United States Office of Research and Development
Environmental Protection Research Triangle Park NC 27711 September 1992
Agency Office of Air and Radiation
Washington DC 20001
vvEPA The 1992 International
Symposium on Radon
and Radon Reduction
Technology:
Volume 3. Preprints
Session VII: Radon Reduction
Methods
Session VIII: Radon Occurrence in
the Natural Environment
Session IX: Radon Surveys
Session X: Radon in Schools
and Large Buildings
September 22-25,1992
Sheraton Park Place Hotel
Minneapolis, Minnesota
-------
The 1992 International
Symposium on Radon and
Radon Reduction Technology
"Assessing the Risk"
September 22-25,1992
Sheraton Park Place Hotel
Minneapolis, Minnesota
Sponsored by:
U.S. Environmental Protection Agency
Air and Energy Engineering Research Laboratory
and
U. S. Environmental Protection Agency
Office of Radiation Programs
and
Conference of Radiation Control
Program Directors (CRCPD), Inc.
Printed on Recycled Paper
-------
The 1992 International Symposium on Radon
and Radon Reduction Technology
Table of Contents
Oral Papers
Session I: Radon-Related Health Studies
Preliminary Radon Dosimetry from the Missouri Case-Control of
Lung Cancer Among Non-Smoking Women
Michael Alavanja, R. Brownson, and J. Mehaffey,
National Cancer Institute 1-1
Rationale for a Targeted Case-Control Study of Radon and
Lung Cancer Among Nonsmokers
• Mark Upfal and R. Demers, Wayne State University and Michigan
Cancer Foundation; L. Smith, Michigan Cancer Foundation I-2
EPA's New Risk Numbers
Marion Cerasso, U. S. EPA, Office of Radiation Programs I-3
Interaction of Radon Progeny and the Environment and Implications as
to the Resulting Radiological Health Hazard
Lidia Morawska, Queensland University of Technology, Australia I-4
Does Radon Cause Cancers Other than Lung Cancer?
Sarah Darby, Radcliffe Infirmary, Oxford, England :. I-5
Measurements of Lead-210 Made In Vivo to Determine Cumulative
Exposure of People to Radon and Radon Daughters
Norman Cohen, G. Laurer, and J. Estrada, New York University I-6
The German Indoor Radon Study - An Intermediate Report After
Two Years of Field Work
Lothar Kreienbrock, M. Kreuzer, M. Gerken, G. Wolke,
H.-J. Goetze, G. Dingerkus, University of Wuppertal;
H.-E. Wichmann, University of Wuppertal and Center for
Environment and Health; J. Heinrich, Center for Environment
and Health; G. Keller, Saar University, Germany I-7
-------
Session II: Federal Programs and Policies Relating to Radon
EPA's Radon Program
Stephen D. Page, U. S. EPA, Office of Radiation Programs 11-1
Revising Federal Radon Guidance
Michael Walker, U. S. EPA, Office of Radiation Programs II-2
Profile of Region 5's Tribal Radon Program
Deborah M. Arenberg, U. S. EPA Region 5 II-3
The Development of the Homebuyer's Guide to Radon
Paul Locke, Environmental Law Institute, and S. Hoyt,
U. S. EPA, Office of Radiation Programs II-4
Mitigation Standards for EPA's Radon Contractor Proficiency Program
John Mackinney, D. Price, and G. L Salmon,
U. S. EPA, Office of Radiation Programs II-5
Consumer Protection and Radon Quality Assurance: A Picture
of the Future
John Hoornbeek, U. S. EPA, Office of Radiation Programs II-6
EPA's Proposed Regulations on Radon in Drinking Water
Janet Auerbach, U. S. EPA, Office of Drinking Water II-7
Session III: State and Local Programs and Policies
Relating to Radon
Radon in Schools: The Connecticut Experience
Alan J. Siniscalchi, Z. Dembek, B. Weiss, R. Pokrinchak, Jr.,
L. Gokey, and P. Schur, Connecticut Department of Health
Services; M. Gaudio, University of Connecticut; J. Kertanis,
American Lung Association of Connecticut 111-1
Trends in the Radon Service Industry in New York State
Mark R. Watson and C. Kneeland, New York
State Energy Office III-2
Targeting High-Risk Areas
Katherine McMillan, U. S. EPA, Office of Radiation Programs III-3
How Counties Can Impact the Radon Problem
Jerald McNeil, National Association of Counties, and D.
Willhoit, Orange County, NC, Board of Commissioners III-4
IV
-------
Innovative Local Radon Programs
Jill Steckel, National Civic League 111-5
Session IV: Creating Public Action
Indoor Radon: A Case Study in Risk Communication
Stephen D. Page, U. S. EPA, Office of Radiation Programs IV-1
Activating Health Professionals at the Local Level
Deborah McCleland, American Public Health Association IV-2
Translating Awareness Into Consumer Action
Mary Ellen Fise, Consumer Federation of America IV-3
Radon Testing and Mitigation as Applied in Corporate Relocations
Richard Mansfield, Employee Relocation Council IV-4
Ad Council Radon Campaign Evaluation
Mark Dickson and D. Wagner, U. S. EPA, Office of
Radiation Programs IV-5
Session V: Radon Measurement Methods
The U. S. Environmental Protection Agency Indoor Radon
Measurement Device Protocols - Technical Revisions
Melinda Ronca-Battista, Scientific and Commercial Systems Corp.;
A. Schmidt and T. Peake, U. S. EPA, Office of Radiation Programs V-1
A Performance Evaluation of Unfiltered Alpha Track Detectors
William Yeager, N. Rodman, and S. White, Research Triangle
Institute; M. Boyd, U. S. EPA, Office of Radiation Programs;
S. Popped, Jr., U. S. EPA-NAREL V-2
An Evaluation of the Performance of the EPA Diffusion Barrier Charcoal
Adsorber for Radon-222 Measurements in Indoor Air
David Gray, U. S. EPA-NAREL; J. Burkhart, University of
Colorado; A. Jacobson, University of Michigan V-3
A Lung Dose Monitor for Radon Progeny
Harvel A. Wright, G. Hurst, and S. Hunter, Consultec
Scientific, Inc.; P. Hopke, Clarkson University V-4
-------
The Stability and Response to Radon of New and Recharged Electrets
William G. Buckman and H. Steen III, Western Kentucky University;
S. Popped, Jr., U. S. EPA-NAREL; A. Clark, City of Montgomery, AL V-5
Design and Performance of a Low-Cost Dynamic Radon Test Chamber
for Routine Testing of Radon Detectors
P. Kotrappa and T. Brubaker, Rad Elec, Inc V-6
Session VI: Transport and Entry Dynamics of Radon
Characterization of 222-Radon Entry into a Basement Structure
Surrounded by Low Permeability Soil
Thomas Borak, D. Ward, and M. Gadd, Colorado State University VI-1
Analysis of Radon Diffusion Coefficients of Concrete Samples
K. J. Renken, T. Rosenberg, and J. Bernardin, University of
Wisconsin-Milwaukee VI-2
Data and Models for Radon Transport Through Concrete
Vern C. Rogers and K. Nielson, Rogers & Associates VI-3
Simplified Modeling for Infiltration and Radon Entry
Max Sherman and M. Modera, Lawrence Berkeley Laboratory VI-4
The Effect of Interior Door Position and Methods of Handling Return Air
on Differential Pressures in a Florida House
Arthur C. Kozik, P. Oppenheim, and D. Schneider, University
of Florida VI-5
Building Dynamics and HVAC System Effects on Radon Transport
in Florida Houses
David Hintenlang and K. AI-Ahmady, University of Florida VI-6
Radon Entry Studies in Test Cells
Charles Fowler, A. Williamson, and S. McDonough, Southern
Research Institute VI-7
Model-Based Pilot Scale Research Facility for Studying Production,
Transport, and Entry of Radon into Structures
Ronald B. Mosley and D. B. Harris, U. S. EPA-AEERL;
K. Ratanaphruks, ACUREX Corp VI-8
VI
-------
Session VII: Radon Reduction Methods
Durability of Sub-Slab Depressurization Radon Mitigation Systems
in Florida Houses
C. E. Roessler, R. Varas, and D. Hintenlang, University of Florida VIM
A Novel Basement Pressurization-Energy Conservation System
for Residential Radon Mitigation
K. J. Renken and S. Konopacki, University of
Wisconsin-Milwaukee VII-2
The Energy Penalty of Sub-Slab Depressurization Radon
Mitigation Systems
Lester S. Shen and C. Damm, University of Minnesota; D. Bohac
and T. Dunsworth, Center for Energy and the Urban Environment VII-3
Design of Indoor Radon Reduction Techniques for Crawl-Space
Houses: Assessment of the Existing Data Base
D. Bruce Henschel, U. S. EPA-AEERL VII-4
Multi-Pollutant Mitigation by Manipulation of Crawlspace
Pressure Differentials
Bradley H. Turk, Mountain West Technical Associates; G. Powell,
Gregory Powell & Associates; E. Fisher, J. Harrison, and B. Ligman,
U. S. EPA, Office of Radiation Programs; T. Brennan, Camroden
Associates; R. Shaughnessy, University of Tulsa VII-5
Two Experiments on Effects of Crawlspace Ventilating on Radon
Levels in Energy Efficient Homes
Theodor D. Sterling, Simon Fraser University; E. Mclntyre, Hughes
Baldwin Architects; E. Sterling, Theodor Sterling & Associates VII-6
Session VIII: Radon Occurrence in the Natural Environment
Indoor Radon and the Radon Potential of Soils
Daniel J. Steck and M. Bergmann, St. John's University VIII-1
Nature and Extent of a 226-Radium Anomaly in the Western
Swiss Jura Mountains
Heinz Surbeck, University Perolles, Switzerland VIII-2
Radon Potential of the Glaciated Upper Midwest: Geologic and
Climatic Controls on Spatial Variation
R. Randall Schumann, U. S. Geological Survey VIII-3
VII
-------
EPA's National Radon Potential Map
Sharon Wirth, U. S. EPA, Office of Radiation Programs VIII-4
Session IX: Radon Surveys
Comparing the National and State/EPA Residential Radon Surveys
Jeffrey L Phillips and F. Marcinowski, U. S. EPA, Office of
Radiation Programs IX-1
Radon Testing in North Dakota Day Care Facilities
Arlen L. Jacobson, North Dakota State Department of Health IX-2
Ventilation, Climatology and Radon Activity in Four Minnesota Schools
Tim Burkhardt, E. Tate, and L. Oatman, Minnesota
Department of Health IX-3
Estimates From the U. S. Environmental Protection Agency's
National School Radon Survey (NSRS)
Lisa A. Ratcliff, U. S. EPA, Office of Radiation Programs;
J. Bergsten, Research Triangle Institute IX-4
Session X: Radon in Schools and Large Buildings
EPA's Revised School Radon Measurement Guidance
Chris Bayham, U. S. EPA, Office of Radiation Programs X-1
Radon in Commercial Buildings
Harry Grafton and A. Oyelakin, Columbus, Ohio Health Department X-2
Iowa Mult!residential Building Radon Study
James W. Cain, Iowa State University Energy Extension X-3
Airflow in Large Buildings
Andrew Persily, U. S. Department of Commerce X-4
Meeting Ventilation Guidelines While Controlling Radon in Schools
Eugene Fisher and B. Ligman, U. S. EPA, Office of Radiation
Programs; T. Brennan, Camroden Associates; W. Turner,
H. L Turner Group; R. Shaughnessy, University of Tulsa X-5
Radon Reduction in a Belgian School: From Research to Application
P. Cohilis, P. Wouters, and P. Voordecker, Building
Research Institute, Belgium X-6
VIII
-------
Multiple Mitigation Approaches Applied to a School with
Low Permeability Soil
D. Bruce Harris, U. S. EPA-AEERL; E. Moreau and R. Stilwell,
Maine Department of Human Services X-7
General Indoor Air Investigations in Schools with Elevated Radon Levels
Terry Brennan, Camroden Associates; G. Fisher and B. Ligman,
U. S. EPA, Office of Radiation Programs; R. Shaughnessy,
University of Tulsa; W. Turner and F. McKnight,
H. L. Turner Group X-8
Comparison of ASD and HVAC System Control in School Buildings
Bobby Pyle, Southern Research Institute; K. Leovic, T. Dyess,
and D. B. Harris, U. S. EPA-AEERL X-9
Effectiveness of HVAC Systems for Radon Control in Schools
Kelly W. Leovic, B. Harris, T. Dyess, and A. B. Craig,
U. S. EPA-AEERL; Bobby Pyle, Southern Research Institute X-10
Radon Prevention in Construction of Schools and Other Large
Buildings - Status of EPA's Program
A. B. Craig, K. Leovic, and D. B. Harris, U. S. EPA-AEERL X-11
Session XI: Radon Prevention in New Construction
The Effect of Radon-Resistant Construction Techniques
in a Crawlspace House
David L. Wilson and C. Dudney, Oak Ridge National Laboratory;
T. Dyess, U. S. EPA-AEERL XI-1
Performance of Slabs as Barriers to Radon in 13 New Florida Homes
James L Tyson and C. Withers, Florida Solar Energy Center XI-2
HVAC Control of Radon in a Newly-Constructed Residence with
Exhaust-Only Ventilation
Michael Clarkin and T. Brennan, Camroden Associates;
T. Dyess, U. S. EPA-AEERL XI-3
A Simplified Analysis of Passive Stack Flow Rate
Pah I. Chen, Portland State University XI-4
Factors that Influence Pressure Field Extension in New Residential
Construction: Experimental Results
Richard Prill, Washington State Energy Office; W. Fisk
and A. Gadgil, Lawrence Berkeley Laboratory XI-5
IX
-------
Evaluating Radon-Resisant Construction Practices in Florida
John Spears, H. Rector, and D. Wentling, GEOMET Technologies XI-6
Laboratory Investigations for the Search of a Radon-Reducing Material
Lakhwant Singh, J. Singh, S. Singh, and H. Virk, Guru Nanak
Dev University, India XI-7
Session XII: Radon in Water
Risk Assessment Implications of Temporal Variation of Radon and
Radium Well Water Concentrations
Alan J. Siniscalchi, C. Dupuy, D. Brown, and B. Weiss,
Connecticut Department of Health Services; Z. Dembek, M. Thomas,
and N. McHone, Connecticut Department of Environmental
Protection; M. v.d. Werff, U. S. EPA Region 1 XII-1
Seasonal Variability of Radon-222, Radium-226, and Radium-228 in
Ground Water in a Water-Table Aquifer, Southeastern Pennsylvania
Lisa A. Senior, U. S. Geological Survey XII-2
Radon in Tap Water from Drilled Wells in Norway
Bjern Lind and T. Strand, National Institute of Radiation Hygiene XII-3
A Rapid On-site Detector of Radon in Water
Lee Grodzins, Massachusetts Institute of Technology, and
S. Shefsky, NITON Corporation XII-4
-------
Poster Papers
Session II Posters: Federal Programs and Policies Relating to Radon
Radon Measurement Proficiency Program: New Exam and Listing
for Individuals
G. Lee Salmon and P. Jalbert, U. S. EPA, Office of
Radiation Programs IIP-1
Social and Economic Considerations in the School Evaluation Program
Jed Harrison, U. S. EPA, Office of Radiation Programs IIP-2
The Health of the Radon Industry - Survey and Program Results
from Radon Proficiency Program Analyses
James Long, U. S. EPA, Office of Radiation Programs IIP-3
Session III Posters: State and Local Programs and Policies
Relating to Radon
The Radon Health Effects Committee Report and Its Consequences:
Getting Results in Radon Policy Development
Kate Coleman, E. Fox, and F. Frost, Washington State
Department of Health IHP-1
Washington State's Innovative Grant: School Radon Action Manual
Linda B. Chapman, Washington State Department of Health IIIP-2
Teaming Up on Local Radon Issues
Robert Leker, State of North Carolina IIIP-3
Session V Posters: Radon Measurement Methods
A Decision-Theoretic Model for Evaluating Radon Test Procedures
Based on Multiple Short-term Measurements
Harry Chmelynski, S. Cohen & Associates VP-1
Operational Evaluation of the Radon Alert Continuous Radon Monitor
Emilio B. Braganza, III and R. Levy, U. S. EPA-LVF VP-2
A New Design for Alpha Track Detectors
Raymond H. Johnson, Key Technology, Inc VP-3
XI
-------
Measurements of Indoor Thoron Levels and Disequilibrium Factors
Yanxia Li and S. Schery, New Mexico Institute of Mining
and Technology; B. Turk, Mountain West Technical Associates VP-4
Comparison of Continuous and Occupancy Time Radon Measurements
in Schools Using Programmable E-Perms
Marvin Haapala, C. DeWitt, R. Power, and R. Fjeld,
Clemson University VP-5
Indoor Radon in New York State Schools
Susan VanOrt, L Keefe, W. Condon, K. Rimawi, C. Kunz,
and K. Fisher, New York State Department of Health VP-6
Session VI Posters: Transport and Entry Dynamics of Radon
Simplified Modeling of the Effect of Supply Ventilation on Indoor
Radon Concentrations
David Saum, Infiltec; M. Modera, Lawrence Berkeley Laboratory;
K. Leovic, U. S. EPA-AEERL VIP-1
Determination of Minimum Cover Thickness for Uranium Mill
Tailings Disposal Cells
Jeffrey Ambrose and D. Andrews, CWM Federal
Environmental Services, Inc VIP-2
A Mathematical Model Describing Radon Entry Aided by an Easy
Path of Migration Along Underground Tunnels
Ronald B. Mosley, U. S. EPA-AEERL VIP-3
Radon Diffusion Studies in Soil and Water
Manwinder Singh, S. Singh, and H. Virk, Guru Nanak
Dev University, India VIP-4
Stack Effect and Radon Infiltration
Craig DeWitt, Clemson University VIP-5
Relative Effectiveness of Sub-Slab Pressurization and
Depressurization Systems for Indoor Radon Mitigation:
Studies with an Experimentally Verified Model
Ashok J. Gadgil, Y. Bonnefous, and W. Fisk,
Lawrence Berkeley Laboratory VIP-6
XII
-------
Session VII Posters: Radon Reduction Methods
Radon Mitigation Systems - A Liability in Cold Climate Homes?
Kenneth D. Wiggers, American Radon Services, Ltd VHP-1
Why We Like Diagnostics
John W. Anderson, Jr. and J. Bartholomew, Jr.,
Quality Conservation VIIP-2
An Approach to Computer-Assisted Radon Mitigation
Hormoz Zarefar, P. Chen, and P. Byrne, Portland State University;
C. Eastwood, Bonneville Power Administration VIIP-3
Radon Control - Field Demonstrations: Diagnostic and Mitigation
Techniques Used in Twenty-Six Radon Field Workshops
Craig E. Kneeland and M. Watson, New York State Energy Office;
W. Evans, Evanshire Company, Ltd.; T. Brennan,
Camroden Associates VIIP-4
Radon Mitigation at Superfund Remedial Action Sites: Field
Experience and Results
Jean-Claude Dehmel, S. Cohen & Associates; R. Simon,
R. F. Simon Company, Inc.; E. Fisher, U. S. EPA, Office of
Radiation Programs VIIP-5
Dose and Risk Projection for Use of Sub-Slab Radon Reduction
Systems Under Realistic Parameters
Larry Jensen, U. S. EPA Region 5; F. Rogers and C. Miller,
Centers for Disease Control VIIP-6
Session VIII Posters: Radon Occurrence in the Natural Environment
Influence of Meteorological Factors on the Radon Concentration
in Norwegian Dwellings
Terje Strand and N. B0hmer, Norwegian National Institute
of Radiation Hygiene VIIIP-1
Soil Radon Potential Mapping and Validation for Central Florida
Kirk K. Nielson and V. Rogers, Rogers and Associates;
R. Brown and W. Harris, University of Florida; J. Otton,
U. S. Geological Survey VIIIP-2
XIII
-------
Correlation of Indoor Radon Screening Measurements with Surficial
Geology Using Geographic Information Systems
Charles Schwenker, J-Y Ku, C. Layman, and C. Kunz,
New York State Department of Health VIIIP-3
Analysis of Indoor Radon in New Mexico Using Geographic
Information Systems (GIS)
Richard A. Dulaney, Lockheed Engineering and Sciences Co VIIIP-4
A Radon "Pipe" (?) in the Brevard Fault Zone Near Atlanta, Georgia
L. T. Gregg and J. Costello, Atlanta Testing & Engineering VIIIP-5
Session IX Posters: Radon Surveys
Summary of Regional Estimates of Indoor Screening
Measurements of 222-Radon
Barbara Alexander, N. Rodman, and S. White, Research Triangle
Institute; J. Phillips, U. S. EPA, Office of Radiation Programs IXP-1
Texas Residential Radon Survey
Charles Johnson, G. Ramirez, and T. Browning, Southwest Texas
State University; G. Smith, P. Breaux, and V. Boykin, Texas
Department of Health IXP-2
Radon Survey of Oregon Pubic Schools
Ray D. Paris and G. Toombs, Oregon Health Division IXP-3
Quality Assurance in Radon Surveys
William M. Yeager, R. Lucas, and J. Bergsten, Research Triangle
Institute; F. Marcinowski and J. Phillips, U. S. EPA, Office of
Radiation Programs IXP-4
Radon in Houses Around the Plomin Coal Fired Power Plant
N. Lokobauer, Z. Franic, A. Bauman, and D. Horvat,
University of Zagreb, Croatia IXP-5
A Radon Survey at Some Radioactive Sites in India
Jaspal Singh, L. Singh, S. Singh, and H. Virk, Guru Nanak
Dev University, India IXP-6
Islandwide Survey of Radon and Gamma Radiation Levels
in Taiwanese Homes
Ching-Jiang Chen, C-W Tung, and Y-M Lin, Taiwan Atomic
Energy Council IXP-7
XIV
-------
Session X Posters: Radon in Schools and Large Buildings
Solar Fresh Air Ventilation for Radon Reduction
Monty Holmes, Intermountain Radon Service, and Kelly
Leovic, U. S. EPA-AEERL XP-1
Characteristics of School Buildings in the U. S.
Kelly Leovic, U. S. EPA-AEERL; H. Chmelynski, S. Cohen
& Associates XP-2
Radon in Schools in Wisconsin
Conrad Weiffenbach and J. Lorenz, Wisconsin Bureau
of Public Health XP-3
Investigation of Foundation Construction Details to Facilitate Subslab
Pressure Field Extension in Large Buildings
Michael E. Clarkin, Camroden Associates; F. McKnight,
H. L Turner Group; K. Leovic, U. S. EPA-AEERL XP-4
Radon Measurements in the Workplace
David Grumm, U. S. EPA, Office of Radiation Programs XP-5
Radon Survey of Oregon Public Schools
George L. Toombs and R. Paris, Oregon Health Division XP-6
Session XI Posters: Radon Prevention in New Construction
Model Standards and Techniques for Control of Radon in New Buildings
David M. Murane, U. S. EPA, Office of Radiation Programs XIP-1
Combined Ventilation and ASD System
David Saum, Infiltec, and F. Sickels, New Jersey Department
of Environmental Protection XIP-2
Evaluation of Passive Stack Mitigation in 40 New Houses
Michael Nuess, Washington State Energy Office XIP-3
Radon Remediation and Life Safety Codes
Lyle Sheneman, Chem-Nuclear Geotech, Inc XIP-4
A Passive Stack System Study
Geoffrey Hughes and K. Coleman, Washington State
Department of Health XIP-5
xv
-------
Session XII Posters: Radon in Water
Radon in Water Measurements Using a Collector-Bubbler
Robert E. Dansereau and J. Hutchinson, New York State
Department of Health XIIP-1
Measurements of Radon in Water via Sodium Iodide Detectors
Paul N. Houle, East Stroudsburg University, and D. Scholtz,
Prosser Laboratories XIIP-2
Continuous Measurement of the Radon Concentration in Water Using
Electret Ion Chamber Method
Phillip K. Hopke, Clarkson University, and
P. Kotrappa, Rad Elec, Inc XIIP-3
Performance Testing the WD200 Radon in Water Measurement System
George Vandrish and L Davidson, Instruscience Ltd XIIP-4
Temporal Variations in Bedrock Well Water Radon and Radium, and
Water Radon's Effect on Indoor Air Radon
Nancy W. McHone and M. Thomas, Connecticut Department of
Environmental Protection; A. Siniscalchi, Connecticut Department
of Health Services XIIP-5
XVI
-------
Session VII
Radon Reduction Methods
-------
VIM
DURABILITY OF SUB-SLAB DEPRESSURIZATION RADON MITIGATION SYSTEMS
IN FLORIDA HOUSES
C. E. Roessler and R. Varas
Dept. of Environmental Engineering Sciences and
D. E. Hintenlang, Dept. of Nuclear Engineering Sciences
University of Florida, Gainesville, FL 32611
ABSTRACT
Durability of sub-slab depressurization radon mitigation
systems installed in demonstration projects in nine North Florida
slab-on-grade houses is being studied. Installations include
several combinations of single suction point and two suction
point, single fan systems. All the systems were operating
without any reported equipment failures after post-installation
periods ranging from 22 to 37 months. In terms of indoor radon
concentrations during the six months of follow-up (Fall-Winter
1991-92), eight of the nine systems were effective in maintaining
levels below 4 pCi/L in houses that had pre-mitigation values in
the range of 9 to 25 pCi/L and seven of the systems were as
effective as or more effective than when initially installed. In
one house that had a pre-mitigation concentration of 30 pCi/L,
the six-month follow-up concentration was 4.4 pCi/L as compared
to the early post-mitigation concentration of 2.5 pCi/L. Evi-
dence of a sub-slab "short-circuit" at this house is being
investigated.
This work was funded in part by Florida State
University System, Board of Regents Radon
Research Program Contract 089037, by the
Florida Department of Community Affairs Radon
Research Program, and by U.S. Environmental
Protection Agency Assistance Agreements CR
814925 and CR 817367.
This paper has been reviewed in accordance
with the U.S. Environmental Protection Agen-
cy's peer and administrative review policies
and approved for presentation and publica-
tion.
-------
DURABILITY OF SUB-SLAB DEPRESSURIZATION RADON MITIGATION SYSTEMS
IN FLORIDA HOUSES
INTRODUCTION
This is a preliminary report of a follow-up and durability
study of sub-slab depressurization (SSD) radon1 mitigation sys-
tems installed in nine existing North Florida houses during 1989-
90.
In North Central Florida, the housing stock is primarily
slab-on-grade and the sub-slab medium typically consists of
native soil and sand. SSD systems were installed in nine houses
between May 1989 and August 1990 as part of a demonstration-
research project to evaluate radon mitigation techniques for this
region.
Several questions were posed concerning durability for
systems operating under Florida conditions. Would continued
operation impact the sub-slab environment in a manner that
affects the continued effectiveness of the system? If there were
effects on the sub-slab environment, would these have structural
effects on the building? Since system flow rates are typically
low because of the low permeability of the sub-slab medium/ would
continued performance of the fans be compromised by the low flow
and by the high temperature in Florida installations?
Early system performance was followed through December 1990
(for post- installation periods ranging from 4+ to 19+ months).
A follow-up study of 1-year duration was initiated September
1991; at the time of this writing, two calendar quarters of data
were available.
BACKGROUND: SUMMARY OF HOUSES AND MITIGATION SYSTEMS
HOUSES
•The study involved nine slab-on-grade houses, six in
Gainesville (Alachua County) and three in Ocala (Marion County).
The general location is shown in Figure 1; characteristics of the
study houses are summarized in Table 1. Seven had simple
rectangular floor plans with slab areas on the order of 150 to
200 m2 (1600 to 2100 ft2) and two had more complex, L-shaped
designs with slab areas on the order of 200 to 210 m2 (2100 to
2200 ft2) . Several combinations of slab and wall type are
this report, the terms "radon" and "Rn" are used to
designate the radon-222 isotope.
-------
Tallahassee
Gainesville (Alachua Co.)
Ocala (Marion Co.)
Jacksonville
Miami
Figure 1. Location of Study Houses.
-------
TABLE 1. FEATURES OF STUDY HOUSES & RADON MITIGATION SYSTEMS.
House Number
House Characteristics:
Age (yrs)
Slab footprint
Size, sq m
Size, sq ft
Type
Wall Construction
Air Handler Location
Return Air Route
Sub-slab Communication
Pre-mitigation Rn, pCi/L:
Subslab
Indoor
Rn Mitigation System Characteristics:
Installation Date:
System A: Fan
Pit #1 diameter, m
Pit #2 diameter, m
System B: Fan
Pit diameter, m
Continuous Data Acquisition
Post-mitigation Indoor Rn, pCi/L
O-l
14
Rectang.
167
1800
F
CB
UTIL
ATTIC
Poor
3880
20
May '89
R-150
0.92
0.92
—
—
no
1.5
O-2
17
Rectang.
164
1760
S
FR
GARAGE
ATTTC+TW
Good
3940
11
May '89
R-150
0.92
0.82
—
—
no
2.1
270
G-l
9
Rectang.
164
1760
s,
FR/BV+CB
GARAGE
TW
Poor
2820
12
Jul '89
R-150
0.92
0.84
—
—
yes
3.5
7.641
G-2
14
Rectang.
194
2087
F
CB/VN
GARAGE
ATTIC
Poor
4270
26
Nov '89
R-150
0.92
0.52
—
—
yes
2.5
G-3
19
Rectang.
158
1700
F
CB
CLOSET
TW(Door)
Good
8600
9
Oct '89
T-l
0.92
—
—
—
yes
1.4
G-4
20
Rectang.
181
1950
F
CB/BV
CLOSET
ATTIC
Fair
3260
20
May '90
R-150
0.61
0.66
—
yes
2.0
O-3
14
Rectang.
149
1608
P
CB
ATTIC
ATTIC
Fair
9260
30
Aug '90
R-150
0.71
0.71
no
2.0
G-5
9
L-shaped
195
2100
S
FR+STV
GARAGE
ATTIC+TW
Fair
4300
25
Jul '90 *
R-150
0.86
0.71
EXP
0.56
no
2.5
G-6
13
L-shaped
203
2188
F
CB
GARAGE
TW
Good
3400
10
Ju! '90 *
R-150
0.61
0.81
...
no
2.5
Slab type:
Wall construction:
Return air:
Sub-slab communication:
Pressure Extension:
Fan Type:
O = Ocala, (Marion Co.), G = Gainesville, (Alachua Co.).
F = floating, S = sealed stem-wall, P = partially sealed stem-wall
FR = wood frame, CB = concrete block, BV = brick veneer, STV = stone veneer, VN = vinyl siding
TW = through wall with no ducting, ATTIC = ducts running through attic
Poor = pressure extension of 0-0.3 m (0-1 ft), Fair = 0.3-3 m (1-10 ft), Good = greater than 3 m (10 ft).
The farthest distance at which a pressure differential of at least 0.25mm of water can be measured when a pressure
of 500 Pa is applied at a suction hole.
R-150=Fantech model R-150, 270 cfm (4.47 cubic m/min), 2150 rpm, 1/20 hp (37.3 J/s);
T-l = R.B. Kanalflakt Turbo model, 158 cfm (4.47 cubic m/min), 2800 rpm, 1/40 hp (18.6 J/s); EXP = Experimental.
* G-5 and G-6 systems turned on in July'90, became fully effective in October '90.
-------
represented. All had central heating/air conditioning systems
and several return air configurations are represented.
Diagnostic testing prior to mitigation system design
included sub-slab communication measurements. Effective
permeabilities of the sub-slab media, determined in-situ from
observations of flow vs suction pressure, were on the order of
10"11 to 10"10 m2. Pressure extension was defined as the farthest
distance at which a pressure differential of at least 0.25 Pa
could be measured when a pressure of 500 Pa was applied at a
suction hole. Communication values were equally divided among
poor (pressure extension of 0-0.3m or 0-1 ft), fair
(pressure extension 0.3 -3m or 1-10 ft) and good (pressure
extension >3 m or >10 ft). Thus in a number of houses, the SSD
systems had to deal with communication distances considerably
less than half the major dimension of the house.
Sub-slab radon concentrations as measured by grab samples
during diagnostic testing ranged from 2800 to 9300 pCi/L.
Average pre-mitigation indoor radon concentrations ranged from 9
to 30 pCi/L.
MITIGATION SYSTEMS
The systems were installed between May 1989 and August 1990;
the installation schedule is diagramed in Figure 2. Character-
istics of the SSD systems as installed are summarized in Table 1.
Systems with two suction points were installed in eight of the
houses. Only a single suction point was installed in house
Gainesville-3 — it was of smaller dimensions (slab area = 160 m2
or 1700 ft2) , had good sub-slab communication, and had a con-
venient location for a central suction point. A single-point,
single-fan system was installed in addition to a two-point,
single-fan system in house Gainesville-5 — this house had a
complex floor plan with several slab rectangles and only fair
sub-slab communication.
Each suction point consisted of a roughly hemispherical pit
approximately 0.5 to 1.0 m in diameter. PVC pipes 10 cm (4 in.)
in diameter were installed through the slabs and 5 cm (2 in.) PVC
pipes were routed to the fans which were located in the attic
with discharge vertically through the roof.
INITIAL SYSTEM PERFORMANCE
Initial system performance is summarized in Table 2 (also
see Figure 3). The systems were generally effective in reducing
indoor radon concentrations from pre-mitigation levels in the
range of 10-30 pCi/L to post-mitigation values of <4 pCi/L.
Levels were reduced to values on the order of 2 pCi/L or less in
four houses, (see post-mitig. column in Table 2).
-------
EARLY MONITORING PERIOD
FOLLOW-UP PERIOD
Fall Win. Sor. Sur
system not effective until Oct. '90
system not effective until Oct. "90
1989
1990
1991
1992
Figure 2. Radon Mitigation System Installation and Follow-up Time Line.
-------
TABLE 2. INDOOR RADON CONCENTRATION IN STUDY HOUSES.
House
No.
G-l
G-2
G-3
G-4
G-5*
G-6*
O-l
O-2
O-3
House
Shape
Rectang.
Rectang.
Rectang.
Rectang.
L-shaped
L-shaped
Rectang.
Rectang.
Rectang.
Number
of
Suction
Points
2
2
1
2
3
2
2
2
2
Date
System
Began
Operation
Jul-89
Nov-89
Oct-89
May-90
Jul-90
Jul-90
May-89
May-89
Aug-90
Months of
Operation
as of
Jun-92
35
31
32
25
23
23
37
37
22
Radon Concentrations in pCi/L
Pre-
Mitig.
12
26
9
20
25
10
20
11
30
Post-
Mitig.
3.5
2.5
1.4
2.0
2.5
2.5
1.5
2.1
2.0
Fall-91
80-day
A-T
2.2
2.0
1.6
2.5
4.3
3.3
0.6 #
1.8 @
6.3
Wint-92
90-day
A-T
1.6
1.3
0.6
1.4
2.1
1.8
0.6 #
1.2
2.5
Fall-
Winter
Average
1.9
1.7
1.1
2.0
3.2
2.6
0.6
1.5
4.4
Icey to House No.: G = Gainesville, (Alachua Co.), O = Ocala, (Marion Co.)
* G-5 & G-6 systems turned on in July; installation fully effective in October 1990.
# O-l Indoor readings for Fall '91 and Winter '92 are from a single 170-day reading.
@ O-2 : Monitored for the last 35 days of the Fall '91 period.
-------
APPROACH AND METHODS
EARLY PERFORMANCE MONITORING (1989-90)
Following the phased-in installation and tuning of the
mitigation systems between May 1989 and August 1990, performance
was followed through the end of 1990 (See Figure 2). Indoor
radon was monitored by continuous2 and/or integrating3 radon
monitors, and house visits were conducted to observe system
operation, measure system pressures and flows, service radon
monitoring equipment, and receive and respond to homeowner
questions and problems. Post-installation monitoring was per-
formed according to the following general schedule:
• Stage 1 (in service <6 months) - Continuous and/or
integrating radon monitoring; house visits on a bi-weekly
basis.
• Stage 2 (in service 6 to 12 months) - Houses without data
loggers were visited on a monthly schedule. For houses with
data loggers, data acquisition was continued, data were
reviewed, and visits were performed as necessary.
• Stage 3 (in service >12 months) - Visits on an approximately
6-month schedule to inspect the systems, measure pressures
and flows, and deploy radon monitors for week-long
measurements.
FOLLOW-UP STUDY (1991-92)
The durability follow-up phase, initiated in September 1991
for a 1-yr duration, involved (1) indoor radon monitoring to
determine continued effectiveness in radon mitigation and (2)
quarterly house visits to observe system operation, measure
system pressures and flows, observe house conditions, and receive
and respond to homeowner questions and comments.
Radon monitoring is being performed, with passive integrating
(alpha track) detectors4 deployed for four consecutive calendar
2Five houses were instrumented with continuous data
acquisition systems (Model 21X Micrologger®, Campbell Scientific,
Inc., Logan, UT); radon detectors were At Ease® Model 1021 radon
monitors (Sun Nuclear Corporation, Melbourne, PL). In the four
houses without continuous data acquisition systems, At Ease®
Model 1023 radon monitors were deployed as stand-alone monitors.
3Electret ionization chambers (E-PERM®, Rad Elec Inc.,
Fredrick, MD).
4Rad-trak®, Tech/Ops Landauer, Inc., Glenwood, IL.
-------
quarters. Two detectors are deployed in each house, one in each
of two separate rooms in the living area.
RESULTS
EARLY PERFORMANCE
This early post-installation phase of durability monitoring
from the time of installation until the end of 1990 involved
durations of 4 to 19 months for the various houses. During this
limited period, the following were observed:
1. With the transient exceptions noted below, the systems
settled to relatively constant performance after an initial
adjustment/stabilization period and retained their
effectiveness in maintaining reduced indoor radon
concentrations.
2. In one house, failure of caulking of a leakage crack near a
suction point resulted in "short-circuit" flows. This was
remediated by re-caulking with a more durable material.
3. With the advent of cool weather in 1989, condensation from
moist exhausted sub-slab air occurred in three systems that
had undrained low points in horizontal attic runs. This
resulted in an audible gurgling noise and compromised
performance (reduced flow and fluctuations in indoor radon
concentrations). This was alleviated in these systems by
installing condensate drain lines from the "traps" and was
prevented in subsequent installations by ensuring self-
draining through avoidance of low points where possible
and/or the installation of drain lines.
4. No fan failures were observed. Any effect of low flow on
fan life was not expressed in this limited sample during the
available observation period.
5. No structural effects were observed.
6. With the exception of the "gurgling" associated with the
water condensation before installation of traps and drains,
there were no homeowner complaints of noise or other
annoyances.
FOLLOW-UP STUDY (1991-92)
This 1-yr monitoring phase began in September 1991 after the
systems had been in operation for periods ranging from 13 to 28
months. At that time, all the systems were operating and there
were no homeowner complaints.
As of June 1992, house visits had been performed for the
Fall, Winter, and Spring quarters and the systems had been in
operation for periods ranging from 22 to 37 months. Indoor radon
data were available for two quarterly monitoring periods. Radon
-------
data are presented and compared to pre-mitigation and early post-
mitigation concentrations in Table 2 and Figure 3. The post-
mitigation and follow-up study concentrations are presented on an
expanded scale in Figure 4.
A striking feature of these data is the fact that all the
indoor radon concentrations for the Fall quarter were greater
than those for the winter quarter — Fall/Winter ratios ranged
from 1.4 to 2.7 and averaged 1.9. This is the inverse of the
seasonal effect seen in unmitigated houses in two previous
Gainesville studies (1,2) and in other Florida cities (2).
Measurements of pressure and flow did not indicate any apparent
system differences between the two quarters. Comparison of exit
air radon concentrations did not indicate any consistent differ-
ences between quarters that could be associated with the Fall-
Winter indoor radon differences. To date, no satisfactory
explanation has been proposed.
In all houses, the follow-up indoor radon concentrations (6-
month average) remained significantly less than the pre-mitiga-
tion values. In most houses (8 of 9), the concentrations were <4
pCi/L. In 7 of the 9 houses, the concentrations were comparable
to or lower than the early post-mitigation reference concentra-
tions; this suggests that these systems are at least as effective
as when initially installed.
In the two exceptions, both mitigated in the second half of
1990, the Fall quarter concentrations were >4 pCi/L and
noticeably higher than the early post-mitigation values. How-
ever, these follow-up/reference comparisons should be interpreted
with reservation. The early post-mitigation reference values for
1990 may be biased by seasonal effects — especially regarding
Gainesville-5, Gainesville-6, and Ocala-3 for which the post-
mitigation observation period was confined to Fall (or late Fall)
and December 1990.
Ocala-3, a rectangular house with a 2-point, single-fan
system, was the only house with a Fall-Winter average exceeding 4
pCi/L. The average of 4.4 pCi/L, based on a Fall concentration
of 6.3 pCi/L and a Winter value of 2.5 pCi/L, was greater than
the 2.0 pCi/L early reference value from observations during
September-December 1990. At this house, the post-mitigation flow
rate at one of the suction points was higher than typical for
these systems and about 20 times that at the other suction point;
this suggested some "short-circuiting". At the follow-up, the
flow imbalance was even greater and the radon concentration in
the exit air was less than 100 pCi/L. This is strong circum-
stantial evidence for intensified "short-circuiting" with
excessive dilution by house air drawn through a slab leakage
and/or ambient air drawn through a pathway to the foundation
perimeter. This is being investigated.
-------
4 pd/L max.
O1 G3 O2 G2 G1 G4 G6 G5 03
Pre-Mitig.
Post-Mitig.
Fall '91
Winter'91-'92
Figure'3. Pre-mitigation and Post-mitigation Indoor Radon Concentrations.
-------
4pCi/L EPA suggested maximum concentration
Ay:*V^«OT:-.V/;Xl«ySy-lY/_V:ViffiSS»S5^
PRE-MITIGATED
CONCENTRATIONS
O1 G3 O2 G2 G1 G4 G6 G5 O3
Post-Mitigation I I Fall '91
Winter'91-'92
Figure 4. Post-mitigation and Follow-up Indoor Radon Concentrations.
-------
In the remaining house, Gainesville-5, the Fall 1992 value
was slightly greater than 4 pCi/L and the Fall-Winter average of
3.2 pCi/L was slightly greater than the late 1990 early reference
value of 2.5 pCi/L. These comparisons may or may not be sig-
nificant — the early post-mitigation value was based on limited
observations during late Fall and December of 1990 and the
follow-up study phase was characterized by a 1991-92 Fall/Winter
seasonal ratio of >1.0 when a value of <1.0 would have been
expected from previous experience.
System flows and pressures and other observations will be
presented in a forthcoming report.
SUMMARY AND CONCLUSIONS
Durability studies are being performed on several configura-
tions of SSD systems installed during 1989 and 1990 in nine North'
Florida slab-on-grade houses. Early observations were made from
the respective times of installation to the end of 1990 and a
follow-up study of 1-yr duration was initiated in the Fall of
1991. Follow-up information to date includes indoor radon
monitoring (continuous integrating sampling) data for the Fall
1991 and Winter 1991-92 quarters and other observations into June
1992, and represents post-installation periods ranging from 22 to
37 months for the various houses. The study demonstrated that:
1. All the systems remained in operation without any reported
failures.
2. All of the systems continued to provide significant
reductions from the pre-mitigation concentrations.
3. Eight of the nine systems (89%) were effective in main-
taining indoor radon concentrations below 4 pCi/L in houses
that had pre-mitigation values in the range of 9 to 25
pCi/L.
4. In terms of indoor radon concentrations, seven of the nine
systems (78%) were as effective as or more effective than
when initially installed.
5. In the eighth house (Gainesville-5), the system was
effective in that the Fall-Winter average concentration was
3.2 pCi/L as compared to the pre-mitigation value of 30
pCi/L. However, comparison to early effectiveness (2.5
pCi/L) was inconclusive since this comparison may be con-
founded by year-to-year differences in seasonal effects.
6. The ninth system (Ocala-3), a two-suction point system,
showed evidence of a sub-slab "short circuit" that com-
promised performance. This house had a Fall-Winter 6-month
-------
average indoor radon concentration of 4.4 pCi/L as compared
to the initial post-mitigation reference value of 2.0 pCi/L.
This case is being further investigated.
7. During the first cold period following some of the 1989
installations, unintended trapping of moisture condensation
in horizontal attic runs compromised performance and drains
had to be installed in these systems. Lessons learned were:
• Moisture condensation in the SSD system during cool
weather is a potential problem, even in Florida.
• New installations should be designed to be self-
draining through avoidance of low points ("traps")
where possible and/or by installation of drain lines.
• Maintenance should include inspections for these and
other inadvertent effects.
8. The Fall 1991 indoor radon concentration was greater than
the Winter 1991-92 value in all of these houses; the average
ratio was 1.9. This is a reversal of the "conventional
wisdom" about seasonal effects in unmitigated Florida
houses.
REFERENCES
1. Roessler, C.E., Revell, J.W. and Wen, M.J. Temporal Patterns
of Indoor Radon in North Central Florida and Comparison of
Short-term Monitoring to Long-term Averages. In; Proceed-
ings: The 1990 International Symposium on Radon and Radon
Reduction Technology: Volume 1. EPA-600/9-91-026b (NTIS
PB91-23444). U.S. Environmental Protection Agency, Research
Triangle Park, North Carolina, 1991. P. 3-131.
2. McDonough, S., Williamson, A. and Sanchez, D.C. Correlation
Between Short- and Long-term Indoor Radon Concentrations in
Florida Houses. In; Proceedings: The 1991 International
Symposium on Radon and Radon Reduction Technology: Volume 3.
(EPA-600/9-91-037C (NTIS PB92-115377) U.S. Environmental
Protection Agency, Research Triangle Park, North Carolina
1991. p. PNL3-21.
-------
VI1-2
A NOVEL BASEMENT PRESSURIZATION-ENERGY CONSERVATION
SYSTEM FOR RESIDENTIAL RADON MITIGATION
by: K. J. Renken and S. J. Konopacki
University of Wisconsin-Milwaukee
Radon Research Laboratory
Department of Mechanical Engineering
P.O. Box 784
Milwaukee, Wisconsin 53201
U.S.A.
ABSTRACT
Basement pressurization is one method of radon mitigation
which has vast potential. It has been theorized that
pressurization could provide significant reduction in radon
levels for a house with a "tight" basement. The largest drawback
of the technique is the energy penalty associated with increased
air infiltration during the heating season.
This paper presents the results of an investigation on a
system that produces radon reduction by basement pressurization
while conserving heating fuel and providing improved air quality
as well as human comfort. A secondary heat exchanger that
increases the efficiency of a conventional residential furnace is
modified to provide fresh air heat exchange and pressure
regulation. The system has subsequently been named "The Radon
Buster". A Wisconsin home with an initial average radon level of
33.5 pCi/L was chosen for the retrofit. Following the
installation of the unit, the average radon concentration
diminished to 5.7 pCi/L and the heating bill reduced enough to
operate the system at an estimated zero cost.
Measurements of radon reduction levels, fuel usage, and
environmental factors that affect radon migration are documented.
A state-of-the-art radon data acquisition system with
accompanying instrumentation are also described.
This paper has been reviewed in accordance with the U.S.
Environmental Protection Agency's peer and administrative review
policies and approved for presentation and publication.
-------
INTRODUCTION
The method of basement pressurization has received very
little attention in the radon mitigation literature. The method
prevents radon soil gas from entering a building by providing a
positive pressure shield which causes air to flow through cracks
and holes in the basement out into the soil. The method
simultaneously reduces radon concentrations by diluting the radon
laden air with fresh air supplied by a blower. Long term data
has not been published to show the effectiveness and reliability
of the technique, its relationship with energy costs or its
lengthy structural effects.
A review of pertinent investigations on the subject reveals
only a handful of papers that can be associated with this study.
One of these is a study of natural basement ventilation by
Cavallo et al. (1). They have shown the method to be effective
as a radon reduction measure when the basement is slightly
pressurized and highly diluted. However, they discovered that
during the heating and cooling seasons a severe energy penalty
was assessed since the outdoor air must be either heated or
cooled to the indoor air temperature.
The Heat Recovery Ventilator (HRV) method is another proven
technique that reduces radon levels by combining basement
pressurization with dilution. Heat Recovery Ventilation employs
an air-to-air heat exchanger to recover energy from stale
radon-laden indoor air with fresh outdoor air. The reduction
will solely be attributed to dilution if the inlet and outlet air
streams are of equal flow rate. Basement pressurization can be
realized if the inlet air stream is at a higher flow rate than
the outlet air stream. However, this method imposes a demanding
energy penalty during the heating and cooling seasons. A study
utilizing a balanced HRV by Nazaroff et al. (2) concluded that
radon reduction below the EPA guidelines could be reached with
this method. Another study employing a balanced HRV by Holub et
al. (3) determined the same method could reduce radon levels by
about a factor of six.
Brennan et al. (4) achieved radon reduction through basement
pressurization by sealing the return ductwork and creating an
opening in the supply ductwork. Their results showed that the
method was as effective as soil depressurization at controlling
indoor radon levels.
The objectives of our research were (1) to develop an
effective system of radon mitigation using basement
pressurization with the additional benefit of heating energy
conservation and (2) to characterize the environmental effects
(e.g. pressure differentials, soil and ambient temperatures,
-------
precipitation, barometric pressure, wind velocity and direction,
humidity, etc.) that influence indoor radon entry and egression.
The final product of our research efforts has been termed "The
Radon Buster."
DESCRIPTION OP RADON BUSTER
A secondary heat exchanger is the main component of the
Radon Buster design. In general, the purpose of a secondary heat
exchanger is to improve the heating efficiency of industrial
boilers, unit heaters and residential conventional furnaces by
increasing the heating capacity of the existing heating
equipment, thus reducing fuel consumption.
In our design, the secondary heat exchanger is arranged in
counter flow orientation with unmixed air and flue gas streams.
The flue gas is distributed through a copper header (20" long, 3"
O.D.) into 23 sinusoidal copper tubes (5' straight length, 7/8"
O.D.), recombined in another header and exhausted to the outdoors
with an exhaust blower. The external dimensions of the heat
exchanger are 40" x 20" x 10". A fan is required to move the
flue outdoors because it is no longer hot enough to rise through
the entire length of the chimney, therefore requiring the chimney
to be capped off. The return air moves through the heat
exchanger and is preheated before entering the furnace whdre it
is then heated to a much higher temperature than the normal feed
air. This results in an elevated supply air temperature causing
a reduction in furnace operation, which decreases fuel
consumption and heating costs. A backdraft damper is placed on
the outlet end of the flue pipe to prevent outdoor air from
entering.
The conventional furnace/secondary heat exchanger system
operates in the following manner. The home thermostat calls the
furnace for heat and closes an electrical path to the exhaust
blower. The blower switches on and creates a vacuum within the
secondary heat exchanger and subsequently closes the vacuum
switch. When the vacuum switch closes it allows the electronic
ignition to ignite the pilot and the furnace operates normally.
The vacuum switch is a safety feature which will not allow the
furnace to operate if the exhaust blower fails. As the
thermostat opens and calls for the furnace to cease operation the
furnace blower continues to circulate air through the house until
the flue inlet header in the secondary heat exchanger cools to
135°F. At this point a thermostat on the flue inlet header opens
and cuts power off from the furnace blower.
A Radon Buster is a modified secondary heat exchanger with
the same internal and external dimensions which is attached to
the furnace in the same way. The major difference is that the
-------
outdoor air is the supply air instead of the return air. Cold
outdoor air moves through the Radon Buster and is preheated then
deposited into the basement. The thermostat on the flue inlet
header is electrically connected to the air side blower on the
heat exchanger instead of the furnace blower. A check valve is
situated at the air side outlet obstructing the flow of air to
the outdoors while maintaining positive basement pressure. The
Radon Buster is illustrated schematically in Figure 1.
DESCRIPTION OF EXPERIMENTAL SITE
The experimental site was the residence of Peter and Judy
Roidt located in Hartland, Wisconsin (35 miles west of Milwaukee,
Wisconsin). The house is a 35 year old conventional single floor
ranch style over a full basement, slightly shielded by trees and
other houses on three sides and open in front. The basement is
adequately sealed from the upstairs living area by a tight
plywood sub-floor, and consists of a poured concrete floor,
hollow-block walls, two sump wells, and four tightly sealed well
windows. A water drainage system exists along the east end of
the basement where a thin slice of the slab was removed and weep
holes were drilled. The weep holes and the sump wells are
assumed to be the major entry routes for the indoor radon gas.
Installed in the basement was a Thermopride gas fired furnace
with electronic ignition and a heating capacity of 80,000 Btu/hr.
The upstairs storm windows and doors as well as the basement
windows and stairway were tightly sealed by adding weather
stripping and industrial silicone. This measure preserved the
positive pressure created by the Radon Buster. The house volume
was measured to be approximately 19,200 ft3. A whole house
blower door test determined the natural ventilation rate to be
0.30 ACHnat and the effective leakage area to be 122 in2 (Lawrence
Berkeley Laboratory Air Infiltration Model).
INSTRUMENTATION
The test site was continuously monitored for radon
concentration, differential pressures, and indoor/outdoor
environmental parameters under normal living conditions from
October 1991 to April 1992. Figure 2 shows schematically the
locations of the various sensors that were used during our
measurements.
