United States       Office of Research and Development

              Environmental Protection  Research Triangle Park NC 27711     September 1992
              Agency          Office of Air and Radiation

                          Washington DC 20001
vvEPA       The  1992  International

              Symposium on  Radon

              and  Radon Reduction

              Technology:



              Volume  5.    Preprints


              Poster Papers

                Session II   Federal Programs and Policies
                            Relating to Radon
                Session III   State and Local Programs and
                            Policies Relating to Radon
                Session V   Radon Measurement Methods
                Session VI   Transport and Entry Dynamics
                            of Radon
                Session VII   Radon Reduction Methods
                Session VIII  Radon Occurrence in the
                            Natural Environment
                Session IX   Radon Surveys
                Session X   Radon in Schools and Large
                            Buildings
                Session XI   Radon Prevention in New
                            Construction
                Session XII   Radon in Water
             September 22-25,1992
             Sheraton Park Place Hotel
             Minneapolis, Minnesota

-------
     The 1992 International
  Symposium on Radon and
 Radon Reduction Technology

        "Assessing the Risk"
         September 22-25,1992
        Sheraton Park Place Hotel
         Minneapolis, Minnesota
             Sponsored by:

    U.S. Environmental Protection Agency
Air and Energy Engineering Research Laboratory

                 and

    U. S. Environmental Protection Agency
       Office of Radiation Programs

                 and

      Conference of Radiation Control
      Program Directors (CRCPD), Inc.
                               Printed on Recycled Paper

-------
           The 1992  International Symposium on Radon
                  and  Radon Reduction Technology
                            Table of Contents


                              Oral  Papers


Session  I:  Radon-Related Health Studies

Preliminary Radon Dosimetry from the Missouri Case-Control of
   Lung Cancer Among Non-Smoking Women
      Michael Alavanja, R. Brownson, and J. Mehaffey,
      National Cancer Institute	1-1

Rationale for a Targeted Case-Control Study of Radon and
   Lung Cancer Among Nonsmokers
      Mark Upfal and R. Demers, Wayne State University and Michigan
      Cancer Foundation; L. Smith, Michigan Cancer Foundation	     I-2

EPA's New Risk Numbers
      Marion Cerasso, U. S. EPA, Office of Radiation Programs	I-3

Interaction of Radon Progeny and the Environment and Implications as
   to the  Resulting Radiological Health Hazard
      Lidia Morawska, Queensland University of Technology, Australia	I-4

Does Radon Cause Cancers Other than Lung Cancer?
   Sarah  Darby, Radcliffe Infirmary. Oxford, England	 I-5

Measurements of Lead-210 Made In Vivo to Determine Cumulative
   Exposure of People to Radon and Radon Daughters
      Norman Cohen, G. Laurer, and J. Estrada, New York University	I-6

The German Indoor Radon Study - An Intermediate Report After
   Two Years of Field Work
      Lothar Kreienbrock, M. Kreuzer, M. Gerken, G. Wolke,
      H.-J. Goetze, G. Dingerkus, University of Wuppertal;
      H.-E. Wichmann, University of Wuppertal and  Center for
      Environment and Health; J. Heinrich, Center for Environment
      and Health; G. Keller, Saar University, Germany	1-7
                                    in

-------
Session II:  Federal Programs and Policies Relating to Radon

EPA's Radon Program
      Stephen D. Page, U. S. EPA, Office of Radiation Programs	  11-1

Revising Federal Radon Guidance
      Michael Walker, U. S. EPA, Office of Radiation Programs	  II-2

Profile of Region 5's Tribal Radon  Program
      Deborah M. Arenberg, U. S.  EPA Region 5	   II-3

The Development of the Homebuyer's Guide to Radon
      Paul Locke, Environmental Law Institute, and S. Hoyt,
      U. S. EPA, Office of Radiation Programs	H-4

Mitigation Standards for EPA's Radon Contractor Proficiency Program
      John Mackinney, D. Price, and G. L Salmon,
      U. S. EPA, Office of Radiation Programs	II-5

Consumer Protection and  Radon Quality Assurance:  A Picture
   of the Future
      John Hoombeek, U. S. EPA, Office of Radiation Programs	II-6

EPA's Proposed Regulations on Radon in Drinking Water
      Janet Auerbach, U.  S. EPA,  Office of Drinking Water	II-7
Session III:  State and Local  Programs and  Policies
              Relating  to  Radon

Radon in Schools: The Connecticut Experience
      Alan J. Siniscalchi, Z.  Dembek, B. Weiss, R. Pokrinchak, Jr.,
      L Gokey, and P. Schur, Connecticut Department of Health
      Services; M. Gaudio, University of Connecticut; J. Kertanis,
      American Lung Association of Connecticut	111-1

Trends in the Radon Service Industry in New York State
      Mark R. Watson and C. Kneeland, New York
      State Energy Office	 III-2

Targeting High-Risk Areas
      Katherine McMillan, U. S. EPA, Office of Radiation Programs	III-3

How Counties Can Impact the Radon Problem
      Jerald McNeil, National Association of Counties, and D.
      Willhort, Orange County, NC, Board of Commissioners	 IH-4
                                      IV

-------
Innovative Local Radon Programs
      Jill Steckel, National Civic League	 111-5
Session IV:   Creating Public Action

Indoor Radon:  A Case Study in Risk Communication
      Stephen D. Page, U. S. EPA, Office of Radiation Programs	IV-1

Activating Health Professionals at the Local Level
      Deborah McCleland, American Public Health Association	 IV-2

Translating Awareness Into Consumer Action
      Mary Ellen Fise, Consumer Federation of America	IV-3

Radon Testing and Mitigation as Applied in Corporate Relocations
      Richard Mansfield, Employee Relocation Council	IV-4

Ad Council Radon Campaign Evaluation
      Mark Dickson and D. Wagner, U. S.  EPA, Office of
      Radiation Programs	IV-5
Session V:  Radon Measurement Methods

The U. S. Environmental Protection Agency Indoor Radon
   Measurement Device Protocols - Technical Revisions
      Melinda Ronca-Battista, Scientific and Commercial Systems Corp.;
      A. Schmidt and T. Peake, U. S.  EPA, Office of Radiation Programs	V-1

A Performance Evaluation of Unfiltered Alpha Track Detectors
      William Yeager, N. Rodman, and S. White, Research Triangle
      Institute; M. Boyd, U. S. EPA, Office of Radiation Programs;
      S. Poppell, Jr., U. S. EPA-NAREL	  V-2

An Evaluation of the Performance of the EPA Diffusion Barrier Charcoal
   Adsorber for Radon-222 Measurements in Indoor Air
      David Gray, U.  S. EPA-NAREL; J. Burkhart, University of
      Colorado; A. Jacobson, University of Michigan	V-3

A Lung Dose Monitor for Radon Progeny
      Harvel A. Wright, G.  Hurst, and S. Hunter, Consultec
      Scientific, Inc.; P. Hopke, Clarkson University	V-4

-------
The Stability and Response to Radon of New and Recharged Electrets
      William G. Buckman and H. Steen III, Western Kentucky University;
      S. Popped, Jr., U. S. EPA-NAREL; A. Clark, City of Montgomery, AL	V-5

Design and Performance of a Low-Cost Dynamic Radon Test Chamber
   for Routine Testing of Radon Detectors
      P. Kotrappa and T. Brubaker, Rad Elec, Inc	V-6
Session VI:  Transport and Entry Dynamics of Radon

Characterization of 222-Radon Entry into a Basement Structure
   Surrounded by Low Permeability Soil
      Thomas Borak, D. Ward, and M. Gadd, Colorado State University	VI-1

Analysis of Radon Diffusion Coefficients of Concrete Samples
      K. J. Renken, T. Rosenberg, and J. Bemardin, University of
      Wisconsin-Milwaukee	VI-2

Data and Models for Radon Transport Through Concrete
      Vem C. Rogers and K. Nielsen, Rogers & Associates	VI-3

