop
P pressure
q total volume air flow rate
R2 coefficient of determination of regression
Re Reynolds number
R total radius of air flow path
r0 effective radius of basement
r, radius of suction pipe
p density
v kinematic viscosity
-------
REFERENCES
1. T.A. Reddy, K.J. Gadsby, H.E. Black III, D.T. Harrje, and R.G.
Sextro, "Simplified Modeling of Air flow Dynamics in SSD Radon
Mitigation Systems for Residences with Gravel Beds," report
submitted to U.S. EPA, February 1990.
2. K.J. Gadsby, T.A. Reddy, R. de Silva, and D.T. Harrje, "A
Simplified Modeling Approach and Field Verification of Airflow
Dynamics in SSD Mitigation Systems," presented at the 1990
International Symposium on Radon and Radon Reduction
Technology, February 19-24, Atlanta, 1990.
3. T.A. Reddy, K.J. Gadsby, D.F. Anderson, and R. Gafgen,
"Experimental Laboratory Tests on Vulcan Crushed Aggregate,"
Report submitted to AEERL, U.S. EPA by the Center for Energy
and Environmental Studies, Princeton University, in January
1991.
4. M. Muskat, The Flow of Homogeneous Fluids Through Porous
Media. McGraw-Hill, 193 7.
5. EPA Manual: Reducing Radon in Structures, 2nd Edition, United
States Environmental Protection Agency, Office of Radiation
Programs, Washington, D.C., 1989.
6. W.W. Nazaroff, B.A. Moed, and R.G. Sextro, "Soil as a Source
of Indoor Radon: Generation, Migration, and Entry," Chapter 2,
Radon and Its Decay Products in Indoor Air, W.W. Nazaroff and
A.V. Nero (Eds.), John Wiley & Sons, 1988.
8-73
-------
TABLE 1. QC ANALYSIS OF VULCAN MATERIALS CRUSHED AGGREGATE AS SUPPLIED BY THE COMPANY
Loose
Vulcan % Passing Screen Size Bulk Void
No. Description 1-1/2 1 3/4 1/2 3/8 3/16 3/32 Density Volume
600 3/8 in. clean 100 90 23 5 91.8 49.3
601 1/2 in. clean 100 83 27 3 1 89.4 50.6
602 3/4 in. clean 100 93 27 7 2 95.1 47.5
603 1 in. clean 100 94 40 6 2 1 93.0 48.6
-------
TABLE 2. COMPARATIVE RESULTS OF POROSITY TESTS
Porosity
Vulcan
Nominal
From
Present
No.
Size (in.)
Supplier
Tests
600
3/8
0.493
0.403
(0.006)*
601
1/2
0.506
0.419
(0.011)
602
3/4
0.475
0.436
(0.015)
603
1
0.486
0.452
(0.009)
*Values in parentheses are standard deviation values.
8-75
-------
TABLE 3. SUMMARY OF EXPERIMENTAL DATA
Number
of Range of Number Regression Permea-
Vulcan Nominal d, Samples Flow Reynolds of Obser- Model bility Exponent
No. Size Tested Range Numbers vations R2 k x 10' b
(in.) (mm) (1/min) (m2)
600 3/8 2.36 3 6 - 21 5 - 18 20 0.91 104.4 1.14
(2.5)* (7.0)
601 1/2 3.03 2 19 - 28 20 - 30 11 0.99 193.6 1.15
(0.4) (3.5)
602 3/4 4.37 2 19 - 28 29 - 42 10 0.92 336.3 1.29
(1.2) (10.9)
603 1 5.62 3 19 - 30 37 - 58 16 0.84 683.9 1.17
(1.2) (12.0)
* Values in parentheses are coefficient of variation (CV) values where CV = (Mean value/Standard error of
the mean) x 100
-------
Figure I. Sketch of the flow apparatus. I >st of components is attached
8-77
-------
Key for Figure 1
(1) Main Column - Schedule 40, PVC, 8 in. Pipe
(2) Test Rig Panel
(3) Support Blocks
(4) Test Material Cavity
(5) Lower Sleeve
(6) Movable 1/4 in. Aluminum Screen
(7) Upper Sleeve
(8) Collection Bin
(9) Assembly Bolts
(10) Short Pipe Section
(11) End Cap
(12) "O" Ring
(13) Flow Straighteners
(14) Pressure Taps
(15) Urethan Tubing
(16) Tubing Union
(17) Miniature Solenoid Valves
(18) Solenoid Panel
(19) Valve Selector Switch
(2 0) 12 Volt DC Power Supply
(21) Manifold
(22) Manifold Panel
(23) Transducer Selector Switches
(24) Pressure Gauges
(25) Interchangeable Connector Tubing
(2 6) Gauge and Flowmeter Panel
(27) High Pressure Transducer
(28) Low Pressure Transducer
(29) DVM Connector
(30) Digital Voltmeter (DVM)
(31) Wiring Harness
(32) Air Supply
(33) Shutoff Valve
(34) Flow Control Valve
(35) Flowmeters
(36) Tubing from Flowmeter to Column
8-78
-------
Air flov.'
path
f fsV/W
Mitigation
/pipe
Sol
Easement slab
Gravel bed
Figure 2. Disc model of a SSD mitigation system in a basement house.
8-79
-------
o
o__
UJ
or
S
w
LJ
SLAB-ON-GRADE
VoCO
V601
V603
20 30
SLAB RADIUS (m)
so
CL,
UJ
Of
tu
Ł
o
BASEMENT
V600
V601
V602
V503
BASEMENT RADIUS (m)
Figure 3. Pressure field extension for case
-------
SLA0-ON-GRADE
a
LtJ
ctr
13
00
in
LJ
OZ
Q_
Z
o
a
3
m
6
tt:
~
o
bJ
q;
1M-
11.2-
11-
10B-
ms-
10.4-
10.2-
V600
V601
V602
V603
10—i i "T"" J—I—I—i i—| 1—r—I—i—j—i 1 i—i—f i—i 1—r
20 30
SLAB RADIUS (m)
50
ias-
BASEMENT
o
LlJ
(Z
ZD
V)
U1
3
in
Q
LlJ
O
10.4-
ia.3-
102-
V600
V601
V603
V602
BASEMENT RADIUS (m)
50
Figure 4. Pressure field extension for case (ii) : ka «= 10"10 mz for
the four gravel mixes tested. Soil flow path length is
1 m for slab-on-grade and 3 m for basement houses.
8-81
-------
90-
SLAB-ON-GRADE
GO
LxJ
I—
<
CŁ
O
i 1 1 1 j 1 1 1 1 1 1 1 1 r
20 30
SLAB RADIUS (m)
-i 1 r
50
BASEMENT
u_! 1 1 1 1 1 1 1 " 1 1 1 1 1 1 1 1 1 1 1 1 r—t r—i
0 10 20 30 40 50
BASEMENT RADIUS (m)
Figure 5. Total air flow rates for case (i) : k5 = 10"* m2, computed from
Eq. (2).
8-82
-------
Session IX
Oral Presentations
Radon Occurrence in the Natural Environment
-------
COMBINING MiTTfiaTTrw ft nsnr.nnv;
INDOOR RflDON REDUCTION BY ACCESSING THE SOURCE
Stephen T. Hall
Radon Control Professionals, Inc.
Reston, Virginia 22094
ABSTRACT
Soil radon testing has shown that radon sources are concentrated
in narrow linear areas congruent with local geology in the Eastern
Piedmont, which should also hold true in any folded mountain belt
region with heterogenous geology.
In existing buildings, If microroanometer tests indicate poor
communication in the sub-slab environment, soil radon concentration
gradients can be mapped with instantaneous sub-slab radon measurements.
By then orienting these difficult-to-mitigate homes on a geologic map,
we have been able to predict the location of the radon source adjacent
to foundation walls. Tapping these source areas with a multi-duct sub-
slab depressurization system has been shown to be effective in
achieving optimum radon reductions.
By using this method of radon soil testing for the construction of
new large buildings, such as schools, to locate areas of sub-slab
depressurization, maximum indoor radon reductions can be achieved with
minimal installations.
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.
9-3
-------
in large buildings, such as schools and office buildings, and
in homes without good sub-slab air-flow communication, e.g. no
aggregate, we have achieved significant indoor radon reductions by
sub-slab ventilation at the source of maximum soil radon
concentrations using quantitative diagnostic tests which Incorporate
the correlation between radon soil testing and local geology.
Recent measurements of soil radon availability numbers by the
author (1) have yielded correlations between indoor radon levels in
homes, office buildings, and schools and the various geologic units
in the Coastal Plain, Piedmont, and Mesozolc Basin. The radon
availability number was determined using the equations of Nazaroff,
et al (2) and Tanner (3), whereby radon availability number is a
function of soil radon content, permeability, and diffusion
coefficient. The equipment used consists of a Pylon radon monitor
with attached Lucas cell and soli probe developed by the author.
The probe has an in-line flow meter and pressure gauge (which must
have an appropriate range for the permeability values Inherent lii
the particular soils being measured) and a drying tube, cut-off
valve, and Swaglok connector which attaches to the in-ground section
of the probe assembly. This in-ground section consists of a three
foot long metal tubing surrounded near its base by an inflatable
packer to prevent atmospheric dilution.
Because soil permeabilities in the Piedmont, Coastal Flain, and
the Mesozoic Basin of Northern Virginia and Maryland are low enough
that radon migration is predominately diffusion driven, it was
decided to calculate the radon availability number based upon the
soil radon concentration and diffusion coefficient of the soil.
Soils were then tested around a number of basement homes and schools
remediated by RCP. Therefore good data was available on the
original radon values and construction characteristics.
Determined radon availability numbers, plotted against indoor
radon levels, revealed two distinct populations (Figure 1). The
lower population (i.e. those with a higher radon availability number
to Indoor radon ratio) consists entirely of buildings having one or
more of the following four factors:
1. A vented crawlspace,
2. A tight or sealed slab-wall contact,
3. A controlled fill around basement walls that has a low radon
availability number,
4. No basement.
9-4
-------
200 -
o
O uo"
R
R 140-
A
D
0 120-
N
» m
P 100-
C
j*7% / with mitigatingJactgrs_
— «!JL .
lo 20 ao 40 so *0 70 to
RADON AVAILABILITY NUMBER, KBq/m2
Figure 1. Correlation between soil test results (radon availability number)
and indoor radon concentration.
Interestingly, the author had previously discovered In the
George Mason University (GMU) Radon Study, that in the same geologic
setting, basement homes vith cravlspaces tended to have lower indoor
radon than those without cravlspaces. Apparently this is a function
of the fact that cravlspaces are normally attached at one end of the
basement and have fresh air vents to the outside, at least in the
local area studied. Cravlspaces are also usually separated from the
rest of the basement by a wall, thereby acting as a decoupled unit
without the decrease in indoor air pressure that the basement
experiences.
9-5
-------
The upper population (i.e. those with a lower radon
availability number to indoor radon ratio) consists entirely oŁ
hones and schools with none of the factors Inherent in the lover
population. It is th13 trend that one would want to use to predict
magnitudes of Indoor radon problems, based upon soil tests for homes
or buildings without any radon mitigating factors. Figure 2-7
(highest values darkened), illustrates that both slab-wall
separation radon measurements (interior semi-circles) In partially
completed schools have corroborated the location of maximum radon
potentials determined from soil tests.
In most cases, elevated sub-slab radon levels and soil test
results have been shown to be concentrated in linear areas for the
various geologic units around the DC metro area, which should also
hold true in any folded geologic region with heterogenous geology.
These linear areas or "bands" can be one foot to a few tens of feet
wide. Importantly, the orientation of the high radon potential
llneations correlate well with the trend of local rock layers
(generally n30°E), or with the trend of local shear fractures
(generally N45°W to N60"W). For example, a boundary between high
and low radon potentials is shown in Figure 2, along a N60°W
fracture trend and in Figure 3, along a H45°w fracture trend.
Figure 4 shows a diagonal band through the central area of the
school along a N45°W trending fracture pattern. Figures 5 and 6
show correlations between high radon potentials and N30°E trending
rock layers; both revealing a linear band through the interior area
of the school.
xfp
\
\
\
\
\
\
\
\
\
\
\
\
fa n n
\
\
©
y
\
Figure 2. Footprint plan of a school showing highest soil radon potentials
(RAN) and indoor slab/wall joint radon concentrations darkened.
Highest soil radon potentials follow N60°W fracture trends.
9-6
-------
-- ©"~k
Figure 3. Footprint plan of a school showing highest soil radon potentials
(RAN) and Indoor slab/wall joint radon concentrations darkened.
Highest soil radon potentials follow N45°W fracture trends.
/
/ N 1 I » N
Figure 4. Footprint plan of a school showing highest soli radon potentials
(RAN) and Indoor slab/wall joint radon concentrations darkened.
Highest soil radon potentials follow N45°W fracture trends.
-------
<3
Figure 5. Footprint plan of a school showing highest s<.>il radon potentials
(RAN) and Indoor slab/wall joint radon concentrations darkened.
Highest soil radon potentials follow N30aE trending rock layers.
Figure 6. Footprint plan of a school shoving highest soil radon potentials
(RAN) and Indoor slab/vall joint radon concentrations darkened.
Highest soil radon potentials follow N30°E trending rock layers.
9-8
-------
For the construction of new large buildings, such as schools,
radon soli testing has proven valuable In locating th h sources of
maximum radon availability. By locating sub-slab ventilation points
In the vicinity of these areas maximum indoor radon reductions can
be achieved with minimal sub-slab ventilation Installations.
In existing homes with a footprint area of less than 2000 ft',
If sub-slab mlcromanometer tests indicate good air-flow permeability
(good sub-slab communication), the location of ventilation points Is
not critical because one fan with one penetration will draw radon
from everywhere under the slab. However, when sub-slab
communication is poor, sub-slab and blockwall radon concentration
gradients can be mapped with instantaneous radon measurements to
determine the orientation and location of high radon potential
lineatlons under the building, based on a knowledge of the local
geology.
For example, Figure 7 shows a home with all the high sub-slab
and blockwall radon levels along the NW side, indicating a linear
source at the NW end oriented N30°E, parallel to the local rock
layers. Mlcromanometer tests Indicated negligible sub-slab
communication. A sub-slab ventilation system, installed with one
wall and three slab penetrations along the NW end, brought Indoor
radon levels down to 0.5 pCi/1.
Highest radon
potential
rock layers
wall suction
250 y point
H
698
3 n h - s 1 a b
407
ventilation
; y s t. a m •
©
77
Q FAN
W
G
0
-/TV.
G Sub-slab radon, pCi/1
-O-Wall penetration radon, pCi/1
Figure 7. Footprint plan of a home showing numerical values of sub-slab
and blockwall radon concentrations that indicate that the radon
source Is following N30°E rock layers, delineated by dash lines.
Sub-slab ventilation systems penetrations are shown as darkened
circles.
9-9
-------
Figure 8 illustrates a similar house situation where high radon
potentials are parallel to N30°E rock layers and generally increase
toward the SE. Sub-slab ventilation as shown brought Indoor radon
levels from 30 pCl/1 to 1.5 pCl/1.
<0 474
-->¦ - ~" / s / s / /
ff) 1825
(D59
• / / / // / /¦¦¦/ / / / ' / // '', Footing
Sub-slab
ventilation
system
/ //
® 1325
/- /
jL
®62
/ ///// /' / ///>/
1255 ®
Highest radon
potent ial
rock layers
Footing
Figure 6. Footprint plan of a home shoving numerical values of sub-slab
and blockvall radon concentrations that indicate that the radon
source is following N30°E rock layers, delineated by dash lines.
Sub-slab ventilation systems penetrations are shown as darkened
circles.
Figure 9 shows a workplace building with a footprint area less
than 2000 ft2 where micromanometer readings indicated no sub-slab
communication because the slab was poured directly on compacted
clay. Construction material radon levels tested negative. However,
sub-slab radon levels increase towards the SE, congruent with local
rock layers oriented N30°E. Thirteen slab penetrations with 2" pipe
were necessary to deplete most of the source from the SE end of the
building because the negative pressure field around each penetration
was so small due to the very poor communication. Indoor radon,
which initially was measured as high as 120 pCi/1, was reduced to
less than 4 pCl/1.
9 -io
-------
Sub-slab
ventilation
system
FAN i— —
Sub-slab radon levels, pCi/1
^ Wall penetration radon levels, pCi/1
Figure 9. Footprint plan of a home shoving numerical value? of sub-slab
and blockvall radon concentrations that indicate that the radon
source Is following N30°E rock layers. Sub-slab ventilation
systems penetrations are shown as darkened circles.
Therefore, by knowing local geology and by tapping those high
radon potential source areas with a multi-duct, single fan, sub-slab
ventilation system opt law# radon reductions can be achieved in
buildings with poor sub-slab communication. Likewise by combining
geologic knowledge with sub-slab and blockwall radon measurements In
large buildings such as schools, the radon source can be located to
determine where to place sub-slab ventilation systems that will
achieve maximum radon reductions.
9-11
-------
REFERENCES
1. Hall, S. Correlation of soil radon availability number
with indoor radon and geology in Virginia and Maryland.
In: Proceedings of the EPA/USGS Soil Gas Meeting,
September 14-16, Washington, D.C.
2. Nazaroff, W. Moed, B., and Sextro, R. Soil as a source oŁ
Indoor radon: generation, migration, and entry, in: W.
Nazaroff and A. Nero (ed.), Radon and Its Decay Products
in Indoor Air. John Wiley and Sons, Nev York, N.Y., 1988.
pp. 82-90.
3. Tanner, A. E, Measurement <">Ł radon availability from
soil. In: Geologic Causes of Natural Radionuclide
Anomalies, Proceedings of the GEORAD Conference, St.
Louis MO, Apr. 21-22. 1987, MO Dept. of Nat. Resources,
Dlv. of Geology and Land Survey, Spec. Pub. No. 4. pp.
139-146.
9-12
-------
PRELIMINARY RADON POTENTIAL MAP
OP THE IINTTFD STATES
Linda C.S. Gundersen, R. Randall Schumann, James K. Otton,
Russell F. Dubiel, Douglass E. Owen, Kendell A. Dickinson
U.S. Geological Survey, MS 939 Federal Center
Denver, Colorado 80225-0046
R. Thomas Peake and Sharon J. Wirth
U.S. Environmental Protection Agency
Washington D.C. 24060
ABSTRACT
The geologic radon potential of the United States has been the subject of a one-year project by the U.S.
Geological Survey and the Environmental Protection Agency, in cooperation with the Association of American State
Geologists. Indoor radon data from the Stale/EPA Indoor Radon Survey and from other sources were compared with
bedrock and surficial geology, aerial radiometric data, soil properties, and soil and water radon studies. A numeric
radon index and confidence index have been developed as part of this project to quantify and standardize geologic radon
potential assessment on a regional scale. Publications from this study will be released in the fall of 1991 and will
include an annotated 1:7.5 million-scale map of the United States that delineates the radon potential of 55 geologic
provinces. Detailed radon potential books containing numerous indoor radon maps, geologic maps, and extensive
radon bibliographies for each state will also be available.
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
Congress directed the U.S. Environmental Protection Agency (EPA) to identify areas of the United States
with the potential lo produce potentially harmful levels of indoor radon (Indoor Radon Abatement Act of 1988,
Public Law 100-551). These characterizations were to be based on geological data and data on indoor radon levels in
homes and other structures. The information is to be used to define high radon potential areas for schools,
workplaces, and residences. The EPA was also directed to develop model standards and techniques for new building
construction that would provide adequate prevention or mitigation of radon entry. They also acknowledge that some
areas may not require any radon resistant features (1).
As part of an Interagency Agreement between the EPA and the U.S. Geological Survey (USGS), the USGS
is preparing radon potential estimates for the United States in order to help identify and prioritize areas toward which
states could target their resources and in which the different building code options would be most appropriate. The
U.S. Geological Survey is developing several products that will provide geologically-based characterizations of radon
potential at both regional and national scales. Booklets are being developed for each state that will present a general
overview of the geology and known or predicted correlations between geologic features and radon in soils. The
booklets will serve as a more detailed guideline for states than the national scale products, with data presented at the
county scale. In the booklets, some generalizations have been made in order to estimate the radon potential of each
county unit as a whole. Variations in geology, soil characteristics, climatic factors, and homeowner lifestyles can be
quite large within any particular county, therefore these reports cannot be used to estimate or predict the indoor radon
concentrations of individual homes or housing tracts. Indoor radon surveys used for the evaluations are also not
consistently statistical at the county scale.
A geologic radon potential map of the United States is also being produced, a preliminary version of which
is presented in this paper (Figure 1). The final published version of the map will be presented at a 1:7.5 million
scale and will include extensive text and annotation. This map illustrates the radon potential of the country by the
9-13
-------
Geologic Radon Potential of the United States of America
Radon Potential Ratings
Indoor Radon Average
( High >4 pCI/L
I I Moderate 2-4 pCI/L
~ Low <2 pCI/L
Km 700
Figure 1.
-------
major geologic provinces of relevance to the radon problem. It incorporates geologic evaluations of radioactivity,
soils, surficial materials, and bedrock lithology in the context of screening indoor radon measurements for the
country. The following sections describe the geologic environment of radon, the data and methodology used to
evaluate radon potential, and a brief summary of radon potential of the United States by region.
RADON FORMATION AND MIGRATION
Radon-222 (^Rn) is produced from the radioactive decay of radium-226 (226Ra), which is, in turn, a
product of the decay of uranium-238. Other isotopes of radon occur naturally, but are of less importance in terms of
indoor radon risk because of their short half-lives and less common occurrence. The exception to this is thoron
(radon-220), which occurs in high enough concentrations to be of concern in a few local areas. In general, the
concentration and mobility of radon in a soil are dependent on several factors, the most important of which are the
soil's radium content and distribution, porosity, permeability to gas movement, and moisture content. These
characteristics arc, in turn, determined by the character of the bedrock, glacial deposits, or transported sediments from
which the soil was derived, as well as the climate and the soil's age or maturity.
Radon transport in soils occurs by two processes, convective or advective flow and diffusion (2). Diffusion
is the dominant radon transport process in soils of low permeability (generally less than 10"7 cm^), whereas
convective transport processes tend to dominate in highly permeable soils (generally greater than 10"^ cm^) (3).
Radon transport distance is limited in low-permeability soils because of the short distance radon may travel during its
half life.
When radium decays in the soil, not all radon produced will be mobile. The portion of radium that can
potentially release radon into the pores and fractures of rocks and soils is called the emanating fraction. When a
radium atom decays to radon, the energy generated is strong enough to send the radon atom a distance of about 40
nanometers (nm, equal to 10"^ meters), or about 2xl(H> inches—this is known as alpha recoil (4). Depending on
where radium is distributed in the soil, much of the radon produced could end up imbedded in adjacent soil grains
rather than becoming mobile in the pore space between grains. Moisture greatly affects the recoil distance a radon
atom will travel. Because water is more dense than air, a radon atom will travel a shorter distance in a water-filled
pore than in an air-filled pore, therefore moisture lessens the chance of a recoiling radon atom becoming imbedded in
an adjacent grain. However, too much moisture can block soil pores and impede radon movement out of the soil.
Soil-gas radon concentrations can vary in response to climatic and weather variations on hourly, diurnal, or
seasonal time scales. Schumann and others (S) and Rose and others (6) recorded as much as order-of magnitude
variations in soil-gas radon concentrations between seasons in Colorado and Pennsylvania. The most important
factors appear to be precipitation (as it affects soil moisture conditions), barometric pressure, and temperature.
RADON ENTRY INTO BUILDINGS
A driving force (reduced atmospheric pressure in the house relative to the soil) and entry points must exist for
radon to enter a building from the soil. The negative pressure caused by furnace combustion, ventilation devices, and
the stack effect during cold winter months are common driving forces. Cracks and other penetrations through
building foundations, sump holes, and slab-to-foundation wall joints are common entry points.
Basement homes generally have higher average indoor radon levels than nonbasement homes because
basement homes provide more entry points for radon, commonly have a more pronounced stack effect, and typically
have lower air pressure relative to the surrounding soil than nonbasement homes. Elevated levels of indoor radon
occur in both basement and nonbasement homes, however. The term "nonbasement" applies to slab-on-grade or
crawl space construction. In many hones with basements, the main floors have radon levels similar to those on the
main floors of nonbasement homes, implying that occupants of basement homes are at greater risk from radon
exposure only if they spend a significant amount of time in their basements.
9-15
-------
METHODS AND SOURCES OF DATA
The assessment of geologic radon potential for the United States was made using five main types of data: (1)
geologic (lithok>gic), (2) radiometric, (3) soil characteristics, including soil moisture and permeability, (4) indoor
radon data, and (5) building architecture (specifically, whether homes in each area are built slab-on-grade or have a
crawl space versus homes with basements). These elements were integrated to produce estimates of radon potential.
Geologic data
Information on the type and distribution of lithologic units and other geologic features in an assessment area
is of primary importance. Rock types with naturally high uranium concentrations, usually greater than 2 parts per
million (ppm), that are most likely to cause indoor radon problems include carbonaceous black shales, glauconite-
bearing sandstones, some fluvial sandstones, phosphorites and phosphatic sediments, chalk, some carbonates, some
glacial deposits, bauxite, lignite, some coals, uranium-bearing granites and pegmatites, metamoiphic rocks of
granitic composition, felsic and alkalic volcanoclastic and pyroclaslic volcanic rocks, syenites and carbonatitcs, and
many sheared or faulted rocks. Rock types least likely to cause radon problems include marine quartz sands, non-
carbonaceous shales and siltstones, some clays and fluvial sediments, metamorphic and igneous rocks of mafic
composition, and mafic volcanic rocks. Exceptions exist within these general lithologic groups because of the
occurrence of localized uranium deposits. The most common sources of uranium and radium are the heavy minerals
such as zircon, titanite, and monazite, iron-oxide coatings on rock and soil grains, and organic materials in soils and
sediments. Less common are phosphate and carbonate complexes, and uranium minerals.
Although many cases of extreme indoor radon levels can be traced to high radium and (or) uranium
concentrations in bedrock and sediments, some structural features, most notably faults and shear zones, have been
identified as sites of localized uranium concentrations and have been associated with some of the highest known
indoor radon levels. Two of the highest known indoor radon occurrences in the United States are associated with
sheared fault zones in Boyertown, Pennsylvania (7), and in Clinton, New Jersey (8,9).
NURE AERIAL RADIOMETRIC DATA
Aerial radiometric data are used to describe the radioactivity of rocks and soils. Equivalent uranium (eU) data
provide an estimate of the surficial concentrations of radon parent materials (uranium, radium) in rocks and soils.
Equivalent uranium is calculated from the counts received by a gamma-ray detector in the wavelength corresponding
to bismuth-214 (^^Bi), with the assumption that uranium and its daughter products are in secular equilibrium. It is
expressed in units of parts per million (ppm) of uranium. Gamma radioactivity may also be expressed in terms of a
radium concentration; 3 ppm eU corresponds to approximately 1 picocurie per gram (pCi/g) of radium-226.
The aerial radiometric data used for assessing radon potential in this study were collected as part of the
National Uranium Resource Evaluation (NURE) program of the 1970s and early 1980s. The purpose of the NURE
program was to identify and describe areas in the United States having potential uranium resources (10). The NURE
aerial radiometric data were collected by aircraft in which a gamma-ray spectrometer was mounted, flying
approximately 122 m (400 ft) above the ground surface. Smoothing, filtering, recalibrating, and matching of
adjacent quadrangle data sets were performed to compensate for background, altitude, calibration, and other types of
errors and inconsistencies in the original data set (11).
Although radon is highly mobile in soil, and its concentration is affected by meteorologic conditions (4,12-
14), relatively good correlations between average soil-gas radon concentrations and average eU values for some soils
have been noted (7,15,16). The shallow (20-25 cm) depth of investigation of gamma-ray spectrometers, either
ground-based or airborne (17,18), suggests that gamma-ray data may sometimes provide an underestimate of radon
source strength in soils in which some of the radionuclides in the near-surface soil layers have been transported
downward through the soil profile or depleted by other processes. The redistribution of radionuclides in soil profiles
is dependent on a combination of climatic, geologic, and geochemical factors. Given sufficient understanding of the
factors involved, these regional differences may be predictable.
9-16
-------
SOIL SURVEY DATA
Soil surveys prepared by the U.S. Soil Conservation Service (SCS) provide data on soil characteristics. The
reports are commonly available in county formats, and occasionally in State summaries, and usually contain both
generalized and relatively detailed maps of soils in the area. Because of time and map-scale constraints, it was
impractical to examine county soil reports for each county in the United States, so more generalized summaries at
appropriate scales woe used where available. For State or regional-scale radon characterizations, soil maps arc
compared to geologic maps of the area, and the soil descriptions, shrink-swcll potential, depth to seasonal high water
table, permeability, and other relevant characteristics of each soil group noted. One of the best summaries of soil
engineering terms and the national distribution of technical soil types is the Soils Map of the National Atlas (U.S.
Geological Survey, National Atlas of the United States, sheet 38077-BE-NA-07M-00).
Uranium aid radium in soils are most often located on the surfaces of clays, with metal-oxides, especially
iron oxides, with calcium carbonate, and with organic matter. As soils form, they develop distinct layers, or
horizons, that are cumulatively called the soil profile. The A horizon is a surface horizon containing a relative
abundance of organic matter but dominated by mineral matter. The B horizon underlies the A horizon. Important
characteristics of B horizons include accumulation of clays, iron oxides, calcium carbonate or soluble salts, and
organic matter complexes. In drier climates, a horizon may exist within or below the B horizon that is dominated by
calcium carbonate, often called caliche or calcrete. This carbonate-cemented horizon is designated the K horizon.
The C horizon underlies the B and is a zone of weathered parent material, it is generally not a zone of leaching or
accumulation.
Soil permeability is typically expressed in SCS soil surveys in terms of the speed, in inches per hour (in/hr),
at which water soaks into the soil, as measured in a soil percolation test Although in/hr are not truly units of
permeability, these units are in widespread use and are referred to as "permeability" in SCS soil surveys. The
permeabilities listed in the SCS surveys are for water, but they are generally related to gas permeability.
Permeabilities greater than 6.0 in/hr may be considered high, and permeabilities less than 0.6 in/hr may be
considered low in terms of soil-gas transport.
Many well-developed soils contain a clay-rich B horizon that may impede vertical soil gas transport Radon
generated below this horizon cannot readily escape to the surface, so it would instead tend to move laterally,
especially under the influence of a negative pressure exerted by a building. Depth to seasonal high water table can
also be an important parameter to consider in some areas. Because water in soil pores inhibits gas transport, the
amount of radon available to a home is effectively reduced by a high water table. Areas likely to have high water
tables are river valleys, coastal areas, and some areas overlain by deposits of glacial origin (for example, loess).
Shrink-swell potential is an indicator of the abundance of smectitic (swelling) clays in a soil. Soils with a
high shrink-swell potential may cause building foundations to crack, and thus create pathways for radon entry into
the structure. In addition, swelling soils often crack as they dry; as a result, they provide additional pathways for
soil-gas transport and effectively increase the gas permeability of the soil (4).
INDOOR RADON DATA
Data from the EPA/State Indoor Radon Survey were used for the 34 States that have participated in this
program between 1986 and 1990 (19,20); for the remaining 16 States, data from the University of Pittsburgh's
Radon Project (21,22) and radon data from individual State indoor radon surveys were used where available.
The State/EPA Radon Survey
The State/EPA Radon Survey has provided statewide characteristics of indoor radon distributions since 1987.
To date, 34 States have participated in the program, with 7 additional States slated to complete the survey in 1991.
This survey has compiled short-term data results in approximately 47,000 homes, representing a total population of
over 21 million homes in more than 2000 counties in those participating States. Elevated screening levels have
been found in every State, with 1 in 5 homes surveyed falling above the Environmental Protection Agency's Action
Guideline of 4 pCi/L. The term "Action Guideline" is used as a recommendation to homeowners that remedial action
should be considered at or above this level.
9-17
-------
Each home surveyed in this program represents a much larger population of homes. The ability to achieve
such statistical validity is a result of the survey's design. To obtain the goal of creating a representative measure of
the distribution of radon levels in houses for major geographic areas within a State and for each State as a whole,
two levels of representativeness were established: first, that a representative sample of the houses in each State be
selected and, serond, that the measurements within each house be representative of actual concentrations. The
measurements involved short-term charcoal canister screening devices as listed in current EPA protocols (EPA-520/1-
89-009). This measurement strategy could be used to: a) quickly determine if an individual house has a potential
radon problem and, b) efficiently identify houses in a multiple-house survey that have elevated screening
concentrations. This survey design creates a database that can reflect large geographic areas of radon distribution,
characterizing States or regions within a State on a statistically valid level.
Table 1 is a summary of the EPA/State Survey data. The arithmetic means for the states range from a high
of 8.8 pCi/L in Iowa to a low of 0.1 pCi/L in Hawaii. In IS States, 20 percent or more of those homes measured
were above 4 pCi/L, whereas 10 States had 10 percent or less of those homes measured above the action guideline.
Although a Stale survey may reflect a low percentage of homes with elevated levels of radon, this does not always
suggest that an insignificant number of homes could be affected by elevated levels. In California, for example, only
2.4 percent of the homes measured were above 4 pCi/L. This small percentage is misleading, however, because it
represents an overall total of more than 247,000 homes that may be effected by elevated levels.
Table 1. Indoor radon data from the State/EPA Indoor Radon Survey
STATE
Average
% > 4 pCi/L RANK
Alaska
Alabama
Arizona
California
Colorado
Connecticut
Georgia
Hawaii
Iowa
Idaho
Indiana
Kansas
Kentucky
Louisiana
Maine
Massachusetts
Michigan
Minnesota
Missouri
North Carolina
North Dakota
Nebraska
New Mexico
Nevada
Ohio
Oklahoma
Pennsylvania
Rhode Island
South Carolina
Tennessee
Vermont
Wisconsin
West Virginia
Wyoming
1.7
1.8
1.6
0.9
5.2
2.9
1.8
0.1
8.8
3.5
3.7
3.1
2.7
0.5
4.1
3.4
2.1
4.8
2.6
1.4
7.0
5.5
3.1
2.0
4.3
1.1
7.7
3.2
1.1
2.7
2.5
3.4
2.6
3.6
7.7
6.4
6.5
2.4
41.5
18.5
7.5
0.4
71.1
19.3
28.5
22.5
17.1
0.8
29.9
22.7
11.7
45.4
17.0
6.7
60.7
53.5
21.8
10.2
29.0
3.3
40.5
20.6
3.7
15.8
15.9
26.6
15.7
26.2
25
29
28
32
5
17
26
34
1
16
9
13
18
33
7
12
23
4
19
27
2
3
14
24
8
31
6
15
30
21
20
10
22
11
9-18
-------
Vendor data from The Radon Project
Dr. Bernard Cohen of the University of Pittsburgh developed The Radon Project that has produced over
170,000 data points since 1986. The data collected from this radon project are not considered statistically valid due
to the utilization of volunteer data and other non-random selection methods, and therefore cannot be used to
accurately predict radon levels, although bias reduction was applied to the data set used in this report by eliminating
certain values based on responses to questionnaires (22). The data are useful because they provide information about
areas both covered and not covered by the State/EPA Radon Survey. The Radon Project collected data using charcoal
canister measurement devices, similar to that used in the State/EPA Survey.
Individual State Surveys
A number of states, including Delaware, Florida, Maryland, New Jersey, New York, New Hampshire,
Virginia, and Utah, have conducted their own indoor radon surveys and have collected extensive volunteer data.
These data arc compiled at several different scales and include zip-code, township, and county groupings. Some of
these surveys woe designed to be statistically valid while others may be biased toward high levels because of the
volunteered nature of the data.
radon index and confidence index
Many of the geologic methods used to evaluate an area for radon potential require subjective opinions based
on the professional judgement and experience of the individual geologist However, these evaluations are based on
established scientific principles that are universal to any geographic area or geologic setting. This section describes
the methods and conceptual framework used by the U.S. Geological Survey to evaluate areas for radon potential
based on the geologic factors discussed in the previous sections. The scheme is divided into two basic parts, a Radon
Index (Rl), used to rank the general radon potential of the area, and the Confidence Index (CI), used to express the
level of confidence in the prediction based on the quantity and quality of the data used to make the determination.
This scheme works best if the areas to be evaluated arc delineated by geologically-based boundaries (geologic
provinces) rather than political ones (State/county boundaries) in which the geology may be inconsistent across the
area.
Table 2 presents the Radon Index (Rl) matrix. Five main categories are evaluated and a point value of 1,2,
or 3 is assigned to each category. These categories were selected because die factors they represent are considered to
be of primary importance in controlling radon potential and because at least some data for these factors are
consistently available for every geologic province. Because each of these main factors encompass a wide variety of
complex and variable components, the subjective professional judgement and experience of the geologists performing
the evaluation arc heavily relied upon in assigning point values to each category.
Indoor radon is evaluated using unweighted arithmetic means of the combined (basement and first-floor)
indoor radon data for each county or for each geologic area to be evaluated. Other expressions of indoor radon levels
in an area could also be used, such as weighted averages or annual averages.
Aerial radioactivity data used in this report are from the equivalent uranium map of the conterminous United
States compiled from NURE aerial gamma-ray surveys (11).
The geology factor is complex and actually incorporates many geologic characteristics. In the matrix,
"positive" and "negative" refer to the presence or absence and distribution of rock types known to have high uranium
contents and to generate elevated radon in soils or indoors. Examples of "positive" rock types include granites, black
shales, and phosphatic rocks. Examples of "negative" rock types include quartz sands and some clays. The term
"variable" indicates that the geology within the region is variable or that the rock types in the area are known or
suspected to generate elevated radon in some areas but not in others due to compositional differences, climatic effects,
localized distribution of uranium, or other factors. Geologic information indicates not only how much uranium is
present in the rocks and soils but also gives clues for predicting general radon emanation and mobility characteristics
through additional factors such as structure (notably the presence of faults or shears) and geochemical characteristics.
9-19
-------
"Soil permeability" refers to several soil characteristics that influence radon concentration and mobility,
including soil type, grain size, structure, soil moisture, drainage, and permeability. In the matrix, "low" refers to
permeabilities less than about 0.6 in/hr, "high" corresponds to greater than about 6.0 in/hr, in SCS standard soil
percolation tests.
Architecture type refers to whether homes in the area have mostly basements, mosdy slab-on-grade
construction, or a mixture of the two. Split-level and crawl space homes fall into the "mixed" category.
To add additional weight to the geologic facta1 in cases where additional reinforcing or contradictory geologic
evidence is available, Geologic Field Evidence (GFE) points are added or subtracted from an area's score. Relevant
geologic field studies are important in enhancing our understanding of how geologic processes affect radon
distribution. In some cases, geologic models and supporting field data reinforce an already strong (high or low)
score; in others, they can provide important contradictory data. For example, areas of the Dakotas, Minnesota, and
Iowa that are covered with Wisconsin-age glacial deposits exhibit a low aerial radiometric signature and would score
only one point in that category. However, data from geologic field studies in North Dakota and Minnesota (23)
suggest that eU is a poor predictor of geologic radon potential in this area because radionuclides have been leached
from the upper soil layers but are present and possibly even concentrated in deeper soil horizons, generating
significant soil-gas radon. This positive supporting field evidence adds two GFE points to the score which help to
counteract the invalid conclusion suggested by the radiometric data. No GFE points are awarded if there are no
documented field studies for the area.
Confidence Index
Except for architecture type, the same factors are used to establish a Confidence Index (CI) for the radon
potential prediction fa- each area (Table 3). Architecture type is not included in the confidence index because house
construction data are readily and reliably available through surveys taken by agencies and industry groups including
the Bureau of the Census, National Association of Home Builders, and the Federal Housing Administration. The
remaining factors are scored on the basis of the quality and quantity of data used in the RI matrix.
Indoor radon data are evaluated on the distribution and number of data points and on whether the data is
consistent, randomly-sampled data (State/EPA Indoor Radon Survey or other Stale survey data) or volunteered vendor
data (likely to be nonrandom and biased toward population centers and/or high indoor radon levels). The categories
listed in the CI matrix for indoor radon data ("sparse or no data", "vendor data", and "State/EPA data") are intended as
representative examples of levels of sampling density and statistical robustness of an indoor radon data set.
Aerial radioactivity data are available for all but a few areas of the continental United States and for part of
Alaska. In general, the greatest problems with correlations among eU, geology, and soil-gas or indoor radon levels
appear to be associated with glacial deposits. Correlations among eU, geology, and radon are generally sound in
unglaciated areas and are usually assigned 3 CI points. Radioactivity data in some unglaciated areas may be assigned
fewer than 3 points, and in glaciated areas assigned only one point, if the data are considered questionable or coverage
is poor.
For the geologic data factor, a high confidence score is given to an area where a proven geologic model for
radon generation and mobility can be applied. Rocks for which the processes are less well known or for which data
are contradictory are regarded as "variable", and those about which little is known or for which no apparent
correlations have been found are deemed "questionable".
The soil permeability factor is also scored on quality and amount of data. Soil permeability can be
approximated from grain size and drainage class if data from standard, accepted soil percolation tests are unavailable.
Percolation test data and other measured permeability data are more accurate and score a higher confidence level.
Examples of radon potential ratings applied to different geologic terrains in the United States is given in
Table 4. Space does not permit including the rating tables for all of the United States.
9-20
-------
Table 2. Radon Index Matrix
INCREASING RADON POTENTIAL ^
POINT VALUE
FACTOR
1
2
3
INDOOR RADON (average)
< 2 pCi/L
2-4 pCi/L
> 4 pCi/L
AERIAL RADIOACTIVITY
< 1.5 com eU
1.5 - 2.5 Dpm
> 2.5 ppm
GEOLOGY*
negative
variable
positive
SOIL PERMEABILITY
low
moderate
high
ARCHITECTURE TYPE
mostly slab
mixed
mostly basement
^GEOLOGIC FIELD EVIDENCE (GFE) POINTS: GFE points are assigned in addition to points for the
"Geology" factor for specific, relevant geologic field studies. See text for details.
Geologic evidence supporting: HIGH radon +2 points
MODERATE +1 point
LOW -2 points
No relevant geologic field studies 0 points
SCORING:
Rn potential category
LOW
MODERATE
HIGH
POSSIBLE RANGE OF POINTS = 3 to 17
point range probable indoor Rn average
3-7 points
8-11 points
> 11 points
< 2 pCi/L
2-4 pCi/L
> 4 pCi/L
Table 3. Confidence Index Matrix
INCREASING CONFIDENCE
FACTOR
POINT VALUE
1
2
3
INDOOR RADON DATA
Sparsc/no data
Vendor data
State/EPA data
AERIAL RADIOACTIVITY
auestionable/no data
north of glacial limit
south of glacial limit
GEOLOGIC DATA
questionable
variable
proven geol. model
SOIL PERMEABILITY
auestionable/no data
variable
reliable, abundant
SCORING: LOW CONFIDENCE 4-6 points
MODERATE CONFIDENCE 7 - 9 points
HIGH CONFIDENCE 10-12 points
POSSIBLE RANGE OF POINTS = 4 to 12
9-21
-------
Table 4. Radon Index and Confidence Index Examples
Des Moines Superior Michigan
Lobe region Upland Basin
FACTOR
RI
CI
RI
ci
RI
CI
INDOOR RADON
3
3
2
3
2
3
RADIOACTIVITY
1
2
1
2
1
2
GEOLOGY
3
3
2
2
2
3
SOIL PERM.
1
2
2
2
2
2
ARCHITECTURE
3
-
3
-
3
-
GFE POINTS
+2
—
0
_
0
TOTAL
13
10
10
9
10
10
HIGH
HIGH
MOD
MOD
MOD
HIGH
Great
Northern Coast Ranges
Puget
Plains
and Cascade Mtns.
Lowlands
FACTOR
RI
CI
RI
CI
RI
CI
INDOOR RADON
2
3
1
2
1
2
RADIOACTIVITY
2
3
1
2
1
3
GEOLOGY
2
2
1
2
1
2
SOIL PERM.
2
2
2
2
2
3
ARCHITECTURE
3
-
1
-
1
-
GFE POINTS
0
—
0
_
0
_
TOTAL
U
10
6
8
6
10
MOD
HIGH
LOW
MOD
LOW
HIGH
Inner Coastal
Appalachian
Basin and
Plain
Carbonates
Range
FACTOR
RI
CI
RI
CI
RI
CI
INDOOR RADON
2
3
3
3
2
2
RADIOACTIVITY
2
3
2
2
2
3
GEOLOGY
2
3
3
3
2
2
SOIL PERM.