A Hewlett-Packard Vectra QS/16S computer interfaced with a
Hewlett Packard 75000 Series 3 Mainframe provided the means for
full automation of the data acquisition. The mainframe was
fitted with three Hewlett-Packard plug-in modules which supplied
the interface for the instrumentation. The first was a 5 1/2
-------
31 Flue Pipe (Golv. Steel
Supply —]
Air Duct
Gas-Fired
Furnoce
80,000
Exhaust
Blowr
Control
Ponel
Vocuun —
Safety
Sir itch
Heat Exchanger
Iforn Air
Outlet
•a
•n
Condensate •
Drain
• Return
Air Duct
3' Flue Pipe (PVC)—j
(to Outside)
Inlet
Blo*er
300 cfn
8' Cold Air Inlet—'
(Golv, Steel)
(Iron Outside)
Figure 1. Rodon Buster
-------
Enclosure
10
Wind Breok
2,5
Upper Level
1 7
/
J
\
\
(
Ro
Bu
Fu
do
st
rn
n
er
1 I
oce
Conputer ond
Instrunentotion
-Bosenent . . c K 7 0 n
1, 4, b, o, /, 8, 11
Supply Air Duct 1, 12
\
Bosenent B
1 - Tenperoture
2 - Soil Tenperoture
3 - Boronetric Pressure
4 - Relative Hunidity
5
5 - SoiI / fiosenent Pressure
6 - Outdoor / Bosenent Pressure
7 - Upstairs / Basenent Pressure
8 - Basenent / Basenent Pressure
9 - Precipitation
10 - Kind Velocity / Direction
11 - Rodon Concentration
12 - Airflow
Figure 2. Location of Instrunentotion
-------
digit multimeter which connected a 16-channel thermocouple relay
multiplexer used to interface continuous voltage and current
output sensors to the mainframe. The last plug-in module was a
3-channel universal counter employed to connect the voltage pulse
output sensors.
The menu-driven data acquisition software package "Labtech
Notebook" performed all the necessary data collection and
internal conversions of output to true engineering units for each
sensor and executed calculations between channels. Readings were
taken at 5 minute intervals which allowed for two continuous
weeks of data collection before the computer memory was filled.
Excursions to the test site were taken every two weeks to
down-load the data and check on equipment performance.
Radon concentrations in the basement were measured by a
calibrated Pylon AB-5 continuous radon monitor which used a 300
ml scintillation cell. The unit was connected to the 3-channel
universal counter card of the data acquisition mainframe. In an
effort to verify the readings of the Pylon unit for indoor radon
level, several Tech/Ops Landauer, Inc. Long Term Alpha Track
Radon Gas Monitors were logistically placed in the basement and
the upstairs.
Outdoor weather conditions were recorded in an effort to
determine the parameters affecting radon entry. Outdoor
temperature was measured with a Type-T thermocouple probe
enclosed in an environmental chamber that shielded various
sensors from radiation, wind, rain, snow, insects, animals, and
intrusion.
The soil temperature was measured in two locations: on the
east side of the house at distances of about 1' and 20' from the
base of the house. The soil probe was fashioned from a metal
protection tube fastened to an aluminum protection head and
inserted 18" into the ground.
Outdoor relative humidity was measured with a current output
relative humidity transmitter (General Eastern) which was also
placed in the environmental chamber.
A tipping bucket rain gauge (Climatronics Corp.) was
enclosed within a 4' high circular windbreak. The presence of
the windbreak reduced eddy currents and turbulence around the
gauge providing for a more accurate precipitation measurement.
The rain gauge was mounted level, in the center of the windbreak
and located in a clear area away from trees and buildings. A
thermostatically controlled heater mounted on the unit prevented
it from freezing and allowed snow fall to be measured as well as
rain.
-------
An in-line wind speed and direction transducer (Earth and
Atmospheric Sciences, Inc.) was mounted upon a 20' stand away
from obstructions and connected to the data acquisition system.
Barometric pressure was measured with a low differential
pressure transmitter (Earth and Atmospheric Sciences, Inc.). The
transmitter was located in the basement with the high port
residing in the environmental chamber connected by means of 1/4"
copper refrigeration tubing with the other port sealed.
During pre-mitigation and post-mitigation data collection,
conditions within the house were monitored to observe what affect
the operation of the Radon Buster had on them. Relative humidity
in the basement was measured with a relative humidity transmitter
of the same type as the outdoor sensor. Temperature was measured
in both the basement away from the Radon Buster and in the
upstairs living area.
One of the functions of the Radon Buster is to pressurize
the basement to impede radon gas entry. Differential pressure
measurements were made to examine the change in pressure with the
addition of the mitigation system. Pressure differential was
measured with very low differential pressure transducers (Setra
Systems, Inc.) with a range of +/- 0.25 inches H2O (+/- 62 Pa)
between the basement near the Radon Buster (low port) with
respect to other positions (high port).
Pressure measurements were collected between the basement
and the soil using soil probes. Each soil probe was constructed
from a well point fitted to 1/4" copper refrigeration tubing with
a compression fitting, which in turn was sealed to 1/4"
polyethylene tubing and connected to the high port of a
transducer. These transducer units were placed 4' in the ground
in four positions around the perimeter of the house: west, south,
southeast, and east (north side of test site had a one car
garage). The outdoor/basement pressure differential process line
ran from the high side of a transducer to the environmental
chamber with 1/4" polyethylene tubing. Differential pressures
were also measured between the basement and the upstairs living
area as well as between the Radon Buster area and the far side of
the basement away from the unit.
The Radon Buster was instrumented with four thermocouples
which were inserted into the inlet and outlet sides of both the
air and flue gas streams. A hot wire anemometer transducer (TSI,
Inc.) was employed to measure air velocity of the air inlet
stream which was converted to a volumetric air flow rate.
' Finally, a Gas Alert Plus Alarm was installed to check for
carbon monoxide emissions from the furnace.
-------
DISCUSSION OF RESULTS
PRE-MITIGATION RESULTS
Radon screening and follow-up measurements revealed an
average pre-mitigation radon concentration in the basement of the
test site of 33.5 pCi/L as shown in Table 1.
TABLE 1. PRE-MITIGATION AVERAGE RADON CONCENTRATION
Detector Type Date Location Radon FpCi/LI
Charcoal
3/16/91
- 3/18/91
Basement
Family Room
35.0
22.0
Alpha Track 11/12/91 - 1/18/92
Pylon Monitor 9/11/91 - 9/12/91
10/18/91 - 1/17/92
Basement 32.7
Family Room 21.9
Basement 33.5
Basement 33.6
Radon concentration was measured continuously starting
October 18, 1991 with the Pylon monitor. Figure 3 displays the
extremes in radon levels for the period of November 22 through
December 2, 1991. The radon levels range from 10 pCi/L to 60
pCi/L with an arithmetic mean of 33.6 pCi/L.
Differential pressure measurements between the basement and
the surrounding soil indicated the average soil pressure to be
about 3 Pa higher than that of the basement as evidenced in
Figure 4. This magnitude of pressure difference is indicative of
the natural state of basement depressurization.
Environmental parameters were also studied for their effect
on radon entry. During our pre-mitigation measurements, the
following environmental effects displayed no direct correlation
with radon entry: indoor/outdoor relative humidity, soil
temperature, and wind direction. On the other hand, the
following parameters did show noticeable relationships with
indoor radon levels: precipitation, barometric pressure, and
natural ventilation rate.
Precipitation events greater than 1/4" during the period of
December 20, 1991 through January 2, 1992 were analyzed. For 12
independent events where precipitation was a dominant variable,
we found that the radon concentration increased by 10% - 40% with
the most common value being 30%. These events demonstrated a
direct relationship between precipitation and an increase in
radon concentration which supports the work of Montague and
-------
0 1
Time [Day]
Figure 3. Pre-Mitigation Radon Concentration
(11/22/91 - 12/2/91)
CO
-2
345
Time [Day]
8
Figure 4. Pre-Mitigation Soil / Basement Differential
Pressure (12/4/91 - 12/11/91)
-------
Belanger (5). Also observed is a delay of up to four hours
between the initial rise in radon concentration and the onset of
precipitation. A plot illustrating the relationship between
radon concentration and precipitation is shown in Figure 5.
An inverse relationship between barometric pressure and
radon concentration is observed in Figure 6. A fall in
barometric pressure is usually attributed to precipitation, but
for this particular event no precipitation was measured.
Therefore, it is hypothesized that barometric pressure was the
primary driving force for this event. Barometric pressure will
normally communicate through a building relatively quickly. A
decrease in barometric pressure will then allow soil gas to
migrate into buildings and the atmosphere, thus increasing indoor
radon levels. Whereas an increase in barometric pressure will
retard soil gas movement decreasing the radon concentration. A
similar inverse correlation between the indoor radon
concentration and the rate of change of barometric pressure was
also observed in the basement of a house by Hernandez et al. (6).
The natural ventilation rate of a building is a function of
indoor/outdoor temperature differential and wind velocity. The
relationship between the natural ventilation rate and indoor
radon concentration is highlighted in Figure 7. On this
particular day, there was no measurable amount of precipitation.
It has been reported by Nazaroff and Nero (7) that an increase in
the natural ventilation rate will increase the depressurization
of the building which will subsequently increase the rate of
radon entry. However, in the majority of our measurements the
natural ventilation rate was not easily identified as the major
transport mechanism. It did not correlate well with the degree
of radon entry confirming the work of Nero et al. (8) and Doyle
et al. (9).
POST-MITIGATION RESULTS
The Radon Buster was installed on January 18, 1992. The
airflow rate through the unit was measured at 300 cfm (0.94 ACH).
A 10 day period from February 19 through February 28, 1992
was examined for post-mitigation effects. Figure 8 displays
basement radon concentrations for that time period.
Unmistakably, the radon concentration had taken a drastic
reduction from pre-mitigation levels. The post-mitigation levels
ranged from 1 pCi/L to 24 pCi/L with an arithmetic mean of 5.7
pCi/L. This resulted in an average radon reduction of 83%. It
is evident that the fluctuations in radon are more systematic
than that of pre-mitigation levels. The reason is that the
cyclic operation of the Radon Buster is the dominant radon entry
control variable: causing radon egression during the on mode and
-------
c
o
•*—'
CD
•^
'g.
'o
CD
ol
fO.5
12/21/91
Figure 5,
Pre-Mitigation Radon Concentration vs.
Precipitation (12/21/91)
r31.9
O
Q.
c
O
2
•*-
CD
O
C
o
O
c
o
-a
nj
cc
31.6
12/27/91
Figure 6.
Pre-Mitigation Radon Concentration vs
Barometric Pressure (12/27/91)
-------
120
-r 40-
o
Q.
I 1
c
o
75
-*—*
0>
O
C
o
o
I
CO
CE
Figure 7
70
12/23/91
Pre-Mitigation Radon Concentration vs.
Natural Ventilation Rate (12/23/91)
0
1
3456
Time [Day]
8
Figure 8. Post-Mitigation Radon Concentration
(2/19/92 - 2/28/92)
-------
radon migration while in the off mode.
Figures 9 and 10 present radon concentration as a function
of Radon Buster operation on a daily basis (the spiked dotted
line represents Radon Buster operation). It is obvious from
these figures that radon concentration is reduced when the system
is operating. Each day in the 10 day time period was examined
for minimum and maximum radon levels while the Radon Buster was
in the on mode and off mode, respectively. The minimum values
(on mode) revealed a range in radon concentration of 1 pCi/L to 4
pCi/L with an arithmetic mean of 1.8 pCi/L and a standard
deviation of 0.7 pCi/L. This indicates that if the Radon Buster
operates continuously the radon concentration would remain around
1 pCi/L. The maximum values (off mode) were in a range of 6
pCi/L to 24 pCi/L with an arithmetic mean of 12.8 pCi/L and a
standard deviation of 4.4 pCi/L. On a percentage basis, a radon
reduction of 97% is feasible with continuous operation of the
Radon Buster.
Differential pressure measurements of the soil with respect
to the basement for this same time period are displayed in Figure
11. Highlighted in this figure is the positive basement pressure
with respect to the surrounding soil (as indicated by the data
points which are less than 0). This can be credited to Radon
Buster operation as seen in Figure 12 where Radon Buster
operation is plotted versus differential pressure. The
mitigation system is blowing in 300 cfm of outdoor air creating a
positive pressure field within the basement. Typically, while
the Radon Buster is on, the outdoor and soil pressures are about
5 Pa lower than the basement pressure and the upstairs is about 4
Pa lower. The differential pressure measurement across the
basement where the low port was near the Radon Buster and the
high side was in another isolated basement room indicated a
negligible pressure difference.
A concern of this mitigation system was the delivery of cold
air to the basement through the Radon Buster. For this 10 day
period the outdoor temperature ranged from 23°F to 57°F and the
resultant basement temperature was 57°F to 67°F.
Another concern was raising the indoor relative humidity to
a level where mold and rot would begin to develop on the interior
of the building shell. This fear was dispelled since the
basement relative humidity remained stable at about 45%
throughout the entire experiment.
The heat exchanger sensible heat effectiveness was about
0.75. The latent heat effectiveness, and therefore the total
heat effectiveness is unknown since water content in the flue gas
was -not measured, but it was higher than the sensible
effectiveness since a steady flow of condensate drained from the
-------
2/19/92
Figure 9. Post-Mitigation Radon Concentration vs.
Radon Buster Operation (2/19/92)
O
Q.
C
O
"
c
0)
O
c
O
O
c
O
T5
cn
cc
Figure 10,
2/20/92
Post-Mitigation Radon Concentration vs.
Radon Buster Operation (2/20/92)
-------
6-r
1
8
10
234567
Time [Day]
Figure 11. Post-Mitigation Soil / Basement Differential
Pressure (2/19/92 - 2/28/92)
CO
D_
w
CD
CD
W
CD
m
0
Figure 12
O
o
CD
O)
u
m
c
o
-o
CO
DC
2/21/92
Post-Mitigation Soil / Basement Differential
Pressure vs. Radon Buster Operation (2/21/92)
-------
flue gas side.
The initial investment of this radon mitigation system is
about $2000. The system is estimated to cost nothing to operate,
since the cost of electricity to run the fans is offset by a
decrease in heating fuel consumption. This is illustrated in the
following example with an estimate based on the 1991/1992 heating
season in the Milwaukee, Wisconsin area.
Assumptions: Heating Degree Days (HDD) = 6600°F/days
Natural Gas Cost (Fcosl) = 0.55 $/Therm
Electricity Cost (Ecost) = 0.0677 $/KW/hr
Hours of Operation (HO) = 1500 hr/yr
Table 2 presents the Roidt's pre-mitigation natural gas
consumption in fuel usage per heating degree day (Fpre) .
TABLE 2. PRE-MITIGATION NATURAL GAS CONSUMPTION
£>
Date Fpre
11/18/91 0.1018
12/19/91 0.1033
1/23/92 0.1034
The mean value (Fprem) for this 3 month period is 0.1028
Therms/°F day. The post-mitigation natural gas consumption (F^,,)
is 0.0982 Therms/°F day. The net reduction in heating fuel
consumption (F,^,) is given by,
Fred = Fpre.m - Fpojl (1)
which is 0.0046 Therms/°F day.
The cost reduction (CR) from waste heat recovery is given
by,
CR = F^, Fcost HDD (2)
which equals $17.00.
The energy penalty comes from fan operation. On the air
side a Dayton split capacitor blower with a free air capacity of
488 cfm at 1580 rpm and 157 watts (Pair) was used, and on the flue
gas side an ITT Jabsco electric blower with a free air delivery
of 150 cfm at 2240 rpm and 78 watts (Pnue) • It is assumed that
half the power consumed by the air side blower is recovered in
the form of heat from the motor and air imparted to the fan
blades, which will warm the incoming air stream. Hence, the
-------
factor of 0.5 in the following equation describing fan operating
cost (CF).
CF = (0.5 Pair + Pnue) Ecost HO / 1000 (3)
The factor of 1000 converts KW to Watts. The cost of operating
the fans is $16.00.
Based on this experimental site it costs $16 per heating
season to operate the fans and $17 is recovered by the heat
exchanger, therefore there is no estimated net cost to operate
the Radon Buster. With a subsurface ventilation system there
exists no payback mechanism and the customer must forever pay for
fan operation.
CONCLUSIONS
The Radon Buster has proven itself an effective method of
radon mitigation with the additional benefits of energy
conservation and an increased indoor air quality. It is one of
the first radon mitigation systems to boast of these attributes.
The unit reduced radon concentration levels by a combination
of basement pressurization and dilution. Radon gas
concentrations dropped from an average pre-mitigation level of
33.5 pCi/L to an average post-mitigation level of 5.7 pCi/L
during intermittent operation. It is believed that if the system
is run continuously, the radon levels would remain at around 1
pCi/L which is well below the EPA standard of 4 pCi/L.
Fuel consumption decreased even though cold outside air was
blown into the basement. This reduction in fuel costs was enough
to compensate for the energy cost from operating the fans.
Moreover, an additional benefit of the Radon Buster which
was reported to us by the occupants of the test site was that
their home was more comfortable in cold weather now than prior to
the installation of the Radon Buster. This is attributed to the
device's ability to hold a positive indoor pressure which
prevents the infiltration of local cold air draft discomforts and
its introduction of more oxygen (due to the outdoor fresh air
intake) within the living environment.
ACKNOWLEDGEMENTS
Financial support for this research provided by the U.S.
Environmental Protection Agency through the 1991 State Indoor
Radon Grant (SIRG) No. K1995011-02 is greatly appreciated. The
-------
authors would like to thank Peter and Judy Roidt for the use of
their home, Robert C. Brauer for the fabrication of the Radon
Buster, Dr. Conrad Weiffenbach with the Wisconsin Department of
Health and Social Services for his ideas and suggestions, Paul
Tellier with the Waukesha County Department of Health for helping
us locate a test site, and Tech/Ops Landauer, Inc. of Glenwood,
Illinois as well as Gas Alert Plus of Berkeley, Illinois for
their equipment donations.
REFERENCES
1. Cavallo, A., Gadsby, K. and Reddy, T.A. Natural basement
ventilation as a radon mitigation technique. In.:
Proceedings of The 1991 International Symposium on Radon and
Radon Reduction Technology, Philadelphia, Pennsylvania, Vol.
2 Preprints, 1991.
2. Nazaroff, W.W., Boegel, M.L., Hollowell, C.D. and Roseme,
G.D. The use of mechanical ventilation with heat recovery
for controlling radon and radon-daughter concentrations in
houses. Atmospheric Environment. 15: 263, 1981.
3. Holub, R.F., Borak, T.B., Droullard, R.F. and Inkret, W.C.
Radon-222 and 222Rn progeny concentrations measured in an
energy-efficient house equipped with a heat exchanger.
Health Physics. 49: 267, 1985.
4. Brennan, T., Brodhead, W. and Dyess, T.M. Preliminary
results of HVAC system modifications to control indoor radon
concentrations. In; Proceedings of The 1991 International
Symposium on Radon and Radon Reduction Technology,
Philadelphia, Pennsylvania, Vol. 4 Preprints, 1991.
5. Montague, A. and Belanger, W.E. Effects of humidity and
rainfall on radon levels in a residential dwelling. In:
Proceedings of The 1991 International Symposium on Radon and
Radon Reduction Technology, Philadelphia, Pennsylvania, Vol.
3 Preprints, 1991.
6. Hernandez, T.L., Ring, J.W., and Sachs, H. The variation of
basement radon concentration with barometric pressure.
Health Phvsics. 46: 440, 1984.
7. Nazaroff, W.W. and Nero, A.V. Soil as a source of indoor
radon: generation, migration, and entry. In: Radon and Its
Decay products in Indoor Air. John Wiley & Sons, New York.
p. 57.
-------
8. Nero, A.V., Boegel, M.L., Hollowell, C.D., Ingersoll, J.G.
and Nazaroff, W.W. Radon concentrations and infiltration
rates measured in conventional and energy-efficient houses.
Health Physics. 45: 401, 1983.
9. Doyle, S.M., Nazaroff, W.W. and Nero, A.V. Time-averaged
indoor Rn concentrations and infiltration rates sampled in
four U.S. cities. Health Physics. 47: 579, 1984.
-------
VII-3
THE ENERGY PENALTY OF SUB-SLAB DEPRESSURIZATION RADON MITIGATION SYSTEMS
by: David. L Bohac and Timothy S. Dunsworth
Center for Energy and the Urban Environment
Minneapolis, MN 55403
Lester S. Shen
Underground Space Center
University of Minnesota
Minneapolis, MN 55455
Christopher J. Damm
Department of Mechanical Engineering
University of Minnesota
Minneapolis, MN 55455
ABSTRACT
The purpose of this paper was to quantify the energy penalty incurred by
sub-slab depressurization radon mitigation systems in a cold, heating-
dominated climate. Two sets of houses were studied. The first set of houses
consisted of houses with previously installed mitigation systems which had
been in operation for at least one year. For the second set, mitigation
systems were installed in five houses during the fall of 1990 and their fuel
usages were monitored over the subsequent heating season. The average pre-
mitigation consumption of the 11 houses was 1210 ccf. Since the average
percent increase in total gas usage for the combined results was about 6%, the
average gas usage penalty was 76 ccf per year or $38 per year, assuming a cost
of 50$ per ccf of natural gas. With an average annual electrical cost of $37
per year, the average total energy cost was $75.
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.
-------
THE ENERGY PENALTY OF SUB-SLAB DEPRESSURIZATION RADON MITIGATION SYSTEMS
INTRODUCTION
This project measures the impact of the operation of radon mitigation
systems on residential energy use. The project studied single family,
detached houses located in the Twin Cities Metropolitan area, a cold, heating-
dominated climate of 8,050 heating degree days (°F-day, base 65°F). The
mitigation systems were limited to active soil depressurization (ASD) systems.
The systems consisted of sub-slab depressurization or a combination of sub-
slab depressurization with either concrete block wall suction or crawl space
floor depressurization. ASD radon mitigation systems can affect the building
energy performance in several ways. Firstly, there is the inherent electrical
energy cost resulting from the continuous operation of the mitigation system
exhaust fan. Concomitant effects of the system include increased building air
infiltration and substructure heat loss. The augmented infiltration is a
result of house depressurization caused by air flow from the lower level of
the house into the mitigation system. The system also draws outdoor air
through the soil surrounding the substructure. During the heating season,
this results in reduced soil temperatures and increased substructure heat
loss. The mitigation system may also reduce soil moisture throughout the year
which can impact the moisture and latent heat effects in the house.
Previous investigations to quantify the associated energy penalty of
radon mitigation systems have been reported in the literature (1,2,3,4).
These studies have estimated the mitigation energy penalty by measuring the
exhaust flow and extrapolating the added heating costs. With typical
mitigation flows, this corresponds to a heating penalty of $16 to $125 per
year depending on local climate and fuel costs, with an additional fan
operation penalty of comparable size plus a cooling penalty in cases with
central air conditioning. Because the Twin Cities area has a heating-
dominated climate (8050 heating degree days versus 662 cooling degree days in
a typical year) with a low penetration of central air conditioning, the
emphasis will be on heating and fan operation penalties.
To ascertain the magnitude of the energy penalty and moisture effects,
two sets of houses were studied. The first set consisted of houses with
previously installed mitigation systems that had been in operation for at
least one year. Changes in energy use were determined through a pre/post-
mitigation fuel bill analysis employing the Princeton Scorekeeping Method
(PRISM). For the second set, five houses were selected from a preliminary
sample of houses that had performed long-term alpha track detector (ATD) tests
and had measured radon levels greater than 5 picoCuries per liter (pCi/1) in
the lower level of the house. Mitigation systems were installed in these
-------
houses during the fall of 1990. Fuel use was monitored over the subsequent
heating season as the ASD systems were alternately enabled and disabled.
METHODOLOGY
This project was divided into two separate analyses: an energy analysis
of previously mitigated houses and detailed monitoring of five houses which
were mitigated as part of the project. The analysis of the previously
mitigated houses allows the comparison of results from two methods of direct
impact measurement, as well as increases the sample size without a
corresponding increase in monitoring costs. From the fuel bill analysis of
the first set of houses, the impact of mitigation systems on whole house
natural gas use could be determined. This use includes space and water
heating use and for some houses, cooking use and clothes drying. The detailed
monitoring of the second set of houses permitted closer scrutiny of the impact
of the ASD system on electrical and natural gas space heating consumption.
Site visits were conducted on both sets of houses to measure mitigation system
air flows, entrained air flows from the living area, and house and system
moisture levels in an effort to better quantify the heat loss mechanisms
induced by the ASD systems. Results of the air flow measurements will not be
described here but are included in the project final report (5). The
following discussion describes in detail the methodology used for the
investigation of the two sets of houses.
ENERGY ANALYSIS OF THE PREVIOUSLY MITIGATED HOUSES
An analysis of previously mitigated homes was initiated to determine the
magnitude of the whole house energy penalty incurred by the mitigation
systems. A fuel bill analysis comparing the post-mitigation energy use with
the pre-mitigation use provides an indication of changes in the energy
consumption that may have been affected in part by the mitigation system.
Since whole-house natural gas use data is analyzed for these houses, any
changes in consumption due to occupancy effects, building structure, and the
heating system will act to mask the mitigation system effects.
Identification of Candidate Houses
To find houses in the Twin Cities (TC) area that had previously received
radon mitigation work, the eight contractors in the TC area who had passed the
EPA Radon Contractors Proficiency Program (RCPP) exam were contacted to get
access to likely candidates. It was found that the majority of mitigation
jobs performed in the TC area resulted from relocations and that relatively
few established homeowners opted for mitigation work. This made it difficult
to find houses with a year of pre-mitigation data and a year of post-
mitigation data from the same occupants. Nevertheless, three of the
contractors were able to provide 14 houses for analysis. Six of these houses
fit the selection criteria and were included in the study.
-------
Evaluation of Energy Use Using PRISM
Natural gas utility bill records of the pre- and post-mitigation energy
use were obtained for each house. Evaluation of the energy use of the houses
was performed using the Princeton Scorekeeping Method (PRISM) (6). Using the
PRISM computer program is the standard practice for quantifying the energy
usage in residential buildings. The program calculates a weather-normalized
annual energy consumption which is a measure of the fuel consumption of a
house under average weather conditions. The method is based on a model that
assumes that below a certain house-specific reference temperature, the fuel
consumption per day of a house varies linearly with the daily average outdoor
temperature. Above this reference temperature, the daily consumption is taken
as a constant. In reality, non-heating use often shows seasonal variation
that is imperfectly correlated with degree days, but this simplified model has
been found to work extremely well for analyzing natural gas use in heating-
dominated regions of the country. Various studies have also found that
outside temperature is generally an adequate explanatory variable without the
addition of other weather variables, such as wind, humidity, or insolation.
PRISM uses a statistical regression procedure which uses the actual
meter readings of a house and the local daily average outdoor temperatures.
The model calculates the annual fuel use corresponding to long-term average
local daily outdoor temperatures. This index is known as the Normalized
Annual Consumption (NAG). Comparing the NAC value of the post-mitigation year
with the pre-mitigation year should provide a measure of the increased energy
use resulting from the mitigation system, adjusted for differences in weather
conditions of the two years being compared. In general, PRISM has been found
to provide a reliable index of consumption and has been used to measure energy
savings in individual house retrofits as well as long-term trends in
residential energy use (6).
House Screening and Selection
Several criteria were placed on the houses that could be in the project.
The restrictions were instituted to ensure that the relevant energy data was
easily accessible, to simplify the analysis with monthly and bimonthly data,
to eliminate the effect of the change in residents on energy use, and to
provide reliable estimates of energy use. The criteria were:
1. stable occupancy - The occupants of the houses had to have resided
in the houses for a full year before the system was installed and
a full year after with no changes in household size.
2. natural gas or electrically heated homes.
3. no supplemental space heating.
4. good fuel use records.
5. no retrofits to the building structure or the heating system
during this two year period.
6. a record of pre-/post-mitigation radon levels to establish that a
radon problem had existed and that it had been satisfactorily
remedied.
-------
For the evaluation, two additional criteria affected the eligibility of
the houses in the study. A house remained in the study if the PRISM analysis
of the pre- and post-mitigation data resulted in an R2 greater than or equal
to 0.95 and a relative standard error of 5% or less (7).
Six of the 14 houses provided by the contractors fit the selection
criteria and were included in the study. Of the eight houses that were not
included, five used fuels other than natural gas. Two of the homeowners did
not respond to correspondence and one resident did not have an entire year of
data from before the mitigation system installation.
Site Visits
Each of the six houses was visited at least twice by project staff to
inspect the operation of the systems and to determine the physical
characteristics of the houses. The work performed during these visits
included measurement of the dry and wet bulb temperatures in the below-grade
portion of the building and in the exhaust system, air velocity measurements
in the mitigation system exhaust, measurement of the below-grade area
dimensions, recording of the fan specifications, and interviews with the
homeowner on pre- and post-mitigation radon levels and on any general changes
in the house dealing with energy use and perceived changes in interior
humidity levels.
MONITORING OF THE FIVE TWIN CITIES AREA HOUSES
While the first set of houses gives a measure of the general change in
whole house energy use after installation of the radon mitigation systems, a
more detailed monitoring of the mitigation systems is necessary to gain a
better understanding of the energy loss mechanisms of the systems. For this
portion of the project, five houses located in the Twin Cities area were
monitored during the 1990-1991 heating season after an active mitigation
system was installed.
House Selection
The selection process for the five houses to be mitigated and monitored
in this study was begun in March 1990. Likely candidates for the study were
obtained from a pool of households that had voluntarily performed long-term
monitoring using alpha track detectors (ATD) through participation in one of
the following programs: the Minnesota Department of Health survey and
monitoring programs sponsored by the American Lung Association of Minnesota,
Group Health, and CEUE. Those participants that were located in the Twin
Cities area and had ATD results greater than 5 pCi/1 in either the first floor
or lowest living level were informed by their sponsoring agency that they
qualified for inclusion in the EPA Innovative Grant study. Interested parties
were instructed to contact the University of Minnesota.
-------
House Screenino
From the selection process, 21 houses were obtained for the initial
screening. A survey of these houses showed that an ATD test had typically
been performed for at least three months on each of the houses, with many
houses performing a one year ATD. All the houses used natural gas as their
primary heat source. A fuel bill analysis using PRISM was performed on each
house to determine the fitness of the model in analyzing the 1989-90 pre-
mitigation heating season. Based on these results, site visits and homeowner
interviews were performed on 11 of the houses. These visits were mostly
performed during July, with one house done in early September. Two to eight
day E-Perm tests were performed under closed house conditions, when possible,
to verify the ATD results.
Radon Diagnostics and Mitigation
Radon mitigation contractors performed diagnostics on eight of the
houses. Four of these houses were selected for inclusion in the study and
contracts were signed with the participants. Radon mitigation using active
soil depressurization was performed on these houses. Work on three of the
houses was completed during July and August and the fourth house was completed
in the beginning of January. A fifth house was located using contacts from
one of the contractors and was added to the study in October. This house also
received active soil depressurization.
Each of the homeowners was concerned about the high level of radon in
their homes and had an interest in installing a radon mitigation system prior
to volunteering for participation in the study. Two of the homeowners had, in
fact, previously received diagnostic testing and bids from contractors to
perform mitigation services. Under the services of this project, the
mitigation systems were installed at no charge to the homeowner in exchange
for their cooperation in the study.
House Monitoring
Immediately after the mitigation systems were installed, the contractors
placed continuous radon monitors (CRM's) in the houses to document the
successful operation of the systems. After this post-installation radon
monitoring was completed, the systems were turned off until the house
monitoring began. Consequently, the mitigation systems were not in operation
during the fall of 1990. Beginning the first week of January 1991, the
mitigation systems were activated for two weeks. Mitigation fan electric
usage and the whole house gas consumption were recorded for this period. The
natural gas usages were obtained from visual reading of the house gas meters
while the electric use of the mitigation system fans were obtained from
electric meters connected to the fans. After this period, the systems were
deactivated for two weeks. Two week alternating mode experiments' were
performed in this manner for the remainder of the heating system (end of
June). The period of two week duration was chosen to provide an adequate
sample of experimental data and to ensure that the heat loss and moisture
transfer characteristics had reached stability. Discrete time relative
-------
humidity measurements using a sling psychrometer were made in the basement and
living space of each house during the bi-weekly visits.
Radon monitoring was begun on December 3, 1990 with three E-PERM
monitors placed in each house: one in the basement, one in the living area,
and one in the attic. These were collected and replaced every two weeks,
corresponding to the operating mode of the mitigation system. An EPA-
certified radon measurement contractor provided the E-PERM monitors and the E-
PERMs were measured at the contractor's office. The monitors were transported
between the sites by project staff and were placed in the houses by staff
according to approved protocols. Blank and replicate measurements were
performed following the guidelines and protocols established by the Minnesota
QA/QC Plan approved by EPA.
RESULTS
FUEL BILL ANALYSIS OF THE PREVIOUSLY MITIGATED HOUSES
The results of the PRISM analysis of the six previously mitigated houses
are shown in Table 1. All of the houses had mitigation systems installed
during 1989 with the exception of house 11 which had its system installed
during the fall of 1988. Four of the houses have full basements while two
houses are a combination full basement with an adjoining crawl space. The
average floor area for the four basement houses is 1219 ft2. For the
combination basement/crawl space foundations, the average basement floor area
is 600 ft2 and the average crawl space floor area is 335 ft2. Sub-slab
depressurization systems were installed for the basement areas and in the
crawl spaces, floor depressurization systems were installed. The walls are of
concrete block construction in all six houses. Air sealing of the basements
was also performed in these houses. This included sealing of large cracks in
the foundation walls and floors and floor areas open to the earth for water
lines.
The pre-mitigation data period spanned from 10 to 13 months and started
as early as September 1987 and ended as late as November 1989. The specific
dates depended on when the system was installed, occupancy variations, and
robustness of the PRISM analysis. The post-mitigation period spanned from 12
to 13 months, typically incorporating the 1989/90 heating season but also
including a portion of 1990/91 heating season if circumstances warranted.
For each house, the NAC, the coefficient of variation (CV), and R2 are
provided for both the pre- and post-mitigation years with the values shown in
parentheses being the standard errors of the NAC. The standard error can also
be taken to be the standard deviation of an estimated statistic, for example,
a regression parameter such as the NAC value. For the purposes of this report,
a 68% confidence interval about the model estimate is defined as ±1 standard
deviation about the estimate and a 95% confidence interval is defined by ±2
standard deviations. A 95% confidence interval about the estimate means that
with a 95% certainty, the true long-term (or "population") value lies within
-------
the defined interval. 95% confidence intervals will be reported when
confidence intervals are included with the results.
The energy penalty incurred by the mitigation system is computed by the
change in the total NAG (ANAC). Since only natural gas use was studied in
this fuel bill analysis, results are given in ccf of natural gas (which is
very nearly equal to 100,000 Btu or one therm). The ANAC is a gauge of the
energy penalty of the mitigation system on the whole house natural gas use.
Again, the standard deviations for each value are given in parentheses and the
value given below each standard deviation is the p-value which is a measure of
statistical significance. For instance, a p-value of 0.05 means that there is
a 95% confidence level that the calculated change in NAG is different from
zero (a two-tailed test) and does not simply fall within the error bounds of
the model estimates.
Examination of the p-values for each ANAC shows that the calculated ANAC
for houses 12 and 16 are not statistically significant and that statistically
no impact from the ASD system could be observed on their natural gas use. The
ANAC of house 14 has only a 90% statistical significance while the ANAC's of
the remaining houses have at least a 95% significance. Examining the energy
use impact on this set of houses as a group, the average pre-mitigation NAC
(whole house use, including heating and base load use), the normalized whole
house energy use, for the six houses was 1206 ccf while the average post-
mitigation NAC was 1291 ccf. The average increase in NAC was 85 ccf, for an
overall sample increase of 7.0%. The highest increase was 24.6% and one house
had a slight decrease in usage of 2.5%. A paired difference t-test on this
group of houses produced a p-value of about 0.03.
MONITORING OF THE FIVE TWIN CITIES AREA HOUSES
House Screening
Table 2 shows the results of the pre-mitigation radon tests, PRISM
analysis, and house characteristics of the five houses selected for detailed
monitoring. The long-term ATD results show that elevated radon levels
(exceeding 5 pCi/1) were measured in all five houses. With the exception of
site 2 which shows a low radon level, the E-PERM tests for the most part
confirm the ATD results. The short-term E-PERM tests were performed during
the summer and even though an attempt was made to maintain closed house
conditions, greater confidence is placed in the ATD results.
The average pre-mitigation NAC for the five houses was 1251 ccf. Three
of the houses are heated by forced air systems (houses 2, 3, and 5) while the
remaining two houses have hydronic systems (houses 1 and 4). All five houses
have basement foundations, with sites 2 and 4 also having crawl spaces under
adjoining additions. The average floor area of the basement-only houses is
837 ft2 while the combination foundation houses has an average basement floor
area of 835 ft2 and an average crawl space floor area of 228 ft2. Site 4 has
a ppured concrete wall foundation while site 1 has both a stone and concrete
block foundation. The three remaining houses have concrete block foundation
walls. Sites 2, 4, and 5 are all one-story houses with house 2 having a walk
-------
in basement and a tuck-under garage. Site 1 is a two story house and house 3
is a story and a half construction. Each house had a sub-slab
depressurization system installed in the basement area. Site 5 also had a
wall suction system installed while crawl space floor depressurization was
also used in sites 2 and 4.
Radon Levels
To gauge system performance of the ASD systems, E-PERM monitors were
placed in the basement, first floor living space, and attic during each two-
week period of the alternating moAe experiment. Figure 1 shows the average
radon levels for both the system-on and system-off modes for each of the
houses. The systems were all working effectively with the apparent exception
of the system at site 2. The other sites show an average radon reduction of
93% on all floors, which is highly significant.
At site 2, the radon levels were reduced in the first floor when the
system was operating. However, there were numerous elevated levels that were
measured in the basement. A continuous monitor (CRM), that had been cross-
calibrated with several other devices, was placed in the basement adjacent to
the E-PERM. For the two week period when the system was operating, the CRM
recorded radon levels of less than 0.5 pCi/1 when the E-PERM reading was 25.8
pCi/1. Diagnostics of the house with the system running showed a substantial
low pressure field was being generated under the slab. It was concluded that
the system was operating properly. At present, no explanation can be
attributed for the spuriously high E-PERM readings. Two E-PERM measurements
were performed in the same basement location after May 14 (and, therefore,
after the end of the heating season) and both readings were below 1 pCi/1.
While Figure 1 illustrates how well the systems are operating, it is
also interesting to note the radon levels in the first floor living space when
the systems are off. Sites 3 and 5 show a generally gradual attenuation in
radon levels going from the basement to the first floor and to the attic
(except for the attic in house 5) while low levels are observed in both the
first floor and attic in sites 1 and 4. On average, for sites 3 and 5, the
first floor radon levels with the ASD off is 73% of the basement while it is
14% for sites 1 and 4. This pattern is a result of the heating systems of
each house. Sites 3 and 5 have forced air systems while sites 1 and 4 have
hydronic systems. Duct leakage and air mixing from the forced air system have
resulted in the increased transfer of radon gas from the lowest level to the
living space. Indeed, for sites 1 and 4, if the basement spaces are not to be
used as habitable space, the need for installing a mitigation system could be
questioned.
The average attic radon levels for sites 1, 2, and 3 suggest
communication between the living space and the attic. Sites 2 and 3 show
measurable attic radon levels with the mitigation system operating while in
site 1, attic and living space levels are comparable, at about 0.5 pCi/1 when
the system is off. A large attic bypass was found in site 1 while appreciable
ice dams above the kneewall areas of site 3 (even with soffit and roof
venting) also indicate the need for attic bypass sealing. It was learned from
-------
conversations with the occupants of site 2 that a previous history of ice dam
problems resulted in the installation of roof turbines by the homeowners the
previous year. All of these observations and the measured attic radon levels
are indicative of substantial living space to attic air flows.
Natural Gas Usage
A method for converting each energy use value to a residual was used to
compare the monitored energy use with the pre-test PRISM model. The residual
was computed as the difference between the observed use per day in each period
and the use that would be expected based on the pre-test PRISM model and the
test period weather conditions (which can be calculated manually or by using
the PREDICT software developed by CEUE for this type of situation). This
method also implicitly assumes that the pre-test PRISM model is not grossly
different from the (unknown) test period models. To the extent that the two
operating modes have roughly similar distributions of weather, this assumption
can be violated somewhat without completely undermining the validity of the
resulting comparisons. This method also has the advantage of not losing the
information of periods with no heating use (as the use per degree day method
does by excluding such points), nor is this approach liable to exaggerating
any deviations in data with small numbers of degree days (or in the worst
case, a fractional number of degree days per day).
After rescaling the data by any of these methods, a two sample t-test
was used to compare operating modes at each site. Single sample t-tests were
used to compare each mode to the pre-mitigation PRISM model (especially the
system-off mode which provides a measure of how consistent the test period
energy use was with the pre-test use). The two-sample t-tests for all the
sites can be combined by using a log-likelihood method and comparing the
resultant test statistic to a Chi squared distribution. This helps to
determine whether several statistically non-significant but fairly consistent
results add up to a statistically significant overall result. For the
purposes of this discussion, statistical significance is defined by a p-value
less than 0.05.
A summary of the results of the residuals approach are displayed in
Table 3. Tests of the off-mode versus the pre-test models show no apparent
pattern of altered usage since two cases rose non-significantly, two others
decreased non-significantly, and the pooled test was also non-significant.
While this is not a rigorous test for changes in energy use patterns (since
changes in several model parameters could produce residuals in colder weather
that balance opposite residuals in milder weather), it does increase
confidence in the subsequent test results at least somewhat. Tests of the on-
mode versus the pre-test model show one significant increase in use (site 5, p
= 0.017), two non-significant increases, and one non-significant decrease.
The pooled test provides a test statistic at the p = 0.035 level. Two-sample
comparisons between modes have -roughly similar results, with only one
significant individual test (site 5, p = 0.020) and a marginally non-
significant pooled result at the p = 0.081 level.
-------
The average residuals from the analyses above can be multiplied by 365
days and then added to the pre-mitigation NAG to yield estimates of the
approximate annual energy use with the ASD system on and off. A change in
annual energy use can then be computed. As shown in Table 4, this calculation
yields average values of 1216 ccf with the systems off and 1278 with the
systems on, for an average gas use penalty of 62 ccf (or 5.3 ± 10.0 percent)
with a p-value of 0.33. At current gas prices, this translates to an economic
penalty of $31 per year.
An average energy penalty for the two sets of houses can be obtained by
combining the results of the two subsets. The analysis of variance (ANOVA) of
the two subsets of houses revealed that the two sets did not differ
significantly. Consequently, a paired t-test of all ten houses could be
performed. The average pre-mitigation energy use for the ten houses was 1210
ccf while the average post-mitigation use was 1286 ccf. The paired t-test
showed a statistically significant overall energy penalty of 76 ccf or a
penalty of about $38 per year. The standard error of the energy penalty
result was 81 ccf. The p-value for the paired t-test was in the range
0.01
-------
The Chance in Humidity Levels Due to the Mitigation System
In addition to radon gas, the soil around the building is also a large
source of moisture. Typically, the air trapped in the pores of the soil is
saturated with moisture in the form of water vapor. The same mechanisms which
work to draw radon gas into the house also bring in this water vapor. By
reducing radon in the house, the mitigation system also acts to reduce this
important source of moisture in the building. During the heating season, the
mitigation system can act to relieve high moisture levels in a house by
reducing the entry of moist soil air and also by drawing in cold dry outside
air from induced air infiltration. Concurrently, the mitigation system will
have an impact on the latent heating loads of the building.
In the absence of continuous house moisture monitoring, house moisture
levels were measured on a discrete time basis using a sling psychrometer
during the biweekly visits. Average values are given in Table 6. For every
house, the average relative humidity was lower in the basements and living
areas when the mitigation system was in operation. Relative humidities ranged
from about 25% to 45% and were typically higher in the basement than the
living area in all houses but site 2 where the reverse trend was true.
Because relative humidity is a function of both air temperature and
moisture content, a more useful measure of the impact of the mitigation system
on house moisture levels is the humidity ratio (in terms of pound mass of
moisture per pound mass of dry air). The humidity ratio is not temperature-
dependent and provides an absolute measure of the moisture content of the air.
Table 6 also shows the average humidity ratios in each of the houses with the
system on and off. For all five sites, the moisture levels in the basement
decreased an average of 18% when the mitigation system was operating. In the
living space, the average humidity level went down in all the houses but site
4 during system operation. For sites 1 through 4 the humidity ratio was
higher in the living space than the basement when the system was both on and
off. This was also true for site 5 when the system was operating. However,
when the system was off, the basement humidity ratio was slightly higher than
the living space. To test the significance of the variation in humidity
levels, a multivariable regression was performed for the humidity ratio on the
measurement date, operation mode, floor, floor/mode interaction, and site.
The results show that the variation by date, operation mode, floor, and
between some houses are all significant.
The data of Table 6 suggest that for sites 2 through 4, human activities
are probably an equal or greater source of house moisture in the living space
as the soil. The results of sites 1 and 5 indicate that the soil is a
significant source of house moisture and that for both houses, when the system
is operating, basement moisture levels reduce dramatically.
CONCLUDING REMARKS AND RECOMMENDATIONS
The measurement of the impact of the radon mitigation systems on energy
use found an average whole house gas use penalty from the combined results of
-------
the two sets of houses of 76 ccf. This increase was 6% greater than the pre-
mitigation average use level (1210 ccf) and at a cost of 50C per ccf,
represents an average energy penalty of $38 per year. The whole house gas use
includes space and water heating use and for some houses, cooking use and
clothes drying (but any changes are most likely due to the added space heating
load). The test statistic for the paired t test of all ten houses was in the
range of 0.01
-------
REFERENCES
1. Harrje, D.T., Hubbard, L.M., Gadsby, K.J., Bolker, B. and Bohac, D.L.
The Effect of Radon Mitigation Systems on Ventilation in Buildings.
ASHRAE Transactions. 95: 107, 1989.
2. Clarkin, M., Brennan, T., and Osborne, M.C. Energy penalties associated
with the use of a sub-slab depressurization system. In; The 1990
International Symposium on Radon and Radon Reduction Technologies. U.S.
Environmental Protection Agency AEERL and ORD, Atlanta, Georgia, 1990.
3. Turk, B.H., Harrison, J. and Sextro, R.G. Performance of radon control
systems. Energy and Buildings. 17: 157, 1991.
4. Henschel, D.B. Cost analysis of soil depressurization techniques for
indoor air reduction. Indoor Air. 1: 337, 1991.
5. Bohac, D.L., Shen, L.S., Dunsworth, T.S., and Damm, C.J. Radon
Mitigation Cost Penalty Research Project. Underground Space Center,
University of Minnesota, Minneapolis, MN, 1991. 68 pp.
6. Fels, M.F. PRISM: An Introduction. Energy and Buildings. 9: 5, 1986.
7. Dunsworth, T.S. and Hewett, M. Data quality considerations in using the
'PRISM' program. In; Proceedings of the Second National Conference on
Energy Conservation Program Evaluation. U.S. Department of Energy,
Chicago, Illinois, 1985.
-------
11
10
:S 9
T 8
c **
0
1
m
o 6
o
c 5
o
•a
0)
a? 3
fli
< 2
1
85% -Percent
RSI Reduction
1
- i
$$
N$:
$S
>oo< ^Vs >^V< *
1.b f a 2. b
^
S^
^
f
$ft
a
90°y
1
1
I
1
88
3.b
'o
1
I
$£
f
i
I
sVs
a
(
i
345
^
i
I
8$
*.t
<
^
'^ r^
) f a
35°/
W//////M
1
s
1
I
5. b
r
I
1
i
i
SV-,
f
'^ V ^
a
b: basement, f: firstfloor a: attic * _ jnvaljd measurements
^ system on R^l system off
Figure 1. Average Radon Levels for System On and System Off Modes of the Monitored
Houses
-------
TABLE 1. ENERGY USE RESULTS FOR THE PREVIOUSLY MITIGATED HOUSES
House
ID
11
12
13
14
15
16
PRB
Time Period
9/17/87-9/20/88
10/26/88-10/26/89
2/22/88-2/21/89
11/2/88-9/29/89
12/8/88-11/2/89
3/21/88-4/20/89
*of
Period*
9
12
8
11
11
7
MAC
(ccffr)
686
(47)
1255
(24)
1097
(22)
2412
(56)
889
(26)
896
(24)
R«R.
CVf%)
0.950
6.9
0.994
1.9
0.996
2.0
0.991
2.3
0.982
2.9
0.994
2.7
POST
Time Period
4/18/89-5/17/90
12/27/89-12/27/90
8/22/89- 8A22/90
11/28/89-12/3/90
1/9/90-1/10/91
3/21/90-3/21/91
*of
Periods
7
12
11
12
11
11
NAC
<)
24.6
(9.0)
[0.021
2.9
(3.1)
[0.4]
7.8
(3-4)
[0.05]
6.0
(3.7)
[0.11
10.7
(4.1)
[0.02]
-2.5
(2-8)
[0.4]
8.3
9.2
0.08
-------
TABLE 2. RADON MEASUREMENTS, PRISM RESULTS, AND BUILDING
FOUNDATION CHARACTERISTICS OF THE MONITORED HOUSES
House
ID
1
2
3
4
5
ATD
Result
(pd/L)
5.7
12.5
5.0
5.7
10.1
E-Perm
Result
(pCi/L)
4.4
5.6
13
7.1
NA
#of
Periods
12
12
12
12
12
NAC
(cctfyr)
1755
(17)
1459
(25)
728
(16)
1275
(27)
1036
(28)
CVof
NAC
(*)
1.0
1.7
22
2.1
2.7
R*R
0.998
0.995
0.994
0.993
0.986
Below Grade
Tjrpe
full
basement
base/CS
full
basement
base/CS
full
basement
Constr
block
block
block
concrete
block
Space
Heat
System
hydronic
forced
air
forced
air
hydronic
forced
air
base/CS: partial basement plus crawlspace
-------
TABLE 3. SUMMARY RESULTS OF THE RESIDUALS APPROACH COMPARING THE
MONITORED DATA WITH THE PRISM MODEL
House
ID
1
3
4
5
Pooled
average=
sd=
t-stat=
p-value=
average=
sd=
t-stat=
p-value=
average=
sd=
t-stat=
p-value=
average=
sd=
t-stat=
p-value=
chi-sq=
p-value=
On-
Model
-0.042
0.596
-0.141
0.895
0.081
0.213
0.659
0.557
0.397
0.808
0.983
0.381
0.434
0.224
3.875
0.017
16.4877
0.035
Off-
Model
-0.331
0.664
-0.997
0.375
0.123
0.236
1.042
0.356
0.491
0.595
1.650
0.173
-0.089
0.244
-0.730
0.506
9.3424
0.314
On-
Off
0.289
0.446
0.648
0.541
-0.042
0.173
-0.242
0.818
-0.094
0.502
-0.187
0.857
0.523
0.167
3.158
0.020
14.0359
0.081
TABLE 4. ENERGY USE RESULTS FROM THE ALTERNATING MODE EXPERIMENT
COMPARED TO THE PRISM MODEL
House
ID
1
3
4
5
average=
sd=
t-stat=
p-value=
NAC (ccf/year)
Pre
Model
1755
728
1275
1036
1199
433
System
Off
1634
773
1454
1004
1216
397
System
On
1740
758
1420
1195
1278
413
Differences (ccf/year)
On-
Model
-15
30
145
159
79
86
Off-
Model
-121
45
179
-33
18
127
On-
Off
105
-15
-34
191
62
106
1.164
0.329
-------
TABLES. MITIGATION SYSTEM EXHAUST FAN ELECTRICAL USAGE, FLOW RATES,
AND STACK PRESSURES
House
ID
1
2
3
4
5
average=
Fan Power (Watts)
Rated
90
90
90
100
90
92
Measured
59
72
72
94
70
73
Mew/Rated
(%)
66
80
80
94
78
79
Mitigation Fan Flow (cftn)
Rated:
free air
270
270
270
360
270
288
Measured
10
52
15
46
40
33
Meas/Rated
<*)
4
19
6
13
15
11
Mit. Pipe
Pressure
("water)
1.6
1.2
1.5
1.4
1.1
1.4
TABLE 6. HUMIDITY RATIO AND RELATIVE HUMIDITY MEASUREMENTS IN THE
MONITORED HOUSES
House
ID
1
2
3
4
5
Averages
Location
basement
1st floor
basement
1st floor
basement
1st floor
basement
1st floor
basement
1st floor
basement
1st floor
Average Humidity Ratio
(lb water/lb air)
System
On
0.294
0.463
0.505
0.580
0.613
0.752
0.538
0.747
0.448
0.482
0.480
0.605
System
Off
0.553
0.642
0.574
0.653
0.687
0.797
0.572
0.626
0.551
0.542
0.587
0.652
Difference
Off-On
0.259
0.179
0.069
0.073
0.074
0.045
0.034
-0.121
0.103
0.060
0.108
0.047
Average Relative Humidity
(%>
System
On
38.5
29.4
24.8
26.5
43.8
41.6
37.9
31.5
30.9
30.5
35.2
31.2
System
Off
43.6
38.7
35.1
38.7
45.3
45.2
44.9
34.0
36.2
33.3
41.0
38.0
Difference
Off-Oa
5.1
9.2
10.2
12.2
L5
3.6
7.0
2.5
5.3
2.8
5.8
6.8
-------
DESIGN OF RADON REDUCTION TECHNIQUES
FOR CRAWL-SPACE HOUSES:
ASSESSMENT OF THE EXISTING DATA BASE
by: D. Bruce Henschel
Air and Energy Engineering Research Laboratory
U. S. Environmental Protection Agency
Research Triangle Park, NC 27711
ABSTRACT
An evaluation has been completed of the alternative technolo-
gies for reducing radon concentrations in the living areas of
houses having crawl-space foundations. Sub-membrane depressuriza-
tion (SMD) is consistently the most effective technique, often
providing radon reductions of 80-98% in the living area. Forced
crawl-space depressurization is second, giving 70-96%. Crawl-space
depressurization is less well demonstrated than is SMD, and
performance will vary with crawl-space tightness and weather, but
it will be a primary option when high radon reductions are needed
in inaccessible crawl spaces. Natural crawl-space ventilation and
forced crawl-space pressurization each provides roughly 50% reduc-
tion or lass in the living area. Crawl-space sealing and barriers
(as stand-alone methods) usually give little or no reduction.