Simplified Modeling for Infiltration and Radon Entry
      Max Sherman and M. Modera, Lawrence Berkeley Laboratory	VI-4

The Effect of Interior Door Position and Methods of Handling Return Air
   on Differential  Pressures in a Florida House
      Arthur C. Kozik, P. Oppenheim, and D. Schneider, University
      of Florida	 VI-5

Building Dynamics and HVAC System Effects on Radon Transport
   in Florida Houses
      David Hintenlang and K. AI-Ahmady, University of Florida	VI-6

Radon Entry Studies in Test Cells
      Charles Fowler, A. Williamson, and S. McDonough, Southern
      Research Institute	VI-7

Model-Based Pilot Scale Research Facility for Studying Production,
   Transport, and Entry of Radon into Structures
      Ronald B. Mosley and D. B. Harris, U. S. EPA-AEERL;
      K. Ratanaphruks, ACUREXCorp	VI-8
                                      VI

-------
Session VII:  Radon  Reduction Methods

Durability of Sub-Slab Depressurization Radon Mitigation Systems
   in Florida Houses
      C. E. Roessler, R. Varas, and D. Hintenlang, University of Florida	VII-1

A Novel Basement Pressurization-Energy Conservation System
   for Residential Radon Mitigation
      K. J. Renken and S. Konopacki, University of
      Wisconsin-Milwaukee	VII-2

The Energy Penalty of Sub-Slab Depressurization Radon
   Mitigation Systems
      Lester S. Shen and C. Damm, University of Minnesota; D. Bohac
      and T. Dunsworth, Center for Energy and the Urban Environment	VII-3

Design of Indoor Radon Reduction Techniques for Crawl-Space
   Houses: Assessment of the Existing Data Base
      D. Bruce Henschel, U. S. EPA-AEERL	 VII-4

Multi-Pollutant Mitigation by Manipulation of Crawlspace
   Pressure Differentials
      Bradley H. Turk, Mountain West Technical Associates; G. Powell,
      Gregory Powell & Associates; E. Fisher, J. Harrison, and B. Ligman,
      U. S. EPA, Office of Radiation Programs; T. Brennan,  Camroden
      Associates; R. Shaughnessy, University of Tulsa	VII-5

Two Experiments on Effects of Crawlspace Ventilating on Radon
   Levels in Energy Efficient Homes
      Theodor D. Sterling, Simon Fraser University; E. Mclntyre, Hughes
      Baldwin Architects; E. Sterling, Theodor Sterling & Associates	VII-6
Session VIII:  Radon  Occurrence in the Natural Environment

Indoor Radon and the Radon Potential of Soils
      Daniel J. Steck and M. Bergmann, St. John's University	VIII-1

Nature and Extent of a 226-Radium Anomaly in the Western
   Swiss Jura Mountains
      Heinz Surbeck, University Perolles, Switzerland	VIII-2

Radon Potential of the Glaciated Upper Midwest:  Geologic and
   Climatic Controls on Spatial Variation
      R. Randall Schumann, U. S. Geological Survey	VIII-3
                                      VII

-------
 EPA's National Radon Potential Map
       Sharon Wirth, U. S. EPA, Office of Radiation Programs	vill-4
 Session IX:  Radon Surveys

 Comparing the National and State/EPA Residential Radon Surveys
       Jeffrey L. Phillips and F. Marcinowski, U. S. EPA, Office of
       Radiation Programs	  IX-1

 Radon Testing  in North Dakota Day Care Facilities
       Arlen L Jacobson, North Dakota State Department of Health	  ix-2

 Ventilation, Climatology and Radon Activity in Four Minnesota Schools
       Tim Burkhardt, E. Tate, and L Oatman, Minnesota
       Department of Health	  IX-3

 Estimates From the U. S. Environmental Protection Agency's
    National School Radon Survey (NSRS)
       Lisa A. Ratcliff, U. S. EPA, Office of Radiation Programs;
       J. Bergsten, Research Triangle Institute	  IX-4
 Session  X:   Radon in  Schools and Large Buildings

 EPA's Revised School Radon Measurement Guidance
       Chris Bayham, U. S. EPA, Office of Radiation Programs	  X-1

 Radon in Commercial Buildings
       Harry Grafton and A. Oyelakin, Columbus, Ohio Health Department	X-2

 Iowa Multiresidential Building Radon Study
      James W. Cain, Iowa State University Energy Extension	X-3

Airflow in Large Buildings
      Andrew Persily, U. S. Department of Commerce	  X-4

Meeting Ventilation Guidelines While Controlling Radon in Schools
      Eugene Fisher and B. Ligman, U. S. EPA, Office of Radiation
      Programs; T. Brennan, Camroden Associates; W. Turner,
      H. L. Turner Group; R. Shaughnessy, University of Tulsa	X-5

Radon Reduction in a Belgian School: From Research to Application
      P. Cohilis, P. Wouters, and P. Voordecker, Building
      Research Institute, Belgium	  X-6
                                     VIII

-------
Multiple Mitigation Approaches Applied to a School with
   Low Permeability Soil
      D. Bruce Harris, U. S. EPA-AEERL; E. Moreau and R. Stilwell,
      Maine Department of Human Services	X-7

General Indoor Air Investigations in Schools with Elevated Radon Levels
      Terry Brennan, Camroden Associates; G. Fisher and B. Ligman,
      U. S. EPA, Office of Radiation Programs; R. Shaughnessy,
      University of Tulsa; W. Turner and F. McKnight,
      H. L Turner Group	 X-8

Comparison of ASD and HVAC System Control in School Buildings
      Bobby Pyle, Southern Research Institute; K. Leovic, T. Dyess,
      and D. B. Harris, U. S. EPA-AEERL	 X-9

Effectiveness of HVAC Systems for Radon Control in Schools
      Kelly W. Leovic, B. Harris, T. Dyess, and A. B. Craig,
      U. S. EPA-AEERL; Bobby Pyle, Southern Research Institute	 X-10

Radon Prevention in Construction of Schools and Other Large
   Buildings - Status of EPA's Program
      A. B. Craig, K. Leovic, and D. B. Harris, U. S. EPA-AEERL	X-11
Session XI:  Radon Prevention in  New Construction

The Effect of Radon-Resistant Construction Techniques
   in a Crawlspace House
      David L Wilson and C. Dudney, Oak Ridge National Laboratory;
      T. Dyess, U. S. EPA-AEERL	XI-1

Performance of Slabs as Barriers to Radon in 13 New Florida Homes
      James L. Tyson and C. Withers, Florida Solar Energy Center	 XI-2

HVAC Control of Radon in a Newly-Constructed Residence with
   Exhaust-Only Ventilation
      Michael Clarkin and T. Brennan, Camroden Associates;
      T. Dyess, U. S. EPA-AEERL	XI-3

A Simplified Analysis of Passive Stack Flow Rate
      Pah I. Chen, Portland State University	 XI-4

Factors that Influence Pressure Field Extension in New Residential
   Construction: Experimental Results
      Richard Prill, Washington State Energy Office; W. Fisk
      and A. Gadgil, Lawrence Berkeley Laboratory	 XI-5
                                      IX

-------
Evaluating Radon-Resisant Construction Practices in Florida
      John Spears, H. Rector, and D. Wentling, GEOMET Technologies	XI-6

Laboratory Investigations for the Search of a Radon-Reducing Material
      Lakhwant Singh, J. Singh, S. Singh, and H. Virk, Guru  Nanak
      Dev University, India	XI-7
Session XII:   Radon in Water

Risk Assessment Implications of Temporal Variation of Radon and
   Radium Well Water Concentrations
      Alan J. Siniscalchi, C. Dupuy, D. Brown, and B. Weiss,
      Connecticut Department of Health Services; Z. Dembek, M. Thomas,
      and N. McHone, Connecticut Department of Environmental
      Protection; M. v.d. Werff, U. S. EPA Region 1	XII-1