2
3
2
2
2
2
ARCHITECTURE
1
-
3
-
2
-
GFE POINTS
1
_
3
—
0
—
TOTAL
10
12
16
10
10
9
MOD HIGH HIGH HIGH MOD MOD
9-22
-------
RADON POTENTIAL IN THE UNITED STATES
Areas of die United States that are geologically similar can be grouped and delineated on a map. Each region
is characterized by a basic geology and climate that determines its radon potential. By examining and correlating
available geologic, aerial radiometric, soil radon, and indoor radon data, generalized estimates of die radon potential of
each region can be made. The following is a discussion of major geologic features and rock types and their known or
expected radon potential for each geologic/physiographic region. In each case, large-scale, well-known, or highly
anomalous features are discussed; this list is by no means exhaustive. Rather, it is intended to give the reader a
general feeling for the geologic features in each area that are likely to produce elevated indoor radon values, point out
important rock units or other geologic features where they are known, and act as a general guide for using geology to
predict radon potential on a regional scale. The numbered Regions designated in the text correspond with those in
Figure 1.
REGION 1,2, AND 3: THE OUTER AND INNER COASTAL PLAIN AND PHOSPHATIC AND LIMESTONE
DEPOSITS OF FLORIDA
The Coastal Plain of the eastern and southern United States consists of a systematic progression of
predominantly marine and fluvial sediments deposited during the evolution of the Atlantic and Gulf Coasts. The
oldest rocks exposed in the Coastal Plain are Cretaceous in age and consist predominantly of glauconitic sandstones,
chalks, and clays as well as some non-glauconitic quartz sandstones and fossiliferous limestone. These are overlain
by lower Tertiary (Paleocene, Eocene and Oligocene) sands and clays, which are often glauconitic, and upper Tertiary
(Miocene) fossiliferous chalks, clays, and thin sands. The youngest Tertiary sediments (Pliocene) are dominated by
gravelly sands, clayey sands, and thin clay beds. Because of the consistency in the general stratigraphy of the
Coastal Plain, many of the lithologic sequences are similar from state to state. Soil radon, surface radioactivity,
uranium and radium concentrations, permeability, and soil grain size distributions have been measured along more
than 1600 km of transects in five states underlain by Coastal Plain sediments (24,25). In general, the data suggest
that the Inner Coastal Plain, (Region 2), which is comprised of Cretaceous and Lower Tertiary sediments, has
higher radon potential than the Outer Coastal Plain (Region 1), which is comprised of Middle to Upper Tertiary and
Quaternary sediments. Grab samples of radon in soil gas collected at a depth of one meter averaged 700-1000 pCi/L
for the Coastal Plain as a whole. The two highest soil radon measurements were taken in Inner Coastal Plain
sediments; 16,226 pCi/L was measured in the glauconitic sands of the Nevasink Formation in New Jersey and 6333
pCi/1 was measured in the carbonaceous shales of the Eagle Ford Group in Texas. Radon in soil gas greater than
1000 pCi/L associated with phosphatic fossil layers and glauconitic sands and clays in the Aquia, Brightseat, and
Yorktown Formations from Maryland to Virginia have also been reported (26). Localized concentrations of uranium,
found in roll-front uranium deposits in Texas and in marine sands and and heavy mineral deposits from Virginia to
Georgia, have produced some locally high indoor radon occurrences. Heavy mineral deposits found throughout the
Coastal Plain also have the potential for creating scattered local anomalies and are a potential source of thoron as
well.
Comparisons with indoor radon data from the State/EPA Indoor Radon Survey (winter screening
measurements from 1986-89) and other data sources show good correlations among soil radon, radionuclide data, and
indoor radon data. The average for indoor radon concentrations is 1 pCi/L or less over different parts of the outer
Coastal Plain. Areas underlain by Cretaceous chalks, carbonaceous shales, phosphatic sediments, and glauconitic
sandstones of the Inner Coastal Plain average 2.3 pCi/L and have the highest radon potential.
Region 3 in Florida outlines the general extent of the uraniferous phosphatic deposits which cause abundant,
moderate radon problems and locally high radon problems. The geologic units thought responsible for these
problems include the Hawthorn Formation (27) and the Alachua and Bone Valley Formations. The southern part of
Region 3 in Dade County, Florida, may be moderate in radon potential due to the Key Largo Limestone.
REGIONS 4,5,6, AND 7: THE NORTHERN APPALACHIAN MOUNTAINS INCLUDING NEW ENGLAND,
THE TACONIC ADIRONDACK AND GREEN MOUNTAINS, AND NORTHERN APPALACHIAN PLATEAU
Region 4 is made up of Proterozoic and Paleozoic melamorphic and igneous rocks of moderate radon
potential, the main source of radon being uraniferous minerals and faults. Volcanic units in this region are low in
radon potential while the schists, gneisses and granites are dominantly felsic in composition and produce moderate
9-23
-------
radon concentrations. Glacial tills and gravels have compositions derived from both local and northern rock sources
and create locally high radon due to permeability and a moderate uranium source.
Region S is also an area of crystalline bedrock with locally derived glacial tills and gravels. The granites of
this area, particularly the Conway Granite, have very high uranium concentrations. Not only do the granites and
associated Proterozoic mctamorphic rocks cause high indoor radon but the highest radon in ground water, with
domestic well water concentrations in the 1 million pCi/L range (28) occur in the fractured granite aquifers.
Region 6 is a highly variable terrain that includes the metamorphic and igneous rocks of the Green
Mountains and the Paleozoic rocks of the northern Appalachian Mountains, Appalachian Plateau, and Taconic
Mountains. The Paleozoic rocks of the Appalachians includes highly deformed shales, sandstones, and carbonates.
Some of the carbonates and shales cause locally moderate to high indoor radon. Gravels and tills in the Albany area
also cause some local high radon.
Region 7, the Adirondacks, is a region of metamorphic and igneous rocks of contrasting radon potential.
The Marcy Anorthosite complex, which forms the ewe of the Adirondack Mountains, is low in radon potential,
whereas the metamorphic schists and gneisses that blanket the rim of the Adirondacks are locally high in radon
potential due to uraniferous minerals, uranium deposits associated with several of the magnetite deposits, and shear
zones. Gravels and tills cause locally high radon.
REGION 8: CENTRAL AND SOUTHERN APPALACHIAN MOUNTAINS, INCLUDING THE PIEDMONT,
BLUE RIDGE, AND VALLEY AND RIDGE.
The eastern part of the Appalachian mountains, known as the Piedmont and Blue Ridge, is underlain by
Proterozoic and Paleozoic-age metamorphic and igneous rocks. These rocks have moderate radon potential, with
localized areas of high potential. Studies thus far have yielded an average of 1000 pCi/L for rocks of granitic
composition and an average of 600 pCi/L for rocks of mafic composition. Over a thousand indoor and soil-gas radon
dara have been averaged for geologic units in the Appalachian region of PA, NJ, MD, and VA,and indicate that on
the average, the indoor radon concentration is approximately one percent of the soil radon concentration (29).
Permeability and emanating power are the main factors affecting this relationship.
Paleozoic-age sediments cover an extensive area of the western Appalachians, known as the Valley and Ridge,
and consist of sandstones, siltstones, shales, and carbonates. The carbonate soils, black shale soils, and black shale
bedrock can generate moderate to high levels of indoor radon. Carbonate soils derived from Cambrian-Ordovician
rock units of the Valley and Ridge Province cause known indoor radon problems in eastern Tennessee, western New
Jersey, western Virginia, eastern West Virginia (30) and central and eastern Pennsylvania The carbonate rocks
themselves are tow in uranium and radium; however, the soils developed on these rocks arc derived from the
dissolution of the CaC03 that makes up the majority of the rock. When the CaC03 has been dissolved away, the
soils are enriched in the remaining impurities, predominantly base metals, including uranium. Rinds containing
high concentrations of uranium and uranium minerals can be formed on the surfaces of rocks involved with CaC03
dissolution. Ground water derived from these areas, however, often contains radon concentrations of 1000 pCi/L or
less (31,32). Carbonates also form karst topography, characterized by solution cavities, sinkholes, and caves, which
increase the overall permeability of the rocks in these areas and may induce convective flow of radon.
In the Appalachians, the highest indoor, soil, and water radon values arc most often associated with faults and
fractures in the rock (28,33,34, 35).
REGION 10: THE NON-GLACIATED PORTION OF THE APPALACHIAN PLATEAU
The Appalachian Plateau Region contains areas of moderate, and some locally high radon potential. The
carbonate soils and shales found in the Paleozoic-age domes and basins characteristic of this part of the United States
have moderate to high radon potential. Of specific interest are the uranium-bearing Chattanooga shale in Kentucky
(Region 9) and Tennessee (36), the Devonian-Mississippian black shales in Ohio, Pennsylvania, New York, and
Indiana, and the Ordovician, Mississippian, Permian, and Pennsylvanian-age carbonates and black shales in Alabama,
Indiana, Tennessee, Kentucky, Michigan, Illinois, Missouri, Iowa, Arkansas, and Oklahoma.(37). Although
exposed in a limited area, Precambrian granites of the St Francis Mountains in southeastern Missouri are among the
most highly uraniferous igneous rocks in the United States (38). Granites, rhyolites and related dike rocks in the
9-24
-------
Wichita Mountains of Oklahoma have also been evaluated as having moderate to high radon potential (39). A large
area of low radon potential is underlain by the Pennsylvanian Pottsville Sandstone, which extends from eastern Ohio
through West Virginia, eastern Kentucky, east-central Tennessee, and northern Alabama. Moderate indoor radon is
also associated with uraniferous coal deposits in Pennsylvania and West Virginia.
REGION 9, 11, 12,13: NORTHERN GREAT PLAINS AND GREAT LAKES
The Northern Great Plains-Great Lakes region is underlain by Wisconsin and pre-Wisconsin-age glacial
deposits and loess. Much of North and South Dakota, western and southern Minnesota, and northern Iowa (Region
11) are underlain by deposits of the Des Moines lobe. Des Moines lobe tills are silty clays and clays derived from
the Pierre Shale and from Tertiary sandstones and shales which have relatively high concentrations of uranium and
high radon emanating power. Included within this region are clay and silt deposits of glacial lakes Agassiz, Dakota,
and Devil's Lake, which generate some of the highest radon levels in the area. Southwestern North Dakota is
underlain by unglaciated Tertiary sandstones, siltstones, and shales, some of which include uraniferous coals and
carbonaceous shales (Region 14).
In Region 9, glacial deposits in southern Wisconsin, northern Illinois, and western Indiana are primarily
from the Green Bay and Michigan lobes. These tills range from sandy to clayey and are derived mostly from
sandstones and carbonate rocks of southern Wisconsin and the Illinois Basin. Eastern Indiana and western Ohio are
underlain by tills derived from the Ohio and New Albany black shales. Black shales extend south of the glacial
limit, forming an arcuate pattern in northern Kentucky. They also underlie and provide source material for glacial
deposits in a roughly north-south pattern through central Ohio, including the Columbus area, and extend eastward
into southern New York. The overall radon potential of this area is high.
In Region 12, the Michigan Basin area includes silty and clayey tills in northern Michigan and surrounding
Lake Michigan. Source rocks for these tills are sandstones, shales, and carbonate rocks of the Michigan Basin which
are generally poor radon sources. Exposed crystalline rocks in the central part of the Upper Peninsula of Michigan
cause locally high indoor radon levels. This area has a moderate overall radon potential.
Hie Superior Upland of Region 13 includes glacial deposits of the Lake Superior lobe in northern
Minnesota and northern Wisconsin. The underlying source rocks for these tills are volcanic rocks and mafic
metamorphic and granitic rocks of the Canadian Shield that have relatively low uranium contents. The sandy tills
derived from these rocks have relatively high permeability, but because of their lower uranium content and lower
emanating power, they have a moderate radon potential. In central Wisconsin, uraniferous granites of the Wolf
River and Wausau plutons are exposed at the surface or covered by a thin layer of glacial deposits and cause some of
the highest indoor radon concentrations in the State.
Glaciated areas present special problems for assessment because bedrock material is often transported hundreds
of km from its source. Glaciers arc quite effective in redistributing uranium-rich rocks; for example, in Ohio,
uranium-bearing black shales have been disseminated over much of the western part of the state, now covering a
much larger area than their original outcrop pattern, and display a prominent radiometric high on the radioactivity
map of the United States. The physical, chemical, and drainage characteristics of soils formed from glacial deposits
vary according to source bedrock type and the glacial features on which they are formed. For example, soils formed
from outwash or ground moraine deposits tend to be more poorly drained and contain more fine-grained material than
soils formed on moraines or eskers, which are generally coarser and well-drained. In general, soils developed from
glacial deposits are poorly structured, poorly sorted, and poorly developed, but are generally moderately to highly
permeable and arc rapidly weathered, because the action of physical crushing and grinding of the rocks to form tills
may enhance and speed up soil weathering processes (40). Clayey tills, such as those underlying most of North
Dakota and a large part of Minnesota, have high emanation coefficients (41) and usually have low to moderate
permeability because they are mixed with coarser sediments. Tills consisting of mostly coarse material tend to
emanate less radon because larger grains have lower surface area-to-volume ratios, but because these soils have
generally high permeabilities, radon transport distances are generally longer, and structures built in these materials
arc able to draw soil air from a larger source volume, so moderately elevated indoor radon concentrations may be
achieved from comparatively lower-radioactivity soils (42,43).
9-25
-------
REGION 14 AND 15: THE UNGLACIATED GREAT PLAINS
The Great Plains extends from eastern Montana to central Texas. The area is mostly underlain by the
Cretaceous Pierre Shale and by Tertiary continental sandstones, siltstones, and shales including the White River,
Ogallala, and Arikaree Formations. The lower part of the Pierre Shale has an overall higher uranium content than
the upper part, and locally contains black shales. Members of the White River Formation are significant radon
producers in the northern and central Great Plains, whereas the Ogallala and Arikaree Formations are principal
sources for indoor radon in the central and southern part of the region from Colorado to west Texas. Carbonaceous
shales and uranium-bearing coals in the White River Formation of unglaciaied southwestern North Dakota generate
locally very high radon levels. Other Tertiary sedimentary units, including die Green River, Wasatch, and Fort
Union Formations and their equivalents, are also exposed in the area from Colorado to eastern Montana, but are of
less importance in terms of radon potential. Also included in this area are the Black Hills of southwestern South
Dakota, which are underlain by Precambrian granitic and metamorphic rocks and Paleozoic sedimentary rocks of
moderate radon potential. Overall, the Great Plains has a moderate radon potential.
The Sand Hills (Region IS), an area of windblown quartz sands in northern Nebraska, have a low radon
potential.
REGION 16: ROCKY MOUNTAINS AND PARTS OF THE WESTERN GREAT PLAINS
The Rocky Mountains have a similar high radon potential to the Appalachian region for many of the same
reasons. The metamorphic and igneous rocks in the Rocky Mountains are generally similar in composition, degree
of defamation, and intrusion by granites to those of the Appalachians. However, the Rocky Mountains have
undergone several periods of intense and widespread hydro thermal activity creating vein deposits of uranium which
cause local, high concentrations of indoor radon and radon in water in Colorado and Idaho (44-46). Colluvium and
alluvium derived from crystalline rocks of the Rocky Mountains covers much of the plains along the Front Range
from New Mexico to Canada and causes known moderate to high indoor radon problems in Colorado and Idaho (46,
47) .The Northern Rocky Mountains comprise the northeast and north-central part of Washington and northern and
central Idaho. This area is underlain by Precambrian sedimentary rocks. Paleozoic sedimentary rocks, and Mesozoic
metamorphic rocks, all intruded by Mesozoic and Tertiary granitic rocks. Hie largest intrusive body, the Idaho
Batholith, is a complex of granitic rock units ranging from diorite to granite. Uraniferous late Cretaceous to early
Tertiary granites occur throughout the Northern Rocky Mountains. An extensive, though dissected, veneer of
Tertiary volcanic rocks underlies much of the central Idaho portion of the Northern Rocky Mountains. Included in
Region 16 are the Permian marine limestones and other marine sediments of eastern New Mexico and the west Texas
Panhandle. These units, in particular the Cutler Formation, Sangre de Cristo Formation, and San Andreas
Limestone, have the potential to create moderate and locally high indoor radon problems (48). Also included in
Region 16 is an apron of Tertiary and Cretaceous sediments with high radon potential that contain local uranium
deposits and are overlain partly by colluvium from the Proterozoic rocks of the Rocky Mountains.
REGION 17: COLORADO PLATEAU, WYOMING BASIN
The Colorado Plateau and Wyoming Basin are underlain by sedimentary rocks ranging in age from
Pennsylvanian to Tertiary. The majority of the sedimentary uranium deposits of the United States are located in this
region and high indoor radon in Utah, Colorado, Wyoming and New Mexico appear to correspond to the uranium
deposits. Dominant rock types include arkosic conglomerates, marine limestones and shales, marginal marine
sandstones and shales, and fluvial and lacustrine sandstones, shales, and limestones. Uranium occurs to some extent
in mast of these rock types. The most significant uranium deposits occur in Mesozoic sediments, with the Jurassic
sandstones being the most common host for uranium ore. Localized sandstone-type uranium deposits are hosted in
particular by the Triassic Chinle, the Jurassic Morrison and the Cretaceous Dakota Formations in this region. Mine
tailings from these sedimentary deposits caused some of the earliest detected indoor radon problems (49). A small
part of the Colorado Plateau is included in Region 18 and is underlain by sediments with moderate rather than high
radon potential.
In the Wyoming Basin, the Permian Phosphoria Formation has moderate to high radon potential. It covers an
area of 350,000 km^ in southeastern Idaho, northeastern Utah, western Wyoming, mid southwestern Montana and
has a uranium content that varies from 0.001 to 0.65 percent. Other rocks with high radon potential in the
9-26
-------
Wyoming Basin are the Cretaceous Mancos Shale, which is uranifcrous in places, and Tertiary sandstones,
siltstones, and shale, which host uranium deposits and uranium bearing coals.
REGION 17, 18, 19 20: BASIN AND RANGE
The Basin and Range (regions 18,19,20) is comprised of Precambrian metamorphic rocks, late
Precambrian and Paleozoic metamorphosed and unmetamorphosed sedimentary rocks, Mesozoic and Tertiary intrusive
rocks, and Tertiary sedimentary and volcanic rocks. Hie region is structurally complex, with the aforementioned
rocks forming the mountain ranges and alluvium derived from the ranges filling the basins. The sedimentary rocks
include marine carbonates and shales, cherts, quartzites, and sandstones, as well as fluvial and continental sandstones,
siltstones, and shales. As with the Colorado Plateau, local uranium deposits occur throughout the sedimentary
rocks. Areas with moderate and locally high radon potential include the Tertiary volcanic rocks, particularly the
Miocene and Pliocene age rocks that arc found throughout the Basin and Range Province, Precambrian gneiss in
southern Nevada, and the Carson Valley alluvium, which is derived from the Sierra Nevada uraniferous granites. In
Utah, Sprinkel (SO) has indicated that the Wasatch fault zone and some geothermal areas have the potential to
produce elevated radon. The southern part of the Basin and Range (Region 19) has fewer Tertiary volcanic rocks and
is notably lower in radon potential.
The Snake River Plain in the northern part of Region 20 forms an arcuate depression in southern Idaho
underlain by basaltic volcanic rocks. Alluvium from adjacent mountains and luffaceous sedimentary rocks underlies
much of the upper Snake River Valley and the western end of the Snake River Plain. Those areas underlain by
basalt have low to locally moderate radon potential, however, those areas underlain by tuffaceous sedimentary rocks
and alluvium along the Snake River Valley have high overall radon potential. Overall, the area has a moderate radon
potential.
REGIONS 21, AND 22: SIERRA NEVADA, GREAT VALLEY, AND SOUTHERN COAST RANGES
The Sierra Nevada (Region 21) is underlain by Paleozoic and Mesozoic metamoiphic rocks with the
metamorphic rocks dominant in the northern part of the range and the granites dominant in the southern part of the
range. Tertiary volcanic rocks are also found in the northern part erf the range. The granites of the Sierra Nevada
Mountains are very high in uranium and have high radon potential as does die colluvium formed on the eastern and
western flanks of the mountains. The granite and colluvium are associated with high indoor radon in Nevada as well
as California.
The Southern Coast Ranges include the Franciscan Formation, a complex assemblage of metamorphosed
marine sedimentary rocks and ultramafic rocks, Cretaceous and Tertiary sedimentary rocks, and Mesozoic
metamorphic and igneous rocks. The Tertiary marine sediments and Mesozoic igneous and metamorphic rocks are
uranifcrous and have moderate indoor radon associated with them. In particular the Rincon shale in Santa Barbara
County may be the source for 75% of the homes having indoor radon greater than 4 pCi/L (SI).
The Great Valley is made up of alluvium and colluvium derived from both the Coastal Ranges and the
Sierra Nevada. Its radon potential is moderate overall but is controlled locally by source rock and permeability.
REGIONS 23,24,25: COLUMBIA PLATEAU.PUGET LOWLAND, CASCADE MOUNTAINS NORTHERN
COASTAL RANGES, KLAMATH MOUNTAINS, AND WILLAMETTE VALLEY
A comprehensive radon potential assessment of the Pacific Northwest has been done by Duval and others
(52). The Columbia Plateau (Region 23) is underlain principally by Miocene basaltic and andesitic volcanic rocks,
tuffaceous sedimentary rocks and tuff. The soils formed from these rocks are low in uranium concentration and
indoor radon is generally tow, giving the region an overall low radon potential An extensive veneer of Pleistocene
glaciofluvial outwash, eolian, and lacustrine deposits in the northern part of the Columbia Plateau (Region 24)
contains locally highly permeable soils and relatively high soil uranium levels and has moderate radon potential.
The subprovinces of the Blue Mountains and Joseph Upland in the central Columbia Plateau also include significant
outcrop areas of Jurassic and Triassic sedimentary and volcanic rocks, weakly metamorphosed in many areas, and
younger intrusive rocks which have a low to moderate radon potential.
9-27
-------
The Puget Lowland in the northern part of Region 23 is underlain almost entirely by glacial deposits and
Holocene alluvium. Most of the glacial and alluvial material of the Puget lowland is derived from the Cascades to
the east and the mountains of the Olympic peninsula to the west. The Puget Lowland overall has low radon
potential because of high soil moisture and low uranium content of soils. Most townships from Tacoma northward
average less than 1 pCi/L radon.
The Cascade Mountains (Region 23) extends from southwestern Oregon to northwestern California and can
be divided into two geologic (crranes: a northerly terrane composed principally of Mesozoic metamorphic rocks
intruded by Mesozoic and Tertiary granitic rocks and a southerly terrane composed of Tertiary and Holocene volcanic
rocks that form locally thick volcanic ash deposits east of the Cascade Mountains. Overall, the sparsely populated
Cascade Mountain Province has low radon potential because of the low uranium and high moisture contents of the
soils.
The Coastal Range Province (Region 23) extends from the Olympic Peninsula of Washington south to the
coastal parts of the Klamath Mountains in southwestern Oregon. In Washington, they are underlain principally by
Cretaceous and Tertiary continental and marine sedimentary rocks and pre-Miocene volcanic rocks. In Oregon, the
northern part of the Coastal Ranges are underlain principally by marine sedimentary rocks and mafic volcanic rocks
of Tertiary age. The southern part of the Coast Range is underlain by Tertiary estuarine and marine sedimentary
rocks much of them feldspathic and micaceous. The Klamath Mountains are dominated by Triassic to Jurassic
metamorphic, volcanic, and sedimentary rocks with some Cretaceous intrusive rocks. These metamorphic and
volcanic rocks are largely of mafic composition. Large masses of ultramafic rocks occur throughout the Klamath
area The radon potential of the Coastal Range Province is low overall. Most of the area has high rainfall and, as a
consequence, high soil moisture. Uranium in the soils is typically low. Highly permeable, excessively well drained
soils may cause locally elevated indoor radon levels.
River alluvium and river terraces underlie most of the Willamette River Valley (Region 25); however, many
of the hills that rise above the plains are underlain by Tertiary basalts and marine sediments. The Willamette River
Valley has moderate radon potential overall. Much of the area has somewhat elevated uranium present in soils and
many areas have excessively drained soils and soils with high emanating power. Many townships in the valley have
indoor radon averages between 2 and 4 pCi/L.
REGION 26: HAWAII
The volcanic island chain of Hawaii consists of Recent volcanic rock, predominantly basaltic lavas, ashes
and tuffs, with minor carbonate and clastic marine sediments, alluvium, colluvium, dune sands, and mudflow
deposits. Overall Hawaii has low radon potential. Although some soil radon is greater than 500 pCi/L (53), the
lifestyles of the inhabitants and local architecture contribute to the overall low radon potential of the islands.
REGION 27 AND 28: ALASKA
Alaska is divided from north to south into two main provinces: the Arctic Coastal Plain and the Northern
Foothills comprise one province (Region 27), and the the Arctic Mountains, the Central Province, and the Border
Ranges comprise the other (Region 28). The Arctic coastal plain province (North Slope) consists primarily of
Quaternary sedimentary rocks, mostly alluvium, glacial debris, and eolian sand and silL A belt of Tertiary
sedimentary rocks along the eastern 1/3 of the area separates the coastal plains from the foothills to the south. The
Foothills province is largely composed of marine and nonmarine Cretaceous sandstone and shale, much of which is
folded into westerly trending anticlines and synclines. This area has low radon potential.
The Arctic Mountains province is composed largely of faulted upper Precambrian and Paleozoic marine
sedimentary rocks. The Central Province consists mostly of Precambrian and Paleozoic metamorphic rocks,
Precambrian through Cretaceous mostly marine sedimentary rocks, Mesozoic intrusive and volcanic rocks. Tertiary
and Quaternary mafic volcanic rocks, flat-lying Tertiary basin-fill (nonmarine clastic rocks), and Quaternary surficial
deposits. The central province has several areas of uraniferous granites together with felsic intrusive and volcanic
rocks. The schist that produces high indoor radon near Fairbanks is in this area.
The Border Ranges area includes the Alaska-Aleutian subprovince, the Coastal Trough province, and the
Pacific Border Ranges Province. The area is composed of several mountain belts separated by a series of ^positional
9-28
-------
Cenozoic basins in a manner somewhat similar to that of the Basin and Range Province of the southwestern United
States. Rocks exposed in the area include Paleozoic mafic volcanic rocks; Mesozoic mafic volcanic flows and tuffs,
together with various units of shale, conglomerate, gray wacke, and slate; and Tertiary and Quaternary intermediate
volcanic rocks, Tertiary felsic intrusives, and Quaternary glacial deposits, eolian sand, and silt The Coastal Trough
province contains thick sequences of Tertiary continental clastic and volcanic rocks penetrated by Tertiary intrusive
rocks. Mesozoic sedimentary rocks and Pleistocene glacial deposits are abundant in some areas. Cretaceous and
Jurassic sedimentary and metamorphic rocks with interbedded mafic volcanic and intrusive rocks comprise most of
the Border Ranges rocks. A fairly large area of lower Tertiary sedimentary and volcanic rocks is found in the Prince
William Sound area. In much of this pan of Alaska, annual rainfall is hl#t (up to 170 inches), and water saturation
likely retards gas flow in soils on all but the steepest of slopes. The Arctic Mountains, Central Alaska, and Border
Ranges area have an overall moderate radon potential.
REFERENCES
1. Murane, D.M. Model standards and techniques for controlling radon levels within new buildings. In:
Proceedings of the 1990 International Symposium on Radon and Radon Reduction Technology, Vol. I:
Preprints. U.S. Environmental Protection Agency report EPA/600/9-90/005a, Paper A-I-2,1990, 5 p.
2. Tanner, A.B. Radon migration in the ground: a review. In: Adams. J.A.S. and Lowder. W.M.. eds. The
natural radiation environment Chicago, University of Chicago Press, p. 161-190,1964.
3. Sextro, R.G., Moed, B A., Nazaroff, W.W., Revzan, K.L., and Nero, A.V. Investigations of soil as a
source of indoor radon. Jm Hopke, P.K., ed., Radon and its decay products. American Chemical Society
Symposium Series 331, p.10-29, 1987.
4. Tanner, A.B. Radon migration in the ground: a supplementary review. Im GeseU, T.F., and Lowder,
W.M., eds, Natural Radiation Environment III. Symposium Proceedings 1:5-56,19%).
5. Schumann, R.R., Owen, D.E., and Asher-Bolinder, S. Weather factors affecting soil-gas radon
concentrations at a single site in the semiarid western U.S.. Im Osborne, M.C., and Harrison, J., eds.,
Proceedings of the 1988 EPA Symposium on Radon and Radon Reduction Technology, Volume 2, Poster
Presentations. U.S. Environmental Protection Agency Report EPA/600/9-89/006B, p. 3-1—3-13, 1989.
6. Rose, A.W., Washington, J.W., and Greeman, DJ. Variability of radon with depth and season in a central
Pennsylvania soil developed mi limestone. Northeastern Environmental Science 7:35-39, 1988.
7. Gundersen, L.C.S., Reimer, G.M., and Agard, S.S. Correlation between geology, radon in soil gas, and
indoor radon in the Reading Prong. Jm Marikos, M.A., and Hansman, R.H., eds., Geologic causes of
natural radionuclide anomalies. Missouri Department of Natural Resources Special Publication 4, p. 91 -
102, 1988.
8. Henry, M. E., Kaeding, M., Montverde, D. Radon in Soil Gas and Gamma Ray Activity Measurements at
Mulligan's Quarry, Clinton, New Jersey. Geological Society of America, Abstracts with Programs 21:22,
1989.
9. Muessig, K., and Bell, C. Use of Airborne Radiometric Data to Direct Testing for Elevated Indoor Radon.
Northeastern Environmental Science 7:45-51,1988.
10. U.S. Department of Energy. National Uranium Resource Evaluation preliminary report Prepared by the
U.S. Energy Research and Development Administration, Grand Junction, Colo., GJO-11(76), 1976.
11. Duval, J.S., Jones, WJ., Riggle, F.R., and Pitkin, J A. Equivalent uranium map of the conterminous
United States. U. S. Geological Survey Open-File Report 89-478,1989.
12. Kovach, E. M. Meteorological influences upon the radon content of soil gas. Transactions, American
Geophysical Union 26:241-248,1945.
9-29
-------
13. Klusman, R. W., and Jaacks, J. A. Environmental Influences Upon Mercury, Radon, and Helium
Concentrations in Soil Gases at a Site Near Denver, Colorado. Journal of Geochemical Exploralion
27:259-280, 1987.
14. Schery, S.D., Gaeddert, D.H., and Wilkening, M.H. Factors affecting exhalation of radon from a gravelly
sandy loam. Journal of Geophysical Research 89:7299-7309,1984.
15. Gundersen, L.C.S, Reimer, G.M., Wiggs, C.R., and Rice, C.A. Map showing radon potential of rocks
and soils in Montgomery County, Maryland. U.S. Geological Survey Miscellaneous Field Studies Map
MF-2043, scale 1:62,500,1988.
16. Schumann, R.R., and Owen, D.E. Relationships between geology, equivalent uranium concentration, and
radon in soil gas, Fairfax County, Virginia. U.S. Geological Survey Open-File Report 88-18, 1988,28
PP.
17. Duval, J.S., Cook, B.G., and Adams, J.A.S. Circle of investigation of an airborne gamma-ray
spectrometer. Journal of Geophysical Research 76:8466-8470,1971.
18. Durrance, E.M. Radioactivity in geology: Principles and applications. New York, N.Y., Wiley and Sons,
1986,441 p.
19. Ronca-Battista, M., Moon, M., Bergsten, J., White, S.B., Holt, N., and Alexander, B. Radon-222
concentrations in the United States—Results of sample surveys in five states. Radiation Protection
Dosimetry 24:307-312, 1988.
20. Dziuban, J.A., Clifford, M.A., White, S.B., Bergstein, J.W., and Alexander, B.V. Residential radon survey
of twenty-three States. M Proceedings of the 1990 International Symposium on Radon and Radon
Reduction Technology, Vol. Ill: Preprints. U.S. Environmental Protection Agency report EPA/600/9-
90/005c, Paper IV-2, 1990,17 p.
21. Cohen, B.L., and Gromicko, N. Variation of radon levels in U.S. homes with various factors. Journal of
the Air Pollution Control Association 38:129-134, 1988.
22. Cohen, B.L. Surveys of radon levels in homes by University of Pittsburgh Radon Project. In:
Proceedings of the 1990 International Symposium on Radon and Radon Reduction Technology, Vol. Ill:
Preprints: U.S. Environmental Protection Agency report EPA/600/9-90/005c, Paper IV-3,1990, 17 p.
23. Schumann, R.R., Peake, R.T., Schmidt, K.M., and Owen, D.E. Correlations of soil-gas and indoor radon
with geology in glacially derived soils of the northern Great Plains. Im Proceedings of the 1990
International Symposium on Radon and Radon Reduction Technology, Volume 3: Preprints. U.S.
Environmental Protection Agency report EPA/600/9-90/005c, 1990,14 p.
24. Peake, R. T., Gundersen, L. C. S., and Wiggs, C. R. The Coastal Plain of the eastern and southern United
States-An area of low radon potential. Geological Society of America, Abstracts with Progams 20:A337,
1988.
25. Peake, R. T., and Gundersen, L. C. S. An assessment of radon potential of the U.S. Coastal Plain.
Geological Society of America, Abstracts with Progams 21:58,1989.
26. Otton, J.K., Surficial uranium deposits in the eastern United States. Geological Society of America,
Abstracts with Programs 23:112, 1991.
27. Smith, D.L., and Hansen, J.K. Distribution of potentially elevated radon levels in Florida based on
surficial geology. Southeastern Geology 30:49-58,1989.
9-30
-------
28. Hall, F. R., Boudette, E. L., and Olszewski, W. J., Jr. Geologic controls and radon occurrence in New
England. Idi Graves, B., ed., Radon in Ground Water. Chelsea, Michigan, Lewis Publishers, 1987, p. 15-
30.
29. Gundersen, L.C.S. Predicting the occurrence of indoor radon: A geologic approach to a national problem.
EOS 70:280, 1989.
30. Schultz, A., and Brower, S.D. Geologic and environmental implications of high radon soil-gas
concentrations in carbonate rocks of the Appalachian Great Valley, West Virginia. Geological Society of
America, Abstracts with Programs 23:125,1991.
31. Gregg, L. T., and Coker, G. Geologic controls on radon occurrence in Georgia. Geological Society of
America, Abstracts with Programs 21:18,1989.
32. Schultz, A. P., and Wiggs, C. Preliminary results of a radon study across the Great Valley of West
Virginia. Geological Society of America, Abstracts with Programs 21:65,1989.
33. Smith, D. K. Uranium mineralogy. Jm DeVivo, B., Ippolito, F., Capaldi, G., and Simpson, P.R., eds,
Uranium geochemistry, mineralogy, geology, exploration and resources. The Institution of Mining and
Metallurgy, London, 1984, p. 43-88.
34. Gundersen, L.C.S. Anomalously high radon in shear /.ones, la: Osborne, M., and Harrison, J.,
symposium cochairmen, Proceedings of the 1988 Symposium on Radon and Radon Reduction Technology,
Volume 1, oral presentations. U.S. Environmental Protection Agency publication EPA/600/9-89/006A,
1989, p. 5-27 to 5-44.
35. Gates, A. E„ and Gundersen, L. C. S. The role of ductile shearing in the concentration of radon in the
Brookneal mylonite zone, Virginia. Geology 17:391-394, 1989.
36. Reesman, A. L. Geomorphic and geochemical enhancement of radon emission in middle Tenessee. In:
Marikos, M.A., and Hansman, R.H., eds, Geologic Causes of Natural Radionuclide Anomalies.
Proceedings of the GEORAD Conference, Missouri Department of Natural Resources, Special Publication
4, 1988, p. 119-130.
37. Coveney, R. M., Jr., Hilpman, P. L., Allen, A. V., and Glascock, M. D. Radionuclides in Pennsylvanian
black shales of the midwestem United States. In; Marikos, M.A., and Hansman, R.H., eds, Geologic
Causes of Natural Radionuclide Anomalies. Proceedings of the GEORAD Conference, Missouri
Department of Natural Resources, Special Publication 4,1988, p. 25-42,.
38. Kisvarsanyi, E. B. Radioactive HHP (high heat production) granites in the Precambrian terrane of
southeastern Missouri. Iq; Marikos, MA, and Hansman, R.H., eds, Geologic Causes of Natural
Radionuclide Anomalies. Proceedings of the GEORAD Conference, Missouri Department of Natural
Resources, Special Publication 4, 1988, p. 5-15.
39. Hood, R J-., Thomas, T.B., Suneson, N.H., and Luza, K.V. Radon potential map of Oklahoma.
Oklahoma Geological Survey Map GM-32, scale 1:750,000,1990.
40. Jenny, H. The clay content of the soil as related to climatic factors, particularly temperature. Soil Science
40:111-128, 1935.
41. Grasty, Rl. The relationship of geology and gamma-ray spectrometry to radon in homes. EOS 70: 496,
1989.
42. Schumann, R. R., Peake, R. T., Schmidt, K. M., and Owen, D. E. Correlations of soil-gas and indoor
radon with geology in glacially derived soils of the northern Great Plains. In; Proceedings of the 1990 EPA
9-31
-------
International Symposium on Radon and Radon Reduction Technology, Volume QI. Preprints: U.S.
Environmental Protection Agency report EPA/600/9-90/005c, paper VI-3,1990,14 p.
43. Kunz, C., Laymon, C. A., and Parker, C. Gravelly soils and indoor radon. In: Osborne, M., and
Harrison, J., symposium cochairmen, Proceedings of the 1988 Symposium on Radon and Radon Reduction
Technology, Volume 1, oral presentations. U.S. Environmental Protection Agency publication
EPA/60Q/9-89/006A, 1989,
p. 5-75-5-86.
44. Schumann, R. R.f Gundersen, L. C. S., Asher-Bolinder, S., and Owen, D. E. Anomalous radon levels in
crystalline rocks near Conifer, Colorado. Geological Society of America, Abstracts with Programs
21:A144-A145, 1989.
45. Lawrence, E. P., Wanty, R. B., and Briggs, P. H. Hydrologic and geochemical processes governing
distribution of U-238 series radionuclides in ground water near Conifer, CO. Geological Society of
America, Abstracts with Programs 21:A144,1989.
46. Ogden, A. E„ Welling, W. B., Funderburg, D„ and Boschult, L. C. A preliminary assessment of factors
affecting radon levels in Idaho. Jo; Graves, B., ed., Radon in Ground Water. Chelsea, Michigan, Lewis
Publishers, 1987, p. 83-96.
47. Otton, J. K., Schumann, R. R„ Owen, D. E., and Chleborad, A. F. Geologic assessments of radon
hazards: A Colorado case history. 1m Marikos, M.A., and Hansman, R.H., eds, Geologic Causes of
Natural Radionuclide Anomalies. Proceedings of the GEORAD Conference, Missouri Department of
Natural Resources, Special Publication 4, 1988, p. 167.
48. McLemore, V.T., and Hawley, J.W. Preliminary geologic evaluation of radon availability in New Mexico.
New Mexico Bureau of Mines and Mineral Resources, Open File Report 345, 1988, 31 p.
49. Spitz, H, B., Wrenn, M. E., and Cohen, N. Diurnal variation of radon measured indoors and outdoors in
Grand Junction, Colorado, and Teaneck, New Jersey, and the influence that ventilation has on the buildup of
radon indoors. Im Gesell, T. F., and Lowder, W. M., eds. The Natural radiation Environment III.
Springfield, VA, U. S. Department of Energy report CONF-780422, Vol. 2,1980, p. 1308-1329.
50. Sprinkel, D. A.. Assessing the Radon Hazard in Utah. Utah Geological and Mineral Survey, Survey Notes
22: 3-13, 1988.
51. D. Carlisle, written communication, 1991.
52. Duval, J.S., Otton, J.K., and Jones, WJ. Estimation of radon potential in the Pacific Northwest using
geological data, U.S. Department of Energy, Bonneville Power Administration report DOE/BP-1234,
1989.
53. Cotter, J., and Thomas, D.M. Ground gas radon response to meteorlogical perturbations. EOS 70:497,
1989.
54. Sterling, R., Meixel, G., Shen, L., Labs, K., and Bligh, T. Assessment of the energy savings potential of
building foundations research. Oak Ridge, Tenn., U.S. Department of Energy Report ORNL/SUB/84-
0024/1, 1985.
9-32
-------
TECHNOLOGICAL ENHANCEMENT OF HADON DAUGHTER
EXPOSURES DUE TO NON-NUCLEAR ENERGY ACTIVITIES
J. Kovac, D. Cesar arid A. Bauman
Institute for Medical Research and
Oc o u pa t i ona 1 Hea 1 th
University of Zagreb
Ksavcr 158, P.O.Box 291
41000 Zagreb, Yugoslavia
ABSTRACT
Natural radioactivity is a part of our natural surrounding and concentra-
tions of natural radionuclides in the environment increase with the development
of technologies. This is the case with phosphate ore processing in fertilizer
industry and during coal combustion in coal-fired power plants. A major source
of exposure to the population in the vicinity of non-nuclear" industries results
from inhalation of Rn-222 daughters. Exposure to radon daughters has been also
associated with lung disorders that include cancer' among workers. For that rea-
son the radon daughter concentrations in different atmospheres are discuseu in
thin paper.
Working levels were measured as ''grab samples" for several years at seve-
ral stations on-site and off-site of the coal-fired power plant an well as the
phosphate fertilizer plant, both located in Croatia. The average mean values
of working levels arc presented, and measurement techniques are reviewed.
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.
9-33
-------
INTRODUCTION
The exposure from 'man-made natural sources is called "technologicaly enhan-
ced natural radiation" (TENR)( 1). One of the first sourcesof ur-an! urn and thori-
um which was detected not being connected with the nuclear industry, was found
during energy production using fossil fuels.
Uranium is widely distributed in nature and is a minor contaminant in all
rocks, sand, and soil. Typical values for uranium are in the domain of 12 - bO
Bq/kg. Hence in ordinary back-yard soil there is of the order of 30 tons of
uranium and 10 g of radium per square mile to a depth of h ft. Each cubic yard
of ordinary soil or rook contains the order of JU kBq of radium. This radium
transforms at a constant rate Into its daughter product, radon (??-2Rn), and
maintains a constant activity of about 7Jt kBq of radon per cuniic yard of rock.
Because all rock and soil is slightly porous some radon diffuses out of any
exposed rock or1 soil surface. A typical value for the flow of radon from ordi-
nary surface soils into the atmosphere is 3.7/UBq/sec.cm2, or about 3.7 kBq/
-------
TABLE 1. DEFINITION OF THE "WORKING LEVEL" UNIT (WL)
U1timate
Alpha
Half-life
Number of
alpha energy
Total ultimate
Radionuclide
energy
atoms per
per atom
aloha energy
(MeV)
100 pCi
(MeV)
(MeV/100 pCi)
Ra-222
5.49
3.8 d
1,770,000
excluded
-
Po-218
6.00
3.05 m
977
6.00 + 7.68
0.134 x 105
Pb-214
-
26.8 m
8,580
7.68
0.659 x 105
Bi-214
-
19.7 m
6,310
7.68
0.485 x 105
Po-214
7.69
0.0027 m
0.0008
7.68
0.000 x 105
1.278 x 10D
or
1.3 x 105
Measurements of radon daughters can be converted to working levels by an
exact calculation if the state of daughter equilibrium is known. Several aut-
hors (6) have developed methods to determine the state of radon daughter equ-
ilibrium relative to Po-2l8, by alpha counting a filtered air sample. The most
widely applied measurement technique in the uranium mines is that of Tsivoglou,
than Kusnetz.