The paper also provides a detailed review — based upon
available data — of the effects of individual house design, house
operation, system design, system operation, and geological varia-
bles, on the performance of SMD and of crawl-space depressurization
systems. From this review, the major data gaps are: 1) the
specific conditions under which crawl-space depressurization might
be preferred over SMD, including the relative operating costs of
the two approaches; 2) the optimal method for distributing suction
beneath a SMD membrane (i.e., individual suction pipes vs. sub-
membrane perforated piping) under different conditions; 3) the
optimal degree of SMD membrane sealing required under different
conditions; 4) the conditions under which it may be possible to
leave a portion of the crawl-space floor uncovered by the SMD
membrane, if the crawl space is partially inaccessible; and 5) the
durability of SMD, crawl-space depressurization, and natural crawl-
space ventilation systems.
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
Based upon data presented in the literature and upon contacts
with commercial radon mitigators, experience with reducing indoor
radon concentrations in crawl-space houses is much more limited
than is experience with mitigating basement and slab-on-grade
houses. This relatively limited experience with crawl-space houses
probably results from the facts that: 1) crawl-space houses
represent a relatively limited percentage of the total housing
stock nationally (14% of the total new housing starts in the U. S.
between 1976 and 1983, according to the National Association of
Home Builders); and 2) houses with vented crawl spaces may be less
prone to having elevated radon levels in the living area than are
basements and slabs on grade, because of less direct contact with
the soil. Another contributing explanation seems to be that many
crawl-space houses with elevated radon have an adjoining basement
(or slab-on-grade) wing which is commonly the major radon entry
route into the house. The mitigation systems in such combined-
substructure houses often focus on sub-slab depressurization (SSD)
or sump/drain-tile depressurization (sump/DTD) in the basement; the
crawl space might need to be treated only by isolation from the
livable area, and/or by natural or forced ventilation. Sometimes
the crawl space can be ignored altogether.
Extensive R&D and commercial experience exists in the
treatment of basement and slab-on-grade houses, especially using
active soil depressurization (ASD) techniques such as SSD and DTD.
The U. S. Environmental Protection Agency (EPA) is able to provide
reasonably definitive guidance regarding the design and installa-
tion of SSD and DTD techniques for houses with slabs. But due to
the more limited experience with crawl-space houses, EPA is in a
weaker position to provide guidance for this substructure type.
Although crawl-space houses are only a relatively limited
fraction of the total U. S. housing stock, they represent a signif-
icant fraction of the housing stock in some parts of the country,
especially the Northwest and the Southeast. And although crawl-
space houses may be assumed to be less radon-prone than the other
substructure types, many crawl-space houses have been encountered
having radon levels of 15-20 pCi/L (550-750 Bq/m3) and higher in the
living area. Such levels exist in houses where the crawl space is
a significant (or the sole) source of the radon, and where the
problem thus cannot be addressed by SSD or DTD in an adjoining
basement.
The data review and analysis presented in this paper identify
the areas where further information is needed in order for EPA to
provide adequate guidance in the design and installation of radon
mitigation systems for crawl-space houses.
The discussion here assumes that the crawl space has a floor
consisting of bare earth, or of gravel or a plastic vapor barrier
on top of bare earth. Crawl spaces having a concrete slab (or an
-------
unfinished concrete "wash" floor) would commonly be treated using
SSD, similar to a basement or slab on grade, and are not considered
directly in this paper.
ALTERNATIVE MITIGATION METHODS FOR CRAWL-SPACE HOUSES
SUB-MEMBRANE DEPRESSURIZATION (SMD)
Active SMD — the ASD approach applicable to crawl spaces —
involves installation of a membrane (usually polyethylene sheeting)
over the crawl-space floor, and drawing suction beneath this
membrane using a fan. This is analogous to installing a plastic
"slab" over the floor, and drawing suction beneath this slab using
either individual suction pipes (analogous to SSD) or a segment of
perforated piping beneath the membrane (analogous to DTD). SMD
probably works by the same mechanisms which apply for SSD and DTD.
One major mechanism is depressurization of the sub-membrane region
relative to the crawl space in order to prevent soil gas flow up
into the crawl space (and from there up into the living area) .
Because the membrane will leak if not sealed completely, other
mechanisms can also come into play (ventilation or depressurization
of the crawl space by the SMD fan).
SMD has consistently been found to be the most effective
approach for reducing radon in crawl-space buildings, commonly
providing radon reductions of 80 to 98% in the living area when the
crawl space is the sole source of the indoor radon (References 1,
2, and 3). Commercial radon mitigators working in regions having
a significant number of crawl-space houses have reported similar
success with SMD (References 4, 5, and 6). SMD would be the
primary mitigation method considered in any house where radon
reductions greater than about 50% are needed (thus ruling out
natural ventilation of the crawl space as an option), where the
crawl space is a major (or the sole) radon source, and where the
crawl space is accessible. It may still be the technique of choice
even where reductions below 50% are needed. Where the crawl space
is adjoined by a basement wing, the SMD system will usually need to
be supplemented (or replaced) by a SSD or DTD system in the
basement (References 2, 4, 5, 6, 7, 8, 9, and 10). A national
survey of mitigators indicated that SMD is the technique most
commonly utilized in crawl-space buildings by about one-third of
the mitigators surveyed (Reference 11) .
The installation cost for SMD is typically in the range of
$1,000-$2,500 (Reference 12). This is the highest installation
cost of any of the crawl-space treatment options, with the possible
exception of crawl-space sealing. However, relative to crawl-space
depressurization (where the entire crawl space is depressurized),
the membrane reduces the amount of air drawn out of the crawl
space, and hence the amount of conditioned air drawn out of the
living area overhead. The annual operating costs of SMD (for fan
electricity and for the heating/cooling penalty associated with
-------
treated house air exhausted by the system) will range between $45
and $200, depending upon the particular conditions (Reference 13) .
At "baseline" conditions — 50 cfm (25 L/sec) exhaust, 25% treated
air in exhaust, 6,550 F°-days (3,640 C°-days) , 90 W fan power
consumption — the annual cost would be $89. As discussed later,
this is about half the baseline annual operating cost that might be
expected with crawl-space depressurization, the major competitor of
SMD for mitigating high-radon crawl-space houses.
CRAWL-SPACE DEPRESSURIZATION
With active crawl-space depressurization, an exhaust fan is
mounted to blow crawl-space air outdoors. Installation of the fan
may have to be accompanied by some sealing of the crawl space, to
reduce the crawl-space leakage area and to thus aid in depressuriz-
ing the crawl space. (As a minimum, foundation vents should be
closed.) The objective is to depressurize the entire crawl space
relative to the living area, rather than depressurizing just the
region beneath a membrane over the floor. Crawl-space depressur-
ization works primarily by reversing the natural flow patterns
between the crawl space and the living area (i.e., so that radon-
free living-area air flows down into the crawl space, rather than
radon-containing crawl-space air flowing up into the living area).
Crawl-space depressurization might also be expected to
function in part by increasing the ventilation rate of the crawl
space, diluting the crawl-space radon with air from outdoors and
from the living area. However, the dilution mechanism may not be
important in practice, if in fact the crawl space is being
effectively depressurized. Depressurization of the crawl space
increases the flow of soil gas up into the crawl space, largely or
completely offsetting any dilution effects resulting from the
increased flow of outdoor and living-area air into the crawl space.
As a result, radon concentrations in the crawl space have generally
been found to remain about the same, or to increase, when the crawl
space is depressurized (References 1, 2, and 3), demonstrating that
the dilution mechanism is providing no net benefit.
Crawl-space depressurization has consistently proven second
only to SMD in effectiveness in reducing living-area radon levels.
Living-area reductions ranging from 70 to 96% have been reported
from tests in nine crawl-space buildings where crawl-space exhaust
flows ranged between 50 and 130 cfm (25 and 60 L/sec) (References
1, 2, 3, and 14). Performance will depend on the tightness of the
crawl space and the size of the exhaust fan; i.e., on the degree of
crawl-space depressurization that can be achieved relative to the
living area. Percentage reductions may also depend upon the
weather and upon homeowner activities, since the low degrees of
crawl-space depressurization often achieved relative to the living
area would seem subject to being overwhelmed.
Where pressures have been reported, in buildings without
forced-air ducting in the crawl space, crawl-space depressuriza-
-------
tion has depressurized the crawl space relative to the living area.
by 0 to 0.004 in. WG (0 to I Pa) in two structures (References 3,
13, and 14) . Higher crawl-space depressurizations were observed in
one house having an unusually tight floor and foundation wall
(Reference 14). However, in more typical houses with relatively
leaky floors and foundation walls, it would be expected that crawl-
space depressurizations relative to the living area will commonly
be a couple thousandths of an inch WG (a fraction of a Pa) , at
best, when exhaust flows are in the range of 50 to 130 cfm (25 to
60 L/sec).
Because crawl-space depressurization may increase radon
concentrations in the crawl space, it is important that the system
not be overwhelmed often by weather- or appliance-induced
depressurizations within the living area. A near-zero crawl-space
depressurization relative to the living area increases the threat
that the system will sometimes be overwhelmed.
Maintenance of depressurizations within the crawl space would
be expected to be more difficult when the crawl space has a high
effective leakage area. Forced-air furnace ducting in the crawl
space would be expected to increase the leakage area between the
crawl space and the living area; in addition, low-pressure return
ducting in the crawl space could draw the potentially increased
crawl-space radon levels into the circulating house air. Limited
experience in two houses having return ducting in the crawl space
has shown living-area radon reductions of 70 to 74% (Reference 1) ,
at the low end of the range for crawl-space depressurization; in
neither house was the living-area radon concentration reduced below
EPA's guideline of 4 pCi/L (150 Bq/m3) by this technique. But in
this same study, depressurization of a crawl space containing only
high-pressure forced-air supply ducts achieved living-area radon
reductions greater than 90%. There is also evidence that crawl—
space depressurization might contribute to back-drafting of combus-
tion appliances in the crawl space, where crawl-space depressuriza-
tions above 0.01 to 0.02 in. WG (2.5 to 5 Pa) are achieved.
The national survey of mitigators (Reference 11) indicated
that natural crawl-space ventilation or forced ventilation
(including crawl-space pressurization and depressurization systems)
is most commonly used by 28% of the mitigators when treating crawl-
space houses. Thus, crawl-space depressurization systems are
preferred by fewer than 28%. It is suspected that many mitigators
using an exhaust fan in crawl spaces are doing so to increase
ventilation/dilution, rather than to depressurize the crawl space.
The installation cost of a crawl-space depressurization system
will generally be much lower than that of a SMD system, perhaps
less than $500 if all that is required is the installation of a fan
and perhaps some limited piping. The installation cost would
increase if additional effort is required to seal the crawl space,
to reduce the leakage area.
-------
Since some fraction of the air exhausted by the crawl-space
depressurization fan will be conditioned air drawn down from the
living area — and since the amount of air exhausted will tend to
be greater than that from a SMD system — the heating/cooling
penalty (and hence the operating cost) associated with crawl-space
depressurization will be greater than that for SMD. The annual
operating cost of a crawl-space depressurization system (for fan
electricity plus the heating/cooling penalty) would be in the range
of $60-$280, depending upon conditions (Reference 13). At
"baseline" conditions — 100 cfm (50 L/sec) exhaust, 50% of the
exhaust is treated air, 6,550 F°-days (3,640 C°-days), 90 W fan
power consumption — the operating cost is $167. This is about
twice the baseline annual cost of the SMD system, cited previously,
as a result of the increased heating/cooling penalty.
In summary, crawl-space depressurization will be most applic-
able where: 1) the crawl space is inaccessible, ruling out SMD as
an option; 2) the crawl space is relatively tight and well isolated
from the living area to begin with, reducing the need to seal and
isolate the crawl space; 3) radon reductions of greater than 50%
are required in the living area, ruling out natural ventilation as
an option; and 4) there is no combustion appliance in the crawl
space that might back-draft. But even where conditions 2) through
4) above are favorable for crawl-space depressurization, SMD should
still be considered, due to its potential for greater and more
consistent radon reductions at reduced operating cost.
CRAWL-SPACE PRESSURIZATION
In crawl-space pressurization systems, a fan is mounted to
blow outdoor air (or air from the living area) into the crawl
space. Some sealing or isolation of the crawl space may be needed
in order to achieve effective pressurization. Crawl-space pressur-
ization works by causing the pressure of the crawl space to be
higher than the pressure in the soil, preventing radon-containing
soil gas from entering the crawl space (and thus from entering the
living area). Unlike crawl-space depressurization, pressurization
can also work in part by a ventilation/dilution mechanism, since
radon-free air is being blown into the crawl space. If the crawl
space cannot be effectively pressurized, the ventilation mechanism
may become the more important mechanism.
In limited testing of crawl-space pressurization in five
structures (References I, 3, and 14), where outdoor air was being
blown into the crawl space, crawl-space pressurization was found to
give living-area radon reductions usually in the range of 35 to
80%. These reductions are generally comparable to those obtained
using natural crawl-space ventilation (i.e., simply opening the
foundation vents, with no fan).
This result would suggest that the pressurization system was
probably not effectively pressurizing the crawl space. Indeed, in
one of the few cases where pressure measurements were reported
-------
(Reference 3), the crawl space was still under 0.006 in. WG (1.5
Pa) negative pressure relative to outdoors with the pressurization
system operating; the resulting radon reductions in the overhead
occupied space were only about 40%. The best results with crawl-
space pressurization systems were observed with two atypically
tight crawl spaces (Reference 14), where the crawl spaces could in
fact be pressurized to 0.02 in. WG (5 Pa) and greater relative to
outdoors; in these two houses, indoor radon reductions were over
80%.
The installation cost of a crawl-space pressurization system
would be about the same as that for crawl-space depressurization:
perhaps on the order of $500 (including insulation of water pipes),
unless additional effort is required to seal or isolate the crawl
space, or to install insulation under the overhead floor. The
operating cost for pressurization is difficult to estimate.
Assuming that the pressurizing air is blown into the crawl space
from outdoors, the annual operating cost should be no greater than
that for crawl-space depressurization, and greater than that for
SMD.
In summary, crawl-space pressurization would appear to be
applicable where the crawl space is particularly tight, so that
effective pressurization might in fact be achieved. Otherwise,
pressurization will be applicable in the same situations where
natural crawl-space ventilation (with no fan) is most applicable:
1) the required living-area radon reductions are on the order of
50%, and the crawl space may be a major source; 2) alternatively,
the crawl space is not a major source, and the crawl-space
pressurization/ventilation system is simply supplementing a SSD or
DTD system in an adjoining basement, so that high crawl-space
reductions are not needed; 3) the crawl space is reasonably well
isolated to begin with; 4) there are no water pipes in the crawl
space which the system might cause to freeze during cold weather
(or, alternatively, any existing water pipes can conveniently be
insulated, or the climate is mild); and 5) ideally, there is
already insulation beneath the overhead flooring (or the climate is
mild), reducing occupant discomfort and the heating penalty
resulting from blowing cold air into the crawl space.
Crawl-space pressurization might be selected over natural
ventilation only in crawl spaces where there is not a sufficient
number of properly distributed foundation vents to permit natural
ventilation to be easily implemented. This statement is based
upon: 1) the available performance data, which suggest that
natural ventilation is about as effective as pressurization in
reducing living-area radon, unless the crawl space is particularly
tight and thus amenable to true pressurization; and 2) the simpli-
city and reduced costs associated with opening existing foundation
vents, compared to the installation and operation of a fan. Even
where sufficient vents do not exist, the installation of additional
vents and the implementation of natural ventilation can still be
considered as an alternative to pressurization.
-------
NATURAL CRAWL-SPACE VENTILATION
In general, natural crawl-space ventilation consists of
opening existing vents in the crawl-space foundation wall (and/or
installing new vents), to increase the natural infiltration of
outdoor air. Natural ventilation works through two mechanisms.
First, it tends to neutralize the pressure between the crawl space
and outdoors. By thus reducing crawl-space depressurization during
cold weather, this technique reduces the driving force drawing
radon-containing soil gas into the crawl space (and thence into the
living area). Second, by providing an increased infiltration rate
for fresh outdoor air into the crawl space, the technique dilutes
any radon that does enter the crawl space.
The performance of natural crawl-space ventilation will
undoubtedly depend upon: the number and location of vents; the
tightness of the crawl space; weather conditions, especially wind
and temperature; the presence of vent obstructions (such as
shrubbery); and activities by the homeowner, such as the operation
of depressurizing appliances in the house or crawl space. Because
of the role of weather and homeowner activities, performance will
vary with time.
In limited testing as part of EPA's R&D program, natural
ventilation was found to provide radon reductions ranging from 46
to 83% in the living area of five buildings in which five to eight
existing foundation vents were opened (References 1 and 3). These
reductions are based upon relatively short-term testing, usually
for 2- to 4-day periods both before and after the crawl-space vents
were opened; this short measurement period probably contributed to
the scatter in the results, since performance is expected to vary
with time. In a somewhat longer-term study, testing over 5 to 7
weeks in two houses having 6 to 10 vents, provided indoor radon
reductions of 40 to 45% (Reference 15). Several-week testing in
three Spokane houses provided indoor reductions of 50 to 60%
resulting from natural ventilation in combination with crawl-space
sealing (Reference 16). Other short-term testing of natural vent-
ilation, sometimes in combination with sealing, has demonstrated
living-area reductions of: 21 to 35% in one house with 14 vents
(Reference 15); and 18 to 27% in two houses, and 75% in a third,
where only two vents were opened (Reference 2) . In a mitigation
effort where multiple vents were retrofit into a number of existing
crawl-space houses (to provide 1 ft2 of vent area per 150 ft2 of
floor area, or 0.1 m2 per 15 m2) , living-area reductions consistent-
ly ranged between 0 and 50% (Reference 17).
These limited current data suggest that average indoor reduc-
tions will probably be no greater than about 50% in most cases.
Some measurements suggest that natural crawl-space ventilation
may increase the ventilation rate of the crawl space by a factor of
3 or 4 (Reference 18). This result would suggest that crawl-space
radon concentrations should be reduced by 67 to 75% due to dilution
-------
alone, and by even more due to the reduction in driving force; and,
in fact, crawl-space radon levels are reduced by about that amount
(Reference 1). The fact that natural ventilation reduces living-
area concentrations by less than that amount suggests that the
reduction in crawl-space concentrations is accompanied by an
increase in the flow of crawl-space air up into the house. Perhaps
some radon is also flowing into the living area via a route other
than the crawl space (e.g., through a block fireplace structure).
If natural crawl-space ventilation can be implemented simply
by opening existing foundation vents, the installation cost will be
near zero. Where the crawl space has no vents (or too few vents)
and where vents thus need to be retrofitted, the estimated install-
ation cost is about $100 per vent, assuming that multiple vents are
being installed (Reference 17) . The installation cost would
increase when: 1) effort is required to isolate the crawl space
from the living area; 2) water pipes in the crawl space must be
insulated; or 3) insulation must be installed beneath the overhead,
floor, for occupant comfort or to reduce the heating penalty. The
operating cost associated with natural ventilation will be a
heating and cooling penalty in the house because the crawl space
will now be colder in the winter (and hotter in the summer). This
heating/cooling penalty cannot be rigorously quantified.
In summary, natural crawl-space ventilation will likely be
useful primarily where: 1) only a limited indoor radon reduction
is needed; or 2) the crawl space is not a major radon source, and
crawl-space ventilation is simply supplementing SSD or DTD in an
adjoining basement. See the earlier discussion regarding crawl-
space pressurization. Based upon EPA's mitigator survey (Reference
11), natural and forced ventilation of the crawl space are
preferred by 28% of the mitigators treating crawl-space structures.
While the survey did not inquire specifically, it is suspected that
most mitigators favoring natural ventilation are utilizing it in
houses meeting criterion 1) or 2) above. It is also suspected that
many mitigators using forced ventilation are using the fan to
increase the ventilation mechanism, rather than to pressurize or
depressurize the crawl space.
SEALING AND BARRIERS
There are two primary approaches for applying sealing and
barriers. One approach involves sealing a barrier such as poly-
ethylene sheeting over the crawl-space floor (and perhaps over the
interior face of the foundation walls) , to reduce the flow of
radon-containing soil gas into the crawl space. The second approach
involves sealing openings between the crawl space and the living
area, isolating the crawl space and reducing the flow of radon-
containing crawl-space air into the living area. This second
approach can include: closure of relatively small openings (such
as plumbing penetrations); closure of major openings, such as exist
when the crawl space is completely open to an adjoining basement;
and sealing the seams in any cold-air return ducting in the crawl
-------
space, to reduce the amount of crawl-space air drawn into this low-
pressure ducting and distributed throughout the house by the
forced-air heating/air-conditioning system.
The limited data available suggest that sealing openings
between the crawl space and the living area of existing houses, by
itself, may provide little or no reduction in the radon concentra-
tions in the living area. In tests on four houses (References 2,
15, and 16), attempts to isolate the crawl space resulted in
changes in indoor concentrations ranging from a 10% decrease to a
15% increase. All four houses had forced-air ducts in the crawl
space, complicating the isolation of the crawl space. These
observed changes in radon levels are within the normal variability
of indoor concentrations in a given house.
The effect of a completely sealed barrier of plastic sheeting
over the crawl-space floor and foundation walls has been tested in
only one house, resulting in indoor reductions of 31%, based upon
2 to 3 weeks of testing both before and after installation of the
membrane (Reference 16). Tests by other investigators, without the
membrane's being completely sealed, suggest that essentially no
indoor reductions are achieved when the barrier is not fully sealed
(Reference 1).
The openings between the crawl space and the living area would
appear to be so numerous, widely distributed, and sometimes
inaccessible — and the gas tightness and durability of sealed
barriers are so suspect — that sealing/barriers should never be
relied upon as a stand-alone mitigation method where the crawl
space is an important radon source.
According to the survey of mitigators (Reference 11), 25% of
the mitigators preferred some type of sealing or barrier approach
in treating crawl-space houses. Discussions with mitigators
suggest that, in fact, sealing and barriers in the crawl space are
most commonly utilized as a supplement to SSD or DTD in an
adjoining basement when the crawl space is not a major source.
Sealing can be most important when: 1) the crawl space opens to an
adjoining basement; and 2) depressurization, pressurization, or
natural ventilation of the crawl-space is to be utilized, and some
tightening of the crawl space is necessary in order to allow the
crawl space to be pressurized/depressurized, or to reduce the
heating/cooling penalty in the living area.
Depending upon the nature of the sealing effort undertaken,
sealing/barriers can be potentially time-consuming and expensive.
In summary, a homeowner may wish to try a sealing or barrier
technique as a stand-alone method on a do-it-yourself basis, to see
how well this approach might perform in that particular case.
However, the likelihood of achieving any meaningful radon reduction
with these techniques on a stand-alone basis is so small that a
commercial mitigator would likely never propose sealing/barriers
-------
except in combination with SSD in an adjoining basement, with a
crawl-space ventilation/pressurization/depressurization approach,
or with SMD in the crawl space.
DETAILED REVIEW OF DATA BASE ON CRAWL-SPACE TREATMENT METHODS
The review of the existing literature on radon mitigation
methods for crawl-space houses focussed on the house and mitigation
system variables which could have an impact on system performance.
These variables are listed in Table 1.
TABLE 1. VARIABLES ADDRESSED DURING REVIEW OF THE DATA BASE
FOR CRAWL-SPACE RADON MITIGATION SYSTEMS
House Design Variables
Nature of crawl-space floor (bare earth vs. gravel, evenness)
Crawl-space floor dimensions
Crawl-space accessibility
Nature of the foundation wall (block vs. poured concrete)
Presence of adjoining wings (basement or slab on grade)
Tightness of crawl space
House Operating Variables
Operation of central furnace fan or exhaust fan
Mitigation System Design Variables
Method for distributing suction beneath membrane
Number of suction pipes
Location of suction pipes
Diameter, type of suction pipes
Method of installing suction pipes through membrane
Nature of membrane
Extent of membrane
Degree of sealing of membrane
Mitigation System Operating Variables
Fan capacity
Fan in suction vs. pressure
Geology/Climate Variables
Source term (especially radon concentration in soil gas)
Permeability of underlying soil
Climatic conditions
Mitigation System Durability
Radon reduction performance
System suctions and flows
Equipment durability
-------
The discussion which follows focusses primarily on SMD and
crawl-space depressurization, since these are the techniques
expected to provide the greatest indoor radon reductions in crawl-
space houses.
HOUSE DESIGN VARIABLES
Nature of the Crawl-Space Floor
It might be expected that crawl spaces having gravel floors
would provide good performance with SMD systems having only one
individual suction pipe (even with relatively large floor areas),
without need for sub-membrane perforated piping or for strips of
porous matting beneath the membrane to aid in distributing the
suction. The gravel should enable good suction field extension,
analogous to SSD systems where a good gravel layer exists beneath
the slab. When the floor consists of bare earth, a greater effort
to distribute suction would likely be needed (more suction pipes,
sub-membrane perforated piping, or sub-membrane porous matting).
The limited data in crawl spaces with gravel floors (Refer-
ences 1 and 9) confirm that very good indoor radon reductions (92-
98%) are achieved with SMD, but, because other variables were also
varied, do not verify how large a floor area can be treated by one
individual SMD suction pipe when gravel is present.
Crawl spaces with bare earth floors have tended to give lower
reductions, often 80-97% (References 1, 2, 8, 9, 10, 19, and 20).
But other variables besides the nature of the floor were also being
varied (e.g., floor area, adjoining wing, method for distributing
suction). Thus, a quantitative comparison of gravel vs. bare earth
(in terms of the number of individual SMD suction pipes per square
meter of floor area required to provide comparable indoor radon
reductions) is not possible.
In most reported studies, where the floor is relatively even
and flat (Reference 1, 2, 9, and 12), 0.15 to 0.25 mm thick (or
equivalent) polyethylene sheeting has generally been found to be
sufficient, with strips of heavier material on top of the membrane
along expected traffic routes. Where the floor is irregular ~
e.g., with rock outcroppings (References 10, 18, and 19) — heavier
membrane materials were used, or matting was placed beneath the
membrane.
Crawl-Space Floor Dimensions
With SMD systems having individual suction pipes (i.e., no
sub-membrane perforated piping) where no gravel is present on the
floor, crawl spaces as large as 1,500 ft2 (140 m2) were reduced
below 4 pCi/L (150 Bq/m3) in the living area with a single centrally
located SMD suction pipe, based on results from 12 houses (Refer-
ences 2, 8, 9, and 20). When floor areas are larger than 1,500 to
2,000 ft2 (140 to 190 m2) , multiple individual suction pipes have
-------
consistently been needed with bare-earth floors to reduce levels
below 4 pCi/L (Reference 2), or else sub-membrane perforated piping
has been needed (References 1 and 5).
The number of individual suction pipes required for a given
floor area (or the need for sub-membrane perforated piping) might
be expected to increase in houses where: 1) it is desired to
achieve lower indoor levels (e.g., 2 pCi/L, or 75 Bq/m3) ; 2) there
is a high radon source term; or 3) the crawl space has dimensions
(such as an L shape) or an interior foundation wall which divides
the floor area into two or more segments. The number of pipes for
a given floor area (or the need for perforated piping) would be
expected to decrease when there is gravel on the floor.
Crawl-Space Accessibility
Where the crawl space is largely or completely inaccessible,
SMD will not be a practical option, since it will be impractical to
install a membrane directly over the crawl-space floor. Crawl-
space depressurization may be the only alternative in such cases
when indoor radon reductions much greater than about 50% are
required. Natural ventilation will be an option if only limited
reductions are needed.
In some cases where the crawl space is only partially
inaccessible — e.g., due to limited headroom at one end, or to
large obstructions — it may still be possible to obtain reasonable
reductions by applying SMD to the accessible portion of the floor
(Reference 1). Since only anecdotal results have been obtained to
date, it is unknown how much of the floor might be left uncovered
by membrane under different conditions while still providing
meaningful radon reductions.
Nature of the Foundation Wall
Almost all of EPA's data on 24 SMD systems are for crawl-space
houses having hollow-block foundation walls (References 1, 2, 8, 9,
10, 19, and 20). All but two of these study houses were reduced
below 4 pCi/L (150 Bq/m3) in the living area, and over half were
reduced below 2 pCi/L (75 Bq/m3) . Thus, it is impossible from the
limited data to separate out the effects of block vs. poured
concrete foundation walls. However, to the extent that the blocks
may have been contributing to radon entry, the SMD system would
seem to have been largely addressing that entry route.
On the other hand, in the two crawl-space study houses having
the highest source term (reflected by living-area levels of 88 to
160 pCi/Li or 3,250 to 5,900 Bq/m3), a block-wall depressurization
(BWD) component had to be added to the SMD system in order to
achieve adequate reductions (References 2 and 19) . Thus, separate
treatment of block walls may be important under some conditions
(such as a high source term) when the walls become important entry
routes. Some mitigators report treatment of block foundation walls
-------
using the SMD system by extending the membrane up the interior face
of the wall, and attaching the perimeter edge of the membrane to
the sill plate on top of the foundation wall (References 6 and 20);
in this manner, the SMD suction pipe might be able to treat the
walls as well as the floor.
Presence of an Adjoining Wing
When a basement wing adjoins the crawl space, the basement
will essentially always have to be treated (usually using SSD or
sump/DTD) . Whether or not the crawl space will also have to be
treated (by SMD or some other method) will depend primarily upon
the importance of the crawl space as a radon entry route
(References 2, 4, 6, 9, and 16). Where the crawl space is not
particularly important — usually as determined by grab radon
measurements in the crawl-space soil or air — the crawl space is
often treated solely by isolation from the basement, as needed,
possibly supplemented with some natural ventilation (References 4
and 6) . Good communication beneath the basement slab — suggesting
that SSD or DTD in the basement will be very effective — also
seems to reduce the need for separate crawl-space treatment
(Reference 9).
Where the wing adjoining the crawl space is a slab on grade,
available data are less definitive regarding when one or both wings
need direct treatment.
Tightness of Crawl Space
When the crawl space is tightly isolated from the living area
and from outdoors, the house may be a candidate for crawl-space
depressurization. A low effective leakage area, as determined by
a blower door, would be a quantitative indicator that the crawl
space is tight. Among the visual (qualitative) indicators that the
crawl space may be tight would be: l) the absence of forced-air
supply or return ducting penetrating the flooring between the crawl
space and the living area; or 2) the absence of foundation vents.
Another house design factor that should be addressed when
crawl-space depressurization is being considered is whether there
is a combustion appliance in the crawl space which might be back-
drafted if the crawl space is depressurized.
Another house design factor somewhat related to crawl-space
tightness is the degree of insulation of the crawl space (e.g.,
insulation under the living-area flooring). Such insulation might
make natural crawl-space ventilation (or forced pressurization
using outdoor air) a more viable option where only limited indoor
radon reduction is needed.
HOUSE OPERATING VARIABLES
A central forced-air furnace fan operating in the crawl space
could depressurize the crawl space, possibly helping both to
-------
overwhelm the sub-membrane depressurizations being created by a SMD
system, and to distribute crawl-space radon throughout the house.
Very limited measurements (Reference 18) suggest that a
central fan might depressurize a crawl space by 0.001 in. WG (0.2
Pa) or less. Limited sub-membrane depressurization measurements
with SMD systems indicate that, depending upon SMD design, sub-
membrane suctions can sometimes drop below 0.001 in. WG within 10
to 15 ft (3 to 4.5 m) of an individual suction pipe (Reference 2).
Thus, in theory, operation of a central furnace fan in the crawl
space could overwhelm the SMD system at locations remote from the
suction pipes. In practice, there are no clear data demonstrating
increases in indoor radon levels when the central fan begins
operating.
Tracer gas measurements in two houses (Reference 18) confirm
that operation of a central forced-air furnace fan located in the
crawl space will in fact significantly increase the inter-zonal air
flows between the crawl space and the living area. Thus, if the
SMD system is indeed overwhelmed, the central fan will help
distribute the radon throughout the house, as expected.
MITIGATION SYSTEM DESIGN VARIABLES
Method of Distributing Suction Beneath SMD Membrane
Suction can be distributed beneath the SMD membrane either by
installing individual suction pipes vertically down through the
membrane (analogous to SSD), or by drawing suction on some pattern
of perforated piping laid beneath the membrane (analogous to DTD).
Installation of perforated piping may increase installation cost by
perhaps $85 (Reference 12). However, since the measurable sub-
membrane suction field may not extend more than about 10 to 15 ft
(3 to 4.5 m) from individual suction pipes when the floor is bare
earth (Reference 2), sub-membrane piping could be a significant aid
in distributing the suction field. In addition, if leaks develop
in the membrane near the suction pipe penetration, distribution of
the suction through perforated piping would be expected to reduce
the amount of short-circuiting resulting from such membrane leaks.
Data from SMD systems are too limited — and too many other
house and system variables were being varied simultaneously — to
enable definitive guidance regarding the conditions under which one
approach or the other would be more cost-effective or otherwise
preferred. Intuitively, sub-membrane piping would be less
necessary when: 1) there is gravel on the floor, giving good sub-
membrane communication; and/or 2) the crawl space is smaller than
about 1,OOQ to 1,500 ft2 (90 to 140 m2), increasing the likelihood
that one individual suction pipe would be sufficient; and/or 3) the
source term is not high, and reduction of indoor levels below 2
pCi/L (75 Bq/m3) is not the objective.
' Among EPA's SMD study houses, all seven which used sub-
membrane piping were reduced below 2 pCi/L (75 Bq/m3) in the living
-------
area, even though one was as large as 2,700 ft2, or 250 m2 (Refer-
ences If 7, and 10). By comparison, none of the study houses with
individual-pipe SMD systems were reduced below 2 pCi/L, except
where the crawl-space was a relatively small wing adjoining a
basement that was being treated by SSD or DTD (References 8 and 9) .
Some commercial mitigators use the individual-pipe approach
almost exclusively, unless the crawl space is unusually large or
complex (References 5 and 6). Others use sub-membrane perforated
piping routinely (Reference 4).
Number and Location of Suction Pipes
With individual-pipe SMD systems — when there is no gravel on
the floor to aid in distribution of the suction field — one
suction pipe is generally sufficient to reduce indoor concentra-
tions to 4 pCi/L (150 Bq/m3) and less with crawl-space floors as
large as 1,500 ft2 (140 m2) . With crawl spaces of 2,000 ft2 (190 m2)
and larger, two or more pipes are needed (Reference 2) . The number
of pipes needed for a given crawl-space floor area would increase
if the source term were high, or if indoor levels below 2 pCi/L (75
Bq/m3) were desired. The number would likely decrease if there were
a gravel floor.
One mitigator has reported treating bare-earth crawl-space
floors as large as 2,500 ft2 (230 m2) with a single individual
suction pipe (Reference 6). This is accomplished by elevating much
of the membrane above the floor, creating an air space underneath
to facilitate suction field extension. Permeable matting beneath
the membrane might accomplish a similar objective, but would add
significantly to the cost.
With individual-pipe SMD systems, the suction pipe almost
always penetrates the membrane at a central location. Multiple
pipes are spaced uniformly.
With SMD systems using sub-membrane piping, one straight or
snaked length of perforated piping is commonly laid down the middle
of the crawl space, generally parallel to the long walls of the
house (References l, 4, and 8). More extensive interior patterns
have sometimes been tested in research studies (References 1 and
10). In a few cases, the perforated piping has been looped around
the crawl-space perimeter (Reference 1). Piping looped beside the
foundation walls might improve the chances that the system will
treat radon entry routes associated with block walls; however,
since the suction will be drawn immediately beside the perimeter
seam of the membrane, any leaks at that seam could degrade system
performance. Data are currently inadequate to enable guidance
regarding the preferred piping pattern under different
circumstances. In commercial practice, good experience has been
obtained with a single central length of piping (Reference 4).
-------
Diameter and Type of Suction Piping
The flows commonly encountered in SMD systems operating with
90-W fans are: 20 to 100 cfm (10 to 50 L/sec) for individual-pipe
systems (with the higher flows being obtained when there is gravel
on the floor); and 100 to 130 cfm (50 to 60 L/sec) for systems with
sub-membrane piping (and no gravel). At these flows, the commonly
utilized 4-in. (10-cm) diameter piping should be sufficient to
avoid undue suction loss due to flow friction.
Either Schedule 40 or lightweight polyvinyl chloride (PVC)
piping can be considered for the rigid piping above the membrane.
Flexible piping might be considered for the sub-membrane piping,
although use of rigid perforated pipe below the membrane could
reduce the risk of the piping's being crushed underfoot.
Method of Installing Suction Pipes Through Membrane
Individual SMD suction pipes must be installed through the
membrane in a manner which prevents their open end beneath the
membrane from resting flush against the dirt floor, thus restrict-
ing air flow into the pipe. Also, it is desirable to create some
form of cavity beneath the membrane where the pipe penetrates —
analogous to the sub-slab pit often excavated beneath SSD suction
pipes — to reduce suction losses as the sub-membrane gas acceler-
ates to pipe velocity.
Researchers and mitigators have utilized several creative
designs for accomplishing these goals. One simple approach is to
install an inverted tee fitting on the bottom of the pipe, with the
cap of the tee resting on the soil and the two open ends of the cap
directed horizontally beneath the membrane. The sub-membrane
cavity is created by attaching the membrane to the vertical suction
pipe (with a hose clamp and caulk) perhaps 1 ft (0.3 m) above the
tee, causing the membrane to be elevated above the floor at that
point. Another approach (References 2 and 5) is to: excavate a
pit beneath the membrane at the point where the suction pipe
penetrates; place a sheet of plywood over the pit beneath the
membrane; and install the pipe through a hole in the plywood with
the open end of the pipe suspended in the pit. Other mitigators
have used supports beneath the membrane (such as an inverted bucket
or a specially fabricated wooden flange) to support the vertical
suction pipe with its open end above the floor, while at the same
time elevating the membrane above the floor to create a cavity
(References 6 and 9).
There are currently insufficient data to suggest the condi-
tions under which one of these approaches might be better than the
others. The sub-membrane pit/plywood approach may be slightly more
complicated, but has the advantage of keeping the membrane flat
against the crawl-space floor, reducing the risk that an elevated
section of membrane might get ruptured by foot traffic. The sub-
membrane pit might occasionally also expose fissures or permeable
-------
strata beneath the soil surface, which might aid in extending the
suction field. The approaches that elevate the membrane may be
easier to implement, and can aid in extending the suction field
beneath the membrane by creating an air space under part of the
membrane.
Nature of Membrane
In almost all cases, 6- to 10-mil (0.15 to 0.25 mm) thick
polyethylene sheeting or equivalent is used as the SMD membrane
material. Most mitigators use at least 8-mil (0.20 mm) equivalent.
Cross-laminated polyethylene equivalent to 8- to 10-mil (0.20 to
0.25 mm) thickness is somewhat more expensive, but is often used
due to improved puncture resistance. Material stabilized to resist
ultraviolet (UV) radiation is preferred. The cross-laminated
sheeting can be sealed effectively at seams and at the perimeter
wall using urethane caulk; standard polyethylene does not bond well
with urethane caulk, so that some other adhesive must be used.
Heavier material, such as rubberized roofing material, should
be laid over the membrane along expected traffic routes.
Extent of Membrane
Mitigators almost universally cover the entire crawl-space
floor with the membrane when installing SMD systems. Tests in two
houses under favorable conditions (each having gravel on the floor
and having sub-membrane perforated piping) showed that good indoor
radon reductions could be achieved even when only a portion of the
floor was covered (Reference 1) . Further testing is needed to
determine under what conditions portions of the floor might be left
uncovered.
Degree of Membrane Sealing
Most mitigators usually seal the SMD membrane completely,
including the seams between sheets, and the junction between the
sheeting and the perimeter foundation wall and interior support
piers (References 4, 5, and 6). An analysis of SMD installation
costs (Reference 12) suggests that sealing can add $60-$120 to seal
seams between sheets; $100-$250 to seal those seams and to seal the
perimeter using a simple bead of caulk or adhesive; and up to $600
to seal sheet seams and to seal the perimeter by wrapping the edge
of the membrane around a furring strip which is then attached to
the wall. Many mitigators seal the perimeter using a simple bead
of caulk or adhesive; such bonds are reportedly so good in some
cases that the membrane can rupture before the bond is broken,
appearing to make the added expense of a furring strip unnecessary.
Data are currently insufficient to enable guidance regarding
the conditions under which complete sealing of the membrane can be
avoided.
-------
Limited testing in six houses with and without complete
sealing of the membrane (References 1 and 9) suggests that sealing
can sometimes provide limited additional reductions in living-area
radon concentrations achieved by the SMD system, and can reduce
system flows. However, in all six cases, very good indoor
reductions were achieved even without the membrane sealed. In
other crawl-space study houses, the degree of sealing was not
varied; the SMO systems were tested with the membrane either
completely sealed (References 2, 8, and 10) or not sealed
(Reference 2). The number of these other houses is too few, and
the number of other variables being varied simultaneously is too
great, to permit any assessment of the effects of membrane sealing.
However, all but one of the houses without membrane sealing were
reduced below 4 pCi/L (150 Bq/m3) .
MITIGATION SYSTEM OPERATING VARIABLES
SMD Fan Capacity
Most research installations and most mitigators usually use
the 90-W in-line duct fans common in the radon mitigation industry
(270 cfm, or 125 L/sec, at zero static pressure, and about 1.5 in.
WG, or 370 Pa, at zero flow) for SMD systems. Tests have shown
that good radon reductions can be achieved with less fan capacity
when sub-membrane communication is good; i.e., when there is gravel
on the floor (Reference 1) or when the membrane is elevated to
provide an air gap beneath the membrane (Reference 6). However,
smaller fans will probably often achieve less of a radon reduction,
especially when the membrane is not fully sealed (Reference 1) . At
least one mitigator (Reference 5) occasionally uses the larger,
100-W in-line fans (410 cfm, or 190 L/sec, at zero static pressure)
to help extend the sub-membrane suction field. No one has reported
testing low-flow/high-suction fans for SMD installations — 25 to
60 cfm (10 to 30 L/sec) at zero static pressure, and 25 in. WG
(6,200 Pa) or greater at zero flow. Such fans might be applicable
in some low-flow SMD cases, but could suffer a dramatic reduction
in performance if air flow increased over time due to leaks.
GEOLOGY/CLIMATE VARIABLES
Data on two houses having high source terms (References 2 and
19) confirm the expectation that an increased source term may
necessitate greater care in the design of a SMD system (e.g.,
addition of a BWD component). Where the underlying native soil has
relatively high permeability, sub-membrane communication may be
improved, reducing the number of individual suction pipes needed
for a given floor area. In two houses with permeable native soil
(Reference 2), "SMD" performed reasonably well without a membrane
(i.e., suction pipes embedded in the bare earthen floor).
MITIGATION SYSTEM DURABILITY
Only very limited data are available on the durability of SMD
systems. In six SMD installations by EPA (References 1, 7, and 9),
-------
no degradation in indoor radon reduction performance was measured
over periods of 1 to 3 years following installation. For
commercial SMD installations, mitigators report that they have not
generally had to revisit SMD installations under their warranty
programs, and have generally not received complaints from home-
owners regarding reduced performance or other problems with their
SMD systems (References 4, 5, and 6).
There have not been significant reports of fan failures with
SMD installations, beyond those reported for other types of ASD
systems. A couple cases have been reported in Iowa and Florida
where rodents seriously damaged the membrane within perhaps 6 to 12
months after the SMD system was installed; however, other
mitigators (References 4, 5, and 6) report that such membrane
damage has not been common in their areas.
CONCLUSIONS REGARDING DATA GAPS
The major data gaps regarding radon mitigation in crawl-space
houses appear to be:
1. The conditions (in addition to inaccessibility of the crawl
space) under which crawl-space depressurization might be
preferred over SMD (e.g., the tightness of the crawl space in
quantitative terms, and the exhaust fan capacity). The
operating cost penalty of crawl-space depressurization vs. SMD
needs to be defined, as a function of crawl-space tightness
and fan capacity/radon reduction performance.
2. The optimal method for distributing suction beneath the
membrane (i.e., the individual-pipe approach vs. sub-membrane
perforated piping in various configurations), under different
conditions (e.g., crawl-space floor area, nature of floor).
3. The optimal degree of membrane sealing required under
different conditions.
4. The conditions under which it may be possible to leave some
portion of the crawl-space floor uncovered by membrane, if a
portion of the crawl space is inaccessible.
5. The durability of SMD, crawl-space depressurization, and
natural crawl-space ventilation systems, in terms of both
radon reduction performance and materials lifetime.
REFERENCES
1. Findlay, W. 0., A. Robertson, and A. G. Scott, "Testing of
Indoor Radon Reduction Techniques in Central Ohio Houses:
Phase 2 (Winter 1988-89)," EPA-600/8-90-050 (NTIS PB90-
222704), May 1990.
-------
2. Pyle, B. E., and A. D. Williamson, "Radon Mitigation Studies:
Nashville Demonstration, " EPA-600/8-90-061 (NTIS PB90-257791) ,
July 1990.
3. Pyle, B. E., and K. W. Leovic, "A Comparison of Radon Mitiga-
tion Options for Crawl-Space School Buildings," in Proceed-
ings; The 1991 International Symposium on Radon and Radon
Reduction Technology.Volume 2f pages 10-73 through 10-84,
EPA-600/9-91-037D (NTIS PB92-115369), November 1991.
4. Anderson, J. W., Quality Conservation, Spokane, WA, personal
communication, February 6, 1992.
5. Howell, T., and D. L. Jones, Radon Reduction and Testing,
Inc., Atlanta, GA, personal communication, February 5, 1992.
6. Shearer, D. J., Professional House Doctors, Inc., Des Moines,
IA, personal communication, February 6, 1992.
7. Scott, A. G., A. Robertson, and W. O. Findlay, "Installation
and Testing of Indoor Radon Reduction Techniques in 40 Eastern
Pennsylvania Houses," EPA-600/8-88-002 (NTIS PB88-156617),
January 1988.
8. Gilroy, D. G., and W. M. Kaschak, "Testing of Indoor Radon
Reduction Techniques in 19 Maryland Houses," EPA-600/8-90-056
(NTIS PB90-244393), June 1990.
9. Messing, M., "Testing of Indoor Radon Reduction Techniques in
Basement Houses Having Adjoining Wings," EPA-600/8-90-076
(NTIS PB91-125831), November 1990.
10. Dudney, C. S., D. L. Wilson, R. J. Saultz, and T. G. Matthews,
"One-Year Follow-Up Study of Performance of Radon Mitigation
Systems Installed in Tennessee Valley Houses," in Proceedings^
The 1990 International Symposium on Radon and Radon Reduction
Technology. Volume 2f pages 7-59 through 7-71, EPA-600/9-91-
026b (NTIS PB91-234450), July 1991.
11. Hoornbeek, J., and J. Lago, "Private Sector Radon Mitigation
Survey," in Proceedings; The 1990 International Symposium on
Radon and Radon Reduction Technology. Volume 3f pages 4-17
through 4-30, EPA-600/9-91-026C (NTIS PB91-234468), July 1991.
12. Henschel, D. B., "Cost Analysis of Soil Depressurization
Techniques for Indoor Radon Reduction," Indoor Airf 1(3) : 337-
351, September 1991.
13. Henschel, D. B., "Indoor Radon Reduction in Crawl-Space
Houses: Evaluation of the State of the Art," submitted to
Indoor Air. April 1992.
14. Turk, B. H., Santa Fe, NM, personal communication, April 3,
1991.
-------
15. Nazaroff, W. W., and S. M. Doyle, "Radon Entry into Houses
Having a Crawl Space," Health Physics. j48(3): 265-281, 1985.
16. Turk, B. H., R. J. Prill, W. J. Fisk, D. T. Grimsrud, B. A.
Moed, and R. G. Sextro, "Radon and Remedial Action in Spokane
River Valley Homes. Volume 1," LBL-23430, Lawrence Berkeley
Laboratory, Berkeley, CA, December 1987.
17. Fisher, E. J., Office of Radiation Programs, U. S. Environmen-
tal Protection Agency, Washington, D. C. , personal communica-
tion (January 9, 1992).
18. Matthews, T. G. , D. L. Wilson, R. J. Saultz, and C. S. Dudney,
Oak Ridge National Laboratory, Oak Ridge, TN, personal
communication, May 1989.
19. Nitschke, I., "Radon Reduction and Radon Resistant Construc-
tion Demonstrations in New York," EPA-600/8-89-001 (NTIS PB89-
151476), January 1989.
20. Osborne, M. C., D. G. Moore, R. E. SoutherIan, T. M. Brennan,
and B. E. Pyle, "Radon Reduction in Crawl Space House," J.
Environ. Engineering. 115(3); 574-589, 1989.
-------
MULTI-POLLUTANT MITIGATION BY MANIPULATION OF CRAWLSPACE PRESSURE DIFFERENTIALS
by: Bradley H. Turk
Mountain West Technical Associates
105 E. Marcy St. , Rm. 109
Santa Fe, NM 87501
Gregory Powell
Gregory Powell and Associates
707 Galisteo St.
Santa Fe, NM 87501
Eugene Fisher, Jed Harrison, and Bryan Ligman
U.S. EPA Radon Division
401 M St. S.W.
Washington, DC 20460
Terry Brennan
Camroden Associates
RD #1, Box 222
Oriskany, NY 13424
Richard Shaughnessy
University of Tulsa
2643 E. 22 St.
Tulsa, OK 74114
ABSTRACT
Two demonstration case studies of manipulating the pressure gradient in
buildings with crawlspaces to control both indoor radon levels and outdoor air
ventilation rates, while maintaining energy efficiency are discussed. In one
study, the inaccessible crawlspace of a house with vented combustion applianc-
es was pressurized with outdoor air to limit soil gas entry into the crawl-
space. The technique reduced indoor radon concentrations and stopped back-
drafting of the appliances. To avoid cold floors, frozen pipes, and an energy
penalty, the pressurizing air stream was first tempered by passing through an
'earth tube' heat exchange pipe buried in the soil surrounding the house. In
the other study, mechanical ventilation systems were installed in two natural-
ly-ventilated school buildings to inhibit radon-laden crawlspace air from
entering through the structures' floors. In one building, a heat recovery
ventilator was used to depressurize the crawlspace and to pressurize the
classrooms. In the other school building, classrooms were pressurized with an
outdoor air supply fan. Besides source control of radon, the low existing
ventilation rates of the occupied zones were boosted, while in one school
building heat recovery from the depressurization exhaust air was made possi-
ble.
This paper has been reviewed in accordance with the U.S. Environmental
Protection Agency's peer and administrative review policies and approved for
presentation and publication.
-------
INTRODUCTION
The nature of crawlspaces below buildings encourages and sometimes
demands the consideration of innovative alternatives to radon control. Most
successful crawlspace mitigation in houses has employed sub-membrane de-
pressurization (SMD) or forced depressurization of the crawlspace (1,2,3).
However, significant air leakage through the crawlspace surfaces, the presence
of other hazardous materials in the crawlspace, backdrafting of combustion
appliances, and limited access to the crawlspace may require that different
mitigation options be considered. These options can provide an opportunity to
address other issues in the buildings -- such as indoor air quality (IAQ) and
energy use.
In this paper, we discuss two case studies of New Mexico buildings with
crawlspaces and elevated indoor radon levels. The first study is of a house
with shallow, inaccessible crawlspaces where forced pressurization of the
crawlspaces using passively tempered air reduced radon entry into the crawl-
spaces and inhibited backdrafting of vented combustion appliances by slightly
pressurizing the occupied space. The pressurizing air stream was first
tempered by passing through underground 'earth tubes', thereby reducing the
possibility of cold house floors, frozen service pipes, and a heating energy
penalty. The second, ongoing study is of a cluster of elementary school
buildings with natural ventilation and asbestos-containing material in the
crawlspaces. To investigate radon mitigation options for buildings having
crawlspaces with asbestos and for improving ventilation and indoor air quality
with a minimum energy penalty, different control techniques were installed in
two buildings. A heat recovery ventilator (HRV) was installed in one building
to both depressurize the crawlspace and ventilate/pressurize the classrooms,
while a system to pressurize the classrooms with respect to the crawlspace was
installed in another building.