Seasonal Variability of Radon-222, Radium-226, and Radium-228 in
    Ground Water in a Water-Table Aquifer, Southeastern Pennsylvania
      Lisa A. Senior, U. S. Geological Survey	XII-2

Radon in Tap Water from Drilled Wells in Norway
      Bjorn Lind and T. Strand, National Institute of Radiation Hygiene	XII-3

A Rapid On-site Detector of Radon in Water
      Lee Grodzins, Massachusetts Institute of Technology, and
      S. Shefsky, NITON Corporation	XII-4

-------
                             Poster  Papers
Session II Posters:  Federal Programs and Policies  Relating to Radon

Radon Measurement Proficiency Program: New Exam and Listing
   for Individuals
      G. Lee Salmon and P. Jalbert, U. S. EPA, Office of
      Radiation Programs	 IIP-1

Social and Economic Considerations in the School Evaluation Program
      Jed Harrison, U. S. EPA, Office of Radiation Programs	IIP-2

The Health of the Radon Industry - Survey and Program Results
   from Radon Proficiency Program Analyses
      James Long, U. S. EPA, Office of Radiation Programs	IIP-3
Session III  Posters:  State  and Local Programs and Policies
                     Relating to Radon

The Radon Health Effects Committee Report and Its Consequences:
   Getting Results in Radon Policy Development
      Kate Coleman, E. Fox, and F. Frost, Washington State
      Department of Health	  IIIP-1

Washington State's Innovative Grant: School Radon Action Manual
      Linda B. Chapman, Washington State Department of Health	IIIP-2

Teaming Up on Local  Radon Issues
      Robert Leker, State of North Carolina	  IIIP-3
Session V Posters:   Radon  Measurement Methods

A Decision-Theoretic Model for Evaluating Radon Test Procedures
   Based on Multiple Short-term Measurements
      Harry Chmelynski, S. Cohen & Associates	VP-1

Operational Evaluation of the Radon Alert Continuous Radon Monitor
      Emilio B. Braganza, III and R. Levy, U. S. EPA-LVF	VP-2

A New Design for Alpha Track Detectors
      Raymond H. Johnson, Key Technology, Inc	VP-3
                                     XI

-------
Measurements of Indoor Thoron Levels and Disequilibrium Factors
      Yanxia Li and S. Schery, New Mexico Institute of  Mining
      and Technology; B. Turk, Mountain West Technical Associates	VP-4

Comparison of Continuous and Occupancy Time Radon Measurements
   in Schools Using Programmable E-Perms
      Marvin Haapala, C. DeWitt, R. Power, and R. Fjeld,
      Clemson University	 VP-5

Indoor Radon in New York State Schools
      Susan VanOrt, L. Keefe, W. Condon, K. Rimawi, C. Kunz,
      and K. Fisher, New York State Department of Health	 VP-6
Session VI Posters:  Transport and Entry Dynamics of Radon

Simplified Modeling of the Effect of Supply Ventilation on Indoor
   Radon Concentrations
      David Saum, Infiltec; M. Modera, Lawrence Berkeley Laboratory;
      K. Leovic, U. S. EPA-AEERL	VIP-1

Determination of Minimum Cover Thickness for Uranium Mill
   Tailings Disposal Cells
      Jeffrey Ambrose and D. Andrews, CWM Federal
      Environmental Services, Inc	  VIP-2

A Mathematical Model Describing Radon Entry Aided by an Easy
   Path of Migration Along Underground Tunnels
      Ronald B. Mosley, U. S. EPA-AEERL	VIP-3

Radon Diffusion Studies in Soil and Water
      Manwinder Singh, S. Singh, and H. Virk, Guru Nanak
       Dev University, India	VIP-4

Stack Effect and Radon Infiltration
      Craig DeWitt, Clemson University	  VIP-5

Relative Effectiveness of Sub-Slab Pressurization and
   Depressurization Systems for Indoor Radon Mitigation:
   Studies with an Experimentally Verified Model
      Ashok J. Gadgil, Y. Bonnefous, and W. Rsk,
      Lawrence Berkeley Laboratory	  VIP-6
                                       XII

-------
Session VII  Posters:  Radon Reduction Methods

Radon Mitigation Systems - A Liability in Cold Climate Homes?
      Kenneth D. Wiggers, American Radon Services, Ltd	 VIIP-1

Why We Like Diagnostics
      John W. Anderson, Jr. and J. Bartholomew, Jr.,
      Quality Conservation	VIIP-2

An Approach to Computer-Assisted Radon Mitigation
      Hormoz Zarefar, P. Chen, and P. Byrne, Portland State University;
      C. Eastwood, Bonneville Power Administration	VIIP-3

Radon Control - Field Demonstrations:  Diagnostic and Mitigation
   Techniques Used in Twenty-Six Radon Field Workshops
      Craig E. Kneeland and M. Watson, New York State Energy Office;
      W. Evans, Evanshire Company, Ltd.; T. Brennan,
      Camroden Associates	VIIP-4

Radon Mitigation at Superfund Remedial Action Sites: Field
   Experience and Results
      Jean-Claude Dehmel, S. Cohen & Associates; R. Simon,
      R. F. Simon Company, Inc.; E. Fisher, U. S. EPA, Office of
      Radiation Programs	VIIP-5

Dose and Risk Projection for Use of Sub-Slab Radon Reduction
   Systems Under Realistic Parameters
      Larry Jensen, U. S. EPA Region 5; F. Rogers and C.  Miller,
      Centers for Disease Control	VIIP-6
Session VIII Posters:   Radon Occurrence in the Natural  Environment

Influence of Meteorological Factors on the Radon Concentration
   in Norwegian Dwellings
      Terje Strand and N. Bohmer, Norwegian National Institute
       of Radiation Hygiene	VIIIP-1

Soil  Radon Potential Mapping and Validation for Central Florida
      Kirk K. Nielson and V. Rogers, Rogers and Associates;
      R. Brown and W. Harris, University of Florida; J. Otton,
      U. S. Geological Survey	VIIIP-2
                                      XIII

-------
 Correlation of Indoor Radon Screening Measurements with Surficial
    Geology Using Geographic Information Systems
       Charles Schwenker, J-Y Ku, C. Layman, and C. Kunz,
       New York State Department of Health .............................. VIIIP-3

 Analysis of Indoor Radon in New Mexico Using Geographic
    Information Systems (GIS)
       Richard A. Dulaney, Lockheed Engineering and Sciences Co ......... VIIIP-4

 A Radon "Pipe" (?) in the Brevard Fault Zone Near Atlanta, Georgia
       L. T. Gregg and J. Costello, Atlanta Testing & Engineering ............ VIIIP-5


 Session IX Posters:  Radon Surveys

 Summary of Regional Estimates of Indoor Screening
    Measurements of 222-Radon
       Barbara Alexander, N. Rodman, and S. White, Research Triangle
       Institute; J. Phillips, U. S. EPA, Office of Radiation Programs ........... IXP-1

 Texas Residential Radon Survey
       Charles Johnson, G. Ramirez, and T. Browning, Southwest Texas
       State University; G. Smith, P. Breaux, and V. Boykin, Texas
       Department of Health ............................................  IXP-2

 Radon Survey of Oregon Pubic Schools
       Ray D. Paris and G. Toombs, Oregon Health Division ................ IXP-3

 Quality Assurance in Radon Surveys
      William M. Yeager, R. Lucas, and J. Bergsten, Research Triangle
      Institute; F. Marcinowski and J. Phillips, U. S. EPA, Office of
      Radiation Programs ...........................................     IXP-4

 Radon in Houses Around the Plomin Coal Fired Power Plant
      N. Lokobauer, Z. Franic, A. Bauman, and D. Horvat,
      University of Zagreb, Croatia ......................................
A Radon Survey at Some Radioactive Sites in India
      Jaspal Singh, L Singh, S. Singh, and H. Virk, Guru Nanak
      Dev University, India ..........................................     IXP_6