The Thomas-Tsivoglou method for calculation of radon daughter concentra-
tions is inconvenient for field use. The irregular counting times require
manual control of the scaler with consequent probabilities of error, and an
error renders the complete data set useless. The 30-min counting period limits
the processing rate to two samples an hour, so at least two scalers are requ-
ired if rapid changes in daughter concentrations are to be measured. With the
method developed by Scott (7) and our equipment it is possible to transfer a
filter from air pump to portable scaler within 40 sec, and next 15 sec is
ample time to note down the scaler reading and restart. Our procedure is there-
fore to take an air sample from 0 to 5 min, and then count the filter from 6
to 11 minutes (the M count), and from 11.25 to 16.25 min (the R count). These
are the only fixed counting times. The third 5-min count (K count) is made on
the filter at a time between 45 and 90 min. The rapid estimation of WL is:
R
where "R" is the total, number of alpha counts, "V" is the sample flow rate
(liters/min), and "E" is the counting efficiency. The value for the average
daughter ratio is 5539 counts, which is rounded to 5550 for convenience.
The radon monitor consists of an alpha scintillator (ZnS/Ag), photomulti-
plier tube, a light-tight outer housing for the detector with passive air entry
and an electronic package to convert the measured pulses to a digitally recor-
9-35
-------
ded signal, all battery operated for field use.
For estimating WLs, paralel with alpha counting we used for a long time a
single beta-counting of air sampler filters, using the method developed by
Iiolmgreen (8), based on the Eberline Air Particulate Monitor and total low-
level beta counting sistem, battery operated for field use. Since the method is
unjustly forgotten, here is a reminder of the basis for WL calculation.
The method for calculation of WLs from total beta activity concentrations
is based upon Table 1, using two simplifying assumptions;
1. Since at equilibrium Pb-214 and Bi-21U account for 90% of the total
ultimate alpha energy, a WL estimation based on Pb-21M and Bi-214 concentrati-
ons would approximate 90% of the actual value, so a factor "F" may be introdu-.
ced to compensate for the exclusion of the Po-218 contribution as a result of
counting only beta activity.
2. The radon daughter concentration ratios 1:0.65:0.35 (Po-2l8:Pb-214:
Ri-214) are employed in the model.
The ultimate energy assigned to an atom of Po-218 is "3.68 MeV, the energy
of its own alpha plus the alpha energy of Po-214, its great-granddaughter.
Also, Pb-214 and Bi-214, although only beta emitters, are assigned the alpha
energy of Po-211!, as Po-214 will ultimately be produced from either of these
atoms. The energy contribution of Po-21*4 present in the 1 litre volume is near-
ly zero, because of the small population of the extremely short-lived Po~21«
atoms. Equation |2| defines the WL unit:
(13.68MeV/atom,) (N,) + (7.68MeV/aton„ r) (N,,+Nr,)
I TT A A ^4" U 1) V r~\ ,
WL = = \?)
1.3 x WJ MeV/Wl
where is number of atoms of Po-218, "N^" number of atoms of Pb-214, and
"N^" is * number of atoms of Bi-214. The numbers of atoms of each daughter can
beudetennined from Table 1. Substitution of numbers of atoms of each daughter
into equation [2| yields:
WL = 0.001028(pCiA/l.) + 0.005069(pCiB/l.) + 0.003728(pCic/l.) (3)
Based upon two assumptions given above, equation (3| may be modified to become:
WL = F|0.005069(pCig/l.) + 0.003728{pCic/l.)| (l|)
Also: pCig/l. = 0.65 Cg (5)
and pCic/l. = 0.35 Ca (6)
where C is the total measured beta activity concentration (pCi/1.). Substitu-
tion into equation |4| of equations |4| and |5| and factoring, and taking into
account that parameter "F" has an empirically determined value of 1.25, substi-
tution into equation |4| gives:
WL = C (0.00575)
3.
9-36
(7)
-------
In all our measurements we used glass fiber filters (General Electric),
even we tried with molecular filters, but they were not convinient in very
dusty atmosphere.
WL IN COAL-FIRED POWER PLANT
As the combustion of coal increases, so will the magnitude of environmen-
tal and human health hazards associated with trace elements and radionuclides
mobilized by the coal fuel cycle. The large fraction of coal ash that does not
find a commercial application is usually dumped in the vicinity of the coal-
fired power plant (CFPP). When the dumping is finished, most dry ash dumps are
covered by topsoil and converted into areas for agricultural or recreational
use, but not yet in Yugoslavia (9).
The coal ash may contain enhanced levels of the natural radionuclides in
the uranium and series, especially fly ash. Among the decay products are the
racon isotopes, Rn-222, Rn-?20 and Rn-219, which are noble gases and thereby
pose special problems in assessing the radiological hazard of fly ash. The
fractional amount of radon lost from the parent-containing material is called
the emanation coefficient or emanating power. It is important to stress the
difference between radon which escapes the physical confines of the parent-
containing material (emanation) and that which occurs in a gas atmosphere which
may be sampled (emanation + diffusion). Beck measured the emenation coeffici-
ents of coal ash obtained from three different, power plants (10). For all sam-
ples he studied, the emenation coefficients were less than 0.1. Gamma radiation
from a tailings dump is in general not a serious problem. Radiation levels 1 m
from the pile surface tend to be less than 0.01 mGy/h and average around 0.005
mGy/h though "hot spots" with much higher dose rates have been reported (9).
As with radon emenation, higher surface dose rates are to be expected over the
tailings from higher grace coal, such as in the investigated case.
For all that reasons, investigations of the hazards were undertaken in the
CFPP in Croatia, because the anthracite coal used for combustion has an average
10% sulphur and a variation of uranium. In the seventies the uranium content in
coal was between 500 - 1200 Bq/kg. After 1980 it declined to an average 250
Bq/kg due to opening of an different vein in the coal mine. This requested a
thorough monitoring programme which included measurements of activity concen-
tration of radionuclides in coal and ash samples, and measurements of WL. First
measurements of WL were carried out at 1977. In the CFPP seven locations have
been chosen, because of long-time occupational exposure, and five on-site in
places with natural air flow. Measurements have been repeated in 1983, when
CFPP used coal with lower uranium content. In 1977 we used only Holmgreer/s
method, and in 1983 we used both, Holmgreen's and Scott/s method. Tables 2 and
3 summarize the estimated WL values, together with occupancy time limit.
9-37
-------
TABLE 2. WL OF OCCUPATIONALLY EXPOSED PERSONS INSIDE THE CFPP
Work place
mWL *
(1977)
Occupancy
time limit
mWL
(1983)
Occupancy
time limit
1. Conveyour belt (coal)
8.0
42 h/week**
unlimited
7.0
42 h/week
unlimited
2. Conveyour belt (coal)
15.0
24-42 h/week
6.0
42 h/week
unlimited
3- Below the automatic
control (ash hooper)
30.0
21 h/week
12.0
24-42 h/week
4. Below the automatic
control (ash hooper)
60.0
42 h/week
12.0
24-42 h/week
5. Waste pile fresh
30.0
21 h/week
-
-
6. Waste pile old
-
-
60.0
J*2 h/week
7. Bottom ash
RO.O
21 h/week
20.0
24-42 h/week
TABLE 3- WL ON-SITE IN PLACES
WITH NAi
URAL AIR FLOW
Work place
mWL*
(1977)
Occupancy
time 1 imit
mWL
(1983)
Occupancy
time limit
1. Area around the steam
generator building
6.0
unlimited
6.0
unlimited
2. Under the stack
5.0
unlimited
6.0
unlimited
3. Near the furr.ice
5.0
unlimited
6.0
unlimited
4. Office building
(500 m from the CFPP)
3-0
unlimited
5. 10 km from the CFPP
3.0
unlimited
6.0
unlimited
* -3
raWL = 1 x 10 WL. All WL values are an arithmetic mean of 3 measurements.
**
42 ft/week was taken as the occupancy time limit to comply with the US
general population standards, since the workers in the CFP? were never
considered as people occupationally exposed to radiation.
The WLs have shown great variations between two measurements depending on
the radioactivity of the coal and combustion products present at the time of
the measurements in the CFPP. Places on-site with good ventilation had 3-6
raWL. The highest WL was besides the bottom ash and fresh waste pile where even
an occupancy time limit should be considered. The values for the WL are chan-
ging, so that the new data are lower than these presented in Table 2 and 3.
Tabic 4 summarizes the estimated WL values measured in 1990, when we used only
Scott's method.
9-38
-------
TABLE 4. WL MEASURED ON-SITE AND OFF-SITE CFPP IN 1990.
Location
mWL
1. Coal storehouse
6.0
2. Below the automatic
control (ash hooper)
11.0
3. Area around the steam
generator building
6.0
4. Slag and ash pile
6.0
5. Strmac
6.0
6. Vozilici
5.0
7. Stepcici
5.0
8. Luka Plomin
4.0
9. Rabac
3-0
There were no differences in WLs between measurements done by one or the
other method. As we expected, the highest values were obtained on-site of the
CFPP. Locations 5-9 were at different directions and distances from the CFPP,
chosen in dependency on the wind-rose (Table 5).
TABLE 5. ALTITUDES, DISTANCES AND DIRECTIONS FROM THE CFPP
Location
Altitude (n)
Distance (km)
Direction
Strmac
120
3
SW
Vozilici
100
5
NW
Stepcici
80
2
W
Luka Plomin
10
1
SE
Rabac
0
20
S
The most interesting case is the location Strmac, where a hamlet was built
on a ninety years old tailing site, where already the second and even the third
generation of same families are dweling in the same houses.
At the location Rabac, which is at the sea shore the WL is slightly lower,
since the radon levels over the sea and the ocean are much lower than over the
land, due to the lower Ra-226 content of the sea. For this reason, radon levels
in the atmosphere at coastal sites are very dependent on whether the wind is
blowing from the land or the sea.
9-39
-------
WL IN FERTILIZER INDUSTRY
Three years after the beguining of the WI. measurements at the CFPP, the
same type of investigations has started in a fertilizer plant.
The activity mass concentrations of natural radionuclides in phosphate ore
for a given radionuclide and type of fertilizer vary markedly from one country
to another, depending on the origin of the components. General features are that
the activity mass concentrations of K-40 and Th-232 and its decay products are
always low and that the activity mass concentrations of the radionuclides of the
U-238 decay series are 5 - t>0 times higher than in normal soil. The degree of
radioactive equilibrium between U-238 and its decay products in a given type of
fertilizer depends essentially on the relative contribution of phosphoric acid,
since phosphoric acid usually has a very low Ra-226 concentration. For the pur-
pose of this, it is assumed that Th-230 and U-234 are in radioactive equilibrium
with U-238 and that Pb-210 and Po-210 are in radioactive equilibrium with Ra-226.
A typical concentrations of U-238 and Ra-2?6 in sedimentary phosphate depo-
sits are 1500 Bq/kg, which are generally found to be in radioactive equilibrium.
When these rocks are processed into fertilizer most of the uranium and some of
the radium accompanies the fertilizer, and than in the fields through crops
enters the food chain.
In the production of fertilizers, phosphate rocks are used in two different
ways. The first method, the acidulation of phosphate rocks was ensured by sul-
phuric acid, where phosphoric acid and gypsum result as normal superphosphate.
The second method, where the phosphate rock is treated by nitric acid, the final
product is phosphoric acid and gypsum as residue, which contains most of the
radium (11).
Almost all of Ra-226 originally in the phosphate ore is discharged in the
piles. The concentration of Ra-226 in piles is about TOO Bq/kg. Since the rate
of radon production equals the rate of radium decay, the rate of radon produc-
tion can be readily calculated. The answer is, 1 g of Ra-226 (this is also 1 Ci
or 37 GBq of Ra-226) produces 7H kBq/sec of Rn-222. Thus the radon production
rate in piles containing 700 Bq/kg od Ra-226 is 1.4 mBq/kg/sec. The density of
dry piles is about 0.7 g/cm3, which means that the production rate of radon per
unit volume is about 1 m3q/m3/sec.
The highest occupational radiation exposure during the process are to be
expected in the fertilizer production or in the fertilizer storehouse. To check
the level of radiation dose, a monitoring programme was introduced, including
the determination of specific activities of natural radionuclides in ambient
air, phosphate ore, phosphate fertilizers, waste products, trickling and well
waters. Measurements of WLs were carried out at ten locations, twice a year for
the last ten years. Five of them were inside the phosphate fertilizer plant,
one on the gypsum's pile. The off-site locations were at four different direc-
tions and distances, chosen on the basis of the wind-rose. Results are presented
in Table 6.
9-40
-------
TABLE 6. WL MEASURED ON-SITE AND OFF-SITE THE FERTILIZER PLANT
Location
mWL
1. Phosphate ore storehouse
12.0
2. KC1 storehouse
4.6
3- Fertilizer package store (NPK)
21.0
4. Inside the fertilizer production
9-4
5. Phosphoric acid production
4.4
6. Gypsum's pile
3.0
7. Off-site locations
1.2
All values are an arithmetic mean of ten years measurements performed in
summer and winter, always three times on each location. During the first year
only beta measurements (Holmgreen) were done, and later only alpha measure-
ments (Scott)(7,8). There were no significant differences observed during the
years. For the comparison in Table 7 one year data are presented measured once
by alpha and once by beta measurement.
TABLE 7. WL MESURED BY DIFFERENT METHODS
Location
Holmgreen
mWL
Scott
1. Phosphate ore storehouse
3-1
3-2
2. KC1 storehouse
2.5
1.1
3. Fertilizer package store (NPK)
3-5
4.2
4. Off-site locations
1.4
1.2
The WL rate differs slightly not because of different measuring methods,
but also due to different phosphate ore origin.
CONCLUSION
This paper introduces WL measurements in industries where TENR is present.
The CFPP is a specific case with the appearance of natural radioactivity which
was very similar to open pit uranium mining, where WL measurements are routi-
nelly done. For that reaso WL measurements were applied also in this case. When
some places of occupational exposure in the CFPP were detected, the authors
have tried to find out if the same problem also exists in the fertilizer indus-
try. The appearence of places with an increase of natural radioactivity in non-
nuclear industries have left the legislator, at present without a ready soluti-
9-41
-------
on in Yugoslavia, how to systematize occupationally exposed workers, especially
after the Chernobyl accident, when the public become sensitive to radiation of
any origin.
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.
RFFF RF.NCFS
1. Gesell, ?.F. and Pritchard, H.M. The technologically enhanced natural radi-
ation environment. Health Physics. 28: 361, 1976.
2. Fry, R.M. Radiation hazards in uranium mining and milling. Atomic Energy in
Australia. 18(4); 1, 1975.
3. Costa-Ribeiro, C., Thomas, J., Drew, R.T., Wrenn, M.E. and Fisenbud, M.
A radon detector suitable for personnel or aera monitoring. Paper presented
at the Uranium Mining Health Physics Workshop, Denver, Colorado. June
21-22, 1970.
4. Lasseur, C. Apparatus for selective continuous measurements of each solid
short-lived daughter products of radon. Paper presented at the Journess
d^Slectronique do Toulouse, France. March 4-8, 1968.
5. Evaluation of occupational and environmental exposures to radon and radon
daughters in the united states. MCRP Report No. 78. National Council on
Radiation Protection and Measurements, Bethesda, Md, 1985. 83 pp.
6. Manual on radiological safety in uranium and thorium mines and mills. IAEA.
Safety series No. 43. International Atomic Energy Agency, Vienna, Austria,
1976. 16 pp.
7- Scott, A.G. A field method for measurement of radon daughters in air.Health
Physics. 41: 403, 1981.
8. Holmgreen, R.M. Working levels of radon daughters in air determined from
measurements of RaB + RaC. Health Physics. 27: 141, 1974.
9. Bauman, A., Horvat, D., Kovac, J. and Lokobauer, N. Technologically enhanced
natural radioactivity in a coal-fired power plant. In: K.G. Vohra, K.C.
Pillai, U.C. Mishra and S. Sadasivan (ed.), Natural Radiation Environment.
Wiley Eastern Limited, Bombay , 1982. p. 401.
10. Beck, H.L., Gogolak, C.V., Miller1, K.M. and Lowder, W.M. Perturbations on
the natural radiation environment due to the utilization of the coal as an
energy source. Paper presented at the DOE/UT Symposium on the Natural Radi-
ation Environment III, Houston, Tx. 1978.
11. Kovac, J., Cesar, D. and Bauman, A. Natural radioactivity in a phosphate
fertilizer plant. In: Proceedings of the XIV Regional Congress of IRPA.
Current Problems and Concerns in the Field of Radiation Protection.
Yugoslav-Austrian-Hungarian Radiation Protection Meeting, Kupari, Yugosla-
via, 1987. p. 69.
9-42
-------
A SITE STUDY OF SOIL CHARACTERISTICS AND SOIL GAS RADON
Richard Lively, Minnesota Geological Survey, 2642
University Ave. St. Paul, MN 55114, and Daniel Steck,
Dept. of Physics, St. John's University, Collegeville,
MN 56321
ABSTRACT
In regional surveys, indoor radon is usually the parameter of
interest, but occasionally soil gas radon at depths of 1 meter or
less is also measured. At statewide scales, even limited data sets
can be used to infer relationships between geology and soil gas or
indoor radon. However, predicting the radon potential of a single
house or even an area the size of a neighborhood is more
difficult. As the size of a surveyed area decreases, site-specific
variables become more significant.
We recently completed a study of two residential
neighborhoods within 7 kilometers of each other near Rochester,
Minnesota. Eight holes were augered into glacial sediments to
maximum depths of 4.5 meters and samples collected for grain-size
analysis, measurement of radon parent/daughter nuclides and radon
emanation. A total of 65 homes in the areas were provided with two
alpha-track registration detectors for 1 year of indoor
monitoring.
Positive correlations were observed between the average soil
radon, the average indoor radon, and the precursor/daughter
radionuclides. The study area with the most topographic relief
also had the highest radionuclide contents, the most variability
with depth, and some variation with time and soil moisture; these
results were not observed at the low-relief site. The type of
study described would best be applied to site-specific
preconstruction screening, rather than to predicting radon in
existing structures.
9-43
-------
INTRODUCTION
This project was designed to collect data on soil type and soil
characteristics, radon and other related nuclides at several
depths, and porosity and permeability. At the same time, radon
levels in basements and living areas of homes built on the soils
were also measured.
Two areas were chosen for the pilot study (Figure 1). St. Marys
Hills on the west side of Rochester consists of modern, single-
family homes on 1/2-acre to 2-acre lots on the west side of a
bedrock hill composed of St. Peter Sandstone, Decorah Shale, and
Galena limestone, with a total vertical relief of about 40 meters.
The bedrock surface is covered by 2 and 6 meters of glacial
sediment and loess. Essex Park, about 6.5 kilometers northeast of
St. Marys Hills is a mix of modern, single-family and multiple-
residence homes on 1/2-acre lots. The topography is subdued, with
about 9 meters of relief. Depth to the bedrock (Prairie du Chien
Group) is between 3 and 18 meters. The profiles for each area and
locations of the sample holes are shown in Figure 2.
Sixty-four owners of single-family homes participated in the
study, 45 from Essex Park and 19 from St. Marys Hills. Each
received two radon detectors, one for the basement and one for a
first-floor living area. Exposures lasted from 9 to 12 months.
METHODOLOGY
The test holes were drilled in October 1988 using a truck-
mounted Giddings soil auger with a 5-cm-diameter bit and core
tube. Sediment samples collected during drilling were placed in
sealable plastic bags.
The following is a summary of the analyses and methods used to
study the sediment samples.
1. Moisture content and bulk density: the wet weights were
measured within 2 days after collection. Soils were dried for a
minimum of 24 hours at 70°C and reweighed. The results are only
approximate, because they do not reflect moisture lost prior to
measurement.
2. Solid particle density: these measurements were based on a
procedure from Luetzelschwab and others (1). These results
combined with the wet and dry bulk densities can be used to
approximate the pore volume in a sample of soil.
3. Grain-size fractions: the soils were screened into
fractions consisting of a bulk sample (undifferentiated as to
grain size), >149 |1 (sand and gravel), 149-63 [1 (very fine sand),
and <63 jl (silt and clay by wet sieving) .
4. Mineralogy: the mineralogy was determined by examining the
>14 9 Jl grain-size fraction with a binocular microscope.
5. Radon: radon emanation was measured from the bulk soil
samples, the <63 Jl, and the 63-149 Jl fractions using a charcoal
trap system modified from an unpublished report by Dr. J.N.
9-44
-------
Andrews, University of Bath, England. Scatter in the bulk fraction
is thought to result from inhomogeneous radium in the sediment.
The reproducibility of the other duplicate analyses was very good,
and replicate analyses of radium standards varied by less than
10% .
6. 2l0Po — 210P}-,. polonium-210 was extracted from the sediment
with a leaching technique modified from Eakins and Morrison (2),
Blake and Norton (unpub.), and D.R. Engstrom (unpub.). The 210po
was assumed to be in radioactive equilibrium with the 21®Pb.
7. Radium and thorium: 1-kilogram sediment splits from each
depth were analyzed for 226Ra and 232Th by gamma-ray spectroscopy
using a high-resolution germanium detector. The measured
activities reflect total radium and thorium in the sediments.
8. Radon concentrations in the soil at multiple depths were
measured by (1) pumping air from isolated intervals through a
liquid scintillation cocktail (active sampling) and by (2)
extended monitoring of isolated intervals with alpha-track
detectors (passive sampling). Inflatable rubber packers on the
outside of hollow PVC pipe were used to isolate each collection
point. Each alpha track detector was wrapped in Saran Wrap® to
keep out water vapor but still allow diffusion of radon. Initial
data from alpha-track detectors is not included in the tables
because of large variance in the calibration constant for the
detectors used at that time and our doubts about the integrity of
the original packers. In 1989, a redesigned system for both the
active and passive sampling was used with more reliable packers
and flexible barriers, which prevented vertical air movement if a
packer failed.
SOIL CHARACTERISTICS
The sediments within the Rochester area are the result of
glacial processes and include tills, outwash, colluvium, and
loess. Loess, ranging from 0.6 to 2.8 meters thick, covers all of
the sample sites except Hole B in St. Marys Hills. The glacial
tills below the loess are oxidized; some show the reddish-brown
colors of ferric iron to depths of about 4.3 meters (Figure 3).
Moisture ranged from a low of 6.8 weight percent in the loess
to a high of 20 weight percent, also in loess. Soil moisture
increased slightly with depth, but not in all holes and not more
than a few percent. Between the initial sampling in 1988 and
measurement of radon in 1989, Hole B (St. Marys Hills) collected
water in the bottom meter. This could have been due to seepage
from the sediment or water infiltrating from the surface.
There is a fairly broad range of grain-size distributions in
the sediment samples, but the means within and between the sites
were not statistically different. The available data do not allow
us to distinguish between the relative effects of deposition and
post-depositional soil development on the grain-size distribution.
9-45
-------
Mineralogically the sediments are very similar, being
predominantly composed of quartz, feldspar, biotite, and
muscovite. Rock fragments form up to 50% of the >149 fl size
fraction and include granite, limestone, quartzite, sandstone, and
metamorphic and volcanic rocks. Varying percentages of magnetite,
pyrite, hematite, and limonite were also observed.
RADIOMETRIC RESULTS
Radon concentrations in the soil gas at St. Marys Hills
generally increase with depth and range from 17 to 71 kBq/m3 (Table
1), with an average of 44 ± 13. In Hole A, both active and passive
radon samples were collected. Below 1 meter, the two methods gave
concentrations that were, within error, identical. The lower radon
value at sample point A1 using the active monitor was probably due
to leakage around the original packer. A second group of passive
monitors was placed in Hole B during August-November and produced
results that were significantly lower than the July-August
measurements. Hole B also contrasts with the other St. Marys Hills
data in that radon decreases with depth. These trends appear
related to increased water retention in the clayey soil of Hole B
as well as collection of water in the lower meter.
In Essex Park the radon levels range from 3 to 42 kBq/n3 with
an average of 26 ± 7 kBq/m3 (Table 2). The level of 3 kBq/m3 was
obtained at a depth of 0.2 meters in Hole G; at a depth of one
meter the lowest concentration was 13 kBq/m3. Some of the holes
show an increase in radon with depth; others show relatively
uniform levels. Some of the radon concentrations measured by the
active sampling are as much as 30% lower than concentrations
measured with the passive monitors. However, the means are not
statistically different.
A second set of measurements in Hole G during August-November
showed lower radon levels than during July-August and correspond
to the decrease observed in Hole B at St. Marys Hills. Although
the decrease can be attributed to higher water retention in the
soil during a rainy fall it is difficult to be sure as only one
hole was measured within each area during the late fall.
Radon emanation was measured on the bulk samples, the 63-14 9
and the <63 J1 fractions as described above. Replicate analyses
gave reproducible results with standard deviations comparable to
the error associated with counting statistics. A number of
factors, such as moisture, radium content, and location of radium
either on grain interiors or secondary coatings, control the
amount of radon emanated (3); however, on the average, higher
radon emanation would be expected to produce higher radon
concentrations in the soil gas. In Tables 3 and 4 the emanation
results are given relative to the mass of the sample. The <63-p.
fraction includes both silt and clay, and the 63 149 (1 (very fine
sand) represents about a third of the total sand fraction (up to 2
mm in size). In estimating total emanation, the results from the
9-46
-------
63-149 Ji. measurements were considered representative of all the
sand-size fractions.
Differences between the radon emanation rates of the two sites,
were comparable with those of the radon concentrations. The
average radon emanating from the soils in St. Marys Hills is just
over twice that emanating from the Essex Park soils. The
difference in means is statistically significant at the 0.025
confidence level. Although emanation rates in St. Marys Hills were
divided fairly evenly between the 63-149 jl and <63 size
fractions, in Essex Park the emanation rates for 9 out of 10
samples was highest in the <63 |i fraction. The variation of
emanation rates in Essex Park was much smaller than in St. Marys
Hills, in accordance with the more uniform radon concentrations in
Essex Park.
The sum of the emanation rates from the grain-size fractions
should be comparable with the emanation rate measured from each of
the bulk samples. In Essex Park this was the case, but in St.
Marys Hills, although most were comparable, some soils, such as
A6, had a bulk emanation rate that was larger than either the
individual or the sum of the fractional emanations. The overall
agreement between the bulk and weighted fractional emanations
indicates that the assumption that the 63-149 |i. size represents
the total sand fraction is reasonable for these samples.
Other radionuclides measured included 232Th, 226f ancj 210po
(Tables 5 and 6). Both the mean and standard deviation of 232Th are
equivalent for both sites. Radium and 210Pb values were higher in
St. Marys Hills than in Essex Park and were also more variable
both within and between sites than was thorium. If post-
depositional migration altered the radionuclide distributions, it
did not affect thorium, which is not mobile under near-surface
geochemical conditions. Uranium isotopes, however, respond to
weathering and changing oxidation/reduction environments, leading
to separation from daughter radionuclides and altered distribution
patterns. The relatively uniform distributions of radionuclides in
Essex Park sediments are consistent with little post-depositional
migration, whereas the uneven distributions in St. Marys Hills
indicate significant migration, possibly related to enhanced
weathering of the sediments on the hill slope.
The activity ratio 210Pb/226Ra in the sediment can be a useful
indicator of relative radon loss. A ratio smaller than one implies
that radon has moved away from the radium source, resulting in
less 210Pb activity relative to 226Ra. All but one of the samples
(A6) have activity ratios less than unity; in fact the overall
activity ratio is about 0.5, with St. Marys Hills having a
somewhat higher mean value (significant at the 0.05 confidence
level). Lower activity ratios could also result from only partial
recovery of polonium from the sediment, with the apparent effect
of reducing the Pb/Ra activity ratio. Sample A6, with an activity
ratio of 1.25, is at present an anomaly because the individual
9-47
-------
activities of 210Pb and 226Ra, as well as the activity ratio, are
much greater than those of the other samples.
Contrary to expectations, the lowest activity ratios were not
always near the surface where radon could easily escape into the
atmosphere. The sandy soils in Essex Park with the lowest activity
ratios imply that radon has moved away from its source even at
depths of 3 meters. Disequilibrium between 226Ra and 210Pb could
also result from downward migration of radium during weathering of
the sediments or could, as noted above, be partially related to
inefficient extraction of polonium from the sediment. We were not
able to compare the radon directly with either 226Ra or 210Pb
because the units were different (volume vs. mass), and the
samples did not always correspond in depth.
We also used a 1-inch Nal detector to measure the total gamma
activity at 2-foot intervals in several of the holes. In general
the activity versus depth relationship followed the pattern of
radium and polonium in the sediment except near the surface, where
there may have been accumulations of potassium. Total gamma
activity in the sediment appeared higher in St. Marys Hills, in
accordance with the other measurements, but not all holes were
measured. The results do indicate that subsurface gamma activity
is a potentially useful and simple screening technique, which
could be improved by using a spectroscopy system that determines
the energy of the radiation and identifies the isotopes present.
INDOOR RADON
The summary of the indoor radon information is given in Table
7. Of the 64 homeowners who were given the two detectors, 48
returned them. Of those, 17 were from the St. Marys Hills area and
31 from the Essex Park area. The mean indoor radon levels of St.
Marys Hills and Essex Park are different and significant at the
0.025 level for a two-sided t-test. The higher average indoor
radon in St. Marys Hills corresponds to the higher average
radionuclide contents in the sediments of St. Marys Hills and to
the higher radon emanation rates. The range of indoor radon
concentrations is similar for both areas; each has levels that
exceed 37 0 Bq m-3 and levels that are less than 37 Bq m~3- Although
this reduces the probability of predicting radon levels for
individual homes, there is a good correlation between the average
soil radon concentrations, parent/daughter radionuclides, and
indoor radon levels.
CONCLUSIONS
Our primary objective in this study was to measure, in two
different areas, radionuclides related to and including radon at
several depths within unconsolidated sediments, and to see what,
if any, correlation existed between the characteristics of the
sediment and indoor radon levels. All of the measurements were
made in glacially derived or related material. None were obtained
9-48
-------
from the limestone bedrock, which was encountered in only three
holes. We think that within the study areas the glacial sediments
are the primary source of indoor radon and that bedrock probably
is not a significant source. A more extensive study is needed to
determine which homes were built on or near bedrock and collect
additional data on the radon and other radionuclide levels.
Radium-226, 210Pb, radon emanation, and downhole radon levels
all have statistically higher averages in St. Marys Hills
sediments than those in Essex Park. Indoor radon levels also were
statistically higher in St. Marys Hills, and had a positive
correlation with the radionuclides in the soil. The mineralogy,
moisture levels, and bulk densities were similar in both areas and
did not correlate with the radionuclide distribution. Texturally
the sediments were variable but showed similar average contents of
gravel, sand, and silt/clay; however, more work is needed before
firm conclusions can be made about the effect of grain-size
distribution on the radionuclide content and distribution within
the sediments.
All of the techniques used to assess the radon potential were
consistent with each other and could be applied individually or
collectively to other areas. We believe that at these sites near
Rochester mineralogical characteristics of the sediments and the
location of samples within the stratigraphic column were only
partially responsible for the observed distribution of
radionuclides. We suggest that post-depositional transport of
uranium and radium related to weathering processes contributed to
the observed distributions. The redistribution of radionuclides
was more extensive in the St. Marys Hills area probably owing to
the greater vertical relief. In both areas 226Ra/210Pb activity
ratios indicate migration of radon independent of the
parent/daughter movement.
Predicting radon source areas in regions where sediments are
more than a couple of meters thick should not be be based solely
on identification of geological materials or on near-surface radon
measurements. Evidence for the secondary transport and
redistribution of radionuclides is not shown on geologic maps, and
near-surface radionuclide characteristics may differ from those at
basement depth. The data from this study, although limited in
area, indicate that measurement of radon or related radioactive
nuclides in soils can be a useful preconstruction indicator of
potential indoor radon problems. Survey methods could involve
active measurements at depths greater than 1 meter of soil gas
radon, subsurface gamma spectroscopy and 226Ra in the sediment.
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.
9-49
-------
REFERENCES
1. Luetzelschwab, J.W., Helweick, K.L., and Hurst, K. A. Radon
concentrations in five Pennsylvania soils. Health Physics. 56:
181, 1989.
2. Eakins, J.D., and Morrison, R.T. A new procedure for the
determination of Lead-210 in lake and marine sediments.
International Journal of Applied Radiation and Isotopes. 29:
531, 1989.
3. Nazaroff, W.W., and Nero, A.v., Jr. Radon and its decay
products in indoor air. J. Wiley & Sons, 1988. 518 pp.
4. Olsen, B.M. Bedrock geology, Plate 2 of 9. Balaban, N.H.,
editor. Geologic atlas of Olmsted County, Minnesota Geological
Survey County Atlas Series, Atlas C-3, Scale 1:100,000.
5. Ilobbs, II. C. Surficial geology, Plate 3 of 9. Balaban, N.H.,
editor. Geologic atlas of Olmsted County, Minnesota Geological
Survey County Atlas Series, Atlas C-3, Scale 1:100,000.
Partial funding for this project was provided the the Center for
Urban and Regional Affairs, University of Minnesota, Minneapolis.
9-50
-------
TABLE 7. SUMMARY OF RADON LEVELS (Bq/ra3) IN HOMES WITHIN THE STUDY AREA
Geometric Arithmetic
Mean Mean Min. Max.
St. Marys Hills
Rn Index No.*
180
* +
1.8
220
70
610
Rn Basement
250
* +
2.1
270
40
1100
Rn 1st Floor
130
3.0
160
30
400
Essex Park
Rn Index No.
60
* +
2.5
90
10
390
Rn Basement
80
*1-
2.4
130
15
650
Rn 1st Floor
40
2.7
70
10
280
fThe radon index number is a weighted average of the two radon measurements in the house. The weighting factor
for each floor was an estimate of the amount of time an occupant spends on each floor.
9-51
-------
TABLE 1. DOWKHOLE RADON MEASUREMENTS, ACTIVE & PASSIVE - ST. MARYS HILLS
Sample No.-
Depth (m)
Date
mo/yr
Active
(kBq/m3)
Date
mo/yr
Passive
(kBq/m3)
Date
mo/yr
Passive
(kBq/m3)
A1-1
10/88
17 ± 2
7-8/89
29 ± 4
_
..
A2-2
10/88
40 ± 3
7-8/89
40 ± 6
-
-
A3-3
10/88
61 ± 4
7-8/89
57 ± 8
-
-
A4-4
10/88
68 ± 6
7-8/89
71 + 9
-
-
B1-1
-
-
7-8/89
42 ± 6
8-11/89
25 ± 2
B2-2
-
—
7-8/89
41 ± 6
8-11/89
28 ± 2
B3-3
-
-
7-8/89
29 ± 4
8-11/89
20 ± 1
B4-4
-
-
7-8/89
water
8-11/89
5 ± 0.4
C1-1
—
-
7-8/89
26 ± 3
-
-
C2-2
-
-
7-8/89
44 ± 6
-
-
D1-1
-
—
7-8/89
44 ± 6
-
D2-2
-
-
7-8/89
46 ± 6
-
—
D3-3
-
-
7-8/89
53 ± 7
-
-
Averaqe
-
44 ± 13
-
Error values are one standard deviation based on counting statistics.
TABLE 2. DOWNHOLE RADON MEASUREMENTS, ACTIVE & PASSIVE - ESSEX PARK
Sample No.-
Date
Active
Date
Passive
Date
Passive
Depth (m)
mo/yr
(kBq/m3)
mo/yr
(kBq/m3)
mo/yr
(kBq/m3)
E1-1
8/89
22
+
2
7-8/89
15
±
2
—
—
E2-2
8/89
21
+
2
7-8/89
22
±
3
-
-
E3-3
8/89
21
+
2
7-8/89
18
+
3
-
-
F1-1
8/89
15
+
2
7-8/89
21
+
3
-
-
F2-2
8/89
26
+
3
7-8/89
32
+
4
-
-
F3-3
8/89
27
+
3
7-8/89
24
+
3
-
-
F4-4
8/89
27
±
3
7-8/89
36
±
5
-
-
G1-1
(0.2)*
8/89
3
±
0.3
7-8/89
18
±
3
8-11/89
20 + 1
G2-2
(1.2)*
8/89
23
±
2
7-8/89
31
+
4
8-11/89
28 ± 2
G3-3
(2.2 )'
8/89
24
±
2
7-8/89
33
±
4
8-11/89
23 ± 2
G4-4
(3.2)*
8/89
31
±
3
7-8/89
42
±
6
8-11/89
collapsed
H1-1
8/89
13
±
1
7-8/89
23
±
3
—
-
H2-2
8/89
21
±
2
7-8/89
25
±
4
-
-
H3-3
8/89
19
±
1
7-8/89
25
±
4
-
-
H4-4
8/89
17
±
1
7-8/89
23
±
3
-
-
Average
22
+
5t
26
±
7
-
'Depth (meters) of active radon measurements in Hole G. Error values are one standard deviation based on counting
statistics.
t Average does not include sample from depth 0.2 meters.
9-52
-------
TABLE 3.
EMANATION RESULTS -
ST. MARYS
HILLS
Sample No.
Bulk Emanation
Sum of Emanation
Emanation
Emanation
Depth (m)
from Sand &(Silt+Clay)*
63-149 ^
<63 n
(Bq/kq)
(Bq/kq)
(Bq/kq)
(Bq/kq)
A1-1.3
7.6 ± 0.4
14.7 ± 0.7
16.7 ± 0.9
f14.6 ± 1.0
A2-2.1
1-10.6 ±0.7
14.3 ± 0.7
11.0 ± 0.8
f16.9± 1.0
A3-2.9
20.1 ± 1.0
18.0 ± 0.9
15.9 ± 0.9
21.3 ±0.9
A4-3.5
20.2 ± 0.9
22.4 ± 1.1
19.1 ± 1.0
26.8 ± 1.2
A5-4.0
11.0 ±0.6
12.3 ± 0.6
11.4 ±0.7
14.5 ± 1.0
A6-4.6
f38.2 ± 3.3
17.5 + 0.9
f13.7 ± 0.8
t20.3± 1.3
B1-1.9
15.6 ± 0.7
9.5 ± 1.2
6.6 ±0.6
t11.2± 1.7
B2-3.4
11.1 ± 0.8
12.2 ± 0.6
20.2 ± 0.9
11.9 ±0.6
B3-4.6
13.8 ± 0.7
11.1 ± 0.6
t15.0 ± 0.8
+10.9 ± 0.6
C2-1.8
18.8 ± 1.0
11.4 ±0.6
21.7 ± 1.0
11.2 ±0.6
C2-2.9
11.0 ±0.6
12.0 ± 0.6
19.5 ± 0.8
f11.2 ± 1.7
D1-1.2
18.2 ± 0.7
12.1 ± 0.6
32.4 ± 2.0
11.4 ±0.7
D2-2.7
34.0 ± 2.0
27.1 ± 1.4
21.8 ± 1.0
30.9 ± 1.0
Averaae
17.7 + 9.2
-
-
-
"Emanation measured on 63-149 n size and applied to total sand fraction; emanation from <63 size includes both
silt and clay. Sum is the measured emanation times the weight percent of each size fraction.
fThe number is mean of replicate measurements; error is the standard deviation about the average.
TABLE
4. EMANATION RESULTS
i - ESSEX PARK
Sample No.
Bulk Emanation
Sum of Emanation
Emanation
Emanation
Depth (m)
from Sand &(Silt+Clay)*
63-149 n
<63 n
(Bq/kg)
(Bq/kq)
(Bq/kq)
(Bq/kq)
E1-1.9
t6.3 ± 0.6
6.8 ± 0.7
NSi
16.5 ± 1.0
E2-3.1
8.2 ± 0.5
12.3 ± 1.2
11.4 ± 0.7
14.9 ±0.8
F1-1.8
9.8 ± 0.7
12.7 ± 1.2
35.5 ± 1.4
10.1 ±0.6
F2-3.1
11.8 + 0.6
9.3 + 0.6
5.0 ± 1.0
9.4 ± 1.0
F3-4.3
8.8 ± 0.6
8.0 ± 0.8
t2.2 ± 0.6
10.0 ±0.7
G1-1.9
7.4 ± 0.6
5.6 ± 0.8
3.1 ± 0.4
10.6 ±0.6
G2-3.3
7.2 ±0.5
6.3 ±0.8
2.7 ±0.4
9.5 ±0.7
G3-4.4
f10.5 ± 0.7
6.7 ± 0.8
t1.2± 0.5
|10.7 ± 0.8
H1-1.8
5.6 ± 0.7
5.6 ± 0.8
2.8 ± 0.4
9.4 + 0.6
H2-3.3
4.7 ± 0.8
5.2 ± 0.8
2.3 ± 0.4
9.7 ±0.7
H3-4.3
5.2 ± 0.5
4.8 ± 0.7
2.0 + 0.4
10.0 ±0.7
Averaqe
7.8 ± 2.3
-
-
-
'Emanation measured on 63-149 n size and applied to total sand fraction; emanation from <63 ji size includes both
silt and clay. Sum is the measured emanation times the weight percent of each size fraction.
fThe number is mean of replicate measurements; error is the standard deviation about the average.
jNS indicates insufficient sample for measurement.
9-53
-------
TABLE 5. RADIUM-226, LEAD-210 AND THORIUM-232 IN ST. MARYS HILLS
Sample No.
Ra-226
Pb-210
Th-232
210Pb/226Ra
Depth (m)
(Bq/kq)
(Bq/kq)
(Bq/kq)
±>=10%
A1-1.3
82
±
10
f24.9
± 0.7
32
+
14
0.30
A2-2.1
77
±
10
1-18.8
± 2.6
48
+
18
0.24
A3-2.9
74
±
09
37.8
± 0.6
29
±
15
0.51
A4-3.5
64
±
09
37.2
+ 0.5
55
+
19
0.58
A5-4.0
45
4;
07
26.4
± 0.5
23
±
17
0.59
A6-4.6
117
+
12
t146
± 46
38
+
18
1.25
B1-1.9
30
+
6
23.7
+ 0.6
33
+
16
0.79
B2-3.4
47
±
8
26.4
± 0.5
55
+
15
0.56
B3-4.6
42
+
7
27.0
± 0.7
55
+
14
0.64
C1-1.8
60
±
9
39.9
± 0.8
45
+
19
0.67
C2-2.9
39
+
7
25.8
+ 0.6
34
+
16
0.66
D1-1.2
39
+
7
32.3
± 0.7
42
+
17
0.83
D2-2.7
79
±
9
t54.9
± 7
44
+
13
0.69
Averaqe
61
±
25
40
± 33
41
±
11
0.64 ± 0.25
fThe number is the mean of replicate measurements; error is the standard deviation about the average.
TABLE 6. RADIUM-226, LEAD-210 AND THORIUM-232 IN ESSEX PARK
Sample No.
Ra-226
Pb-210
Th-232
210Pb/226Ra
Depth (m)
(Bq/kq)
(Bq/kq)
(Bq/kq)
±= 10%
E1-1.9
27
+
6
16.1 ± 0.4
29
+
10
0.60
E2-3.1
21
+
5
10.9 ± 0.4
32
+
11
0.52
F1-1.8
39
4;
7
19.4 ± 0.5
49
±
17
0.50
F2-3.1
36
±
7
18.4 ± 0.8
64
+
20
0.51
F3-4.3
33
±
7
16.1 ± 0.7
50
+
17
0.49
G1-1.9
26
±
5
13.3 ± 0.3
30
+
14
0.51
G2-3.3
25
+
7
14.4 ± 0.4
37
+
12
0.58
G3-4.4
28
±
6
tl1.9 ± 0.2
29
±
13
0.42
H1-1.8
30
+
5
12.8 ± 0.4
33
15
0.43
H2-3.3
23
±
5
6.5 + 0.3
32
±
14
0.28
H3-4.3
25
±
6
9.2 ± 0.4
31
+
10
0.37
Average
28
±
6
14 ± 4
38
+
12
0.47 ± 0.09
fThe number is the mean of replicate measurements; error is the standard deviation about the average.
9-54
-------
Essex Park
|||llII|i|IIlliiHW
^ v l,w!
- Rochester "
• • •••• \. \ ^ > i*
St. Marys Hills
Scale
I L.
i *r
_l
Figure 1. Map showing study areas near Rochester,
Olmsted County, Minnesota
9-55
-------
Plan View
St. Marys Hills
Elevation (meters asl)
370
Loess —
111 [Galena Ls
C.D
360
=st
Glacial Till
Decorah Shale;
350
Platteville Fm
340'
Glenwood Shale
? Contact ?
St. Peter Sandstone
? Contact ?
Prairie du Chien
Elevation (meters asl):
320
310
Essex Park
Plan view
ground surface
Glacial Till
300
Prairie du Chien Group (Ls)
0 St 193 «
.. . , , i i Vertical exageralion ~ 8x
Horizontal Scale ' 1 '
Figure 2. Profiles and location of sample holes. Modified from
Plates 2 and 3, Olmsted County Atlas (4,5).