SITE DESCRIPTIONS
HOUSE STUDY
The house (HEP1) involved in the first study was part of the U.S. EPA's
(EPA) Office of Radiation Programs' (ORP) House Evaluation Program. The
structure is small (57 m2), with four rooms, and exterior walls constructed of
adobe block. It is located approximately 20 miles (32 km) north of Santa Fe.
Three different additions were made to the house over its approximately 80
year life. Each addition is built over a separate, shallow (0.2 to 0.3 m)
crawlspace that has no access. The floor joists, in at least the kitchen
area, appear to be resting directly on the soil. The crawlspace exterior
walls are relatively air tight with no air vents and stucco plaster extending
below grade. The building has a flat, built-up roof with no attic space.
Three vented, LPG-fired space heaters and one domestic water heater are
located in the living space. Indoor radon levels during the winter before
mitigation were approximately 26 pCi/L.
SCHOOL STUDY
The three Santa Fe school buildings in the second study are being
jointly investigated as part of the EPA ORP's School Evaluation Program and a
-------
University of Tulsa study of indoor air quality in schools. The buildings are
located together on the same campus and were built approximately 50 to 60
years ago. Each building has a combination crawlspace and basement. Only the
basement in building SF1 contains classrooms. The crawlspaces in these
buildings also have rather air-tight exterior walls covered with stucco. They
are not vented to the outdoors. Each crawlspace is accessible through a
hatch(es), with space heights ranging from 0.3 to 1.8 m. Asbestos-containing
material and debris from an earlier heating system renovation were found in
all crawlspaces. All buildings are single story, naturally ventilated with
operable windows, and heated by wall convectors from hot water boilers.
Exterior walls are constructed of brick, clay tile, and concrete, while the
flat roofs have some small attic spaces. Other basic building information is
summarized in Table 1.
TABLE 1. INFORMATION ON 3 NEW MEXICO SCHOOL BUILDINGS
BUILDING
FIRST FLOOR
FLOOR AREA
[# rooms]*
CRAWLSPACE AREA
(X of substructure)
BASEMENT AREA
(X of substructure)
[# rooms]*
RANGE OF
PRELIMINARY
SHORT-TERM'
SCREENING
(pCi/L)
RANGE OF
PRELIMINARY
LONG-TERM*
SCREENING
(pCi/L)
SF1 6940/645 39 61 1.7-8.6 0.6-3.6
(CONTROL) [8] [3]
SF2 6090/576 82 18 0.7-39 0.5-16
[6]
SF3 5600/520 93 7 2.0-22 0.7-9.6
[8]
* Number of rooms includes commonly occupied classrooms, offices, libraries, and lounges.
' Charcoal canister measurements from 3/1/91 - 3/4/91 and 3/22/91 - 3/25/91.
* Alpha track measurements from 3/12/91 - A/19/91.
Buildings SFl and SF2 are connected by a heated hallway, although the sub-
structures do not appear to interconnect. As seen from the table, only
buildings SF2 and SF3 had classrooms with significantly elevated short-term
and long-term radon measurements during screening. After being alerted to the
presence of elevated radon levels, the school has adopted an interim policy of
opening windows at 7:00 am, before students arrive, and keeping windows at
least partially open throughout the school day.
STUDY DESIGNS
HOUSE STUDY
An initial attempt to provide radon control by forced crawlspace
depressurization caused flue gases from the combustion appliances to backdraft
into the living area. However, during the several hours of its operation, the
technique caused radon levels to be reduced below approximately 2 pCi/L. Sub-
floor pressure extension between the crawlspaces during operation of the
system and during subsequent diagnostic measurements indicated that forced
pressurization of the crawlspaces should inhibit soil gas transport into the
crawlspaces. In addition, tests suggested that this type of system would push
air up through existing openings in the house floor, thereby pressurizing the
-------
living space and forcing the proper draft on the combustion appliances.
Because operation of this system blows outdoor air directly into the
crawlspace, cold outdoor temperatures could cause cold house floors, an
additional heat load for warming the air, and the possibility of freezing
water pipes in the crawlspace when air temperatures dropped below freezing.
To alleviate this problem, the air was tempered by passing through O.lm
(4 inch) PVC 'earth tube' heat exchanger pipes buried approximately 1.2 m deep
in the soil surrounding the house (4,5). This earth-air heat exchanger system
has been most often used to cool air coming into a building (6,7). Figure 1
depicts the layout of this earth tube system. The large thermal storage of
the soil warms the air in the pipe during the heating season and cools the air
during the summer.
Two Porollel 4" Sched. 40 Earth Tubes Approx. 70' Long:
4 Ft. Below Grade, Rising to Within 12" of Grade
at Building Perimeter Where it is Protected by R—20
Rigid Insulation.
1/4" Condensate
e«p Holes
:Pressurization Fan
and Filter Box
Figure 1. Crawlspace pressurization system and earth tube configuration showing the details of pipe layout,
pipe penetrations (designated 'M'), and test hole openings (designated 'IF') into the crawlspaces in the
New Mexico study house.
An insulated box containing the in-line centrifugal fan was located on
the roof of the house to filter the incoming air and to reduce the noise
associated with fan operation. Once inside the house, the pressurization pipe
branched to three penetrations -- one for each crawlspace. The flow and
pressure in each crawlspace branch were adjusted by an in-line ball valve.
The total air flow delivered to all crawlspaces is approximately 0.042 m3/s
(90 cfm).
A Femto-Tech model R210F radon monitor was installed to continuously
record hourly radon data during the course of the study (12/90 through 4/91).
Flows, differential pressures (APs), and radon in grab samples were periodi-
cally measured as the system flow rates and distribution of flows were varied.
An electronic digital micromanometer (Neotronics model MP20SR) was used to
measure APs and flow pressures (with a pitot tube). Grab samples were
collected in evacuated 300 cc alpha scintillation flasks and the sample
activity was counted using a portable counter/sealer (Pylon model AB-5).
-------
After the earth tube heat exchanger was installed, an on-site data logger
continuously collected average air temperature data over consecutive 30-minute
intervals for approximately two weeks. Air temperatures were monitored in the
house and outdoor air, and in the earth tube air stream at two locations: 1)
at grade level where the pipe entered the soil, and 2) after leaving the soil,
where the pipe entered the house.
The continuous temperature data were used in a mathematical model with
empirically-derived parameters to predict the air temperature at the outlet of
the earth tube system as the outdoor air temperature changed. A thermal
network was established for each 3.1 m (10.0 ft) length of earth tube (7,8).
This network consisted of concentric cylinders of air, pipe, and soil around
the pipe (see Figure 2).
[8] ANNUAL AVG. SOIL TEMP
[7] 0.43 M (17.0 IN) SOIL
[6] 0.43 M (17.0 IN) SOIL
[5] 0.20 M (8.0 IN) SOIL
[4] 0.10 M (4.0 IN) SOIL
[3] 0.05 M (2.0 IN) SOIL
[2] PIPE WALL
[1] AIR IN PIPE
Figure 2. Cross section of the earth tube and soil showing the thermal network
temperature nodes and arrangement.
For a given inlet air temperature and soil temperature surrounding each
pipe section, a matrix of standard heat transfer equations was solved for the
eight temperature nodes. The new air temperature in the pipe was then used
for calculations in the succeeding pipe section. Thus, for each 30-minute
monitoring interval, outdoor air would progressively gain (or lose) heat until
it reached the end of the earth tube system, where the calculated outlet
temperature was compared to the measured temperature. The length of the pipe
section was determined primarily by the time to perform the extensive computa-
tions. Values for the numerous parameters required for the calculations are
from handbooks (7,9,10,11,12) and manufacturers and are shown in inch-pound
units in Table 2. The air flow rate in the pipe was assumed to be 0.044 m3/s
(94 cfm).
-------
TABLE 2. FIXED PARAMETER VALUES USED IN THERMAL NETWORK CALCULATIONS
MATERIAL
Air - Single Pipe
Air - Two Pipes
PVC Pipe
Soil
CONDUCTANCE
(Btu/h-ft3-°F)
3.454
1.985
--
CONDUCTIVITY
(Btu/h-ft-°F)
--
--
0.1201
1.50
iterated
DENSITY
(Ib/ff)
0.063*
0.063*
70
0.20
iterated
SPECIFIC HEAT
(Btu/lb-°F)
0.2447
0.2447
0.45
0.20
* Site elevation is approximately 1680 m (5500 ft) above sea level.
Because some of the parameter values are dependent on specific site conditions
(soil density, soil thermal conductivity, and average temperature of the
surrounding soil), iterative calculations were performed using different
values for these three parameters. Calculated and measured tube outlet
temperatures over the entire two-week monitoring period were compared graphi-
cally for each iteration. The values identified in Table 2 as 'iterated'
resulted in the best graphical agreement. The optimum average temperature of
the soil at node eight (8) and beyond was found to be 8.6 °C (47.5 °F).
SCHOOL STUDY
Blower door tests of SF2 and SF3 indicated that crawlspace depressuriza-
tion and reduction of indoor radon levels could be achieved with less than
approximately 0.94 m3/s (2000 cfm) of exhaust flow from each crawlspace.
However, concern over disturbing asbestos particles with air flow through the
crawlspaces prompted the school district to have the asbestos removed from SF2
in February 1992. Asbestos-containing material remained in SF1 and SF3. This
situation provided an opportunity to examine different radon reduction
techniques for buildings with and without existing asbestos problems (SF3 and
SF2, respectively). In addition, techniques were considered that could
improve ventilation and general IAQ, and reduce the energy penalty associated
with the additional ventilation (SF2). Because building SF1 had satisfactory
initial indoor radon levels, it was selected to serve as a study control --no
radon or IAQ mitigation work will be conducted. Mitigation work began in
March 1992 and.was substantially completed by mid-April 1992.
In building SF2, where asbestos had been removed, the crawlspace is
being depressurized to reverse the pressure gradient across the classroom
floors using the exhaust fan of an air-to-air heat exchanger (HRV). The heat
exchanger core scrubs available heat from the exhaust air stream and returns
it in the supply air stream to the upstairs classrooms and offices. The
supply air is routed to the upstairs via newly-installed uninsulated ducts in
the crawlspace. The supply air is then delivered to slightly pressurize the
rooms through hot water coils in the existing wall convectors. To meet the
ASHRAE Standard 62-1989 recommendation of 15 cfm (0.007 m3/s) of outdoor
ventilation air/occupant (13), the supply air delivery was designed to provide
approximately 450 cfm (0.21 m3/s) of outdoor air to each classroom, and
approximately 2800 cfm (1.32 m3/s) of total outdoor air to the building. In
order to maintain satisfactory energy efficient operation of the heat exchang-
-------
er core, the exhaust air flow rate was designed at approximately 3000 cfm
(1.42 m3/s). Since balancing and adjustments have not been completed, air
flow rates at this time tend to be lower than designed. New heating system
valves and thermostats have been installed to assure that the supply air is
delivered at a suitable temperature.
The asbestos-containing material still in the crawlspace of SF3 elimi-
nated the selection of any mitigation system that involved either the movement
of air or work in the crawlspace. Therefore, a supply air system that
pressurizes the occupied portion of the building to prohibit radon-laden air
from entering from the crawlspace has been installed. The system was
designed to provide a total of approximately 3600 cfm (1.70 m3/s) of outdoor
air (using ASHRAE 62-1989 recommendations) from a roof-top fan through
insulated ducts to all classrooms. Because the system does not have a return
side, ventilation is achieved by supply air exiting the classrooms through air
leaks in the ceilings, walls, and floors, and around doors and windows. The
incoming supply air is warmed by passing through hot water coils in each room.
The coils are fed by a new glycol/water loop that was added to the hot water
heating system, and the temperature of the supply air is controlled by a
sensing bulb and three-way valve at each coil. Doors and windows must be kept
closed to maintain a pressurized condition in the classrooms. To promote the
closed condition by providing better temperature regulation, new heating
system thermostats and valves were installed on the existing wall convectors.
Two classrooms in each of the three buildings were selected to have
sensors installed to continuously monitor temperature, relative humidity,
carbon dioxide (C02), differential pressures, and radon. Carbon dioxide is
passively monitored by a Gaztech model 1070 Ventostat, while radon in the
classrooms is monitored with the Femto-Tech model R210F. Differential
pressures are measured across the floor to the crawlspace and across the
exterior wall (both referenced to the classroom) with a Setra model 264
bidirectional pressure transducer with a range of ± 25 Pa. Radon is monitored
at one crawlspace location in each building with a Pylon model AB-5 monitor
with a passive scintillation flask. Temperatures are also monitored in each
crawlspace. Air flows in the ducting of the mitigation systems are determined
from stationary pitot tubes attached to pressure transducers (as above).
Outdoor wind speed and direction, temperature, humidity, and C02 are monitored
at one location on SF1. A data logger in each building is connected to all
sensors and stores 30-minute average values. This instrumentation was in-
stalled from February 1992 through May 1992 and will remain in place into the
heating season of 1992-93. Ventilation rates were measured in the three
buildings under conditions of windows/doors closed and windows/doors open
using an SF6 dilution/decay technique before mitigation systems were in-
stalled.
The initial diagnostic measurements showed that reversal of the
pressure gradient in the crawlspaces of SF2 and SF3 will, by itself, control
radon entry. Calculations indicate that the additional ventilation from the
mitigation systems installed in each building provides redundant radon
control. By switching the supply and exhaust fans on and off, this provides
an opportunity to study four different mitigation conditions in SF2: 1) crawl-
space depressurization alone, 2) classroom pressurization with ventilation, 3)
classroom ventilation alone, and 4) all (1 through 3) combined. These and
other tests will be conducted during the summer of 1992 and heating season of
1992-93.
-------
RESULTS AND DISCUSSION
HOUSE STUDY
Data from flow, pressure, and radon measurements are summarized in
Figure 3. Figure 3B exhibits the flow measurements made in each pipe branch
and the main pipe out of the fan. Figures 3C to 3H present data from radon
grab samples and AP measurements at each of the 6 test holes drilled through
the floor into the crawlspace. The AP data shown on the figure are for the
change in AP measurements made between the crawlspace and outdoors with the
system on and the system off.
The figure shows changes in the various measurements as the crawlspace
pressurization system was modified or turned off ('baseline' periods).
Columns 2 to 4 represent the 'initial' pressurization system that used
untempered outdoor air and one penetration into the crawlspace. Flow rates
designated as '1/2', '2/3', and '5/6' of maximum flow were from the position
of the fan speed switch. The system was then turned off (column 5, 'post rait
1 baseline') to prevent freezing of water pipes. All remaining columns (6 to
14) represent changes to the pressurization system with three penetrations
(Ml, M2, M3) and the earth tube heat exchanger in place. Adjustments in the
ball valves to change air flows through different pipe penetrations are
denoted in columns 7 to 13 by the term 'modify'. Changes to flows in individ-
ual pipes are also noted.
By directing more flow to various pipes, APs and radon concentrations
are affected differently in each of the crawlspaces. The most efficient
operation of the mitigation system was found to involve closing pipe M2 and
adjusting the flows in pipes Ml and M3 to be equal. The data show that indoor
radon levels were reduced to approximately 4 pCi/L after optimization of the
crawlspace pressurization system. The inability to reduce indoor radon levels
below 4 pCi/L, may be due to several factors. First, it is possible that
diffusion from the soil floor -- not convective flow of soil gas --is the
more important mechanism for radon entry into the crawlspaces of this house.
This could explain why radon concentrations in the crawlspaces remain elevated
even when crawlspace pressures are high enough to stop radon entry via
convective flow. In this explanation, reduction in indoor radon levels would
be due to dilution of the crawlspace radon from the ventilation air provided
by the pressurization system. A second explanation involves the construction
details of the wood floors to the crawlspaces. Since at least the kitchen
crawlspace is divided into separate zones by the joists that are resting
directly on the soil, air from the pressurization pipe that is forced into one
end of the crawlspace must pass through soil below the joists to reach the
zone farthest from the pipe. Figure 4 illustrates this process.
-------
00
c
H
ID
Cf
rr
a hi
n a>
M en
o v>
l-l C
Sil
O N
3 O
Cf
cn
r^ tt
S §
•*«
INITIAL
CRAWL
PASEMN.E
INITIAL
CRAWL
INITIAL MODIFY MODIFY
CRAWL EARTH MODIFY Ml UP MODIFY M 1 =• M3
|f flfl ALLIES MOD.FY M^N TO M^WN IfeW ^2=0^ S^NG
FLOW BASELINE fyjj. MJ DOWN M3 UP M3 DOWN SEALJKJG SEALING SEALING BIG FAN BASELINE
(A)
(B)
24-
o-
:
I
20
, r"l r^ H1
•H r"-.
4.4 4.6 1 1
1 1 1 1 1 1
20 --
MITIGATION PIPE FLOW
M1
+
M2
+
M3 TOTAL
1
(O g^|
0.1
TEST HOLE RADON & DELTA P
IF1
RADON DELTA P
H 1 h
100 -C
i.o
0.1
100 ^
IF2
IF3
n
n
i
•14
12
10
8
6
•12
•10
•8
6
•4
2
•14
12
10
8
6
4
2
g
•
S
100
»
io.o
, s .
(G) g'-
0.1
IF5
n
n
^ &
H
12
10
•4
2
14
12
10
8
4
2
IF6
n
9.8
:!24?
10 "^
B »-
6 iS
'* ui
2 Q
-------
HOUSE INTERIOR
Wood Floor: Joists on Soil—v
Decking —y \
Crawlspaee
Pressure. -
Mitigation
Pipe
i i i—
SOIL
Figure 4.
Air flowing through the soil below the crawlspace floor joists may pick up radon despite the high
crawlspace pressures.
The farthest zone would be pressurized, but would have acquired radon from the
air passing through the soil. In both explanations, we found that crawlspace
radon levels are highest at test holes farthest from the penetration of the
pressurization pipe, and that this radon would be forced into the first floor
living area.
A number of cracks and holes were sealed to reduce the air leakage in
the floor between the house and the crawlspace - those periods are denoted by
'sealing'. There was no observable change in APs, or in radon levels in the
crawlspaces or living area. To increase air flows in the pipes, a larger fan
was installed for one test period ('big fan'). Although crawlspace APs were
affected, there was no significant change in indoor radon levels.
Air forced from the crawlspace by the pressurization system caused the
first floor living area to be slightly pressurized (0.5 Pa) with respect to
outdoors, at floor level. With the pressurization system off, a typical AP
across the living area exterior walls of this house was approximately -3 Pa.
This change in living area pressures caused the flues of the combustion appli-
ances to draft properly -- reducing flue gas entry into the house.
The pressurizing air appeared to slightly increase the house ventilation
rate from 0.3 ach (estimated from blower door test data). On two separate
occasions with the mitigation system operating, ventilation rates were
estimated by observing the radon dilution rate for several hours after the
system was turned on. The ventilation rates were calculated to be 0.4 ach for
a low pressurization flow condition and 0.6 ach for a maximum flow condition.
Earth Tube Temperature and Energy Use Predictions
Figure 5 shows the actual measured temperatures for the earth tube
-------
system and the calculated outlet temperatures using the optimal parameter
values in the thermal network model. Note that the calculated and actual
outlet temperature curves converge after approximately eight days, then
diverge again after a break in the data (because of instrument failure) from
March 24 to March 25. The convergence is due to the time required for the
calculational procedure to establish a realistic temperature profile for the
soil close to the earth tube. The divergence probably results from the
calculational procedure not having the history of the missing outdoor air
temperatures to accurately predict tube and soil temperatures starting on
March 25, In general, though, the calculated temperatures very closely
predict the measured temperatures. The effectiveness of the earth tube heat
exchanger in moderating incoming air temperatures is demonstrated by the
stability of the outlet temperature despite large changes in the inlet
(outdoor) air temperatures.
65
C 18 . 3D
60
n C 1 5 . 6}
t? 55
Oi C 1 2 . BD
0)
-o 50
c 10 , 03
LL
45
nj *
« C ^ . 23
i_i 40
w C 4 • 4}
cr
3 35
< C 1 • ^}
cr
HI 30
a
2 C - 1 , 1D
LU
t- 25
C- 3 . 93
20
£
5
i
s jfc-k^ <*.
"^SS^*"^^
I
', .. 2:
} \\ ||:
: ' !.!':!
\ 'f, s
\K ' \
\
\
£
^
^
\^
i
i
s
::
- *• _/
»". r
• •
i :
? *: '^^vi'^
•i _SrT^ ._jl-* X^. V**^V ^ ~**f
s:^^*1^* ^^ y
/~ '
\ '
f il
i
"*
:
: =
f :
S
*^s
-. :. ";
i- • *d i
::
u.
7 i
/-IvLr^S. fi .j
f ^^ A ^o^yV^^^v"'
^
.
;
\
x :
~ :
"^~
*_
"
V.
Cv/
N/^
**'
i\
•
, \
;*:
^
'
0:00 0:00 0:00 0:00 0:00 0:00 0:00 0:00 0:00 0.0
14Mgr 16Mar 18Mar 2QMar 22Mar 24Mar 26Mar SBMar 30Msr 1Ap
DATE
MEASURED TUBE OUTLET TEMPERATURE
CALCULATED TUBE OUTLET TEMPERATURE
MEASURED TUBE INLET TEMPERATURE
Figure 5. Measured inlet and outlet temperatures for the earth tube are compared with predicted outlet temperatures,
Note the large diurnal temperature swings in the tube inlet temperature.
The energy saved by tempering the incoming air with the earth tube heat
exchanger was computed using simulated daily average outdoor temperatures for
an entire year (base data from the nearby U.S. National Weather Service
-------
reporting station at Espanola) . The thermal network model (with optimized
parameter values) was then applied to the year-long outdoor temperature
profile. Figure 6 shows the simulated outdoor temperatures and corresponding
earth tube outlet temperatures that were calculated. Differences in the daily
average energy necessary to heat the tempered and untempered air are also
shown for a heating season from September 26 through May 17 assuming that all
pressurizing air would be forced to the upstairs living area. The heating
season difference in total energy usage (3160 kwH for outside air vs. 2490 kwH
for earth tube air) is approximately 670 kwH (0.25 x 10" J) or approximately
$35 (using a thermal conversion efficiency of 70% and propane at $0.26/L
The minimum tube outlet temperatures that are likely to be encountered
were calculated for unusually cold conditions that were simulated for a 10-day
period in January. With simulated outdoor temperatures dropping as low as
-28.9 °C (-20 °F) for this period, the calculated tube outlet temperatures
never went below 4.9 °C (40.8 °F) . This suggests that it is unlikely that
tube outlet temperatures will be cold enough to freeze pipes in the crawl -
space. All tube outlet temperatures calculated for the above predictions were
based on the assumption that air would be blown through the earth tube system
for the entire year. In this way, energy from the warm air during the summer
would be stored in the soil to be recovered during the colder months.
I
LU
in
D
a.
LU
Z
LU
<
O
<
o
25
JUL AUG SEP OCT NOV DEC JAN FEB MAR APR MAY J UN
DATE
CALCULATED TUBE OUTLET TEMPERATURE
SIMULATED OUTSIDE TEMPERATUPE
ENERGY USAGE USING EARTH TUBE
— ENERGY USAGE USING OUTSIDE AIR
Figure 6. Daily average temperatures predicted using the thermal network model and simulated outdoor
temperatures. The daily energy necessary to heat tempered air from an earth tube and untempered
outdoor air are shown as separate curves.
-------
There are several drawbacks to the system installed in this house.
First, forced pressurization of the crawlspaces can possibly push other
pollutants that are specific to the crawlspaces (such as termiticides and
bioaerosols) up into the occupied areas. Although no measurements were made
of these pollutants in the occupied zone, there was no evidence to suggest the
presence of other important pollutant sources in the crawlspaces. The second
disadvantage involves the installation of the buried earth tube piping.
Because a large mass of soil is necessary for the heat reservoir, a sizable
space near the house is disturbed by excavation equipment. Care must be
exercised during the excavation to avoid disturbing underground utilities and
to assure an adequate slope for condensate to drain to weep holes. If
condensate is permitted to collect in the pipe, biological growths (e.g.,
molds and fungi) may flourish and be transported to the indoor air.
SCHOOL STUDY
Because installation of mitigation systems was not substantially
completed until mid-April, only preliminary data from a 7-day pre-mitigation
period and a 5-day post-mitigation period are presented here. Figure 7 is an
example of data from Rm. 15 and the crawlspace in SF2 (with the HRV) that
shows continuous plots for radon and APs during the pre-mitigation and post-
mitigation periods.
O
a
z
O
a
+
•- 8
•- 12
<
a
LU
Q
0:00 0:00 0:00 0:00
2Apr 5Apr BApr 11Apr
0:00
0:00
17Apr
0:00
20Apr
0:00
23Apr
0:00
2BApr
DATE
RM 15 RADON
CRAWLSPACE RADON
RM 15 CRAWL-1st FLOOR DELTA P
• RM 15 OUTS IDE-1st FLOOR DELTA P
Figure 7. Continuous data during pre- and post-mitigation periods for Rm. 15 and the crawlspace in SF2.
-------
Each of these periods includes a weekend when the school buildings were
closed and unoccupied (April 4 and 5, and April 23 and 24). The pre-mitiga-
tion data shows that the open-windows policy is effective at reducing indoor
radon levels during the school day. However, operation of the HRV system that
both depressurizes the crawlspaces and pressurizes/ventilates the classrooms
lowers radon levels even more during the school day and keeps the levels low
during unoccupied periods. With the HRV operating, the AP data clearly show
that the crawlspace is depressurized approximately 7 Pa with respect to the
first floor, and that the first floor is pressurized approximately 3 Pa with
respect to the outdoors. When windows and doors are opened during the school
day of the post-mitigation period, the first floor pressurization is reduced.
These data are summarized on Table 3 along with C02 measurements and
data from SF1 (control) and SF3 (classroom pressurization). Results are shown
for the two classrooms in each building with continuous monitoring equipment.
During the pre-mitigation period in all three buildings, the indoor
radon concentrations over the weekend tend to be slightly higher than for the
week nights (data not shown), and much higher than for the school day. This
result illustrates the limitation of using radon monitors that continuously
integrate over 24-hour periods (or longer) to accurately measure the occupant
exposure in schools (or other intermittently occupied structures).
The pressurization system in SF3 is very effective in controlling indoor
(and crawlspace) radon levels. The pressurization of the building is quite
robust, even though the mitigation system was not delivering designed air
flows during the post-mitigation period. Pressurization is significantly
diminished during the school day when doors and windows were opened by the
faculty and students. The data show that indoor radon levels also fell in the
control building (SF1) during the warmer post-mitigation period, although
crawlspace radon levels were not reduced as much as in the two buildings with
mitigation systems operating. Part of the change is probably due to the lower
driving forces (APs) resulting from milder outdoor temperatures during the
second period.
Average C02 concentrations during the occupied hours were as high as
1900 ppm in Rm. 17 (SF2), with a maximum 30-minute average of 4500 ppm and a
maximum single measurement of 4800 ppm. Carbon dioxide levels are signifi-
cantly reduced in the four classrooms of SF2 and SF3 by operation of the radon
control systems, while C0a levels in SF1 change very little during the second
monitoring period. These data suggest that the systems, despite not operating
at designed capacity, are improving the general IAQ, as well as reducing
indoor radon levels, in SF2 and SF3.
-------
Table 3. Summary of Selected Pre- and Post-Mitigation Data from Santa Fe Schools
.(April 1992).
Building
SF1*
PRE-mhig.
POST-mitig.
SF2
PRE*4Dltlg.
POST-mitig.
SF3
PRE-autig.
POST-mhig.
Period
Statistic
Weekend:*
WeekDajr.t
Weekend:
WeekDay:
Mean
Period (fan)
Mean
Period (bn)
Mean
Period (hrs)
Mean
Period (hrs)
Weekend:
WeekDay:
Weekend:
WeekDay:
Weekend:
WeekDay.
Weekend:
WeekDay:
Mean
Period (hrs)
Mean
Period (his)
Mean
Period (fars)
Mean
Period (hrs)
Mean
Period (hrs)
Mean
Period (hrs)
Mean
Period (hrs)
Mean
Period (hrs)
Radon Concentration
Rm9
5.0
60.5
2.6
40.0
0.9
60.5
0.5
24.0
Rml5
19.0
60.5
8.0
40.0
OS
603
0.6
233
Rml
28.
45.5
9.4
293
0.4
60.0
0.6
23.5
Rml3 I
55
60.5
2.7
40.0
13
60S
0.7
24.0
Km 17 |
63
60S
2.6
40.0
0.4
603
OS
24.0
1 Rm6
12.
603
4.1
40.0
0.6
603
0.6
24.0
(pCi/1)
Crawl
29.
603
17.
40.0
20.
603
7.6
24.0
Crawl
41.
603
29.
40.0
11.0
603
10.
24.0
Crawl
68.
603
49.
40.0
0.4
603
10.
24.0
Differential Pressure (Pa)*
CrawHst Ft
Rm9 .
0.4
573
-0.1
40.0
0.2
573
-02
24.0
RmlS
0.7
573
03
40.0
-6.9
573
-73
233
Rml
03
573
03
40.0
-4.1
573
0.8
24.0
Outdoor-lst Fir
Rm9
23
563
1.1
40.0
13
573
0.6
24.0
RmlS
0.9
573
03
40.0
-32
573
-2.1
24.0
Rml
2.1
57.0
0.6
40.0
-6&
573
-OS
24.0
Carbon Dioxide (ppm)
Rm9 |
450
573
1100
253
400
57.0
1250
24.0
RmlS
750
573
1000
25.0
350
573
750
23.0
1 Rml
600
573
950
253
350
573
600
24.0
Rml3
350
603
750
253
350
603
650
24.0
Rml7
550
573
1900
253
400
573
850
23.0
Rm 6
600
573
700
253
300
573
350
24.0
* Building SF1 is the control building where no mitigation system was installed.
f The 'weekend* is defined as beginning Friday at 17:00 and ending Monday at 05:00.
The 'weekday' is defined as beginning at 07:00 and ending at 15:00 for a day falling on Monday through Friday.
* All differential pressures are referenced to the first floor classroom.
-------
Ventilation rate measurements made before mitigation are summarized in
Table 4. Although ventilation rates have not been measured since installation
of the mitigation systems, estimates of ventilation for individual classrooms
in SF2 and SF3 have been made based on measured air flow from the supply
diffusers. At the air flow rates for the post-mitigation period, ventilation
rates would be approximately 4.3 and 4.8 ach for Rm. 1 and Rm. 6, respective-
ly, in SF3; and 1.6 and 1.4 ach for Rm. 15 and Rm. 17, respectively, in SF2.
Clearly, ventilation rates have been greatly increased in SF3 and have
probably been slightly increased in SF2. Larger increases are expected after
completion of system balancing and adjustments.
TABLE 4. VENTILATION RATE MEASUREMENTS BEFORE INSTALLATION OF MITIGATION
SYSTEMS
BUILDING
SF1:
SF2:
SF3:
Building Avg.
Classroom Range
Building Avg.
Classroom Range
Building Avg.
Classroom Range
WINDOWS/DOORS OPEN
(ACH)
0.8
1.0 - 1.6
1.1
0.7 - 1.1
1.7
1.6 - 2.1
WINDOWS/DOORS CLOSED
(ACH)
0.3
0.5 to 0.6
0.2
0.1 to 0.3
0.4
0.3 - 0.4
SUMMARY
The two case studies presented in this paper suggest that alternative
techniques for radon control exist that may also improve the general air
quality in a building. Furthermore, the additional energy usage to provide
these benefits can be lessened through the use of innovative or existing 'off
the shelf technologies. The specific control technique used in the house
study may not be applicable to a large number of other houses, since its
crawlspace construction is not often found in other regions of the United
States. However, the backdrafting of appliances by depressurization of the
crawlspace may occur occasionally - and will require consideration of control
methods such as the one described here. In addition to providing passive
heating of ventilation air during the winter months, the earth tube heat
exchanger can also furnish some space cooling and dehumidification during the
summer months.
The systems demonstrated in the school study may have more widespread
application. For schools with asbestos in the crawlspace or basement, the
classroom pressurization technique can not only significantly reduce indoor
radon levels, but also boost classroom ventilation rates (for buildings with
natural ventilation) and create an improvement in general IAQ, without
disturbing the asbestos. This technique also may be an effective radon and
IAQ control technique for buildings with slab-on-grade construction. The HRV
-------
installed in the other school building features a commercially-available unit
that not only provides radon control and improved IAQ, but also will reduce
the amount of energy necessary to heat the additional incoming outdoor air.
Because the mitigation systems in the schools represent newer and more complex
technology, the maintenance staff must be properly trained so that the systems
continue to be effective over the long term.
ACKNOWLEDGEMENTS
The authors gratefully appreciate the technical assistance of Ron Simon, John
Anderson, and Francis Offermann; the identification of study sites by Ron
Mitchell of the State of New Mexico Environment Department; the administrative
support of Mike Miller of the EPA's Region VI Radiation Office and John
MacWaters of SC&A; engineering and construction management of M&E Engineering;
the interest and efforts of the construction contractors; and the cooperation
of the homeowner in the house study, and the faculty, staff, school district
officials, and student body at the Santa Fe Public Schools. These studies
were funded by the U.S EPA Office of Radiation Programs' Radon Division.
REFERENCES
1. Henschel, D.B. Indoor radon reduction in crawl-space houses: evaluation
of the state of the art. Submitted to Indoor Air. 1992.
2. Pyle, B.E. , and Leovic, K.W. A comparison of radon mitigation options
for crawl space school buildings. In: Proceedings: The 1991 Interna-
tional Symposium on Radon and Radon Reduction Technology, Vol. 2, EPA-
600/9-91-037b (NTIS PB92-115369), 1992, pp. 10-73 to 10-84.
3. Pyle, B.E., and Williamson, A.D. Radon mitigation studies: Nashville
demonstration. EPA-600/8-90-061 (NTIS PB90-257791), U.S. Environmental
Protection Agency, Research Triangle Park, NC, 1990.
4. Nuess, R.M. Personal communication. 1991.
5. Nuess, R.M., and Prill, R.J. Radon control - towards a systems ap-
proach. In: Proceedings: The 1991 International Symposium on Radon and
Radon Reduction Technology, Vol. 2, EPA-600/9-91-037b (NTIS PB92-
115369), 1992.
6. Francis, C.E. Cooling with earth tubes. Solar Age. January 1984, pp.
30 to 33
7. Zoellick, W. Predicted and observed performance of a buried earth-air
heat exchanger cooling system. Publication reference unknown.
8. Kohler, J.T., and Sullivan, P.W. TEANET user's manual. Total Environ-
mental Action, Harrisville, NH, 1979.
9. Avallone, E.A., and Baumeister, T. Marks' Standard Handbook for
Mechanical Engineers. Ninth Edition. McGraw-Hill, 1986, pp. 4-56, 8-
198.
-------
10. American Society for Heating, Refrigeration, and Air Conditioning Engi-
neers. 1989 ASHRAE Handbook of Fundamentals. ASHRAE, Atlanta, GA,
1989, pp. 3.14, 22.21, 22.8.
11. American Society for Heating, Refrigeration, and Air Conditioning Engi-
neers. 1991 ASHRAE HVAC Applications. ASHRAE, Atlanta, GA, 1991, pp.
11.4.
12. Chemical Rubber Company, gRC Handbook of Chemistry and Physics 62nd
Edition. CRC Press, Boca Raton, FL, 1982, pp. E-2, F-49.
13. American Society for Heating, Refrigeration, and Air Conditioning Engi-
neers. ASHRAE Standard, Ventilation for acceptable indoor air quality,
ASHRAE 62-1989, ASHRAE, Atlanta, GA, 1989, p. 10.
-------
VII-6
TWO EXPERIMENTS ON EFFECTS OF CRAWLSPACE VENTILATING ON
RADON LEVELS IN ENERGY EFFICIENT HOMES.
by: T.D. Sterling1*; E.D. Mclntyre2, E.M. Sterling3
Professor, School of Computing Science,
Faculty of Applied Sciences,
Simon Fraser University,
Burnaby, B.C., Canada, V5A IS6.
Telephone: (604)291-4685/733-1348/681-2701,
Facsimile: (604)681-2702.
Author for correspondence.
2 Hughes Baldwin Architects,
300-1508 West 12th Avenue,
Vancouver, B.C., Canada, V6J 1H2.
President, Theodor D Sterling and Associates
Ltd.,
250-1122 Mainland Street,
Vancouver, B.C., Canada, V6B 5L1.
ABSTRACT
The reduction of radon levels in the living area of a
tightly sealed energy-efficient unoccupied home with a naturally
ventilated crawlspace in comparison with an older, traditionally
constructed home with a tightly closed, un-ventilated basement,
suggests that mechanical ventilation of a crawlspace might be an
effective mean to reducing radon levels in the home. Radon
concentrations were measured in the main living area of
unoccupied energy-efficient residences. The houses were of
woodframe construction with crawlspaces, a common construction
technology in many parts of North America. Radon levels were
measured under three conditions in the test homes: (a) crawlspace
completely enclosed, thereby preventing ventilation; (b)
crawlspace naturally ventilated with square vents located on
opposite sides of the structure; and (c) mechanical ventilation
of the crawlspace using a 120 cfm supply fan, providing
approximately two air changes per hour to the crawlspace.
Results showed a 60% reduction in indoor radon levels in the test
-------
home with the mechanically ventilated crawlspace compared to the
test home with the un-ventilated crawlspace, and a 30% reduction
compared to the test home with the naturally ventilated
crawlspace. 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.
-------
INTRODUCTION
Indoor radon levels are directly related to the ventilation
or fresh air exchange rate measured in air changes per hour
(ach). However, comparing indoor radon levels for several
ventilated houses rarely results in a significant or meaningful
correlation because of wide variations in the rate of entry of
radon into the indoor environment, even within the same community
(1). On the other hand, for a given rate of entry, indoor radon
levels are higher in energy conserving houses with low
ventilation rates below 0.5 ach than for conventional houses with
0.5-1.5 ach. A study of 21 energy-efficient homes and 14
conventional homes in New York State found indoor radon levels in
energy-efficient homes to average 2.7 pCi/L on a year round basis
compared to 0.89 pCi/L for conventional homes.(2) Nazaroff
estimated on the basis of measurements in energy-efficient houses
and crude estimates of average source levels for radon that, 36%
of English and 75% of American residences would have indoor radon
levels exceeding 0.02 WL if they were retrofitted to reduce
ventilation to 0.1 ach.(3)
The increased likelihood of radon contamination of energy-
efficient dwellings and the impact on occupant health makes it
important to explore ways and means to decrease exposure to radon
in energy-efficient dwellings through appropriate building
design.
EXPERIMENT #1
The effectiveness of direct outdoor ventilation of a
crawlspace or basement beneath an energy-efficient house in
reducing occupant exposure to radon contamination was evaluated
by comparing indoor radon levels in three immediately adjacent
residences during the time period October to December.
House 1: a new, energy-efficient home erected over an
unheated crawlspace with a vapor barrier consisting of 3/4 inch
plywood between the crawlspace and the house. The crawlspace was
naturally cross-ventilated by vents located in the foundation
wall.
House 2: a 67-year old, well built, non-energy-efficient
house with a standard non-mechanically ventilated basement with
operable windows. Furnace and central air supply ducts were on
-------
the basement level. Walls and ceilings were insulated. The
house was empty during the test period, with windows and doors
tightly closed.
House 3: a 50 year-old house similar in design to House 2
This house was occupied during the test period.
All three houses were two-storeyed homes of woodframe
construction.
Terradex Trac Etch Type SF Radon Detectors were placed in
the crawlspace or basements and living areas of Houses l and 3
for 88 days and of House 2 for 75 days. In all locations, the
Trac Etch Detectors were pinned to interior vertical partition
walls a minimum of 9 feet away from exterior windows and midway
between the floor and ceiling. The detectors were then collected
and pCi/L values determined.
RESULTS
Table 1 summarizes the pCi/L levels for the three houses.
The low radon level in the ventilated crawlspace is as expected,
as constant ventilation prevents buildup of any contaminant.
Despite being tightly enclosed, House 1 registered 0.82
pCi/L in the living area, approximately half as much as House 3,
which was a standard, non-mechanically ventilated structure, even
though House 3 was ventilated to the comfort of the occupants.
House 2, which was not tightened for purposes of energy
conservation but had been unoccupied with windows and doors
constantly closed, registered 2.04 pCi/L in the living area.
The effectiveness of ventilated crawlspaces plus a vapor
barrier between the crawlspace and house can be evaluated by
comparing Houses 1 and 2. While neither structure had been
occupied during the test period, each differs to some extent in
its rate of air infiltration. There was little, if any,
ventilation in House 1, which was a new, specially tightly
constructed structure. However, there was some air exchange in
House 2, which was a 67-year old structure that had not been
built to be airtight. Thus, an estimate of effect of ventilated
crawlspace and bottom vapor barriers errs toward the conservative
side.
-------
The effect of ventilation on radon levels emerges from a
comparison of Houses 2 and 3. For similar houses, normal
ventilation associated with occupancy seems to reduce pCi/L by
approximately 25%. If the same reduction due to normal
ventilation is assumed to apply to House 1, then radon levels in
that house would be approximately 0.61 pCi/L when the ventilation
system in that house was turned on.
Assuming the values measured in Houses l, 2, and 3 are
typical, one can estimate that the use of a ventilated crawlspace
and a bottom-side vapor barrier can reduce indoor radon levels by
approximately 65%.
Table 1 also shows expected yearly WLMs for occupants under
conditions represented by Houses 1, 2, and 3. Estimates make the
assumption that an employed occupant spends 13.4 hours indoors
(4,891 hours per year) and a housekeeper spends 20.5 hours
indoors (7,482.5 hours per year).(4) One conclusion flowing from
such a comparison is that exposure to radon by housekeepers may
approach levels deemed excessive by EPA or Canadian authorities
if they should: (a) live in geographic regions with high
radiation background, (b) live in tightly closed houses that (c)
lack ventilated crawlspace and bottom-side vapor barriers.(2)
EXPERIMENT #2
This experiment investigated the effect of mechanically
ventilating crawlspaces as a means of reducing radon levels in
houses. A new duplex (i.e. two adjoining but separate buildings)
was constructed adjacent to the houses for which we had already
obtained indoor radon measurements. Each unit in the new duplex
structure was built on a crawlspace. The crawlspaces of each
unit were separated by an insulated, woodframe firewall
incorporating an air vapor barrier. The two new units were
identical in all relevant parameters and size to the previously
constructed airtight homes with ventilated crawlspaces. It was
thus possible to compare indoor radon levels as a function of
different crawlspace ventilation in an unoccupied, new, sealed
building, wherein:
Condition 1: The crawlspace is completely enclosed,
preventing any ventilation;
Condition 2: The crawlspace is naturally ventilated; and
-------
Condition 3: The crawlspace is mechanically ventilated.
Condition 1 was accomplished by blocking the supply and
exhaust vents in the crawlspace exterior walls. Blocking off the
supply and exhaust vents was accomplished with duct tape and 6
mil polyethylene. Mechanical ventilation for Condition 3 was
accomplished by installing a fan with a capacity of 120 cfm and
ducted to the 6-inch diameter supply vent. A 6-inch diameter
exhaust vent was installed at the opposite end of the crawlspace.
Insofar as the crawlspace contained 3,600 cubic feet, the use of
this small fan ensured two air changes per hour. Under Condition
2, a 2-foot-square supply vent and a 2-foot-square exhaust vent
were located on opposite sides of the crawlspace (see figure 1).
During each measuring period, one unit had the supply and
exhaust vents in its crawlspace blocked for a period of 30 days,
while the other house had the ventilation ports opened and the
fan operating 24 hours a day. At the end of each measuring
period, the ventilation conditions were reversed, and new
monitors installed.
Again, radon levels were estimated during four measuring
periods using etch-type radon detectors. Detectors were kept in
place for 30 days. In all units during each measuring period one
recorder was installed on each floor of the unit and one recorder
was installed in the crawlspace. In all locations, the detectors
were pinned each time in the same position to interior vertical
partition walls a minimum of 9 feet away from exterior windows
and midway between the floor and ceiling.
RESULTS
Table 2 gives the average values observed for Conditions 1,
2, and 3.
Under Condition 1 with the crawlspace un-ventilated, the
mean level of radon measured on the main floor was 1.94 pCi/L.
Under Condition 2 with the crawlspace naturally ventilated, the
mean radon level on the main floor was 0.82 pCi/L. Under
Condition 3 with the crawlspace mechanically ventilated, the mean
radon level on the main floor was 0.57 pCi/L. Compared to a non-
ventilated crawlspace, mechanically ventilating a crawlspace led
to an approximate reduction of 70% in indoor radon levels.
Mechanically ventilating the crawlspace with a small fan (no more
than 2 ach), yielded a 30% reduction on radon levels over what
would be achieved by having a naturally ventilated crawlspace.
-------
Table 2 also shows expected yearly WLMs for occupants under
Conditions 1, 2, and 3. Estimates were based on the assumption
that an unemployed occupant, such as a homemaker or a very small
child, spends 20.5 hours per day indoors (or 7,482.5 hours per
year).(4)
In conclusion, such a comparison shows that exposure to
radon by homemakers or small children may approach levels
declared excessive by EPA or Canadian authorities, if they should
(a) live in geographic regions with high radiation background;
and (b) live in sealed houses without ventilated crawlspaces.
Homemakers or small children are least exposed in houses where
ventilation of the crawlspace is enhanced by mechanical means.
(Retired persons and invalids would be exposed at a similar rate
as homemakers, but exposure would have relatively less severe
consequences for older than for younger persons.)
While our findings are based on a small sample of
observations in Vancouver and on relatively short measuring
periods, they are in line with expectations based on the effects
of mechanical ventilation and, to a limited extent, on the
findings of others. Indoor radon levels have been found to be
higher in energy conserving houses with low ventilation rates,
below 0.5 ach, than for conventional houses with 0.5 - 1.5 ach.
A study of 21 energy-efficient homes and 14 conventional homes in
New York State found indoor radon levels in energy-efficient
homes to average 2.7 pCi/L on a year round basis, compared to
0.89 pCi/L for conventional homes.(2) Insofar as one major entry
of radon into a structure is through gas emanating from the soil
beneath the structure, an air/vapor barrier between crawlspaces
and living areas, combined with a mechanically ventilated
crawlspace, should ensure minimal entry of radon gas into a
residence. Extra care in placement of pipes and other direct
connections from the soil into the building is of additional
importance.
Insofar as a link between radon exposure and lung cancer
seems to be well established, inclusion of mechanical ventilation
in the crawlspaces might be highly recommended by construction
standards organizations.
In light of the very low radon levels measured, longer
measuring periods than 30 days would have been desirable.
However, the arrangements that had been made to keep the houses
unoccupied permitted measurements only over a 120-day period.
-------
ACKNOWLEDGMENTS
This study was made possible, in part, through the Canadian
governmental policy of granting special research tax credits to
research-oriented Canadian corporations. Buildings were
financed, constructed by, and all research supported by TDS
Limited. David Mclntyre designed the buildings and supervised
their construction. Elia Sterling served as research architect.
Part of the study was supported by a special grant from the
Council for Tobacco Research.
-------
REFERENCES
1. Nero A.V., Boegel M.L., Hollowell C.D., Intersoll J.G.,
Nazaroff W.W. Radon concentrations and infiltration rates
measured in conventional and energy-efficient houses. Health
Phy.. 45:401, 1983.
2. Fleischer R.L., Turner L.G. Indoor radon measurements in the
New York capital district. Health Phv.. 46:999, 1984.
3. Nazaroff W.W., Boegel M.L., Hollowell C.D., Roseme G.D. The
use of mechanical ventilation with heat recovery for
controlling radon and radon-daughter concentrations in
houses. Atmos. Envir.. 15:263, 1981.
4. Szalai A. The use of time. The Hague: Mouton, 1972.
-------
TABLE 1. RESULTS OF RADON MEASUREMENTS IN THREE HOUSES IN VANCOUVER
Yearly WLM*
Unoccupied sealed house
with ventilated crawlspace.
Unoccupied house with
partially occupied basement.
Occupied house with
basement.
pCi/L
Crawlspace 0.07
1st Floor 0.82
Basement 0.38
1st Floor 2.04
Basement 0.82
1st Floor 1.57
WL*'
0.0041
0.0102
0.0079
Employed *
0.029
0.073
0.056
Unemployed*1
0.045
0.112
0.086
' 1 WLM = 680 hours of exposure to 1 WL.
1 WL = 200 pCi/L, assuming an equilibrium between radon and its daughters of 0.5. U.S. EPA
remedial action if WL > 0.015.
* Estimated average of 13.4 hours per day spent indoors at home (4,891 hours per year).
** Estimated average of 20.5 hours per day spent indoors at home (7,482.5 hours per year).
-------
TABLE 2. COMPARISON OF EXPOSURE TO RADON IN MAIN
LIVING AREAS OF HOUSES WITH CRAWLSPACES
(1) WITHOUT VENTILATION; (2) NATURALLY VENTILATED;
AND (3) ENHANCED VENTILATION VIA MECHANICAL MEANS
Average radon level.
WL
WLM for employed occupant
(based on 4,891 hours per year).
WLM for unemployed occupant
(based on 7,482 hours per year).
Crawlspace
Sealed
1.94
0.0097
0.069
0.112
Crawlspace
Ventilated
Naturally
0.82
0.004
0.029
0.043
Crawlspace
Ventilated
With Fan
0.57
0.0028
0.020
0.031
-------
SUPPLY
VENT
EXHAUST VENT
Unit A
FAN
SUPPLY
VENT
FAN
EXHAUST VENT
UnitB
Figure 1. Crawlspace ventilation in experimental homes.
-------
Session VIII
Radon Occurrence in the Natural Environment
-------
VIII-1
INDOOR RADON AND THE RADON POTENTIAL OF SOILS
by: D.J. Steck and M. J. Bergmann
Department of Physics
St. John's University
Collegeville, MN 56321
ABSTRACT
Subsurface soil samples, collected adjacent to 110 Upper
Midwest houses, were measured to determine radon emanation,
diffusivity, and permeability. The soilborne radon
characteristics of the samples of glacial till that covers most
of the region varied by a factor of ten or more. A Monte Carlo
model was used to calculate indoor radon concentrations in
idealized basements sited in a homogeneous medium whose radon
properties matched the soil sample properties. Measured basement
radon concentrations correlate with the model predictions. The
correlation is more useful to assess clusters of homes than
individual dwellings.
-------
INTRODUCTION
Direct airborne radon measurements can be a time-consuming
and costly way to survey the potential radon problems in a large
sample of homes. At sites were no home exists, no direct indoor
measurements are possible. Thus, it would be useful to find an
alternative, indirect technique for predicting indoor radon
concentrations in a home, a cluster of homes, or a site for a
future home. Indoor radon concentrations can vary significantly
from place to place. The Upper Midwest region of the United
States contains broad areas of modestly elevated indoor radon
with widely scattered small towns. Within this region, localized
areas of high radon concentrations (hot spots), often the size of
neighborhoods, have been found (1). Despite widespread publicity,
few homes have been measured for radon. Thus, radon
concentrations are poorly known in most populated areas outside
the major cities. In such an environment, an inexpensive,
effective radon assessment technique would be quite useful to
public health officials.
Since soils have been identified as a primary source for
indoor radon (2), it seems reasonable to base a radon potential
assessment protocol on soil characteristics. This approach is not
new. In fact, soil-based assessment protocols have been applied
to a limited number of dwellings with varying degrees of success
(3). Since many of these methods include intensive, in situ,
measurements they are not cost- competitive with direct airborne
surveys (3). In the hopes of finding an efficient screening
method for homes with elevated radon, we have investigated the
utility of simple, inexpensive soil sample tests that are used
with homeowner-supplied information to predict annual-average
radon concentrations in homes.
METHODS
INDOOR RADON MEASUREMENTS
More than 230 homes, located in Minnesota, Wisconsin, and
Michigan, were monitored for indoor radon concentrations and
sampled for radon source material (soil and water). Airborne
radon concentrations in basements were measured for a period of 1
yr.or longer with alpha-track detectors. Typical uncertainty in
these data is 15%. Waterborne radon concentrations were measured
with a liquid scintillation technique. Waterborne concentrations
were too low (median value 9 kBq m"3 ) to be considered a
significant radon source. Other experimental details have been
reported previously (4).
-------
SOIL SAMPLES
Participating homeowners were asked to collect soil samples
from a depth greater than 30 cm at four separate locations within
1 m of their basement walls. We selected 110 representative soil
samples for analysis. In the lab, the samples were mixed,
categorized into types (sand, silt, clay) by microscopic
examination of grain sizes, and their porosities were measured.
In order to replicate in situ soil characteristics, water was
added until the concentration equaled the average of the wilting
and saturated concentrations ( 2% to 20% by weight). Local
building practices call for soils near homes to be compacted to
90% Proctor density (5). We packed the moist samples into PVC
columns (5 cm diameter, x 6 cm) and compacted them with a
instrument patterned after ASTM Standard 78(6).
EMANATION AND 222Rn CONCENTRATION
The soil columns were sealed and allowed to age. Then the
pore gas was extracted into a scintillation cell. The gas in the
cells was allowed to age for 3 hours, counted, and recounted
after 12 hours to help eliminate any contimantion from thoron
daughters. Results from this system were reproducible to within
10% with a total uncertainty of approximately 15%.