Islandwide Survey of Radon and Gamma Radiation Levels
   in Taiwanese Homes
      Ching-Jiang Chen, C-W Tung, and Y-M Lin, Taiwan Atomic
      Energy Council .........................................           IXP-7
                                    XIV

-------
Session X Posters:  Radon in Schools and Large Buildings

Solar Fresh Air Ventilation for Radon Reduction
      Monty Holmes, Intermountain Radon Service, and Kelly
      Leovic, U. S. EPA-AEERL	  XP-1

Characteristics of School Buildings in the U. S.
      Kelly Leovic, U. S.  EPA-AEERL; H. Chmelynski, S. Cohen
      & Associates	 XP-2

Radon in Schools  in Wisconsin
      Conrad Weiffenbach and J. Lorenz, Wisconsin Bureau
      of Public Health	XP-3

Investigation of Foundation Construction Details to Facilitate Subslab
   Pressure Field  Extension in Large  Buildings
      Michael E. Clarkin, Camroden Associates; F. McKnight,
      H. L Turner Group; K. Leovic, U. S. EPA-AEERL	XP-4

Radon Measurements in the Workplace
      David Grumm, U. S. EPA, Office of Radiation Programs	XP-5

Radon Survey of Oregon Public Schools
      George L Toombs and R. Paris, Oregon Health Division	XP-6
Session XI Posters:  Radon Prevention  in New  Construction

Model Standards and Techniques for Control of Radon in New Buildings
      David M. Murane, U. S. EPA, Office of Radiation Programs	 XIP-1

Combined Ventilation and ASD System
      David Saum, Infiltec, and F. Sickels, New Jersey Department
      of Environmental Protection	XIP-2

Evaluation of Passive Stack Mitigation in 40 New Houses
      Michael Nuess, Washington State Energy Office	 XIP-3

Radon Remediation and Life Safety Codes
      Lyle Sheneman, Chem-Nuclear Geotech, Inc	XIP-4

A Passive Stack System Study
      Geoffrey Hughes and K. Coleman, Washington State
      Department of Health	 XIP-5
                                     xv

-------
Session XII Posters:  Radon in Water

Radon in Water Measurements Using a Collector-Bubbler
      Robert E. Dansereau and J. Hutchinson, New York State
      Department of Health	  XIIP-1

Measurements of Radon in Water via Sodium iodide Detectors
      Paul N. Houle, East Stroudsburg University, and D. Scholtz,
      Prosser Laboratories	XIIP-2

Continuous Measurement of the Radon Concentration in Water Using
   Electret Ion Chamber Method
      Phillip K. Hopke, Clarkson University, and
      P. Kotrappa, Rad Elec, «nc	XIIP-3

Performance Testing the WD200 Radon in Water Measurement System
      George Vandrish and L Davidson,  Instruscience Ltd	XIIP-4

Temporal Variations in Bedrock Well Water Radon and Radium, and
   Water Radon's Effect on Indoor Air Radon
      Nancy W. McHone and  M. Thomas, Connecticut Department of
      Environmental Protection; A. Siniscalchi, Connecticut Department
      of Heatth Services	  XIIP-5
                                    XVI

-------
      Session II Posters

Federal Programs and Policies
      Relating to Radon

-------
                                                                   IIP-1
              Radon Measurement Proficiency Program
               ••v Bxaa and  Listing for Individuals


                                by

                  G.  Lee Salmon and Phil Jalbert
               U.S. Environmental Protection Agency
                   Office of Radiation Programs
                             ANR-464
                        401  M Street, S.H.
                      Washington, D.C. 20460


                             ABSTRACT

The U.S. Environmental  Protection Agency has developed a
new proficiency listing status  for individuals in the Radon
Measurement Proficiency (RMP) Program to assist consumers in
identifying Competent contractors providing on-site residential
radon measurement  services.  Since its inception in 1986, the RMP
Program has evaluated radon measurement firms and organizations.
The Agency has now developed a  measurement exam to better assess
the proficiency of individuals  alone.

Contractors Who wish to obtain  this new proficiency designation
must pass a written national examination on radon measurement and
meet other requirements.  The exam was first offered in January,
1992 and evaluates six  major areas of practice including a
contractor's knowledge  of radon health effects and risks, types
of measurement devices,  special procedures to ensure valid
measurements as part of real estate transactions, interpretation
of test reports,  strategies  for mitigation work to reduce indoor
radon levels^ and  professional  standards of conduct and ethics.

Upon obtaining listing  status,  individuals receive an EPA
identification card to  assist  consumers in identifying them as
knowledgeable* proficient measurement contractors.

This paper discusses the exam,  relevant EPA policies addressed
therein, listing  requirements,  recommended training through EPA's
Regional Radon Training Centers and the national distribution of
RMP proficient measurement  contractors.  It will also discuss use
of the exam and fcMP  listings as part of requirements in  state
radon certification programs,  and explore  issues relating to
combining KPJi^s proficiency programs under one common
administratiVa structure.

-------
                                                                          IIP-2
Title:  Social and Economic Considerations in the School Evaluation Program

Author: Jed Harrison, U. S. EPA, Office of Radiation Programs

      This paper was not received in time to be included in the preprints, and
the abstract was not available. Please check your registration packet for a
complete copy of the paper.

-------
                                                                     IIP-3
   THE HEALTH OF THE RADON INDUSTRY - SURVEY AND PROGRAM RESULTS
              FROM RADON PROFICIENCY PROGRAM ANALYSES
             By: James Long
                 U.  S. EPA
                 Office of Radiation Programs (ANR-464)
                 401 M Street SW
                 Washington,  D.  C.  20460
                             ABSTRACT
     From data collected to support the Radon User Fee Rule, it is
possible to characterize the radon industry in terms such as:

             o   Device availability by type of device and location

             o   State radon certification programs and
                 participation in these programs

             o   Participation levels in state and federal
                 proficiency programs with keys to how long a firm
                 has been in the radon measurement and/or
                 mitigation business

             o   Etc.

     The paper will  detail, in terms of statistics, the types of
programs currently in operation and how voluntary and mandatory
programs have impacted the types of services available for
measurements and mitigations.  Types of firms and individuals can
be characterized in  terms of:  how long they have been in
business; what types of services they offer; how many measurements
or mitigations they  perform monthly and yearly; what prices they
charge for their services; and who will, likely, be left and what
types of services might be available in the 1990s.

-------
         Session III Posters

State and Local Programs and Policies
         Relating to Radon

-------
                                                                   IIIP-1
            THE  RADON HEALTH  EFFECTS  COMMITTEE  REPORT

AND ITS CONSEQUENCES; GETTING RESULTS IN RADON POLICY DEVELOPMENT

               by:  Kate Coleman, Ed Fox and Floyd Frost
                    Washington State Department of Health
                    Olympia,  Washington 98504-7825
                             ABSTRACT

     In the  overall  effort to develop a  public  health policy on
radon, the Washington State Department of  Health  conducted a study
as a  background report  for  the  Health Effects  Committee  of the
state's Radon Task Force.  Finding that the results of residential
radon health effects studies are  inconclusive,  the  report warns
that  recommendations  for  action  be stated  with  an  explicit
discussion of areas of uncertainty.  The Committee finds no current
direct evidence that radon causes or does  not cause lung cancer at
concentrations below 4 pCi/1, but recommends that the Department of
Health accept the U.S. Environmental Protection Agency action level
of  4  pCi/1  for  residential exposures.   The Committee  also
recommends the use of the BEIR IV risk estimates,  which  assume that
all levels of radon pose a risk and that  this risk increases with
increasing   radon  concentrations.     The  Committee  urges  the
Department  of  Health to  encourage  residential radon  testing,
particularly in areas  known  to have high radon  levels.   One
important outcome of this work is the state's adoption of interim
radon resistive construction standards for residential buildings.
These standards are effective for two years, during which time the
Department of Health will  conduct further research—a state-wide
survey of  residential radon exposures, and an evaluation of the
field performance of passive radon vent stacks.  New standards to
replace the  interim  standards will  incorporate  new  evidence and
will attempt to reduce the areas of uncertainty.