9-56
-------
Loess
Oxidized glacial
sediment
Depfi
m
Hole A
A A
ft A
AAA
Sample
Interval
St. Marys Hills
Hole B Hole C
Hole D
Unoxidized
glacial
sediment
Essex Park
HoleE
HoleF
HoloG
HoleH
Loess
Clay
Sandy glacial
sediment
Unoxidized
glacial
sediment
Sample
Interval
PVV'
•yv;
'•'^3
L%-VVJ
:-.w
'S'S'N
'S'VS
W;V
v.%^-
.ŁŁ¦
tftse
¦St
EUUJ
Figure 3. Borehole lithology and sample locations
based on field identifications.
9-57
-------
GEOLOGICAL PARAMETERS IN RADON RISK ASSESSMENT - A CASE HISTORY
OF DELIBERATE EXPLORATION
Donald Carlisle and Haydar Azzouz *
Department of Earth and Space Sciences
University of California Los Angeles
Los Angeles, California 90024
ABSTRACT
Geological exploration has identified an unsuspected radon-prone belt
in southern California. Detailed analysis of aeroradiometric (NARR)
data in relation to geological units, soil-gas radon, soil permeability,
and finally indoor radon has identified the Rincon Shale and soils
derived predominantly from the Rincon Shale in Santa 3arbara County as
anomalous in uranium and radon. Roughly 7 6% of our screening tests to
date from houses on the Rincon Shale exceed 4 pCi/1 and 26% exceed 20
pCi/1. Measurements under "normal" living conditions show 50% exceeding
4 pCi/1. An estimated 4,000 plus houses are at this level of risk;
extensive new construction on the Rincon Shale is limited only by
domestic water supply.
Unusually good correlations between aeroradiometry, soil-gas radon at
75 cm depth adjusted for soil-gas permeability, geology, and indoor
radon concentrations reflect the unmetamorphosed character of
sedimentary host rocks and the tendency for anomalous uranium
concentrations to be disseminated throughout a geological unit rather
than in erratic mineralized zones. Under these circumstances,
deliberate geological exploration can be a more efficient approach to
indoor radon risk identification than simple random sampling or non-
random testing of houses and by the same token geological parameters can
facilitate radon risk assessment on undeveloped lands. Attention is
drawn to multiple populations within radon test samples and the
consequent problems in estimating regional parameters.
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.
* Present address of Haydar Azzouz: GeoSyntec Consultants, Huntington Beach,
California
9-59
-------
RATIONALE
Two very different questions can be asked about the incidence and
distribution of indoor radon concentrations:
1. What is the probability distribution of indoor radon
concentrations among the entire stock of houses in a given region
or in the country as a whole?
2. What is the probability of occurrence, and the location, of
radon-prone areas within this given region? This latter question
is of much greater interest to individuals.
Answers to the fir3t kind of question, regional probability or
frequency distribution, are usually estimated by statistical analysis of
measurements from a simple random sample or probability sample of houses
in the area of interest. Simple random sampling aims ideally to avoid
bias by making every house equally selectable and, as a consequence,
obliterating differences among sub-populations which may, or may not,
exist within the whole. Ex-post-facto analysis of existing measurements
from private and/or public sources is a less expensive and less reliable
substitute. Aggregate regional frequency distributions so obtained are
usually shown as approaching log-normality with characteristic arithmetic
means from 0.8 to 11.3 pCi/1, geometric means from 0.6 to 3.3 pCi/1, and
geometric standard deviations from 2.1 to 3.4 pCi/1. In reality the
distributions are commonly very irregular, particularly at higher
concentrations, and undoubtedly represent multiple populations each with
its own characteristic frequency distribution.
The alternative approach, which we among others have taken, is to
purposely explore for radon-prone areas using geological reasoning along
with inexpensive, practical techniques modified from mineral exploration,
engineering, or research methodology already in use or in the literature.
In other words we have directly addressed the second kind of question:
the probability of occurrence and the location of radon-prone sub-
populations. There is now an extensive literature on sources,
distribution, and measurement of radon in soils, and on its contribution
to indoor radon which we will not cite in detail: a recent primary
reference is Nazaroff, et al. (1) . The very brief summary on the next
two pages establishes the principal assumptions for our work.
Given that the overwhelming preponderance of indoor radon is derived
from underlying soils and rocks, and ultimately from U-238, the detection
of anomalous natural radon sources is in large measure the detection of
anomalous uranium concentrations and therefore quite analogous to the
exploration for mineral deposits in general. Uranium is a ubiquitous
element, present in Lrace amounts in all soils and rocks in
concentrations ranging from as little as 0.2 ppm (parts per million) or
less in sandstones, from 1 to 10 ppm or more in common igneous rocks, and
to as much as 200 ppm, rarely 500 ppm in black shales. Contents in
excess of 5 ppm are considered anomalous by some investigators. Ore-
grade concentrations average from 500 to 28,000 ppm, locally much higher,
9-60
-------
across mineralized zones which are comparatively small, erratic and
difficult to find.
Geological controls for uranium distribution are reasonably well known
.in principle. Lithologica 1. and geochemical methods of exploration,
including radiometrics, are particularly useful but must take into
account the differing geochemical properties of the several isotopes
within the U-238 decay series and the consequent likelihood of departure
from secular equilibrium. Uranyl ions produced during oxidative
weathering, for example, are extremely soluble and mobile in contrast
with the insoluble Th-234 and Th-230 compounds. Ra-226 behaves as an
alkaline earth: its daughter Rn-222 is an inert gas. Natural secular
disequilibrium is commonly found 1) where the deposition of uranium-
bearing sediments or rocks (or the later introduction of uranium into
host sediments or rocks) has taken place very recently - crudely less
than ten million years or so - in which case the radioactive decay
products will not have "grown in" completely and will be "deficient"
relative to uranium content and 2) where the soluble uranyl ion has beer,
leached from the near surface by weathering, while thorium has not, in
which case the uranium decay products, and their associated
radioactivity, may appear to be excessive in relation to uranium. Radium
too can migrate from its source under near-surface conditions, in r.he
same way as other alkaline earths. Radon gas, of course, moves easily
unless confined which is why the radon content of soil gas close to the
surface of the ground approximates that of ambient air even though radon
content at depths of a meter or so ranges up to hundreds, thousands, or
tens of thousands of pCi/1. Nevertheless, radon anomalies in soil-gas
samples taken from comparable depths are often reasonably indicative of
uranium anomalies nearby.
The predominant source of gamma radiation in the U-238 series
detected by ground or airborne scintillometry is Bi-214 and accordingly
this isotope is the most widely used geochemical pathfinder in uranium
exploration. The fact that 3i-214 is separated from Rn-222 in the decay
series by only two extremely short-lived isotopes, Po-218 and Pb-214
helps to maintain a correlative relationship between radon and the
observed Bi-214 gamma radiation in spite of the fugacity of radon and in
spite of the fact that gamma radiation is essentially blocked by 20 cm or
so of typical soil. Airborne gamma-ray spectrometry is an excellent
uranium reconnaissance tool. Standard practice is to calculate an
apparent uranium concentration from Bi-214 gamma-ray intensity as if
secular equilibrium actually obtained and to report this apparent
concentration as "equivalent uranium" (eU). This same technique and
terminology is equally useful for concentrations of radon (eRn), radium
(eRa) or other precursors of 3i-2l4 in upper layers of the soil.
Geological controls influencing near-surface radon concentration,
given the distribution of radium in the underlying soil must take into
account: 1) the proportion of Rn-222 newly-produced from Ra-22 6 able
to escape from the solid mineral phase into soil gas - the "emanating
fraction," 2) the distribution of fractures, shear zones or other
pathways which facilitate upward radon migration and 3) soil-gas
permeability. Of these three, soil-gas permeability is the most easily
quantified for a particular site. Gas permeability of soils ranges over
eight orders of magnitude in the extreme case of gravels and clays,
although in typical soil categories the range is reduced to about four or
five orders of magnitude: roughly 5 x 10-10 cra^ for silt - clay mixtures
9-61
-------
to 5 x 10 ' cm2 for coarse sand. Variations in the emanating fraction
are too costly to evaluate since they depend heavily on local soil
moisture content among other factors and in any event they are relatively
insignificant for purposes of radon exploration. Fractures or other
pathways are essentially impossible to quantify as controls but sometimes
are recognizable visually or by geological inference as confounding
factors in particular sites. The seasonal cracking of montraorillonite-
rich soils to depths of as much as a meter, as in the case of Rincon-
derived soils, may be characteristic of an entire formation and more
significant than soil permeability since it tends both to reduce the
radon content of otherwise impervious soil and to facilitate transfer of
soil gas into a building. Moisture content is a major non-geological
variable - though not the only one - because of its affect on both the
emanating fraction and soil-gas permeability. An optimum moisture
content for combined radon emanation and migration has been observed by
Stranden, et al. (2) at about 25%.
Recognizing all these complexities, and lesser ones, it can
nevertheless be argued 1) that radium concentration or soil-gas radon
concentration is a good measure of the "source strength" for radon in the
soil and 2) that soil-gas permeability is a first approximation, but
only a first approximation, of the rate at which radon-bearing soil gas
can reach the foundation of a building. A radon index number (RIN) which
includes only the two parameters, soil radium concentration and soil-gas
permeability was first suggested by DSMA Atcon, Ltd. (3). Tanner (4)
subsequently proposed a radon availability number (RAN) defined as the
product of soil-gas radon concentration, mean radon migration distance,
and soil porosity. At about the same time Kunz, et al. (5) developed a
simplified RIN based empirically on comparisons with indoor radon
measurements in New York State. The Kunz formulation which we have
adopted here is:
RIN = 10 ( C ) ( K )1/2
where: C is soil-gas radon concentration (pCi/1)
K is the soil-gas permeability (cm2)
The factor 10 was inserted by Kunz, et al. merely to
make the RIN roughly comparable with their typical
indoor radon levels.
A "depth factor," less than or equal to 1, is added by
Kunz for areas where the depth to the water table,
bedrock or substantially less permeable soil is known
to be less than 10 feet.
The purpose of our work has been to test the practicality of
deliberate radon exploration using aeroradiometri.es, extrapolated within
reason, as a reconnaissance tool followed by geology, soil-gas
measurement, application of the Kunz, et al. formulation and, where
appropriate, by detailed site studies all in a large populated portion of
Southern California where the indoor radon risk inferred from random and
non-random home tests has been purported to be very low.
For our study we chose the part of Southern California encompassed by
the Los Angeles Sheet of the Geological Map of California. This is a
9-62
-------
one-degree by two-degree sheet ( 34° to 35° latitude, 118® to 120®
longitude) including roughly the northern half of metropolitan Los
Angeles and extending from about 8 miles east of Pasadena to about 18
miles west of Santa Barbara.
We were fortunate in that shortly after initiating our analysis of
aeroradiometric data, the California Department of Health Services began
a three-month alpha track survey in a random sample of home.: in a port.ion
of our study area .in northwestern Los Angeles - southeaster" Ventura
counties. Through DHS efforts 82 homeowners (out of a total of 171 DKS
participants) made their properties available to us for brief
examination, for soil and soil-gas sampling adjacent to the house and for
surface gamma-ray measurements. The indoor radon measurements became
available for 79 of the test houses after our radiometric compilations
and our field measurements were complete and these became the basis for
evaluation of our methodology (6). In the following discussion we refer
to this 82-home area as the Primary or Baseline Subarea.
METHODOLOGY
AERORADIOMETRIC RECONNAISSANCE DATA
Beginning in the mid 1970's, the U.S. Department of Energy sponsored
the National Airborne Radiometric Reconnaissance (NARR) to provide a
semi-quantitative evaluation of radioactive element distribution in the
United States as part of the National Uranium Resource Evaluation (NURE)
Program. Gamma-ray data were collected on K-40 for potassium, Tl-208 for
thorium, and Bi-214 (using the 1.7 6 MeV photopeak) for uranium typically
by means of helicopter at an average of 400 feet or less above ground
surface fitted with a gamma-ray spectrometer and large crystal detectors.
Primary flight lines are typically oriented east-west about 3 miles
apart; tie lines are typically north-south and about 12 miles apart.
Radiometric data were corrected for live time, aircraft and equipment
background, cosmic background, Compton scatter, altitude, barometric
pressure and temperature. In California the corrected data were
statistically evaluated in terms of individual geological units as shown
on the The Geologic Map of California. Statistical data were reported on
eU ppm, eTh ppm, K %, and on their ratios to 0.01. The compiled data
were also presented as pseudo-contour maps, stacked profiles, anomaly
maps, and "geological histograms" which are frequency distributions of
the "equivalent" isotope concentrations and their ratios for each
geological unit.
Stacked profiles show eU along each flight line and are the most site-
specific graphic presentation of the data. Extrapolations can be made to
specific sites between flight lines with varying degrees of reliability.
We have done this using the appropriate geological histograms and eU
anomaly maps as a basis for making the extrapolations. The wide spacing
of flight 1ines, particularly the north-south tie lines, and the small
scale of the graphical reproductions introduce large uncertainties in the
longer extrapolations. A striking case in point is the community of
Summer]and which is predominantly on radon-rich Rincon Shale but also
9-63
-------
entirely between flight lines and therefore not shown as eU anomalous on
the aeroradiorr.etric diagrams or maps. The point we would make, however,
is that our geological approach led us directly to Suirmerland, among
other areas, as soon as we had identified the Rincon Shale as eU and eRn-
anomalous in the areas which were covered by the flight lines.
Aeroradiometric data for the Los Angeles Sheet were plotted by
geological unit and informally subdivided by us into six categories
having mean values between 1.0 - 1.9 pprn eU and 6.0 - 6.9 ppra eU.
Perhaps the main value of extrapolation between flight lines is that
it makes possible the comparisons of aeroradiometric data with soil-gas
radon concentrations and with indoor measurements both of which we
discuss next.
SOIL-GAS RADON CONCENTRATION
We used the method of Reirr.er (7) to obtain soil gas samples: A
stainless steel probe, 0.80 cm O.D., 0.16 cm I.D., is pounded into the
soil to a depth of 75 cm, an O-ring fitting with a septum is attached and
three successive 10 cc samples of soil gas are extracted by hypodermic
syringe through the septum after purging the probe. The small diameter
of the probe ensures minimal disturbance of the subsurface environment.
The 7 5 cm depth is a compromise between probe refusal or bending and the
more ideal depth of 1 to 1.5 m where radon concentration in soil-gas
tends to approach an equilibrium value. It also enables sampling from
the lower B or upper C horizon. Samples were always collected during the
day, not during periods of unstable weather or strong winds, and not
after precipitation until dry conditions were allowed to return. Special
care was taken to sample natural soil away from any filled zone. The
time of sampling was recorded.
Soil-gas samples were then taken to the laboratory for radon
measurement by injecting the sample through a valve and septum device
into a Lucas cell radon/radon daughter detector ( RDA 200, manufactured
by EDA Instruments, Inc.). Measurements were made 3 to 24 hours after
soil ga3 sampling, more than sufficient time for decay of Rn-220
(thoron). Earlier experiments showed that radon daughters in the
original soil-gas sample or generated up to the time of injection into
the Lucas cell are plated out in the hypodermic syringe. Since there is
also some adsorption of radon, particularly on the syringe plunger, and
other potential complications, the entire assemblage of components was
calibrated as configured during analysis against known radon-bearing gas
samples from the SPA operated chamber at Las Vegas, Nevada. Each Lucas
cell was partially evacuated to a standard pressure prior to sample
injection. Early experiments also showed that truly consistent results
require 25 minutes of counting in the Lucas cell detector. Counts up to
5 minutes were judged to be insufficiently consistent for the research
stage of our work reported here and were omitted from the calculations.
The formula used to calculate radon concentration is:
C = {N30 - N5 - CB) / (T x SV x DF x CF)
where: C = Soil-gas Rn-222 concentration in pCi/1
9-64
-------
n30
N5
CB
T
SV
DF
CF
Counts in 30 minutes
Counts in 5 minutes
Cell background (counts) prior to sample injection
Counting period: 25 minutes
Sample volume: 0.01 liters
Radioactive decay factor for radon for elapsed
sampling-to-counting time: from standard table.
"Cell Factor" in cpm/pCi from calibration of
configured apparatus using known radon chamber
samples.
Even at moderate concentrations of radon this 30-minute long counting
period results in appreciable contamination of the Lucas cells, as Reimer
points out, and for reconnaissance purposes much shorter counting periods
may be acceptable. In exploration subsequent to the research reported
here we have tested and adopted a counting protocol of 5 arid 10 minutes
with appropriate adjustment of the formula.
Prior to sampling in the study area, the following indications of
precision and reproducibility were obtained from a test plot:
1) Radon measured from soil-gas samples at 75 cm depth from 9 probe
sites within a square meter yielded a Gaussian distribution, arithmetic
mean = 1,071 pCi/1, standard deviation = 106 pCi/1 (coefficient of
variation =10 %).
2) Radon measured from five soil gas samples consecutively drawn from
one probe site in the same plot yielded a Gaussian distribution,
arithmetic mean = 1,052 pCi/1, standard deviation = 85 pCi/1
(coefficient of variation = 8 %).
Standard procedure in the Primary Subarea was to occupy three probe
sites within 4 meters of each house and to take three soil-gas samples at
75 cm depth from each site. If the first two radon analyses from a giver,
probe site were within ten percent of each other, the third sample was
discarded, otherwise it was measured. The soil-gas radon value reported
for each house location is the mean of the three probe sites. The
variation coefficient for each house location ranged from 1.1 % to 63.3 %
with a mean of 18.5 %.
SOIL-GAS PERMEABILITY
Recent studies including those of Kunz and Tanner previously cited
have adopted a quantitative determination of soil-gas permeability based
upon measured gas flow under measured differential pressure during
pumping of gas from the ground or pumping of air into the ground.
However this apparently rigorous method rests upon assumptions about the
size and geometry of the soil volume which the gas is drawn from or
pumped into even in the case of uniform soil profiles. It is unclear
even in presumably homogeneous soils whether the geometry of the soil
volume should be assumed to be spherical, hemispherical or some other
shape, for example. Many soils are not only layered but are randomly
penetrated by fractures, root cavities or animal burrows or contain
irregular zones of varying permeability.
9-65
-------
In-situ measurement of water permeability is not truly indicative of
in-situ gas permeability nor are gas permeability measurements on
reconstructed soils.
Therefore bearing in mind cost and practicality, we have adopted the
technique of permeability estimation based upon grain size distribution
determined by sieving the dried soil. This method is eminently suitable
for reconnaissance work and perhaps as good as any for detailed follow-on
studies. It has the advantage not only of simplicity but also of
avoiding the potentially large errors that might arise during measured
pressure-volume pumping due to openings in the soil or zones of varying
permeability which are entirely site-specific on a small scale and
therefore not necessarily representative of the area under consideration.
Soil samples taken at depths of 25 to 35 cm from the same sites as
soil-gas samples were oven dried overnight at 100 degrees Celsius and
sieved by mechanical shaker. Soil types were categorized and
permeabilities were assigned on the basis of published tables, e.g.
Sextro, et al. (8). Because of the large number of samples involved we
did not perform wet separation of clay and silt grain sizes. Nor did we
measure moisture content, the degree of compaction or the cementation of
the soil. Perhaps the permeability assigned in this way should be called
pseudo-permeability or at best equivalent permeability but the method is
as likely as any to provide a reproducible basis for comparison of soils
at reasonable cost. Moreover it should be noted that permeability
appears in the RIN formulation, above, only as its square root.
RESULTS FROM THE PRIMARY (BASELINE) SUBAREA IN NORTHWESTERN LOS ANGELES -
SOUTHEASTERN VENTURA COUNTIES
HOUSE CHARACTERISTICS IN THE PRIMARY SUBAREA
Mention was made above of the 82 test homes in the DHS three-month
alpha track survey which were made available to us for soil-gas radon and
soil permeability measurements. Questionnaires about house design and
use were completed by all participants. None of the 79 houses (out of
82) for which indoor radon results became available had indoor
measurements in excess of 2.9 pCi/1 and we were not able to show that any
of the adjacent soils had more than normal gamma-ray activity at the
ground surface using a hand-held scintillometer with a one cubic inch T1
activated Nal crystal. None of the houses have a basement and only 4 %
have crawl space construction: the remaining 96 % are slab-on-grade.
Ventilation patterns were approximately similar during the test period
which was characterized by mild dry coastal and near-coastal Southern
California weather. It would appear that in this survey building
characteristics and meteorological factors probably had less-than-average
influence on indoor radon levels.
9-66
-------
n-squared: .757
4500
4000
3S00
% 3000
cc 2500
(3 2000
000
500
2.5 3 3.5
Airborne Equivalent Uranium (ell) ppm
4.5
5.5
Figure 1. Airoorne equivalent, uranium vs soil-gas radon in the Primary Subarea
R-squared: .503
Indoor Rn
pCi/1
o8.o° °
0 500 1000 1500 200C 250C 30C0 3500 1000 4600
Soil-Gas Rn pCi/l
Figure 2. Indoor radon vs sell- gas radon in t he frixary 3-ibarea
R-squared: .683
Indoor Rn
pCi/1
CEO O
R I N
Figure 3. Indoor, radon vs RIN value in the Primary Subarea
9-67
-------
AERORADIOMETRIC DATA, SOII.-GAS RADON, RIN VALUES AND INDOOR RADON IN THE
PRIMARY SUBAREA: CORRELATIVE RELATIONSHIPS
Figure 1 shows the relationship observed between airborne equivalent
uranium (eU) and soil-gas radon concentrations obtained by soil probe
from the vicinities of the 82 houses in the Primary Subarea. The
strength of the correlation probably reflects the proximity of Rn-222 and
3i-214 in the decay series but was somewhat surprising and encouraging
nevertheless.
Figure 2 shows the relationship observed between indoor radon and
soil-gas radon for the 7 9 houses with indoor radon measurements in the
Primary Subarea. Again the correlation is very encouraging in spite of
some scatter. We believe that it probably reflects in part the relative
uniformity of house and weather parameters noted above. However it may
also reflect the relatively well-ventilated character of nearly all homes
in the region because, even though some soil-gas radon concentrations
were between 2,000 and 4,500 pCi/1, the maximum indoor radon level was
only 2.9 pCi/1.
Figure 3 shows the relationship observed between indoor radon and the
RIN value. The essential difference between soil-gas radon and RIN is
that soil-gas permeability is taken into account in the latter. It is
not surprising therefore that the correlation of indoor radon with RIN
value is even better than with soil-gas radon.
Sample statistics are as follows. Equivalent uranium extrapolated
from the NARR data in the Primary Subarea shows a range of 1.5 to 5.8
ppm eU, an arithmetic mean (AM) of 2.9 ppm, a geometric mean (GM) of 2.8
ppm, and an arithmetic standard deviation (ASD) of 0.86 ppm. Soil-gas
radon concentrations have a range from 206 to 4,390 pCi/1, and AM =
1,388 pCi/1, GM - 1,162 pCi/1 and ASD = 859 pCi/1. Indoor radon shows
a range of 0.2 to 2.9 pCi/1, and AM = 1.2 pCi/1, GM = 0.99 pCi/1, ASD =
0.70 pCi/1. All three frequency distributions are 3kewed toward a log-
normal pattern but are quite irregular. Soil-gas radon concentrations
most closely approach log-normality as discussed later in the paper
(Figure 4) .
DISCOVERY OF THE RADON-PRONE RINCON SHALE BELT
AERORADIOMETRIC INDICATIONS
In the entire northwestern Los Angeles - southeastern Ventura region,
which includes the Primary Subarea, there are only four elongate patches
in which the geological units have aeroradioraetric signatures in category
5.0 to 5.9 ppm eU, the second highest of our categories, and these
patches are predominantly in undeveloped terrain. None of the 82 test
houses happen to be on a geological unit with a truly anomalous mean eU
level: the highest category is 4.0 to 4.9 ppm eU. However the excellent
9-68
-------
correlations between eU, soil-gas radon and indoor radon at the low
levels observed demanded that we re-examine the aeroradiometric map.
To the west of the Primary Subarea, passing through major parts of
Santa Barbara city and vicinity, there is a pronounced east-westerly
trending belt with eU averaging 6.0 to 6.9 ppm (up to 14 ppm in one
place) and an adjacent belt in category 5.0 to 5.9 ppm eU. Both of
these belts coincide with the generalized unit "ML" (Lower Miocene) as
shown on the Los Angeles Geological Map Sheet. "ML" in this area could
encompass parts or all of three geological formations and several
members. Lithology pointed to two possible uranium-rich candidates: the
dark, moderately organic Rincon Shale and the lower Monterey Formation
which is locally organic and locally phosphate-bearing. Uranium tends in
general to associate with organic matter and with phosphate. Probe soil-
gas sampling and analysis, followed by indoor radon screening tests in a
handful of houses, very rapidly identified the Rincon Shale as the
undoubted source of most, if not all, of the anomalous airborne
signature. There is no Rincon Shale in the Primary Subarea.
SOIL-GAS RADON IN THE RINCON SHALE
To date we have made soil-gas radon measurements at 66 locations, each
with from 1 to 4 probe sites, on the Rincon Shale and soils derived
predominantly from the Rincon Shale extending from near the easterly
boundary of Santa Barbara County to Gaviota on the west; a strike length
of about 48 miles. The range of soil-gas radon concentrations is from
1,100 to 20,350 pCi/1, with an AM of 6,400 pCi/1, GM of 5,045 pCi/1,
and an ASD of 4,486 pCi/1. The AM is more than four and one-half times
higher than the mean of soil-gas radon concentrations in the Primary
Subarea or in non-Rincon units in Santa Barbara County and elsewhere in
Southern California that we have tested to date.
The frequency distributions of soil-gas radon values in the Rincon
Shale and in non-Rincon formations each approach lognormality but the two
distributions are distinctly different and represent two separate
populations. Figure 4 shows two near-Gausian curves, one derived from
the logarithms of soil-gas values from the Rincon formation, the other
similarly derived from non-Rincon soil-gas values. If the two different
populations are lumped together on a single histogram one also sees a
crudely lognormal distribution but the statistics of the lumped data are
not truly representative of either population.
It may be geologically significant that the highest Rincon soil-gas
radon concentrations appear to be near Santa Barbara itself, though much
more data need to be obtained both to the east and the west to confirm
this pattern.
INDOOR RADON MEASUREMENTS IN HOUSES ON THE RINCON SHALE
We have now tested 85 houses in Santa Barbara County, the great
majority in easterly Santa Barbara city, Montecito and Summerland.
Three 3 to 7-day ACs for the screening test and one AT for the longer
9-69
-------
20-
1
Non-Rincon Formations
Rincon Shale
4.6
4.2
4.4
2.6
2.8
3
3.2
3.6
3.8
2.2
2.4
3.4
4
Logarithm of Soil-Gas Radon Concentrations
figure 4. Frequency distributions of log soil-gas radon concentration in the Rincon Shale
and in non-Rincon formations
R-squareo .305
O
Q.
C
oc
o
o
x>
c
O Rincon House
-2500
"T 1 ' ¦» 1 « T
2500 5000 7500 5 0000 -2500 J5000 17500 20000 22'
500
Soil-Gas Rn pCi/l
Figure S.
Indoor radon in relation to soil-gas radon in the Primary Subarea
and in houses on the Rincon Shale regardless of their age.
R-squared: .702
e
e
Houses More Than5yreOkJ on Rincon Shale
Primary Subarea Houses
a. I * I ¦ 1 ¦ 1 " T 1 I » " I * T
-2000 C 20C0 4000 6000 0000 10000 12000
Soil-Gas Rn pCi/l
Figure 6. Indoor radon in relation to soil-gas radon in the Primary Subarea
and in Rincon Shale houses more than 5 years old.
9-70
-------
terrr. measurement were placed in each house and retrieved by us following
SPA standards. Thirty eight of these houses are on the Rincon Shale or
on soils derived predominantly from Rincon Shale. Slightly over 76 % of
the houses on the Rincori Shale have screening test results in excess of
4 pCi/1: twenty six percent exceed 20 pCi/1 (Range 1.6 to 58 pCi/1; AM =
16 pCi/1; ASD = 14 pCi/1; GM = 11 pCi/1). Follow-on alpha track
measurements of from one to six months duration, under normal living
conditions show 50 % exceeding 4 pC.i/1 and none exceeding 20 pCi/i
although some of the worst situations were mitigated prior to completion
of the follow-on measurement (Range 1 to 19 pCi/1; AM = 6 pCi/1; ASD = 5
pCi/1; GM = 4 pCi/1). An estimated 4,000 plus houses are at the
indicated level of risk: extensive new construction on the Rincon Shale
is limited only by domestic water supply.
On".y one house not on Rir.con Shale has a follow-on measurement
exceeding 20 pCi/1 and that house is on a slide area with a badly cracked
slab-on-grade at the time of measurement and, almost certainly, deeply
cracked soil underneath.
Two subsequent independent surveys support our conclusions: one by
Keller of the University of California Santa Barbara who providing
single activated charcoal detectors to roughly 100 homeowners on and off
the Rincon Shale (9) and a second by the California Department of Health
Services also based on single AG's placed by the homeowner (10) . In both
cases homeowners were instructed to ensure screening test conditions
IRREGULARITIES IN THE RELATIONSHIP BETWEEN SOIL-GAS RADON AND TNDOOR
RADON ON THE RINCON SHALE: SOIL FAILURE AND AGE OF HOUSE
If one combines the data on indoor radon and soil-gas radon from the
total study area - i.e. the Primary Subarea and Santa Barbara Rincon
Shale areas - on a single plot (Figure 5) the general correlation of
soil-gas radon and indoor radon observed previously is clear enough but
the scatter within the Rincon Shale portion of the data is very large. A
plot of the Rincon Shale portion alone is likewise strongly scattered and
shows only a weak positive relationship between indoor radon and soil-gas
radon.
The failure of indoor radon to reflect soil-gas radon concentrations
arises in a group of houses which have a range of high soil-gas radon
concentrations around the house, often greatly exceeding 6,400 pCi/1 -
the mean value for Rincon Shale - but have indoor results around only 4
pCi/1. All of these houses were less than five years old at the time of
tesLing. Though there are unfortunate exceptions, newer houses arc
generally built to a higher standard than those of many years ago;
concrete slabs and footings are of better design even though not
specifically "radon resistant," arid such older practices as leaving a 1
to 2 square foot opening in slabs for plumbing access, for example, are
happily discontinued. Many newer crawlspace-design houses have oversized
and well ventilated crawlspaces often with subfloor insulation and vapor
barriers. Moreover newer houses have not yet had time to develop
concrete cracks or other openings which might allow soil gas to enter
more easily.
9-71
-------
Rincon-derived soils are montmorillonite-rich and prone to soil
heaving and, on occasion, to soil flow and fracture associated with slope
failure all of which disrupt the structural integrity of a house. In
general the older the house, the greater the likelihood of cracks and
openings for soil-gas entry and the more closely does the indoor radon
value reflect soil-gas radon content. The tendency mentioned earlier for
Rincon-derived soils to develop abundant large dehydration soil cracks
may account for unexpected high indoor values in at least two houses in
our sample but it does not account for unexpectedly low indoor values.
To test the "age-of-house" hypothesis we subtracted the data on all of
those houses less than 5 years old: the result (Figure 6) is a striking
improvement in the correlation between indoor radon and soil-gas radon.
The improvement is equally obvious on a Rincon-only plot.
Statistics for our screening tests of houses on the Rincon Shale are:
Range 1.6 to 58.4 pCi/1; AM = 16 pCi/1; GM = 11 pCi/1; and ASD = 14
pCi/1. Measurements under normal living conditions have a range of from
1 to 19 pCi/1; AM = 6 pCi/1; GM = 4.3 pCi/1; and ASD = 5.2 pCi/1.
DISCUSSION
Figure 6 also serves once again to illustrate the existence of the two
separate populations, one for houses on the Rincon Shale and the other
for houses on non-Rincon formations. The existence of sub-populations,
in any large region, needs to be kept in mind whenever attempts are made
to use regional frequency distributions of radon occurrence as a basis
for predicting national or regional radon risk. It is quite possible for
simple random samples of houses to miss Rincon Shale environments
especially when the random samples are small.
The correlations that we have obtained regionally between
aeroradiometry, soil-gas radon at 75 cm depth adjusted for soil-gas
permeability, geology, and indoor radon concentrations are apparently
better than those reported from some other places. In our opinion this
probably reflects the fact that the unmetamorphosed sedimentary rocks
that we find in much of coastal and central California tend either to
have, or to not have, anomalous uranium concentrations disseminated more
or less throughout the unit. This is not to say that erratic uranium-
rich zones do not occur in these same rocks: there are many examples of
small quite rich uranium concentrations in the Monterey, in the Sespe
and several other unmetamorphosed formations. The locations of these
erratic concentrations are extremely difficult to predict but, unlike
some situations in metamorphosed terrains for example, such
concentrations are few and far between. Anomalous amounts of uranium
disseminated throughout a rock unit can affect a large number of houses
and it is especially in these circumstances that deliberate geological
exploration can be a more efficient approach to radon risk
identification than simple random sampling or non-random testing of
homes. Judging by experience in the mineral industry, exploration based
upon geological models of occurrence is infinitely more likely to find
anomalous occurrences than is random sampling. Models for geological
predictability can also contribute to radon risk assessment on
undeveloped tracts of land.
9-72
-------
REFERENCES
1. Nazaroff, W. W. and Nero, A. V. (eds.). Radon and its Decay
Products. John Wiley & Sons, Inc., 1988.
2. Stranderi, E, Kolston, A. K. and Lind, B. Radon exhalation:
moisture and temperature dependence. Health Physics. 47: 480 -
484, 1984.
3. DSMA Atcon Ltd. Review of existing instrumentation and evaluation
of possibilities for research and development of instrumentation to
determine future levels of radon at a proposed building site.
INFO-0095, Atomic Energy Control Board, Ottawa, Ontario, 1983. 30
pp.
4. Tanner, A. B. Measurement of radon availability from soil. In:
M. A. Markos and R. H. Hansman (eds). Geological Causes of Natural
Radionuclide Anomalies. Missouri Department of Natural Resources,
Division of Geology and Land Survey. Spec. Pub. No. 4, 1987. p.
139 - 146.
5. Kunz, C., Laymon, C. A. and Parker, C. Gravelly soils and indoor
radon. Presented at EPA Symposium on Radon and Radon Reduction
Technology, Denver, CO. Oct. 17 - 21, 1988.
6. Azzouz, H., Geological parameters in the estimation of indoor
radon potential in Los Angeles, Ventura and Santa Barbara Counties,
California. Unpub. MSc thesis, UCLA. 1990.
7. Reimer, G. M. Reconnaissance techniques for determining soil-gas
radon concentrations: An example from Prince Georges County,
Maryland. Geophysical Research Letters. 17: 809 - 812. 1990.
8. Sextro, R. G., Moed, B. A., Nazaroff, W. W., Revzan, K. L. and
Nero, A. V. Investigations of soil as a source of indoor radon.
Jja: P- Hopke (ed.), Radon and its Decay Products: Occurrence,
Properties and Health Effects. American Chemical Society,
Washington, D.C., 1987. p. 10 -29.
9. Burns, M. Radon may lurk in 2,000 homes, tests find. Santa
Barbara News-Press. February 4, 1991.
10. Burns, M. Radon tests urged for South Coast. Santa Barbara
News-Press. April 18, 1991.
9-73
-------
Session X
Oral Presentations
Radon in Schools and Large Buildings
10- 1
-------
SEASONAL VARIATION IN SHORT-TERM AND LONG-TERM
RADON MEASUREMENTS IN SCHOOLS
Anita Schmidt1, John T. MacWaters2, and Harry Chmelynski2
1U.S. Environmental Protection Agency
Washington, D.C. 20460
2S. Cohen and Associates, Inc.
McLean, VA 22102
ABSTRACT
During the summer of 1989, the Environmental Protection Agency initiated Phase II of
the School Protocol Development Study, a year-long, in-depth radon measurement
study conducted in school buildings. The purpose of the study has been to gather
additional data to support and to refine the Agency's guidance and procedures for
measuring radon in schools, entitled "Radon Measurements in Schools - An Interim
Report'.
Radon measurements were made in 21 public schools (in 7 States) selected from 130
schools previously tested in the winter of I989 (Phase I). In each school, radon levels
were measured in all frequently occupied rooms in contact with the ground. Both
short- and long-term seasonal measurements were made using a variety of radon
measurement devices.
Results and analyses of the seasonal variation in short- and long-term tests conducted
in this study are presented and discussed.
This paper has been reviewed in accordance with the U.S. Environmental Protection
Agency's peer and administrative review policies and approved for presentation and
publication.
10- 3
-------
INTRODUCTION
In April 1989, the Environmental Protection Agency (EPA) released interim guidance
on radon measurement approaches for schools (1). The guidance and procedures
outlined in the report were based primarily on data from an intensive study of radon
conducted in I988 in 5 schools in Fairfax County, Virginia (2). Prior to development of
this document, there was little information available to school officials on procedures
for radon measurement in schools.
In early I989, EPA also initiated the School Protocol Development Study in order to
refine this interim report into a final guidance document. This study has consisted of
two phases. Phase I, conducted in February I989, was a screening study of 130
schools in 16 States (3). This was not a statistically representative survey of radon
levels in schools. The schools were not randomly selected, but rather were selected
based on factors such as geographic location, proximity to areas with elevated radon
levels, and school interest in participation. All frequently occupied rooms in contact
with the ground were screened for radon.
Based on the screening results, as well as other criteria, 21 schools in 7 States
were selected for Phase II. This paper describes a portion of the various results and
findings from this study. The results of other measurements will be presented in a
subsequent paper.
STUDY OBJECTIVES
Phase II of the School Protocol Development Study evaluated various short- and
long-term, seasonal radon measurements and various factors that may influence these
measurements (4). Of particular interest is the extent of seasonal variation in
measurements and the implications for testing recommendations. The Agency's
current guidance recommends that schools test "in the colder months (October
through March) when windows and doors as well as interior room doors are more
likely to be closed and the heating system is operating"{1). When the interim
guidance was published, this recommendation was based primarily on what was
known about testing houses under "closed conditions-, with limited information
available on schools.
This study specifically gathered data in order to evaluate: the seasonal variation in
shorter term measurements from 2-day open-faced charcoal canisters (OF-CC); the
seasonal variation in a longer term measurement from 3-morrth alpha track detectors
(ATD); and the seasonal variation in the relationship of these two measurements.
These evaluations are presented in this paper.
10- 4
-------
The 21 schools were located in:
New Jersey - 3 schools
Minnesota - 4 schools
Kansas City - 4 schools
KS/MO
Georgia - 3 schools
New Mexico - 3 schools
(7 small buildings)
Washington - 4 schools
In June 1989, 3-month ATD measurements were begun according to the following
schedule. ATDs were placed in all frequently occupied rooms in contact with the
ground.
Summer I989 June-August (schools
unoccupied)
Fall I989 September-November
Winter 1989-90 December-February
Spring I990 March-May
Short-term measurements with 2-day OF-CC were made during the same four
seasons in the same rooms with ATDs. Three (3) rounds of "2-day" tests were made
over a 9-day period: on two weekends and during the intervening week. Each round
covered approximately 64 hours, extending from 4 pm of the first day to 8 am on the
fourth day. This was done to minimize interference with classroom hours and to
accommodate normal weekend hours for school personnel conducting the
measurements. The 9-day periods for each season were conducted beginning:
Summer I989
Fall I989
Winter 1989-90
Spring
NJ
7/21
10/27
1/5
3/30
GA
7/14
missing
1/9
5/11
MN
7/28
11/12
2/2
5/4
NM
7/21
10/20
1/19
4/20
KS/MO
7/21
10/13
212
4/20
WA
8/4
10/20
1/19
4/20
The purpose of the 3 rounds was to study the influence of school ventilation systems
on short-term radon measurements by operating these systems differently during each
round:
Round 1 Weekend
To measure the effect of minimum ventilation on radon concentrations, all
ventilation systems are turned off, or set to minimum operation, during
the day with a normal night-time set-back of the system
[Note: Most of the schools had provision (manual or clock-operated
switches) for "night-time set-back" which keeps the building above some
10- 5
-------
minimum temperature (e.g., 55 degrees) to prevent freezing and to allow
quick morning warm-up. This set-back usually includes exhaust fans.
Night-time set-back is usually used on weekends as well.]
Round 2 Weekdays
To measure a weekday period with normal school activity, the ventilation
systems are operated on a normal, day-time setting, including normal
night-time set-back of the system.
Round 3 Weekend
To measure the effect of maximum ventilation on radon concentrations,
the ventilation systems are operated on a normal daytime setting, without
the normal night-time set-back.
* Round 3 is equivalent to the recommended 2-day weekend approach
outlined in the EPA's interim guidance document on radon
measurements in schools.
MEASUREMENT RESULTS AND ANALYSIS OF SEASONAL VARIATION
Measurement results used in this evaluation of seasonal variation are summarized
and presented in Tables 1 and 2 for ATDs and Tables 3 and 4 for OF-CC. For OF-
CC, note that the number of rooms varies from season to season because of
occasional problems with operation of ventilation systems in a few schools. For OF-
CC and for ATDs there were also small amounts of missing data in all States in all
seasons due to a variety of reasons.
ATP Results
A summary of the seasonal ATD data for each State is presented in Table 1 and a
summary by school in each State is in Table 2. Two types of seasonal comparisons
were done. First an analysis was done to determine if measurements within a State
varied significantly from season to season, and, secondly, to determine to what extent
measurements within a State for a given season deviated from a 4-season annual
average measurement and from a 3-season average representing the school year (i.e.,
no summer measurements included).
For the first analysis, the results of a non-parametric one-way Analysis of Variance
(ANOVA) test for significant differences by season are shown in the final column of
Tables 1 and 2. This was done school-by-school and State-by-State. The one-way
ANOVA test compares variation of room measurements for a State within each season
to the observed variations across 4 seasons (or for 3 seasons) for that State. If the
within-season variation is approximately equal to or greater than the across-season
variation, the test reports that the seasonal means are ogj significantly different.
Alternatively, if the within-season variation is small compared to the across-season
10- 6
-------
variation, the test reports that the observed difference in season means is significant.
(A 0.05 probability level is used for all ANOVA tests.)
Tables 1 and 2 show that there was significant seasonal variation in radon
measurements using ATDs for most of the schools and States.
For the second analysis, a 4-season (and 3-season) average of school
measurements (weighted by the number of rooms tested in each season), was
computed for each State. The percent deviation of each seasonal average (e.g.,
winter) for a State from the 4-season (and 3-season) ATD average for that State is
shown in Figures 1 and 2. (Note: data from schools in Georgia were not included in
the analyses for Figures 1 and 2 because data from the fall season were not available
to compute an across season average.)
An additional line has been added to Figures 1 and 2 to represent the average
percent deviation for each season from the 4-season (and 3-season) average across 6
States. The seasonal average was computed by averaging the percent deviation for
each State in that season.
In Figure 1, there is no single, distinct pattern of seasonal variation present. The
measurements in New Jersey school rooms were lower in the summer (-8%) and fall
(-8%) than the annual average from four seasons of measurements for the schools in
that State, but were higher in the winter (12%) and spring (+4%). For Kansas City
schools in Kansas and Missouri, the summer and winter measurements were 12% and
8% higher, respectively, than the annual average. In the winter and spring, the
measurements were 10% and 9% below the average. In Washington, measurements
were lower in summer (-20%) than the 4-season average, but higher in fall (+18%)
and winter (+4%), and lower once again in spring (-2%). A similar pattern is present
for the New Mexico schools. For Minnesota, measurements alternated from lower to
higher than the 4-season average in summer and fall, respectively, but were only 1%
below the annual average in winter and spring.
In Figure 2, where seasonal averages are compared to a 3-season, school year
average, there is no single seasonal pattern. However, as in Figure 1, there appears
to be an overall trend (for 6 States combined) of higher fail and lower spring
measurements, with winter measurements not consistently above or below the long-
term averages. Overall summer measurements in Figure 1 were lower than the annual
average.
Another general observation can be made from Figures 1 and 2, for all States in
the study, the variations in seasonal radon averages are within a range of
approximately ± 20 % of the 4-season and the 3-season averages.