Gamma spectroscopy was used to measure the total soilborne
222Rn concentration to an accuracy of 10%. (4)
DIFFUSION
The effective, textural diffusivity was measured using a
scintillation chamber patterned after Cohen, et al. (7). Soil
columns served as leakage pathways for a continuously-monitored
scintillation chamber filled with a high concentration of radon.
The time dependence of the count rate over a 10 to 16 hour period
was fit to extract the diffusion constant. This procedure yielded
effective diffusion lengths to an accuracy of 8%.
PERMEABILITY
We developed a computer-interfaced permeameter capable of
measuring permeabilities in the range from 10"10 m2 to 10"15m2. In
this device, pressure gradients and flow rates, close to expected
field values, were measured by electronic sensors. Usually, five
separate sets of pressure - flow data were collected and then fit
to extract the permeability for each soil sample. Although the
accuracy of the fit is generally better than 10%, a comparison
between the experimental and theoretical permeabilities for two
columns constructed with glass beads ( 1 mm and 0.1 mm), suggest
that the permeabilities are accurate to with a factor of two.
-------
Since we have observed order of magnitude variation in a single
sample's permeability as a function of moisture and compaction, a
factor of two in uncertainty seems acceptable.
TRANSPORT MODELS
Ninety percent of the homes that we surveyed have basements
that normally extend 2 meters or more into the ground. We have
limited our analysis to basement-type homes with no known nearby
bedrock. In order to find a tractable solution to the radon
transport problem, we have made a number of simplifying
assumptions, namely that:
• soil is the primary source of indoor radon;
• the soil is homogeneous;
• the indoor and outdoor airborne radon concentrations are low
compared to the pore gas concentrations;
• the effects of soil moisture are represented by the transport
characteristics of moist soil samples;
• the radon entry can be separated into a component through the
basement walls and a component through the floor;
• radon enters the basement walls via diffusion and
pressure-driven flow;
• radon enters the basement floor via diffusion;
• the diffusive component is governed by Pick's Law and the
pressure-driven component is governed by Darcy's Law (1)
• the basement pressure is uniform.
After analyzing our data with two previously published
models (8,9), we developed a Monte Carlo model simulation to see
if we could improve the correlation between the theoretical
predictions and the measured indoor radon concentrations. We used
a simple, two-dimensional geometry (see Figure 1), where the
solution to Laplace's equation yields a simple vector field for
the pressure-driven velocity, namely:
a f\ ,..,
~T° (D
where k is the permeability, P is the pressure difference between
the outside and the basement, eis the porosity, p. is the
viscosity of air, and r is the distance from the basement wall -
ground intercept. We model diffusion as a macroscopic random
walk. The bulk diffusion constant, D, generates a step size, d,
as:
d=^6DAt (2)
In order to assure that each test atom can take at least 30
steps before reaching the basement, we determine the time
-------
increment, At, between each random step as the minimum of three
characteristic times of convection, diffusion and decay, e.g.
At =
MIN'-f,
where T is the mean life of radon. The model is used to
approximate the probability, P(x,y), that a radon atom generated
in the soil at (x,y) will reach the basement. We define the
effective thickness, A, of soil that contributes to the radon
entry as:
P(x,y)dxdy
4)
where F is the basement depth. Then the radon flux, F, can be
expressed as:
F = E ATw
where w is the width of the basement.
p=o
(5)
Figure 1. Monte Carlo model geometry, pressures, and streamlines
A sample trajectory is shown.
-------
Two dimensionless parameters, p and 5, are used to characterize
the transport distances (in units of basement depth: T) due to
pressure-driven and diffusive flow respectively:
We populated the parameter space with 63 data points
distributed log - uniformly in the region bounded by 0.001< p<
30 and 0.001 < 8 < 1. We interpolated values of A between these
solution points. In order to calculate the probability for each
case, we placed a large (>30,000) number of particles in a
truncated region of space next to the basement. After the initial
solution has been calculated, the region is extended, a new
solution calculated and the process repeated until the
probability changed by less than 3%. Each solution required 100
CPU minutes a Silicon Graphics 4D25 workstation. We compared our
solutions at extremum points (p,8 = 0) with the Tanner model (9) .
House characteristics, such as basement pressure differential,
volume, air exchange rate, must be estimated in order to generate
the model predictions. Although we have collected and coded some
information from homeowners that may be useful in estimating
these parameters, we chose to include only the depth of the
basement in the first analysis. The standard basement depth used
by the model was 2.5 m. For the other input parameters, we
selected a "standard" environment of 8 Pa basement
depressurization, basement dimensions of 10 x 15 x 2 . 5 m, and an
air exchange rate of 0.3 ach. We calculated the diffusive flux
through the basement floor using a simple one dimensional
solution (2) and recently - published slab diffusion coefficients
(10).
RESULTS
Table 1 summarizes the key statistics of the basement radon
and soil source characteristics. Figures 2 through 4 show the
complete distributions.
The basement radon concentrations, shown in Figure 2,
indicate that elevated indoor radon concentrations are common in
the Upper Midwest . Although the observed soil source parameters
span a broad range, they generally are somewhat below the average
values that have been previously reported (2) . We discovered that
classifying soil types by grain size was not a significant
distinction vis a vis the source or transport characteristics.
For example, even though the mean diffusion length and
permeability for clays were smaller than sands, the range of
variation within the clays overlapped the range of variation of
-------
Table 1. Airborne and soilborne radon characteristics.
Measure
Basement Rn
Emanation
Diffusion Length
Permeability
Effective Thickness*
Unit
Bq nf 3
kBq m°
cm
nr
cm
Mean
150r
II4
120
10-12F
140
SD
2*
6
60
10r
60
Range'
45
3
45
10
20
to
to 3
to
600
2
155
~15 to 10'"
to
200
* 95% cumulative fraction interval
+ Log normal distributions; geometric mean and standard deviations listed.
± This corresponds to an emanation fraction of 0.4 and an equilibrium radon
concentration in the available pore spaces of approximately
40 kBq m° (lOOOpCi/L)
s Monte Carlo model generated parameter
sand parameters. This may reflect the wide variety of grain sizes
that occur in a typical sample of our glacial till. We note that
while the permeability spans four orders of magnitude, the rest
of the variables (including the basement Rn and the effective
source thickness) only span one order of magnitude or less.
.99
z
o
M
H
CJ
HI
>
H
_J
D
X
O
-1 1—r
1—i—r
i
r
i fe
.01 b
i -I—L_L-
100
RADON CONCENTRATION (Bq m~-3)
_
4
5
•
u
-i
0_0_3
1000
Figure 2 .Basement radon concentration distribution.
-------
ETT
.99 p-
r r-rr
•n-i-T-i-rrr;T3
J
z
o
a
<
a:
u.
.
D
O
•sp
-1*
E
.01
i i I i i i i i i i i i I i i i i I i i
5 10 15 20
EMANATION (kBq m~-3)
«J
25
Figure 3. Soil sample emanation distribution.
2
O
H
0
tr
u.
1 1 i 1
.99 f-
\~
u
P
.9 r-
1
L
L
LU
D
O
T—r~i—r~r
I—[-
,01
_
-
r
20
.*
•
l i l 1 l 1 i i
40 60 80 100
12O 140 160
DIFFUSION LENGTH (cm)
Figure 4. Soil sample diffusion length distribution,
-------
.99
z
o
o
<
GC
u_
HI
>
l-l
U
. 1
01
IT
l.E-15 l.E-14 l.E-13 l.E-12 l.E-11 l.E-10
PERMEABILITY ( nT 2 )
Figure 5 Soil sample permeability distribution
Previously published models failed to produce a significant
correlation between the predicted and observed radon
concentrations. Our Monte Carlo model produced a better
correlation including an order of magnitude agreement between the
observed and predicted indoor radon concentrations. While the
probability that this fit could result from a random set of data
is small ( p<0.01), the percent of the variation explained by the
model ( 6%; R2= 0.06) is also small. The dimensionless parameters
for pressure (p) and diffusion (5) calculated from the soil
samples are shown in Figure 6. Figure 7 illustrates the Monte
Carlo model-generated dependence of the effective source
thickness (A) on p and 5. The effective thickness distribution
of the soil samples is shown in Figure 7. A comparison between
'the model predictions and measured radon in all 110 homes is
shown in Figure 8.
DISCUSSION
The radon environment in the Upper Midwest is rich and
complex. Although median basement concentrations are 150 Bq m~3,
-------
10
2 1
PU
0)
OH .1
.01
-*. . r . r
lit
Diffusion Parameter (<5)
Figure 6 Soil sample pressure (p ) and diffusion (5) parameters
0.30
-0.30
(P)
-x.so
-2.50
tog (A)
-0.2
-2.50 -Z.UO -T.5O -l.DO -O.SO 0.00
(6)
Figure 7 Effective thickness (A)contour plots.
-------
concentrations in excess of 4000 Bq rrf3 have been observed. Yet,
the glacial till surrounding the houses generally is not high in
radon content or transport properties. However, the climate
encourages construction of basements, heating of homes from
October to May, and cooling of homes from June to September.
Thus, soilborne radon can enter through a large, relatively
permeable shell (most basement walls are cement block or
fieldstone) which is depressurized relative to the soil for a
substantial fraction of the time. The region receives ample
precipitation ( approximately 700 mm annually) that sustains
forests in the east and agricultural prairies in the west. Since
soil moisture plays a pivotal role in radon availability (11),
its seasonal and cyclic variation can dramatically influence
indoor radon. In particular, the role of snow and frost (which
often penetrates several meters into the ground in some years) is
not well documented. In addition, the static water table is often
near the surface and varies cyclically. These factors make
predicting indoor radon concentrations a difficult task in our
region. For example, approximately half of the indoor basement
radon measurements used in this study were made during years of a
severe drought in the north central part of the region. Yet, the
moisture content of the sample material was selected to
approximate historical average soil conditions.
Although our model results do indicate a good correlation
between the measured and predicted indoor radon concentrations,
numerous factors associated with the particular environment of
each home complicate any attempts to simply model the radon entry
and retention. Below is a partial list of those factors, not
included in our model, that could be expected to affect the
indoor radon in a particular basement:
• source variability
* soil inhomogeneity
» structural versus textural transport properties
* vertical and horizontal moisture distributions
* near surface water tables
* compactness of soil adjacent to basement walls
• structural variability
* transport properties of subsurface structure/holes
* depressurization: HVAC systems
» basement - superstructure mixing
» local meteorological conditions; wind, precipitation
* superstructure infiltration / ventilation rate
We are presently trying to incorporate information about the
individual homes into our model in order to improve its
performance. Unfortunately, a preliminary analysis of the utility
-------
.99
cc
L.
HI
>
_J
D
U
. 1
.01
;! I I 1 [ I I I I
I I I I I I I I I I I I I
: l I i i I i I i i I
i i I i i i i I i i
50 100 150 200
EFFECTIVE THICKNESS ( cm )
250
Figure 8 Soil sample effective thickness distribution.
cn
*E
CD
c
cr
a
tu
800
600
400
cc
UJ
In 200
o
0 1=
100 20O 300
PREDICTED Rn (Bq/m~3)
400
Figure 9 Model predictions compared to observed radon
concentrations. The solid line is the regression fit
-------
of the information supplied by homeowners showed no significant
correlations between any categorical responses and indoor radon.
However, we may be able to improve the model by adjusting a
combination of parameters, e.g. the basement pressure, entry
area, ventilation rates, and shell transport characteristics.
Since a number of these factors might not vary over a small
geographic locale, we tested the model for 20 localized clusters
of homes by calculating the average predicted indoor radon
concentration and comparing it with the average measured radon
concentration for each cluster. An acceptable cluster was defined
as more than 3 houses located within a 5 km radius. The
comparison is shown in Figure 10. The model is significantly more
successful for clusters (R2 = 0.34 ) than individual houses. In
fact, if we define the coefficient of variation (COV) of a
prediction as the average deviation of the data from the fit,
expressed as a percentage, then the COV for our model is 100% for
individual homes but only 35% for clusters. These COV's compare
favorably with COV's of screening surveys based on short-term
airborne measurements using a single detector in each house (1).
400
T 300
E
\
IT
CQ
200
Q
111
>
QC
HI
S
o
100
1
100
PREDICTED Rn
200
(Bq/nrT3)
300
Figure 10 Average radon concentrations for clusters of houses as
predicted by the model compared to the measured
concentrations.
-------
CONCLUSIONS
Indoor radon concentrations in basements can be predicted
from soilborne radon parameters using a simple model. At present,
the model is most useful for predicting average concentrations in
localized clusters of houses. Work is underway to investigate
whether the utility of the model can be improved through the
modelling of additional transport, entry, and retention factors.
AC KNOWLEDGEMENTS
We would like to thank Dr. Carol Meger, Norm Tyrell, Fred
Monette, Adam Whitten, Dr. Phil Byrne, and Greg Taft for
assistance with various parts of the instrument development, data
collection, and analysis. Special thanks is given to Dr. Tom
Kirkman who provided valuable suggestions and criticisms
throughout the project.
This work was supported, in part, by a grant from the
Minnesota Private College Research Foundation with funds provided
by the Blandin Foundation of Grand Rapids, Minnesota.
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. Steck, D.J. Spatial and temporal indoor radon variations.
Health Physics . 62: 351, 1992
2. Nazzaroff, W.W., Moed, B., Sextro, R.G. Soil as a source of
indoor radon: generation, migration, and entry, in: W.W.
Nazaroff and A.V. Nero (eds.) Radon and its Decay Products in
Indoor Air. Wiley, Interscience New York, N.Y., 1988
3. Scott,A.G. Site characterization for radon supply potential: a
progress review. Health Physics 62: 422, 1992
-------
4. Steck, D.J. Variations of radon sources and indoor radon along
the southwestern edge of the Canadian Shield. Environment
International 15: 271, 1989
5. Gary Traut, private communication, Braun Engineering, St.
Cloud, MN
6. American Standard Testing Methods for moisture - density
relation of soil and soil-bearing aggregate D 698, 1978
7. Cohen, B.L., Rakowski, J., and Nason, R. A simple compact
apparatus for measuring diffusion properties of radon through
soils and other materials. Health Physics 50: 133, 1986
8. Nazaroff, W.W. and Sextro, R.G. Technique for measuring indoor
222Rn source potential of soils. Environmental Science and
Technology 23: 157, 1989
9. Tanner, A.B. A tentative protocol for measurement of radon
availability from the ground. Radiation Protection Dosimetry
24: 79, 1988.
10.Nielsen, K. K. and Rogers, V.C. Radon entry into dwellings
through concrete floors. In: Proceedings of The 1991
International symposium on Radon and Radon Reduction
Technology. U.S.E.P.A. Philadelphia, Pa. 1991
11.Rogers, V.C. and Nielsen K. K. Correlations for predicting air
permeabilities and 222Rn diffusion coefficients of soils.
Health Physics 61: 225, 1991
-------
NATURE AND EXTENT OF A 226-RA ANOMALY IN THE WESTERN SWISS JURA MOUNTAINS
by : Heinz Surbeck
Federal Office of Public Health
Rad.Surv.Sec. (SUeR)
c/o Physics Institute, University
Perolles
CH-1700 Fribourg, Switzerland
ABSTRACT
High indoor radon concentrations (up to 5 kBq/m3 in living rooms)
are abundant in the Western Swiss Jura Mountains, a Karst terrain. The
search for the radon sources has led to the discovery of very high 226Ra
activities (up to 850 Bq/kg) in soils covering the low activity limestone
bedrock.
Nature and extent of this radium anomaly are presented and possible
origins are discussed.
A striking feature of this anomaly is a strong disequilibrium in the
23aU series. The 236Ra is in equilibrium with its parent nuclide 230Th
but the precursors of the 230Th are present at far lower levels only.
Many soils of the Western Jura Mountains show a loess-like grain
size distribution, pointing to an aeolian origin. The dust has probably
been deposited at the end of the Hurra glacial period. Leaching of this
very fine material may have led to the observed disequilibrium in the
2aaU series.
-------
INTRODUCTION
Those of you mainly interested in radon remedial actions may consider
the reserach presented in this paper as sheer luxury. But sometimes it
pays off to invest in a better understanding of the nature and extent of
a radon source prior to mitigation. If, as an example, it can be shown
that there is a widespread radium anomaly of natural origin, you will not
waste your time and money to clean up tiny artificial sources.
High indoor radon concentrations are frequently found in the Western
Swiss Jura Mountains (1,2,3). These mountains consist of at least several
hundred meters of Jurassic and partly Cretaceous limestone. They have
been folded up during the late phase of the alpine orogeny. The peaks are
at about 1600 m above sea level. Karst features like sinkholes or caves
are abundant. The thin soil cover on the mountain heights (above about
900 m) is younger than the last glacial period (Wurm) and contains no or
only very few alpine material. Windblown (aeolian) material of unknown
origin is present in many of the soils (4).
Despite the low 226Ra concentration in the limestone bedrock and the
lack of alpine material with possibly higher activities, high 22SRa
activities have been found in many soil samples from the Western Jura
Mountains.
It was tempting to blame the watch industry for the enhanced radium
concentrations; an important part of the Swiss watch industry is situated
in this region. Until about 1965 this industry processed large quantities
of 226Ra. Radium activated luminous paint was applied onto watch dials
and hands.
Radium contaminations are clearly present by they do not explain the
high activities found in soils sampled at the top of the mountains. A
strong evidence for the natural origin is also the presence of enhanced
230Th activities in the high-radium soil samples. 230Th is the precursor
of the 226Ra in the Z3BU series. 230Th has not been found in samples of
luminous paint used by the watch industry and therefore can be used as an
indicator for the natural origin of the 226Ra (5,6).
Whereas it seems to be clear that the watch industry can not be made
responsible for the widespread radium anomaly in the Western Swiss Jura
Mountains, the origin of this anomaly reamains unclear.
-------
In this paper I present what is already known about the extent and
the nature of this anomaly, how this information has been gained by
gamma spectrometry and how it may help to reveal the origin of the
enhanced 226Ra activities.
The list below summarizes important characteristics of the anomaly.
(1) The highest 32SRa activities found is nearly 900 Bq/kg dry weight.
(2) The 22SRa is in equilibrium (within experimental error) with it's
natural precursor 230Th.
(3) There is a strong disequilibrium in the "au series. The first
member with enhanced activity is the 230Th.
(4) There is an extremely large radon emanation from the soils with
the enhanced aasRa activity. At least half of the "3Rn produced
can escape to the pore space and is thus available for transport.
(5) The specific 2a*Ra activity as a function of the grain size
peaks at intermediate grain sizes (near 200 um). This is in clear
contrast to the activity vs. grain size pattern found in "normal"
soils from the Swiss Plateau. In "normal" soils the specific
activity decreases monotonically with increasing grain size.
(6) Clearly enhanced activities (>150 Bq/kg dry weight) have so far
only been found at altitudes higher than 900 m above sea level.
(7) The anomaly extends over a strip of at least 100 km lenght and
some km width. There is a tendency to higher values towards the
south-west.
Figure 1 helps to locate the sites mentioned in this paper.
-------
Western Swiss
Jura Mountains
Chasseral
Tete de Ran
Chasseron
Ml. Tendre
Marchairuz^
*V
La
Figure 1 . Swiss map with (he locations mentioned in the text and in the other
Figures. The dashed lines roughly represent the axis oF the
respective geographic unit.
-------
MATERIALS AND METHODS
SAMPLING
In general soil samples are taken to a depth of 10 cm and are split
into two subsamples of 0 to 5 cm and 5 to 10 cm. The standard sample
surface is 10 cm x 10 cm. Some of the samples have been taken with a 7 cm
diameter soil auger down to a depth of 50 cm. At most places in the Jura
Mountains the soil (A and B horizon) has a thickness of hardly 30 cm, but
in small depressions the soil can be as deep as 2 m.
SAMPLE PREPARATION
The soils are dried under a heating lamp. The temperature is kept
below 60 °C to avoid destruction of the clay mineral structure. During
drying the soils are carefully stirred by hand to avoid the formation of
large hard blocks. After drying, a part of the grain size fraction < 32
urn is separated by dry sieving.
The separation of the fractions larger than 32 urn is done by wet
sieving in two step. About 500 g can be treated in one batch. For the
first step, the coarse separation, a vibrating sieve machine is used.
About 50 liters of water flow through the soil during this step. The
fractions are then carefully washed by hand. A sharp water jet is used to
dissociate the agglomerates that formed during drying. Gentle rubbing
with the fingertips accelerates this process. The clay fraction staying
suspended and the floating organic material are discarded. The separated
fractions are then dried under a heating lamp. During the second sieving
step there are again about 50 liters of water flowing through the soil
sample of about 500 g. The quantitiy of water used for both steps
corresponds to about the quantity of rainwater that has passed through
the soil during the last 10 years. The water used for wet sieving is
ordinary tap water, a non-chlorinated groundwater. It shows, apart from
about 3 Bq/liter of 222Rn, no measurable radioactivity.
Repeated washing with large quantities of water showed that at least
for these soils from the Jura Mountains there is no measurable leaching by
wet sieving.
-------
Just before measurement, each grain size fraction is saturated with
water and dried under a heating lamp within about 1 hour. This procedure
removes any radon that has escaped to the pore space. The radon remaining
in the sample after this treatment is the part that could not escape to
the pore space.
MEASUREMENT
If there is sufficient material, 40 cm3 of the dried sample are
packed into a flat polystyrene box. In this case the gamma spectrum is
taken with a 25% relative efficiency HP-Ge-detector. Smaller samples
(grain size fractions) are measured with a well-type HP-Ge-detector.
Construction and shielding materials of this detector have been selected
for low activity. This makes it possible to reliably determine natural
and artificial radionuclides in samples of only some grams and
measurement times of 2 days.
The calibration of both detectors is based on measurements of IAEA
reference samples (SOIL-6, SDN-2, SDA-1, RG-series). The calibration
points for the 25% detector are fitted with a single 5th-degree
polynomial (log(effficiency) vs. log(energy)) over the energy range from
20 keV to 2 MeV. The well-type detector is calibrated for each gamma line
individually.
The gamma spectra are evaluated with a interactive program developed
by Intertechnique (INTERGAMMA). Mainly below 100 keV the operator has to
improve the marking of the peaks. This is very important for a reliable
determination of the 230Th from it's weak 67.6 keV peak.
22
-------
Therefore the fraction of the radon produced that can escape under
the most favorable conditions to the pore space and is thus available for
transport can be determined from the difference between the 22SRa
activity and the activities of the radon daughters in a freshly dried
sample.
RESULTS AND DISCUSSION
Figure 2 shows the disequilibrium in the 23aU series for some samples
from the Harchairuz-Nt.Tendre region. Whereas 230Th and 22SRa are in
equilibrium (within measuring errors) the 23*Th activities are at only a
fraction of the 32*Ra levels. The equilibrium between 226Ra and it's
precursor 230Th is a strong evidence for the natural origin of the 226Ra.
In samples of luminous paint used by the watch industry I have found no
230Th (less than 1/10 of the 226Ra activity) (5,6).
According to alpha spectrometry measurements carried out on some of
my sample by a different laboratory, 23
-------
1200
s 1000
^m
w
"3
* 800
•o
I 600
o-
ffi
jj 400 •
3
=
« 200
a)
a
0
2
•D
$ 5
I ™
O 5
9
0 g
ffi
o
I I I I I
1200 •
1000 -
-
800 -
-
600
•
400 -
200 -
0.
<
b)
>
T
o
o
o
r n
I I I
.* o to t o
z s s ™ z
i A
4
{
.1 *
0
0 I I I I I
i 0 to t o
M N «^ t}I CSI
f £ a) A A
£ £ GC Q. 0.
a
d
o
•_
T3
m
**
3
u
Ml. Tendre #0 , 0-5 cm depth
d> Mt.Tendre *0 , 5-10 cm depth
-------
this fallout.
The 236Ra activity concentration as a function of the grain size
(Fig.3) peaks at very small and at intermediate grain sizes. This is
different from the behaviour in "normal" soils from the Swiss Plateau,
where the activity decreases monotonically with increasing grain size
(3,9).
.A hint to understand this strange pattern found in the soils from the
Jura Mountains may come from the activity vs. grain size for 137Cs and
13*Cs. Both cesium were brought in by the rainwater and have been
adsorbed by the soil from solution. In soils from the Swiss Plateau the
13TCs activity vs. grain size shows that the cation adsorption or
exchange capacity (c.e.c.) is highest for the smallest grains (<32 ym).
This can be attributed to the presence of clay minerals (< 2 jim) in the <
32 urn fraction. The 137Cs activity pattern in the Jura Mountains soils
shows the presence of a high-c.e.c. substance not only in the smallest
fraction but also in intermediate grain sizes. The similarity between the
13'rCs and the 5*26Ra acitivity vs. grain size patterns points to a common
origin. aasRa may therefore have been adsorbed from solution too. This
would also explain why such a large part of the radon produced can escape
to the pore space. The nature of the high-c.e.c. substance in the large
grains remains unclear. That this exchange capacity is still present can
be seen .from the -l3*Cs activities vs. grain size. 13'*Cs has its origin in
the fallout after the 1986 Chernobyl accident.
Mainly the soils from Marchairuz contain many dark grains of
different diameters. With the skill and the patience of a gold panner
grain size fractions larger than 64 urn of these soils can be separated
into a lightweight dark and a heavier light part. The lightweight part
consist of dark, very hard grains. They are not an artefact of the drying
process, they are already present in the fresh soil. The light part
consist mainly of small pieces of limestone and some clear calicte
crystals. The lower part of figure 3 shows that the activity
concentrations are far higher in the dark grains than in the light
grains.
Dark, nearly indestructible grains in soils from the (French) Jura
Mountains have also been described by Gaiffe (10). She has found that
they are composed of organic material and clay minerals held together by
a cement containing iron hydroxide. Either these clay minerals or the
iron hydroxide may be the reason for the strong adsorption capacity of
the grains in my soil samples.
Enhanced a26Ra concentrations have so far only been found in the
Western Swiss Jura Mountains and only at altitudes higher than 900 m
above sea level (11). This altitude corresponds to about the upper ice
margin at the end of the last glacial period. It is therefore tempting to
attribute the radium anomaly to an aeolian deposition during the retreat
of the glaciers, with the Jura Mountains free of ice. According to Pochon
-------
1UUU •
O)
f 800 -
3
•a 600 •
ci
o 400
.&
I 200
o
n J
o .
MfMf
* t 0 0 o
• 0 • " * • 3
B •
0 0 0 0 0 0 0
"ill III
CM m in o o o o
n to CM in o o o
V i *- CM in o o
CM ' ' ' T ^
rn ro in C9
to CM in o o
«• CM o a
. in o
1
( 1000 -
§ 800 -
"3
3
* 600 •
^g
5
9 400 -
00
| 200
a
0.
f ^
: \
B
^ 0
1
|_
b^
N
M
c
2
d
250
f 200-
3
>
5 150
*
a 100
f 50
•3
a
0
t *
r f i
. A T j,
^ T * *
I I I I
CM n in o
n e CM in
v t *• N
CM ' '
ro w in
(O N
^™
I
/"
1 250
•£ 200
3
•o 150-
d
m 100
1 50 •
a
n .
A
I
I
*
f*
i*
*
i i
e o
o a
in o
i
0 '
in o
CM 0
in
J
t
I
*
I -
e E
e i
o *.
*? o
I l^
O 7
0 _
o .E
fc
^
M
^ c
I I
Figure 3 . Activities as a function of the grain size for the soil sample
Harchairuz *91.2 .0-10 cm depth. Error bars are +/-2 sigma.
O : Th-234 . • : Ra-226 . • : Pb-214 . D : Pb-210
X : ftc-228 . A : Cs-137 , A : Cs-134 x 10 . + : K-40 .
The Cs-134 activity in the light grain part is below the
detection limit of 4 Bq/kg dry weight.
-------
Ra-226 activity
I Bqflcg dry weight 1
< Error approx. +/- 15 %
anticlinal
axis
1450m
above
sea level
10m
E
e
o
in
a>
500 m
b>
Tele de Ran
Marchairuz Chasseron Chassera
(390)
1680m
@@ # @ @ ©
1450m 1610m 1425m 1550m
1683 m Altitude aboue sea leuel
120km
c>
fl
Figure 4 . Spatial variation oF the Ra-226 activity in soil samples from the
Western Swiss Jura Mountains. Soils have been dried bul
otherwise untreated, a) Region oF Ht. Tendre, 1680 m above
sea level. b> Region oF Marchairuz. all sampling points within
+/-20 m at 1360 m above sea level, c) Samples taken near the
top oFthe mountains, in small depressions. The values shown are
the highest Found in several samples oFthe respective location.
-------
many soils in the Western Jura Mountains contain aeolian material from
this period, but the origin and the original composition of this
windblown material are unknown (4).
The activities in the soil samples from even nearby points can vary
considerably. Even on the 10 m scale spatial variations can be a factor of
2 as can be seen in figure 4. This figure also shows the variations on the
medium (km) and the large (100 km) scale.
The anomaly extends over a strip of at least 100 km lenght and some
km width. There is a tendency to higher values towards the south-west.
Soil samples from farther west would be very interesting. Unfortunately
"farther West" lies in France and until know it has been impossible to get
soil samples from this part of the world.
It would be a really big surprise if the radium anomaly and the
extreme disequilibrium in the 23au series in the Western Swiss Jura
Mountains would be unique. There are similar geologic and climatic
conditions all over the world. Have a closer look at your soils !
ACKNOWLEDGEMENTS
I would like to thank Giovanni Ferreri, Christophe Murith and Georges
Filler for their help to sample and prepare the soils and H.R. von Gunten
for the alpha spectrometry data.
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. Lauffenburger, T. and Auf der Maur, A., The Concentration
of Radon in a Town where Radium-Activated Paints were
used. In; Proceedings of the 6 th Int. Congress of IRPA,
Berlin (West), May 1984
2. Surbeck, H., The Search for Radon Sources, a Multi-
disciplinary Task.
Rad.Prot.Dosim.,24,1/4 (1988) 431-434.
-------
3. Surbeck, H. and Filler, 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, 1988, Environmental Protection
Agency Report EPA-600/9-89-006, Research Triangle Park, 1989
4. Pochon, M. Origine et evolution des sols du Haut-Jura
suisse. Memoires de la Societe Helvetique des Sciences
Naturelles. Vol.XC. 1976.
5. Surbeck, H., Radium und Radon im Boden, messtechnische und
geologische Aspekte, In: Tagungsbericht Kolloquium Messung
von Radon und Radon-Folgeprodukten, May 6-7, Berlin, FS-
91-56T, Fachverband fuer Strahlenschutz, 1991, ISBN 3-8249-0028-9.
6. Surbeck, H., The Search for Radon Sources, a Multi-
disciplinary Task, In : Proc. of the 2nd Workshop on Radon
Monitoring in Radioprotection, Environmental and/or Earth
Sciences, November 25 - December 6, Trieste, Italy, in press.
7. Tanner A.B., Radon Migration in the Ground:a Supplementary
Review. In : Gesell, T.F. and Lowder, W.M. (eds.),
Proc.Nat.Rad.Env.Ill, 5, Conf-780422, US Dept. of Commerce,
Nat.Techn.Inf.Serv., Springfield,VA, 1980
8. Fleischer, R.L., Moisture and Rn-222 Emanation,
Health Phys. 52/6, 797-799, 1987
9. Surbeck, H. and volkle, H., Radionuclide Content vs. Grain Size
in Soil Samples. The Science of the Total Environment 69 (1988)
379-389.
10. Gaiffe, M., Processus pedogenetiques dans le karst jurassien,
analyse de la complexation organo-minerale en ambiance calcique,
Ph.D. thesis, Univ. de Franche-Comte, Besancon, France, 1987,
ISSN 0150-679X.
11. Surbeck, H., Volkle, H., Zeller, W., Radon in Switzerland, ^n :
Proc. of The 1991 Internat. Symposium on Radon and Radon Reduction
Technology, Philadelphia PA, April 2-5, 1991, Environmental
Protection Agency Report EPA-600/9-91-37a-d, National Technical
Information Services (NTIS), Springfield VA, 1992.
-------
VIII-3
RADON POTENTIAL OF THE GLACIATED UPPER MTDWF.ST-
GEOLOGIC AND CLIMATIC CONTROLS ON SPATIAL VARIATION
R. Randall Schumann
U.S. Geological Survey
MS 939 Federal Center, Denver, CO 80225-0046
Many areas of the United States underlain by soils derived from continental
glacial deposits generate elevated indoor radon levels (£ 4 pCi/L). For example,
Iowa (71 percent), North Dakota (63 percent), and Minnesota (46 percent) have
some of the highest percentages of homes with elevated indoor radon levels in the
State/EPA Indoor Radon Survey. Determining the radon potential of glaciated
areas is complicated by several problems: 1) surface radioactivity is generally
uncharacteristically low in glaciated areas and does not appear to correlate well
with indoor radon values; 2) because glaciers redistribute the bedrock they
override and entrain, the composition and physical properties of till soils do not
necessarily reflect those of the underlying bedrock; and 3) where glacial cover is
thin (less than about 10 m), the radon potential may be a complex product of the
glacial cover and the underlying bedrock. Data from field studies in North Dakota
and Minnesota suggest that radium is leached from the near-surface soil
horizons and accumulates deeper in the soil profile, providing a radon source at
depth which is not readily detectable by surface radioactivity. Crushing and
grinding of rocks by glaciers increases the mobility of uranium and radium in the
resulting tills, allowing them to move readily downward through the soil profile
with other mobile ions as the soils are leached. Soil-gas and indoor radon values
in North Dakota are generally higher than those in northern Minnesota. The
higher radon levels in North Dakota are likely due to enhanced emanation and
relatively higher radium concentrations in the clay-rich soils, whereas the
generally higher permeability and deeper weathering profiles of the sandy till
soils in Minnesota allow soil gas to be drawn into structures from a larger source
volume, enhancing indoor radon levels in these areas. Some of the highest indoor
radon levels in North Dakota are associated with deposits of glacial Lake Agassiz
and other glaciolacustrine deposits. In contrast, glacial lake deposits in
Wisconsin and Michigan are typically associated with low radon levels.
Differences in source-rock composition and soil moisture conditions are likely
responsible for this dissimilarity.
-------
VIII-4
Title: EPA's National Radon Potential Map
Author: Sharon Wirth, U. S. EPA, Office of Radiation Programs
This paper was not received in time to be included in the preprints, and
the abstract was not available. Please check your registration packet for a
complete copy of the paper.
-------
Session IX
Radon Surveys
-------
ABSTRACT
Comparing the National and State/EPA
Residential Radon Survevs
by: Jeffrey L. Phillips
Frank Marcinowski
U.S. Environmental Protection Agency
Office of Radiation Programs
Washington, D.C. 20460
The purpose of this paper is to examine the scope and
findings of two extensive EPA radon surveys, the National
Residential Radon Survey (NRRS), and the State/EPA Residential
Radon Survey (SRRS). Both surveys provide estimates of radon
distribution across the U.S. while utilizing different survey
designs and concentrating on different units of analysis. This
paper will attempt to provide a comparative account which
outlines both complimentary and deviant characteristics of each
data set.
Statistical parameters, survey goals and design, survey
implementation, and survey analysis will provide points of
discussion and comparison.
The NRRS was initiated in 1989 and completed in 1991. The
SRRS was initiated in 1986 and is still ongoing, with a total of
40 States participating to date.
-------
FINAL REPORT
RADON TESTING IN NORTH DAKOTA
DAY CARE FACILITIES
JUNE 1992
Prepared By:
Arlen L. Jacobson
Radiation Control Program
Division of Environmental Engineering
North Dakota State Department of Health
and Consolidated Laboratories
-------
ABSTRACT
The North Dakota State Department of Health and Consolidated
Laboratories' survey of day care facilities in North Dakota has
gathered data that will be useful in the mitigation of radon in
those facilities. While many of the county, city, and facility
arithmetic means were less than 4 picocuries per liter, a large
number were above this level. The data will enable the Department
to map the concentrations in day care facilities and prove useful
in predicting radon "hot spots."
INTRODUCTION
Background
During the second and third years of a three-year grant from
the U.S. Environmental Protection Agency (EPA), the North Dakota
State Department of Health and Consolidated Laboratories (NDSDHCL)
began testing for radon in 2,000 licensed day care facilities.
The day care facilities are licensed by the North Dakota
Department of Human Services.
During January 1991, the Department mailed 2,019 application
packets to day care operators. A second mailing was sent to
1,596 day care facilities in April 1991. Of the 2,000 day care
facilities solicited, 411 facilities responded.
Sampling
The Department followed EPA guidance for sampling protocol.
Every third facility was sent a duplicate and every seventh
facility received a field blank.
Determination of Sampling Locations
1. At least one detector was placed on each level of the house.
2. After locating one detector on each level, the locations are
prioritized (based upon room occupancy).
Highest Priority Rooms: living room, recreation room, dining
room, and family room.
Lower Priority Rooms: bedroom, sleeping room, TV room, toy
room, kitchen, entry, etc.
Not tested: cloak room, bathroom, pantry, furnace room, etc.
-------
The Department followed EPA guidance to formulate its QA Plan
and used the National Database format for recording collected
data.
Precision
Precision was determined by utilizing a coefficient of
variation computer program.
Accuracy
Accuracy was determined from a series of samples that were
exposed in radon chambers in laboratories at the Office of
Radiation Programs - National Air and Radiation Environmental
Laboratory/ Las Vegas, Nevada facility or Montgomery, Alabama
facility. In all, 24 sampling vials were sent to Las Vegas and
181 sampling vials to Montgomery.
Data was verified by use of a QA computer program.
Type of Sample
A "Type of Sample" field was added to record the sample type.
Selections for this field were from the following:
1. NR A normal or regular sample.
2. FB A field blank that is to remain closed in a room that is
not being tested.
3. OD Original duplicate.
4. CD Complimentary duplicate.
5. AC Accuracy check, a sample that has been placed in a
chamber of known radon concentration.
PARAMETERS
Geological Regions
Table 1 contains the arithmetic and geometric means for
selected day care parameters. The geological region with the most
samples had the lowest average radon concentration.
Basement Sump
Many buildings are built on flood plains and a large number
have sumps in basements. Sumps were especially prevalent in
cities built near rivers. It has been noted previously, by the
-------
Department, that basements with sumps generally have higher radon
concentrations. The radon concentration can usually be lowered by
covering the sump. A surprising 440 samples, about 26%, were from
homes containing sumps.
Mobile Homes
Of 545 day care facilities, 29 (5.32%) were mobile homes.
Mobile home day care facilities were given one sample to be placed
in the living room unless the facility was scheduled for a
duplicate sample or field blank sample. Only 2.79% of the total
samples were from mobile homes. The average radon concentration
for mobile homes was near 4 pCi/1.
Heat Distribution
An "other" category was added to the Federal Database for
heat distribution systems to include combination systems and wood
burners that are only used during the winter heating season.
Forced air was the most popular type of heat distribution system
(72.12%). Individual heating units should be further
investigated.
Test Location Floor
The most popular sampling floors were the first floor
(61.63%) and the basement (31.49%).
Type of Facility
The facility breakdown is as follows. The family type of
facility (64.29%), Group (28.29%), Center (5.28%), In Home (0%),
Multiple (1.01%), Half-Day (0.65%), Pre-School (0.12%), and Public
Approval (0.36%). Note, the family and group facilities are more
than 90% of the samples.
Basement, Crawl Space, or Slab-on-Grade
The basement, crawl space, and slab-on-grade building styles
can be found in combinations in some older homes. Older homes
generally have crawl spaces for small houses and basements for
larger houses. Some houses are a combination of all three types.
Slab-on-grade is not a popular building style in North Dakota,
comprising only 3.68% of the samples.
Type of Building
Single family buildings comprise 85.05% of the samples.
Following Federal guidelines, the Department categorized the
buildings such that most buildings were contained in the first
-------
four types. The most popular other type of building selected was
churches.
Room Type
The relatively large average radon concentration of bedrooms
prompted the following investigation:
Arith. Mean
Radon Cone. % of % of % of % of
Room Type in pCi/1 Samples Basement 1st Floor 2nd Floor
Bedroom 5.10 29.89 26.19 58.33 4.63
Family Room 5.61 12.63 66.20 31.46 2.35
Living Room 4.57 20.88 6.25 91.48 2.27
Other Room 5.31 11.63 22.96 72.96 4.08
The remaining room types all contained less than 5% of the
samples. The large percentages of bedrooms and family rooms in
the basement is a contributing factor to their high radon
concentration.
Maximum Number of Children/Type of Facility
The most prevalent maximum number of children found in day
care licenses are 7 (59.31%), followed by 12 (13.05%), and 6
(5.22%). The maximum number of children licensed for a center is
140 children. Arithmetic and geometric means are contained in
Table 2.
Testing Company
Two different radon laboratories were used to analyze day
care samples. The first laboratory analyzed 1054 samples and the
second company analyzed 1568 samples. Differences between
companies are explained in the Variance section.
Two Testing Seasons
Two different radon testing seasons were involved,
October 1990 through April 1991 and October 1991 through
April 1992. A case can be made concerning the two different
seasons, yet a case can also be made about differences within a
season. In North Dakota it makes a difference whether one samples
in October, April, or January.
Figure 1, "North Dakota County Data" contains the essential
arithmetic means, standard deviation, number of samples,
geological region, and weather station. For county data see
Table 3 and for city data see Table 4.
-------
QUALITY ASSURANCE
Field Blanks
Nine field blanks with the accuracy checks were sent to the
second testing company. The results were 5 samples with 0.0 pCi/1
and 4 samples with 0.1 pCi/1 radon concentration. Of the 93 field
blanks sent to day care facilities most were 0.1 pCi/1, however,
some were greater than 1 pCi/1. It is assumed the day care
operator opened the field blank for those greater than 1 pCi/1.
Accuracy
The first company's mean accuracy for a 5.9 pCi/1 standard
was 72.7% with a standard deviation of 4.73. Twelve samples of
11.9 pCi/1 concentration averaged 93.5% with a standard deviation
of 13.44.
The second company averaged 110.7% (Std. Dev. 12.37) for
85 samples for a standard of 5.4+/-0.4 pCi/1. The other standard,
12.8+/-0.8 pCi/1 averaged 110.5% with standard deviation of 4.86.
Coefficient of Variation
Eighty-one duplicate pairs were analyzed during the first
radon season. The coefficient of variation arithmetic mean was
16.74% with a standard deviation of 29.61.
During the second radon season, the average coefficient of
variation was 9.49%, standard deviation 11.42.
VARIANCE
Utilizing S.A.S. PROC NESTED software, the variance of the
parameters were calculated. The results follow:
-------
PARAMETER
County
City
Building
Geological Regions
Basement Sump
Mobile Homes
Heat Distribution
% OF TOTAL
VARIANCE
21.5
23.3
90.6
6.2
9.2
1.2
3.1
PARAMETER
Test LOG. Floor
Type of Facility
Bsmt Crawl Space
Slab-on-Grade
Type of Building
Type of Room
Maximum Children
Maximum Children
Type of Facility
% OF TOTAL
VARIANCE
5.2
1.90
1.5
5.8
0.8
4.7
5.6
Testing companies
program.
were separated by a PROC UNIVARIATE
Initial Testing
Company
Final Testing
Company
N
593
1093
ARITH.
MEAN
4.57
5.63
ARITH.
STD.DEV.
6.05
7.29
LOG 10
MEAN
2.83
3.63
LOG 10
STD.DEV.
2.73
2.50
CROSS VARIANCE IN % OF TOTAL VARIANCE
Maximum Children 0
Type of Facility 5.62
City 23.28
Geological Region 0
County 16.06
Building 74.69
Basement Sump 0
County 16.06
Geological Region 0
Building 74.69
City 10.20
Building 80.46
Basement Sump 0
Building
Geological Region
City
Basement Sump
County
City
Basement Sump
City
Geological Region
Building
80.46
0
0
50.44
12.29
0
44.24
10.20
0
80.46
-------
There are 47 counties, 104 cities, 545 buildings,
3 geological regions, and Yes or No indicating whether or not the
facility contains a basement sump. Major contributors to variance
are the county, the city, the building, and the sump.
WEATHER ANOMALIES AND RAINFALL
The parameters the Department uses from the three national
weather stations at Bismarck, Fargo, and Williston are a departure
from the 30-year temperature average, inches of rainfall, inches
of snow, average wind speed, percent of possible sunshine, and
weather conditions, e.g., hail, fog, rain, snow, etc.
Figure 2 indicates the data is log normal. The results as
computed by the S.A.S. PROC UNIVARIATE program are:
N 1686
Mean 3.32
Std.Dev. 2.60
W: Normal 0.98441
Median 3.30
SUMMARY AND CONCLUSIONS
Conducting a survey using the National DataBase is labor
intensive. Often the instructions to testing applicants are
either not understood or not followed. In future surveys, the
Department should conduct organizational meetings in which the
applications can be completed and questions answered.
More effort and expertise is needed in the area of the
mapping of radon concentrations.
This survey shows day care facilities have less of a radon
concentration than do homes. The problem is that both average
concentrations in North Dakota are greater than 4 pCi/1, day care
facilities (5.18) and homes (7.00).
The Department should continue to work with day care
operators to help them mitigate their radon problems.
-------
Table 1
DAT CARE
Arithnetic and Geometric Means for Selected Par*
iters
Arithmetic and Geometric Me
Parameters
% Of
Samples
ans for TO and OP Samples in pCi/1
Arithmetic
Mean
Mean
Std.Dev.
Geometric
Mean
Mean
Std.Dev.
Geological Region
1
2
3
All
11.33
35.94
52.73
100.00
5.24
6.99
4.08
5.26
8.34
8.87
4.26
6.89
3.14
4.26
2.84
3.32
Basement Sump
Yes | 26.10
No | 73.90
7.52
4.46
9.36
5.56
4.55
2.97
2.51
2.77
2.41
2.60
2.74
2.48
Mobile Home
Yes
NO
Heat Distribution
Hot Air
Hot Hater
Radiant
Individual Units
Other
2.79
97.21
72.12
14.18
4.21
6.88
2.49
3.81
5.30
4.96
6.94
5.43
3.64
4.18
7.59
4.76
6.82
4.08
3.27
11.77
5.70
2.02
3.37
3.50
2.32
2.81
4.31
3.41
2.96
2.58
2.54
2.61
2.71
2.79
2.08
Test Location Floor
Basement
First
Second and Above
31.49
61.63
6.88
6.91
4.62
3.42
8.07
6.36
3.32
4.65
2.89
2.46
2.46
2.59
2.26
Type of Facility
Family License
In Home License
Multiple
Half-Day
Preschool
64.29
1.01
0.65
0.12
4.96
3.40
3.72
5.15
5.45
2.80
2.27
1.95
3.34
2.01
2.85
4.77
2.46
3.22
3.23
1.49
-------
ArithMtic and G«<
Table 1
OAT GARB
»tric Means for Selected Pan
(Continued)
iters
Arithmetic and Q«<
^— ^^— — ~— — —
Parameters
aaetric M<
^^SSS^S^^^^^^^S
% of
samoles
tans for
.
MR and "" fl*nples in pci/l
Arithmetic
Mean
Mean
Type of Facility (Cont'd)
center
Group
Public Approval
Type of Basement
Basement
Crawl Space
Slab on Grade
other
5.28
28.29
0.36
17.16
11.51
3.68
7.06
3.79
6.37
1.05
5.54
4.65
4.69
3.44
Std.Dev.
Geometric
Mean
1
1 Std.Dev.
I
5.30
9.62
0.84
2.15
3.69
0.75
6.66
9.27
7.35
3.39
3.63
2.50
2.54
2.32
2.67
2.76
2.30
2.52
2.75
2.82
2.56
Type of Building
single Family
Multi-Family
Business
School
Other
85.05
4.86
3.32
2.43
4.33
5.64
3.03
3.84
3.01
2.52
7.26
1.75
5.70
4.39
1.87
2.54
2.53
1.76
1.78
1.87
3.63
1.89
3.27
2.71
2.32
Room Type
Bedroom
Family Room
Living Room
unfinished
Basement
Office
Classroom
other
Finished Basement
Kitchen
Recreation Room
Dining Room
29.89
12.63
20.88
3.44
0.24
4.98
11.63
1.48
3.02
2.49
2.49
5.10
5.61
4.57
6.85
5.68
3.29
5.31
7.13
5.37
7.00
5.31
6.60
6.20
7.22
6.17
4.46
4.12
6.99
4.73
8.82
6.36
7.65
3.33
3.92
2.74
5.11
3.50
2.03
3.37
4.92
3.21
4.90
3.20
2.50
2.34
2.71
2.09
3.03
2.67
2.49
2.99
2.46
2.39
2.60
10
-------
Arithaetic and Ge<
Table 1
DAT CARE
trie Mean* for Selected Par.
(Continued)
Arithmetic and Geometric Means for MR and OD 8
Parameters
% Of
Samples
Arithmetic
Mean
Mean
Std.Dev.
aaples in pCi/1
~^ —
Geometric
Mean
Mean
Std . Dev .
Room Type (Cont'd)
Playroom
TV Room
Sleeping Room
Nursery
3.91
0.36
1.72
0.77
7.01
4.80
6.45
8.02
7.57
3.29
7.34
15.16
4.46
3.67
3.91
2.84
2.62
2.14
2.65
4.09
11
-------
Table 2
DAT CARE
Maximm Number of Children and Type of Facility
Arithmetic and Geometric Means for HR and OD S»«ple« in DCi/1
Maximum Number of
children/Type of Facility
6 Family
7 Family
8 Group
9 Group
10 Group
11 Family
1 1 Group
12 Group
13 Group
14 Family
14 Group
14 Half-Day
15 Group
16 Pre-School
1 6 Group
1 7 Group
18 Group
18 Half -Day
24 Pre-School
30 Public Approval
32 Center
35 Center
45 center
5 0 center
55 Center
56 Center
58 center
% Of
samples
5.22
59.31
1.01
1.42
0.71
0.30
1.60
13.05
0.65
0.12
2.08
0.12
1.84
0.06
0.53
0.18
4.21
4.74
0.06
0.12
0.06
0.47
0.24
0.95
0.30
0.12
0.30
Arithmetic
Mean
Mean
3.75
5.05
4.71
5.33
2.78
8.68
3.81
7.04
6.14
4.10
4.77
0.85
6.06
3.20
17.63
6.97
4.87
4.36
7.10
2.20
3.20
7.11
1.95
2.48
1.04
12.90
1.56
Std . Dev .
2.95
5.63
2.47
3.84
1.88
1.27
5.47
9.55
4.36
1.10
3.66
0.25
4.72
0
30.48
12.46
0.80
2.01
0
0.20
0
5.16
0.15
0.77
0.37
12.40
0.37
Geometric
Mean
Mean
2.62
3.39
4.09
3.94
2.17
8.59
2.21
4.26
4.61
3.94
3.56
0.81
3.57
3.20
3.34
3.31
4.84
3.77
7.10
2.19
3.20
3.53
1.94
2.38
0.96
3.56
1.51
Std . Dev .
2.53
2.45
1.71
2.23
2.03
1.15
2.60
2.74
2.20
1.32
2.29
1.35
3.57
0
5.27
2.96
1.12
1.78
0
1.10
1
4.35
1.08
1.32
1.51
7.11
1.32
12
-------
Table 2
DAT CJUtB
of Children and Type of Facility
(Continued)
AriJ-haetie and Geometric MMP« for HR and OD Savple* in DCi/1
Maximum Number of
children/Tvce of Facility
60 Multiple License
61 Multiple License
64 Multiple License
65 Center
75 center
80 Center
83 Multiple License
85 Center
95 Center
108 Center
116 center
120 Center
123 Center
135 center
140 Center
% Of
Samples
0.59
0.12
0.30
0.30
0.12
0.71
0.12
0.24
0.30
0.30
0.06
0.24
0.18
0.24
0.30
Arithmetic
Mean
Mean
1.35
2.75
0.40
1.68
3.80
15.42
6.50
1.05
2.64
3.08
0.70
3.28
7.73
1.20
1.12
Std.Dev.
0.05
0.35
0.06
0.19
0
8.42
0.30
0.18
0.77
0.55
0
0.50
6.58
0.19
0.24
Geometric
Mean
Mean
1.35
2.73
0.39
1.67
3.80
11.69
6.49
1.03
2.53
3.03
0.70
3.24
4.90
1.18
1.09
Std.Dev.