        The  work  described  in this  paper  was  partially
        funded   by   the  United   States  Environmental
        Protection Agency and is  in  their review process.
        Therefore,  the contents do not necessarily reflect
        the views of  the  agency and no official endorsement
        should be inferred.

-------
            THE RADON HEALTH EFFECTS COMMITTEE REPORT

AND ITS CONSEQUENCES: GETTING RESULTS IN RADON POLICY DEVELOPMENT

                           BACKGROUND


     Radon is produced by the decay of naturally occurring uranium.
Uranium is  present,  at  least  in low levels, in  all  soils and is
abundant in soils in  certain geologic regions. Radon can accumulate
indoors  at  levels   which  some  health   experts   believe  may
substantially  increase  the risk of lung cancer. Radon  has been
shown  to  cause  lung cancer  among underground  miners.  Recently,
there has been considerable concern that  the accumulation of radon
in homes is a  cause  of lung cancer.

     Lung cancer, primarily caused by cigarette smoking,  is now the
leading cause of cancer death  in Washington State.  Lung cancer
death rates in Washington State  have increased from 7.8 deaths per
100,000 population in 1940 to 52.5 deaths per 100,000 population in
1987. Smoking  is believed to be  the cause of 90%  of lung cancer in
men and 78% of lung  cancer in women.  Radon may also interact with
cigarette  smoking to increase  the  risk  of  lung cancer among
smokers.

     The Radon Health Effects  Committee,  a subcommittee of the
Radon Task Force, was formed by the Washington Department of Health
(DOH)  to review the  scientific  literature  on  radon-caused health
effects, evaluate whether radon may be causing human health effects
in this state, and advise the  department  on a public health policy
for radon.  The Committee met  three times in late 1990 and early
1991 to develop the  recommendations that follow.

                   FINDINGS AND  RECOMMENDATIONS


      (1)  The  Committee finds there is no current direct evidence
          either that radon causes  or does not cause lung cancer at
          concentrations  below  4  pCi/1.   Some members  of the
          Committee  believe  there  is  uncertainty  as   to  the
          existence  of health effects at a level of 10 pCi/1.

      (2)  The  Committee believes that for a conservative approach
          to public health protection,  it is prudent to  accept the
          risk estimates put forth in reports of major scientific
          bodies.  More  specifically,   the  Committee recommends
          utilizing  the National Academy of Science Committee on
          Biological Effects of Ionizing Radiations (BEIR  IV) risk
          estimates.

-------
     (3)   The Committee recommends that the Washington Department
          of Health accept the U.S.  Environmental Protection Agency
          (EPA)  action level of 4 pCi/1 for residential exposure.
          However, according to BEIR IV, all  levels  of radon are
          assumed to have a risk, which increases with increasing
          radon concentration.

     (4)   The Washington  Department of  Health should encourage
          occupants,  owners or managers of homes and buildings to
          test for radon, particularly  in areas known to have high
          radon  levels.  Even   in areas   where   average  radon
          concentrations are low,  testing  may be  advisable since
          some houses  in these  areas  have been  identified  with
          levels above 4 pCi/1.

     (5)   Where indicated,  repairs should be made to bring annual
          average indoor levels down to or below 4 pCi/1. Follow-up
          testing and mitigation  information is found in the EPA's
          "A Citizen's Guide to Radon".

     (6)   The Committee recommends  that cigarette smokers, whether
          or not exposed to indoor  radon, be informed that smoking
          cessation should be their highest priority for reducing
          lung cancer risk.

     (7)   Since  a number  of  radon  health  effects  studies  are
          ongoing and the results of these studies may affect the
          Findings and  Recommendations of this Committee,  it is
          recommended  that the  Washington Department  of Health
          reconvene the Radon Health Effects Committee annually.

     The following voluntary participants composed the Radon Health
Effects Committee and endorse the Findings and Recommendations:

     MEMBERS                  AFFILIATION

     John Beare               Spokane Co. Health District
     David Bodansky           University of Washington
     Fred Cross               Battelle, Pacific NW Labs
     Kenneth Jackson          University of Washington
     Jerry Leitch             US EPA
     Ahmad Nevissi            University of Washington
     Sam Reed                 General Public
     Maurice Robkin           University of Washington
     James Matsuyama          NE Tri-County Health District
     Larry Jecha              Washington    Association   of
                              Public Health Officials

                    DEPARTMENT OF HEALTH  STAFF

            Kate Coleman, Chair,  Ed Fox,  Floyd Frost,
              Pat McLachlan, Sam  Milham,  Bob Mooney

-------
     The  attached  report  reflects  the  literature  review  and
discussion  of  the   Committee.  All  of  the  comments  from  the
Committee have not been incorporated,  thus the  report  should be
viewed as a DOH staff report, which formed the background for the
Committee   Findings  and  Recommendations.   For  organizational
purposes, the report is  divided into sections on risk, exposure and
cost.

                      RADON HEALTH EFFECTS


     Alpha-emitting radionuclides can pose a risk to human health,
possibly inducing cancer and birth defects. These  radionuclides are
generated by both human and natural sources. Natural sources make
the  largest contribution to  human exposure,  especially  through
inhaled  radon  and radon decay  products.  This  exposure  occurs
primarily when radon concentrates indoors. The U.S. Environmental
Protection  Agency  has estimated  that human  exposure to naturally
occurring radon is  responsible for approximately 16,000 U.S. lung
cancer deaths  each year (1). Radon-222 is  an  inert  gas which is
derived from uranium-238.  Its abundance in soils varies, depending
on   the   concentration  of   uranium-238   in  the  soil  and  the
permeability of the soil.  Because the air  pressure  in buildings is
often lower than the atmospheric  pressure, radon  can be drawn into
buildings from the soil. To a much lesser  extent,  radon can also be
released  from  water used in the building and from burning natural
gas  which contains low levels of  radon. Certain building materials
can  also  contribute to the indoor radon concentration.  Most human
exposure  to radon gas and radon decay products occurs from indoor
air  exposure.

MINING AND  LUNG CANCER

     The relationship between mining  and elevated lung cancer risk
has  been known for over  one hundred years.   Uranium  miners are
heavily   exposed   to   soils  and   rocks   with  high  uranium
concentrations.  Because of  this, they receive high  exposures to
radon  and radon progeny.  The evidence for  lung  cancer resulting
from radon  exposure  comes  from  cohort  mortality  studies  of
underground miners who were  exposed to differing  concentrations of
radon-222  progeny (2,3).  The lung cancer hazard for underground
miners  was  first  described  in  1879   in  Europe  (4) .  Excess
lung-cancer risk has been observed for uranium miners in  the United
States,  Czechoslovakia, France  and  Canada  (3,5-7). Excess lung
cancer risk has also been demonstrated for other underground miners
such as  Newfoundland   fluorspar miners,  Swedish metal  miners,
Cornish tin miners, Norwegian niobium miners and Chinese  tin miners
(8-11).

     A casual  relationship between lung cancer and radon exposure
was  first suggested in the 1930's (5). It was not  until after World

-------
War II that radioactivity was seriously considered to be the cause
of  the  excess lung  cancer among  miners.  During the  1950's and
1960's, several new epidemiological studies were begun to determine
the hazard of  exposure  to radon progeny in mines (12,13). Unlike
earlier studies,  the recent studies more accurately measured radon
exposure.  Some also  collected data  on other lung cancer  risk
factors such as smoking. Although studies revealed that underground
miners had a very high smoking prevalence  (14),  the number of lung
cancers among  these miners exceeded  the number expected based on
smoking alone  (15). One study of  Navajo Indian miners, who smoke
very little, demonstrated that radon exposure could increase the
risk of lung cancer even among non-smokers (16). Recently, radon
exposure has also been demonstrated to be a lung carcinogen among
laboratory animals (17).