Qpen-Faced Charcoal Canister Results
A summary of the seasonal OF-CC data is presented in Table 3 by State and in
10- 7
-------
Table 4 by school. These data were analyzed in the same manner as the ATD data.
The results of a non-parametric one-way ANOVA test for significant difference by
season are shown in the final column of both tables. These results also show that, for
the most part, there is significant seasonal variation in "2-day" measurements by
school and by State.
The measurement data were next analyzed for the percent deviation of
measurements from each State by season from a 3-season average (school year) for
that State. For the purpose of investigating seasonal variation, only results from
Round 3 are presented (Figure 3). This is the round that approximates the current
recommendation in EPA's school guidance (weekend, with weekday HVAC settings,
no night set-back). As was observed for the seasonal data from ATDs. there is no
specific pattern of seasonal variation for OF-CC that applies to all schools or to all
States measured in this study. Data for Washington and Missouri show deviations
from the 3-season average for Round 3 going in opposite directions in fall and spring.
For these two States, the range of variation in seasonal averages from a 3-season
average is larger than for ATDs, approximately +. 50% for Washington in fall and
spring and -23% for fall and +40% in spring for the two Kansas City schools in
Missouri. (Note: insufficient data were available from Round 3 for the Kansas City
schools in Kansas and schools in Georgia.) For New Jersey, Minnesota, and New
Mexico the range of variation for all 3 seasons is somewhat smaller, +10% and -16%.
Overall, the smallest variation for schools in these 5 States was in winter, with a range
of +9% to -16%. As in Figure 2, there is an overall trend of higher measurements in
the fall relative to a 3-season average, and lower in the spring.
Figure 4 examines the seasonal variation of Round 3 data from the 3-season
average using ATD measurements. For each school room with a complete data set,
the percent deviation of the OF-CC measurement in each season from the 3-season
average was calculated. These room-by-room deviations were then averaged for each
season for each State. (Note: complete data were not available for Kansas, Missouri,
or Georgia.) These results are very similar to those in Figure 3. Again there is no
single pattern for all States. Based on the combined data from the 4 States, Round 3
OF-CC results were similar to the long-term average for fall and winter. The spring
measurements with OF-CC for all 4 States are lower than the 3-season ATD average.
Seasonal Variation in the Ratio of Open-faced Canister and ATD Results
The deployment of collocated "2-day" canisters and 3-month ATDs in a large
number of rooms provided the data needed to compare a set of short-term and long-
term radon measurements from schools. This comparison is of interest because It is
useful to know to what extent a 2-day measurement may under or over estimate a
longer term measurement. In addition these are two of the most commonly used
devices in schools which makes the comparison of further interest.
The measurements were directly compared for each season by calculating the
ratio of the '2-day" to the 3-month ATD results for each room, for each round, for each
10- 8
-------
season. In general, for the resulting ratios for each round for each season, there was
a wide spread in the "2-day"/3-month ratios in rooms with low radon levels, while at
higher radon levels the spread was much smaller. The mean ratio of all
measurements by season was then calculated. If the average ratio exceeds 1.0. this
indicates that the short-term measurements for that season were higher, on average,
than the 3-month measurements for that season. Ratios less than 1.0 indicate that the
"2-day" measurements were lower than the 3-month measurements on average.
Average ratios for each of the 3 rounds, for each season, are presented in Figure 5.
In Table 5, the average ratios are presented along with additional information on the
number of room measurements and the standard deviations for the ratios.
In Figure 5, the average ratios of short to long-term measurements were generally
much greater than 1.0 in summer (20% to 93% above). In the winter, the mean ratios
for Rounds 2 and 3 are very close to 1.0 (1% below, 7% above). In the spring,
Rounds 2 and 3 OF-CC measurements are 23% and 17%, respectively, lower than the
ATD results for that season. In the fall, Rounds 2 and 3 are 16% below and 7% above
the ATD results. This suggests that these "2-day" results, conducted under normal
weekday, occupied conditions (Round 2) or under maximum ventilation during
unoccupied conditions (Round 3) are very similar on average, especially in winter, to
the longer term measurements in each season or across all 4 season as indicated by
the weighted yearly averages. Averaged over all seasons, Round 1 canisters were 50
% higher than the 3-month ATDs. In Round 2 and 3 the canisters are only 1 to 5 %
higher than the ATDs across all seasons.
Quality Assurance
The precision and accuracy of the OF-CC and ATDs was measured by the use of
a number of quality control detectors. For charcoal canisters each school had 10
percent duplicates samples and 5 percent blanks. For ATDs each school had 20
percent duplicates and 5 percent blanks. The devices were placed side by side over
identical measurement intervals in each season. Charcoal canisters and ATDs were
also exposed in chambers with known concentrations of radon and submitted to the
analysis laboratory without being identified as spiked samples.
The overall quality control results for the study are good. The precision of
duplicate canisters is considered very good, with an average coefficient of variation
(COV) of 6%. This compares favorably with the suggested 10% in EPA's
measurement protocols (5). The COV for duplicate ATDs was 6% for fall and winter
and 19% for spring. These results are particularly good for fall and writer and
considered acceptable for the spring. EPA's measurement protocols for ATDs
recommend that the COV for precision should not exceed 20%.
The overall accuracy of these devices in this study were very good as measured
by field blank detectors and spiked detectors. For OF-CC field blanks, the average
radon level was below 0.1 pCi/L For ATD field blanks, the average radon level was
below 0.8 pCi/L Based on results from spiked devices, the accuracy of the OF-CC is
10- 9
-------
considered very good (average bias of +5%). For ATDs, the bias ranged from good
(-5%) to acceptable (-19%).
CONCLUSIONS
This assessment of seasonal variation using "2-day" and 3-month measurements
indicates the following:
1. Seasonal variation was observed in school radon measurements from season to
season for both "2-day" and 3-month results. This variation was significant for most of
the schools and States in the study (Tables 1-4). This seasonal variation was present
even though the radon levels were generally under 4 pCi/L in most of the schools.
2. There is no clear, single, pattern to this seasonal variation in the data across all the
schools or States for either OF-CC or ATD results. !n Figures 1 -4, the average
percent deviation of measurements for a given season from a long-term average vary
from State to State, season to season. Patterns for any two States may go in
opposite directions. Tables 2 and 4 also indicate that an apparent seasonal pattern
for a State may not be consistently observed at the school level for that State. Given
the wide geographical range and climatic variation, the lack of a well-defined seasonal
pattern may not be unexpected.
3. In Figures 1 -4, there is a genera! seasonal trend observed when all data for a
season are combined across States. On average, summer and spring measurements
for both OF-CC and ATDs tend to be lower than the annual or school year averages,
higher in the fall, and close to the long-term averages in the winter. However, New
Jersey and Missouri measurements did not follow this pattern for fall and spring.
4. Figure 5 suggests that a "2-day" radon measurement taken in the winter months,
on average, may approximate a 3-month measurement. An exception is
measurements made under conditions of minimum ventilation (Round 1). The "2-day"
measurements in the fall also were dose to the 3-month average radon levels. In the
summer, the OF-CC measurements may over predict what a 3-month ATD
measurement would indicate.
Figures 1-5 suggest that measurements taken in the winter months of the year
may provide, on average, the best measurement of school year (or annual) average
radon levels. These results may be due to the more uniform and stable conditions in
winter compared to fall and spring. These latter two seasons usually span both cold
and warm weather. Therefore, during some weeks of warm weather, windows and
doors may be open (heating systems not operating). During other weeks of much
colder weather, schools may be more closed up (heating systems are operating). In
the winter months, the heating systems in these schools were operated routinely and
doors were likely to be closed more consistently over this 3 month period than in fall
and spring. The reason for the overall higher measurements in the fall compared to
10-10
-------
the lower results for the spring relative to the longer term averages is not readiiy
apparent.
These evaluations are generally supportive of EPA's current recommendation to
test schools during the colder months of the year. Perhaps a more definitive
recommendation would be to test schools during the coldest months of the year for
that geographic area when the heating system of the school is routinely operated.
These results also suggest that winter measurements, whether taken over a few days
or 3 months, may on average be a good indicator of the school year average radon
levels in a school.
ACKNOWLEDGEMENTS
The authors would like to recognize the contributions of the school, State, and EPA
Regional personnel in the planning and conduct of this study, and without whom this
project would not have been possible.
REFERENCES
1) U.S.E.P.A. Radon Measurements in School: An Interim Report, EPA
520/1-89-010, March 1989, U.S. Environmental Protection Agency, Office
of Radiation Programs, Washington, D.C. 20460.
2) Peake, R.T., MacWaters, J., Chmelynski, H., Mollyn, G., and
Maconaughey, K. Radon and radon progeny measurements in five
schools, in: Proceedings of the 1988 Symposium on Radon and Radon
Reduction Technology, Denver, CO. Vol. 1, EPA 60iO/9-89-006A.
3) Peake, R.T., Schmidt, A., MacWaters, J., and Chmelynski, H. Radon
measurements in 130 schools: Results and implications. In:
Proceedings of the 1990 Symposium on Radon and Radon Reduction
Technology, Atlanta, GA. EPA 600/9-90/005.
4) Schmidt, A., Peake, R.T., MacWaters, J., and Chmelynski, H. EPA's
school protocol development study-Phase II. In: Proceeding of the
1990 Symposium on Radon and Radon Reduction Technology, Atlanta,
GA. EPA 600/9-90/005.
5) U.S.E.P.A. Indoor Radon and Radon Decay Product Measurement Protocols.
EPA 520-1/89-009, March I989, U.S. Environmental Protoection Agency,
Office of Radiation Programs, Washington, D.C. 20460.
10-11
-------
TABLE 1. ANALYSIS OF SEASONAL VARIATION IN ROOM RADON LEVELS
BY STATE USING ATD MEASUREMENTS
I
« I
' i
* i
T I
t I
seasonal meau (pci/D aio or koms cm
W*
-------
TABLE 2. ANALYSIS OF SEASONAL VARIATION IN ROOM RADON LEVELS
BY STATE AND BY SCHOOL USING ATD MEASUREMENTS
J * 1 I
T c 1 I
* * | SEASONAL KEA* (pCi/t) AJfC MUK8E1 Of WOS CN} j A*E D! FTESENCS ACISSS
T 0 I I SEASONS SISUIflGUfT
I 0 I SUWEI rut WIlTII SPIIkC I AT THE 0.05 tfVEL?
L M MEAN H MCA* • KAM « MEAN |
1 1
35
1.53
40
1.41
39
1.59
39
1.52 |
HO
2 j
31
1.85
31
2.03
31
2.76
18
3.22 |
TES
3 |
34
0.82
34
0.78
35
0.88
33
0.80 j
no
1 }
50
2.84
50
1.73
47
1.70 |
TES
2 |
40
1.96
4C
1.92
.
38
1.64 |
¦O
3 |
39
6.84
23
7.76
•
•
36
4.03 |
TES
0 |
11
0.63
16
1.09
17
0.82
17
1.22 |
TES
2 |
21
1.17
30
1.79
31
1.13
31
1.31 |
TES
4 |
20
1.97
22
2.49
22
2.46
22
2.10 |
HO
6 {
27
1.Z3
•
•
30
1.31
37
1.90 |
TES
1 |
36
0.83
30
1.30
35
1.33
32
0.99 |
TES
2 j
21
1.16
20
2.25
20
2.02
19
1.44 |
TES
3 I
6
0.87
9
1.39
9
1.24
9
1.07 |
TES
4 |
20
1.24
19
1.27
15
1.42
11
1.15 |
HO
5 j
6
1.23
7
0.94
8
0.73
8
0.55 |
TIS
6 |
6
1.42
5
1.80
.
.
6
1.50 j
MO
7 j
6
1.10
6
0.62
•
•
6
0.67 |
MO
2 |
17
1.30
17
1.48
16
1.24
16
1.25 |
MO
5 |
3i
3.08
31
2.64
34
1.95
37
2.22 |
TIS
® I
20
1.10
21
1.28
20
1.06
21-
0.90 |
MO
2 |
11
1.«
17
1.52
17
1.J9
18
1.16 |
MO
a |
20
2.41
20
3.18
20
3.18
17
2.82 |
MO
s |
It
2.96
19
3.00
21
1.5)
20
0.72 |
TIS
* j
23
^.n
22
2.90
23
3.13
22
3.55 |
MO
7 j
24
0.69
25
2.04
26
1.13
26
1.88 |
TES
(.) n{t«
-------
TABLE 3. ANALYSIS OF SEASONAL VARIATION IN ROOM RADON LEVELS
BY STATE AND BY ROUND USING OF-CC MEASUREMENTS
s a
t 0
I
I
I
SEASONAL «A* (pCI/l) ANO WMEI OF iOWS <*>
1
1
| m oirntCHCt Aotoss
A U
I
( SEASONS SICK)FJCANT
T N
| am*
FAIL
VtKTtt
speixc
j AT T« 0.05 LIVIL?
E 0
I -
N
WAN
¦
***
V
MEAN
1
UJ 1
[ 103
2.24
105
1.60
40
1.95
71
3.10
i TW
2
| 103
1.37
102
1.32
70
2.33
70
2.17
| TIS
3
| 104
1.42
101
1.52
105
1.78
106
1.63
1 no
OA 1
| 125
*.20
121
4.19
122
3.45
| TIS
2
I 125
5.6?
.
,
116
3.31
110
1.72
j TIS
3
I ™
3.07
•
•
119
3.17
87
0.88
| TIS
MN 1
| 104
2.47
1C6
1.69
103
1.86
106
1.98
| TI5
2
| ICS
1.73
106
1.40
106
1.48
106
0.88
j TtS
3
1 105
1.28
105
1.97
105
1.82
78
1.54
j TIS
MM 1
J 100
1.83
106
1.52
106
1.12
108
1.50
J TIS
2
10$
1.05
105
1.58
104
1.14
108
0.99
j ns
3
I «7
1.12
102
1.38
97
1.26
108
1.17
j no
rs 1
I 52
2.95
50
2.24
52.
2.91
36
3.15
1 no
2
I 52
2.74
52
2.02
51
2.07
37
1.67
| TIS
3
I 52
3.71
52
3.47
52
2.87
•
•
j TIS
MO 1
I 39
1.60
39
1.87
18
1.66
18
1.49
1 m
2
I 39
1.05
38
1.23
17
1.16
18
0.31
| TIS
3
I S'
1.07
39
1.73
37
1.71
18
3.47
| TIS
M I 1
I 91
2.37
89
1.73
70
1.59
54
2.60
| TIS
NO 2
I 51
2.01
90
1.88
*8
1.85
55
1.23
i TIS
3
1
2.58
fl
2.73
•9
2.39
18
3.47
I «0
UA 1
i «
4.42
87
3.52
13
s.ti
87
4.54
| fts
2
I "
2.34
•7
1.92
IS
2.55
86
1.28
| Tli
3
1 *
l.iO
87
IM
«S
1.(3
87
e.9o
i TIS
ALL 1
1 •"
3.3S
m
l.fi
S23
3.00
548
2.83
1 VII
2
1 *«
2.*a
m
1.17
$49
2.12
S35
1.35
| ft«
3
j 420
1.88
486
2.06
600
2.08
484
1.32
| fit
(,) Missing Data
10-14
-------
TABLE 4. ANALYSIS OF SEASONAL VARIATIONS IN ROOM RADON LEVELS
BY STATE, BY SCHOOL, AND BY ROUND
USING OF-CC MEASUREMENTS
I
0
y
«ŁASO*AL MUH CpCi/U ANC KMEI Of iOQMS (>!
UKI
II «AN
mi
I MEAN
VIKTft
i MEAN
$MJ*5
¦ ICAN
arc oif««*cs anoss
SEASONS StCklflCAIlT f
Hi
1 1
I 5S
2.61
39
1.94
40
1.95
40
1.85 |
K0
2
I «
1.69
39
1.12
40
1.66
40
1.50 j
MO
3
I *0
1.30
37
0.99
40
1.24
40
1.09 |
NO
2 t
I 31
2.94
31
2.15
•
•
31
4.72 |
Til
2
I »
1.14
29
2.47
30
3.23
SO
3.07 |
HO
3
I 31
2.29
31
3.14
31
3.65
31
3.21 |
HO
I.'
1 1
I 34
1.18
35
0.75
•
.
.
* 1
TIS
2
I *
0.60
34
0.37
•
•
»
• i
HO
1
I 33
0.74
33
0.19
34
0,72
35
0.86 |
TIS
a
1 1
I *9
4.77
49
2.51
49
2.40 |
TES
2
I V
6.19
«
45
1.92
35
1.22 j
TIS
3
I 49
2.40
m
•
48
2.33
47
0.S9 |
TIS
u
2 1
) 39
2.92
•
•
38
2.47
36
2.19 |
MO
2
I
3.18
¦
•
57
2.43
39
0.97 |
TtS
3
I «
1.77
*
34
1.13
-
* I
NO
a
3 1
I V
11.16
•
S4
8.51
17
6.06 |
its
2
I »
7.66
«
•
34
6.11
36
3.03 |
Ttl
3
I »
S.It
•
*
37
5.77
40
1.22 |
TES
MN
8 1
I w
2.61
17
e.«s
15
1.61
17
2.16 |
Ttl
2
i »
i.te
17
0.48
18
8.68
17
6.48 |
Til
3
J *
• .41
16
0.66
16
8.f1
17
0.98 |
TtS
m
I 1
1 »
2.26
SO
1.11
SO
1.37
SO
1.44 |
¦0
2
I »
1.12
SO
1.27
SO
1.12
SO
0.67 |
TIS
3
I 30
1.73
SO
2.01
SO
1,74
2
3.00 ;
¦0
(.) Missing Data
(c
o n r. i n i; e
'-D
10- 15
-------
TABLE 4 CONTINUED
s s » ! |
T C 0 I |
A N u I SEASONAL HEAN (pCi/l) AMD KWSEt OF tttWS {*) |
t o n | | a*e oirmfNCS acsoss
E 0 0 I SUWE2 FALL WINTER SPRING I SEASONS SiWiFICAkT 7
L I N MEAN M KAN N MEAN N MAN I
KN
4
1 1
21
3.30
22
2.74
22
2.96
22
2.25 |
YES
2 1
22
2.11
22
1.97
21
2.62
22
1.30 |
TIS
3 1
22
0.63
22
3.10
22
3.36
22
1.05 |
TIS
NX
6
1 I
37
2.20
37
1.64
36
1.69
37
2.18 |
NO
2 I
36
1.93
37
1.49
37
1.39
37
0.99 |
TIS
3 i
37
1.68
37
1.85
37
1.36
37
2.01 |
HO
NN
1
1 I
31
1.24
37
0.74
37
1.07
37
1.19 |
TIS
2 I
36
0.91
37
1.20
37
0.95
37
0.76 |
TES
3 I
36
1.67
34
0.93
36
1.13
37
0.72 |
TIS
**
2
1 1
22
3.07
22
3.36
22
1.76
22
1.39 |
TES
2 I
20
1.62
22
2.69
20
1.70
22
1.21 I
TES
3 i
22
0.58
21
3.35
22
2.04
22
2.53 |
TES
MM
3
1 1
7
1.70
7
1.67
8
1.11
9
0.87 |
TES
2 1
9
1.19
7
1.21
6
0.86
9
1.01 |
TIS
3 1
9
0.84
7
1.51
•
«
9
1.21 |
TES
MM
4
1 1
20
1.60
20
1.H
19
0.92
20
2.75 |
TES
2 1
20
0.12
19
1.57
19
1.40
20
0.97 |
TES
3 1
20
0.52
20
0.72
20
1.09
20
0.92 |
TIS
MM
5
1 1
1
1.55
•
1.14
•
0.41
t
0.44 |
TES
2 1
S
0.69
I
0.65
•
0.58
8
0.8S |
ns
3 1
•
0.54
«
0.40
•
0.60
8
0.26 |
TIS
m
6
1 I
*
2.22
6
2.17
*
1.27
6
2.50 |
TfS
2 I
6
1.33
*
1.10
*
1.58
6
2.12 |
nt
3 I
*
1.85
6
1.38
6
1.45
6
1.38 J
m
( C ft lit i n U C: (i )
10- 16
-------
TABLE 4 CONTINUED
« ¦ I I
CO) |
N U j SU.$dML KŁAk (pCi/1) AW KMft Of »0C*S («) |
0 « | [ AXE DI"CIŁnŁS AOIOSS
o p | Kmi »au wirrn vimc [ sums iig*imca*t i
L I H (CJL* I MEAN I HUH I (CAM I
m
7
i
1
4
1.47
8
0.47
6
0.40
4
8.98 |
us
2
i
6
0.72
6
0.60
6
0.38
6
0.78 |
ns
3
1
*
1.02
4
0.45
5
0.38
6
CM |
TIS
(S
2
1
1
15
1.38
14
1.06
15
1.53
.
• I
ns
2
1
15
0.73
15
1.20
u
1.09
•
' I
TES
3
1
15
1.S0
15
2.73
13
1.88
*
• I
res
cs
3
1
1
37
3.59
36
2.7C
37
3.47
34
3.15 |
NC
2
1
37
3.35
37
2.35
37
2.45
37
1.67 |
TIS
3
1
37
4.61
37
3.76
37
3.27
1
TES
MC
0
1
1
21
1.37
21
0.54
•
*
•
' 1
TIS
2
1
21
0.9C
20
1.35
•
•
•
' 1
ao
3
1
21
1.01
21
1.60
20
1-29
*
• I
MO
PC
2
1
1
18
1.88
18
1.48
18
1.66
18
1.49 1
•0
2
i
It
1.Z2
18
1.09
17
1.16
18
0.31 |
ns
3
1
1S
1.14
18
1.93
17
2.21
18
3.47 |
NO
UA
0
1
I
20
4.4«
20
4.94
18
4.07
19
5.48 |
*0
2
1
19
e.#9
20
2.7»
19
2.57
19
2.04 |
ns
S
1
19
t.n
20
t.M
18
•.81
19
8.39 |
ns
w
s
1
1
20
7.48
20
1.11
19
3.29
20
8.42 |
Til
2
1
1
20
4.24
20
1.S4
20
1.44
20
8.93 |
m
3
19
z.n
19
1.89
20
4.81
20
8.59 |
Til
IM
«
1
1
23
3.80
22
1.81
Z3
i.99
23
8.U |
iti
2
1
a
2.57
23
2.19
22
4-47
a
1J0 {
Til
J
1
23
8.41
23
4.87
22
2.15
a
2.02 |
ns
W
r
1
1
2S
2.49
23
2.82
Z3
4.88
a
J-H 1
ns
2
t
24
1.84
24
1.24
24
1-33
24
8.94 |
ns
J
1
25
e.77
23
3.04
2S
1.55
a
SJ4 |
ns
(.) Missing Data
10- 1.7
-------
TABLE 5. SEASONAL RATIOS OF OF-C
SEASON ROUND RATIO MEAN RATIO STD * FAIRS
SUMKER 1 1,93 1.47 571
2 1.35 1.05 574
3 1.20 1.16 577
FALL 1 1.04 .44 433
2 .84 .40 431
3 1.07 .59 428
WINTER 1 1-44 .77 368
2 .99 .39 398
3 1.07 .69 448
SPRING 1 1.4 5 .81 507
2 .77 .45 495
3 .83 .84 441
10-18
-------
©
A
V
E
R
A
G
E
P
E
R
C
E
N
T
D
E
V
I
A
T
I
O
N
50
Data were incomplete for Georgia.
-50
-100
17 18 18
-22 "20
NJ
KS/MO
6 STATES COMBINE
D
SUMMER
FALL
WINTER
SPRING
FIGURE 1. Seasonal Variation of ATD Schoolroom
Measurements from 4-Season ATD Average
-------
N>
©
A
V
E
R
A
G
E
P
E
R
C
E
N
T
D
E
V
I
A
T
I
O
N
50
0
-50
-100
Data were incomplete for Georgia.
MN
[Oil WA
NJ
KS/MO
fc- 6 STATES COMBINE
D
FALL
WINTER
SPRING
FIGURE 2. Seasonal Variation of ATD Schoolroom
Measurements from 3-Season ATD Average
-------
©
to
A
V
E
R
A
G
E
P
E
R
C
E
N
T
D
E
V
I
A
T
I
0
N
50
-50
-100
Test conditions: HvAC systems a 11 on
weekend measurements.
Data were incomplete for QA and KS
FALL
WINTER
NM
5 STATES COMBINE
SPRING
D
FIGURE 3. Seasonal Variation of Round 3 OF-CC
Schoolroom Measurements from 3-Season
OF-CC Average
-------
©
K)
A
V
E
R
A
G
E
P
E
R
C
E
N
T
D
E
V
I
A
T
I
O
N
50
-50
-100
Test conditions: H\MC systems all on;
weekend measurements.
Data were incomplete for QA and KS/MO.
25
] NM
-65
WA 4 STATES COMBINED
FALL
WINTER
SPRING
FIGURE 4. Seasonal Variation of Round 3 OF-CC
Schoolroom Measurements from 3-Season
ATD Average
-------
2.2
1.93
2.0
1.8
1.6
1.50
1.45
1.44
1.35
1.4
1.2
1.2
1.07
1.07
1.05
1.04
1.01
0.99
1.0
0.84
0.83
0.77
0.8
0.6
0.4
0.2
0.0
SUMMER
WINTER
SPRING
WEIGHTED AVERAGE
EZ1 round i X777X round 2 K3 round 3
FIGURE 5. MEAN SHORT-TERM / LONG-TERM RATIO OF OF-CC
MEASUREMENTS TO ATD MEASUREMENTS BY ROUND, BY SEASON
-------
Diagnostic Evaluations of Twenty Six US Schools -
EPA's School Evaluation Program
Gene Fisher
Bob Thompson
EPA
Office of Radiation Programs
Washington, D.C.
Terry Brennan
Camroden Associates
Oriskany, New York 13424
William Turner
H.L. Turner Group
Harrison, Maine 04055
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.
ABSTRACT
As part of a coordinated radon in schools technology development effort,
EPA's School Evaluation Team has performed on-site evaluations of twenty six
schools in eight regional locations throughout the United States. This paper
presents the results and preliminary conclusions of these evaluations. This
represents the largest data bank of schools that have been diagnosed with
consideration for interactions of the building with both sub-slab and heating,
ventilating and air conditioning (HVAC) characteristics. Occupied classroom carbon
dioxide concentrations and building shell tightness are reported. These
measurements help to judge the existing outside air ventilation rates and the
potential for building pressurization. Besides these technical issues, physical and
institutional problems that affect the selection and implementation of radon control
systems in schools are identified. Both soil depressurization and use of existing
HVAC equipment were evaluated as mitigation approaches for each school. Results
of this two year study suggest that the EPA should consider a new direction in large
building radon abatement - a holistic approach that considers the broader issues of
indoor air quality, comfort, cost and energy issues.
KEY WORDS: Radon, Holistic, Schools, Large Buildings, Airtightness, Ventilation, Outside
Air, EPA, Carbon Dioxide, IAQ, Indoor Air Quality, Diagnostics, Measurements.
10-25
-------
INTRODUCTION
The School Evaluation Program (SEP) was originally conceived in the
summer of 1989 as a technical assistance program in response to an emerging need
for information on diagnostic techniques and mitigation strategies applicable in
schools with elevated levels of radon. The program was designed to incorporate all
appropriate state of the art radon diagnostic procedures that weresuccessfully being
used in residential investigations. Occupant density is approximately 7 times greater
for schools than for residential settings. Therefore, recommended or mandated
ventilation rates for school rooms are several times that for residences. Accordingly,
industry accepted evaluation methods for non-residential air handling systems were
incorporated into the program.
The original goal of the SEP was to develop school diagnostic procedures that
would consistently provide sufficient information to enable school officials and
private sector contractors to choose the most effective mitigation strategy for their
school. When completed, the diagnostic procedures will become part of EPA's
Technology Transfer Program, and will be made available for training purposes
through EPA funded Radon Regional Training Centers. Information collected also
would be used to update the EPA school mitigation guidance publication.
The selection of SEP schools was based on four criteria:
1. Schools with radon screening measurements greater than 4 pCi/L.
2. Geographic location/climatic conditions.
3. Structure Type
4. A willingness by the school to mitigate based on the results of the
evaluation and the recommended remediation strategy.
As a result of the selection process three schools in Washington state and
three schools in New Mexico were evaluated, and results were reported at the 1990
International Radon Symposium [1]. In fiscal 1990, twenty additional schools were
evaluated: three each in the states of Georgia, Iowa, North Dakota, Illinois, and New
Jersey, and five in Maine.
The SEP Team in 1990 consisted of the authors of this paper - a team that
represents expertise in the areas of residential radon investigation and large
building HVAC design, installation, and operation.
The evaluations performed on all twenty six (26) schools have strengthened
the investigators opinion and judgement that:
• Schools are more complicated than residential homes in every respect.
• All aspects of the building dynamics and purpose must be taken into
account when selecting and planning a radon control strategy.
• Low ventilation rates appear to be a serious problem in many of the
nations classrooms.
10- 26
-------
An optimum control strategy would depend on each schools' specific needs
and would incorporate active sub slab depressurization and HVAC mitigation
techniques, independently or collectively, as required. Through extensive field
experience gained by the SEP, a holistic approach to radon control and general
indoor air quality in schools became apparent.
METHODOLOGY
The conceptual model for radon entry into buildings has directed field
investigations of indoor radon problems toward two areas. The first area is the ways
in which the foundation and underlying materials affect radon entry and could be
used to control entry by soil depressurization. The second area is an investigation of
the ways in which the operation of the mechanical equipment in the building affects
radon entry and could potentially be used to control radon.
Both approaches, HVAC and soil depressurization, prevent soil air entry by
managing the air pressure differential relationships between the air in the soil and
the air in the building. In addition, dilution by increased ventilation can have a
significant role in the HVAC approach and sometimes plays a minor role in a soil
depressurization approach.
RADON ENTRY INVESTIGATION
The radon entry investigation is divided into three parts, review of radon
measurements, identification of radon entry points, and soil depressurization tests.
Each of these are summarized in the following paragraphs.
Radon Measurement Data
Radon measurement data for the schools was obtained through the screening
results taken as part of EPA's Phase I and Phase II school survey [2 & 3]. Several
schools had initiated their own testing program and offered their measurement
results for acceptance into the SEP. Most of the schools had tested many of their
classrooms using the current EPA radon measurement protocol. A number of
schools had performed confirmation measurements using alpha track detectors.
Radon measurements in schools have been found to vary considerably both
spatially and temporally [1]. Figure 1 shows the mean classroom radon levels in the
schools evaluated. Error bars of one standard deviation are given to illustrate room
to room variation for simultaneous measurements. Temporal variation can be
seen by using continuous radon monitors or by the sequential use of passive
integrating monitors. These topics will be discussed later in the paper. Both
temporal and spatial variation are the result of varying radium concentrations,
locations, and transport pathways under the building. Also, variations are caused by
air pressure differences resulting from the dynamic interaction of occupants,
building, mechanical equipment and outside weather conditions [4].
10- 27
-------
The maintenance personnel and school safety officers proved to be invaluable assets
in the investigation of every school. They provided a wealth of information about
the operation and maintenance of the school building and the mechanical
equipment within the a school. Often they could provide blueprints for the school
and in all cases were able to supply fire exit plans.
The radon screening measurements for a school were plotted room by room
on a floorplan of the school, such as the fire exit floorplans available for most public
buildings. By plotting radon screening measurements on the fire exit plans, the
pattern of radon concentrations could be studied. Often this was not particularly
enlightening. For example, in many schools, screening measurements ranged
between 2 and 8 pCi/L. Given the temporal variation of school room radon levels,
it is difficult to make a meaningful distinction between a 3 pCi/L screening
measurement and a 6 pCi/L screening measurement. This amount of difference
might easily occur in the same room at different times.
In some schools, the pattern of radon levels was more helpful. For example,
a wing in a school that has relatively uniform, elevated concentrations in all the
rooms might well have a widely dispersed entry mechanism that the other wings do
not. Anecdotal observations included: 1) the air handler that supplied conditioned
air to the classrooms in one zone was drawing in soil gases and delivering it with
the conditioned air, 2) all wings are built on a permeable sand but two of them have
stone pebbles beneath the slab, creating a dilution break at that layer, while one wing
has a slab poured directly on the site material providing a radon source at each hole
in the slab. Clues like these formed the basis for the radon source and entry
diagnostics, to be covered in the next section.
Radon Entry Points
The predominant school foundation type was slab on grade, with several of these
having perimeter and internal utility tunnels under the slab. A few schools had
crawlspaces and basements. The major entry points of the SEP schools were:
• joints at the edge of slabs
• water pipe penetrations and drains
• trenches containing heating pipe penetrations
• trenches used as conditioned air supply and returns
• crawlspaces with all of the above
The radon concentrations beneath the slabs of the schools were measured.
The results are shown in Figure 2, with a summary of the HVAC equipment in the
schools. As can be seen, levels varied from a low of 200 pCi/L to a high of 8000 pC/L
with a mean of 1500 pCi/L. These are relatively low sub slab concentrations when
compared to sub slab measurements in some residential buildings with elevated
radon.
10-28
-------
Soil Depressurization Tests
In order to assess the potential for radon control by soil depressurization, a
vacuum suction test was made in most schools. This test allowed visual
identification of the sub slab material and measurement of how easily air could be
drawn from beneath the slab. Figure 2 has a column that lists the type of sub-slab
material and another that lists the amount of air that could be pulled from under
the slab using the vacuum. Generally speaking, the lower the amount of air that can
be drawn from under the slab, the more difficult it is to extend a low pressure field
beneath the slab [5]. Low airflows generally indicate more suction points are needed
for a successful soil depressurization system. In a few schools the vacuum could not
compete with the suction exerted on the building by the exhaust fans.
HVAC CHARACTERIZATION
System Operation
Review of mechanical system plans is the first step in evaluating the expected
effect of the HVAC system's effect on radon. Equipment observations must be
performed in order to determine what equipment is actually being operated, its
operation schedule, the control sequence, and whether the air flows are near the
design quantities. Air flow observations may be as simple as visually observing the
position of an outside air damper blade or tile operation of an exhaust fan and the
direction of air flow with a chemical smoke pencil, or may be more involved.
Measurement of actual exhaust flow rates with an air balancing hood has proven
valuable for determining the operational status of exhaust systems and for
conducting a building shell tightness test. Observations of closed outside air intakes
or inoperative make-up air supply fans were typical indicators of a HVAC system
not being operated or maintained as designed [5].
The HVAC systems in the SEP schools were characterized in terms of heating
and cooling, ventilation, and control strategy, and the resulting potential impact of
these parameters on indoor radon levels. A wide variety of ventilation systems
were found in the school buildings. The HVAC features are summarized in Figure
2.
A summary of the HVAC system types is as follows:
• 96% had a mechanical ventilation system
• 38% had a single ventilation system
• 27% had three or more ventilation systems
• 12% had no mechanical outdoor air supply
As indicated above, most of the schools had some form of mechanical
ventilation system, but the majority of the system were not operating properly. One
of the prominent features of the HVAC systems was the number of problems found.
These problems covered a range that included inoperative equipment (broken belts,
fans, controls), equipment that had never been wired, poorly maintained equipment
(disabled damper linkage, dampers painted shut), poorly designed equipment
(ventilators too small or not used because they are too noisy), and unwitting
10-29
-------
modification of the ventilation system (replacing the rolled steel sash window walls
with insulated wall). The extent to which these types of problems were present was
remarkable. Every school visited suffered from at least one of these problems.
Ventilation Air Delivery Rate
A continuous carbon dioxide monitor was used as an indicator of ventilation
rates. In an occupied classroom the C02 level is a function of the number of students
and the ventilation rate. These measurements are invaluable when deciding
whether a radon control approach that increases the ventilation rate is appropriate
or not. Carbon dioxide measurements were made in the classrooms in the mid
afternoon before students were let out. Levels over 1000 ppm indicate that the
current ASHRAE guideline [6] of 15 cfm (7 L/s) of outdoor air per student is not
being met. Figure 3 shows C02 levels measured in occupied classrooms of nine
schools. Most rooms in these schools were above the ASHRAE guideline. The
mean C02 level was 1780 ppm. Figure 4 shows radon screening measurements
plotted versus the C02 levels for a number of rooms. It can be seen from this data
that a room that has elevated C02 levels may not have elevated radon levels and
vice versa. This chart should be viewed with caution as the radon measurements
were made at a different time and in different mode than the carbon dioxide
measurements. This data forms the basis for the SEP Team feeling that broader
indoor air quality issues need to be considered when investigating radon in schools.
Building Shell Pressure Relationship
A measurement of the pressure relationship between the inside of the
building and outside is one of the key parameters which needs to be investigated.
This can be accomplished on a not too windy day by simply making measurements
through a crack in a closed doorway or window utilizing a sensitive electronic
micromanometer (pressure transducer). Readings of slight positive pressure
indoors (+.001 to +.010 inches water (0.25 Pa to 2.5 Pa)) will help to keep radon out
during operation of the HVAC system, while negative readings will increase radon
entry. By monitoring the pressure differentials while modifying HVAC operation,
the potential of using existing equipment for mitigation can be evaluated.
Building Shell Tightness Test
To determine how much make-up air would be needed to slightly pressurize
a building, a fan door pressure test, or equivalent, can be performed. The results of
this test will reveal the practicality of "building pressurizationM to mitigate an
observed radon problem. All the schools tested were within the normal leakage
range for buildings of this size.
10-30
-------
RECOMMENDATIONS
In each of the school buildings both a soil depressurization and an HVAC
approach to controlling indoor radon levels was addressed. An HVAC system
approach was considered the first choice in 23 of 26 schools. This high fraction
reflects the number of schools whose ventilation rates did not meet current
guidelines and whose radon levels were low enough that meeting ventilation
guidelines would likely control the radon through both dilution and building
pressurization. Soil depressurization was often listed as a second option in the
event that increasing the ventilation rate to meet guidelines did not lower the
radon sufficiently,, or if the school needed to respond to the radon levels more
quickly than modifying the HVAC system would allow.
Any radon control strategy that is used in school buildings must comply with
state and local codes or requirements for the design and installation of mechanical
systems. This work typically involves a professional engineer.
CONCLUSIONS
Conclusions from this work are:
• radon in school rooms must be seen in light of other health concerns - a
holistic approach should be taken.
• there appears to be a serious problem with under ventilated schoolrooms
in US schools.
• in order to use HVAC systems to control radon in schools a team
approach must be used that incorporates school maintenance personnel,
professional engineers, and a radon professional.
• when HVAC systems are used to control indoor radon, increased
emphasis must be placed on scheduled maintenance and periodic
inspection by knowledgeable staff.
• interpreting radon measurements made in classrooms is difficult because
of the wide variety of activities that occur in schools.
It is the opinion of the authors that the current under ventilated status of
many school rooms is the result of ignorance on the part of the general public and
school officials. If the public knew that our children were trying to learn in
situations where the air they breathe is contributing to drowsiness, absenteeism,
poor focus of attention, and headaches, we believe this would be remedied. Once
these groups understand the importance of good indoor air quality, then and only
then will the funding for school design, construction, operation, and maintenance
reach a level that will ensure adequate ventilation for children.
10- 31
-------
REFERENCES
1. Fisher, G., Brennan, T.B., Turner, W. The School Evaluation Program, iru The
1990 International Symposium on Radon and Radon Reduction Technology:
Volume V. Preprints. 1990. Atlanta. EPA.
2. Peake, R.T., Schmidt, A., MacWaters, J., Chmelynski, H., Radon Measurements in
130 Schools: Results and Implications, in: Proceedings of the 1990 International
Symposium on Radon and Radon Reduction Technology. Atlanta. EPA 600/9-
90/005.
3. Schmidt, A., MacWaters, J., Chmelynski, H. Seasonal Variation in Short-Term
and Long-Term Measurements in Schools, in: Proceedings of the 1991 International
Symposium on Radon and Radon Reduction Technology. Philadelphia. EPA.
4. Brennan, T., Fisher, G., Turner, W., Thompson, R. Extended Heating,
Ventilating and Air Conditioning Diagnostics in Schools in Maine, in Proceedings
of the 1991 International Symposium on Radon and Radon Reduction Technology.
1991. Philadelphia. EPA
5. Brennan, T., Turner, W., Fisher, G. Building HVAC/Foundation Diagnostics For
Radon Mitigation in Schools and Commercial Buildings: Part 1. in Indoor Air '90
The Fifth International Conference on Indoor Air Quality and Climate. 1990.
Toronto, Canada.
6. ASHRAE, 62-89 Ventilation For Acceptable Air Quality. 1989, American
Society of Heating Refrigeration and Air Conditioning Engineers.
10- 32
-------
Figure 1 - Mean Radon Levels from SEP Schools
Error Bars are 1 standard deviation
X
liiii
I
S1 S2 S3 LA1 LA2 LA3 AT1 AT2 AT3 IA1 IA2 IA3 ND1 ND2ILL1 ILL2 ILL3 NJ1 NJ2 NJ3 ME1 ME2ME3ME4ME5
School ID
-------
Figure 2 - Summary of School HVAC and Sub Slab Characteristics
Ventilation (cfm) (nuMirad) Sub Slab Vac. Window Floor Num. Num.
8ch. ID Healing Cooling Exhaust Malta up cfm/per. Control Mate rial" Rn pCI/L cfm Retrofit? Area Student Staff
81
Warm air
AC
1000*
1 70/OA
10
T stat
FS
600*200
6
82
Hydronlc
None
5000*
OA/off
11
Tstat
CS
11001200
27
83
Warm Air
AC
NT
VAV
NA
CPU man.
NT
NT
LAI
Hydronlc
None
10000
400/UV
14
Tstat
S&G
500*50
20
LA2
Hydronlc
None
NT*
None
NA
Tstat**
SP
900*100
38
LA3
Warm Air
AC
NT*
NT/UV
NA
Tstat*
NT
NT
QA1
HP WA
HP
3600*
OAF/off
5
Tstat/Clock
8P(1/4')
800
43
unk.
49390
680
40
GA2
HP WA
HP
NT/lnop.
OAF/off
NA
Tstat/Clock
SP/Clay
1220-8000
25
unk.
44251
unk.
unk
GA3
Warm Air
AC
NTflnop.
OA/off
NA
Tstat/Clock
SP
4000*1000
40
unk.
80818
729
61
IA1
Hydronlc
None
2000
Wind.
5
Man.
FS/Clay
700*200
2-18
blocked
28000
215
25
IA2
Hydronlc
Wall AC
400*
Wind.
3
Man.
FS
500*200
18
unk.
14000
110
10
IA3
Hydronlc
Wall AC
5000
Wind.
34
Man.
S&G
1300*200
17-40
retrofit
87564
115
30
ND1
Warm Air
None
2300
OA/off
8
Tstat/Man.
CS
800*500
20
blocked
47500
350
36
ND2
Steam
None
•800
UV/WInd
18
Tstat/Man.
8P
800*100
17-48
retrofit
60800
400
27
ND3
Many
None
NT
OA/HRV
NA
Tstat/Clock
gap
700*150
48
unk.
largest
unk.
unk
ILL!
Warm Air
None
8357
OA/UVoff
14
Tstat/Clock
gap
200*100
60
blocked
41570
430
20
ILL2
Steam
None
5100*
OA/UVoff
10
Tstat/Clock
FS
700*150
7
no
36000
507
19
ILL3
Hydronlc
None
NT*
OA/UVoff
NA
Tstat/Clock
Crawl
20
NA
limited
28000
328
13
NJ1
Hydronlc
AC
NT
OA/UVoff
NA
Tstat/Clock
S&G/SP
2000/380
48
no
58900
457
68
NJ2
Hydronlc
AC
NT
OA/UVoff
NA
Tstat/Clock
CS.SP
1500/300
20
no
27752
400
30
NJ3
Hydronlc
AC
NT
OA/UVoff
NA
Tstat/Clock
SP
150*20
NA
no
32500
480
50
ME1
Hydronlc
None
5800
UVoff
13
Tstat/Clock
S&G
4500*1500
2&42
no
34000
3 74
44
ME2
Hydronlc
None
Bill
OA/UVoff
NA
Tstat/Clock
S&G
3500*1000
48
no
108500
540
46
ME3
Electric
None
11300
OA/UVoff
9
Tstat/Clock
8&G.SP
3500*2000
10-45
no
167000
1200
60
ME4
Hydronlc
None
NT
UVoff
NA
Tstat/Clock
SP
2500*1000
32
no
10500
85
ME5
Steam
None
NT
OA/UVoff
NA
Tstat/Clock
S&G
2500*1000
25
no
120000
unk.
unk.