1.04
1.14
1.18
1.12
1
2.35
1.05
1.19
1.34
1.20
1
1.15
2.76
1.18
1.27
13
-------
Ttbl* 3
COUNTY
DAYCARE AND 1988 STATE STUDY
RNCOWCinBD PERCE8TGT4 PCTGT20 RNCONC1998 PCTCT41988 PCTGT2088
ADAMS
BARNES
BENSON
BILLINGS
BOTTINEAU
BOWMAN
BURKE
BURLEIGH
CASS
CAVALIER
DICKEY
DIVIDE
DUNN
EDDY
ENMONS
FOSTER
GOLDEN VALLEY
GRAND FORKS
GRANT
GRIGGS
HETTINGER
KIDDER
LAMOURE
LOGAN
MCHENRY
MCINTOSH
MCKENZIE
MCLEAN
MERCER
MORTON
MOUMTRAIL
NELSON
OLIVER
PEMBINA
PIERCE
RAMSEY
RANSOM
RENVILLE
HIGHLAND
ROLETTE
SARGENT
SHERIDAN
SIOUX
SLOPE
STARK
STEELE
STUTSMAM
TOWKER
TRAILL
WALSH
WARD
WELLS
WILLIAMS
ALLCOUNTIES
13.30
8.29
0.00
0.00
3.58
20.62
2.27
3.76
7.37
3.95
4.32
3.89
2.23
3.63
3.38
0.00
1.81
10.98
9.05
3.09
2.44
3.02
3.97
2.60
4.94
6.26
2.70
3.84
3.44
3.70
3. 10
3.20
0.00
4.23
2.43
3.06
10.38
3.03
1.82
2.40
3.00
0.00
0.00
0.00
3.41
6.87
4.12
6.33
6.30
13.33
3.23
2.79
2.44
3.18
83.33
73.76
0.00
0.00
41.67
76.92
0.00
29.15
62.91
45.83
60.00
25.00
0.00
30.00
31.58
0.00
12.50
83.61
100.00
37.03
22.22
20.00
44.44
7.14
33.71
80.00
0.00
16.67
.23.00
28.73
29.17
0.00
0.00
25.00
10.00
37.50
78.57
33.33
94,34
3.00
40.00
0.00
0.00
0.00
27.69
100.00
49.26
100.00
51.72
68.42
22.05
33.71
13.04
41.40
16.67
13.13
0.00
0.00
0.00
38.46
0.00
0.95
4.69
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
10.07
0.00
0.00
0.00
0.00
0.00
0.00
7.14
0.00
0.00
0.00
0.00
1.25
4.17
0.00
0.00
0.00
0.00
0.00
7.14
0.00
0.00
0.00
20.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
31.38
0.00
0.00
0.00
2.97
9.30
8.00
7.20
8.70
6.00
0.70
2.<0
4.90
7.90
3.70
3.50
8.80
8.70
4.60
6.60
3.70
4.00
1.70
8.10
3.30
7.20
4.00
5.20
3.70
3.60
7.10
3.50
.00
.10
.60
.30
.30
.40
.40
.80
.60
.40
.10
.00
.50
.50
.00
.90
.20
.00
.20
.80
.10
.80
0.50
4.20
3.60
4.60
7.00
78.30
78.90
62.30
77.80
60.60
71.00
0.00
48.50
69.00
42.90
81.80
100.00
75.00
50.00
CO. 00
57.10
42.90
90.10
60.90
20.00
61.30
50.00
60.00
76.90
33.30
44.40
50.00
70.60
73.90
53.50
60.00
69.30
84.20
72.90
29.40
77.80
57.10
66.70
60.90
50.00
70.00
60.00
50.00
42.90
60.10
42.90
65.00
50.00
65.40
75.50
45.50
27.30
47.10
63.00
13.00
7.90
12.50
11.10
3.00
6.50
0.00
1.00
1.00
0.00
0.00
0.00
10.70
0.00
6.70
0.00
0.00
11.00
4.30
0.00
6.50
0.00
0.00
0.00
0.00
11.10
0.00
0.00
6.50
3.00
10.00
0.00
0.00
13.60
0.00
0.00
0.00
0.00
4.30
0.00
0.00
0.00
0.00
0.00
4.90
0.00
2.30
10.00
3.80
10.20
1.50
9.10
0.00
4.00
Legend
RNCONCNROD - Arithmetic mean of day care radon
concentrations
RNCONC1988 - Arithmetic mean of the 1988 State Study
radon concentrations
PERCENTGT20 - Percent of day care samples
greater than 4 pCi/1 and 20 pCi/1,
respectively.
PCTGT41988, PCTGT2088 - Percent of 1988 State Study
samples greater than 4 pCi/1 and
20 pCi/1, respectively.
PERCENTGT4,
14
-------
Table 4
Day Care Arithmetic and Geometric Means
and Standard Deviations
COUNTY
ADAMS
BARNES
BENSON
BILLINGS
BOTTIMEAU
BOWMAN
BURKE
BURLEIGH
CASS
CAVALIER
DICKEY
DIVIDE
DUNN
EDDY
EMMONS
FOSTER
GOLDEN VALLEY
GRAND FORKS
CITY
HETTINGER
LITCHVILLE
VALLEY CITY
BOTTINEAU
BOWMAN
SCRANTON
BOWBELLS
LIGNITE
BALDWIN
BISMARCK
LINCOLN
MENCKEN
WING
CASSELTON
FARGO
HARWOOD
KINDRED
MAPLETON
WEST FARGO
WHEATLAND
LANGDON
ELLENDALE
CROSBY
KILLDEER
NEW ROCKFORD
HAZ ELTON
LINTON
BEACH
EMERADO
GRAND FORKS
LARIMORE
THOMPSON
N
6
6
33
4
29
12
12
13
10
3
7
5
2
422
3
383
26
3
4
213
5
134
11
5
3
53
2
24
24
5
5
8
8
8
8
4
4
19
4
15
8
8
139
3
132
1
3
ARITH
MEAN
13.30
13.30
8.80
6.33
9.14
0.00
0.00
3.35
3.35
22.74
27.15
8.03
2.27
2.28
2.25
3.72
2.37
3.77
2.13
6.43
8.40
7.58
9.70
7.72
5.78
5.22
1.53
8.04
5.60
3.80
3.80
4.32
4.32
3.84
3.84
2.13
2.13
3.03
3.03
3.55
1.85
4.01
0.00
1.75
1.75
11.38
7.73
11.63
0.90
7.43
AMSTD
DEV
7.42
7.42
7.90
0.65
8.37
0.00
0.00
1.81
1.81
22.37
23.80
0.26
0.24
0.28
0.05
3.64
0.19
3.68
1.31
0.56
5.97
9.89
3.01
8.65
5.80
0.27
0.54
13.84
1.40
2.37
2.37
2.37
2.37
0.64
0.64
0.73
0.73
1.78
1.78
1.95
1.48
1.80
0.00
1.89
1.89
10.97
6.58
11.15
0.00
0.25
GEO
MEAN
9.91
9.91
6.12
6.29
6.10
0.00
0.00
2.86
2.86
12.28
13.95
8.03
2.26
2.26
2.25
2.71
2.36
2.76
1.78
6.41
6.01
4.61
9.26
4.83
2.73
5.21
1.45
4.49
5.42
2.73
2.73
3.50
3.50
3.79
3.79
2.01
2.01
2.45
2.45
3.01
1.44
3.66
0.00
1.28
1.28
8.41
4.90
8.69
0.90
7.43
GEOSTD
DEV
2.53
2.53
2.37
1.10
2.51
0.00
0.00
1.77
1.77
3.14
3.59
1.03
1.11
1.13
1.02
2.23
1.09
2.23
1.86
1.09
2.37
2.80
1.39
2.77
4.55
1.05
1.39
2.66
1.29
2.66
2.66
2.01
2.01
1.17
1.17
1.38
1.38
1.96
1.96
1.84
1.93
1.50
0.00
1.96
1.96
2.22
2.76
2.17
1.00
1.03
15
-------
Table 4
Day Care Arithmetic and Geometric Means
and Standard Deviations
COUNTY
GRANT
GRIGGS
HETTINGER
KIDDER
LAMOURE
LOGAN
MCHENRY
MCINTOSH
MCKENZXE
MCLEAN
MERCER
MORTON
MOUNTRAIL
NELSON
OLIVER
PEMBINA
PIERCE
CITY
N
ARITH AMSTD
MEAN DEV
ELGIN
COOPERSTOWN
HANNAFORO
NEW ENGLAND
MOTT
REGENT
STEELE
EDGELEY
LAMOURE
CACKLE
NAPOLEON
TOWNER
UP HAM
VELVA
ASHLEY
WISHER
WATFORD CITY
GARRISON
MAX
TURTLE LAKE
WASHBURN
BEULAH
STANTON
HEBRON
MANDAN
NEW SALEM
PARSHALL
PLAZA
NEW TOWN
STANLEY
LAKOTA
CAVALIER
ST. THOMAS
2
2
27
23
4
9
3
3
3
5
5
9
5
4
14
5
9
14
6
4
4
20
4
16
10
10
24
7
3
8
6
8
5
3
80
2
74
4
24
4
3
5
12
1
1
16
12
4
10
9.05
9.05
3.31
2.70
6.78
2.54
1.70
4.83
1.10
3.02
3.02
4.07
1.08
7.80
2.64
2.52
2.70
4.94
7.83
2.50
3.03
6.16
6.80
5.99
2.79
2.79
3.60
1.60
0.90
7.15
2.55
3.45
4.26
2.10
3.70
4.95
3.67
3.58
4.90
1.95
19.73
3.44
2.79
3.20
3.20
0.00
4.23
4.76
2.65
2.45
0.75
0.75
2.02
1.38
1.55
1.92
1.24
1.09
0.57
0.92
0.92
3.35
0.19
0.47
0.93
0.37
1.12
5.77
7.82
0.07
1.56
2.72
0.57
3.01
0.72
0.72
3.75
1.06
0.00
4.60
0.73
1.38
1.14
0.14
3.17
1.05
3.29
0.13
5.95
1.57
2.95
0.46
2.12
0.00
0.00
0.00
4.11
4.63
0.32
2.57
GEO GEOSTD
MEAN DEV
9,02
9.02
2.74
2.35
6.60
1.80
1.27
4.72
0.98
2.87
2.87
2.58
1.06
7.79
2.50
2.49
2.51
3.38
4.91
2.50
2.60
4.69
6.78
4.28
2.70
2.70
2.40
1.29
0.90
5.89
2.43
3.20
4.12
2.10
2.87
4.84
2.79
3.57
2.69
1.17
19.53
3.41
1.96
3.20
3.20
0.00
2.77
2.82
2.63
1.68
1.09
1.09
1.89
1.74
1.26
2.40
2.19
1.25
1.60
1.39
1.39
2.72
1.20
1.06
1.36
1.15
1.45
2.17
2.64
1.03
1.75
2.72
1.09
3.00
1.30
1.30
2.41
1.92
1.00
1.83
1.37
1.47
1.30
1.07
2.07
1.24
2.11
1.04
3.16
3.20
1.15
1.14
2.51
1.00
1.00
0.00
2.44
2.79
1.13
2.38
16
-------
Table 4
Day Care Arithmetic and Geometric Means
and Standard Deviations
COUNTY
RAMSEY
RANSOM
RENVILLE
HIGHLAND
ROLETTE
SARGENT
SHERIDAN
SIOUX
SLOPE
STARK
STUTSMAN
STEELE
TOWNER
TRAILL
WALSH
WARD
CITY
RUGBY
WOLFORD
DEVILS LAKE
STARKWEATHER
ENDERLIN
LISBON
GLENBURN
BARNEY
COLFAX
HANKINSON
WAHPETON
WYNDMERE
DUNSEITH
ROLETTE
ROLLA
ST. JOHN
FORMAN
KILNOR
BELFIELD
DICKINSON
SOUTH HEART
JAMESTOWN
MONTPELIER
STREETER
HOPE
CANDO
BUXTON
HATTON
HILLS BORO
HAYVILLE
PORTLAND
GRAFTON
BERTHOLD
MI NOT
N
7
3
8
5
3
14
3
11
9
9
53
1
6
4
38
4
20
5
9
2
4
5
1
4
65
3
57
5
136
129
4
3
3
3
3
3
29
3
5
6
6
9
19
19
68
1
67
ARITH
MEAN
1.97
3.57
3.80
4.26
3.03
10.99
1.80
13.49
3.14
3.14
1.83
0.10
0.98
0.80
1.99
3.05
2.40
3.50
2.32
3.80
0.50
7.50
25.50
3.00
0.00
0.00
0.00
3.41
4.17
3.32
3.98
4.18
3.97
8.43
7.70
6.87
6.87
6.13
6.13
6.46
2.70
8.58
4.02
8.57
6.77
14.66
14.66
3.24
2.90
3.25
AMSTD
DEV
0.22
4.48
2.17
2.54
0.93
6.37
0.80
4.71
1.41
1.41
1.40
0.00
0.38
0.07
1.46
0.82
1.44
1.70
0.53
0.00
0.07
9.22
0.00
2.23
0.00
0.00
0.00
2.61
0.26
2.65
2.77
2.48
2.33
2.11
0.94
3.43
3.43
1.25
1.25
3.47
0.22
3.91
0.91
3.84
2.37
13.15
13.15
2.88
0.00
2.90
GEO
MEAN
1.96
1.17
3.36
3.66
2.90
8.20
1.63
12.74
2.69
2.69
1.36
0.10
0.93
0.80
1.52
2.95
1.91
3.21
2.25
3.80
0.49
3.82
25.50
2.38
0.00
0.00
0.00
2.60
4.16
2.52
2.72
3.27
3.11
8.18
7.64
6.13
6.13
6.02
6.02
5.58
2.69
7.58
3.93
7.45
6.25
9.29
9.29
2.28
2.90
2.28
GEOSTD
DEV
1.13
4.54
1.60
1.69
1.35
2.50
1.55
1.40
1.87
1.87
2.24
1.00
1.38
1.09
2.14
1.28
2.11
1.47
1.30
1,00
1.15
3.04
1.00
1.91
0.00
0.00
0.00
2.15
1.07
2.14
2.63
2.28
2.27
1.27
1.13
1.59
1.59
1.21
1.21
1.72
1.09
1.68
1.23
1.77
1.52
2.67
2.67
2.43
1.00
2.44
17
-------
Table 4
Day Care Arithmetic and Geometric 'leans
and Standard Deviations
COUNTY
WELLS
WILLIAMS
CITY
HARVEY
ALAMO
RAY
WILDROSE
WILLISTON
N
14
14
46
4
5
3
34
ARITH
MEAN
2.79
2.79
2.42
4.23
1.62
7.67
1.86
AMSTD
DEV
1.73
1.73
2.32
1.15
1.70
5.35
0.97
GEO
MEAN
2.31
2.31
1.66
4.08
0.78
2.36
1.62
GEOSTD
DEV
1.83
1.83
2.52
1.29
3.50
9.34
1.70
18
-------
riouu 1
North Dtkoti County Data
Top Line: Geolo|K»l Re|ion. Weitber Statioa
Middle Un< Numb«r of Stmptei
Bottom Une Arithmetic Mcu, Standard Devittioa
Daycare Radon Study 1992
100.01
-
-
G
o
5
£
—
G
-
CJ
c
O 1.0
o
c
1
£
: .
—i—
K
—i—
.:
20 30 40 SO 60 70
Percentage Less Than
M
19
-------
Ventilation, Climatology &£ Radon Activity in Four Minnesota Schools
Tim Burkhardt, Earnest Tare, & Laura Oatman
Indoor Air Quality Unit, Minnesota Department of Health
Minneapolis, MN 55414
ABSTRACT
Radon concentrations were measured in two Minnesota schools and analyzed by examining the
relation between HVAC system operation and radon levels; differences between short-term, long-
term, and continuous testing devices; and associations between atmospheric pressure and radon
levels. Multiple monitoring was performed in the schools using charcoal canisters, alpha-track
detectors, and Honeywell continuous monitors. Climatologic data was examined for correlations
to radon concentrations. HVAC operation (on versus off) was found to significantly alter radon
concentration; long-term alpha-track tests were found to give lower results than short-term charcoal
canisters or continuous monitors; atmospheric pressure appeared to be inversely related to radon
concentration.
INTRODUCTION
In early 1989, the U.S. Environmental Protection Agency (EPA) conducted a screening program
to assess radon levels in homes and schools across the country. Sixteen schools were screened in
Minnesota. EPA then assisted the Minnesota Department of Health (MDH) in conducting a
comprehensive study of those schools with high screening levels. The four schools with the highest
screening levels (range: 5.5-10.7 pCi/L) were chosen for an extensive one year study. Selected
results of the national study including the Minnesota schools were reported at the 1991
International Radon Symposium in Philadelphia, PA (1,2,3). This paper presents a further
analysis of selected findings in two of the four Minnesota schools.
* Present address: Health & Environment Digest, Freshwater Foundation, Wayzata, MN.
-------
The two schools are Gideon Pond Elementary in Burnsville, MN (EPA number MNO) and
Sweeney Elementary in Shakopee, MN (EPA number MN4). Gideon has two occupied floors, 12
occupied classrooms in contact with die ground, and is 27 years old; Sweeney has one occupied
floor, 24 occupied classrooms in contact with the ground, and is 23 years old.
Bodi buildings are of slab-on-grade construction with the lowest level partially below grade. First
floor construction is of concrete in both buildings. Utility pipes are located mostly or
completely overhead.
The schools were both equipped with air handlers, univents and radiant heating but differ in the
configurations of their heating, ventilation and air conditioning (HVAC) systems. For example,
some rooms were not ventilated by central air handlers; sizes of fresh air intakes were varied.
Where relevant to radon levels, these differences are discussed below.
METHODS
Four schools were selected for extensive study based on their high screening levels and their
proximity to the Minnesota Department of Health. Detailed methods for the study were
developed by EPA and are given elsewhere (4).
Each classroom in contact with die ground was tested using charcoal canisters, alpha-track
detectors, and Honeywell continuous monitors. The year-long testing period was divided into
four seasons. In each season a nine-day period was selected and divided into diree "rounds" of
about three days each, with different ventilation conditions in effect during each round (TABLE
1). Round 1 was over a weekend with the HVAC system off; Round 2 was during the week under
normal HVAC conditions (on during die day and set back during the night); Round 3 was during
the second weekend with die HVAC system on continuously.
During each round of each season, charcoal canisters were placed in all occupied classrooms that
were in contact widi the ground. In other words, in each season, one charcoal canister was used for
each of three three-day periods (rounds). Alpha-track detectors were placed and collected every
diree months. Continuous monitors were set for 24-hour intervals except during die three rounds
when diey were set for four-hour intervals. Three to four continuous monitors were used in each
school and were placed in different rooms. The continuous monitors did not remain in any one
room for the entire twelve months due to needs of the school and a desire by the researchers to
collect data from as many rooms as possible.
-------
The detectors were provided by the EPA and placed in the schools by MDH with the assistance of
officials from each school. Continuous monitors were provided by Honeywell. Climatologic
data was collected at the Minneapolis/St. Paul International Airport by the National Oceanic and
Atmospheric Administration (NOAA).
RESULTS
Data collection was interrupted during each season and in each school for various reasons such as
temporary power shut down, monitor moved from original location, monitor tampered with,
results illegible. In addition, power shortages during Season 1 in August 1989, prevented
fulfillment of the Round 3 requirement of continuous HYAC operation in two of the four schools
studied. Adherence to the schedule during the other three seasons is not clearly documented and
test results from these periods are not analyzed here.
Results presented here are for Gideon and Sweeney schools, Season 1, Rounds 1-3. Gideon and
Sweeney were selected from the four schools studied for the quality of data available, especially
adherence to the nine-day HVAC schedule discussed above. Season 1 was chosen because good
data were available and because the ground remained unfrozen during the entire testing period
(August through October), allowing for better analysis of precipitation data.
VENTILATION EFFECTS
Effects of changing ventilation pattens on measured radon concentrations can be seen in Figures 1
and 2. For the charcoal canister and continuous monitor data, each column represents a three-day
average for all the rooms tested in that school.
In general, radon levels in both schools were greatest when the HVAC system was off, least when
the system was continuously on, and intermediate when the system was in normal operation (i.e, on
during the school day and set back during the night). Although the absolute radon levels differ, the
same trend is seen whether measured either by charcoal canisters or continuous monitors. On
average, a four- to six-fold decrease in radon concentration was seen with HVAC continuously on
compared to completely off. Long-term alpha-track concentrations are provided for reference
and are approximately equal to the concentrations measured by the short-term testing devices when
the HVAC system was on continuously.
The trend with respect to HVAC status was also seen in a third school (Cleveland Elementary-
Secondary School; MN2) that did not follow the HVAC operating schedule. Here, the HVAC
-------
system was not turned on continuously during Round 3 and tadon concentrations were lowest
during normal operation (Round 2) and elevated in both Round 1 and Round 3. In a fourth
school (Lincoln Elementary School; MN6), however, where the HVAC schedule also was not
followed, the trend was not observed. Specifically, Round 1 concentrations were greater than
Round 2, as expected, but Round 3 concentrations were the lowest of the three when they were
expected to be approximately equal to Round 1 and greater than Round 2.
COMPARISON OF TESTING DEVICES
A comparison of selected results from charcoal canisters, continuous monitors and alpha-track
detectors is given in Figure 3. The long-term alpha-track reading is considerably lower than the
readings from the two short-term measurement devices. In the data shown, charcoal canister
readings were from two to four times greater than the long-term alpha-track reading. Particularly
for Rooms 9 and 11, the continuous monitor data highlights the potential daily changes in radon
concentration in a given classroom.
Results from Round 2 (HVAC normal) were selected for Figure 3 because they represent the
building under normal operating conditions (with students present and warm weather ventilation).
Results from Round 1 (HVAC off) were similar. Results for Round 3 (HVAC on continuously)
were not included in the comparison because the very low radon levels measured are too near
detection limits to assure significance.
All of the readings are specific for each room. Continuous monitor data represents an average of
the six four-hour readings given during a 24-hour period. The charcoal canister reading is a three-
day average and the alpha-track reading is a 70-day average that includes the 9-day test period.
In addition to the differences among the measurement devices, it is notable that radon levels in
different rooms varied from each other considerably, thus emphasizing the importance of testing
every room. It is notable also that rooms 9 and 11 are both basement rooms (a full level below
grade) and room 11 has no windows or univents.
CLIMATOLOGIC EFFECTS
A plot of radon concentration (Gideon Pond Elementary School, Room 1) versus atmospheric
pressure is given in Figure 4. Rainfall is also noted. The trend suggested here supports the
observation that radon levels increase with decreasing pressure and vice versa. The two highest
radon readings in this time period occurred simultaneously with the second- and fourth-lowest
pressure readings (days 16 and 21). Low radon readings occurred on a number of days that also
-------
had some of the highest pressure readings (days 1-3, 8, 11-15, 18-19, 22). While these results do
not explain or make certain the relationship between radon concentration and pressure, they add
positively to the other data collected on this subject.
From the year's worth of survey data collected, this time period (May 27, 1990- June 19,1990)
was selected for the large number of precipitation events and pressure changes that occurred, and
because it is a period when the ground is completely unfrozen. This time period also does not
overlap with the HVAC schedules described in TABLE 1, thus avoiding confounding from the
intentional ventilation changes in the school.
Climatologic data was collected at the Minneapolis/ St.Paul International Airport,
approximately 20 miles from Gideon Pond Elementary. Because atmospheric pressure varies with
altitude, the absolute pressure reading at the school is slightly different from that reported here.
However, because the difference is both slight and systematic, it does not change die effect shown
here (5).
DISCUSSION
The results discussed here add useful information to die knowledge base of radon behavior in
schools. In particular, this study found that HVAC operation can directly and significantly affect
the radon concentration in a school classroom. Improving ventilation to reduce high radon levels
is therefore recommended as another method of improving indoor air quality in schools.
Comparison of testing devices suggests that a long-term (in this case, 70-day) alpha-track test
gives a lower reading than short-term charcoal canister tests. Continuous monitor readings vary as
expected and on any single day are likely to be higher than the long-term alpha-track data.
Assuming alpha-track results represent die true radon concentration most accurately, these results
support the use of a long-term radon test over a short-term test. This is consistent with Minnesota
Department of Health radon testing recommendations. The observed variation in radon
concentration among different rooms in a given school also demonstrates the importance of testing
each occupied classroom in contact with the ground.
Finally, radon concentration appears to be inversely associated with atmospheric pressure. While
the relationship is not fully understood, the data collected in this study support the previously
suggested association. A better understanding of this effect might be useful for improving radon
testing and mitigation protocols.
-------
ACKNOWLEDGEMENTS
The Minnesota Department of Health appreciates the cooperation and assistance received from the
staffs of Gideon Pond, Sweeney, and Lincoln Elementary Schools and from Cleveland
Elementary-Secondary School. We also thank Gregory Spoden of the Minnesota Department of
Natural Resources State Climatology Office for providing climatologic data and Honeywell for
the loan of their radon continuous monitors.
The work described in this paper was partially funded by the United States Environmental
Protection Agency but has not been reviewed in accordance with dieir peer and administrative
review process and therefore the contents do not necessarily reflect the views of die Agency and no
official endorsement should be inferred.
REFERENCES
1. Schmidt, A.L. The results of EPA's school protocol development study. In: Proceedings of
die 1991 International Symposium on Radon and Radon Reduction Technology. United States
Environmental Protection Agency, Philadelphia, PA, April 1991, Volume 5, Session 10.
2. Fisher, G. Diagnostic evaluation of 26 U.S. schools— EPA's school evaluation program. In:
Proceedings of the 1991 International Symposium on Radon and Radon Reduction Technology.
United States Environmental Protection Agency, Philadelphia, PA, April 1991, Volume 5,
Session 10.
3. Schmidt, A. L., MacWaters, John T., and Chmelynski, Harry. Seasonal variation in short-term
and long-term radon measurements in schools. In: Proceedings of the 1991 International
Symposium on Radon and Radon Reduction Technology. United States Environmental
Protection Agency, Philadelphia, PA, April 1991.
4. School Protocol Development Study. United States Environmental Protection Agency,
Office of Radiation Programs, June 1989.
5. Spoden, Gregory. State Climatology Office, Minnesota Department of Natural Resources.
Personal communication, May 1992.
-------
TABLE 1. MINNESOTA SCHOOLS RADON TESTING SCHEDULE.
SEASON
l
(SUMMER)
2
(FALL)
3
(WINTER)
4
(SPRING)
ROUND
1
2
3
1
2
3
1
2
3
1
2
3
HVAC STATUS
OFF
NORMAL
ON
OFF
NORMAL
ON
OFF
NORMAL
ON
OFF
NORMAL
ON
DATE
7/28/89-7/31/89
7/31/89-8/2/89
8/4/89-8/7/89
l_l 1/3/89-11/6/89
11/6/89-11/8/89
11/10/89-11/13/89
2/2/90-2/5/90
2/5/90-2/7/90
2/9/90-2/12/90
5/4/90-5/7/90
5/7/90-5/9/90
5/11/90-5/14/90
-------
CHARCOAL CANISTER
2.5 -
2 :
0.41
OFF NORMAL ON
HVAC STATUS (THREE 3-OAY PERIODS: 7/28/89-8/7/89)
CONT1NOUS MONITOR
OFF NORMAL ON
HVAC STATUS (THREE 3-OAY PERIODS: 7/28/89-8/7/89)
ALPHA-TRACK DETECTOR
70-DAY AVERAGE (6/23/89-8/30/89)
Figure 1. Radon concentration and HVAC status: Gideon Pond Elementary School. Concentrations
shown represent an average of all rooms tested with each testing device during the time period
specified.
-------
CHARCOAL CANISTER
3.5
3
Z.5
[Rn]. 2
pCi/L 1-5
1
0.5
0
3.0S
OFF NORMAL ON
HVAC STATUS (THREE 3-DAY PERIODS: 7/28/89-8/7/89)
CONTINUOUS MONITOR
OFF NORMAL ON
HVAC STATUS (THREE 3-DAY PERIODS: 7/28/89-8/7/89)
ALPHA-TRACK DETECTOR
0.56
7
70-OAY AVERAGE (6/23/89-8/30/89)
Figure 2. Radon concentration and HVAC status: Sweeney Elementary School. Concentrations shown
represent an average of all rooms tested with each testing device during the time period
specified.
-------
2.5 -
2 -
7/31/89
ROOM 1
8/1/89
8/2/89
7/31/89
ROOM 9
8/1/89
8/2/89
[Rn].
pCi/L
7/31/89
ROOM 11
8/1/89
8/2/89
Alpha-track
Charcoal canister
Continuous monitor
Figure 3. Comparison of radon measurement devices in Gideon Pond Elementary School. Charcoal
canister and alpha-track results are three-day and 70-day averages, respectively.
-------
3
2.5
2
pCI/L
1
0.5
^B
q
-
,1 — 1,1 |,
j 4 j i+j L(_ -^
1234
\
\
5
/
/
\/
V
R
/
R
-H -t • " — i '
6 7
8
\ ,
•
, — |
R
9
B.--*-,
/
_...
,n,
10 11
/
V
12
/
p
/
13
DAY (May 2 7- June 19. 1990)
1
Radon concentration
\
14
R=
\
P
•!,
__—
V
R
/
/
/
i —
R
15 16 17
'RAINFALL
p
/
18
I
.
I — 1
19
\
|
R
V
R
/
20 21
Atmospheric pressure
•
/
—
22
R
r 29.4
29.2
29
28.8
28.4
28.2
to
23
WINCHES
Figure 4. Radon concentration and atmospheric pressure in Gideon Pond Elementary School, Room 1.
Pressure data was collected at the Minneapolis/St. Paul International Airport.
-------
ESTIMATES FROM THE IT. S. ENVIRONMENTAL PROTECTION AGENCY'S
NATIONAL SCHOOL RADON SURVEY (NSRS)
By: Lisa A. Ratcliff
U. S. Environmental Protection Agency
Office of Radiation Programs
401 M Street, S. W.
Washington, D. C. 20460
Jane Williams Bergsten
Research Triangle Institute
P. O. Box 12194
Research Triangle Park, NC 27709
ABSTRACT
During the winter of 1990-91, a nationwide probability sample
of approximately 1,000 public schools participated in the NSRS.
In approximately 100 of the schools, both long-term and short-term
measurements were made in all schoolrooms, yielding alpha track
and diffusion barrier charcoal canister measurements for about
4,000 schoolrooms. In abut 900 schools, short-term radon
measurements were made in all schoolrooms having ground contact,
yielding diffusion barrier charcoal canister measurements for
about 28,000 schoolrooms. National and regional estimates of
radon levels in schools are presented.
Approximately 78 percent of the nation's public schools have
no rooms with screening radon levels over 4 pCi/L, approximately
11 percent have either 1 or 2 such rooms, and an additional 11
percent have 3 or more rooms over 4 pCi/L. Approximately 3
percent of the nation's public schools have one or more rooms with
screening levels over 10 pCi/L.
-------
Session X
Radon in Schools and Large Buildings
-------
X-1
Title: EPA's Revised School Radon Measurement Guidance
Author: Chris Bayham, U. S. EPA, Office of Radiation Programs
This paper was not received in time to be included in the preprints, and
the abstract was not available. Please check your registration packet for a
complete copy of the paper.
-------
X-2
RADON IN COMMERCIAL BUILDINGS
By: Harry Grafton and Ayotunde Oyelakin
Columbus Health Department
Environmental Health Division
181 S. Washington Blvd.
Columbus, OH 43215
ABSTRACT
Radon screening and followup data reported for 250 Municipal
Buildings are presented. Data are organized to group buildings by
structural characteristics and by occupant activity.
Prioritization guidelines for large building testing projects are
suggested. Commercial building radon levels are contrasted to
residential radon levels to demonstrate the significance of radon
testing in the workplace.
-------
X-3
IOWA MULTIRESIDENTIAL BUILDING RADON STUDY
by: James w. Cain
Iowa State University Energy Extension
Ames, Iowa 50011
ABSTRACT
This study tested radon levels in duplex and other multiresidential
ouildings. Fifty-nine units in twenty-six buildings were tested. Long-term
(9 month) and short—term (3 day) tests were used to evaluate radon levels.
Twenty units were tested using both test devices. The averages for the duplex
units are 1.8 pCi/L for the long-term tests and 2.9 pCi/L for the short-term
tests. The averages for the multiresidential buildings are 3.8 pCi/L for the
long-term tests and 4.3 for the short-term tests. Radon levels above 4 pCi/L
are found on every level of multiresidential buildings and duplexes. Fifteen
percent of the second floor units tested, and twenty percent of the third
floor units, were above 4 pCi/L. This study finds elevated radon levels in
all types of buildings and on all the floors, and recommends that testing be
encouraged in all dwelling units regardless of the location in a building, or
the type of building.
-------
IOWA MULTIRESIDENTIAL BUILDING RADON STUDY
A substantial amount of testing has been conducted on single family
homes in Iowa. A study conducted by the Iowa Radon Project including 7100
dwellings found 70% of the dwellings have tests of 4 pCi/L or greater.1 The
U.S. Environmental Protection Agency-Iowa Department of Public Health survey
of 1381 homes found 71% of dwellings have test results of 4 pCi/L or greater
(1989). The findings of these two studies are essentially the same. These
screening surveys indicate a large percentage of Iowa's single family homes
should have additional testing and are potential candidates for mitigation.
Indeed, to date, no Environmental Protection Agency supported state survey has
found as high a percentage of dwellings at or above the action level.
The Iowa Department of Public Health, the Iowa Radon Project, and other
Iowa organizations have expended considerable effort to convince home owners
to test. An assessment of public awareness conducted by the Iowa Radon
Project in 1990 indicates some knowledge about radon is possessed by a very
large part of the population.2 Although the portion of the population which
has tested is small, the message appears to be getting to people. The
message, that every home should be tested, clearly means single family homes.
This message is based on solid research. Little research has been conducted
in Iowa on multi-family structures. Therefore, this study focuses on multi-
residential buildings.
A target population of fixed income families was chosen as a mechanism
to gain entree into multiresidental buildings. First, fixed income families
living in duplexes and multi-family buildings were identified. Selection for
participation was structured to produce a sample that included a variety of
multiresidential building types—2 family, 4 family, 6 family, 12 family, and
larger buildings.
The study was conducted in two phases. The first phase is long-term
radon testing conducted with alpha-track detectors. The testing period was
nine months. The number of units in the first phase is smaller than the
second phase; not all units were included in phase one.
The second phase is short-term radon testing. The short-term tests were
conducted with charcoal canisters with a typical exposure of three days. The
intent was to include all of the participants from the first phase plus
additional units.
As the study proceeded, the opportunity arose to include a 48-unit
building. This building, the Project, was not available for inclusion in the
first phase. Therefore, no long-term tests were conducted in the building.
During phase two, we were able to conduct short-term tests in 31 (34?) units
of the Project, in the office, and in the common room.
^reiner, T.H.; Hodges, L.; and Cain, J. Radon in Iowa. Paper presented at
1989 International Winter Meeting, American Society of Agricultural Engi
neers, New Orleans, Louisiana. December 12-15, 1989. p. 2.
2Cain, J.W., Radon Public Awareness: A 1990 Survey of Iowa Residents. Iowa
Radon Project, Iowa State University Energy Extension, Ames, Iowa, 1991. p.6.
-------
The
and multi
agencies
Agencies,
groups.
a number
asked to
SAMPLE SELECTION AND LOSSES
first task was to identify fixed income families living in duplexes
-family buildings. This was accomplished by contacting a number of
which provide services to fixed income families: Community Action
elderly meal sites, housing authorities, and similar community
These service providers were given an explanation of the program and
of return post cards for potential participants. Each agency was
inform eligible families and give them a post card to send in.
The post card asked the potential participant to provide an address,
indicate the type of building they lived in (mobile home, duplex, apartment
building, or single family), and affirm they were a fixed income household.
Not counting the responses from people living in single family homes and
mobile homes, 52 post cards were returned in time for selection. From these
responses a sample of participants was selected.
The original pool included 11 families in duplexes and 31 families in
other multi-family buildings. Two samples were drawn from the pool. The
first sample was for the long-term tests (alpha tracks), and the second sample
was for the short-term tests (charcoal canisters). The short-term pool
includes all the units in the long-term pool. Two households were eliminated
because they were not in central Iowa.
For phase one a sample of 24 units was desired, so an initial sample of
30 households was selected. The extra households were included to cover
anticipated losses. Some families withdrew during the placement of the detec-
tors, and additional selections were made to replace those households. A
variety of building sizes (sized by number of units) was desired in the
sample. Therefore, it was determined that no more than 25% of the units would
be duplexes. Households were selected to produce a sample with a structural
profile to match the goal.
The short-term sample included all the qualified families, including
everyone in the long-term sample. The profile of buildings in the pool
assured that a variety of building sizes are represented in the sample.
Among the families that withdrew during the first phase, long-term
testing, were three families that resided in the same building. These
individuals indicated management would not permit them to participate. After
some communication with the management, an agreement was reached to include
the three units in the short term sample and to test the remainder of the
building. This building is the Project.
TABLE I. SAMPLE SELECTION AND LOSSES
Duplexes
Multi-Family
Project
TOTALS
LONG TERM SAMPLE
Drawn
9
23
32
Tests
6
15
21
Losses
3
8
11
SHORT TERM SAMPLE
Drawn
10
22
32
Tests
8
19
31*
58
Losses
2
3
5
FINAL
9
19
31*
59
* Does not include tests conducted in non—living areas
-------
Sample losses in the long-term phase include families that withdrew and
those households where the testing could not be completed. Two families
moved, two families threw the alpha track away, and one family lost the detec-
tor. Sample losses in the short-term phase include the families that could
not be reached to arrange testing.
Table I summarizes the sample selection and loss information. Combined,
the final sample includes 9 families in duplexes and 19 families in multi-
family buildings plus the additional units in the Project.
TESTING PROCEDURES
ALPHA TRACK DETECTORS
The placement of the long-term, alpha track, detectors started on May
29, 1990 and continued until June 28, 1990. The devices were placed by staff
from the Iowa Department of Public Health (IDPH). The IDPH staff contacted
each household by phone for an appointment to place the detectors. The IDPH
staff tried to contact each of the families at least twice while the detectors
were in the homes to remind them about the study and answer potential ques-
tions. Because of these efforts the number of detectors/families lost was
minimal.
Iowa Department of Public Health staff began to collect the alpha track
detectors on March 22, 1991 and had completed the collection by March 29,
1991. The long term tests were exposed for an average 291 days (9.7 months)
over three seasons—summer, fall, and winter.
CHARCOAL CANISTERS
The placement of the short-term, charcoal detectors, started on November
26, 1990 and was completed by January 14, 1991. The canisters were placed by
staff from the Iowa Department of Public Health (IDPH). The IDPH staff
contacted each household by phone to set an appointment. Those households in
the long-term phase of the study were also contacted for an appointment.
IDPH staff collected the charcoal canisters—starting on November 30,
1990 and completed the pick-ups by January 18, 1991. The typical test was
three days in duration. Except for one over-exposed canister (24 days), the
longest exposure was five days and the shortest was three days.
The Project residents were notified by management of the day and time
staff would be in the building to place detectors. Cooperation among the
tenants was very good. All but thirteen apartments were tested. Those not
included were not home the morning the canisters were placed. In addition to
the apartments, tests were conducted in the boiler room, the office, beauty
shop, and the community room. There were no long-term tests in the Project.
OTHER TESTING ISSUES
The schedule for this research was set to produce the best test results
within the grant period. The long-term tests were placed as soon as possible
in an effort to get as long a test period as possible. By placing the
canisters in June, the long-term devices were exposed for an average of 9.7
months. The plan for the short-term detectors was to wait until the heating
season was well underway before placement began. So short-term testing was
started after Thanksgiving. The expectation was that all the short-term
testing would be completed before Christmas. However, a few short-term tests
were performed in January.
-------
No attempt was made to select which unit in a building was to be used in
this study. The tested unit(s) could be located anywhere in the building,
from basement to top floor. Placement of detectors was in living areas, not
necessarily the lowest livable area of the structure. In the case of most
multiresidental buildings, this distinction is not particularly important.
However, four of the duplexes have basements. The only basement test conduct-
ed in a duplex was where the basement was a living area, within dwelling
units the test devices were placed in bedrooms or living rooms.
The research plan included duplicate tests on ten percent of the
dwellings. Six of the thirty alpha tracks sent to the laboratory for process-
ing are duplicates, 20% of the total. Nine of the seventy-seven charcoal
canisters sent for processing are duplicates, 11.7% of the total. The
duplicate detector results were essentially the same as those for the primary
detectors, most variation in results is within three or four tenths of a pico
curie per liter.
The research plan also called for 5 percent of the tests to be blanks.
Two blank alpha tracks were sent to the laboratory, 6.6% of the total detec-
tors analyzed. Six blank charcoal canisters were sent for analysis, 7.8% of
the sample. The results for the unexposed canisters are as expected: one,
two, and three tenths of a pico curie per liter.
RESULTS
Summary results for all testing is presented in Table II. Twenty-one
units were tested with alpha track detectors. The highest alpha track result
is 14.4 pCi/L and the lowest is 0.7 pCi/L. The average is 3.3 pCi/L.1 Nearly
24% of all the units tested above 4 pCi/L.
TABLE II. SUMMARIZED TEST RESULTS
Number of
Units
Average
Result
Highest
Result
Lowest Re-
sult
% Below 4
pCi/L
% Above 4
pCi/L
ALPHA TRACK TESTS
ALL TESTS
21
3.3
14.4
.7
76.2
23.8
CHARCOAL TESTS
ALL TESTS
62
4.1
31.1
.5
62.9
37.1
WITHOUT PROJECT
UNITS
27
4.7
31.1
.5
59.3
40.7
'There is a 95% confidence that the mean for the alpha tracks for these types
of dwellings is between 1.4 and 3.8 pCi/L.
-------
Comparing the results from the short-term charcoal tests with those for
the long-term tests, presented in Table II, shows the short-term tests are
similar although a bit higher. Short-term tests were conducted on 62 units.
The aggregate average short-term result is 4.1 pCi/L1; with the lowest being
.5 pCi/L and the highest 31.1. About 37% of these units tested above 4 pCi/L.
While these are not screening tests, the difference between these results and
screening test studies of single family homes is substantial. When all the
extra units in the project are removed from the sample, the percent of
charcoal tests above 4 pCi/L increases.
DUPLEXES
The results for the tests conducted in the duplexes are presented in
Table III. Six duplex units completed the alpha track test phase of the
study. The highest score is a 5 pCi/L and the lowest is .7 pCi/L. The
average for the six duplexes is 1.8 pCi/L. All but one of the tests was
conducted on the first floor. The one test conducted in a basement is the
high reading, the only unit to test above 4 pCi/L.
Eight duplex units participated in the short-term phase of the study.
The highest is 7 pCi/L and the lowest is .5 pCi/L. The average is 2.9 pCi/L.
Two units have tests above 4 pCi/L, and one of those is a basement test.
Except for the basement unit, testing was conducted on the first floor of the
duplexes. Only one alpha track test and two charcoal tests did produce
results above the action level.
TABLE III. DUPLEX RESULTS
Number of Units
Average
High
Low
Above 4 pCi/L
ALPHA TRACK
6
1.8
5.0
0.7
16.7 %
CHARCOAL
8
2.9
7.1
0.5
25.0 %
MULTI-FAMILY BUILDINGS
Tests were conducted in 17 different multi-family buildings. Sixteen of
these buildings are discussed here. The Project had a multiple number of
tests and is discussed separately later. These buildings had only one unit
tested. A summary of information for these multi-family buildings is present-
ed in Table IV. Except for the project, this figure shows the data broken
down according to buildings and the size of the building, in units.
The long-term and short-term tests were conducted in the same units. A
wide range of building sizes are represented in the sample. Apartments to be
tested were not selected by location in the building. Only in the case of the
four unit and twelve unit buildings are there more than one building of any
size in the sample.
JThere is a 95% confidence that the mean for the charcoal tests for these
types of dwellings is between 2.6 and 4.4 pCi/L.
-------
Four different buildings have units with long-term tests above the
Action level. Two of the buildings are four unit structures. One is a twelve
unit building and one is a twenty-two unit building.
Eight buildings have short-term tests above the action level. Four of
these buildings are four unit buildings. One is an eight unit building. One
is a twelve unit building. One is a twenty-two unit building, one is twenty-
four units. All of the buildings which had long-term tests above the action
level have short-term tests above the action level.
TABLE IV. MULTI-FAMILY BUILDINGS BY SIZE
SIZE IN UNITS
48
24
22
20
12
8
6
5
4
Totals
# OF
BUILDINGS
1
1
1
1
2
1
1
1
8
17
# OF
UNITS
34
1
1
1
2
1
1
1
8
50
ALPHA
TESTS
*
2.7
14.4
1.5
4.3
0.7
2.1
1.1
2.0
5.5
1.7
2.5
1.1
0.7
3.4
14.0
CHARCOAL
TESTS
*
4.9
31.1
0.8
6.8
1.6
4.5
1.2
1.8
1.2
5.5
4.4
3.7
0.8
1.6
6.1
15.2
Summarized separately.
Table V summarizes the results for the Project tests and the rest of the
buildings in the study, the regular sample. The Project results will be
discussed later. Table V shows that for the regular sample, both long-term
and short-term tests were taken in 15 of the buildings. One building had only
a short-term test. The average for the long-term tests is 3.8 pCi/L, with a
high of 14.4 pCi/L and a low of .7 pCi/L. The average for the short-term
tests is 5.4 pCi/L, with a high of 31.1 pCi/L and a low of .8 pCi/L. Most of
the tests, 13 of 19, were conducted in first floor apartments. Three were in
basement units, one in a second floor unit and two in third floor apartments.
Four of the buildings have basements, the remaining twelve are built slab-on-
grade.
-------
The charcoal tests results are slightly higher than those for the alpha
track tests, see Table V. The average of all the alpha track tests is 3.8
pCi/L and the charcoal test average is 5.4 pCi/L. Ten units had higher
charcoal tests results than alpha track results, three were lower, and two
were the same. The percentage of units above 4 pCi/L is much higher for the
charcoal tests, 42%, than the percentage of alpha track units, 27%.
TABLE V. MULTI-FAMILY BUILDINGS
ALPHA TRACK TESTS
Highest
Lowest
Average
Above 4 pCi/L
CHARCOAL TESTS
Highest
Lowest
Average
Above 4 pCi/L
REGULAR
SAMPLE
15
14.4
0.7
3.8
27 %
19
31.1
0.8
5.4
42 %
PROJECT
35
9.5
0.9
3.7
34 %
TOTAL
54
31.1
0.8
4.3
39 %
THE PROJECT
The Project is a two and three story building, built in the general
shape of a U. The two legs of the U are primarily two stories and the
connecting base of the U is three stories. The building is constructed on a
slab. The building contains 48 dwelling units, an office, a community room,
the boiler room, and a beauty shop area. Thirty-six of these spaces were
included in the testing program. On both the first and second floors 13 areas
were tested. On the third floor 9 apartments were tested.
All of the tests on The Project building are short-term tests. The
tests were all conducted between Monday, December 10 and Thursday December 13,
1990. Within the project building a variety of results were obtained. The
range extends from a low of 1.1 pCi/L to a high of 9.5. The average reading
for the building is 3.7 pCi/L.
The boiler room test is 3.3 pCi/L. Nine of the 15 areas tested on the
first floor have readings above 4 pCi/L (60 % of the areas tested). First
floor tests range from a low of 1.8 pCi/L to a high of 9.5 pCi/L.
Thirteen second floor areas in The Project were tested. The average of
the second floor readings is 2.3 pCi/L. The lowest second floor reading is .9
pCi/L and the highest is 6.2 pCi/L. Eight of the test results are below 2
-------
pCi/L (53%). Two second floor (15 %) readings are above the action level (4.6
and 6.2 pCi/L).
The Project's third floor tests range from a low of 2.2 pCi/L to a high
of 6.2 pCi/L. Six of the nine third floor tests (66%) are between 2 and 3
pCi/L. Two third floor units (22 %) have readings above the action level (4.1
and 6.2 pCi/L). The average of the third floor readings is 3.2 pCi/L.
The third floor readings are slightly higher than those on the second
floor. The second floor average is 2.3 pCi/L while the third floor average is
3.2 pCi/L. The lowest second floor reading is .9 pCi/L while the lowest third
floor reading is 2.2 pCi/L. Eight second floor readings are below 2 pCi/1,
but no third floor readings are below 2 pCi/L. On the other hand, the highest
reading on either the second or third floor is the same, 6.2 pCi/L.
RESULTS BY BUILDING LEVEL LOCATION OF UNIT
One hypothesis expects the radon levels to decline as one goes up from
one level to another. In this theory radon levels in multi-family buildings
are expected to be lower on upper floors because forced air heating is not
common in such buildings and because apartments are separate units. However,
this pattern does can occur in single family homes even when heated with a
forced air system. There are many routes in multi-family buildings through
which radon can migrate from the lowest level to the top floors; elevator
shafts, plumbing and chimney chases, and electrial wiring chases or cavities.
The test results by floor are summarized in Table VI. In considering the
radon levels on various floors of buildings all tests are included, including
the Project.
Basements
Five tested areas are located in basements. Four of these areas have
both long-term and short-term tests. On the short-term tests four of the five
areas are above the action level. One of these apartments tested at 4.9 pCi/L
with the short-term test, but is 2.7 pCi/L on the long-term test, considerably
below the action level.
TABLE VI. RESULTS BY FLOOR
FLOOR
Basement
First
Second
Third
TOTAL
UNITS
4
34
12
10
ALPHA TRACK TESTS
Units
4
16
0
1
% > 4 pCi/L
75.0
12.5
0.0
CHARCOAL TESTS
Units
4
32
12
10
% > 4 pCi/L
100.0
43.8
16.7
20.0
# WITH
BOTH TESTS
3
15
0
1
First Floors
The first floor test results can be divided into those on units in
buildings constructed directly on the ground, slab-on-grade, and those with
-------
units in buildings with basements and/or crawlspaces. A total of 35 first
floor areas were tested. Eight of the 35 first floor test areas are in
duplexes and 27 are in multi-family buildings.
Long-term tests were conducted in 16 first floor areas (duplexes and
multi-family buildings). The highest of these long-term test results is 14.4
pCi/L and the lowest is .7 pCi/L. The average long-term reading is 2.6 pCi/L.
Two of the units are above the action level (5.5 and 14.4 pCi/L). Of the 14
tests below 4 pCi/L ten of them are below 2 pCi/L, 62.5% of all the long-term
first floor tests.
The short-term tests display a greater variability with a high of 31.1
pCi/L and a low of .5 pCi/L. The average short-term reading for the first
floor units 4.6 pCi/L. There are 15 units (44%) with short-term test results
above 4 pCi/L and the remaining 19 (56%) are below the action level.
Six of the first floor tests were taken in buildings with basements and
two with crawlspaces. Four of the units over basements have long-term tests.
The average is 1.3 pCi/L and none of the long-term results is over the action
level. The average for the short-term tests on all six of the basement units
is 2.4 pCi/L, one unit exceeded the action level. The two units built over
crawlspace are similar in their results; neither unit is over the action
level, the long-term test average is 1.4 pCi/L and the short-term test average
is 1.5 pCi/L.
More than half (27) of the first floor test areas are in slab-on-grade
buildings. In 11 cases long-term tests were conducted. The average long-term
reading is 3.2 pCi/L. Two of the 11 are above 4 pCi/L. In 26 cases short-
term tests were conducted with an average of 5.3 pCi/L. A total of 14 of the
26 (54 %) are above the action level.
Second Floors
There are no long-term tests on any second floor apartments, but there
are 13 short-term tests. All of the second floor tests results are from the
Project. The average short-term test for second floor apartments is 2.3
pCi/L. Eight (61.5%) of the test results under 2 pCi/L. Only two units
(15.4%) have tests above the action level (4.6 and 6.2 pCi/L). The highest
test is 6.2 pCi/L and the lowest is .9 pCi/L.
Third Floors
There is one long-term test for a third floor unit and ten short-term
tests. All but one of the third floor test results are from the Project. The
long-term test (0.7 pCi/L) is well below the action level. The average of the
short-term tests is 3.1 pCi/L. The short-term test results range from a low
of 1.6 pCi/L to a high of 6.2 pCi/L. Six of the third floor units (60%) have
readings between 2 and 3 pCi/1; seven (70%) have readings less than 3 pCi/L.
Two units (20%) have test results above the action level.
FINDINGS
DUPLEXES
Data was collected on nine duplex units occupied by fixed income
families. The duplex averages are below the action level; long-term average
1.8 pCi/L, short-term average 2.9 pCi/L. One long-term test and two short-
term tests are above the action level. Although the sample is small, elevated
radon levels can occur and the percentage of scores above the action level
-------
(16.7% of long-term and 25% of short-term) suggests the number of duplexes in
the housing stock might be substantial.1 A recommendation to test dwelling
units in duplexes is prudent.
MULTI-FAMILY STRUCTURES
The results of testing in multi-family buildings indicates that testing
all units is an_appropriate activity. Tests were conducted in 17 different
buildings and nine buildings have at least one test above the EPA action level
of 4 pCi/L. The project building has several apartments with short-term
readings above the action level. Four of the 15 long-term tests (27%) are
above the action level, and 19 of the 50 short-term tests (38%) are above the
action level.
The average long-term readings are lower than expected, but four of the
readings are above the action level. While 17 buildings is not a large
sample, many different sizes of buildings are represented. Readings above the
action level are found in different sizes of buildings.
BY FLOOR
Long-term tests in basement apartments were above the action level in
three of four long-term tests and four of five short term cases. Two of 16
(12.5 %) first floor apartments have long-term test results above the action
level and 44 % of the short-term first floor tests are above the action level.
In two second and two third floor apartments the short-term tests are above
the action level. Although most of the tests that are above the action level
were conducted in basement or first floor apartments, these results clearly
indicate elevated levels can occur on the upper floors of multi-family build-
ings. Testing of units in multi family buildings, including the second and
third floor apartments, is recommended.
The data connecting radon levels with basement and crawlspace construc-
tion practices are not conclusive. Although none of the first floor units
over basements or crawlspaces have readings above the action level, there are
only 8 units in this category. The results for slab-on-grade buildings are
somewhat skewed by the fact that the Project accounts for more than half of
the readings above the action level. However, it is clear that with slab-on-
grade construction elevated radon levels are possible.