     Since many of the  earlier studies of underground  miners did
not adequately measure personal radon exposure levels,  deriving a
quantitative relationship between  exposure and risk was possible
only from the more recent  epidemiological studies.   These studies
allowed  relative  risks  for   lung  cancer  to  be  calculated  or
estimated  per  "Working Level  Month"   of  radiation exposure.  A
"Working Level" is a measure of radon progeny concentration in the
air where each liter of air has a given amount of potential alpha
energy derived from radon progeny. A "Working Level Month"  (WLM)
represents an  exposure  to a "Working Level" of radon progeny for
170 hours.

DERIVING A RELATIONSHIP BETWEEN EXPOSURE AND DISEASE

     Based on the findings in miner studies, concerns were raised
over the hazard of residential exposure to radon and its progeny.
In the 1970's the Public Health Service discovered that high levels
of  radon  were accumulating in certain houses. The hazards for
people living in houses  built  over uranium mill tailings or on land
reclaimed from phosphate mines in Florida were discovered. In 1984,
a nuclear-power engineer  set off personnel contamination monitors
at his job because of  his exposure  to high  radon concentrations in
his Pennsylvania home (18) .

     To assess the risk  from indoor radon exposures, a committee of
the National Academy of Sciences (NAS) was formed to evaluate the
evidence for health effects and estimate the risk to citizens. This
committee, known as the Fourth Committee on Biological Effects of
Ionizing Radiation (BEIR IV) (18), developed a model relating radon
exposure to health effects.

     In  constructing  the  model   or dose-response  relationship
between radon exposure and lung cancer risk, BEIR IV used only four
miner cohort studies  (18) which had adequate  exposure data. The
model assumes  that any  radon  exposure  increases lung cancer risk
and that the increasing  cumulative  exposure linearly increases the

-------
 risk.   Radon exposure is also assumed to multiply  the  underlying
 risk of lung cancer. Therefore the higher the underlying risk,  the
 higher the increase in absolute  risk.

      Because cigarette smokers have a high background risk for lung
 cancer, the BEIR IV model estimates that cigarette smokers exposed
 to radon will experience a greater absolute risk from  radon than
 will non-smokers  (18).  The data  supporting  this assumption  are
 limited. Data on tobacco use were available  for only two of  the
 miner cohort studies.  Data  on the  time  since cessation  of tobacco
 use and inhalation practices were unknown. There were also very few
 non-smokers  in  these  miner populations.    The  findings of  the
 studies suggest  that  radon exposure  and cigarette  smoking  may
 interact by producing  a higher risk than simply the  sum  of the  two
 risks.

 CONCERNS WITH THE RADON RISK EXTRAPOLATION

      A  number  of  concerns  have been  raised  about  the  derived
 estimates  of risk.  Some experts  question whether  a linear risk
 extrapolation is justified.  Health effects for miners  exposed to
 more than  100 "Working Level Month"  (WLM) of  radon during  their
 mining work are extrapolated to residents exposed to less than  100
 WLM spread throughout  a  lifetime.  It is possible that  either  the
 lower  doses or the lower rates of exposure produce fewer cancers
 per unit of exposure.  Other experts question whether the exposure
 estimates  for residences, measured during the winter in  the lowest
 living area accurately estimate  lifetime exposure.

     The epidemiological studies of underground miners  yield data
 that relate  cancer incidence to  radon  exposure.   However,  the
 confidence  in  any  one  postulated relationship between  radon
 exposure and cancer incidence is  limited. Epidemiological data  are
 not sufficient  to rule  out  alternative relationships  between
 disease and exposure.  The epidemiological  studies are limited  by
 the study  designs and  the available data. Some studies  restricted
 the period of follow-up and thus did not  identify disease occurring
 many years after exposure.  Some  studies had  uncertain or missing
 exposure  levels  for  many   individuals.  Some  were  plagued   by
 unadjusted confounding factors such  as  smoking.  Since smoking  is
 likely  to  be the most important cause  of lung  cancer,  even  for
 miners,  not knowing which members of the cohorts smoked and how
 much they smoked seriously limits the precision of the derived risk
 estimates.

     In extrapolating  risks for  miners  to  residences, several
 adjustments  must be made.  Radon exposure to  underground miners
 differs in three important ways from residential exposure.  First,
most  underground  miners  have   short  duration,  high   intensity
 exposure to  radon, occurring during the 8 hour workday for 5 days
per week.   This  pattern of  exposure may produce different health

-------
effects  than the lower level but longer duration radon exposures
which occur  in residences. Second, since mining involves strenuous
physical  work,  miners breath in more  air  and thus have a higher
exposure  to  airborne pollutants.  The radon progeny may adhere to
these  other  pollutants and be more  efficiently carried into the
lungs of miners. Third, little is known about the long-term effects
of radon  exposure  for children.

RESIDENTIAL  RADON  HEALTH  EFFECTS STUDIES

     Questions about the BEIR IV risk model would be less important
if the lung cancer risk could be  empirically determined for people
living in houses with moderately elevated radon levels. Attempts
have  been made  to directly  evaluate  the  lung  cancer  risk from
residential  radon  exposures.  However,  since the magnitude of the
expected  increased  risk  is  relatively small,  the size  of the
population in any study capable of detecting such an effect must be
large. To evaluate small elevated  risks,  epidemiologists prefer
case-control   studies  rather  than   cohort  studies.   A  radon
case-control study compares  the radon exposure of lung cancer cases
to a comparable  people without  lung cancer. A radon cohort study
calculates  the  incidence  of  lung  cancer  in a  well  defined
population with elevated radon exposure. The lung cancer incidence
is then  compared  to the general  population or to a  comparable
unexposed  population.  When  studying rare  diseases,  such  as lung
cancer, case-control  studies  have a much better  chance
of detecting an elevated  risk.

     A 1979  Swedish  study examined whether individuals with lung
cancer were  more  likely  to live in  stone or  wood houses  (19) .
Significantly more people with lung cancer lived in stone houses.
A relationship between estimated  radon levels and lung cancer risk
was  noted in  other  Swedish case-control  studies  (20,21).  One
cross-sectional study calculated  the  lung  cancer mortality rates
for areas with high  and  low background radon  levels  (20).  This
study found that areas with high background radon levels had higher
lung cancer mortality  rates  (21).

     A small follow-up study of lung cancers in Maryland revealed
no significant  differences  in lung cancer  risks by housing type
(22). A  small cohort study  in Ontario,  Canada found  a  twofold
increase in lung cancer risk among residents of houses built with
"radioactive  materials"  (12,23).  This  increased  risk  was  not
statistically significant.   In New Jersey,  a case-control study
found that the lung cancer risk was twice as high for people living
in houses with radon  levels  above  4  pCi/1  compared to  those with
less than 4 pCi/1  (24).

     A 1988  Swedish  case-control study  compared the residential
building  characteristics  (and  thus   indirectly radon  exposure
levels) of 177  people with  lung cancer  and  677 controls  (people

-------
without lung cancer).  A smoking history which included active and
passive smoking was obtained for each case and control. The lung
cancer risk was found to be double for rural residents who lived in
the houses most likely to have high radon levels.  No increased risk
was  observed for  urban residents  who lived in  houses  with high
predicted radon levels (25).

     A recent case-control study of  308  Chinese women  with lung
cancer was conducted by the  U.S.  National Cancer Institute (26) .
This study found no association of radon levels above 4 pCi/1 and
lung cancer risk.  The  lack of association was consistent for both
smokers and non-smokers. No association with  radon  exposure was
apparent for any lung  cancer cell type. This study is the largest
residential case-control  study of lung cancer and radon published
to  date. The median period of residence was 24  years.  The radon
exposures were  relatively low with only  20% of the houses having
levels above 4  pCi/1.  It is also possible that other  exposures
found  in China, such as high levels of other indoor air pollutants,
may have obscured a radon effect.