* many fans Inoparatlva Not* : The Kitchen Exhaust Is Included In the total for most schools making cfm/person high
'• an Installed enrgy management control had been disabled
* the controls for the Unit Ventilators were In each classroom and teachers
reported they didn't operate them because they were too noisy
*> F8 3 Fine 8and C8 » Coarse 8and S&G = Sand and Gravel
SP = Stone Pebbles NT = Not Tested
-------
Figure 3 - Histogram of Carbon Dioxide
Measurements Made in Occupied
Classrooms SEP-1990
— ASHRAE 62-89
TA-IA
LA-IA
Bfl-ND
CC-ILL
Ti-ILL
HA-ILL
GH-ME
GM-IA
RC-GA
L
400 800 1200 1600 2000 2400 2800 3200 3600 4000 4400
Carbon Dioxide (ppm)
Note : Measurements made in rooms with open
windows or less than five
people are not included in this data set.
10- 35
-------
Figure 4 • Room Carbon Dioxide vs. Radon Screening
Measurements in School Rooms
35-
30
25
2 20
u
a
c
o
¦o
ASHRAE Q2-8Q
fC'C02
15-
10 -
~~
~
~
TA-IA
LA-IA
BR-ND
CC-ILL
TI-ILL
HA-ILL
•
GH-ME
¦
GM-IA
~
RC-GA
r^r1"
s EPA Indoor Radcn
Action Level
1000 2000 3000
C02 ppm
4000
5000
10- 36
-------
EXTENDED HEATING. VENTILATING AND AIR CONDITIONING
DIAGNOSTICS IN SCHOOLS IN MAINE
Terry Brennan
Camroden Associates
RD#1 Box 222
Oriskany, New York 13424
Gene Fisher
Robert Thompson
USEPA
Office Of Radiation Programs
Washington, DC
William Turner
H.L. Turner Group
Harrison, Maine
ABSTRACT
An extensive effort to assess the effects of HVAC system operation on the indoor
radon levels was conducted. Many schools in the EPA School Evaluation Program
have been found to have disabled or malfunctioning outside air on the ventilation
system. The outside air in the Maine schools had been disabled. This condition was
corrected using professional HVAC and control contractors. Measurements were
made of radon levels, total and outside airflows, pressure differentials across the
building shell and sub-slab radon levels. Exhaust ventilation, built up air handlers and
unit ventilators were investigated. A heat recovery ventilator was added to a room that
had leaky window sash as the outside air supply for a passive roof vent system. The
passive vents have been blocked off.
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.
10-37
-------
INTRODUCTION
In August, 1990, extended radon diagnostics were performed in two Maine
Schools. The purpose was to assess the effects of returning the heating, ventilation
and air conditioning (HVAC) system to the original operating specifications would
have on indoor radon levels. This effort was part of the 1990 School Evaluation
Program[1]. Measurements of radon, air pressure differences across the building shell
and carbon dioxide levels[2] were made to help judge the system changes. While a
large amount of data was collected, these measurements were open to a number of
interpretations because the radon levels found in the schoolrooms during the
extended diagnostics week were much lower than were found by the screening
measurements made in April, 1990.
In December of 1990, followup measurements were made at the Gray High
School and Russell Elementary School in Gray, Maine. The purpose of these
measurements was to provide a basis upon which to judge the effect of the HVAC
improvements on radon levels, air pressure relationships and carbon dioxide
concentrations in occupied rooms. December was a good time to make this
assessment because it represented a worst case scenario. That is, the outside air
dampers in the unit ventilators and built up air handlers were closed to minimum and
the competing stack effect was at the maximum. Both conditions are the result of the
low outdoor temperatures found in Maine at that time of year. The measurements were
carried out by a team of people. The team included : Gene Fisher and Bob Thompson
USEPA Office of Radiation Programs, Washington, D.C. ; Bruce Harris, USEPA,
AEERL, Radon Branch, Research Triangle Park, NC; Bill Turner, Fred McKnight, H.L.
Turner Group, Harrison, Maine; Terry Brennan, Camroden Associates, Oriskany, New
York; and Gene Moreau, Bob Stillwell, Maine Department of Health Engineering,
Augusta, Maine.
A special note of thanks is extended to the Maine Department of Health for their
active participation in this evaluation.
PROCEDURE
The evaluation consisted of a visual inspection and measurement of key
performance related variables in the Gray High School and the Russell Elementary
School.
10-38
-------
An extensive set of measurements were made in the High School.
The following measurements were made :
continuous radon (pulse ionization and semi-conductor)
continuous air pressure differences (variable capacitance)
carbon dioxide survey (infrared spectrometer)
Continuous radon monitors were placed in rooms 2, 7, 17, 31, 32, 33, the
Guidance Office and the Conference Room. The monitors used were eight Honeywell
continuous radon monitors and two femto-Tech continuous radon monitors (room 33
and room 7). The Honeywell units provide mean radon levels for 4 hour intervals and
the femto-Techs for 1 hour intervals. Air pressure differences were monitored across
the floor slab in rooms, 33, 7, the Conference Room and the Guidance Office. Variable
capacitance chambers manufactured by Setra were connected to a data logger
provided by EPA to collect pressure difference data. Calibration curves were made for
each sensor using a micromanometer. Ventilation rates, outside air fractions and
ventilation effectiveness were estimated by making a survey of carbon dioxide levels
in the occupied classrooms. These could then be compared to carbon dioxide
measurements made in the same rooms at the end of the previous school year. Data
was collected from 12/18/90 until 1/16/91. This afforded the opportunity to see the
classrooms operated both normally and with school in recess for the Christmas
Vacation.
Additionally, measurements of sub slab radon were made in the High School and
the nearby Middle School. A carbon dioxide survey was also made in the Middle
School. The Middle School is very close to the High School but does not seem to have
nearly the elevated radon levels that the High School does. These measurements
were made to determine whether the Middle School radon levels were lower due to
lower source term, construction characteristics or HVAC operation and design. The
radon levels under both schools were in the range of 2000 to 4000 pCi/L. There is no
evidence that the source strength is the variable causing the large difference in the
radon levels in the two schools.
RESULTS
Overview Of Results
The results of this investigation can be briefly summarized in a few lines. The
evidence supporting these conclusions are then presented.
1) average radon levels that do not distinguish between occupied and unoccupied
10-39
-------
conditions can be misleading
2) the operation of the air handlers, both outside air and exhaust only, has a definite
reducing effect on the radon concentrations in the rooms
3) the decay rate of the radon after the air handler turns on is less than would be
expected given the amount of outside air that is introduced because the radon is still
entering due to negative building air pressure
4) repairing the outside air functions of the air handler made dramatic improvements in
the carbon dioxide levels in the rooms where outside air was introduced.
5) while effective and reliable at solving radon problems, soil depressurization in
rooms with inadequate ventilation leaves children sitting in high concentrations of C02
and other indoor air contaminants for which C02 levels are an indicator.
Effect Of Outside Air Improvements Qn Radon Levels And Dynamics
Introduction-
Continuous radon levels were monitored in eight rooms of the High School.
Rooms 33 and 7 are going to be used to illustrate the effects of the air handler
operation on radon levels in classrooms. The resolution of the femto-Tech units in
these rooms allows one hour radon levels to be used in the analysis. These rooms are
representative of the two different air handling systems - exhaust fans only and unit
ventilators with passive relief. Room 33 is in the new wing of the high school, contains
a unit ventilator and has repeatedly shown the highest average radon levels and
spikes. Room 7 is in the old wing, which has exhaust only ventilation and has shown
high radon levels. The only fan powered outside air that can potentially enter Room 7
is from the gym air handlers, when they are running. Otherwise., outside air to Room 7
consists of whatever is drawn in through leakage in the building shell, window wall
and corridor.
The next two major sections will examine first Room 33, the unit ventilator room
and then Room 7, the exhaust only room, in detail.
Room 33 - Unit Ventilator Ventilation-
The results of the continuous monitoring in Room 33 are shown in Figure 1.
Notice that the "rain spike" in this room on Christmas eve rises from 8 to 90 pCi/L and
10-40
-------
drops again to 16 pCi/L in a 24 hour period. This is far more severe than in other
monitored rooms, indicating that a substantial amount of radon is available to enter
this room. As in Room 7, the radon levels in this room drop quickly when the ventilation
turns on. This can be seen at the points labeled "Air Handler On" in Figure 1. Notice
that on Christmas eve during a rain storm there is large spike in the radon
concentration. This spike is seen in every room monitored and is interpreted as a rain
spike.
The dynamics of the drop in radon that occurs when the unit ventilator comes on
is illustrated by Figure 2. This graph shows the 24 hour period of December 19, 1991.
Between midnight and 6 AM the radon level hovers around 17 pCi/L. At 6 AM when
the unit ventilator is turned on by a timeclock control, the radon level drops in an
exponential decay until it reaches a minimum of around 2 pCi/L in the late afternoon.
An exponential decay of contaminant level is expected when dilution air is introduced
into the room. After the unit ventilator is turned off, the radon levels begin to climb until
they reach a level of 7 pCi/L again at midnight. The mean radon concentration for this
24 hour period is 8.9 pCi/L and for the occupied time it is 6 pCi/L. However, for the
lowest nine hour period the mean radon level is 3.8 pCi/L. This means that the dose
delivered to the occupants could be reduced 37% by starting the unit ventilator three
hours earlier.
NOTE:A correction for built up radon decay products in the continuous monitor is not
required for the pulse ionization device used because the decay products are
collected using an electric field without being counted. However, due to diffusion lag
into and outof the sensitive volume, a one hour time delay is observed in the radon
dynamic.
10- 41
-------
30
Spike to 90 pCi/L
Air Handler On
Air
Handler
On
25
20
1 5
Q.
Christmas Vacation
i Air Handler Off
288 Mon360
Dec 31
216
432
504
720
576 Sat 648
Sat
Dec 18,1990
Time (Hours)
1 - Radon Levels in Room 33
-------
25
.. Unit Ventilator on
20
7 AM
O
CL
c
o
"O
CC
1 0
24 hour mean = 8.9 pCi/L "
Mean occupied
time = G pCi/L
Lowest 9 hour
mean = 3.8 pCi/L
24 28 32 36 40 44 48
Time (Hours)
Figure 2 - Radon Dynamics in Unit Ventilator Room 33
10- 4.3
-------
While for this one day, the 19th of December the mean radon level for the
occupied time period was 6 pCi/L, it was not so for other occupied days. In fact, the
average occupied time radon level for the entire monitored period shown in Figure 1 is
a higher 7.8 pCi/L. This is still 28% lower than the 10.8 pCi/L mean for the entire time
period.
Another approach to understanding this dynamic is to apply tracer decay theory.
This has been done in the analysis shown in Figure 3. Figure 3 was created by taking
the decay curves for all the occupied days during the monitoring period and plotting
them on a single graph. The time scale has been changed from consecutive hours to
hours after the unit ventilator turns on. The result is a scattergram that plots all the
decay data for all the occupied days on top of each other.
If a given amount of contaminant is released into a room and then allowed to be
removed by dilution with ventilation air, it is expected that the concentration of the
contaminant will decay exponentially with time[3]. The rate at which it decays is
described by the solution to the continuity equation. This is given as the following :
1) C(t) = C(0) x eNt
where : C(t) = concentration at time t
C(0) = concentration at the start of the decay
N = airchange rate in air changes per hour
t = time in hours
By fitting an exponential decay curve to the data in Figure 3, the decay rate and
the air exchange rate for the average day during this monitoring period can be
determined. It is obvious from this curve that if the radon level at the start of the day is
greater than about 8 pCi/L, the mean level during the day would not get below 4 pCi/L.
The curve fit yields an air exchange rate of 0.13 air changes per hour (ACH). By direct
measurement of outside air, it is known that the air exchange rate in the room is 1
ACH. This discrepancy is explained in the following way. In order for equation 1) to
describe radon concentrations, the entry rate of radon after the start of the decay must
be zero. The introduction of outside air has not stopped radon from entering the room.
This is easily verified by a glance at the air pressure difference between the room air
and the sub slab air. The room air was at a lower pressure than the sub slab air during
the entire monitoring period. When the unit ventilator turned on, this difference became
smaller, but the room was still negative relative to the sub slab. The radon
10-44
-------
y = 15.416 * eA(-0.13157x) R= 0.83454
y = 15.416 * eA(-1x) R= 1
r. o
I- &
rt
V.'
* - H
v' $¦
)
p H -
,"! ~ -C-
0
w
I
W S 2 8 v
§ | r §'- H. n
H o
-2 0 2 4 6 8 10 12
Time After Air Handler Comes On (Hours)
' Measured Decay-all days
- *- Decay Curve if radon entry were stopped
Figure 3 - Reduction Rate of Radon in Room 33
Air Handler On - All Days Combined
10- 45
-------
entry rate may have been reduced but it certainly was not stopped. If the room was
pressurized by the unit ventilator then the radon concentration would have dropped
according to the lower curve in Figure 3. The radon concentration would be below 4
pCi/L in a matter of an hour.
In fact, it is likely that this is the case in this room during the spring and fall when
the outside temperature is warmer than in January. This is expected for two reasons.
One, warmer outside air means a reduction in the air pressure differences induced by
the stack effect. Two, when the outside air is warm enough gains from body heat will
overheat the room and cause the outside air dampers to open more. This will increase
the outside air volume and contribute to pressurizing the room.
Lastly, the room could potentially be pressurized even under the worst case
condition represented by these test results. This could be accomplished by air sealing
the room so that the minimum outside air flow rate would pressurize the room. Not only
would this control the indoor radon but it also would result in energy savings by
reducing air infiltration.
Room 7 - Exhaust Only Ventilation-
Figure 4 shows the continuous radon data in Room 7. The data begins on
December 18, 1990. Christmas vacation began on December 20, 1990 and ended
January 2, 1991. The radon levels in this room plummet whenever the rooftop exhaust
fans turn on (see the points labeled "Air Handlers On" in Figure 1). This effect is
repeatable. The radon levels drop in spite of the fact that operation of the exhaust fans
drives the air pressure difference between room 7 air and the sub slab air 3 pascals
lower. It is likely that the amount of radon entering the room increases when the fans
turn on. Although more soil air is being drawn in by the operation of the fans, the
dilution effect of the increased ventilation from above grade overwhelms the increased
radon entry. Unfortunately, the increased entry is not overwhelmed enough so that the
occupied radon levels are below 4 pCi/L, but are instead 7.1 pCi/L.
Figure 5 shows the agglomerated radon data for the occupied days in Room 7.
This graph was generated in the same way that Figure 3 was for Room 33. The
general trend of decreasing radon levels after the exhaust fan turns on is obvious.
There is a great deal more scatter in this data than there was in the data from Room 33
(the unit ventilator room). The curve fit to this data shows an effective ventilation rate of
only 0.065 ACH, while the measured exhaust rate informs us that there is actually 0.63
ACH (shown as the theoretical curve in Figure 5). The data from Figure 3 and Figure 5
are combined in a single graph in Figure 6. This figure highlights the similarities and
differences between the dynamics of the two rooms. Notice that the theoretical curves
for the two rooms almost coincide, even though the fan powered air exchange rates
10-46
-------
o
Q_
c
o
*o
(0
(T
25
20
1 5
1 0
0
-5
School in Session
Christmas Vacation
Atf-Handlers |-0n
Rain Spikes
' Sal ' '
0 72 144 216 288 Mon360 432 504 576 648 720
Dec 18, 1990 Dec 31
Time (hours)
Figure 4 - Radon Levels in Room 7
-------
Rm 7
¦ Theory rm 7
- I ¦ ' (J/ ' <
12
10
G
Q.
; s
O
"O
to
G
6
4
2
0
-2 0 2 4 6 8 10 12
Time
y - 9.6569 * eA(-0.065616x) R- 0.4982
- - y = 9.67 * eA(-0.63x) R- 1
Figure 5 -Reduction Rate of Radon in Room 7
Exhaust Fan On-All Days
( ?
{ > 1
P S
r, - n
> ¦ ¦
10-48
-------
25
20
o
s
* 10
5
0
-2 0 2 4 6 8 10 12
Time
y = 15.416 * eA( 0.13157x) R- 0.83454
- y - 15.416 * eA(-1x) R- 1
y - 9.6569 * eA(-0.065616x) R- 0.4982
y - 9.67 * e*(-0.63x) R= 1
Figure 6 - Reduction Rate of Radon
Air Handlers On - Rms 7 & 33
: )
¦
Rm 33
Theory rm 33
Rm 7
Theory rm 7
•r
+;
I
o
$
j.-
o
4
<¦>
b
'f 9
*
4:
H
.
:
r>i H-
tr?
V
-jh |
10- 49
-------
are quite different (1 and 0.63 ACH). This is largely due to the difference in source
terms. Room 7 begins the average occupied day at around 10 pCi/L while Room 33
begins the average occupied day at just over 15 pCi/L.
It is tempting to attribute the differences in radon dynamics in these two rooms to
the difference between exhaust only and fan powered outside air ventilation. But, two
rooms, no matter the depth of study provide anecdotal, not conclusive evidence. The
results of these measurements do support the current model of radon entry and control
as follows :
• entry is dominated by air pressure driven mechanisms
• exhaust ventilation can lower radon concentrations, but not as effectively
as powered outside air ventilation
To these two basics we can add a further hypothesis :
• unless fan powered outside air ventilation stops radon entry, the reduction
rate of radon will not be as great as expected from dilution alone
and a corollary :
• exhaust only ventilation will never lower radon concentrations as quickly
as would be expected from dilution alone because it does not stop the
entry of radon
It is important to understand that these two suggestions apply only to dynamic
radon behavior and not to steady state conditions. This only applies to the rate at
which radon levels change.
Effect Of Outside Air Improvements On Carbon Dioxide Measurements
Introduction--
The reason we breathe is to get oxygen to the cells in our bodies and to remove a
number of the byproducts of respiration. Carbon dioxide and water vapor are the most
plentiful products of respiration. Carbon dioxide levels in outgoing breath are several
thousand parts per million. Carbon dioxide measurements made in occupied rooms
can be used as a surrogate for levels of indoor air contaminants that are produced by
the occupants themselves and routine activities of occupants. If a simplifying
assumption is made about the generation rate of C02 being constant then they also
can be used to estimate the outside air ventilation rate [4], The ventilation guidelines of
10-50
-------
15 cfm/person in the publication ASHRAE 62-1989 Ventilation for Acceptable Indoor
Air Quality should result in a steady state 1000 ppm of carbon dioxide in an occupied
classroom.
Carbon Dioxide Measurements-
Carbon dioxide measurements were made in the High School and the Russell
School (pre and post radon control) and in the Middle School. The pre radon control
measurements were made in early June of 1990 and the post measurements were
made in December of 1990.
Carbon Dioxide Measurements in the High School-
A histogram is shown in Figure 7 that differentiates between the pre and post
carbon dioxide measurements. Only measurements from occupied rooms with closed
windows are shown. The distribution of C02 levels has been very clearly pushed to
the lower levels by the repairs made to the ventilation system. The pre radon control
C02 levels had a mean of 1402 ± 450 ppm and the post level mean was 1042 ± 394
ppm. This represents a 33% decrease in the mean. From a health, comfort and
alertness perspective, this is a great improvement over the situation before the
ventilation equipment was repaired. Although the mean is now nearly the level
recommended in the ASHRAE guidelines[4], half the rooms in the post control sample
would still be considered underventilated by the current guideline. Eight percent of
them (2 rooms) are above 1700 ppm, which would reflect an outside air exchange rate
of 5 cfmiperson. By contrast, all the rooms in the pre mitigation set of measurements
were above the current guidelines (1000 ppm) and 27% of them (3 rooms) were above
1700 ppm.
10- 51
-------
CO
e
o
o
or
©
o
F
3
hi ill III I II
III I JTjTTT'jTn
Post Ventilation
Pre ventilation
0
400 800 1200160020002400280032003600
Carbon Dioxide (ppm)
Figure 7 - Pre and Post Control C02 Histogram for High School
Carbon Dioxide Measurements in the Russell School-
A bar graph is shown in Figure 8 that differentiates between the pre and post
carbon dioxide measurements and between ventilation and radon control type.
Measurements are from occupied rooms with closed windows except the pre control
measurements in the exhaust only ventilation - soil depressurization rooms. These
rooms had open windows during the June measurements. The number of open
windows is shown on the bar graph.
The C02 levels have been very clearly lowered by the repairs made to the unit
ventilators (rooms 5, 9, and 6) and by the installation of the heat recovery ventilator
(located in room 1, with no powered ventilation). Pre control C02 levels were not
available for some rooms with unit ventilators (rooms 7, 8, 10 and 11) but post control
measurements were. The mean post control C02 levels for all the rooms in which unit
ventilators were repaired (5, 6, 7, 8, 9, 10, and 11) was 1350 ± 408 ppm.
Rooms 1, 2a, 2b, 3 and 4 are in the oldest wing, where there is no fan powered
ventilation. Rooms 2a and 2b show slight increases in C02 levels, averaging 1500
ppm C02, as compared to Room 1 which has dropped from over 1250 ppm to 925
10- 52
-------
2500
-4-
2000
E
CL
^Q.
0)
~g
x
o
Q
c
o
_Q
CO
O
1500
1 000
500
~1 f r-' "4-
i
-+-
Number of windows open during pre test* = i 3 wo
!
Unit Ventilators
*
(
i .
%
j HRV
One
window
No
Powered
Ventilation
I :
a r\
1 o
. !-;<
I ['
i
1_
' i
'/A
3 L
2
4 WO j 3 WO ; 4 wo
1 wo
2 wo
Exhaust Ventilation
with Soil Depressurization
! !
4 wc -
<¦/-
x''t
lz
I
I
i i1
LH
a
'A
/•'j
:JL
1
T
l
1
//*¦
K>:;
'/A
i
2
Pre C02 | ¦;
„ , n Post co21
21
20 22 24 26 25 23
Room Number
*Ncte No windows were open in the Unit Ventilator rooms during the C02 tests
Figure 8 - Carbon Dioxide Levels Pre and Post Radon Control at the Russell School
-------
ppm. This is expected considering that no changes in the ventilation of rooms 2a
and 2b have taken place, but a heat recovery ventilator has been added to Room 1.
Rooms 20, 21, 22, 23, 24, 25, and 26 are in the exhaust only wing, in which soil
depressurization has been used to control the radon. The radon levels in these rooms
(except for the library, which is around 7 pCi/L) are averaging between 1.4 and 3.5
pCi/L. The pre control C02 levels in these rooms must be interpreted cautiously
because at least one window was open in each room when these measurements were
made. The post control C02 levels had a mean of 1857 ± 376 ppm.
None of the exhaust only rooms meet the current ASHRAE guideline for
ventilation rates. In fact, none of them meets the ASHRAE ventilation guideline for the
year in which they were constructed. While it is clear that soil depressurization will
control indoor radon, it is also clear that it has little impact on other indoor air
contaminants.
Histograms of the C02 data from the Russell School are not presented because
there is so little pre control data that did not have windows open.
CONCLUSIONS
Conclusions for this work contribute to interpretation of radon measurements
made in school rooms (and other non-residential settings) where a wide range of
occupant activities and the operation of air handlers can have important effects on
radon measurements. Radon measurements in the Maine Schools show that average
radon levels that do not distinguish between occupied and unoccupied conditions can
be misleading when the effect of air handlers is unknown.
The operation of both types of air handlers, outside air and exhaust only, has a
definite reducing effect on the radon concentrations in the rooms. Unless radon is
prevented from entering, the radon concentration does not drop as quickly as
expected given the known amount of outside air that is being introduced. Only fan
powered outside air has the chance of doing this. In the High School it is not doing so
during the coldest months. It is likely that there are times during the spring and fall
when the outside air dampers are open wider and the stack effect is reduced that the
unit ventilator rooms are pressurized enough to prevent radon entry. Exhaust only
ventilation can have reducing effects, but will always be drawing some soil air into the
building. It is possible that for given source strengths and slab/building shell leakage
characteristics exhaust ventilation could be good enough to control radon, but that is
not so in the Gray High School.
Clearly many, if not all the classrooms investigated, were underventilated for the
number of occupants. The carbon dioxide data gives plenty of evidence for this
contention. Repairing the outside air functions of the air handler made dramatic
10- 54
-------
improvements in the carbon dioxide levels in the rooms where outside air was
introduced. However, while effective and reliable at solving radon problems, soil
depressurization in rooms with inadequate ventilation leaves children sitting in high
concentrations of C02 and other indoor air contaminants for which C02 levels are an
indicator.
1. Fisher, G., Brennan, T.B., Turner, W. The School Evaluation Program, in The
1990 International Symposium on Radon and Radon Reduction Technology : Volume
V. Preprints. 1990. Atlanta: EPA.
2. Brennan, T., Turner, W., Fisher, G. Building HVAC/Foundation Diagnostics For
Radon Mitigation in Schools and Commercial Buildings : Part 1. in Indoor Air '90 The
Fifth International Conference on Indoor Air Quality and Climate. 1990. Toronto,
Canada
3. Charlesworth, R, Air Exchange Rate and Airtightness Measurement Techniques
- An Applications Guide. 1988, Coventry, Great Britain: Air Infiltration and Ventilation
Centre.
4. ASHRAE, 62-89 Ventilation For Acceptable Air Quality. 1989, American Society
of Heating Refrigeration and Air Conditioning Engineers.
10- 55
-------
MTTTflftTTflN TU AfiNnBTTr.B; TUP. NEF.n FOR UNDERSTANDING
BOTH HVAC AND GEOLOGIC EFFECTS IN SCHOOLS
Stephen T. Hall
Radon Control Professionals,
Reston, Virginia 22094
Inc.
ABSTRACT
Experience In the remediation of schools has shown that In
some, highest indoor radon levels vere located near large central
HVAC return ducts and vere attributed to the predominance of and
the proximity to negative HVAC pressure. Successful sub-slab
depressurization systems vere Installed, however, in rooms with
lover indoor but greatest sub-slab radon levels, closest to the
source. This shows the inadequacy of using indoor radon levels
alone as a basis for remediation. Wings of other schools with
radon problems have window heating units in rooms of equal size
and no central HVAC system. Highest indoor radon levels
correlated well wit!) highest sub-slab radon levels due to the
equivalent effects of the window units and the predominance of
geology.
Diagnostic tests in other schools have revealed: blockvall
radon transport to upper floors; elevated blockvall radon
adjacent to sub-slab sources; and elevated indoor radon above a
cravlspace caused by HVAC-induced negative pressure.
The vork 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.
10-57
-------
In the past three years the author has conducted radon soil analyses at
approximately 20 school and numerous other construction sites in the Washington,
DC area (Northern Virginia and Montgomery County, MD) to predict indoor radon
potentials. Previous soil gas surveys shoved correlations with indoor radon in
existing buildings (1) and revealed that radon sources occur along narrow linear
trends within footprint confines of a single building, correlative with geologic
structures in metamorphic and sedimentary rock terrains (2). In addition, Radon
Control Professionals has performed radon remedial diagnostics and remediation in
20-30 schools and other large buildings.
Our experience has shown the importance of the effects of both the location
of geologic sources and HVAC-induced distribution of indoor radon. In general,
elevated radon in areas of schools with evenly distributed HVAC pressures are
correlated with maximum soil radon emanations. However, strong or unequal HVAC
effects can redistribute Indoor radon to areas aw.iy from the direct source.
Effective remediation required a complete understanding of both contributions.
In some schools with central hvac systems, highest Indoor radon levels were
located near large return ducts. However, highest sub-r-lab radon measurements
were often located in neighboring rooms with lower indoor radon levels indicating
that the negative pressure created by the return ducts had a more important
contribution to elevated indoor radon than source strength (Figures 1, 2, and 3;
In all figures, although some alpha track measurements were available, indoor
radon levels, shown in the center of each room, are two-day charcoal tests
performed during the same winter season for comparison. Both sub-slab radon
levels, adjacent to circles, and block vail, radon levels, adjacent to semi-
circles, are underlined.) Successful sub-slab depressurization systems were
installed in rooms with lower indoor but greatest sub-slab radon levels, closest
to the source. This shows the inadequacy of using indoor radon levels alone as a
basis for remediation.
10-58
-------
SPRINGBROOK HIGH SCHOOL
69
n
O
1168
3,5
2.8
FiCUKK 1. Kprlnqbrook High School indoor ridon levels not C'>r t <: i w! *: n ?.uh •
slat' radon levels due tc HVAC effertr. fr«*'!«>i»in.u:t over ycologlc sourr»
• • ~ • • - j - - -- -. - .t - - •> i - - • ~ ; •- i i... a . h ^ » e?.Ch
effects. In ail Fiyures, indent racon
room. Both s-.if.-3 lab ryJ^r. levels, aOjacn!
r.jijor: !evfl.-, -tdM'enl to semi -cirri*?, -ire
i ii
s a:
t. o r ' c c 1
n t e i
blot
10- 59
-------
799
12
WHITTER WOODS ELEMENTARY SCHOOL
170 O
HVAC Return
Duct
0554
FIGURE 2. Whitter Woods Elementary School - Indoor radon levels not
correlated with sub-slab radon levels due to HVAC effects
predominant over geologic source effects.
-------
The school shown In Figure 3 h?s a plenum celling with openings for return
air. The room with 3.2 pCi/1 has no windows or return openings in the plenum
ceiling. Differential pressure measurements between this room with the door
closed and the hallway showed no significant difference until a nearby outside
door was opened and hallway air rushed outside (Table 1). We suggested sub-slab
depressurization for this room because it had the potential for higher radon
levels if openings were added in the return plenum ceiling or doors were opened,
because both would depressurize the room.
TABLE 1. RIDGEVIEW JUNIOR HIGH SCHOOL - AP EFFECT FROM OFEN DOORS
INDOOR/HALLWAY, AP,
ROOM 119: TIME, SEC. INCHES HaO COLUMN
HVAC ON 30 -.001
60 -.001
90 -.001
120 -.001
ADJACENT OUTSIDE 150 +.017
DOOR OPENED-HALLWAV 180 +.020
AIR RUSHED OUTSIDE 210 +.020
Wings of two other schools with radon problems have equivalent window fan
co-il units in rooms of equal size and no central HVAC system. Highest Indoor
radon levels correlated well with highest sub-slab radon levels due to the
equivalent effects of the window units. (Figures 4 and 5). This was verified by
an outside corner room in Francis Scott Key High School (Figure 4) with 1.0 pCi/1
indoor radon and 132 pCi/1 sub-slab radon, the lowest source strength found.
Sub-slab/indoor radon ratios were approximately 100/1. The rooms with elevated
radon are aligned along a N60°W trend, correlative with local shear fractures
(2). In Cannon Road Elementary School (Figure 5), rooms with elevated radon
levels are aligned along a N30°E trend, correlative with local rock layers or
foliation (2). Thus in schools with equivalent HVAC effects, geologic source
appears to dictate indoor radon concentrations.
10- 61
-------
14
yyQoH70 |
H i2-5 :—
—
TJ
¦ i
-r "
TJ®Z56
I iggjj4
11.4
FIGURE 4. Francis Scott Key High School - Indoor radon levels
proportional to sub-slab radon levels due to equivalent
HVAC effects and predominant geologic control.
-------
3.4
312
• —o
u
16 12
FIGURE 5. Cannon Road Elementary School - Indoor radon levels
proportional to sub-slab radon levels due to equivalent
HVAC effects and predominant geologic control.
-------
Martin Luther King Junior High School (Figure 6} revealed indoor radon
migration through blockvalls from thr first floor to th* second floor. Rooms
near the center of the school and in the southeast corner had both first and
second floor radon levels equivalent to adjacent blockvall radon levels, shoving
that second floor radon problems were caused by vertical migration through
blockvalls. Sub slab depressorization vlth appropriately placed blockvall
penetrations remediated the school.
10-64
-------
Two schools (Figures 7 and 8) showed approximately equivalent hlock-
wall/sub-slab radon concentrations revealing radon migration into blockvalls
directly from the sub-slab source. This shows the need to assess blockwall radon
measurements to determine when blockwall penetrations are required based upon
high blockwall/sub-slab radon ratios.
10- 65
-------
Wall
0
i
01
05
penetration
38—
— P - 0 I I | ! 5 gl
Wall
6-8 I penetration
Q
h±i j ©
enetration
Drain
n
FIGURE 6. Martin Luther King Junior High School - Blockwall radon
transport to second floor. Second floor slab-on-grade is
outlined in bold with rooms in dashed lines. Where the
second floor is above a first floor, radon levels are
encircled.
-------
109
O 734
1210
534
FIGURE 7. Springbrook High School - Blockwall radon concentrations
correlating with adjacent sub-slab radon levels.
10-67
-------
-O,
276
3.2
ill o
1010
Blockwall
penetration
FIGURE 8. Ridgeview Junior High School - Blockwall radon
concentrations correlating with adjacent sub—slab radon
levels.
-------
in one school radon problems existed over one end of a room (F104) underlain
by the unvented end of a cravlspace (Figure 9). Table 2 shows the results of
indoor/outdoor ^ P measurements with a micromanometer. A Tygon tube was run from
the high pressure port of the micromanometer to outside a window, sealed shut
with tape, while the low pressure port was open to first the room and then the
crawlspace. An aquarium stone was attached to the high pressure tube outside to
minimize wind effects. The differential pressures were then measured In both the
room and the crawlspace by turning the central HVAC system on with the exhaust
fan off and then with the exhaust fan on. Results reported in Table 2 show that
the HVAC system created a negative pressure in the room resulting in radon levels
nearly as high as a two-day average within 60 seconds. The exhaust fan, blowing
from the room into the crawlspace, diminished this effect. In the cravlspace,
the HVAC system created an equal negative pressure with the exhaust off but
higher radon levels. However, the exhaust fan created a positive pressure in the
crawlspace greatly diminishing the radon levels. Theoretically pressurizing the
crawlspace with outside air would optimally reduce crawlspace radon levels.
However warm summer outside air entering the cool crawlspace causes condensation
problems so remediation was achieved by adding another crawlspace vent below the
problem room and running an exhaust line from a roof-mounted fan into the
crawlspace, as shown in Figure 9, to draw radon from the crawlspace at a high
enough rate to overcome the increase in radon levels from depressurizatlon.
WHITE OAK MIDDLE SCHOOL
existing
vent
exis ting
exhaust
fan
... T..
L_
•
2.1
1.3
1
1
.
.pressurization
fan
FIGURE 9. White Oak Middle School - Crawlspace area outlined with dashed line,
existing exhaust fan exhausts indoor air into crawlspace.
10- 69
-------
TABLE 2, WHITE OAK MIDDLE SCHOOL - A V AND RADON DEPENDENCY ON HVAC AND
EXHAUST FAN
TIME, SEC.
INDOOR/OUTDOOR, A P,
INCHES H2O COLUMN
Rn, pCi/1
ROOM F104:
HVAC ON,
EXHAUST FAN
OFF
15
30
60
-.005
-.008
-.010
4.5
HVAC ON,
EXHAUST ON
15
30
60
120
O
0
-.002
-.005
<0.1
CRAWLSPACE:
HVAC ON, 15 -.005
EXHAUST OFF 30 -.008
60 -.010 13.0
HVAC ON, 15 +.050 1.4
EXHAUST ON 30 +.100 <0.1
10- 70
-------
REFERENCES
1. Hall, S. Correlation of soil radon availability number with indoor radon
and geology in Virginia and Maryland. In: Proceedings of EPA/USGS Soil
Gas Meeting, September 14-16, 1988, Washington, D.C.
2. Hall, S. Combining mitigation and geology: indoor radon reduction by
accessing the source. In: Proceedings of the Annual Meeting of the
American Association of Radon Scientists and Technologists, October 15-
17, 1989, Ellicott City, MD.
10- 71
-------
A COMPARISON OF RADON MITIGATION OPTIONS FOR
CRAWL SPACE SCHOOL BUILDINGS
Bobby E. Pyle
Southern Research Institute
Birmingham, AL 35255
Kelly W. Leovic
U.S. Environmental Protection Agency
Air & Energy Engineering Research Laboratory
Research Triangle Park, NC 27711
ABSTRACT
School buildings that are constructed over crawl spaces can present
unique challenges to radon mitigation since they are often quite
large (at least 4,000 ft2 in area) and may contain support walls
with footings that extend below the soil surface. The perimeter
walls in the crawl space can also be extensive (on the order of 500
to 1,000 lineal ft). In this research project, natural ventilation
using the existing vents in the foundation walls, depressurization
and pressurization of the crawl space, and active soil
depressurization under a polyethylene liner covering the soil were
compared in a wing of a school building in Nashville, Tennessee.
The wing has four classrooms constructed over a crawl space area of
4,640 ft2. The building and crawl space were monitored throughout
each mitigation phase with continuous sampling devices that
recorded radon levels both in the crawl space and in the rooms
above, in addition to environmental conditions such as temperatures
and pressure differences in the building.
Results showed that active soil depressurization was the most
effective technique for reducing radon levels in both the crawl
space and the rooms above. Crawl space depressurization was also
very effective in reducing radon levels in the rooms above the
crawl space; however, radon levels in crawl space increased during
depressurization.
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!
10- 73
-------
INTRODUCTION AND BACKGROUND
This 29,266 ft2 (refer to Table 1 for metric conversion factors)
Nashville school building was originally constructed in 1954, with
subsequent additions in 1957 and 1964. The original building and
the first addition are slab-on-grade construction, and the 19 64
four-classroom addition is constructed over a crawl space connected
to the slab-on-grade section by a walkway. Initial charcoal
canister measurements in this school in 1989 indicated that the 18
slab-on-grade rooms measured presented the most severe radon
problems, averaging 34.1 pCi/L with a standard deviation of 7.5
pCi/L. In fact, levels over 100 pCi/L were subsequently measured
in some of the slab-on-grade rooms. Radon levels in the four
classrooms constructed over the crawl space were relatively much
lower, averaging 9.7 pCi/L with a standard deviation of 0.7 pCi/L.
As a result, initial remediation efforts during the summer of 1989
focussed on reducing levels in the slab-on-grade wings with active
subslab depressurization (1,2). Post-mitigation measurements
during February 1990 indicated that levels in the slab-on-grade
rooms averaged below 2 pCi/L, and at this time plans were initiated
to research the effectiveness of various mitigation techniques in
the crawl space wing.
The crawl space is approximately 4,640 ft2 in area, and the
height ranges from 46 to 80 in. with a total air volume of
approximately 25,500 ft3. The plan view of the crawl space is shown
in Figure 1. Access to the crawl space is excellent and the
surface of the soil is not complex (i.e., no inaccessible areas,
rock outcroppings, or large piles of soil). The floor of the
classrooms over the crawl space is a suspended concrete slab poured
over corrugated steel sheets supported by a network of steel
trusses. There are two internal concrete block support walls in
the crawl space that extend below the soil. These walls do not
penetrate the slab overhead; however, the walls effectively
subdivide the crawl space into three sections, as shown in Figure
1. This type of construction is quite different from that found in
residential houses. In many existing houses, the floor is composed
of wood decking (either 1 by 6 in. boards or plywood sheathing)
supported by wooden floor joists. This type of house construction
has been shown to be quite leaky and nearly impossible to seal all
the openings between the crawl space and the rooms overhead (3,4).
Since the crawl space does not contain any heating, ventilating,
and air-conditioning (HVAC) ductwork or any asbestos, it was of
interest to determine if the crawl space in this school building
could be sealed well enough to permit pressurization or
depressurization of the crawl space volume as a mitigation option.
The crawl space is ventilated naturally with eight block vents
(four each on the east and west sides of the building). Each of
these foundation wall vents has a screened opening with the same
gross area as a concrete block (8 by 16 in.) or approximately 128
in.2 Fan door leakage tests carried out on the crawl space
according to ASTM E 779-87 resulted in an effective leakage area
10-74
-------
(ELA) at 0.016 in. WC of pressure difference of 251 in.2 with the
vents open and 83 in.2 with the vents sealed (using closed-cell foam
board and caulking). Thus, the vents were providing approximately
168 in.2 of total open area, or about 21 in.2 per vent. This value
is consistent with that measured in houses using similar techniques
(5). The important point is that the leakage area independent of
the block vents is very low (83 in.2) compared to that measured in
15 houses in the same geographic area which ranged from 198 to 424
in.2 with a mean of 262 in.2 (5) . Thus, this building was thought
to be an ideal candidate to test a variety of possible mitigation
techniques.
METHODOLOGY
Mitigation systems typically installed in crawl space houses
include: isolation of the crawl space from the rooms above,
isolation and depressurization or pressurization of the crawl
space, isolation and ventilation of the crawl space (either natural
or forced), and active soil depressurization either directly in the
soil or under a plastic membrane (SMD) covering the exposed soil
(4). Each of these mitigation techniques (with the exception of
the forced ventilation and direct soil depressurization techniques)
was tested in this school crawl space in an effort to compare their
effectiveness when applied to a building having a larger size and
a different construction type (concrete slab over the crawl space).
Initial baseline testing was carried out before any
modifications were made to the building. Following the baseline
measurements, the accessible openings (e.g., utility penetrations)
from the crawl space to the upstairs rooms were sealed with a
combination of closed-cell foam and urethane caulking. The block
vents were also sealed with rigid closed-cell foam board and
caulking. Following testing with the vents closed, a network of 4
in. PVC ducting was installed as shown in Figure 1. The fan
installed is rated at 200 cfm at 1.5 in. WC. The fan and the air
distribution network were used to test the effectiveness of crawl
space pressurization and depressurization as mitigation options for
the building. After the crawl space depressurization and
pressurization tests were completed, two suction pits approximately
24 in. in diameter and 12 to 18 in. in depth were excavated in each
of the three sections of the crawl space for a total of six suction
pits as shown in Figure 1. Each suction pit was covered with a
piece of 36 in. square by 1 in. thick marine grade plywood. The
plywood covers were supported at the corners by four common bricks.
Both the suction pits and the exposed soil were covered with two-
ply high-density polyethylene sheeting. The plastic film was
installed in three pieces, one in each section of the crawl space.
No attempt was made to seal the plastic to the outer or inner
foundation walls. The edges of the plastic were cut approximately
12 in. wider than necessary in the event that sealing to the walls
was necessary. The excess material was then simply folded up the
walls or allowed to fold back upon itself. The network of PVC
ducting was connected to the suction pits to complete the active
soil depressurization systems, as in previous house research (3).
10-75
-------
Throughout the entire testing period, several parameters were
monitored continuously using a datalogging device. The parameters
monitored include: pressure differentials between Room 116 and
outside the building on the east and west sides; pressure
differentials between Room 116 and the crawl space interior;
pressure differentials between Room 116 and the sub-poly region
during the SMD testing; temperatures outdoors, in Room 116, in the
crawl space, and in the soil; wind speed and direction; the outdoor
relative humidity and rainfall; and the radon levels in both Room
116 and the crawl space. Each of these parameters was sampled
every 6 seconds and averaged or totaled at the end of every 30
minute interval. These measurements and their locations are
summarized in Table 2. The data were accumulated in the datalogging
device and periodically downloaded to a personal computer and
stored on magnetic disks for later analysis. Initial testing of
the building began on March 1, 1990, and continued through July 20,
1990, for a total of 152 days (3648 hours). The datalogger was
reinstalled from December 18, 1990, to January 17, 1991, in order
to evaluate the mitigation systems during winter conditions. The
most significant results are described in the following sections
for both the spring/summer and winter measurements.
RESULTS OF SPRING/SUMMER MEASUREMENTS
Baseline Measurements
The baseline radon measurements made with the block vents open
averaged 5.1 pCi/L in Room 116 and 10.8 pCi/L in the crawl space,
as shown in Figure 2. Figure 3 shows the averaged pressure
differences between the crawl space and outdoors and between Room
116 and outdoors during each phase of the mitigation. Also plotted
in Figure 3 are the average temperatures outdoors, in the crawl
space, and in Room 116 averaged over the testing period. Following
closing and sealing of the block vents and sealing the major
openings from the crawl space to the classrooms above, the average
radon levels in the classroom increased by about a factor of 3.3 to
17.1 pCi/L and the crawl space levels by a factor of 8 to 87.2
pCi/L. During this time the average pressure difference in the
classroom increased by a factor of 1.6 to -4.7 Pa, and the crawl
space pressure increased by a factor of almost 4 to -3.9 Pa. It is
obvious that closing up the crawl space greatly enhanced the
depressurization produced mainly by the stack effect. Also, the
temperature differences between the interior of the building and
the outdoors were much larger than during the other testing
periods, thus increasing the stack effect. These results clearly
indicate the effect on radon when the vents of crawl spaces are
closed for energy conservation purposes.