CONCLUSION
In summary, this study concludes that the testing of all dwelling units
in duplexes and multi-family structures is advised. Despite the fact that
multi-family buildings and duplexes seem to have lower frequency of units
above the action level than found in Iowa single-family homes, apartments in
these type of structures can have radon above the 4 pCi/L level. Testing
should be encouraged in all structures and on all floors.
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.
One ineligible unit adjacent to a participant was tested in the short-term
phase. In this unit the short-term readings are 15.6 pCi/L in the basement
and 14.9 pCi/L on the first floor. The other unit in this building tested at
3.3 pCi/L on the first floor.
-------
REFERENCES
1. Greiner, T.H., Hodges, L., and Cain, J. Radon in Iowa. Paper presented
at 1989 International Winter Meeting, American Society of Agricultural
Engineers, New Orleans, Louisiana. December 12-15, 1989.
2. Cain, J.W. Radon Public Awareness: A 1990 Survey of Iowa Residents. Iowa
Radon Project, Iowa State University Energy Extension, Ames, Iowa, 1991.
-------
X-4
Title: Airflow in Large Buildings
Author: Andrew Persily, U. S. Department of Commerce, Building and
Fire Research Laboratory
This paper was not received in time to be included in the preprints, and
the abstract was not available. Please check your registration packet for a
complete copy of the paper.
-------
X-5
MEETING VENTILATION GUIDELINES WHILE CONTROLLING
RADON IN SCHOOLS
By: Gene Fisher and Bryan Ligman
U. S. Environmental Protection Agency
Office of Radiation Programs
401 M Street S. W.
Washington, D. C. 20460
Terry Brennan
Camroden Associates, Inc.
RD #1, Box 222
Oriskany, NY 13424
William Turner
H. L. Turner Group
Concord, NH 03301
Richard Shaughnessy
University of Tulsa
Tulsa, OK
ABSTRACT
As part of a continuing radon in schools technology
development effort, EPA's School Evaluation Team has performed
radon mitigation in schools by the method of
ventilation/pressurization control technology. Ventilation rates
were increased, at a minimum, to meet the American Society of
Heating, Refrigeration and Air Conditioning Engineers (ASHRAE)
standard, "Ventilation for Acceptable Indoor Air Quality" (ASHRAE
62-1989).
This paper presents the results and the preliminary
evaluations which lead to the Team's decision to implement this
technology. Factors considered include energy penalties, comfort,
indoor air quality (IAQ), building shell tightness, and equipment
costs. Cost benefits of heat recovery ventilation were also
considered. Earlier results of the SEP Team's efforts have
indicated a severe ventilation problem within the schools of this
nation. A holistic approach to radon mitigation in schools and
other large buildings, which controls radon as well as improving
overall IAQ, should be the goal of radon remediation where
practical.
-------
X-6
REDUCTION IN A BELGIAN SCHOOL : FROM RESEARCH TO APPLICATION
by : P. Cohilis, P. Wouters and P. Voordecker
Belgian Building Research Institute (CSTC-WTCB)
Rue d'Arlon 53/10
1040 Brussels, Belgium
ABSTRACT
Several kind of measurements were performed in a school, situated in
the province of Luxembourg (south of Belgium), in which radon
concentrations higher than 1000 Bq/m3 were measured. These experiments
included tracer gas techniques, pressurization techniques and CO2 and radon
concentration measurements, in order to define appropriate remedial
actions. As a consequence of the results of these studies, two mechanical
ventilation systems producing an over-pressure in the rooms and improving
the global ventilation were installed and tested in two rooms of the
school. Their use was completely successful in achieving radon reduction.
The possibility of using a sub-slab depressurization system was also
evaluated, and it was shown that this method could also be applied to
reduce radon levels in the school. However, it was decided to consider the
solving of the radon problem in a larger context and to suggest, to the
school authorities, the use of the method based on the pressurization of
the rooms. Indeed, as many Belgian schools have general indoor air quality
problems (too low air change rates), solutions aiming to solve several
problems at the same time should be considered in first instance.
-------
INTRODUCTION
The average indoor radon concentration in Belgium is almost 50 Bq/in3,
with a clear distinction between, roughly, the north (mean value :
39 Bq/m3) and the south (mean value : 11 Bq/m3) parts of the country, due
to geological factors (1). The highest indoor radon concentrations may be
found in the south : the available data indicates that a very high
proportion of the buildings expected to have radon concentrations of more
than 400 Bq/in3 (the European action level for existing buildings) are
situated in the province of Luxembourg.
A number of radon concentration measurements were made in schools of
this province, during the last two years, on the initiative of the
educational responsibles. These measurements allowed the identification of
a few school buildings with an important radon problem (2).
The present paper describes the investigations which were performed in
one of these schools in order to define the most appropriate remedial
actions. In this school, the radon concentrations measured in some
classrooms were higher than 1000
EVALUATION OF THE CHARACTERISTICS OF THE BUILDING
PRESENTATION OF THE BUILDING
Figure 1 presents a schematic view of the studied school. It is a
modern building of the middle of the eighties. The rooms are distributed
among the basement, the ground level and the first floor. All the
investigations were performed on the right part of the building (see arrows
in figure 1), which is the zone where high radon concentrations were
measured. In fact, the occupied rooms presenting a radon problem were rooms
CL.l to CL.9, situated at the ground level. All those rooms have a floor
heating system and small opening windows at the top of fixed windows. They
also have large sash-windows usable for intensive ventilation. But because
of comfort problems these windows are not frequently used during most part
of the year.
It is important to note that only some of the rooms of the ground level are
situated above the basement (rooms CL.l to CL.4 and part of the central
hall). The other rooms of this level are constructed directly on the soil
(ground-supported floor), with a gravel layer (30 cm thick) immediately
below the slabs.
RADON ENTRY PATHS AND TRANSPORT
The use of tracer gases (N2O, SFg, ...) is an attractive way to study
the radon transport within a building. For this purpose, a tracer gas (N2O)
was injected in the basement hall and N2O concentrations were measured, in
function of time, in all the rooms of the measure zone defined by the
arrows in figure 1. The MATE-System (Multi-purpose Automated Tracer gas
Equipment) (3) was used for the monitoring.
-------
J:
FIRST FLOOR
L r
CL. D
BASEMENT
Figure 1. Schematic view of the school. The arrows
investigations were performed.
indicate the zone of the building where the
-------
These tracer gas measurements allow, among other things, to check the
hypothesis following which the basement could be the principal source of
the high radon (222Rn) concentrations measured in the ground level rooms.
In order to make this check, the evolution with time of the N2O
concentrations measured in the rooms must be compared to radon
concentrations measured during the same periods, in the same rooms.
Four measurement periods were defined, corresponding to different
experimental conditions and dates (see table 1).
TABLE 1. MEASUREMENT PERIODS AND RELATED CONDITIONS FOR THE TRACER GAS
EXPERIMENTS
Period
Date
(Jul. Day)
Conditions
226-229
229-233
233-236
236-240
basement entry
room doors
room windows
basement entry
room doors
room windows
basement entry
room doors
room windows
basement entry
room doors
room windows
closed
open
closed
open
open
closed
open
closed
closed
open
closed
closed except in
CL.2 and CL.8
The results of the tracer gas measurements are given in table 2, where
we present a "pollution indicator" obtained by dividing, for each room, the
measured concentration by the concentration measured in the injection room
(basement hall), and averaging over each period. More detailed results may
be found elsewhere (4).
TABLE 2. N2O POLLUTION INDICATORS (X 100) FOR SEVERAL ROOMS, AVERAGED
OVER EACH MEASUREMENT PERIOD
Room
Basement hall
Ground floor hall
CL.l to CL.4 (mean value)
Period
1
100
2
1.5
2
100
20
19
3
100
42
3
4
100
30
6 (CL.l, 3 and 4) ;
CL.5 to CL.9 (mean value) 1 16 5
First floor (mean value) 2 19 29
0.5 (CL.2)
(CL.5, 6, 7 and 9);
0 (CL.8)
18
-------
Radon concentrations were measured with charcoal canisters exposed
during 24 hours in different rooms (one measurement per period). Table 3
presents the results of these measurements.
TABLE 3. RADON CONCENTRATIONS IN SEVERAL ROOMS (Bq/m^) DURING THE TRACER
GAS EXPERIMENTS
Room Period
1 2 3 4
Basement hall 1150 1080 990 1160
Ground floor hall 650 530 870 550
CL.8 515 460 440 40
CL.2 560 550 160 < background
The results of the N2O and radon concentration measurements (tables 2
and 3) indicate :
that the closing of the access to the basement has no influence on the
radon concentrations measured in the rooms but has a net effect on the
N20 concentrations measured in the same rooms.
- that the closing of the doors of the rooms allows the reduction of the
radon concentration in room CL.2 (located above the basement) but not in
room CL.8 (located directly on the soil). The conclusion is different
for N2O : when the doors of the rooms are closed, the N2O concentrations
in the rooms become negligible.
that the opening of the small windows of rooms CL.2 and CL.8 highly
reduces the radon concentrations measured in these rooms. The same
conclusion comes from the N2O concentration measurements : the opening
of the windows of the rooms (doors closed !) is sufficient to ensure
enough ventilation to reduce dramatically the gas concentrations.
The above remarks indicate that the basement does not act as a major
"source room" for the rooms located at the ground level. The high radon
concentrations measured in this school (essentially in the occupied rooms
of the ground floor) are related to openings (cracks, joints, ...) between
the rooms and the underlying soil.
AIRTIGHTNESS CHARACTERISTICS OF THE BUILDING
The degree of permeability to air of the studied zone of the building
was measured with the pressurization method. For an indoor/outdoor pressure
difference of 50 Pa, an air change rate of 4.0 h~l (^50 value) was
measured. This value corresponds to a "rather well airtight" building in
the belgian context.
The permeability to air of each one of the various rooms is also an
important parameter which may have an influence on the choice of an
appropriate remedial action. For example, it will be vain to try to apply
the pressurization method on a very leaky room!
-------
In order to calculate the total air-flow Qtot across all the walls of
a specific room, for a given value of the inside/outside pressure
difference AP for that room, it is necessary to make a quantitative
evaluation of its airtightness characteristics. Pressurization measurements
allow us to make such evaluations. In fact, it is interesting to separate,
for each room, the air-flow QOut across the walls separating the room from
outside the building ("outer walls") from the air-flow Q^n across the walls
separating the room from other rooms in the building ("inner walls") :
Qtot = Qsut + Qin
U)
- cc« . |APCUJN"" + c
lfl
where COut anc* Nout characterize the outer walls, Cin and Nin characterize
the inner walls, APout is the pressure difference between the room and the
world outside the building and APin is the pressure difference between the
considered room and the remainder part of the building.
The knowledge of Qtot (and of Qout and Qin) for the rooms of the
building is rather important. Let's take again the example of the
pressurization method : a low value for Qin (comparatively to QOut)» for a
particular room, means that the production of an over-pressure in that room
will have a weak influence on the pressure in an adjoining room. In other
words, the over-pressure produced in a given room by a ventilation system
will not be significantly affected by what happens in the neighbouring
rooms. By another way, two similar rooms with rather different Qtot values
for the same values for AP may need the use of ventilation
(pressurization) systems with rather different characteristics in order to
produce similar values of AP in those two rooms.
The experimental method employed to determine Cout, Nout, Cin and Nj_n
is based on the use of two variable-flow fans. For each room, one of the
fans controls the pressure difference between the considered room and the
remainder part of the building, and the other one allows the control of the
pressure difference between this remainder part of the building and the
exterior. More details may be found elsewhere (5).
The values for Cout, Nout, Cin and Nin were determined for some rooms of
the school. It was concluded that the values for Cj.n are low comparatively
to the Cout values, for each one of these rooms : the principal
airtightness defaults in rooms CL.l to CL.9 are situated on their outer
walls. These rooms are well separated from the rest of the building when
the doors are closed.
POSSIBLE REMEDIAL ACTIONS
Several radon reduction techniques may be applied to schools (6) . It
appears, from the above described characteristics of the building, that two
types of remedial actions should be tested in this school-building,
together with the sealing of radon entry routes : the pressurization of the
classrooms and the subslab depressurization. These two techniques have the
same goal : to reverse the driving pressure difference between the soil and
the house, in order to obtain a higher pressure in the lower part of the
-------
building than underneath the entire slab (to avoid the entering of soil gas
\_. . . ^^-vrs^f^^t* T rt T"l I
by convection).
EVALUATION Or THE CLASSROOM PRESSURIZATION METHOD
Two different ventilation systems were installed in the school, one of them
in room CL.7 (ventilation system called "RENS") and the other one in room
CL.8 (system called "VENT"). These systems are air supply systems, then
producing an over-pressure in the rooms.
The air-flow versus pressure difference curves corresponding to these two
systems were measured in our laboratories, in an experimentation chamber.
They are presented in figures 2 and 3. For each system the measurements
were mace for three ventilator velocities (minimum, medium and maximum) and
the entrance grids were always kept completely open. In figures 2 and 3
those measurements are represented by different symbols on solid lines.
As the over-pressures produced in the rooms by these systems also
depend on the airtightness characteristics of the rooms, we also present in
figures 2 and 3 the characteristic curves of the rooms (dashed lines)
calculated using an expression similar to expression (!) with the
parameters Cout' ^out' *-in and Nin defined in the previous paragraph (and
experimentally determined as explained above). The intersection points
between the dashed lines and the solid lines, in figures 2 and 3, indicate
the air flows and the over-pressures expected for rooms CL.7 and CL.8 with
the used ventilation systems.
400
Minimum
Medium
A Maximum
Room CL.7
Figure 2. Characteristic curves for ventilation system RENS (three
ventilator velocities : minimum, medium and maximum) and
for room CL.7 (dashed line).
-------
1000
Minimum
40 60
DP [Pa]
Medium A Maximum
80
Room CL.8
100
Figure 3. Characteristic curves for ventilation system VENT (three
ventilator velocities : minimum, medium, maximum) and for
room CL.8 (dashed line).
Radon concentrations were measured before and after the installation
of the ventilation systems, in order to evaluate the influence of their
operation en the concentrations measured in the rooms where they were
installed and also in the adjoining rooms (rooms CL.6 and CL.9, see
figure 1).
The radon concentrations were measured with charcoal canisters exposed in
the rooms during two to three days. Four measurement periods were defined,
corresponding to different experimental conditions and dates (see table 4).
Table 5 presents the results of these measurements. It can be seen that the
two ventilation systems are very efficient in reducing the radon
concentrations in rooms CL.7 and CL.8. This reduction is a result of two
effects : the increased ventilation of the rooms and the over-pressure
created in these rooms. It is interesting to point out that the
consequences of using these ventilation systems in rooms CL.7 and CL.8 are
different for rooms CL.6 and CL.9. The radon concentration in room CL.6
seems to be independent of the use of the ventilation system (RENS) in the
adjoining room CL.7. The situation is very different for room CL.9 which
shows a large increase of its radon concentration when the ventilation
system (VENT) is ON in room CL.8 (note that the values measured in room
CL.5, which is far away from the other rooms, can be, to some extend, used
as normalisation values). This is probably due to the effect of the
"pressure barrier" in room CL.8, which is much greater than in room CL.7 as
shown in figures 2 and 3.
-------
TABLE 4. MEASUREMENT PERIODS AND RELATED CONDITIONS DURING THE EVALUATION
OF THE PRESSURIZATION METHOD.
Period
Conditions
doors and windows closed in all the rooms
cracks and joints between the floor and the walls
sealed in rooms CL.7 and CL.8
doors and windows as in period 1
cracks and joints in rooms CL.7 and CL.8 as in period 1
rooms CL.7 and CL.8 : ventilation systems installed
(but OFF)
doors and windows as in period 1
cracks and joints in rooms CL.7 and CL.8 as in period 1
room CL.7 : ventilation system (RENS) ON during all
the period, at maximum velocity
room CL.8 : ventilation system (VENT) ON during all
the period, at maximum velocity
doors and windows as in period 1
cracks and joints in rooms CL.7 and CL.8 as in period 1
room CL.7 : ventilation system (RENS) ON during all
the period, at medium velocity
room CL.8 : ventilation system (VENT) ON during all
the period, at minimum velocity
TABLE 5. RADON CONCENTRATIONS IN Bq/m3 ("-" MEANS "NO MEASURE") FOR
PERIODS DEFINED IN TABLE 4
Room
Period
CL.5
CL.6
CL.7
CL.8
CL.9
-
1 335
450
~
190
2 020
1 390
247
610
-
42
24
^
180
1 830
120
19
1 680
It can be said, in conclusion, that the use of these ventilation
systems was completely successful, for both systems. This is in part due
to the fact that the air-tightness of the rooms is rather good (as it was
shown by pressurization measurements) so it was rather easy to produce an
over-pressure of a few Pascals (especially with system VENT). Ventilation
systems similar to one of these two systems, installed in each room, could
be an efficient radon reduction method in this school. Moreover, their use
allows the reduction of the high C02 levels which can be measured in the
-------
rooms during the occupation periods. This is illustrated in figure 4 which
shows, for one of the rooms, the evolution with time of the C02
concentration for two experimental conditions : fan ON and a small window
open (29 April), and fan OFF and all windows closed (30 April).
CO ^
§43
c
a>
o
c
o
o
o
u
4 -
F/sn
Door
Small window
Window
C02
i
0
I
4
I
8
12
16
20
i
0
i
4
i
8
12
16
20
10 14 18 22 2 6 10 14 18 22
Time (hour) 29 and 30 april
Figure 4. Evolution, with time, of the C02 concentration in a
classroom for two experimental conditions.
EVALUATION OF THE SU3SLAB DEPRESSURIZATION METHOD
The subslab depressurization (SSD) technique is an alternative to the
classroom pressurization method. It is well known that the probability of
success of the SSD method depends on several factors, in particular on the
permeability below the slab (7). As some of the rooms of the studied
building are constructed directly on the soil, with a gravel layer
underneath tr.e slabs, it appeared that it could be interesting to install
and test an SSO system in one of. these rooms.
-------
The gravel layer below the slab of room CL.7 was depressurized with
the help of a variable-flow extraction fan. A rectangular opening was
performed through one of the exterior walls, in a parallel direction to and
below the slab, up to the gravel layer. A pipe (100 mm diameter) was
introduced inside this opening, in order to allow the connection of the
fan. Afterwards, the rectangular aperture was sealed with a PUR foam.
To evaluate the effectiveness of the depressurization of the gravel
layer it is necessary to measure the pressure at several points of this
layer. For this purpose several regularly-spaced small apertures were
drilled through the exterior walls of room CL.7 (and also of the adjoining
rooms), below the slab, up to the gravel. This allows the measurement, with
a micro-manometer, of the pressure difference between several points of the
gravel layer and the exterior of the building. All these "measurement
points" are connected to the micro-manometer with the help of a selection
system. Most of the measurement points are located on the perimeter of room
CL.7, so they are very close to the joints between the slab and the walls
of that room. Figure 5 shows the position of all the measurement points
(numbered from 1 to 12) .
-U,
CL.5
Hall
i— 1
CL.6
— 2
3 4
CL. 9
12
CL. 8
CL.7
11
— 5
FAN
10 9 8
Figure 5. Position of the various "measurement points" (numbered
from 1 to 12) for the evaluation of the SSD technique.
-------
We have measured, with the apparatus described above, the pressure
differences AP between each measurement point (in the gravel layer) and
the outside, for different values of the flow Q measured through the fan.
Figure 6 presents the results of these measurements. It shows, for
different Q values, the evolution of the pressure difference AP in
function of the number of the measurement point.
0
-10
-20
-30
-40
-50
-60
-70
11
_^- 29.6 m3/h
5 4 5 6 7 I 9 10
Measurement point
. 45 m3/h _^_ 79 m3/h _^ 101 m3/h _^ 124 m3/h
12
5 6 7 8
Measurement point
10
11
148 m3/h
164m3/h
274 m3/h
Figure 6. Pressure difference between the gravel layer and the
outside in function of the number of the "measurement
point" (see figure 5), for different values of the flow
measured through the fan.
-------
It appears that it is rather easy to depressurize the gravel layer
situated below the slab of room CL."7 : AP is of the order of several tens
of Pa, even for the measurement points which are the most distant from the
extraction fan. The obtained depressions are believed to be sufficient to
prevent the entry of radon from the soil into the room (this is probably
true for most of the values of the flow Q measured across the extraction
fan, but it could be interesting to confirm this by performing some radon
concentration measurements in the room).
The AP values measured at the points 1, 2, 11 and 12 indicate that it is
very difficult to create large depressions below the adjoining rooms CL.6
and CL.8 with an extraction fan connected to the gravel layer situated
below room CL.7. That means that an extraction point should be necessary
for each room of this school. This is probably due to the fact that the
walls of the rooms penetrate deeply in the soil, thus preventing the
connection between the gravel layers situated below the slabs of the rooms.
In conclusion, the subslab depressurization technique seems rather
easy to apply with success when, as it is the case for some of the rooms of
this school, a gravel layer is present below the slab. However, one has to
keep in mind that, for the particular case of the studied building, a
suction point will be necessary for each one of the rooms. Moreover, the
method may not be applied to all the rooms of the ground level.
DISCUSSION AND CONCLUSION
As a consequence of the results of the investigations presented in the
previous paragraphs, the use of the pressurization method was suggested to
the school authorities. It was decided to seal the cracks and joints
between the floors and the walls, and to install mechanical ventilation
systems (air supply !) in rooms CL.l to CL.9. The characteristics of these
systems were defined according to the indications obtained during the
evaluation of the effectiveness of systems RENS and VENT presented above. A
special attention was given to the comfort aspects : noise, draught
problems, ... .
All the systems will be connected to an automatic time-switch and will
be used in the daytime, i.e. during a period of time slightly longer than
the occupation period of the rooms. Indeed, radon concentrations
measurements performed with an active detector indicate that it is more
than enough to start the systems a few hours before the arrival of the
occupants.
It is clear that an over-pressure will be difficult to obtain in a room in
which the occupants let the door and the windows open! This is a
disadvantage of the method : the need of an active co-operation of the
occupants.
In addition to their effectiveness in reducing the radon
concentrations in this school, these air-supply systems will reduce the CO2
levels in the rooms. This is an advantage. In fact, the solving of the
radon problem should be seen in a large context, the final objective being,
in our opinion, to improve the indoor air quality and climate. From this
point of view, several aspects should be taken into account, together with
the radon levels : combustion products levels, odour levels, thermal
comfort, When possible, solutions aiming to solve several problems
at the same time, without creating other problems, should be considered in
first instance.
-------
ACKNOWLEDGEMENTS
This work was partly financed by the Radiation Protection Programme of
the Commission of the European Communities.
The authors are indebted to Messrs. Bossicard, R., Hotton, M, and
L'Heureux, D. for assistance during some of the experiments, and to Dr. A.
Poffijn and Dr. F. Tondeur for fruitful discussions.
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
Poffijn, A., Charlet, J.M., Cottens, E., Hallez, S., Vanmarcke, H. and
Wouters, P. Radon in Belgium : the current situation and plans for
the future. In : Proceedings of the 1991 International Symposium on
Radon and Radon Reduction Technologies. Philadelphia, USA. 1991.
Vol. 5, VI, p. 7.
Poffijn, A., Uyttenhove, J., Drouget, B. and Tondeur, F. The radon
problem in schools and public buildings in Belgium, submitted to
Radiation Protection Dosimetry as part of the Fifth International
Symposium on the Natural Radiation Environment. Salzburg, Austria.
September 22-28, 1991.
Roulet, C.-A. and Vandaele, L. Airflow patterns within buildings :
measurement techniques. IEA, Technical Note AIVC 34. December 1991.
p. 111-72.
Cohilis, P., Kouters, P., Verheyden, J., Vandaele, L., Bossicard, R.,
L'Heureux, D. and Voordecker, P. A case-study concerning radon
problems in schools : Libramont - Belgium. CSTC-WTCB internal report,
November 1990.
5. Cohilis, P., Wouters, P. and Voordecker, P. Intervention du CSTC a
Libramont : mesures de pressurisation et evaluation de systemes de
ventilation dans le cadre de la problematique du radon. CSCT-WTCB
internal report. February 1991.
Saum, D., Craig, A.B. and Leovic, K. Radon mitigation in schools.
ASHRAE journal. January-February 1990.
7. Cohilis, P., Wouters, P. and L'Heureux, D. Prediction of the
performance of various strategies of subfloor ventilation as remedial
action for radon problems. In : Proceedings of the llth AIVC
Conference. Belgirate, Italy. September 1990. Vol. 2, p. 17.
-------
X-7
MULTIPLE MITIGATION APPROACHES APPLIED TO A SCHOOL WITH LOW
PERMEABILITY SOIL
by
D. Bruce Harris
USEPA, AEERL
RTP, NC
and
Eugene Moreau and Robert Stilwell
Maine Department of Human Services
Division of Health Engineering
Augusta, ME
A series of mitigation approaches have been applied to a
amiti-wing elementary school with low permeability soil. Each of
the three wings have been mitigated using a different technique. An
active soil depressurization (ASD) system deployed in one wing was
used to compare the performance of inline axial flow fans and high
vacuum blowers. A high vacuum blower ASD system and a room
pressurzation system using unit ventilators were compared in the
second wing. The third wing used a heat recovery ventilator. The
results of long term monitoring of each system will be discussed.
-------
X-8
GENERAL INDOOR AIR INVESTIGATIONS IN
SCHOOLS WITH ELEVATED RADON LEVELS
By: Terry Brennan
Camroden Associates, Inc.
RD #1, Box 222
Oriskany, NY 13424
Gene Fisher and Bryan Ligman
U. S. Environmental Protection Agency
Office of Radiation Programs
401 M Street S. W.
Washington, D. C. 20460
Richard Shaughnessy
University of Tulsa
Tulsa, OK
William Turner and Fred McKnight
H. L. Turner Group
Concord, NH 03301
ABSTRACT
As part of the EPA School Evaluation Program, several public
schools were visited. The schools were being investigated because
of elevated indoor radon levels. A variety of additional indoor
air measurements and observations were made in the schools. The
goals were to characterize important factors related to general
indoor air quality and to assess the impact of radon control
methods on air quality problems other than radon. Schools were
inspected and measurements were made. Potential sources of
airborne contaminants and conditions that increase the risk of
indoor air problems were cataloged. Measurements of peak room air
carbon dioxide, carbon monoxide, bioaerosols, organic compounds,
and respirable particulates were made. Methodologies for carpet
sampling for fungi and bacteria were developed and tested.
-------
X-9
COMPARISON OF ASP AND HVAC SYSTEM CONTROL IN SCHOOL BUILDINGS
by: Bobby E. Pyle
Southern Research Institute
Birmingham, Alabama 35255
Kelly W. Leovic, Timothy M. Dyess,
and D. Bruce Harris
Air and Energy Engineering Research Laboratory
U.S. Environmental Protection Agency
Research Triangle Park, North Carolina 27711
ABSTRACT
Active soil depressurization (ASD) and heating, ventilating,
and air-conditioning (HVAC) system control are the two most widely
applicable radon mitigation techniques for school buildings.
School officials often need to select which technique is better
suited for a particular building; however, the two techniques have
not been compared previously in the same building.
This paper presents direct comparisons of the radon reduction
capabilities of ASD and HVAC control in two schools. The primary
HVAC systems for both schools are unit ventilators, although one
school also has individual room exhausts and the other has an
office area with a central air handling unit. The schools are
located in South Dakota and Minnesota, and data were collected
using continuous dataloggers. In the South Dakota school the unit
ventilators provided only marginal radon reduction while the ASD
system consistently maintained radon levels below 4 pCi/L. In the
Minnesota school, although both systems maintained radon levels
below 2 pCi/L, the HVAC systems provided slightly better radon
control than ASD when sufficient outdoor air was supplied.
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
Since 1988 the U.S. Environmental Protection Agency's (EPA's)
Air and Energy Engineering Research Laboratory (AEERL) in Research
Triangle Park, NC, has conducted various levels of radon mitigation
research in 50 schools located in 13 states. Mitigation systems
have been installed in 24 of the schools: 14 installations are ASD
systems; 4 are HVAC system control; and 6 are a combination of both
ASD and HVAC control (1).
Initially, AEERL's radon mitigation research in schools
focussed on applying the most successful technique developed for
residential houses — ASD. Because of complicated subslab
structures and subslab fill material that sometimes make ASD
systems expensive to install, and because of indoor air quality
concerns (such as carbon dioxide), radon reduction using HVAC
systems has also been researched. In more recent projects,
continuous datalogging equipment has been used to monitor schools
with HVAC systems — either central HVAC systems or unit
ventilators (UVs) to identify the conditions under which HVAC
systems can be used to control the radon levels. This paper
presents results from two schools in which ASD systems and HVAC
system control were compared in the same building. The schools are
located in South Dakota and Minnesota. Those more familiar with
metric units should refer to the conversion factors listed in Table
1.
SOUTH DAKOTA SCHOOL
Initial radon measurements in the Rapid City area schools were
made from November 1990 to February 1991 using alpha track
detectors (ATDs). Based on these results, eight schools were
identified by AEERL as potential research schools. The school
ultimately were selected for this study, Lincoln Elementary, had
the highest average radon level, 18.8 pCi/L. This school is
located in western Rapid City on a hillside of exposed and
crumbling shale.
Building Description
The original building was constructed in 1951 with 10
classrooms, a cafeteria/gymnasium, and several offices and special
purpose rooms. The original building had approximately 14,280 ft2.
Three classrooms and a library were added to the west side of the
original building in 1957, increasing the total area to 22,132 ft2.
The plan of the building is shown in Figure 1.
The design drawings indicate that the original building has
be low-grade, poured concrete, foundation walls around the perimeter
of the building and under the corridor walls but not between the
-------
classrooms. The slabs are poured on compressed soil covered with
a minimum of 6 inches of gravel. The addition, Rooms 17-20, has a
subslab utility tunnel along the outside walls. However, because
of the lower radon levels in the addition, it was not included in
this investigation.
Floor/wall cracks were observed in many of the classrooms and
are thought to be a major contributor to radon entry.
Unfortunately cabinets and closets typically surround three of the
four walls in each classroom (the fourth wall is the chalkboard)
making access to and sealing of the floor/wall cracks difficult.
One crack observed in Room 16 — a corner room with both the
highest ATD (36.7 pCi/L) and followup measurements (35.9 and 14.1
pCi/L) — was about 0.5 inch wide.
HVAC System Description and Initial Measurements
Each classroom has a UV for heating and ventilating. The
units are located along the outside wall. Most of the classrooms
also have a fan-powered exhaust located in the closet along the
wall parallel to the corridor.
On May 1, 1991, carbon dioxide (CO2) measurements were made in
most of the rooms. The measurements, in most cases, were made
while the rooms were occupied and the hall doors were open. The CO2
concentrations ranged from 700 TO 4000 parts per million (ppm) with
an average value of 1687 ppm. Because the American Society of
Heating, Refrigerating, and Air-Conditioning Engineers (ASHRAE)
recommends CO2 levels below 1000 ppm the researchers recommended
that school personnel make the necessary repairs to bring the UVs
up to design specifications. This was before continuous
datalogging began.
Measurements of UV airflows and differential pressures across
the building shell were made in July 1991. The measurements were
taken under various building operating conditions in all classrooms
in the original building:
• UV speed (high, medium, and low)
• UV outdoor air damper (completely open and completely
closed)
• closet exhaust systems (on and off, each room)
• room-to-corridor door position (open and closed)
For these spot measurements the greatest negative pressure
differentials across the building shell occurred when the hallway
door was closed and the exhaust fan on. With the UV outdoor air
damper open and the exhaust fan operating, the differential
pressure for the classrooms averaged -0.022 inch water column (WC) .
The differential pressure averaged -0.032 inch WC with the UV
outdoor air damper closed.
-------
Operation of the exhaust fans (one for each room) had a
dramatic effect on the pressures in the room and tended to override
any pressurizing abilities of the UVs. The total exhaust measured
for the 10 classroom fans was 5,554 cfm while the maximum outdoor
air capability of the UVs totaled 1,590 cfm — 3.5 times more
exhaust than supply. The exhaust fans are new and draw 630 to
1,021 cfm per room, compared to the old fans that exhausted about
100 cfm per room.
Opening the hallway doors tended to have a neutralizing effect
on the classroom differential pressures. The negative pressures
caused by exhaust fan operation were near zero when the hallway
doors were open.
The UVs were able to pressurize the rooms on average, 0.003
inch WC with the exhaust fan was off and the hallway door closed.
A positive pressure was also observed when the UVs were on, the
exhaust fan was off, and the hallway door was open.
Pre—Mitioation Radon Levels
Two sets of radon measurements were made in May 1991 using
charcoal canisters provided by AEERL. These results are shown in
Figure 1, together with the initial ATD measurements. The values
circled are the Winter 1990-91 ATD results; those in parentheses
were made with the building closed and all UVs off over the weekend
of May 3-5, 1991; the final set of measurements was made with the
building closed and the UVs operating over the weekend of May 17-
19, 1991. The average school radon level with the UVs off was 10.9
pCi/L. The highest level, 35.9 pCi/L, was measured in Room 16.
The lowest level, 1.3 pCi/L, was measured in Room 9. With the UVs
on, the school average radon level dropped to 5.7 pCi/L with a high
of 14.1 pCi/L measured in Room 16 and a low of 0.6 pCi/L in Room
12.
The relative changes in radon levels with the UVs on compared
to UVs off are seen in Figure 2. In half of the rooms (Rooms 1, 3,
11, 13, 16, Office, Gym, Principal's Office, Teacher's Lounge, and
Health Room) the radon levels dropped with the UVs on compared to
the levels measured with the UVs off. This would be expected
because of the outdoor air provided by the UVs. In the other rooms
the levels increased with the UVs operating (Rooms 20, 9, 18, and
the Boiler Room showed the largest increases). The increases in
these classroom radon levels are thought to be due to closed or
inoperable outdoor air intake dampers in those rooms.(As mentioned
previously, the UVs were repaired prior to the continuous
datalogging.) The higher radon level in the Boiler Room might be
due to increased depressurization with the UVs operating.
The outdoor air temperatures were significantly different for
the two measurement periods so these data should be interpreted
cautiously. Over the period May 3-5, 1991 (UVs off) the average
-------
high and low temperatures were 46.7 and 31.7°F, respectively. Over
the period May 17-19, 1991 (UVs on) these average temperatures were
67.3 and 47.3°F. These large temperature differentials during the
two charcoal canister measurements make comparison of the relative
influence of the UVs difficult. The influence of the UVs on radon
levels will be discussed later in this paper based on the results
from the continuous datalogger.
Subslab Pressure Field Extension (PFE) Measurements
PFE measurements were made by AEERL and Southern Research
Institute. The subslab PFE was found to be much better in the
north wing than in the east wing. ASD systems were installed as
shown in Figure 1. Because of the different PFE measured in the
two wings of the building, the systems for the two wings must be
different.
The PFE in the north wing was better than that in the east
wing. As indicated in Figure 1, one suction point was placed on
each side of the corridor, because subslab footings separate the
two areas. One point was placed in Room 14 (to treat Rooms 12, 16,
the Teacher's Lounge, and the Health Clinic), and one point was
placed in Room 11 (to treat Rooms 9, 13, and the Office areas).
Both systems use 8 inch diameter, Schedule 40 polyvinyl chloride
(PVC) piping for both vertical and horizontal pipe runs. Each
suction point uses a fan rated at approximately 300 cfm at 1 inch
WC. The fans are located on the roof, away from any air intakes.
Because of the poorer PFE measured during the diagnostic
testing in the east wing, one suction point was installed in each
of the four rooms. The overhead piping for Rooms 1 and 3 and for
Rooms 2 and 4 were manifolded together so that only two fans were
needed. In Rooms 1 and 3, the piping was run from the rooms across
the hall above the ceiling tile and up through the chase above the
closets in Room 4. The pipes were joined and exited the building
through an old window opening (now closed with plywood sheeting).
The piping from Room 2 was run across the bathrooms and joined to
the piping from the suction point in Room 4. A single pipe was run
around the room to exit through the old window openings above the
closet.
The fan flow rate was expected to be much lower than that in
the north wing, because of the lower PFE. A fan capable of a high
pressure head was also anticipated. Due to the low flow rate and
high pressure expected in these east wing systems, the piping was
much smaller in diameter than the systems in the north wing. All
piping in this wing was 2 inch diameter Schedule 40 PVC. The fans
used in this wing were capable of moving roughly 25 cfm at a static
pressure of around 30-35 inch WC and 50 cfm at 0 inch WC static
pressure.
-------
Datalogger Installation
A computer based datalogger monitoring system was installed in
the school in July 1991. Remote monitoring of the school was
carried out via a telephone modem. Data were downloaded to a
personal computer in Birmingham, Alabama for data analysis and
storage. This system monitored the following parameters every 30
minutes:
• Ten continuous radon monitors measured the radon levels
in Rooms 1, 2, 3, 4, 9, 11, 12, 13, 14, and 16.
• Nine pressure transducers monitored differential
pressures between both the rooms and the subs lab areas in
Rooms 1, 4, 11, and 14. These pressure differentials
were referenced to the pressure in the hallway which in
turn was monitored relative to outdoor pressure.
• The temperatures in Rooms 1, 4, 11, 16, the hallway, and
outdoors were monitored.
• The operation of the UVs in Rooms 1, 4, 11, and 14 was
monitored. This included the on/off times and the
positions of the outdoor air dampers.
• The on/off times of the closet exhaust fans and the
position of the doors (open or closed) for Rooms 1, 4,
11, and 16 were also monitored.
• A weather station was used to continuously monitor wind
speed and direction and precipitation.
Results
The test matrix used to evaluate both the UVs and the ASD
system's effects on radon is shown in Table 2. Testing occurred
November 18, 1991, through January 31, 1992. The results of this
series of tests are summarized in Figures 3, 4, 5, and 6 for Rooms
1, 4, 13, and 16, respectively. In each of these figures the
results are shown first (left) with the ASD off to evaluate the UV
alone and second (right) with the ASD operating to evaluate both
ASD alone and ASD in conjunction with the UVs. The results with
the ASD system off are discussed first, followed by a discussion of
the results with the ASD system on.
Room 1 was the only classroom with a door directly to the
outdoors. In Figure 3 it is quite evident that opening the outside
door was very effective in lowering room radon levels. However,
this approach is not recommended as a practical year-round approach
to radon control. It is interesting to note that opening the hall
door also lowered the radon levels for all modes of operation of
the UV and the exhaust fan. In fact, the radon levels in all four
-------
rooms decreased when the hall door was open. This was most likely
due to increased ventilation due to dilution of the radon
concentrations together with pressure neutralization, reducing the
driving force.
With the ASD systems off (left-hand graphs in Figures 3, 4, 5,
and 6) operation of the closet exhaust fans without the UVs
resulted generally in either lowering the radon levels or having
little effect. One exception is seen in Figure 4 where operation
of the exhaust fan in Room 4 with the hall door open actually
increased the radon levels. Other research has shown that
operation of exhaust fans can either decrease or increase radon
levels, depending on the relative leakage between above grade
(decreases radon levels through infiltration of outdoor air) or
below grade (increases radon levels through infiltration of radon-
containing soil gas).
Operation of the UVs (with ASD and exhaust fan off) reduced
the radon levels in Rooms 1, 4, and 16 by about 10%. However, UV
operation increased the levels in Room 13 by approximately 25%.
These results are not consistent with the charcoal canister results
obtained in May 1991 (shown in Figure 2) . However, during the
period that these continuous radon measurements were taken
(November 1991 to January 1992) the average outdoor high and low
temperatures were 44.2 and 19.9°F, considerably lower than in May,
1991 resulting in less outdoor air supplied through the UV dampers.
It is possible that the UV in Room 13 pulled radon from the
floor/wall crack, distributing it to the room and increasing the
radon levels.
With the ASD systems operating (the right-hand graphs in
Figures 3, 4, 5, and 6), the average radon levels in all classrooms
dropped to less than 3 pCi/L and in some cases to less than 1
pCi/L. The percent reductions ranged from about 80 to 99% with an
average reduction of about 90%. These reductions appear to be
fairly independent of the mode of operation of the UVs and exhaust
fans, indicating that ASD and HVAC together did not outperform ASD
alone. Thus, while the UV and exhaust fan operation may lower the
radon levels in many of the classrooms, the levels were not
consistently reduced to below 4 pCi/L. Only the ASD systems
effectively and reliably maintained radon levels below 4 pCi/L.
MINNESOTA SCHOOL
Nokomis Elementary School was initially researched by the
Midwestern Universities Radon Consortium (MURC) and used in their
May 1991 training course. Because the school presented a good
opportunity to compare ASD and HVAC control in the same school,
AEERL installed a continuous datalogger in the school in January
1992 to monitor the systems installed by MURC.
-------
Building Description and Pre-Mitiaation Radon Levels
Nokomis Elementary is a single story concrete masonry building
with brick veneer and slab-on-grade foundation located in the
eastern part of St. Paul. The total floor area, as shown in Figure
7, is about 15,000 ft2. The average radon level in the school was
4.2 pCi/L measured with charcoal canisters in October 1990. The
highest and lowest concentrations were 6.4 and 2.8 pCi/L,
respectively. Subslab material is packed sand, and the
configuration of the subslab footings is unknown. PFE measurements
indicated that a pressure field could be extended about 15 feet
from the suction point.
The heating and ventilating systems in the school are a
combination of systems. Hot water for these systems is provided by
a gas/oil fired boiler located in the northwest corner of the
building. Each of the nine classrooms (Rooms 101, 102, 103, 104,
105, 106, 107, 108, and 109) are heated with finned-tubes along the
outside walls. In addition, the two classrooms on the south end of
the building (Rooms 107 and 108) have hydronic slab heating (these
classrooms are used as kindergarten rooms). Heating and cooling in
all of the classrooms is provided by UVs located along the exterior
walls. The gym/lunchroom, kitchen, and the office complex are
serviced by a central air handling unit (AHU) with ducted returns.
The unit has provisions for outdoor air intake at roof level.
There are fan powered exhausts in the restrooms and in the kitchen.
A building investigation conducted by MURC indicated that the
outdoor air intake capabilities of the UVs and AHU unit exceeded
the exhaust fan capacities so that building pressurization appeared
to be a viable option for radon control. A fan door test conducted
on the building indicated an effective leakage area of around 500
in.2. This indicates that about 3000 cfm of outdoor air is needed
to pressurize the building adequately to prevent radon entry when
the exhaust fans are on. PFE measurements indicated that a pressure
field could be extended about 15 feet from the suction point.
Implementation of Mitigation Systems
In May 1991 a private contractor of MURC mitigated the school
using three mitigation systems. One system used the UV's in Rooms
101 to 106 to bring in the minimum amount of outdoor air necessary
to pressurize the rooms and to reduce the radon entry. During
normal operation the outdoor air dampers modulate above this
minimum setting as determined by indoor and outdoor temperatures.
This was also attempted in Rooms 107 and 108; however, the radon
levels were not reduced below 4 pCi/L during the initial
evaluation. An ASD system was then installed in these two rooms
with a single suction pit installed in each of the two closets.
The pipes from each point were connected and run to a single, roof-
8
-------
mounted, exhaust fan capable of moving 420 cfm at 0 inch we and
about 180 cfm at 1.6 inch WC. The locations of the ASD points are
shown in Figure 7. Although the Gym/Office area in the north end
of the building has a HVAC system with an outdoor air intake,
mitigation was accomplished through a combination of HVAC control
and another single-point ASD system. The suction point was located
in the Boiler Room very close to the Utility Tunnel. This location
resulted in bypassing the subslab area by pulling air through the
tunnel block walls instead of from the subslab. Post-mitigation
radon levels with all systems operating averaged below 0.5 pCi/L.
Datalogger Installation
In January 1992 a datalogger was installed in the school in
order to determine how the various mitigation systems performed
relative to each other and the interactions between each pair of
systems. Parameters monitored were: radon levels in Rooms 106,
107, 108, and the Principal's office; subslab differential
pressure measurements in Rooms 106, 107, and the Principal's
office; outdoor air damper positions on the UVs in Rooms 106 and
107 and on the office AHU; temperatures in Rooms 106, 107, the
Principal's office, and outdoors; and wind speed and direction.
Results
The test matrix shown in Table 3 was developed to evaluate
each of the mitigation systems. School Officials were unwilling to
turn the ASD systems off while the building was occupied (to test
the HVAC systems alone) . As a result, testing with the ASD systems
off was done over the spring break, April 10 to 20, 1992. These
tests should be representative of winter conditions: the outdoor
temperatures during this testing period ranged from 18 to 43°F, with
an average of 30°F. The testing schedule was:
Test if UVs and AHU at normal damper operation; ASD on: This
test was used to measure the radon levels in the post-
mitigation configuration with the ASD and HVAC systems all
operating. Under normal conditions the outdoor air (OA)
damper modulated based on OA temperature. This is typically
20 to 25% damper open.
Test 2, UVs, AHU, ASD off: To collect baseline (no
mitigation) radon data, all UVs and the AHU were turned off.
The ASD fan outlets on the roof were covered with plastic
garbage bags and sealed with duct tape on Friday afternoon,
April 10, 1992. The ASD systems remained off through Test 6.
Test 3, UVs and AHU on, damper closed: On Monday morning,
April 13, 1992, the UVs in Rooms 101 to 108 and the AHU were
turned on but the outdoor air dampers were kept closed (UVs
and AHU running 24 hours). The school was operated in this
-------
configuration until Wednesday morning, April 15, 1992.
Test 3A, UVs and AHU on, dampers closed, filters out: On
Tuesday afternoon, April 14, 1992, it was found that the
return air filters in the UVs and the AHU were extremely dirty.
The filters were pulled out of the units and the systems continued
to operate as described in Test 3, above.
Test 4, UVs and AHU at 10% damper open: On Wednesday
morning, April 15, 1992, the outdoor air vents were opened to
10% outdoor air on both the UVs and the AHU (running 24
hours). The ASD systems remained off. The school was
operated in this configuration until Friday morning, April 17,
1992.
Test 5, UVs and AHU at 50% damper open: On Friday morning,
April 17, 1992, the outdoor air vents were opened to
approximately 50% outdoor air on both the UVs and the office
HVAC system (UVs and HVAC running 24 hrs) . The school was
operated in this configuration until Saturday afternoon, April
18, 1992.
Test 6, UVs and AHU at normal damper open: On Saturday
afternoon, April 18, 199.2, the UVs the office HVAC system were
changed to normal day/night operation (night setback at 68°F) .
The ASD systems remained off with exhaust pipes covered.
Test 7, UVs and AHU at normal damper operation; ASD on: On
Monday afternoon, April 20, 1992, the ASD system pipes were
uncovered and fans turned on. The UVs and AHUs returned to
the post-mitigation configuration.
Test 8, same as Test 7 (2 weeks later).
The amount of outdoor air introduced into each of the UVs
and the AHU as a function of damper position is shown in Table 4.
Note that in Table 4, the percent damper opening corresponds to
different percentages of outdoor air for each of the four locations
monitored. For example, a 10% damper opening corresponds to 27,
38, 34, and 19% outdoor air for Rooms 106, 107, 108, and the
Office, respectively. For simplicity, the remainder of this
discussion will refer to the percent damper opening rather than the
percent outdoor air. Refer to Table 4 for the specific quantity of
outdoor air for a given location and damper position.
The results of the nine tests are shown in Figure 8. The
average radon levels for the Principal's Office, Room 106, and Room
107 are plotted for each of the test conditions. Radon levels
increase from Test 1 to 2 because of stack effect induced radon
entry. With the ASD systems operating (Test 1), the stack effect
is reversed. The operation of the UVs with outdoor air also helps
10
-------
to reduce the stack effect.
The results shown under Test 3 and 3A are conditions in which
the UVs and AHU are depressurizing the building because the outdoor
air dampers are completely closed. In addition to building
depressurization, it is also possible that the radon levels
increased from Test 2 to Tests 3 and 3A because background radon
levels continued to build up in the building after turning off the
ASD systems (Test 2) .
With minimum outdoor air (approximately 10%, damper opening in
Table 4) the radon levels were reduced below 4 pCi/L during Test 4.
With about 50% damper opening (Test 5) , the radon levels are
reduced to background levels. In fact, the levels during Test 5
were lower than normal UV and AHU operation together with the ASD
system operation. In Test 6 the UVs and the AHU outdoor air
dampers were returned to the post-mitigation setting of
approximately 20 to 25% outdoor air. Turning on the ASD systems in
Test 7 reduced the levels in the Office and in Room 107 but had
little effect in Room 106 since there is no ASD system in Room 106.
Test 8 presents continuous data collected about 14 days after
Test 7. Radon levels in the two areas with ASD systems are
relatively consistent with those observed during Tests 1 and 7.
The radon levels in Room 106 (without the ASD system) are slightly
lower than in Tests 1 and 7.
The lowest radon levels in these three rooms were observed
during Test 5 when the UVs and AHU were operated at 50% damper
position (refer to Table 4 for actual percent outdoor air) and the
ASD systems were off. However, it is often difficult to
consistently depend on 50% damper position when outdoor
temperatures drop below freezing. The UVs do maintain radon levels
below 4 pCi/L in all three rooms as long as the damper position is
set to 10% outdoor air (Test 4).
Results show that radon levels are consistently reduced below
1 pCi/L in Rooms 107 and the Office when the ASD systems for these
areas are operating. During these tests, the radon levels in the
rooms with the ASD systems were always lower than in the room with
the only UV control (Room 106) . The effectiveness of the ASD
system alone is currently being evaluated.
11
-------
CONCLUSIONS
In Lincoln, the UVs alone were not able to consistently
maintain radon levels below 4 pCi/L. The exhaust fans and the
opening of the classroom-to-hallway doors had varying effects on
radon levels. The ASD systems, however, provided excellent radon
control, even better than that predicted during the diagnostic
tests. Operation of the UVs together with the ASD system did not
provide much additional radon reduction over the ASD system alone.
The UVs were effective in reducing the radon levels to below
4 pCi/L in Nokomis if a minimum amount of outdoor air was supplied.
If the damper position was set at 50%, then average radon levels
were reduced below l pCi/L. Radon levels were consistently reduced
below 1 pCi/L in Rooms 107 and the Office with the ASD systems
operating and the HVAC system set to normal outdoor air. During
these tests, the radon levels in the rooms with the ASD systems
were always lower than in the room with the only UV control. The
effectiveness of the ASD system alone is currently being evaluated.
If the PFE measurements indicate that a properly designed ASD
system will be effective, this would be the preferred system for
consistent radon control. If in addition, improvement in indoor
air quality or further radon reduction is desired, the amount of
outdoor air supplied through the HVAC system should be increased.
This will help to approach the long-term national goal of ambient
radon levels in buildings established in the 1988 Indoor Radon
Abatement Act while also improving indoor air quality.
REFERENCE
1. Leovic, K.W., Craig, A.B., and Harris, D.B. Update on
Radon Mitigation Research in Schools. Paper presented at
The 1991 Annual American Association of Radon Scientist
and Technologists National Fall Conference, Rockville,
MD. October 9-12, 1991.
ACKNOWLEDGEMENTS
The authors would like to express their appreciation to the
Rapid City and St. Paul school officials who assisted in making
these projects possible. Thanks also to Bill Angell of MURC who
assisted with the project in St. Paul.
12
-------
TABLE 1. METRIC CONVERSION FACTORS
Non-Metric
cubic foot per minute
(cfm)
degrees Fahrenheit (°F)
foot
inch
inch water column (WC)
picocurie per liter
(pCi/L)
square foot (ft2)
square inch (in.2)
Times
0.47
5/9 (°F-32)
0.305
2.54
249
37
0.093
6.45
Yields Metric
liter per second
degrees Centigrade
meter
centimeter
pascals
becquerels per cubic
meter
square meter
square centimeter
13
-------
TABLE 2. TEST MATRIX FOR LINCOLN ELEMENTARY SCHOOL
Test
No.