CONCLUSIONS

     Without conclusive  and consistent   findings  from  studies
relating lung cancer to residential radon exposure, recommendations
for action must be stated with clear,  explicit discussion of areas
of  uncertainty.  Elevated lung  cancer  risk has been  observed
primarily for miners with more than 100 WLMs of exposure delivered
over  a  relatively  short  time  period. Most  lifetime residential
exposures to radon are lower  but of longer duration. The strongest
data supporting an elevated lung cancer risk from  radon comes from
Sweden  (20,21,25).  It  is possible  that  the higher residential
exposures commonly found  in  Sweden  differ  in risk from  the lower
exposures found in most  of  the U.S. The authors  of  the Chinese
study  report that their findings are not  consistent  with BEIR IV
risk extrapolations. This may be due to the levels of exposure or
other  confounding factors. However,  if there is  non-linearity in
the risk at  low  levels, the effect of this non-linearity might be
that low levels of radon exposure would cause less lung cancer than
predicted by the  BEIR IV model.

     Because all epidemiologic studies are subject to design flaws,
no  single   study   can  refute  or  confirm  the  risk  estimates
extrapolated from underground miner  exposures.   For  this reason,
the risks estimated by  BEIR IV and others have not yet been clearly
confirmed or refuted with direct evidence.  It should also be noted
that most design flaws  or random errors  in exposure  measurements
will tend to produce lower  estimates of  risk than  are actually
present.  These errors  and biases must be  taken  into  account when
studying small but potentially important  elevated risks.

-------
                       EXPOSURE ASSESSMENT


     An  accurate  risk assessment requires  estimates  of both the
health risks  per  unit radon exposure  and  the level of exposure.
Since radiation related  health effects are believed to depend on
the  cumulative radiation exposure,  the  risk from radon exposure
would depend on the cumulative lifetime radon exposure.  Since most
of  the  time  spent  indoors  occurs at home,  the bulk  of radon
exposure  is received  in the home. For that  reason,  most radon
exposure research has  focused on measuring  radon  levels in homes.

THE EXPOSURE COMPONENT OF RISK ASSESSMENT FOR RADON

     Washington  State  has  one  of  the  largest  sets of  radon
measurements  of any  state  in the U.S.  In addition  to  direct
measurements of radon,  accurate geological information  pertinent to
radon is also  available  (27).

GEOLOGY AS A PREDICTOR FOR RADON EXPOSURE

     Although  nuclides of the uranium-decay  series are present in
all soils and  rocks, certain geological  formations are associated
with high  indoor  radon  levels.  Rocks  and  soils  are  the primary
source  of  indoor  radon  (28).  In  addition  to  the  soil  radium
content,  characteristics  of  density,  porosity,  and  dry  gas
permeability of the  soil are also critical  for prediction. Dryer,
loose soils (often associated with sloping  land such as that found
along rivers and streams) facilitates the transport of radon.

     Based  on  the  soil  radium content and soil   characteristics,
predictions of radon  problem  areas can be made  (29) .  The most
heavily  populated areas  of  Washington  State have soil conditions
associated with low  radon levels such as a  low radium content and
low permeability due to the wetness  of  soils.  The  lower population
areas of northeastern  Washington have  soil conditions associated
with high radon levels.

RADON TESTING  OF RESIDENCES

     Testing  of  residences has  focused on  single  family homes.
Nation-wide, 21 percent of all homes screened had  radon levels over
the Environmental  Protection Agency's (EPA) action level of 4 pCi/1
(30).

     In Washington State, a large number of  residences were tested
for  radon  through  Bonneville Power Administration's  (BPA)  radon
testing  program. Eligibility for this  program was  restricted to
single  family homes  that  were  part  of a BPA  sponsored  energy
conservation program.  As of  January, 1991  BPA tested  over 20,000
Washington State homes (31). All the radon measurements were made

-------
by Alpha-Track Detectors  (ATD's). The BPA's radon testing program
required that the detectors be placed in residences for a minimum
of  three winter  months and up  to one year  (31) .  The results  of
these tests help establish both a tentative statewide average radon
level in  houses  and  information on the  distribution  of radon
throughout the state.

      The average radon level  for Washington State, using only BPA
data,  is  1.0  pCi/1  (Fact Sheet,  Appendix 3).  The  percentage  of
houses  with  readings over  4 pCi/1  is 3.3%  compared  with  the
national  figure  of 21%.   This  finding suggests  that  the radon
problem  in Washington State is less severe than it is nationally.

SHORTCOMINGS  OF BPA SURVEY

      It  is important to examine the shortcomings of the BPA survey.
The   BPA  data  were  collected  under   the  BPA's   "Residential
Conservation Programs Radon Monitoring" research program. According
to the BPA:

      "These  data  represent only  those  readings  within the
      service areas of utilities  (and the State of Washington,
      Department of Community Development)  who participate in
      the Residential Weatherization,  Super Good  Cents,  and
      Northwest Energy Code Programs. Several  areas  in the
      Pacific  Northwest are excluded in this  evaluation.  In
      addition,  the  amount of results  received  is  heavily
      weighted  by  the number of homes monitored through large
      utility conservation  programs located in Western Oregon
      and Washington."

      The BPA survey was not designed as  a random population based
survey of residences  in  Washington State.  Since homes  were  not
randomly selected and  especially since some  counties  are over
represented  in the  BPA survey,  the  results  do  not provide   an
accurate  estimate of  radon  exposure  statewide.  For  example,
Snohomish  county,  a low radon area,  has  over 6,000 radon readings
or 29% of  all  radon measurements  in the  survey. Snohomish county
has about  10%  of  the state's population.

WASHINGTON STATE/EPA RESIDENTIAL RADON SURVEY

     To  improve the reliability  of radon exposure estimates and  to
provide more county specific exposure estimates,  the Washington DOH
is conducting  a more  valid statewide radon  survey of  homes. Free
radon testing  has been offered  to  randomly selected households
across the state. This survey  will provide  a valid profile   of
residential radon exposure in Washington State and  for selected
regions within the state.  The objectives of the project are:

      (1)   To estimate the  distribution of residential radon levels

-------
          statewide and

     (2)  To identify areas of elevated radon levels.

     The survey will place activated charcoal canister (CC) radon
testing  devices  in  approximately  2,300 residential  structures
throughout the  state.  A seven day  screening  measurement will be
obtained. A subset of 10% of  the  selected residences will receive

additional Alpha  Track Detectors (ATD)  and additional  CCs to be
used as follows:

     (1)  One ATD will be placed  on each floor of the  structure (up
          to a  maximum of four  floors)  for annual measurements.
          There is a minimum of  2 ATDs per home.

     (2)  Four  charcoal canisters will  be  provided  for seasonal
          measurements.

     The  survey began on February  5,  1991,  and the  first phase,
consisting of approximately 2,300 charcoal canister readings, will
be  completed by May  29,  1991.   The test protocol maximizes the
estimated radon level for the test homes by requiring  that tests be
made in the lowest livable area of the  residence under closed home
conditions  in  the winter months. While  this  testing strategy is
well  founded,   its readings  cannot be  interpreted  as accurate
exposure  levels for the homes. Reasonable adjustments can  be made
to  estimate  actual annual exposure  (32).

SCHOOLS

     The Department of Health's Radon Program developed the  "School
Radon Action Manual" to be used by school administrators as a guide
to  address radon issues.  The Department will  conduct  workshops
statewide on the manual this year.

     Some schools in Washington State have begun testing for radon.
The results  of  these  tests are not  complete,  but some preliminary
findings  are available:

      (1)   Radon levels for schools are high  in  the  same parts of
           the  state  where residential  radon levels are  high.  This
           is primarily in the northeastern portion of the state.

      (2)   Individual  schools can have high radon levels  even in
           counties  that  have  low   radon   average levels   for
           residences.