Crawl Space Pressurization
The next mitigation technique tested was crawl space
pressurization using the fan installed near the roof level of the
building and the network of PVC ducting to distribute the flow with
the crawl space vents closed. During pressurization, the average
fan flowrate was 234 cfm which was equivalent to about 0.6 air
changes per hour (ACH). During this time the average crawl space
10- 76
-------
pressure difference was reduced to -1.5 Pa and the average
classroom pressure difference was reduced to -2.5 Pa as seen in
Figure 3. The average radon levels in the classroom and crawl
space were 10.6 and 29.1 pCi/L, respectively, as shown in Figure 2.
It is apparent that the flowrate of outdoor air into the crawl
space is not sufficient to raise the pressure in the crawl space
above the outdoor pressure and could only negate about 60% of that
produced by the stack effect in the crawl space and about 50% of
that produced in the classroom. It is possible that by doubling
the flowrate (to around 500 cfm) the crawl space and the classroom
could have been pressurized above the outdoor conditions and the
radon levels further reduced. However, this option did not appear
as a desirable year-round solution in view of the fact that
unconditioned air was being used for pressurization.
Crawl Space Depressurization
Following the crawl space pressurization testing, the fan was
reversed so that air was withdrawn from the crawl space and
exhausted above the roof of the building. In this configuration,
the fan flowrate increased slightly to 279 cfm or about 0.7 ACH.
The negative pressures in the classroom were similar. However, the
pressure differential in the crawl space increased by approximately
73% (from -1.5 to -2.6 Pa). The radon levels in the classroom were
reduced by about 94% (from 10.6 to 0.6 pCi/L) even though the
levels in the crawl space increased by a factor 1.8 (from 29.1 to
53.6 pCi/L). Therefore, while depressurizing the crawl space
lowered the levels in the classroom, it nearly doubled the levels
in the crawl space. This was not unexpected since a similar
technique applied to a residential house increased the levels in
the crawl space by about a factor of 3 (4, 5).
Active Soil Depressurization
The third type of mitigation system implemented was active soil
depressurization under a plastic membrane covering the exposed soil
(SMD). The total flowrate exhausted from under the plastic liners
was 260 cfm when using all six suction points shown in Figure 1.
As seen in Figure 2, the radon levels in the classroom were reduced
within a matter of hours to around background (0.5 pCi/L), and in
the crawl space the levels decreased to 3.5 pCi/L. In an attempt
to determine if fewer suction points could be used, the two suction
points in the central sector of the crawl space were disconnected
and the suction pipes to both the fan and the suction pits were
capped. The results are shown in Figure 2. The decrease in the
crawl space levels is probably not significant, and the levels in
the classroom are the same within the level of uncertainty of the
measurement. The results from the SMD mitigation technique are
quite similar to those found when the same method is applied to
residential houses (4, 5, 6), where the area of the exposed soil is
typically in the range of 1,000 to 2,000 ft2. In this building the
area is much larger (4,640 ft2); however, the resulting reduction
in the radon levels using SMD is seen to be as good as that
achieved in smaller crawl spaces. The next important research step
10-77
-------
is to apply the SMD technique to crawl space areas on the order
10,000 ft2 or larger.
RESULTS OF WINTER MEASUREMENTS
The above measurements were repeated during the winter (December
18, 1990, to January 17, 1991) in order to determine if the results
were consistent with the spring and summer measurements. A brief
analysis of the winter data supports the results of previous
measurements and the integrity of the SMD system during cold
weather. These data will be fully analyzed and documented in a
future report. Based upon the initial analysis of the data, the
average cold weather radon levels both in Room 116 and in the crawl
space are shown in Figure 2.
Baseline Measurements
No attempt was made to reproduce the open vent (natural
ventilation) condition as this was felt to be an unusual operating
mode for wintertime conditions. The results for the closed vent
mode in winter were much the same as those obtained in the
spring/summer, with the possible exception that the winter radon
levels in the crawl space were not as high as the previous values
(63.4 pCi/L compared to 87.2 pCi/L). The lower readings could be
due in part to the fact that the winter measurements were carried
out after the soil was covered with the polyethylene liners. The
presence of the plastic liners covering the soil could act as a
partial barrier to soil gas exhalation. The lower readings could
also be due to the fact that the winter measurement period was much
shorter than the spring/summer measurement period.
Crawl Space Pressurization
The wintertime crawl space pressurization levels were much the
same as obtained previously. These results indicate that, with the
amount of unconditioned air used, the radon reductions achieved
with this mitigation technique are still less than desirable.
Crawl Space Depressurization
Using this technique during cold weather conditions gave very
similar results to those obtained in the spring/summer tests. The
wintertime levels in both the classroom and the crawl space were
somewhat higher and could be due to an increased stack effect
normally expected during cold weather. In order for this technique
to be successfully applied year-round, it is obvious that the
installation and testing must be done during extreme temperature
conditions in order to ensure that an adequate amount of air is
exhausted from the crawl space.
Active Soil Depressurization
The wintertime radon levels measured with the SMD system
operative were almost identical to the levels measured previously.
The average level in the classroom was within the uncertainty of
10- 78
-------
the measurement techniques, and the levels in the crawl space were
slightly lower than before. These results clearly indicate that
the SMB technique is not only effective but stable in its ability
to lower the radon levels in both the classroom and the crawl space
under varying weather conditions.
CONCLUSIONS
The results of this project indicate that the SMD technique is
the most effective in reducing elevated levels in both the crawl
space and the classrooms. In this application, the crawl space was
large but fairly simple in geometry. Access to the exposed soil
areas was excellent and, with the exception of the two internal
support walls, did not contain a large number of obstructions such
as support piers or utility pipes lying on the soil. The topology
of the soil surface in this crawl space was relatively smooth.
Other crawl spaces may have some or all of the complications that
were absent in this application (7). Application of the SMD
technique in these more difficult crawl spaces needs further
investigation.
Depressurization of the crawl space is effective in reducing
levels in the classrooms; however, the levels in the crawl space
will be increased by at least a factor of 2 and perhaps as much as
a factor of 3. This could pose a problem in buildings that have
nonsealable openings from the crawl space into the occupied rooms
above (e.g., HVAC ducts in the crawl space, wooden floors over the
crawl space, or doors or other entry openings from the crawl space
into the rooms above) or if the crawl space is occupied on a
regular basis. In this building the overhead floor was a poured
concrete slab with very few openings to the classrooms above that
helped to contribute to the effectiveness of crawl space
depressurization.
Pressurization of the crawl space was found to be less effective
in reducing the radon levels than natural ventilation. This method
may be more effective if larger quantities of air are supplied to
the crawl space; however, this may result in increased energy
losses and perhaps could increase the risk of damage to utility
lines in cold weather.
Natural ventilation of the crawl space also appears to be
ineffective in reducing the radon levels to acceptable levels.
Increasing the ventilation through larger or more numerous vents
may increase radon reduction; however, the effectiveness of this
method depends to a large extent on the wind patterns outdoors.
Also, this method can easily be defeated by closing vent openings
during the colder periods.
The number of school buildings constructed over crawl spaces is
not quantified at the present, although EPA research in over 4 0
schools has shown that only 7 of the buildings contain crawl spaces
(in combination with slab-on-grade substructures). There is little
information available regarding crawl space characteristics, such
as floor construction, number of vents, number of piers and support
walls, and the presence of HVAC ductwork or asbestos in the crawl
10-79
-------
space. While the SMD technique appears to be the method of choice
for reducing levels in both the crawl space and the rooms above,
further investigations need to be carried out in crawl spaces that
are not as simple as the one used in this study to determine if it
can indeed be applied successfully in non-ideal conditions.
REFERENCES
1. Craig, A.B., K.W. Leovic, D.B. Harris, and B.E. Pyle, Radon
Diagnostics and Mitigation in Two Public Schools in
Nashville, Tennessee. Presented at the 1990 International
Symposium on Radon and Radon Reduction Technology, Atlanta,
GA, February 19-23, 1990.
2. Leovic, K. W. , Summary of EPA's Radon Reduction Research In
Schools During 1989-90. EPA-600/8-90-072, U.S. Environ-
mental Protection Agency, Air and Energy Engineering
Research Laboratory, Research Triangle Park, NC, October
1990, (NTIS PB91-102038).
3. Pyle, B.E., A.D. Williamson, C.S. Fowler, F.E. Belzer III,
M.C. Osborne, and T. Brennan, Radon Mitigation Techniques in
Crawl Space, Basement, and Combination Houses in Nashville,
Tennessee. In: Proceedings: The 1988 Symposium on Radon
and Radon Reduction Technology. Volume 1. Symposium Oral
Paper VII-5. EPA/600/9-89/006a, U.S. Environmental Protec-
tion Agency, Research Triangle Park, NC, and Office of
Radiation Programs, Washington, DC, March 1989, (NTIS PB89-
167480).
4. Pyle, B.E. and A.D. Williamson, Radon Mitigation Studies:
Nashville Demonstration, EPA/600/8-90/061, U. S. Environ-
mental Protection Agency, Air and Energy Engineering
Research Laboratory, Research Triangle Park, NC, July 1990,
{NTIS PB90-257791).
5. Findlay, W.O., A. Robertson, and A. G. Scott, Testing of
Indoor Radon Reduction Techniques in Central Ohio Houses:
Phase 2 (Winter 1988-1989), EPA-600/8-90-050, U. S.
Environmental Protection Agency, Air and Energy Engi-
neering Research Laboratory, Research Triangle Park, NC,
May 1990, (NTIS No. PB90-222704).
6. Brennan, T., B. Pyle, A. Williamson, F. Belzer, and M.C.
Osborne, "Fan Door Testing on Crawl Space Buildings,"
Presented at ASTM meeting, Atlanta, GA, April 17-18, 1989.
7. Leovic, K.W., A.B. Craig, and D.W. Saum, Radon Mitigation
Experience In Difficult-To-Mitigate Schools. Presented at
the 1990 International Symposium on Radon and Radon
Reduction Technology, Atlanta, GA, February 19-23, 1990.
10-80
-------
TABLE 1. METRIC CONVERSION FACTORS
Non-Metric
cubic foot (ft3)
cubic feet per minute (cfm)
degrees Fahrenheit (°F)
foot (ft)
inch (in.)
inch of water column
(in. WC)
square foot (ft2)
square inch (in.2)
Times
28.3
0.47
5/9 (°F-32)
0.30
2.54
248
picocurie per liter (pCi/L) 37
0.093
6.452
Yields Metric
liters (L)
liter per second (L/s)
degrees centrigrade (°C)
meter (m)
centimeters (cm)
pascals (Pa)
becquerels per cubic
meter (Bq/m3)
square meter (m2)
square centimeters (cm2)
TABLE 2. SUMMARY OF MEASUREMENTS
Parameter
Differential Pressure
Radon
Temperature
Wind Speed and Direction
Relative Humidity
Rainfall
Location
Room 116 to Outdoors
Room 116 to Crawl space
Room 116 to Subpoly
Room 116
Crawl Space
Room 116
Crawl space
Soil
Outdoors
Outdoors
Outdoors
Outdoors
10- 81
-------
WEATHER STATION POLE
>VENTS
4"PVC
ROOM 116
-TO SLAB-ON-GRADE
pr^
24
ROOM 118
ROOM 119
FAN
VENTS
80'
Figure 1. Plan view of the crawl space and installed ducting
network*
10-82
-------
%
&>ca
V«nt« Op«n V«nt» Clo»«d C/S Prasi c/S Dapraao 6-Pt SMD 4-Pt smd
ClaSS (Spring/Summer) Crawl (Spring/Summer)
1 Class (Winter) Crawl (Winter)
Figure 2. Average radon levels in the crawl space and in
Rcon 116 during each of the mitigation testing
periods (both spring/sumner and winter).
10-83
-------
Vents Open Vents Closed C/S Press OS Depress
Rcom-to-Outdoors t//A Crawl space-to-Outdoor^ Tout
Tcrawl Troon
Figure 3. Average pressure differences between both the crawl
space and outdoors and Room 116 and the outdoors
during each of the mitigation testing periods.
10-84
-------
HVAC SYSTEM COMPLICATIONS AND CONTROLS FOR
RADON REDUCTION IN SCHOOL BUILDINGS
Kelly W. Leovic, D. Bruce Harris, and Timothy M. Dyess
U.S. Environmental Protection Agency
Air and Energy Engineering Research Laboratory
Research Triangle Park, NC 27711
Bobby E. Pyle
Southern Research Institute
Birmingham, AL 35255
Tom Borak
Western Radon Regional Training Center
Ft. Collins, CO 80523
Dave W. Saum
Infiltec
Falls Church, VA 22041
ABSTRACT
School mitigation research to date has emphasized reduction of radon levels using active
subslab depressurization (ASD). Although ASD has proven successful in a number of schools, it is not
reasonably applicable in all school buildings since many schools do not have a layer of clean, coarse
aggregate under the slab or may have many subslab barriers that would require an unreasonable
number of ASD suction points. Additionally, mitigation options that have relatively low installation and
operating costs need to be researched for application to schools with moderately elevated radon levels
(4 to 20 picocuries per liter, pCi/U. Since many schools are designed with heating, ventilating, and
air-conditioning iHVAC) systems that can provide outdoor air to the building, research has been
initiated to determine the feasibility of using HVAC systems to pressurize the building interior to reduce
elevated levels of radon in selected schools.
This paper discusses case studies of four schools where the U.S. Environmental Protection
Agency's (EPA) Air and Energy Engineering Research Laboratory (AEERL) has recently initiated long-
term research on the ability of HVAC systems to reduce elevated levels of radon. The schools are
located in the states of Colorado, Maryland, Virginia, and Washington. Depending on the school
building floor plan and HVAC system design, a specific wing or the entire building was selected for
research. Two of the schools have unit ventilators in the rooms being researched and two have central
air-handling systems. Initial results indicate that, when sufficient outdoor air is supplied by the HVAC
system, radon levels can be reduced. The amount of radon reduction depends on the specific HVAC
system design and operation.
This paper has been reviewed in accordance with the U.S. EPA's peer and administrative
review policies and approved for presentation and publication.
BACKGROUND
Previous research efforts on radon reduction in schools have presented theoretical aspects and
limited short-term data on radon mitigation using HVAC systems (1, 2, 3); however, long-term research
on the feasibility of radon mitigation using HVAC system pressurization is limited. As a result, in the
summer of 1990 AEERL's Radon Mitigation Branch initiated several projects in an effort to better
understand school HVAC systems and their ability to reduce radon levels in schools while also
improving overall indoor air quality.
10- 85
-------
To initiate this research ori radon mitigation using HVAC systems, four schools (or wings of
the schools) were selected. Two of the school wings contain wall-mounted unit ventilators in each
classroom (Maryland and Washington), and two of the schools have central air-handling systems
(Colorado and Virginia). The Maryland and Virginia schools had been part of previous research efforts
by AEERL (1, 4), and the Colorado and Washington schools were identified during field studies in the
summer of 1990. These four schools, in addition to three other schools (one in Maryland and two in
Ohio), will be studied in more detail over the next year. In a few of the schools future research will
also include a comparison of HVAC systems and ASD in reducing elevated levels of radon. Metric
conversion factors are presented in Table 1.
CASE STUDIES
The following case studies discuss four schools located in Colorado, Maryland, Virginia, and
Washington. In addition to background information on each school, each case study includes an HVAC
system description, the results of initial measurements, and future research plans for the school. The
summary characteristics of these schools are displayed in Table 2.
COLORADO SCHOOL
The original building was constructed in 1956 and includes seven classrooms and various other
support offices and storage rooms with a total area of approximately 15,750 ft2 of floor space as
shown in Figure 1. The original building includes a 1,300 ft2 boiler room located in a basement in the
southwest corner. The boiler room is approximately 11 ft below grade and contains the HVAC system.
The remainder of the building is slab-on-grade construction. In 1958 an additional six classrooms, a
kitchen, several restrooms, and support rooms totaling about 9,500 ft2 were added to the original
building. In 1976, a 2,100 ft2 media center was added to the end of the 1958 addition. The last
addition to the building was in 1982 when a 200 ft2 storage area was added to the southwest end of
the multipurpose (gym) room. The total footprint of the building is approximately 29,000 ft2 , with
approximately 27,700 ft2 in contact with the soil.
E-Perm measurements made in all classrooms from January 15 to 17, 1990, averaged 6.6
pCi/L with a minimum of 4.8 pCi/L and a maximum of 12.3 pCi/L. Most of the rooms were remeasured
during followup tests from February 14 to 16, 1990. These later measurements averaged 7.6 pCi/L
with a minimum value of 5.8 pCi/L and a maximum value of 10.2 pCi/L. The results of both sets of
measurements are shown on the floorplans in Figures 1 and 2, respectively.
HVAC System Description
The building HVAC system includes a central air handler with a single fan and individual
controls in each of the rooms. The HVAC system operates by time control with the system operating
approximately 9 hours during the daytime and set back for approximately 15 hours at night. This
schedule is apparently maintained even during the weekend when the school is not occupied. The
HVAC registers are located in the floor and the supply ducts are located below the slabs and are
composed of cylindrical cardboard ducts surrounded by poured concrete. In those areas where these
ducts were visible, large gaps were found between the cardboard tubing and the surrounding concrete.
It is highly likely that in most locations the cardboard tubing has deteriorated to the point that the
supply air is in direct contact with the concrete. Since radon levels may build up in the supply ducts
when the HVAC fan is not operating, these levels will be measured in future studies to determine the
relative contribution to building radon levels.
The return air from each classroom exits through grilles into the hallway with the hallway of
the building serving as a return air plenum. From the hallway the return air is ducted into a centra!
subslab return-air tunnel that leads back to the air handler in the basement. The air in the gym is
returned through floor grilles in the northeast and northwest corners of the room directly into the return
10-86
-------
air tunnel. The tunnel varies in size from about 3 by 3 ft up to 4 by 4 ft in cross section and can be
accessed in the boiler room. The tunnel has numerous penetrations by utility lines that lead to direct
soil contact and probably represent a major radon source. There is a provision for outdoor air to the
air handler located at roof level with the air ducted directly into the HVAC fan chamber through a
control damper. Visual observation of the outdoor air intake damper from inside the fan chamber with
the fan operating indicated that the damper did not open during fan operation. Subsequent
investigation by the school maintenance staff confirmed that the control rod for the fresh-air intake
damper did not operate properly, and this was repaired. However, it is not clear what control system
operates the damper. During the cold winter days the damper may be only partially opened depending
on the outdoor temperature.
Results of Initial Measurements
Room pressure differentials were investigated primarily in the kindergarten room using an
electronic micromanometer. These measurements were made before the outdoor air damper was
repaired. The differential pressure in the kindergarten room relative to the subslab was measured to
be -0.005 in. WC with the HVAC on and the door to the hall open. When the door was closed the
differential pressure dropped to -0.003 in. WC. The differential pressure between the kindergarten
room and the hallway was -0.005 in. WC with the HVAC on and the door closed. The pressure of the
room relative to outdoors was -0.005 in. WC. Differential pressure was not measured with the HVAC
system off. However, it appears that the HVAC system is depressurizing the classroom relative to
both the subslab region and outdoors. This indicates that, even in the warm summer months when
the HVAC system is used for ventilation purposes only, it causes room depressurization which results
in soil gas flow from the subslab regions into the room.
Radon concentrations under the slab and at several possible entry points were measured using
a Pylon AB5 in a "sniff" configuration. The subslab radon levels measured through 0.5 in. diameter
holes drilled through the slab in the kindergarten room and the office in Room 6 were about 700 pCi/L.
Levels of about 300 pCi/L were measured in a crack in the slab adjacent to one of the air supply
registers in the kindergarten room. Sniffing in one of the supply registers in the gym showed levels
of about 15 pCi/L with the air handler off and about 25 pCi/L with the fan on. Measurements in the
wall cracks of the air return tunnel showed levels of between 50 and 100 pCi/L with the fan off.
These levels increased to about 350 pCi/L when the fan was turned on and the tunnel depressurized.
This indicates that the depressurization of the return duct can increase radon entry from the soil
through the cracks and penetrations in the tunnel walls.
Examination of the air handler fan chamber identified a relatively large crack (about 0.1 in.
wide) in the slab. The investigators sealed the accessible part of the crack with duct tape for a length
of roughly 4 ft and sealed the hose of the Pylon under the tape. The levels were measured to be
about 700 pCi/L with the fan off and about 800 pCi/L with the fan on. The AB5 was placed in the fan
chamber to sniff the air in the chamber. The radon levels were about 70 pCi/L with the fan off and
increased to 350 to 700 pCi/L with the fan on, indicating that the slab crack into the fan chamber is
a major radon entry route. It was also observed that the crack was very clean with little or no dust
filling in the crack. Apparently there is sufficient air flow out of the crack (or turbulence in the air
above) to keep the crack clean. The pressure in the fan chamber relative to the boiler room was
measured to be approximately -2 in. WC.
Over the 1990 Christmas break a series of E-Perm measurements were made in all classrooms
of this school with the outdoor air damper for the HVAC system opened and closed. Measurements
were also made in another school in the district that has the same design but has not been shown to
have elevated levels of radon. In both schools the first set of measurements were made with the
outdoor air dampers closed (December 21-26, 1990), and the second set were made with the damper
open (December 27-31, 1990). The weather during the second measurement period was exceptionally
cold and, as a result, it appears that the damper in the school with the radon problem did not open as
10- 87
-------
intended. Because of this, the measurements with the damper open were repeated in this school on
January 1-2, 1991. For each of the two schools, Table 3 presents the average of the radon levels in
all classrooms, the levels in the boiler rooms, and the levels in the return air ducts. School 1 is the
school with the known radon problem, and School 2 is the other school.
As indicated by the results in Table 3, opening the outdoor air damper reduces average
classroom radon levels in School 1; however, it does not bring the average of the average classroom
levels to below 4 pCi/L. The results from School 2 show only a slight decrease in average radon
classroom levels with the outdoor air damper open. The radon measurements in the return air duct
exceed average levels in the classrooms in both schools. These results support the theory that the
return air duct is a major contributor to elevated radon levels, particularly in School 1. Opening the
damper helps to dilute radon levels in the tunnel but not enough to reduce average classroom levels
to below 4 pCi/L.
Future Plans
A datalogger was installed in School 1 in January 1991 to collect continuous radon levels (in
Room 6, the supply, and the return ducts), differential pressure, and meteorological data.
Measurements will also be made to compare the radon source strengths in Schools 1 and 2. Once a
series of baseline data are collected with the outdoor air damper opened and closed, the slab crack in
the fan chamber will be sealed and the measurements repeated.
MARYLAND SCHOOL
This school was mitigated with ASD in 1988 and is discussed in detail in Reference 5.
Previous measurements in a four classroom addition to the school (Building B in Reference 5) indicated
that the unit ventilators in the classrooms could reduce radon levels of over 20 pCi/L to below 2 pCi/L;
however, school personnel had decided to install an ASD system since radon levels increased at night
when the unit ventilators were off. Measurements indicate that radon levels are typically well below
1 pCi/L with the ASD system operating and, as a result, this school presents an ideal opportunity to
compare ASD and unit ventilator pressurization in the same school.
HVAC System Description
The area being studied is a four classroom addition, as shown in Figure 3. Each of the
classrooms has a wall-mounted unit ventilator that has the ability to provide outdoor air when the
damper is open. Although there is a large exhaust fan in the school (3600 cfm), according to school
officials it is never used.
Investigation of the unit ventilators revealed that, although the design drawings called for a
minimum of 16% outdoor air, the outdoor air dampers for two of the four units were not opening at
all. After repairs, flow hood measurements for the units in Rooms 107 and 108 indicated that about
120 cfm of outdoor air was being supplied by each unit with the outdoor air damper in minimum
(roughly 10% open) position. With the restroom exhaust estimated to be 50 cfm, Classroom 107 was
at a neutral pressure. With the outdoor air dampers open to 100% outdoor air, Room 107 was about
+ 0.003 in. WC relative to the outdoors, and the air flow into the unit ventilator was 450 cfm. All
doors and windows in the room were shut during the data collection.
Results of Initial Measurements
A datalogger was installed in Rooms 105 - 108 over the holiday break (December 21 to 31,
1990) in order to collect preliminary data on the unit ventilators operating with the ASD system off.
Measurements were made over successive 3-day periods with the unit ventilators operated as follows:
10-88
-------
1) setback (no outdoor airl, 2) normal operation (with evening setback), and 3} continuous day
operation with no setback (outdoor air provided for entire period). The fans for these units do not run
during setback unless room temperatures drop below 60°F. The radon levels measured in Room 108
during these three conditions are shown in Figure 4. As seen by these data, radon levels remain well
below 4 pCi/L while the unit ventilator is operated continuously but rise above 4 pCi/L during the
setback modes. Note that during the day-plus-setback operation, radon levels rise at night and drop
to about 4 pCi/L during the day.
Future Plans
A datalogger was re-installed in Rooms 107 and 108 to study unit ventilator operation over a
longer time period while the school is occupied. Continuous data being collected include: radon levels,
room to subslab differential pressure, unit ventilator damper position, and indoor/outdoor temperatures.
VIRGINIA SCHOOL
This school was constructed in 1987 in an area with a known radon problem. As a result,
various steps were taken by designers to reduce the likelihood of elevated levels of indoor radon and
to facilitate post-construction mitigation if needed. (The construction of this school is covered in detail
in Reference 4.) Initial post-construction charcoal canister measurements were made in October 1988
in all ground floor classrooms: all measurements were below 2 pCi/L, as shown in Figure 5. These
measurements were repeated in December 1990, and radon levels were consistently higher: 13 of the
rooms measured 4 pCi/L or higher as shown in Figure 6. Note that levels in the east wing of the
school tend to be highest. This is consistent with the higher subslab radon levels measured during
construction (4).
HVAC System Description
This school has eight air-handling units serving eight zones. The units are designed to provide
a total of 72,600 cfm with a minimum of 16,010 cfm outdoor air. Total building exhaust is 9,506
cfm. This design should maintain the building at a positive pressure; however, the HVAC system is
Variable Air Volume (VAV), and outdoor supply is reduced if the temperature drops below a given level.
Results of Initial Measurements
Differential pressure data showed the room to be at a negative pressure relative to the subslab,
thus the air-handling units were not adequately pressurizing the building as intended. A datalogger was
placed in a conference room December 21, 1990, to collect continuous radon, differential pressure,
and temperature data. These results, displayed in Figure 7, show that radon levels are about 5 pCi/L
when the room is at a negative pressure relative to the subslab. Radon levels tend to drop slightly as
the room-to-subslab differential pressure approaches zero.
Future Plans
The datalogger will remain in this school to collect additional continuous data. School
personnel are also considering installation of an ASD system to reduce radon levels on a continual
basis. If the ASD system is installed, its effectiveness in reducing radon levels will be compared with
that of HVAC pressurization.
WASHINGTON SCHOOL
This school has 16 classrooms, a multipurpose room (gym/cafeteria), and several special
purpose rooms and offices. Eight of the classrooms are built over a crawl space, and the remaining
eight are slab-on-grade.
10- 89
-------
Several radon measurements were made over all four seasons (spring, summer, fall, and winter)
under a number of ventilation conditions using 2-day charcoal canisters, short and long term E-perms,
and alpha track detectors. The results of these measurements are to be presented at the 1991
International Symposium on Radon and Radon Reduction Technology in a paper entitled "The Results
of EPA's School Protocol Development Study (6)."
Measurements indicated that the eight rooms built over the crawl space did not have elevated
radon levels. As a result, research focused on four of the eight slab-on-grade classrooms that had
consistently measured above 4 pCi/L. The layout of this part of the school is shown in Figure 8.
These classrooms were located in the northwest wing of the school (Rooms 139-142), and the design
drawings indicated the presence of aggregate under the slab. This school contained several classrooms
additions, and the foundation drawings available were not particularly clear on specific subslab
foundation locations. The subslab foundations included both poured concrete footings and thickened
slab footings.
There was a utility tunnel located under the slab along the perimeters of the classrooms in each
wing. This tunnel was approximately 4 ft wide by 4 ft high with a dirt floor. The walls of the tunnel
were of poured concrete and had numerous penetrations leading to the soil. Accesses to the tunnels
were in Rooms 140 and 141 in the west section and in Rooms 127 and 128 in the east section. The
tunnel contained the steam pipes that connected the boiler with the unit ventilators in each of the
rooms.
HVAC System Description
The HVAC system in this school consists of heating-only, three-speed unit ventilators located
in each room. Each room had an electronic thermostat that controlled the outdoor air damper and the
heating valve in the unit ventilator. Each unit had a low-limit thermostat that shuts off the outdoor air
damper when the supply air temperature falls below 60° F. The units appeared to be in excellent
working order in Rooms 139-142. Rooms 141 and 142 each have a wind-turbine exhaust ducted to
the roof through the storage/coat closets. The turbine for Room 142 was inoperable (not turning)
during the investigation, but school maintenance personnel planned to fix it promptly. A passive
exhaust was located in Room 140, and there was no exhaust in Room 139 (library).
There was no automatic shutoff of the ventilators, nor was there an automatic temperature
setback control. It appeared that each unit fan ran continuously and the unit cabinets and thermostats
were inaccessible without special tools (a hex key); thus the fan speeds and temperature settings could
not be adjusted by the teachers. The unit ventilator fans could be shut off at the electrical panelboard.
The piping was routed to each unit ventilator through tunnels under the slab, as seen in Figure
8. The return air for the unit ventilator was not isolated from the slab over the tunnel, thus any
opening in the slab (e.g., a pipe sleeve, crack) would allow air from the tunnel to enter the unit
ventilator and mix with the room air return and outdoor air. A high radon level in the tunnel could be
the source of elevated radon levels in the room. Some openings were found around pipe penetrations,
and radon levels in the tunnel averaged about 55 to 60 pCi/L.
Results of Initial Measurements
Air flow quantities were measured for each unit ventilator, and static pressure readings (relative
to the outdoor pressure) were taken in Rooms 139-142. The readings were taken for the various
operating modes of each unit ventilators: 1) unit ventilator off; 2) unit ventilator on low, medium, high
fan speed; and 3) unit ventilator with outdoor air damper opened and closed. In addition to these unit
ventilator modes of operation, room static pressure was measured with the hallway door opened and
closed. The results of the differential pressure and flow measurements are shown in Tables 4 through
7, and the results of the differential pressure measurements are displayed graphically in Figures 9
10-90
-------
through 12. These measurements indicate that the optimal operating mode for the reduction of soil
gas infiltration would require the unit ventilator to be on (any speed) with the outdoor air damper in
the open (or 100%) position, and with the hallway door closed. It appears that no other operating
mode, or door position, would allow for pressurization of the room. Only Room 139 (library) could be
pressurized with the outdoor air damper in the minimum (roughly 10% open) position, with the hallway
door closed (probably due to the lack of any exhaust system in the room). With the unit ventilator on,
the outdoor air damper open, and the hallway door closed, pressures in those rooms with wind turbine
or passive exhausts (140-142) ranged from +0.020 to +0.036 in. WC. These pressures should be
adequate to prevent soil gas infiltration into the rooms.
To determine the ability of the unit ventilators to reduce radon levels during normal occupancy,
a datalogger was installed in this school from November 29, 1990, to January 8, 1991. Continuous
radon levels were measured in Rooms 139, 140, and 141, and in the tunnel. Differential pressures,
temperatures, wind speeds and directions, classroom door openings and closings, and unit ventilator
operations were also monitored. These data are currently being analyzed. For a general comparison,
a summary of the data is displayed in Table 8. The results shown in this table were obtained over a
2 week period (December 2 through December 15,1990). During the first week (December 2 through
8, 1990) the classrooms were operated in a normal manner with the classroom doors into the hall
closed about 75% of the time (note that the doors were usually closed after class and throughout the
weekend). During the second week (December 9 through 15) the teachers were asked to keep their
classroom doors closed as much as possible during class. As seen in Table 4, the percent of time the
doors were closed increased to about 90%. The average radon level was reduced by approximately
50% as a result of the pressurization of the classrooms produced by the unit ventilators.
These data indicate that if the classroom-to-hall doors are kept closed, radon levels in the
classrooms can be reduced. The slightly lower levels in Room 139 (the library) are probably due to a
combination of factors including: a lower source strength, no exhaust (passive or turbine), and the
library door is closed more frequently than the classroom doors.
Future Plans
Data collected from the datalogger are being analyzed. Depending on the need to keep the
classroom-to-hall doors closed to achieve adequate mitigation with the unit ventilators, the school will
make a final decision on the mitigation approach.
CONCLUSIONS
The initial data collected in these four schools confirmed that pressurization of classrooms
(using the HVAC system) reduces average radon levels. Pressurization, however, did not consistently
reduce the levels to below 4 pCi/L in all the classrooms studied. The schools used in this study are
a small sample, but the HVAC systems found in these schools are expected to have a great deal in
common with those installed in most school buildings constructed in the U. S. since the 1950s.
Those buildings with central air handling units are designed to be pressurized. It was found
that modifications to the control systems by owners and deterioration of the system components have
resulted in these systems no longer operating to pressurize the classrooms. These systems were
contributing to depressurization of the building interiors, thus increasing the potential for the entry of
radon-laden soil gas. (In one case, it appears that radon entry into the subslab return air duct is also
contributing to elevated radon levels in the building.) A change in the control strategy, returning them
to original operations, should allow for pressurization of the classrooms and a reduction in radon levels.
However, it should be noted that most control strategies will close outdoor air dampers in cold weather
to reduce the likelihood of freezing the heating coil.
Unit ventilators are designed and operated in such a manner that the outdoor air damper is
10-91
-------
modulated based on indoor and supply air temperatures. They were observed in this study to
pressurize a classroom but usually only when the classroom door to the hallway was closed and the
outdoor air damper was open. This may not be sufficient to reduce radon levels consistently below
4 pCi/L without additional efforts to reduce other negative pressures in the building.
Research in these schools and additional schools over the next year will focus on determining
the optimal HVAC system operation for radon reduction. Limitations of HVAC pressurization will also
be studied, and in some of the schools HVAC pressurization will be compared with ASD.
REFERENCES
1. Leovic, K.W., Craig, A.B., and Saum, D.W. The influences of HVAC design and operation on
radon mitigation of existing school buildings. In: Proceedings of ASHRAE IAQ'89. The Human
Equation: Health and Comfort. San Diego, 1989, NTIS PB89-218-762.
2. Turner, W.A., Leovic, K.W., and Craig, A.B. The effects of HVAC system design and operation
on radon entry into school buildings. Presented at the 1990 International Symposium on Radon
and Radon Reduction Technology. Atlanta, 1990.
3. Brennan, T., Turner, W.A., and Fisher, G. Building HVAC/Foundation Diagnostics for Radon
Mitigation in Schools: Parti. In: Proceedings of Indoor Air'90, Toronto, 1990.
4. Witter (Leovic), K., Craig, A.B., and Saum, D.W. New-construction techniques and HVAC
over-pressurization for radon reduction in schools. In: Proceedings of ASHRAE IAQ'88,
Atlanta, 1988.
5. Saum, D. W., Craig, A. B., and Leovic, K. W., Radon Mitigation in Schools: Part 2. American
Society of Heating, Refrigerating, and Air-Conditioning Engineers (ASHRAE) Journal, Vol. 32,
No.2, pp. 20-25, 1990.
6. Schmidt, A. The results of EPA's school protocol development study. To be presented at the
1991 International Symposium on Radon and Radon Reduction Technology, Philadelphia, April
1991.
ACKNOWLEDGMENTS
The authors would like to express their appreciation to all the school officials who have
graciously permitted them to conduct measurements in their school buildings.
10- 92
-------
TABLE 1. METRIC CONVERSION FACTORS
Non-Metric
Times
Yields Metric
cubic foot per minute (cfm)
0.47
liter per second (L/s)
degree Fahrenheit (°F)
5/9 (°F-32)
degrees Centigrade (°C)
foot (ft)
0.305
meter (m)
inch (in.)
2.54
centimeters (cm)
inch of water column (in. WC)
248
pascals (Pa)
picocurie per liter (pCi/L)
37
becquerels per cubic
meter (Bq/m3)
square foot (sq ft)
0.093
square meter (m2)
square inch (sq in.)
0.00065
square meter (mJ)
TABLE
2. SUMMARY OF SCHOOLS'
State
Approximate
Size of Area
Under Study
SQ ft
HVAC
Initial
Radon Levels
oCi/L
Colorado
29,000
central
5-12
Maryland
3,500
unit
ventilators
14-20
Virginia
1,200
central
2-7
Washington
5,000
unit
ventilators
3-21
* Substructure of all schools is slab-on-grade.
TABLE 3. E-PERM MEASUREMENTS IN COLORADO SCHOOLS 1 AND 2
Radon Levels. DCi/L
Dates
Outdoor
Air Damper
Classrooms
1 2
Boiler Room Return Air Duct
12 12
Dec 21-26
closed
10.8 2.9
2.5 2.6 13.5
6.6
Dec 27-31
open in 2
open & closed
in 1
7.0 2.5
3.5 2.0 14.6
7.8
Jan 1-2
open in 1
4.6 -
2.5 - 7.5
-
10- 93
-------
TABLE 4. DIFFERENTIAL PRESSURE AND FLOW MEASUREMENTS IN ROOM 139
DATA TAKEN: August 22, 1990
DIFFERENTIAL PRESSURE MEASUREMENTS. ROOM TO OUTDOORS (in. WC.)
Room-to-Hall Door Closed
Unit Ventilator Speed Setting:
Outdoor Air Damper Position Off Low Medium High
Open (100%)
Closed i10% open)
Room-to-Hall Door Open
Open (100%)
Closed (10% open)
AIR QUANTITY MEASUREMENT (cfm)
Outdoor Air Damper Position
-0.001
-0.001
0.053
0.01
-0.001 -0.001
-0.001 -0.002
0.054
0.009
-0.001
-0.003
0.056
-0.012
0
-0.001
Low
Medium
High
Open (100%)
Outdoor Air
Supply Air
Closed (10% Open)
Outdoor Air
Supply Air
Percent Outdoor Air
Outdoor Air Damper Open
Outdoor Air Damper Closed
Outdoor Air Per Student (cfm
Outdoor Air Damper Open
Outdoor Air Damper Closed
Avg Leak Area (in.2) =
460 470
1175 1306
30 47
N/A N/A
39% 36%
3% 4%
Based on 20 Students)
23 24
2 2
73.2
500
1285
109
N/A
39%
8%
25
5
OBSERVATIONS:
Room 139 (Library) could be pressurized with the unit ventilator, regardless of
the outdoor air damper position, but only when the room-to-hall door was
closed. It does not have an exhaust vent like the other rooms, thus it is easier
to pressurize.
10-94
-------
TABLE 5. DIFFERENTIAL PRESSURE AND FLOW MEASUREMENTS IN ROOM 140
DATA TAKEN: August 22, 1990
DIFFERENTIAL PRESSURE MEASUREMENTS. ROOM TO OUTDOORS tin. WC)
Room-to-Hall Door Closed
Unit Ventilator Speed Setting:
Outdoor Air Damoer Position Off Low Medium High
Open (100%) 0
Closed (10% open) 0
Room-to-Hall Door Open
Open (100%) 0.001
Closed (10% open) 0.001
AIR QUANTITY MEASUREMENT (cfm)
Outdoor Air Damper Position Low
0.02
-0.002
0
-0.004
0.021
-0.001
-0.003
-0.002
0.024
-0.002
-0.001
-0.008
Medium
Hiah
Open (100%)
Outdoor Air
Supply Air
Closed (10% Open)
Outdoor Air
Supply Air
Percent Outdoor Air
Outdoor Air Damper Open
Outdoor Air Damper Closed
361
1200
45
1090
30%
4%
Outdoor Air Per Student (cfm - Based on 20 Students)
Outdoor Air Damper Open 18
Outdoor Air Damper Closed 2
438
1263
23
1135
35%
2%
22
1
449
1380
44
1197
33%
4%
22
2
Avg Leak Area (in.2)
OBSERVATIONS:
101.1
Room 140 could be pressurized with the unit ventilator, only with the outdoor
air damper in the fully open position, and the room-to-hall door closed. This
room has a passive vent and is more difficult to pressurize.
10-95
-------
TABLE 6. DIFFERENTIAL PRESSURE AND FLOW MEASUREMENTS IN ROOM 141
DATA TAKEN: August 22, 1990
DIFFERENTIAL PRESSURE MEASUREMENTS. ROOM TO OUTDOORS (in. WC)
Room-to-Hall Door Closed
Unit Ventilator Speed Setting:
Outdoor Air Damoer Position Qff Low Medium High
Open (100%)
Closed (10% open)
Room-to-Hall Door Open
Open (100%)
Closed (10% open)
AIR QUANTITY MEASUREMENT (cfm)
Outdoor Air Damper Position
0 0.03
-0.003 -0.002
-0.002 -0.005
-0.002 -0.002
0.034
-0.002
-0.001
-0.003
0.036
-0.003
-0.001
-0.003
Low
Medium
High
Open (100%)
Outdoor Air
Supply Air
Closed (10% Open)
Outdoor Air
Supply Air
Percent Outdoor Air
Outdoor Air Damper Open
Outdoor Air Damper Closed
495
1001
72
N/A
49%
7%
Outdoor Air Per Student (cfm - Based on 20 Students)
Outdoor Air Damper Open 25
Outdoor Air Damper Closed 4
580
1097
87
N/A
53%
8%
29
4
657
1160
94
N/A
57%
8%
33
5
Avg Leak Area (in.2) =
OBSERVATIONS:
113.0
Room 141 could be pressurized with the unit ventilator, only with the outdoor
air damper in the fully open position, and the room-to-hall door closed. This
room has a wind turbine exhaust and is more difficult to pressurize.
10- 96
-------
TABLE 7. DIFFERENTIAL PRESSURE AND FLOW MEASUREMENTS IN ROOM 142
DATA TAKEN: August 22, 1990
DIFFERENTIAL PRESSURE MEASUREMENTS. ROOM TO OUTDOORS (in. WC)
Room-to-Hall Door Closed
Unit Ventilator Speed Setting:
Outdoor Air Damper Position Off Low Medium High
Open (100%)
Closed (10% open)
Room-to-Hall Door Open
Open (100%)
Closed (10% open)
AIR QUANTITY MEASUREMENT (cfm)
Outdoor Air Damper Position
0 0.03
-0.003 -0.002
-0.002 -0.005
-0.002 -0.002
0.034
-0.001
-0.001
-0.003
0.036
-0.003
-0.001
-0.003
Low
Medium
High
Open (100%)
Outdoor Air
Supply Air
Closed (10% Open)
Outdoor Air
Supply Air
Percent Outdoor Air
Outdoor Air Damper Open
Outdoor Air Damper Closed
266
1123
150
1078
20%
14%
Outdoor Air Per Student (cfm - Based on 20 Students)
Outdoor Air Damper Open 11
Outdoor Air Damper Closed 8
230
1218
160
1250
19%
13%
12
8
251
1362
184
1306
18%
14%
13
9
Avg Leak Area (in. ) =
46.2
OBSERVATIONS:
Room 142 could be pressurized with the unit ventilator, only with the outdoor
air damper in the fully open position, and the room-to-hall door closed. This
room has a wind turbine exhaust vent and is more difficult to pressurize
although the turbine was inoperable during these measurements. The outdoor
air damper appears not to open fully.
JO- 97
-------
TABLE 8. AVERAGE RADON LEVELS IN WASHINGTON SCHOOL DURING 1 WEEK OF
NORMAL OPERATION AND 1 WEEK OF TESTING OPERATION
Normal Operation Test Operation Subslab Radon
Persent
Persent
Sniff
Average
Time Door
Average
Time Door
Measurement
Radon (max)
Closed
Radon (max)
Closed
(Aug. 1990)
Location*
(pCi/L)
<%)
(oCi/L)
(%)
loCi/L)
Room 139
2.6 (26.7)
76
1.4 (16.5)
97
400
Room 140
5.3 (29.2)
74
3.2 ( 7.4)
92
500
Room 141
4.5 (32.1)
75
2.2 (25.0)
88
700
Average
4.1
75
2.3
92
533
Tunnel
55.6 (202.8)
60.8 (129.2)
N/A
" Data for Room 142 not available; Pylon inadvertently "unplugged."