1
2
3
4
5
6
7
8
9
10
11
Test
Dates
11/18
to
24/91
11/25
to
29/91
12/2
to
6/91
12/9
to
16/91
12/16
to
20/91
12/30
to
1/3/92
1/6
to
10/92
1/13
to
17/92
1/20
to
24/92
1/27
to
31/92
12/23
to
27/91
Class
Room
Doors
Normal1
Normal
Closed4
Closed
Closed
Closed
Normal
Normal
Closed
Closed
Closed
Unit
Ventilator
operation
Normal2
Normal
Normal
Normal
Normal
Normal
Normal
Normal
Normal
Normal
Off
Exhaust
Fan
Operation
Off
On
Off
On
Off
On
Off
On
Off
On
On
Outdoor
Air-Damper
Position
Normal3
Normal
Normal
Normal
Open5
Closed6
Normal
Normal
Normal
Normal
Normal
ASD
System
Operation
Off
Off
Off
Off
Off
Off
On
On
On
On
Off
Notes: 1 Determined by teacher preference
2 On 4:00 am - 8:00 pm weekdays, off over weekend
3 Controlled by outdoor/inside temperature
4 Teachers asked to keep closed as much as possible
5 OA damper mechanically blocked open
6 OA damper mechanically blocked closed.
14
-------
TABLE 3. TEST MATRIX FOR NOKOMIS ELEMENTARY
Test
Number
1
2
3
3A
4
5
6
7
8
uv
(On/Off)
On
Off
On
On
On
On
On
On
On
UV OA
(%>
Normal
Normal
0
0
10
50
Normal
Normal
Normal
AHU
(On/off)
On
Off
On
On
On
On
On
On
On
AHU OA
(%)
Normal
Normal
0
0
10
50
Normal
Normal
Normal
ASD
(On/Off)
On
Off
Off
Off
Off
Off
Off
On
On
Comments
Filters
Out
Two
Weeks
Later
15
-------
TABLE 4. OUTDOOR AIR FLOW VERSUS DAMPER POSITION FOR UVS
AND AHU UNIT IN NOKOMIS ELEMENTARY
Room *
106
106
106
106
107
107
107
107
108
108
108
108
Office
Office
Office
Office
Outdoor Air
Damper Percent
Open
0
10
50
100
0
10
50
100
0
10
50
100
0
10
50
100
Total Unit
Flow (cfm)
900
990
1020
950
926
1109
1050
1009
1050
1190
1200
1205
2025
2160
2280
1960
Outdoor Air
Flow (cfm)
0
265
320
358
0
420
650
660
0
400
513
642
0
420
500
1100
Percent
Outdoor Air
0
27
31
38
0
38
62
65
0
37
43
53
0
19
22
56
Note that Rooms 106,
an AHU.
107, and 108 nave UVs and the Office has
16
-------
To High Pressure
Fans on Roof
Suction
Points
N
Fan (T3B)
On Roof
Figure 1. Floor plan of Lincoln Elementary School, Rapid City, SD,
showing radon levels and mitigation systems.
17
-------
CO
o
o
o
03
EC
LLI
CD
O
o
CC
20
19
18
Library
16
Health Room
14
13
12
11
Boiler Room
Teacher's Lounge
Principal's Office
Gym
Office
4
3
2
1
-100 -50 0 50 100 150 200
CHANGE WITH UV ON COMPARED TO OFF (%)
250
Figure 2. Changes in radon levels with UVs on compared to off
using CCs at Lincoln Elementary on 5/91 (data missing for Rms 12 & 14).
18
-------
fcFFECIS OF ROOM OPERATING CONDITIONS ON
RADON LEVELS IN ROOM 1, ASD OFF. WEEKDAYS ONLY
EFFECTS OF ROOM OPERATING CONDITIONS ON
RADON LEVELS IN ROOM 1. ASD ON. WEEKDAYS ONLY
UV-CkCXO UV-O/EX-1 IN-1/EX-0 UV.1/EX-1
Unrt V«ntlfator/Exh Fan Oporatlon
Outoki* Door CUMOHU Door Cto*«l
O*»U» Door CtoMiittti Door Open
OvMtto Door optnftWI door OP«I
Ou»W« Ooor Cloud/Hall Door Cloud
Ouaidt Door OOM4/HUI Ooor Op«n
Outrtrt Deer Op*n/H«l Door Oo^d
uv.o/Ex-o uv-o/W-i uv-iyix.o tv-i/tx-t
Unit Ventilator/Exh Fan Operation
Figure 3. Comparison of HVAC and ASD in Room 1, Lincoln Elementary.
-------
bF-HtCIS OF ROOM OPERATING CONDITIONS ON
RADON LEVELS IN ROOM 4, ASD OFF, WEEKDAYS ONLY
' I
'
W-CVEX-O UV-WEX-1 W-1/EX-0 LV-1/EX-I
Unit V»ntII«tor/Exh Fan Operation
MMDoorOpwi
EFFECTS OF ROOM OPERATING CONDITIONS ON
RADON LEVELS IN ROOM 4. ASD ON. WEEKDAYS ONLY
,
'
'
u
s
•
I
•
UV-O/tX-O UV-O/EX-1 UV.l/EX-0 UV>1/EX-1
Unit Ventllator/Exh Fan Operation
KWOoorCloMd
HID Doof Op«n
Figure 4. Comparison of HVAC and ASD in Room 4, Lincoln Elementary.
-------
:
EFFECTS OF ROOM OPERATING CONDITIONS ON
RADON LEVELS IN ROOM 13. ASD OFF, WEEKDAYS ONLY
UV-WEX-O UV-WEX-1 W-1/tX-O UV-1/EX-1
Unit Ventilator/Exh Fan Operation
Hd Door Opm
EFFECTS OF ROOM OPERATING CONDITIONS ON
RADON LEVELS IN ROOM 13. ASD ON, WEEKDAYS ONLY
!»
II
M
'
•
,
.
t
^^ ^m jam
~~L_ _T X /j&BK&y X-fJHBKraJr XH
fPfW / / SJy X vSy XH-,C
UV-0/EX.O UV-O/EX-I OV-1/EX-O UV-1/EX-1
Unit Ventilator/Exh Fan Operation
Figure 5. Comparison of HVAC and ASD in Room 13, Lincoln Elementary.
-------
EFFECTS OF ROOM OPERATING CONDmONS ON
RADON LEVELS IN ROOM 18. ASD OFF, WEEKDAYS ONLY
JO
'
i >
• •
UV-CVEX-1 UV-1/EX-0 UV-l/tX-1
Unrt V«n«later/Exh Fan Operation
HMOoordOMd
HaOewOpm
EFFECTS OF ROOM OPERATING CONDITIONS ON
RADON LEVELS IN ROOM 16, ASD ON. WEEKDAYS ONLY
i"
u
:
9
.
Hen Door dowa
UV-O/EX-O UV-O/tX.1 UV.T/EX-0 UV.1/EX-1
Unit Ventilator/Exh Fan Operation
Figure 6. Comparison of HVAC and ASD in Room 16, Lincoln Elementary.
-------
UNIT VENTILATORS
CO
r
-i ^
107
108
c
/
\
)
X
AbU
"---AS
PIT
RR RR
109
i i
DP1T
105
106
103
104
101
11
L
1fl9
GYM
UTILITY
TUNNEL
STOR
KIT
J
I "• I
STOR
OFF
BOILER
ROOM
Asp pn
•2f^^^
^
Ml IR9F
RR
OFF LNG
Figure 7. Floor plan of Nokomis Elementary School, St. Paul, MN,
showing radon mitigation system locations.
-------
Tast Number.
l-UVi and AHU «l NormaJ Damper Operation; ASO On
2-UV». AHU, ASD Off
3-UV* and AHU On. Damper Closed
3A-UV» and AHU On. Dampen Closed. Piters Out
4-UVs and AHU at 10% D»mper Open
S-UV* and AHU at 50% Damper Open
6-UVs and AHU at Normal Damper Open
7-UV. and AHU at Normal Damper Operation: ASD On
8- Same as Te*1 1 (Two Weeks Later)
C
Office
3A 4 S
Test N.aibCT
^^ ROOM 106 2>£
ROOK 107
Figure 8. Radon levels during the testing period at Nokomis
Elementary School.
24
-------
X-10
EFFECTIVENESS OF HVAC SYSTEMS FOR RADON CONTROL IN SCHOOLS
by: Kelly W. Leovic, A. B. Craig, and Timothy M. Dyess
Air and Energy Engineering Research Laboratory
U.S. Environmental Protection Agency
Research Triangle Park, North Carolina 27711
Bobby E. Pyle
Southern Research Institute
Birmingham, Alabama 35255
ABSTRACT
Research has demonstrated that heating, ventilating, and air-
conditioning (HVAC) systems in schools can be used to reduce indoor
radon levels if adequate outdoor air is supplied. The U.S.
Environmental Protection Agency's (EPA's) Radon Mitigation Branch
(RMB) has been researching the ability of HVAC systems to reduce
radon levels since 1988. Continuous dataloggers have been
installed in 10 schools over the past 2 years to better understand
the influences of various HVAC system operating conditions on radon
levels. This paper presents the results from a Columbus, Ohio,
school where a central HVAC system was used to reduce elevated
radon levels. The results from this project, together with other
HVAC research projects, are applied to address three general
questions regarding radon mitigation with HVAC systems:
1) Can HVAC systems be used to reduce radon levels from above
20 to below 4 pCi/L?
2) Can HVAC systems be used to maintain radon levels below 4
pCi/L year round in cold climates?
3) How do other building components (such as operation of
exhaust fans and internal doors) affect HVAC system control of
radon?
This paper has been reviewed in accordance with the U.S. EPA's
peer and administrative review policies and approved for
presentation and publication.
-------
INTRODUCTION
EPA's Air and Energy Engineering Research Laboratory (AEERL)
has conducted radon mitigation research in 50 school buildings
since 1988 (1). Research in many of these schools has addressed
radon reduction using active soil depressurization (ASD) systems.
Results indicate that ASD can be successfully applied for radon
reduction in about 75% of U.S. school buildings.
AEERL has also conducted research on the use of heating,
ventilating, and air-conditioning (HVAC) systems for radon
reduction. Using the HVAC system to control radon can be
beneficial for the following reasons:
• radon reduction in the estimated 25% of U.S. schools where
ASD is not applicable,
• supplemental radon reduction in schools where ASD systems
are installed in order to further reduce radon levels (to
reach the long-term national goal of ambient radon levels
established in the 1988 Indoor Radon Abatement Act), and
• improved indoor air quality in addition to radon reduction
through the introduction of additional outdoor air.
Use of the HVAC system as a radon control technique depends on
the specific building, but in general, it may be considered in any
school that has a HVAC system that supplies outdoor air.
Specifically, use of the existing HVAC system to control radon
levels is not reasonably applicable in schools where:
• the existing HVAC system is not designed to supply
conditioned outdoor air (e.g., exhaust-only system, radiant
heat, or fan coil units),
• the existing HVAC system does not consistently supply
outdoor air during all seasons, and
• radon control/indoor air quality concerns in the school
system are overridden by energy cost concerns.
Since 1990, AEERL has researched HVAC control of radon in 10
schools. Central HVAC systems and unit ventilators have each been
extensively researched in four schools (for a total of eight
schools), and specialized units (a heat recovery ventilator and a
solar ventilator) have been researched in two other buildings.
Dataloggers were installed in these schools to continuously monitor
parameters such as radon concentration, differential pressure,
differential temperature, percent open of outdoor air damper,
operation of exhaust fans, and opening and closing of doors in the
building. AEERL's datalogging system is described in Reference 2.
This paper presents recent data collected in one of the
research schools with a central HVAC system. The school is located
in Columbus, Ohio, and research was conducted from spring 1991 to
-------
spring 1992. These recent data, together with data presented
previously (3) and at this Symposium — Comparison of ASD and HVAC
System Control in School Buildings (4) and Multiple Mitigation
Approaches Applied to a School With Low Permeability Soil (5) —
are summarized in the conclusion of the paper to address three
questions on radon reduction with HVAC systems:
1) Can HVAC systems be used to reduce radon levels from above
20 to below 4 pCi/L?
2) Can HVAC systems be used to maintain radon levels below 4
pCi/L year round in cold climates?
3) How do other building components (such as operation of
exhaust fans and internal doors) affect HVAC system control of
radon?
RESULTS: COLUMBUS, OHIO, SCHOOL
The following sections cover a description of the building,
diagnostic measurements, and analyses of the continuous data
collected in the school. Readers more familiar with metric units
should refer to the conversions in Table 1.
BUILDING DESCRIPTION
Structure
The original school building was constructed in 1966 and
includes 13 classrooms, offices, and a multipurpose room. The area
of this slab-on-grade building is 24,000 ft2. A slab-on-grade
addition was constructed in 1973, bringing the total area to 31,000
ft2. A floor plan of the school is displayed in Figure 1. A subslab
utility tunnel runs under three corridors in the school. The
tunnel walls are constructed of unpainted concrete blocks,
facilitating airflow from the soil to the tunnel.
HVAC Systems
The original building has 23 fan-coil units (FCUs) located in
the subslab utility tunnel. The utility tunnel runs under the
corridors to the east, west, and south, as shown in Figure 2.
Supply air for the FCUs is distributed to the tunnel by a central
fan located in a fan room adjacent to the tunnel. The outdoor air
supplied from the central fan ranges from 0 to 100%. Air in the
tunnel (supplied by the central fan) is then supplied to each
classroom by the FCUs. The air either passes through hot water
coils for heating or bypasses the coils. The air from each FCU is
then supplied to the classroom via a subslab duct which then
distributes the air through registers located along the outside
-------
walls of the classroom. Return air from the classrooms passes
through openings in the classroom-to-corridor wall and into the
corridor. The return air is then pulled into a centrally located
return air grille in the corridor near the central fan (Figure 2).
During normal building operation (7 am to 6 pm) the central fan and
the FCUs run continuously.
This type of HVAC system is sometimes referred to as a "face
and bypass" system. Although a number of schools in the Columbus
area have this type of system, it has not been observed by EPA in
schools in other parts of the country. Central HVAC systems with
subslab supply and/or return air ductwork have, however, been
observed in many of EPA's research schools (1).
The 1973 addition has a central HVAC system for heating and
cooling. All of the ducting for this system is located overhead.
Since premitigation radon measurements indicated that the radon
problem was more urgent in the original building than in the
addition, the addition was not part of this research project.
PreMitiqation Radon Levels
initial radon measurements in the school were made in December
1990 by the Columbus Department of Health. The measurements were
made with E-Perms and averaged 11.1 pCi/L, with a maximum reading
of 20.5 pCi/L. Follow-up charcoal canister measurements were made
the weekend of March 22-25, 1991, when the HVAC system was off.
The average radon level was 4.5 pCi/L with a high of 8.4 pCi/L
measured in the east tunnel. The weather was relatively mild
during this measurement period, with a high of 72°F and an average
of 56°F. One would expect these mild weather conditions to reduce
the stack effect in the building. A reduced stack effect would
result in reduced radon entry compared to colder weather.
DIAGNOSTIC MEASUREMENTS
Initial diagnostic measurements conducted in March 1991
indicated that the utility tunnel and associated air distribution
system were the major contributing factors to elevated radon levels
in the school. As a result, the focus of this research project
was on the effect of HVAC system operational parameters on radon
levels. Pressure field extension (PFE) measurements were also
conducted as part of the diagnostics and will be presented in a
forthcoming EPA report.
Inspection of the building HVAC system indicated that the
system might not be operating per design. Reduced system
maintenance due to budget limitations probably contributed to this
situation. To obtain information on the current operation of the
system, a local testing and balancing company was hired to measure
system airflows in May 1991. A 60 point traverse on the suction
side of the central fan was made with the fan set at its design of
-------
550 revolutions per minute. Results showed that the fan was
running at 27, 264 cfm: 264 cfm above the design specification of
27,000 cfm. The results of the airflow measurements (actual
measurements) for all 23 FCUs, together with the design airflow,
are shown in Table 2. These measurements show that, although the
overall airflow for the 23 FCUs is only 233 cfm above design, there
is wide variation from design for most of the individual units.
During the following winter (1991-92), Columbus Public Schools
maintenance personnel adjusted the FCUs to their original design
airflow (third column of Table 2). The central air handling fan
was set to maximum opening position with no change in operating
speed and locked in place. These adjustments of the HVAC fans
helped to reduce the depressurization in the tunnel. School
maintenance personnel also applied two coats of paint to the tunnel
walls and caulked all floor and walls cracks in the tunnel. The
authors were not informed of these changes (adjustments to the HVAC
system and sealing of the tunnel) beforehand and, as a result, were
not able to quantify the effect of the individual changes on radon
levels.
ANALYSIS OF CONTINUOUS DATA
During the spring of 1991, the school was instrumented with a
continuous datalogger to record:
• radon concentrations (eight locations)
• differential pressure (seven locations)
• temperature (seven locations)
• percent open of return and outdoor air dampers (central fan)
• weather (humidity, wind speed and direction, rainfall)
Initial plans were to collect data during the spring of 1991 with
the central HVAC fan supplying specified quantities of outdoor air.
The test matrix for these measurements is shown in Table 3.
Unfortunately, school officials were unable to adjust the outdoor
air damper as required for the testing, so testing of the
parameters in Table 3 was delayed until the winter of 1991-92.
Note that the percent open of return air and outdoor air
dampers in this paper refers to the actual damper position.
Airflow versus damper position would be expected to be non-linear,
with the greatest increase in airflow in the initial increments of
damper opening. For example, if the outdoor air damper is open to
50%, the quantity of outdoor air likely exceeds 50%. Calculations
to determine the quantities of outdoor air associated with the 50%
damper position have been made and are to be presented at the
Symposium.
Data were collected for the test matrix both before and after
school personnel adjusted the HVAC system and sealed the tunnel.
The results are discussed below.
-------
Effect of Outdoor Air Damper Position on Radon — Before
Adjustments
Figure 3 shows the effect of the outdoor air damper position
on radon levels in the school. These data were collected prior to
adjustments to the HVAC system and sealing of the tunnel; however,
it is possible that some of the work was initiated during the final
test (with 100% outdoor air). Since school officials did not
inform the researchers of the work until it was underway, the exact
timing is not known.
The top graph in Figure 3 shows radon levels in the west and
east tunnels, and the bottom graph shows radon levels in five
classrooms and the teacher's lounge. The average radon levels
during each of the three test conditions in Table 3 (100% return
air, 50% return air/50% outdoor air, and 100% outdoor air) were
calculated from the continuous data collected during the test
condition. Three conclusions are apparent from these data:
(1) With 100% return air, average radon levels are much higher
than the previous premitigation E-Perm and charcoal canister
measurements, averaging over 20 pCi/L.
(2) Average radon levels in the six rooms (bottom graph)
closely track radon levels in the tunnels (top graph). The
levels in the rooms tend to be slightly lower than in the
tunnels, supporting our assertion that the tunnels are the
primary source of radon in the school.
(3) Radon levels are reduced by about 75% when the outdoor air
damper position is increased from 0 to 50%. No additional
reduction is observed when the outdoor air damper position is
increased from 50 to 100%. However, the data collected during
the 50% return air/50% outdoor air damper conditions are
limited to a 24 hour test and, thus, should be interpreted
cautiously.
(4) Although radon levels are reduced by about 75% when the
outdoor air is increased to 50 or 100%, average radon levels
in the classrooms still exceed 5 pCi/L.
Effect of Outdoor Air Damper Position on Radon — After Adjustments
The test matrix was then repeated after school maintenance
personnel adjusted the HVAC system to design specifications and
painted the tunnel walls. The results from these tests are shown
in Figure 4. These data show similar trends in radon levels as
Figure 3 (before adjustments). However, average radon levels are
about 6 pCi/L lower in the west tunnel and 1 pCi/L lower in the
east tunnel after sealing. These reductions in tunnel radon levels
are also reflected in the room radon levels. During the test run
with 50% return air/50% outdoor air, radon levels in the rooms
-------
average about 4 pCi/L. When the outdoor air is increased to 100%
average radon levels in the classrooms are about 2.5 pCi/L.
Figure 5 summarizes the average radon levels in the two
tunnels and six rooms after adjustment of the HVAC system and
sealing of the tunnel. The percent reduction attributed to these
changes is shown above each bar. All reductions exceeded 10% with
a high of 57% reduction when 100% outdoor air was supplied.
Prior to these adjustments, the differential pressure between
the tunnel and the outdoors was negative to neutral (depending on
the percent outdoor air). After the HVAC system was adjusted and
the tunnel sealed, the differential pressures under all three
outdoor air conditions were relatively neutral. This reduction of
negative pressure reduced radon entry into the tunnels and
consequently, radon levels in the school.
CONCLUSIONS
The data from the Columbus school presented in this paper,
together with data from other schools researched by AEERL, are
applied to address three questions on radon reduction with HVAC
systems:
1) Can HVAC systems be used to reduce radon levels from above
20 to below 4 pCi/L? Data from this school showed that radon
levels could only be reduced from above 20 pCi/L to below 4
pCi/L when the HVAC system was adjusted to its design
specifications, 100% outdoor air was supplied to the central
fan, and tunnel walls were sealed. Although radon reductions
were about 75% prior to adjustment of the HVAC system and
sealing of the tunnel, levels were not reduced to below 4
pCi/L even with 100% outdoor air. Data from other schools
(3,4,5) support the conclusion that is it difficult to reduce
radon levels from above 20 pCi/L to below 4 pCi/L using only
the HVAC system.
2) Can HVAC systems be used for radon control year round in
cold climates? The research project in this school indicated
that, when 100% outdoor air is introduced and major radon
entry routes are sealed, radon reduction with the HVAC system
is consistent. However, school personnel indicated that this
condition is difficult to achieve during extremely cold
weather due to concern with energy costs. Similar
observations have been made in other research schools (3,4,5) ,
particularly in schools with unit ventilators.
3) How do other building components (such as operation of
exhaust fans and internal doors) affect HVAC system control of
radon? Other building components were not monitored in this
study. However, research in other schools (3,4,5) have shown
-------
that (a) opening the classroom-to-corridor door affects
classroom radon levels (may increase or decrease depending on
the school), and (b) operation of exhaust fans may increase or
decrease classroom radon levels depending on the building.
REFERENCES
1. Leovic, K.W., A.B. Craig, and D.B. Harris. Update on Radon
Mitigation Research in Schools. EPA-600/D-91-229 (NTIS PB91-
242958). Presented at the 1991 Annual AARST National Fall
Conference, Rockville, MD, October 1991.
2. Harris, D.B. and B.E. Pyle. Data Logging Systems for
Monitoring Long-Term Radon Mitigation Experimental Programs in
Schools and Other Large Buildings. Presented at the 85th
Annual AWMA Meeting, Kansas City, MO, June 21-26, 1992.
3. Leovic, K.w., D. B. Harris, T. M. Dyess, B. E. Pyle, T. Borak,
and D. W. Saum. HVAC System Complications and Controls for
Radon Reduction in School Buildings, Presented at the 3rd
International Symposium on Radon and Radon Reduction
Technology, Philadelphia, PA, April 2-5, 1991.
4. Pyle, B.E., K.W. Leovic, D.B. Harris, T.M. Dyess, and A.B.
Craig. Comparison of ASD and HVAC System Control in School
Buildings. To be presented at the 4th International Symposium
on Radon and Radon Reduction Technology, Minneapolis, MN,
September 1992.
5. Harris, D.B. and E. Moreau. Multiple Mitigation Approaches
Applied to a School With Low Permeability Soil. To be
presented at the 4th International symposium on Radon and
Radon Reduction Technology, Minneapolis, MN, September 1992.
ACKNOWLEDGEMENTS
The authors would like to thank David Sands of the Columbus
Public Schools (Architecture and Engineering Department) for his
assistance with this project; Harry Grafton of the Columbus
Department of Health for conducting radon and differential pressure
measurements; and Bill Angell of the Midwest Universities Radon
Consortium for his assistance during the diagnostic measurements in
Columbus.
8
-------
TABLE 1. METRIC CONVERSION FACTORS
Non-Metric
cubic feet per minute
(cfm)
degrees Fahrenheit (°F)
picocuries per liter
(pCi/L)
square feet (ft2)
Times
0.47
5/9 (°F-32)
37
0.093
Yields Metric
liters per second
degrees Centigrade
becquerels per cubic
meter
square meters
-------
TABLE 2. AIRFLOWS FOR FCUS
Room Served by FCUs
1
2
3
4
5
6
7
8
9
10
11
12
Multipurpose Room
Stage
Teacher's Lounge
Music Room
Principal's Office
Supply Room
Conference Room
Office
Nurse's Office
TOTALS
Measured
Airflow (cfm)
868 & 601
705 & 816
1305
1260
1170
1435
1407
1335
1192
1210
1269
1318
3645
839
541
568
345
361
309
476
308
23,283
Design
Airflow (cfm)
1700
1700
1200
1200
1200
1200
1200
1200
1200
1200
1200
1200
4000
1000
450
700
350
300
300
350
200
23,050
10
-------
TABLE 3. TEST MATRIX FOR SCHOOL
Test
No.
I
II
III
Duration
1 week
1 week
1 week
HVAC System
Operation (hours)
6:30 am - 4:30 pm
6:30 am - 4:30 pm
6:30 am - 4:30 pm
Return
Air (%)
100
50
0
Outdoor
Air (%)
0
50
100
Notes: FCUs operate when HVAC system operates.
HVAC system hours of operation are approximate.
11
-------
5.3
(tunnel)
N
5.3
4.9
<0.5
<0.5
C0.5 \
5.5
<0.5
-------
N
u
Return Air
Damper
^^^_
Main
Air Handling x
Unit
Lt/
•*•**>.
1
•••
Outdoor Air
Damper
r
*•
^sia^^
--* Filters
c=»ol
i ••
,
Fan Coil
J (Typical)
•*«*» Supply Air
(/"
^
J
r^
-*^ Supply Air
(To Fan Coils)
^ Return Air
>s (From Upstairs)
Figure 2. Layout of subslab utility tunnel,
13
-------
20
1
10
100% O A
HVAC Operating Condition
* - Limited data for chit condition
20
-4
c 10
cc
100%RA
100%OA
HVAC Operating Coadinoa
- Limited dan (or this condition
Figure 3. Effect of outdoor air damper on radon levels
(before sealing). Top, tunnels; bottom, classrooms
(OA = outdoor air; RA = return air.)
14
-------
23
.
1
i
1
HVAC Operating Condition
W-Tunn«l
100%OA
2C
§
a is
1
•
10
100%RA
50%RA/50%OA
HVAC Operating Condition
lOCK^OA
Figure 4. Effect of outdoor air damper on radon levels
(after sealing). (OA = outdoor air; RA = return
air.)
15
-------
J 20
0
-
-
t
i
:
:
15
10
* J
17
10O&RA 50%RA/50%OA 100%OA
HVAC Operating Condition
Qassrooms V//J Tunnel
Note: Numbers above bars are percent reduction due to sealing tunnel
Figure 5.
Average radon levels in tunnels and in
classrooms before sealing and percent radon
reduction due to sealing. (OA — outdoor air;
RA = return air.)
16
-------
X-11
RADON PREVENTION IN CONSTRUCTION OF SCHOOLS AND
OTHER LARGE BUILDINGS—STATUS OF EPA'S PROGRAM
by: Alfred B. Craig, D. Bruce Harris,
and Kelly W. Leovic
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 (AEERL) of
the U.S. Environmental Protection Agency (U.S. EPA) started a radon
mitigation research, development, and demonstration program in
1985. Initial studies were on existing and new houses, and the
program was expanded in 1987 to include mitigation studies in
existing schools. AEERL has conducted various levels of radon
mitigation research in 49 existing schools in 13 states since that
date.
In 1990, a decision was made to issue a technical guidance
document that describes how to design and build schools and other
large buildings which will be radon resistant and easy to mitigate.
Tentative design premises were developed, based on studies in
existing schools, and presented at The 1991 International Symposium
on Radon and Radon Reduction Technology, Philadelphia, PA, April
2-5, 1991 (1).
During 1991, an opportunity arose to demonstrate this
recommended technology in a 60,000 sq ft hospital building in
Johnson City, TM. Results of this project are reported herein.
The mitigation system was extremely effective, lowering radon
levels to less than 0.5 pCi/L (detection limit for test used) in
the entire building. Levels greater than 50 pCi/L were measured
in the building when both the air handling and mitigation systems
were shut off.
The technical guidance document is in the final stages of peer
review and will be issued before the end of this year.
This paper also gives examples of oversized or poorly designed
radon mitigation systems which are currently being installed in
some new U. S. schools.
This paper has been reviewed in accordance with the U. S.
Environmental Protection Agency's peer and administrative review
policies and approved for presentation and publication.
-------
INTRODUCTION
The Air and Energy Engineering Research Laboratory (AEERL) of
the Environmental Protection Agency (EPA) initiated an extensive
research, development, and demonstration program on the mitigation
of radon in structures in the U.S. in 1985. initial work was done
on houses, both existing and new. In 1987, this program was
expanded to include schools and other large buildings. Extensive
diagnostics have been carried out in many existing schools, and
mitigation systems have been installed and evaluated in typical
ones. This work was carried out in 49 existing schools in 13
states.
In 1990, a decision was made to develop technical guidance for
the design and construction of new schools and other large
buildings which were radon resistant and easy to mitigate. As a
prelude to developing this technical guidance, the architectural
plans and specifications of all schools mitigated by EPA were
carefully studied to identify those features which affect radon
entry and ease of mitigation. The results of these studies are
being used to develop a technical guidance manual for the design
of new schools which are radon resistant and which can be easily
and inexpensively mitigated if a radon problem is found after the
building is completed. This document will be released by the end
of this year.
DESIGN FEATURES AFFECTING EASE OF MITIGATION
WITH ACTIVE SUBSLAB DEPRESSURIZATION (ASD)
Review of all school pressure field extension (PFE) studies
and examination of architectural plans have led to the
identification of the following features which affect PFE and hence
the effectiveness of ASD:
Subslab barriers (type, size, and location)
Aggregate
Bulk density (or void volume)
Particle size (both average size and particle size
distribution)
Particle shape (naturally occurring stone from
moraine deposits with rounded corners vs crushed
bedrock)
Subslab suction pit size
• Amount of suction applied
SUBSLAB BARRIERS
Schools have been found to fall into one of the following four
categories listed in order of ease of mitigation:
-------
Type 1. No interior walls extend through slab with the roof
load being carried by posts (steel or reinforced
concrete) extending through the slab to footings.
Type 2. Walls between classrooms extend through slab to
subslab footings. Hall walls do not.
Type 3. Hall walls extend through slab to subslab footings.
Walls between classrooms do not.
Type 4. All walls extend through slab to subslab footings.
A recent survey of the characteristics of about 100 schools
chosen at random on a national basis has shown that about 60% of
existing schools in the U.S. are Type 1, the easiest type to
mitigate. Type 1 construction is also increasing in popularity
particularly since air conditioning in schools negates the need for
windows in all rooms. In commercial and industrial construction
where dimensions of buildings are large in both directions (length
and width) , Type 1 (known architecturally as post and beam
construction) is almost always used. It is also believed to be the
most economical of the four types.
AGGREGATE
Aggregate characteristics are also very important to the
mitigation of large slabs. The following conclusions on the effect
of aggregate properties on PFE are believed to be accurate although
the effects of these variables have not been quantified:
1. PFE is proportional to average aggregate particle size
—the smaller the average particle size, the less the PFE
(assuming the same relative particle size distribution).
2. The narrower the aggregate particle size distribution
range, the greater the void volume and the PFE (assuming
the same average particle size).
3. The smoother the shape of the stone, the lower the void
volume; hence moraine stone (with its rounded corners)
has lower void volume and will give less PFE for the same
average particle size and particle size distribution than
crushed aggregate.
Based on these parameters the optimum subslab aggregate for
ASD is believed to be crushed bed rock meeting specifications for
#5 stone as defined in ASTM-33-86 (2).
-------
SUCTION PIT SIZE
Suction pit si2e also increases in importance as the size of
the building increases. Since cost of the suction pit in new
construction is relatively independent of size, the pit as shown
in Figure 1 is recommended. This type pit with a 6 in. riser was
used in the Johnson City hospital described in the next section.
It is inexpensive and easy to install. The interface between the
pit void and the aggregate face in this pit (a key variable in PFE)
is 7 sq ft.
AMOUNT OF SUCTION APPLIED
The larger the suction fan the greater is the PFE attained.
A T3B Kanalflakt fan was used in the Johnson City hospital
described in the next section. This fan moves 510 cfm of air at
no head. Initial and operating costs of this size fan (rated at
0.1 HP) are inconsequential in a large building.
APPLICATION OF OPTIMUM MITIGATION DESIGN FEATURES TO
JOHNSON CITY REHABILITATION HOSPITAL CONSTRUCTION
Late in 1990, an opportunity presented itself to demonstrate
ASD in a large building under optimum conditions in a hospital
being built in Johnson City, TN. The hospital building is one
story with a floor area of about 60,000 sq ft and is slab-on-grade
construction with no foundation walls penetrating the slab (Type
1). Mechanical piping, electrical conduit, and structural columns
penetrate the slab, and the columns sit on footings below the slab.
These columns support steel beams overhead which in turn carry the
bar joists for the roof. This type of construction is referred to
architecturally as post and beam construction. It is used in most
commercial and industrial buildings currently being built in the
U.S. All internal walls are gypsum board on metal studs, and the
exterior walls are metal stud supporting gypsum board on the inside
surface and an exterior insulation finish system on the outside.
The 4 in. slab was poured over a 6 mil vapor barrier underlain with
a 4 in. layer of crushed aggregate which was continuous under the
entire slab. The slab, exterior foundation walls, and footings
were poured monolithically. The slab was divided into about 15 ft
squares by a combination of pour joints (1,000 lineal ft) and
control saw joints (5,000 lineal ft). No expansion joints were
used.
EPA was requested to review the plans and specifications and
to recommend a radon mitigation system since the region was known
to have high radon potential. After this review, the following
recommendations were made to the architect designing the building:
-------
1. Minimum of 4 in. of crushed aggregate meeting the specifi-
cations for #5 stone as defined in ASTM-33-86 "Standard
Specifications for Concrete Aggregate" (2) carefully placed
so as not to include any soil.
2. Sealing of all pour and control saw joints and any slab
penetrations with a polyurethane caulking.
3. Installation of one centrally located subslab suction pit (SP)
of the design shown in Figure l in the approximate center of
the slab with a 6 in. stack leading to the roof capped with
a Kanalflakt T3B turbo fan capable of moving 500 cfm at no
head.
4. Continuous operation of the heating, ventilating, and air
conditioning (HVAC) fans in order to pressurize the building
in all areas except those where negative pressure is necessary
to control odors, noxious chemicals, or infectious diseases
(toilets, kitchen, pharmacy, soiled linens area, isolation
wards, etc).
All of the recommendations were accepted and incorporated into
the building design. Upon completion of the building shell and
sealing of the slab, diagnostic measurements were made to determine
the potential for a radon problem and to test the effectiveness of
the ASD system. Test holes were drilled through the slab at
varying distances from the suction pit, including a series around
the entire perimeter about 5 ft from the slab edge. Radon levels
below the slab were measured by "sniffing" using a Pylon AB-5
continuous monitor. Levels from 200 to 1,700 pci/L were found
under the slab. This is a significant level of radon which could
result in indoor measurements of over 50 pCi/L under some
conditions of building operation.
The depressurization fan was then turned on and subslab
pressure was measured made using a Neotronics roicromanometer. The
fan removed about 200 cfm of soil gas at a vacuum of about 1.5 in.
W. C. Negative pressure was 0.47 in. W. C. in the suction pit,
0.22 in. 50 ft from the SP, and 0.18 in. at the farthest point on
the perimeter (a distance of 185 ft) . This is considered extremely
good PFE. The PFE data are shown in Figure 2 and are plotted in
Figure 3, giving essentially a straight line on semi-log paper.
Extrapolation of these data in Figure 3 indicates that the
mitigation system would mitigate a much larger slab. At 500 ft
from the suction pit, subslab pressure would have been about 0.15
in. W. C., almost two orders of magnitude greater than needed for
mitigation. This indicates that the system as installed could have
mitigated a slab of over 1,000,000 sq ft had it been that large and
built according to the specifications used for the Johnson City
hospital.
-------
Upon completion of the building, radon levels were measured
in half of the building with the HVAC and ASD systems off using
open faced charcoal canisters. Radon levels ranged from less than
0.5 pCi/L (lowest detectable level with the open faced canisters
used) to 52.7 pCi/L. Highest levels were in the bathrooms
associated with the patient rooms. The patient room with the
highest bathroom radon level had a reading of 10.5 pCi/L, higher
than that found in any non-bathroom area in the building.
The entire building was then measured with the HVAC system on
and the ASD system off. Again some of the bathrooms (under
negative pressure) had elevated radon levels as did some of the
patient rooms. The bathroom with the highest radon reading with
the HVAC off was again the highest in the building with the HVAC
operating, testing 16.1 pCi/L.
The final series of tests was made with both the HVAC and ASD
systems operating. The 20 bathrooms with the highest radon levels
in the second series of tests were remeasured. No measurable radon
levels were found in any of the rooms tested. This is not
surprising in view of the large negative pressure under the slab
with the installed ASD system in operation.
INSTALLATION OF ASD AT
DRANESVILLE ELEMENTARY SCHOOL
AEERL was involved in the development of the mitigation
strategy for Dranesville Elementary School, Fairfax County, VA,
when it was being designed in 1987. The school was designed to
control radon using the HVAC system by running the circulation
fans of the HVAC system continuously to keep the building under
positive pressure at all times. As a backup precaution, the
subslab was prepared for the possible use of ASD by installing a
4 in. layer of #57 stone and carefully sealing all of the expansion
joints in the slab. When the school was tested after it was
completed in the summer of 1988, radon in all rooms was below 4
pCi/L with the air handling fans operating. The building was
tested in the middle of the winter 3 years later: all rooms in one
wing tested between 4 and 10 pCi/L and EPA was called back to
investigate the reason for the failure of the radon mitigation
system. After considerable investigation, it was found the HVAC
system had been converted to standard night setback shortly after
Energy Management System (EMS) personnel took over control of the
HVAC system in the building.
Maintenance department management decided to install ASD
rather than try to convince the EMS people to operate the HVAC
system as designed (continuous fan operation). At the time the
building was designed, the importance of subslab barriers had not
been recognized and no attempt was made to influence the foundation
-------
design. Although the design does not fit the definitions for any
of the four types of subslab construction, it is most similar to
Type 3 (see Figure 4) but only one room (the kindergarten room) was
found to have subslab barriers on all four sides. Since one
suction point was expected to reach all of the wing (except
possibly the kindergarten room), the suction point location was
chosen centrally as shown in Figure 5. PFE was measured using the
vacuum cleaner test method and reached all areas of the wing under
investigation including the kindergarten room. The ASD system was
then installed by maintenance personnel. Retesting with the system
operating showed that all of the subslab area of the wing was under
negative pressure as shown in Figure 5. with the ASD system
operating, all rooms tested at less than 1 pCi/L.
The system as installed did not meet the specifications used
in Johnson City with regard to pit size, stack pipe size, and
suction fan size. Dirt removed from the suction pit was only 5
gallons, meaning that the pit was probably only about 1 ft in
diameter. The surface of the interface between the pit void and
the crushed aggregate calculates to be about 1 sq ft compared to
about 7 sq ft at Johnson City. The stack pipe was only 4 in. in
diameter (compared to 6 in. at Johnson City) and about 75 ft long
resulting in considerable pressure loss. The fan used was a
Kanalflakt T2 capable of pulling 260 cfm at no head rather than
500 cfm as specified for Johnson City. To test the effect of fan
size, a Kanalflakt T3B fan, capable of pulling 500 cfm at no head,
was added to the line as it exited the building. With this
arrangement, subslab negative pressure was about doubled at all
test points, indicating that the PFE was about doubled.
INSTALLATION OF ASD AT
DUNN ELEMENTARY SCHOOL
Installation of ASD at an existing school of Type 1
construction in 1991, Dunn Elementary School in Louisville,KY, gave
EPA an opportunity to test the effect of the suction pit size on
PFE. A local mitigator had made diagnostic tests at this school
and had recommended an ASD system consisting of 14 suction points
and three fan systems. EPA was asked to review the mitigation
strategy. Based on the contractor's diagnostic data and a review
of the foundation drawings (showing post and beam construction),
EPA believed that the 53,000 sq ft building could be mitigated with
one suction point. The suction point was evaluated using four pit
sizes: none (just a 6 in. hole through the slab), 1 by 1 ft, 2 by
2 ft, and 2 by 3 ft. Although it was necessary to dig the last two
pits 2 ft deep to get the desired diameter, it is believed that
only that portion of the pit contacting the aggregate layer (top
4 in.) contributed to PFE.
-------
PFE was measured at the points noted on Figure 6 for each pit
size using a fan capable of pulling 510 cfra at no head. Results
are plotted in Figure 7. Increasing the diameter of the suction
pit definitely increases the PFE. If negative pressure for all of
the points is averaged for each suction pit size, consistent
results are obtained. Increasing pit size increases subslab
negative pressure as shown in Table 1.
TABLE 1. EFFECT OF SUBSLAB SUCTION PIT SIZE ON
SUBSLAB NEGATIVE PRESSURE
Pit diam
ft
0
1
2
3
Aggr./void
surf area
sq ff
0.2
0.8
1.6
2.4
Stepwise
increase
aggr/void
ratio t %
400
200
150
Average
SS neg
press
in. WC
0.014
0.024
0.030
0.034
Increase
in SS
nea P.%
70
25
13
neg pressure
incr/interface
ratio incr"
%
17
12
9
* Assumes 4 in. aggregate layer.
b Increase in subslab negative pressure divided by the stepwise
increase in the aggregate to void ratio.
These data show that average subslab negative pressure
continues to increase as the suction pit size is increased but at
a decreasing rate even when corrected for the actual increase in
size of the void to aggregate interface ratio (last column).
However, since (in new construction) cost of the suction pit does
not increase with increasing pit size (within the range used), the
larger size is recommended. Consequently, the choice of a 4 ft
pit (Figure 1) at the Johnson City hospital appears to be a valid
decision.
OTHER INSTALLATIONS IN
NEW CONSTRUCTION
Several other installations have been brought to EPA's
attention where architects and/or mitigators have recommended much
more extensive mitigation systems than necessary in new
construction. This has resulted from the lack of field-
demonstrated technology (by EPA or others). Three examples are
described in this paper in order to illustrate the amount of
redundancy which is currently being used.
-------
SCHOOL 1
This school has an area of about 60,000 sq ft and it is Type
1 construction (post and beam) with the exception of the gymnasium
which is Type 4 (all four walls extend through the slab to
footings). The building is in a radon-prone area, and a decision
was made to rough in an ASD system. The system was designed with
one suction point for about every 5,000 sq ft (ten points were
installed in locations shown on Figure 8). An average of 45 lineal
ft of 4 in. perforated pipe was used beneath the slab at each
suction point to improve PFE (represented by dotted lines in Figure
8). This installation was brought to EPA's attention before -the
building was completed, and permission was obtained to make PFE
measurements.
In order to measure the maximum PFE range of one suction
point, 11 were capped and a fan was temporarily installed on stack
5. (See Figure 9.) With the fan operating, negative pressures
under the slab were measured through small holes in the slab using
a micromanometer. PFE data obtained are shown in Figure 9 as is
the distance from the suction point. Good PFE was found in all
cases except at point O. (The reason that this point was not
reached is not known since it is closer to stack 5 than some of the
points that were reached.) Note that excellent negative pressure
was found at points F and G even though this area (the gymnasium)
is isolated by subslab walls from the location of the active
suction point. Further investigation showed that the footings in
the gymnasium area were poured under very wet conditions and that
coarse crushed aggregate was placed in the trenches before the
footings were poured. This apparently allowed the PFE to reach the
area under the gymnasium slab. Adequate negative pressure was also
found at point H although this point had two subslab walls between
it and the activated suction point.
This evaluation showed that one suction point (rather than
12) would have been sufficient in this school. The additional cost
of the extra points was significant as was the cost of the subslab
perforated pipe which is not needed if a large suction pit is used.
SCHOOL 2
A local mitigator designed the subslab system for this school,
essentially treating it as multiple houses. Twelve suction points,
with 10 in. diameter suction pits in the subslab aggregate, were
installed with 3 in. risers to the attic where a small house type
suction fan was to be added to each stack if a radon problem were
found. After the slab was poured, a professional engineer who
designed the plumbing, mechanical, and electrical systems for the
architect contacted one of the authors to ask if the 3 in. stacks
could be manifolded and larger fans used. The plans were revised
to manifold six stacks to one suction fan and three to each of two
-------
other fans. The use of three fans rather than one was chosen
because of the distance between the stacks. The individual stacks
will have valves for balancing purposes. This will also allow us
to determine the effectiveness of a single suction point in the six
point system with the other stacks shut off. EPA plans to test
this building as soon as the fans are installed. Results should
be available by the time of the symposium.
Examination of the foundation plans of this school disclosed
that it is Type 1 (post and beam). It is believed that this
building could have been mitigated with one properly designed and
located suction point.
LARGE BUILDING 1
In October 1990, one of the authors was contacted by an
architect who had designed a war gaming facility to be built by the
U. S Army Corps of Engineers for the Department of Defense in a
radon-prone area of Pennsylvania. The building was four stories
with a footprint of about 50,000 sq ft. The mitigation system
contained about 2,500 lineal ft of 4 in. diameter perforated pipe
consisting of a loop around the building near its outside walls
with connecting cross lines every 20 ft. Four stacks were to be
installed on the four corners of the pipe system leading to a
penthouse. If the building were found to have a radon problem
after completion, the system was to be activated by the
installation of four suction fans on these four stacks. This plan
was an adaptation of a plan originally developed by the Atomic
Energy Control Board of Canada in the 1970s.
Examination of the foundation plans showed that the building
design was Type 1 (post and beam), and it was recommended that the
proposed mitigation system be replaced with a centrally located
single suction pit and stack as in the Johnson City Hospital.
Unfortunately, the building was on a fast track schedule and the
Corps was unwilling to slow down the construction schedule as long
as the planned system would solve the problem. They were assured
that it would, but that it was more expensive than it needed to be.
A decision was made to proceed with the mitigation system as
designed to eliminate any delay.
Mitigation of these three buildings demonstrates the amount
of over-design being used on mitigation systems in schools and
other large buildings. The reason for this is the uncertainty of
what is needed and the philosophy of erring on the side of
redundancy if an error is made. This is really unnecessary since
additional suction points can readily be added on a retrofit basis
if tests show that a given system is not sufficient to eliminate
a radon problem in an entire building.
-------
It is also recognized that additional field verification of
the system as installed in Johnson City is needed, and EPA is
making an effort to locate new school buildings and other large
buildings being designed where this can be done.
It is hoped that EPA's new construction guide for schools and
other large buildings, to be issued before the end of 1992, will
eliminate the practice of over-design. A system similar to the
Johnson City system is being recommended in this guidance. The
document will also recommend that the HVAC system be designed and
operated to meet ASHRAE 62-1989 ventilation standards (3). Plans
are underway to see that this guidance manual gets wide
distribution to architectural firms designing school buildings.
Meanwhile a two page article has been placed in Architecture/
Research, a publication of the American Institute of Architects/
Association of Collegiate Schools of Architecture (AIA/ACSA) (4).
INSTALLATION COSTS
Incremental costs of the mitigation system at Johnson City
were easily ascertained since the contract for the building had
been let before the mitigation system was added to the design.
The additional cost of the radon mitigation system was covered by
four change orders for which the construction contractor charged
$5,367. Thus, the system cost $0.096 per sq ft of floor space.
Specifications had already called for 4 in. of aggregate under the
slab, and there was no charge for the change in aggregate
specification. The other three change orders covered installation
of the suction pit and stack to the roof, sealing of all pour and
control saw joints with a polyurethane caulking, and installation
of the fan and warning system.
A recent study of costs of mitigation in eight new schools
recently built gave costs from $0.35 to $1.10 per sq ft, as much
as $46,000 for one school building (5). Hence, the installed
mitigation system at the Johnson City Hospital cost only a fraction
of the cost of systems currently being installed in new schools in
the U.S.
CONCLUSIONS
A low cost, single point ASD system, installed during
construction, has lowered radon levels in a new single story
hospital building to near ambient levels of less than 0.5 pCi/L,
the detection limit of the radon test used. Levels greater than
50 pCi/L were measured in the building with both the HVAC and ASD
systems off and as high as 16 pCi/L with the HVAC system operating
and the ASD system off. ingredients of this radon mitigation
system are:
-------
1. Slab-on-grade post and beam construction with no barriers
to soil gas flow below the slab.
2. Continuous layer of coarse, narrow particle size range crushed
aggregate a minimum of 4 in. thick under the slab. (ASTM #5
is the preferred stone.)
3. Careful sealing of all slab cracks and penetrations and the
use of a 6 mil plastic film between the slab and the
aggregate.
4. A specially designed subs lab suction pit having a void to
aggregate interface area of at least 5 sq ft and a 6 in.
diameter stack to the roof.
6. An exhaust fan (on the stack) capable of exhausting a
minimum of 500 cfm at no head.
Incremental cost for the mitigation system was less than $0.10
per sq ft compared to a cost range of $0.35 to $1.05 per sq ft for
eight schools recently built in the U.S. with more complicated
radon mitigation systems.
Mitigation systems being designed for new schools and other
large buildings by architects and/or mitigators are more extensive
and costly than they need to be. This has resulted from the lack
of field-demonstrated technology (by EPA or others). Application
of the technology to be recommended in EPA's new construction
technical guidance manual (under development and named "Radon
Prevention in the Design of Schools and Other Large Buildings") has
the potential to reduce both cost and complexity of these systems
yet achieve very low radon levels.
REFERENCES
1. Craig, A. B., K. W. Leovic, and D. B. Harris, Design of Radon
Resistant and Easy-to-Mitigate New School Buildings. In
Proceedings: The 1991 International Symposium on Radon and
Radon Reduction Technology, Volume l, EPA-600/9-9l-037b (NTIS
PB92-115369), November 1991.
2. ASTH-33-86 "Standard Specifications for Concrete Aggregate,"
American Society for Testing Materials. May 1986.
3. ASHRAE-62-1989. Ventilation For Acceptable Indoor Air
Quality. American Society of Heating, Refrigerating and Air
Conditioning Engineers, Inc. 1989.
-------
4. Architecture/ Research, Volume 1, Number 1, October 1991. A
publication of the AIA/ACSA Council on Architectural
Research.
5. Craig, A. B,, K. W. Leovic, and D. W. saum, Cost and
Effectiveness of Radon Resistant Features in New School
Buildings, Healthy Buildings—IAQ'91, Washington, D. C.,
September 4-8, 1991.
CONVERSION FACTORS
Readers more familiar with the metric system may use the following
factors to convert to that system.
Non-metric Multiplied bv Yields Metric
cfm 0.00047 mx/s
ft 0.30 m
ft2 0.093 m2
gal. 0.0038 mx
HP 7.46 W
in. 0.025 m
in. we 249 Pa
mil 0.000025 m
pCi/L 37 Bq/ra3
-------
SECTION A
• 6" PLASTIC PIPE
4'X4'X3/4- TREATED
PLYWOOD SHEET
TO STACK
TO ROOF
8"X8*X8"
CONCRETE BLOCK
SECTION A
Figure 1. Subslab suction pit
-------
I I
0.180"
-0.180"
185'
-0.193"
J
n
A
102'
H
121'
U-0.186'
•0.188"
173'
-0.204"
-0.225
"
-0.256"
-0.468
"
106'
-0.2061
108'
113'
i
-0.215
Figure 2, Pressure field extension,
Johnson City Rehabilitation Hospital
-------
SUBSLAB PRESSURE, Neg . inch W. C.
Figure 3. Plotted pressure field extension,
Oohnson City Rehabilitation Hospital
-------
Ki nder-
garten
Room
Figure 4. Subslab walls ,
Dranesville Elementary
ConfRm 4 ASD Slab Penetration
\
' »«\ u<
• D -0.021 "\
tu \
\
r ".Q 148" i
V« I *TW |
1L5 •" — ""
ttt
111"
\ / -0.006"
Stack -0.99"
117 US
-0.007" E •
f
• -0.66 " IM I
>•
110 1 m
-0.011" C. 3 H*Xo.005»
Figure 5. Pressure field extension,
Dranesville Elementary
-------
X PFE measurement locations
Suction pit
Exhaust fan
Figure 6. Pressure field extension measurements
Dunn Elementary School
-------
I
II
M
I)
u
II
L
•I
II
CJ!
H
L
c>
>
<
0.69
o.oe-
'.••"•'.•• •
Pit Size (depth x diameter,
2x3
2x2
1 X 1
0x0 (no pit)
<100 100- 149 150-200
Distance From Suction Pit (ft)
Figure 7. Effect of subslab suction pit size,
Dunn Elementary School
>200
-------
1 1
\ X J
\s
/ \
Stack 13
X
\
*» .*
» * *
1 1 V
! \ /
1 *•- 1*
i
\
I— 1 V'*'* v*x /
u "Stack 18 Stack 19
1 /\ j» (Post & beam)
• * ^** 3
Stack 11* p--j
. .
E
1
'*" »«
10' *
i" ^tack 15 / 5
*• « —
IO «, ^ n
la V-' »
•
" .' V C.
.-' ••-. 2.
5
\
*% S UB^BBS^J "V •* *
\/
Stack 12
^>( St.ck\l7
St.'ck 16 P"~J| \
footing
s ?
Z GYM f
i *
(no post *°
& beam)
1
• 4
st,
'
\
\
*
ick It-^
Figure 8. ASD stack and perforated pipe location, School 1
-------
* I -0.006"
210'
. K -0.032"
120'
of 1/4" drilled 'subtUb"tirho>ts -0.029" !
70'
• M fno • *•«.) """*"
1
190'
N -0.003"
I fi52iU?5A'^S*^rT;
S,t»ek IS .•*
\ / - o
• J V ./ -0.171"
-0.164" X 26'
40' •' *« • c
\ -0.120"
42'
-0.144"
'190'.
"i rr . 8znrt>111
UM^J I *ft -0.094
LI,!?-
•r
* 0
0.000
175'
-0,018"-0.OlinJ -0,003"
.•*-<096
..100
omi-o.oc
140', 150
Figure 9. Pressure field extension,
from stack 5, School 1
GOVERNMENT PRINTING OFFICE: 1992—648-003/6O.020
------- |