10-98
-------
KINDERGARTEN
W.RU
(6.9)
STORE
T
(6.8)
(TO)'
V
RU 1
(5.5)
RU 2
(6.0)
RU 3
(8.9)
FOOD
STORE
RMl T"
6H;7j)
KfT-
CHEN
(4.6)
I5°h
B1
CT
F1
RU 5
C'.o)
RU 4
(7.2)
MEDIA
CEHTER
(70)
("7.3)*
CYU S.
CYU
(8-2)
(6.1).
t :: j
V
N
B.T.
-DOWN
C.7.
. JjQ
RU 6
COACH OFF (5.6)
ART TEACH OFF (6.1)
•—DUPLICATE
SCALE
H w
F i cure 1.
^ R
AIDS ~
(6.4)
V
0
w
"N
RU 7
(6.0)
PRI | OFT
RU 8
LOUNCE ' S
(6.6) fT
(7.0)
RU 12
RM 9
(e.1)
(8.8)
RU It
" RU 10
(8.8)
»
. 1 .
(5.0)
Results of January 1990 radon measurements
in Colorado school, pCi/L.
CYU s.
DOWN
FOOD
RU 6
KINDERGARTEN
(7.2)
STORE
KIT
CHEN
RU 7
ST CT
RU 1
RM 2
RU 5
l :: j
(7.5)
(7.6)
J
COACH OFF (5.8)
RM 3
RM 4
(6.1)
(7.0)
SCALE
-
J- SO' -j
W.RM
(»•»>
STORE
MCDIA
Ct^fTER
(8.3)
V
N
PRI | OFF
(6.3) j (6.8)
LOUNCC | 5
(8.2) fT
RM 8
(7.9)
RM 12
RM 9
(7.6)
(SO)
RU 11
RM 10
(7.9)
(¦•5)
_
Figure 2. Results of February 1990 radon measurements
in Colorado school, pCi/L
10-99
-------
Room 108
20.3
Room 107
20.3
1
Room 106
14.4
(1 J
toom 105
13.8
-P
Figure 3. Results of initial radon measurements
in wing of Maryland school. pCi/L.
:c~t!nuC'js Setback
Continuous Day
360
356
358
362
354
1990 Julian Day
Figure 4. Continuous radon measurements
in Maryland school.
10-100
-------
0.6
I I
I I
L3
Jr
gure 5. Res jits of October 1988 radon measurements in Virginia school, pCi/L
2.0 3.0
2.8
3.2
ExEtD
3-. 6
Figure 6. Results of December 1990 radon measurements in Virginia school, pCi/L.
10-101
-------
8
School Starts 1/2/91
5
Radon in Conference Room
2
Subslab Pressure
Positive Pressure over Slab (warn weather)
1 •—
355
358
361
364
2
5
1990 and 1991 Julian Days
Rador, in Room
Subslab Pressure
Figure 7. Continuous radon and differential
pressure in Virginia school.
- UMT VOffUkTOft
142
141
126
127
128
129 !
140
wrurr
IUMCL
Figure 8. Partial floorplan showing utility tunnel
and room locations in Washington school.
10-102
-------
0.06
OFF LOW MEDIUM HIGH
t'V Speed Setting
¦ OAaOOV.DOOR-CLOSED [ID OA/l
-------
;_5 0.06 r
i rvi/'
OFF LOW MEDIUM HIGH
UV Speed Setting
H GAfl 00% )DOOR-CLOSED CHoA(10^)DOOR-CLOSED
52 O A(100% )DOOR - OPEN ~ OAflO-iODOOR-OPEN
Figure 11. Differential pressure measurements
i'n Rooti 14'., August 1990.
I F'»
OFF LOW MEDIUM HIGH
UV Speed Setting
H OA(100%)DOOR-CLOSED ~oA<10%)DOOR-CLOSED
22oA(100%)DOOR-OPEN O0A<109i)D00R-0PEN
Figure 12. Differential pressure measurements
in Room 142, August 1990.
10-104
-------
RADON DIAGNOSIS IN A LARGE COMMERCIAL OFFICE BUILDING
David Saum and Marc Messing
Infiltec
Falls Church, VA 22041
ABSTRACT
Large commercial office buildings present a significant
challenge to the commercial radon mitigator. A radon problem in
a Washington, DC area was recently analyzed with a number of
diagnostic techniques in a attempt to get a quick understanding
of the nature of the problem while operating within a limited
budget. The building has 7 stories, is 5 years old and it has a
VAV type HVAC system with 21 air handler zones. The diagnosis
was carried out using and integrated approach combining: 1)
multiple short term radon screening to look for hot spots, 2)
continuous radon monitoring in a few sites to identify day/night
radon variations, 3) pressure tests across doors to identify
localized depressurization, and 4) continuous pressure in hot
spots monitoring to identify building HVAC cycles. This
integrated approach identified different mitigation solutions in
each zone. Mitigation options have been presented to the
building owners, but a final decision on mitigation has not been
made at the time this paper was written.
If)-105
-------
BACKGROUND
Radon mitigators may need to use a wide variety of
diagnostic tools to analyze radon problems in large office
buildings. These buildings generally have sophisticated HVAC
systems and complex foundation structures that are not generally
found in homes or schools. For quick, cost effective radon
diagnosis in large office buildings, it may b necessary to use a
variety of radon and pressure measurement equipment. This paper
describes an attempt to diagnose a building using: 1) multiple
short term radon screening to look for hot spots, 2) continuous
radon monitoring in a few sites to identify day/night radon
variations, 3) pressure tests across doors to identify localized
depressurization, and 4) continuous pressure monitoring in hot
spots to identify building HVAC cycles.
The ground floor of this Washington, DC Metro area 7 story,
5 year old building is underground except for a loading dock
area. The HVAC system is a VAV type with 3 air handlers on each
floor, supplies in most rooms, and a return plenum overhead.
Figure 1 shows the floor plan of the basement and each of the
three HVAC zones is outlined. There are a number of areas in the
basement with slab-to-slab walls that may cross the boundaries of
the HVAC zones.
Previous radon tests were made with alpha-track monitors
deployed for three months during the summer and winter of 1989.
Rooms indicated on Figure 1 are locations of radon tests. Table
1 lists all of the radon test results. When some radon levels
above 4 p/Ci/L were found, all the building VAV units were
adjusted to supply a minimum airflow of 30%, and booster fans
were installed in the fresh air supply ducts. All of this work
was assumed to guarantee that the building would be under a
positive pressure while the HVAC system was on. No further radon
tests were performed after these modifications, and one of the
goals of the Infiltec work was to determine if the HVAC
modifications have made a change in the radon levels. Additional
goals include a determination of the pressure balances inside the
building and suggestions for mitigation if elevated radon levels
are found.
RADON MEASUREMENTS
In order to determine if the radon levels had changed since
the HVAC modifications were performed, radon tests were conducted
by Infiltec over the period 9/6 to 9/14 with electret passive
monitors in 23 rooms and continuous radon monitors (CRMs) in two
rooms. The electrets were read out every few days to check the
average radon levels and the CRMs recorded hourly data so that
the short term fluctuations could be monitored. Table 1 lists
the electret results and Figures 2 and 3 show the hourly radon
data in 2 zones.
10-106
-------
PERIOD 9/6-9/7
A quick 24 hour test was performed to get a snapshot of the
building and to check out areas such as elevator shafts and HVAC
rooms that had not been tested before. This data is shown in the
first data column of Table 1. No new sources were found but the
shop area which had shown the highest radon levels in previous
tests was not as high as the rooms in zones B and C.
PERIOD 9/6-9/10
A longer electret test (second data column in Table 1) over
the weekend was performed in more rooms with the hope of finding
sources in the building when the HVAC system shut down over the
weekend. Unfortunately, it was found that during the weekend the
HVAC system operates with the same cycling as a weekday because
of partial weekend occupancy. However, the longer tests showed
continued elevated levels of radon in most rooms in zones A and
B, and the shop showed the highest levels.
PERIOD 9/6-9/14
Adding 4 more days to the electret test (third data column
in Table 1) resulted in a surprising lowering of radon levels in
zones A and B, but the shop room stayed at about 6 pCi/L. When
the electret data is analyzed for the levels between 9/10-9/14
(fourth data column in Table 1) it can be seen that the radon
levels have dropped substantially in both of these zones during
this period, while the levels in zone A have not changed very
much.
Figure 2 shows what happened to the radon levels in one room
in zone B which is expected to be representative of most of the
rooms in this zone. On the evening of September 10 the radon
levels fall from about 4 pCi/L to about 2.5 pCi/L and remain
there. The electret data suggest that this is what happened in
all the rooms in zones B and C. One possible explanation is that
the onset of cooler weather on 9/10 may have changed the VAV
settings to bring in more fresh air. At present the reason for
this sudden change in radon levels is unknown but it seems to
have only affected the radon levels in zones B and C. Since
Figure 2 shows that the radon levels in zone B do not show a
day/night fluctuation, it seems that radon is being constantly
pulled into these zones during the day and that when the HVAC
system shuts down at night there is no significant increased or
decreased entry.
Figure 3 shows that the radon levels in the shop area
exhibit extreme day/night fluctuations with peaks up to 30 pCi/L
at night and decreasing to 1 or 2 pCi/L during the day. The
shaded area on this graph shows the radon levels during occupied
hours (7 am to 5 pm), and the average radon during occupied hours
is not very much different from the average levels during
occupied hours because the HVAC system comes on a t 7 am and it
takes several hours to sweep the radon from this room. Some of
this effect may be due to time lag in the CRM response. Note
that Table 1 shows that radon levels in the rest of zone A rooms
are quite low. There seems to be a strong radon source in the
10-107
-------
shop that is suppressed during the day by either positive
pressure or ventilation, but when the HVAC system shuts down this
source raises the levels in the shop very quickly.
PRESSURE MEASUREMENTS
Figure 4 shows a recording of the pressure difference
between the shop and the subslab gravel layer. This data was
measured through a small hole drilled through the slab in the
shop. The graph shows that there is a positive pressure in the
shop (relative to the subslab) during the day of 0.01 to 0.02
inches of water column ("wc) and when the HVAC system is shut
down at night there is still a positive pressure of about 0.002
"wc. The pressure in the shop relative to the hall was measured
at about 0.007 "wc lower than the hall during the day (Table 1)
and Figure 4 suggests that zone A is generally well pressurized
by the HVAC system. It is generally assumed that if there is any
positive pressure in a room relative to the subslab that all
radon entry will be suppressed. Therefore it is surprising that
the shop appears to be at a slight positive pressure even at
night when the radon is entering. This suggests that the radon
source is not in the subslab and that it may be somewhere in the
walls. We have been unable to locate the entry point and it may
be necessary to conduct further investigations when the HVAC
system is not pressurizing the room.
Subslab radon measurements were made through three drilled
holes in the shop floor and levels of 130 to 280 pCi/L were found
(Table 2}. These radon levels are very low. From our experience
we have generally seen subslab radon levels in problem buildings
ranging from 500 to 80,000 pCi/L. It appears that the subslab
radon may be diluted by the positive room pressurization induced
flow or that there is a hot spot somewhere that we have not
located.
Figure 5 shows the pressures measured through a hole drilled
through the slab in room H0228A in HVAC zone B. Again we see
good HVAC pressurization during the day (0.01 to 0.01 "wc) and
nighttime pressure around zero, with the exception of a half hour
negative period (about -0.006 "wc) just before the HVAC system
comes on in the morning. Note that several days of data were
recorded and each daily pressure cycle is almost identical to the
one shown. Table 1 pressure measurements made under the doors in
zones B and C show that the only rooms that are significantly
negative are the HVAC and electrical rooms. When these rooms
were investigated for possible radon sources, drains were found
that had large gaps around them leading directly to the subslab.
When radon measurements were taken in these drains, levels of
about 250 pCi/L were found (Table 2) together with significant
air flow into the HVAC rooms. It seems reasonable to believe
that the negative pressure in the HVAC rooms pulls in radon
during the day and distributes it around zones B and C, and that
when the HVAC system goes down at night this radon does not decay
enough to show any decrease in levels.
Pressure in the HVAC rooms (relative to the halls) in zones
B and C were measured on 9/10 at -0.050 and -0.026 "wc
10-108
-------
respectively. The significant decrease in zone C negative
pressure may be the reason that this zone had lowest radon levels
during the 9/10-9/14 electret monitoring. It is assumed that
this lower pressure was present during that previous time period.
The lower pressure would have reduced the flow of soil gas from
the drain hole in the zone C HVAC room. Zone B radon entry may
not have changed but there may be some communication between the
air in the two zones and the zone B radon reduction may be caused
by zone C.
CONCLUSIONS AND RECOMMENDATIONS
Based on the diagnostic measurements, the following
conclusions and recommendations were made:
1. The radon levels appear to be generally lower now than
they were during the 1989 summer and winter alpha-track
measurements. Of course, these radon measurements may not be
representative of the longer term, since they only covered one
week and we already have seen significant variations that appear
to be due to HVAC changes resulting from weather changes. Long
term (3 month) winter radon measurements are definitely recommend
for confirmation.
2. The building appears to be generally under positive
pressure (relative to the subslab) in most rooms while the HVAC
system is on. Only a few rooms were found to be significantly
negative relative to the hallway and subslab. No continuous
pressure measurements were made in HVAC zone C but all other
measurements suggest that it is just as positive as zones A and
B.
3. At night during HVAC shutdown there appears to be very
little negative pressure, but this may change as the weather gets
colder and the "stack effect" becomes stronger. In order to
investigate this possible effect it would be necessary to do
continuous radon and pressure measurements during cold weather.
If this stack effect causes significant radon entry during the
night, the HVAC system might be turned on earlier in the morning
(e.g. 6 am) to flush out the building. Another option is to run
the basement air handlers continuously during the night to
guarantee a continuous positive pressure over the slab.
3. The negative pressure in the pump room and the HVAC
equipment rooms should be eliminated if possible. Since a very
wide range of depressurization was measured in these rooms (from
0.8 to 0.008" wc) , it is assumed that there is a balancing
problem that could be corrected.
4. The radon source in the shop was not found and it might
be easier to locate when the HVAC system was shut down. It is
difficult to locate it during the day because the positive
pressure in the shop appears to suppress the radon entry.
5. The drain openings in the HVAC equipment rooms should be
sealed to prevent radon and soil gas entry. Sealing could
probably be done with a non-shrink grout or with a pourable
10-109
-------
polyurethane caulk. This may be the primary solution to the
radon problem in zones B and C, but it cannot be guaranteed
because radon tends to build up behind sealing and emerge at
other entry points. A combination of reducing depressurization
and sealing is likely to be most effective. It is not clear
whether the porous block walls in the HVAC rooms are also a
source and it may be necessary to seal them too.
6. The standard radon reduction technique of subslab
depressurization (SSD) may not be necessary in this building if
all rooms can be pressurized, the major soil gas leaks can be
closed, and any radon that enters when the HVAC system is shut
down can be countered by bringing up the HVAC system early enough
to flush it out. The shop area might be treated with SSD if the
source is located, and a small exterior exhaust fan could
probably be located in the bermed area next to the shop.
DISCLAIMER
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.
10-110
-------
Table 1 Radon and Pressure Test Results by Room
Test Type and Date
Room Electret Slectrct Electret Electret Aipha-Trk Alpha-Trk 9/10 Door
or zone S/6-9/7 9/6-9/10 9/6-9/14 9/10-9/14 Summer89 Winter 89 Pressure
Tested (pCl/L) (pCl/L) (pcCl/L) (pCl/L) (pCl/L) (pCl/L) C wc)
HVAC ZONE A
Shop (Pylon)
H000I pump
Custodial
H0138 Locksmith
Elect
Kitchen elevator
Kitchen storage
Freight elevator
cable chase
h'0001 storage
H,168 HVAC -0.800
H0168 electrical -C.120
3.3
6.0
5.9
6.4
CO
CN
rH
16.1
-O.C07
3.2
2.6
2.5
2.4
2.8
-0.098
1.1
1.5
1.3
1.1
-0.022
1.7
1.6
1.3
1.0
2.0
4.0
0.002
0.7
0.5
0.5
0.5
-0.120
0.2
0.2
0.3
0.4
-0.010
0.7
0.8
0.8
0.8
2.3
-0.003
1.2
0.6
0.5
0.4
-0.010
0.6
0.5
0.4
0.2
-0.009
na
1.3
1.3
1.3
0.000
HVAC ZONE 3
HC226 electrical 3.6 3.6 2.7 1.8 -0.008
H0226 HVAC 4.5 3.6 3.3 2.4 -0.050
K0256 na 4.4 3.4 2.6 0.000
H0266 (Pylon) na 4.1 3.4 2.9 5.0 4.5 0.000
H0244 r.a 4.2 2 .4 0.000
H0229B n 4.0 2.2 1.5 3.7 0.000
HVAC ZONE
H0407 electrical
H04C7 HVAC
H0470
H3440
H0450
HC4 95
HC308
H0310
H0318
H0324
4 .3
5.0
5.4
na
na
na
na
na
4.2
4 .6
4 .2
4.2
3.9
3.9
4.2
4.3
2.5
2.7
2.5
2.2
2.0
2.0
2.1
2.7
1.1
1.1
0.9
0.8
1.6
3.3
2.9
3.0
2.9
6.2
6.3
-0 .005
-0.026
O.QCO
na
0.000
0.000
0.000
na
0.000
0.000
Pylon indicates continuous monitoring available for that room
Negative pressure indicates that pressure across door is ower inside room
Table 2 Subslab Radon Test Results By Room
Room
Test
Radon
HVAC
Tested
Date
(pCi/L)
Zone
Shop hole A
9/14
200
A
Shop hole B
9/3 4
130
A
Shop Hole C
9/14
280
A
H3226 HVAC drain
9/18
24C
B
HC4 37 HVAC drain
9/18
27 0
C
Grab sample
test us
ing Pylon A3-5
with
10-111
-------
Custodial
HVAC Room
H0001
Pump Room
berms / i /loading dock
Shop
Kitchen - Elevator
Kitchen - Storage
3 HVAC zones per floor
7 floors
VAV type HVAC
Night shut down (10pm-6am)
Ceiling return plenum
Some walls slab-to-slab
Freight Elevator
H 0 4 7 0
HO 4 95
berms
lUv/v II
HJ.J \ |
HVAC ZONE A j
HO 14 6
below grade
H 0 2 4 4
Locksmith
H0138
r
H0100
front
entrance
H 0 4 0 7
HVAC & Elect
HO 4 4 0
below grade
102293
^4-fofeBSEB
H0266
H0450
H0256
H0226 HVAC & Electrica
H 0 310
Figure 1 Basement Plan of Building
-------
b
Q_
cd
cr
VAVs creatine more dilution due to milder weather?
wv
^ m ll«
v|
1 t ' 1 - <—• - -I ' ! 1
7 3 11 13 15
September 1990 Day
Hourly estimates from Pylon with Lucas cell
Figure 2 Continuous Radon in Zone B, Room H0226
30
X
b
Q-
"O
tvj
cr
Average rcdcr. = 5.9 pCi/L dtnng a1; hours
ccup'ed hours
Average radon = 5.1 pCi/^ during oc<
. Average radon - 6A pCi/L curing uiuccupied hours
SeptemDer 1990 Day
Pylon wHh P3D
to b pm)
Figure 3 Continuous Radon in Zone A, Shop Room
Hourly estimates from Pylon wHh P3D
Occupied hours (7 am to b pm)
10-113
-------
Pressure Across Slab
(measured through drilled hole)
o
i
O-r <- +0.10" WC
lO O
Room Pressurized
co ts;
<- +0.020" wc
fc: <- 0.0 0 0" wc
-0.020
+0.002" wc
pressure
Room Depressurized
, , r H
0.010" wc during night
setback
o- <- -0.10" wc
2-4 Tb 22- Z.I 2.0 l^f I?
9/12/90
9/13/90
<- TIME
Figure 4 Continuous Pressure in Zone A, Shop Room
-------
Pressure Across Slab
(measured through drilled hole)
~i p. <- +0.10" wc
o
i
U1
• • •- - Room Pressurized
I \2i~~ <- +0 . 020" wc
r.-r<- ZERO PRESSURE
<- 0.000" wc
: • \; j— <- -0.020" wc
0.0 0 6" wc
0.010" wc
¦ Negative pressure
•—jpike for
1/2 hour each
morning
Room Depressurized
V t . .
"f".
IT
13 IZ if 10 ? S- 1 6 5^3
9/15/90 <- TIME <-
<- -0.10" wc - -¦¦¦
% I SLflh 11 2.1 20
J2 17 14
9/14/90
Figure 5 Continuous Pressure in Zone B, Room H0226
-------
DESIGN OF RADON RESISTANT AND EASY-TO-MITIGATE
NEW SCHOOL BUILDINGS
Alfred B. Craig, Kelly W. Leovic,
and D. Bruce Harris
U.S. Environmental Protection Agency
Air and Energy Engineering Research Laboratory
Research Triangle Park, North Carolina 27711
ABSTRACT
The Air and Energy Engineering Research Laboratory's (AEERL)
radon mitigation research, development, and demonstration program
was expanded in 1988 to include the mitigation of schools.
Application of technology developed for house mitigation has been
successful in many but not all types of school buildings. School
mitigation studies carried out to date in the AEERL program have
been reviewed in order to determine those architectural features
which affect radon entry and ease of mitigation. This paper
details those features having the most effect and recommends the
design parameters which should be most cost-effective in
controlling radon in new school buildings.
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.
10-117
-------
INTRODUCTION
The Air and Energy Engineering Research Laboratory (AEERL) of
the U.S. Environmental Protection Agency (EPA) has been developing
and demonstrating radon mitigation technology in houses, both
existing and new, since 1985. In 1988, the program was expanded
to include radon mitigation in existing schools. In the
intervening 3 years, detailed diagnostic studies have been carried
out in about 40 schools in 8 states and mitigation studies in 20
of these schools. Walk-through examinations and reviews of
architectural drawings have been conducted in many additional
schools.
Over the past year, architectural features of the schools
studied have been carefully reviewed in an attempt to identify
those features which affect radon entry and ease of mitigation.
Results of the studies are currently being used to develop a guide
for construction of radon resistant and easy-to-mitigate schools.
This new guidance document will be available later this year. The
purpose of this paper is to briefly summarize some of the design
and construction features which have been identified as important
in this study.
Nearly all new schools being built today are slab-on-grade
(SOG), and this paper is limited to this architectural
substructure. However, what is stated for SOG schools normally
applies to schools with basements and is applicable to them. Few,
if any, new schools are being built today with crawl spaces, so
they are not covered in this paper.
DESIGN FEATURES WHICH AFFECT RADON ENTRY
Two design features are known to affect the rate of radon
entry into large buildings—slab cracks and penetrations and
pressure differentials resulting from the building shell
construction and the design and operation of the heating,
ventilating, and air conditioning (HVAC) system.
SLAB CRACKS AND PENETRATIONS
Slab cracks, expansion joints, and penetrations in schools are
similar to those in houses as is their control. These can be
eliminated by a change of building design, or their effects can be
minimized by proper sealing. Great care should be taken in slab
design to minimize slab cracking.
Sealing is even more difficult in existing schools than in
houses since the cracks are frequently hidden and cannot be readily
found. However, this is not true in new school construction where
all cracks and openings in the slab are readily accessible at some
stage of construction.
10-118
-------
DESIGN OF RADON RESISTANT AND EASY-TO-MITIGATE
NEW SCHOOL BUILDINGS
by: Alfred B. Craig, Kelly W. Leovic,
and D. Bruce Harris
U.S. Environmental Protection Agency
Air and Energy Engineering Research Laboratory
Research Triangle Park, North Carolina 27711
ABSTRACT
The Air and Energy Engineering Research Laboratory's (AEERL)
radon mitigation research, development, and demonstration program
was expanded in 1988 to include the mitigation of schools.
Application of technology developed for house mitigation has been
successful in many but not all types of school buildings. School
mitigation studies carried out to date in the AEERL program have
been reviewed in order to determine those architectural features
which affect radon entry and ease of mitigation. This paper
details those features having the most effect and recommends the
design parameters which should be most cost-effective in
controlling radon in new school buildings.
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.
10-J19
-------
INTRODUCTION
The Air and Energy Engineering Research Laboratory (AEERL) of
the U.S. Environmental Protection Agency (EPA) has been developing
and demonstrating radon mitigation technology in houses, both
existing and new, since 1985. In 1988, the program was expanded
to include radon mitigation in existing schools. In the
intervening 3 years, detailed diagnostic studies have been carried
out in about 40 schools in 8 states and mitigation studies in 20
of these schools. Walk-through examinations and reviews of
architectural drawings have been conducted in many additional
schools.
Over the past year, architectural features of the schools
studied have been carefully reviewed in an attempt to identify
those features which affect radon entry and ease of mitigation.
Results of the studies are currently being used to develop a guide
for construction of radon resistant and easy-to-mitigate schools.
This new guidance document will be available later this year. The
purpose of this paper is to briefly summarize some of the design
and construction features which have been identified as important
in this study.
Nearly all new schools being built today are slab-on-grade
(SOG), and this paper is limited to this architectural
substructure. However, what is stated for SOG schools normally
applies to schools with basements and is applicable to them. Few,
if any, new schools are being built today with crawl spaces, so
they are not covered in this paper.
DESIGN FEATURES WHICH AFFECT RADON ENTRY
Two design features are known to affect the rate of radon
entry into large buildings—slab cracks and penetrations and
pressure differentials resulting from the building shell
construction and the design and operation of the heating,
ventilating, and air conditioning (HVAC) system.
SLAB CRACKS AND PENETRATIONS
Slab cracks, expansion joints, and penetrations in schools are
similar to those in houses as is their control. These can be
eliminated by a change of building design, or their effects can be
minimized by proper sealing. Great care should be taken in slab
design to minimize slab cracking.
Sealing is even more difficult in existing schools than in
houses since the cracks are frequently hidden and cannot be readily
found. However, this is not true in new school construction where
all cracks and openings in the slab are readily accessible at some
stage of construction.
10-120
-------
Expansion joints are the largest source of cracks in SOG
construction. Where codes do not require them, they should be
eliminated since, in most cases, they serve no useful purpose. A
slab is at its largest size during curing in the first few hours
after pouring due to the heat of hydration of the cement. As a
result, the only slab which can be larger at a later date
(requiring an expansion joint) is one that is poured and cures in
extremely cold weather. Allowance for shrinkage, the other
function of an expansion joint, is better accomplished using pour
joints (without expansion joints) or control saw joints, both of
which are much easier to seal than are expansion joints. Where
pour joints are used without expansion joints, both slabs should
have a tooled edge to make possible a good polyurethane (PU) seal.
Control saw joints, pour joints, and expansion joints, where
used, should be carefully sealed with a flowable PU caulk applied
according to the manufacturer's specifications. With expansion
joints, the top 1/2-in.* should be removed to make space for a good
PU seal.
A second source of openings in the slab are utility line
penetrations. These can be minimized by running all utility lines,
except sanitary sewer, overhead in the area above the drop ceiling,
a practice found in some existing schools visited in our mitigation
studies. Overhead utility lines are recommended in radon-prone
areas in order to minimize slab penetrations by utility lines.
Utility penetrations, when present, must be carefully sealed. If
any type of wrapping has been put around a utility pipe to protect
it from the concrete, it frequently allows soil gas passage. This
type of wrap must be designed so as to not allow any soil gas
passage or it must be removed after the concrete is set and the
resulting space filled with a PU caulking.
In some design situations, utility pipes penetrate the slab
in groups to enter pipe chases. In these situations, great care
should be taken to design and construct in such a way that no slab
openings are left between the pipes.
HEATING, VENTILATING, AND AIR CONDITIONING SYSTEMS
Most schools being built today are air conditioned. This
usually results in the use of large HVAC systems supplying many
rooms. These large systems are always built with provisions for
ventilation by the addition of outdoor air to the air handling
system. This results in pressurization of the building as long as
the circulating fan of the air handler is in operation and an
adequate quantity of outdoor air is being brought into the system
continuously. Pressurization by this means will significantly
reduce radon-containing soil gas entry as long as the circulating
fan is operating and fresh air is being brought in. When the
(*)Readers more familiar with metric units may use the factors at
the end of this paper to convert to that system.
10-121
-------
circulating fan goes off, as is usually the case during night or
weekend temperature setback, radon-containing soil gas can enter
the building and in some cases has been found to reach high levels
in some classrooms. Once the circulating fan of the HVAC system
starts operating continuously in the morning when heating or
cooling is called for, soil gas entry is stopped and the radon in
the building is diluted over some period of time by the outdoor air
being brought in by the HVAC system. If the radon level reached
during the night is high, this dilution process can take several
hours. Studies underway, some of which are being reported at this
meeting, are aimed at determining under what conditions the HVAC
system can be depended upon to control radon to a satisfactory
level. Viability of HVAC system design and operation as a radon
mitigation approach cannot be determined until these studies are
completed.
Return air ducts have been found to be an entry route for
radon-containing soil gas. These should never be routed below the
floor since they are always under negative pressure when the HVAC
fan is running. Where the ceiling plenum is used as an unducted
return air space, any block walls penetrating the slab and ending
in the plenum should be capped with a solid block. Otherwise
radon-containing soil gas can reach the plenum through the block
wall which is very porous below the slab. Radon levels can also
build up in supply ducts under the slab when the circulating fan
is off and then be brought into the room when the circulating fan
comes back on.
Buildings can also be heated and air conditioned using unit
ventilators (UVs) supplied with hot water or steam from a boiler
and with chilled water furnished from a central chiller. All UVs
are designed for fresh air addition at the unit. Use of
pressurization to control radon in this type of system is similar
to that of a large central HVAC system.
Exhaust systems for large rooms such as kitchens, lunchrooms,
gymnasiums, multipurpose rooms, and shops create special problems
since they can create negative pressure and cause radon-containing
soil gas to be brought in. This can be eliminated by supplying
more conditioned outdoor air than is removed by the exhaust system.
Although this may appear to be an expensive solution, it is the
only known way to ensure no soil gas entry.
Restrooms also contain exhaust fans which frequently cause
elevated radon levels in these rooms. This can be minimized by
keeping the exhaust fan as small as code requirements will allow.
In addition, since the amount of time per day any person spends in
a restroom is presumed small, exposure in this area is relatively
small.
Schools without air conditioning are frequently ventilated by
the use of exhaust fans usually mounted in the plenum above the
10-122
-------
hall ceiling. The use of exhaust fans should be minimized in
radon-prone areas since this will usually result in a radon
problem. Rooms should always be ventilated by bringing in outdoor
air rather than by exhausting room air.
DESIGN FEATURES AFFECTING EASE OF MITIGATION
WITH ACTIVE SUBSLAB DEPRESSURIZATION (ASD)
The most successful mitigation technique for existing schools
has been the use of ASD, the same as in existing houses. This is
true as long as the school has aggregate under the slab. Since the
presence of aggregate can be required in new school construction
(and is, in fact, common practice), then it is logical that, until
similar information for other mitigation options becomes available
for performance and cost comparison, ASD should be the mitigation
system of choice in new schools. Thus the rest of this paper will
dwell on factors which affect the ease of application and the
effectiveness of ASD in new schools.
In a paper which the authors presented at the last symposium
in Atlanta(l), two schools mitigated in Nashville, TN, were
compared. One required 16 suction points to mitigate 15 rooms,
whereas 15 rooms were mitigated to a lower radon level in the
second school with only 1 suction point. This striking difference
in ASD effectiveness was the motivation for these authors'
beginning to review the factors which affect the ease of mitigation
in schools and has led to this paper.
In the authors' experience, pressure field extension (PFE),
is the most valuable diagnostic tool in determining the ease of
application of active subslab depressurization (ASD) to mitigation
of houses, schools, and large buildings. PFE measurements are even
more important in large buildings than in houses since much larger
subslab areas are involved and subslab barriers frequently exist
that are not normally found in houses. For example, PFE
measurements led to the prediction of the difference in ease of
application of ASD to the two previously discussed Nashville
schools which was then confirmed by the results obtained. Thus PFE
is used as a surrogate for ease of mitigation in the subsequent
discussion in this paper.
A review of the PFE measurements that have been made on all
of the schools in EPA's program, examination of their architectural
drawings, and many discussions of the factors affecting flow of
gases through aggregate beds with fellow scientists working on
radon have led to the identification of the following factors which
affect PFE:
Aggregate
Bulk density (or void volume)
Particle size (both average size and particle size
distribution)
Type (naturally occurring stone from moraine
deposits or crushed bed rock)
10-123
-------
Layer thickness and uniformity of thickness
Subslab barriers
Subslab suction pit size
Amount of suction applied
Size and location of openings in slabs (both
planned and unplanned)
These factors are discussed in the following sections.
AGGREGATE
The four properties of aggregate listed above are known to
affect the flow of a gas through stone beds- Bulk density is
actually controlled by particle size distribution and type of stone
(naturally occurring moraine gravel, which is rounded, packs more
efficiently than crushed bedrock with its greater variation in
shape).
The following tentative conclusions are postulated on the
effect of aggregate properties on PFE:
1- PFE is proportional to average particle size—the smaller
the particle size, the less the PFE assuming the same
particle size distribution.
2. The narrower the particle size distribution range the
greater the void volume and hence the greater the PFE.
3. The smoother the shape of the stone, the lower the void
volume; hence moraine stone (with its rounded corners)
will give lower PFE for the same average particle size
and particle size distribution than crushed aggregate.
AEERL is sponsoring work at Princeton University to verify and
quantify these effects. The first report of this work is being
made by Kenneth Gadsby in a poster paper given at this
symposium(2).
SUBSLAB BARRIERS
One of the greatest differences between mitigation of houses
and schools is the presence of subslab barriers which are
commonplace in schools and other large buildings and are rarely
found in houses. PFE measurements made in schools have shown a
very strong correlation with the presence or absence of these
barriers. Their presence is determined by a review of the
foundation plan in the structural drawings. Based on school plans
reviewed to date, foundation designs can be divided into the four
types shown schematically in Figures 1,3,5,and 6. These types
determine the ease of mitigation and the number of suction points
necessary assuming other factors are the same. They are presented
in the order of difficulty to mitigate by ASD starting with the
most difficult.
10-124
-------
l1
The type shown in Figure 1 (schematic) is the most common and
unfortunately the most difficult to mitigate. In this type, all
walls around each room extend below the slab to footings in
undisturbed soil resulting in the same number of compartments under
the slab as number of classrooms above the slab. A section of this
type of wall is shown in Figure 2. PFE measurements made in
Nashville showed that some PFE could be achieved through one
subslab wall but not two. Unfortunately, installation of a suction
point in every other room was not sufficient to mitigate the
intervening rooms, and it is now believed that a suction point will
normally be necessary in every room in this type of school.
Obviously, this is not a recommended footing configuration for new
schools built in radon-prone areas.
In the plan shown in Figure 3, the hall walls extend through
the slab to footings, but the walls between rooms are set on the
slab. The slab under these walls are normally thickened slab
footings as shown in Figure 4. Aggregate continues under these
thickened sections; consequently, they do not adversely affect PFE.
One suction point on each side of the hall will mitigate a number
of rooms in this configuration, the number depending on other
variables which affect PFE (such as type of aggregate). A third
suction point might be needed in the hall but it is unlikely if the
rooms on each side of the hall are adequately mitigated. In this
type of structure, the bar joists for the roof are placed
perpendicular to the hall and rest on the hall walls. The walls
between the rooms do not carry any roof load and consequently can
rest satisfactorily on thickened slab footings.
Figure 5 shows a footing configuration found in three schools
mitigated by EPA. In this configuration, the walls between the
rooms go through the slab to footings but the hall walls set on
thickened slab footings. In this case, the roof bar joists are
placed parallel to the hall and rest on the walls between the
rooms. The aggregate continues under the hall for the full length
of the building; consequently, PFE can be achieved down the hall
and into the individual rooms. With this configuration, the
suction point is best put in the hall, and the number of rooms that
can be mitigated will depend on other variables (such as type of
aggregate).
The final configuration found to date, shown in Figure 6, was
used in the Two Rivers Middle School in Nashville. In this
configuration, no walls go through to footings: all sit on
thickened slab footings. This is referred to architecturally as
post and beam construction and is commonly used in buildings which
are very wide and very long, such as supermarkets. Posts on both
sides of the hall at Two Rivers go through to footings and are tied
together with overhead beams which in turn carry the roof bar
joists. The posts and beams can be either reinforced concrete as
in Two Rivers, or more commonly steel as in supermarkets. In this
configuration, the aggregate is continuous under the entire
10-125
-------
building and, consequently, PFE can reach long distances if other
conditions are proper. At Two Rivers, PFE easily extended 130 ft,
and one suction point mitigated 15,000 ft2 to less than 1 picocurie
per liter (pCi/L). EPA recently arranged to have a hospital
building under construction install a suction point in the center
of a 200 by 250 ft slab (50,000 ft3) underlaid with carefully placed
coarse crushed aggregate (ASTM #5 stone). Some time this spring,
PFE of this slab will be measured, and EPA will have a better feel
for just how much PFE can be achieved under a very large slab with
optimum aggregate and a large suction pit.
SUBSLAB SUCTION PIT SIZE
The importance of the size and geometry of the suction system
under the slab has been the subject of considerable debate and
disagreement over the past 3 years. However, it has been the
authors' experience that, everything else being the same, the
larger the suction pit, the greater the PFE. Although this is not
too important in houses, it becomes much more important in large
slabs such as schools.
In an existing school, the size that can readily be dug
through a hole in the slab is about 40 in. in diameter. However,
in new construction, there is essentially no limit to the size of
the pit which can be installed. It is believed that the
controlling factor in increasing effectiveness is the size of
interface between the hole and the surrounding aggregate. With
this in mind, one of the authors (Craig) designed the suction pit
shown in Figure 7. The pit is constructed by digging out an area
of about 6 ft square where the suction pit is desired. Four
concrete blocks, 8x8x8 in. in size, are placed in a square 4 ft on
a side and covered with a 4x4 ft piece of 3/4-in. treated plywood.
The depth of the hole is such that the top of the plywood is even
with the bottom of the slab to be poured. The aggregate is filled
level with the plywood, allowing it to slope into the hole. The
angle of repose of the stone will be about 30° leaving most of the
hole open. The 6 in. suction pipe is installed under the plywood
as shown in Figure 7 and run to a convenient place for the riser.
This arrangement makes it possible to separate the location of the
suction pit from that of the riser.
The plywood serves only as a form for the slab over the hole.
The strength of the concrete after setting is more than sufficient
to span a 4 ft hole unless the slab has unusually high loading.
In that case, the slab will need reinforcing.
Perforated pipe can also be used in lieu of the suction pit
described above. However, calculations show that the suction pit
has an air to aggregate interface equivalent to about 200 ft of
perforated pipe with 10 holes of 3/4-in. diameter per lineal foot.
As a result, it is believed that the PFE from either system will
be about the same. Tests are planned to compare these two
techniques in new construction. It is believed that the suction
pit is significantly cheaper to install than the perforated pipe.
10-126
-------
AMOUNT OF SUCTION APPLIED
The amount of PFE also depends on the level of suction applied
to the suction pit. The amount of vacuum which can be applied
depends on fan size and the air leakage rate from all sources into
the subslab area. Theoretically, if the subslab aggregate envelope
is completely airtight, very little air will need to be moved to
get very large PFE. The top and sides of the envelope can be well
sealed, resulting in only a small amount of air leakage. However,
the bottom of the envelope, the compacted soil under the aggregate,
has variable permeability depending on composition and compaction.
Consequently, the air infiltration into the envelope from this
source is variable. Given a choice of subaggregate conditions, the
underlayment should be made as impermeable as reasonably possible.
For a given subaggregate, the more the soil is compacted, the less
the resultant permeability. In areas where the subaggregate fill
is highly permeable, such as with sand in Florida or with near-
surface moraine in many areas, it may be necessary to overlay the
permeable material with a compacted layer of impermeable clay.
The size of the suction fan needed can best be determined
experimentally. Table 1 lists the performance characteristics of
various sizes of one commercial exhaust fan (Kanalflakt). Note
that the larger fans can achieve a higher negative pressure than
the smaller ones. One wing of the Two Rivers Middle School (15,000
ft3) was mitigated by a T3B fan which had a flow of 150 cfm at 1.97
in. WC when installed in this system. In choosing a fan size, it
is better to err on the high side rather than the low side.
SIZE AND LOCATION OF OPENINGS IN SLABS
Expansion joints, pour joints, control saw cracks, and pipe
penetrations are discussed in an earlier section. Several other
types of slab penetrations can also affect radon entry. One such
source is an open sump connected to perforated pipe installed under
the slab for groundwater protection. All sumps must be sealed in
order to keep out soil gas which may contain radon. One good
solution for this is the use of a sewage ejector pit as a sump pit
since they always have vaportight lids.
Floor drains can also be a source of radon entry if connected
to a septic system (which is rare in the case of schools but they
do exist). In this case, care should be taken in the design to
make sure that the floor drain is trapped and will always be full
of water. Lines of conventional sewer systems have not been found
to contain radon since they are tightly sealed.
If electrical conduit is routed under the slab, care must be
taken to make sure that any conduit connections under the slab are
vaporproof. The same is true for any other subslab conduit.
10-127
-------
CONCLUSIONS
Study of the architectural features, diagnostic studies, and
mitigation results for the existing schools that have been
mitigated as part of the AEERL school mitigation program has
resulted in identifying many factors which affect radon entry and
ease of mitigation. Results of these studies have led to tentative
conclusions on how to design new schools which are radon resistant
and easy to mitigate. Many of these findings can be considered as
sufficiently sound that they can be recommended for incorporation
in new school buildings. Others need field verification in schools
either currently under construction or in the design phase. Work
is underway to accomplish this in the next 2 to 3 years.
1. Craig, A.B., K.W.Leovic, D.B. Harris, and B.E. Pyle, Radon
Diagnostics and Mitigation in Two Public Schools in Nashville,
Tennessee. Presented at the 1990 International Symposium on
Radon and Radon Reduction Technology, Atlanta, GA, February
19-23, 1990.
2. Gadsby, K.J., T.A. Reddy, D.F. Anderson, R. Gafgen, and A.B.
Craig, The Effect of Subslab Aggregate Size on Pressure Field
Extension. To be given at the 1991 International Symposium
on Radon and Radon Reduction Technology, Philadelphia, PA,
April 2-5, 1991.
Readers more familiar with the metric system may use the following
factors to convert to that system.
REFERENCES
CONVERSION FACTORS
Nonmetric
Multiplied bv
0.00047
Yields Metric
cfm
ft
ft2
HP
in.
0.30
0.093
7.46
0.025
mss
m
in. WC
249
m2
W
m
Pa
10-128
-------
Figure l. All walls are load-bearing.
Figure 2. Section of load-bearing wall
10-129
-------
Figure 3. Hall and outside walls are load-bearing.
Figure 4. Section of wall resting on thickened slab footing.
10-130
-------
Itooa vails
subml&i: footings
Figure 5. Walls between rooms and outside walls are load-bearing.
;*rt
Veil-
subsl&to footings
JJ
Figure 6. Outside walls and posts are load-bearing.
10-131
-------
S«Ct'on k
6" P'astit o'P»
Tr««tad plywood shMt,
V
To stack
to roof
Concr«t» block
1 * o . ^ ^ '
Sect^cn A
Figure 7. Design for large suction pit.
TABLE 1. KANALFLAKT FAN PERFORMANCE
,, _nCT ^ I™ Am FI,.QW fcfip) V? STATIC PFFr?SVRE fa. WQ PI?E
NiODEL HP RPM 0 1/8 1/4 3/8 1/2 3/4 1 1-1/2 2 DIA.
Tl Turbo 5
1/40
2800
158
143
125
114
90
45
T2 Turbo 6
1/20
2150
270
255
235
200
180
140
110
T3A Turbo 8
1/15
2150
410
375
340
28S
225
180
135
T3B Turbo 8
1/10
2300
520
500
470
445
415
310
230
200
T4 Turbo 10
1/6
2400
700
670
640
612
582
470
410
250
T5 Turbo 12
1/8
1250
900
801
718
624
557
456
359
254
5"
6"
8-
8"
115 lO-
ll*
10-132
------- |