WORLD-WIDE RADON STANDARDS

      Since one of the charges of the Committee  was to recommend

-------
adoption  of a goal  for  residential radon levels,  the Committee
considered  the  range  of  standards  or  action  levels  adopted
internationally. National radon  standards or action levels range
from 4 pCi/1  to  20 pCi/1 for structures.  In  some countries there
are variations in  standards by type of occupancy (residential vs.
workplace), and whether or not the structures are new construction
or  existing  buildings  (33). The  International  Commission  on
Radiological  Protection  (ICRP)  has  recommended a standard of 10
pCi/1 for existing  buildings and 5  pCi/1 for new construction. It
has  been reported that both international  recommendations  and
individual  country standards  may soon be  revised downward toward
the U.S. EPA  action level of 4 pCi/1 (34).

   ESTIMATION OF RADON INDUCED LUNG CANCERS  - WASHINGTON STATE
     The following is a methodology used to calculate excess lung
cancers from residential radon exposure in Washington State.

ESTIMATES OF RISK PER WLM EXPOSURE

     BEIR IV estimates that the lifetime lung cancer risk for males
is  506  lung cancers per million  males  per WLM exposure  and 186
deaths per  million  females  per WLM of exposure.  For a combined,
equally divided male and female population, the estimated risk is
350 deaths per million population per WLM exposure.

CONVERSION FROM PCI/L TO WLM/YEAR

     Assuming that 75% of time  is  spent  in the home, the average
person would spend 3.22  times the  number of hours at home than the
average miner spends at work ([75% of 8760 hours/yr]/[170 working
hours per month]). Therefore 12  months of exposure at home results
in  (3.22  * 12) =38.6  times  the monthly hours  of exposure  for a
miner. With an equilibrium factor  of 0.5 for indoor radon, 1 pCi/1
results  in 0.005  WL of  exposure. With  a duration  for  yearly
household exposure 38.6 times the duration of  a monthly workplace
exposure, an exposure of 1 pCi/1 for one year residential exposure
translates to 0.193 WLM (38.6 * 0.005).

RISK PER PCI/L EXPOSURE

     (1)  Exposure to 1  pCi/1 for  one million males results in the
          following  risk:   (0.193  WLM  per  pCi/1   *  506  lung
          cancers/WLM) =97.7

     (2)  Exposure to 1 pCi/1 for  one million  females results in
          the following  risk:   (0.193 WLM per  pCi/1 *  186  lung
          cancers/WLM) = 35.9

     (3)  For a combined population of  500,000 males and 500,000

-------
           females  the risk  is:   (0.193  WLM  per pCi/1  * 346  lung
           cancers/WLM) =  66.8

     Washington State's average residential  radon level, based on
BPA data,  is 1.0 picocuries/liter. With  a  population of  4,657,000
the calculated number of lung cancers due to  radon  is 311  annually.

                                COSTS


     This  section provides current costs of residential testing and
mitigation.    The  two most  common commercially  available  radon
detectors  are the  charcoal canister and  the  Alpha Track  Detector.
Charcoal canisters cost $10-15 and Alpha Track Detectors cost $20-
50.  The cost of  mitigation  provided by the  U.S. EPA Office of
Radiation  Programs is listed below:


 TABLE  I.            ESTIMATED REMEDIATION  COSTS
                                  SELF
CONTRACTOR
 SIDING
 MINOR HOLES & CRACKS        $  50-100               $  100-150

 FORCED  VENTILATION          $  50-200               $  400-800

 WALL VENTILATION            $ 200-300               $ 1500-2000

 SUBSLAB SUCTION             $ 200-300               $  900-2500

Testing for radon is included in the 1990 new energy code passed by the state legislature.
The State Building Code Council (SBCC) has the responsibility for setting construction
standards. SBCC has a contract to develop a cost/benefit model to use in policy making.
When available in the summer of 1991, this cost/benefit model report will address the cost
and benefit of residential radon testing and mitigation.

-------
                                  APPENDIX 1

                     HISTORIC LUNG CANCER DEATH RATES

                              WASHINGTON STATE

       Lung cancer is the leading cause of cancer death for both males and females in
Washington  State. Lung  cancer  death rates in Washington State have  increased
dramatically since the beginning of the century. The rate has increased from 7.8 deaths per
100,000 population in 1940 to 52.5 deaths per 100,000 population in 1987. Time trend
data (Table 1) show a rapidly increasing mortality rate from the late 1940's to the
mid-1950's and again a rapid increase from the late  1960's to the late 1970's. These
increases show the effects of the epidemic initially on men in the 1940's and 1950's and
later on women in the 1960's and 1970's. Each of these epidemics correspond, with a 20
year latency period, to a dramatic increase in cigarette smoking first among men and later
among women.

       Since lung cancer is generally fatal, incidence and mortality from lung cancer are
closely related. The fraction of all cases with localized disease  at time of diagnosis has
remained stationary over time at about 20%. National data show that one year survival
rate, increased from 20% in 1950 to 31% in the early 1970's.

       The lack of a significant number of lung cancers during the  early part of the
twentieth century has provided evidence against a significant contribution of radon to the
overall lung cancer risk. Although the diagnosis of lung cancer has improved since the early
part of the century, it is believed that lung cancer should have been easily diagnosed since
the introduction of the X-ray into routine medical practice. This generally occurred during
the 1930's.

       Alternatively, a number of factors may have reduced the  contribution of radon to
the overall lung cancer risk during the early part of the century. If radon exposure interacts
with cigarette smoke exposure to cause lung cancer, then the  combination of the two
exposures produce a greater risk than the sum of the two risks. This means that much of
the increased radon related risk will occur among cigarette smokers and would not have
been observed earlier in the century.

       Indoor radon levels may have increased with time. From limited evidence on the
frequency of air exchange in older houses, many believe that less radon  would have been
drawn into older houses. As the drafts were sealed, lower air pressure  inside relative to
outside would have  occurred during the heating season. Therefore, residential radon
exposure would  have been lower during  the early part of the  century.  In  Sweden,
measurements suggest that indoor radon levels have increased fourfold since the 1950's due
to house tightening and better construction (33).

-------
                                     TABLE 1
                               Lung Cancer Mortality
                      (crude mortality per 100,000 population)
                          (Washington State - both sexes)
Year

1940
1941
1942
1943
1944
1945
1946
1947
1948
1949
1950
1951
1952
1953
1954
1955
Death
Rate
7.8
7.9
8.4
9.2
8.5
7.8
9.3
12.3
11.9
13.3
12.3
13.6
14.8
16.0
16.5
17.9
Year

1956
1957
1958
1959
1960
1961
1962
1963
1964
1965
1966
1967
1968
1969
1970
1971
Death
Rate
19.1
18.5
21.6
19.6
22.5
22.4
22.9
22.9
23.9
23.9
27.2
29.5
31.7
34.0
34.8
37.5
Year
Rate
1972
1973
1974
1975
1976
1977
1978
1979
1980
1981
1982
1983
1984
1985
1986
1987
Deat

34.4
37.9
39.1
38.9
41.1
42.7
41.6
43.4
47.8
46.6
46.5
45.2
49.8
50.5
52.3
52.5
Respiratory cancer defined according to International Classification of Diseases 7th
revision codes 160-164.

-------
                                  APPENDIX 2

                           LUNG CANCER MORTALITY

                         WASHINGTON STATE COUNTIES
  Age-adjusted death rates were calculated for Washington State counties for the period
of 1950 to 1987. Age-adjustment is necessary when comparing death rates for different
populations. Since the age distribution of populations often differ, the population with
proportionately more old people can be expected to have higher death rates. The question
of interest is usually not whether older people are at a higher risk of death, but rather
whether people of similar age residing in different areas differ in their risk of death. Instead
of comparing the death rates for each age group, an age-adjusted death rate is used as a
standardized summary measure.

  To derive an age-adjusted death rate, the age-specific mortality rates by sex for each
county  are  calculated. These age-specific county rates are  applied  to  the 1940 U.S.
population to estimate how many citizens nationwide would have died in 1940 if the entire
country had the age-specific mortality rates of this county. The age-adjusted death rate is
calculated as the number of these expected deaths divided by 1940 U.S. population counts.
For purposes of county comparisons, death rates for whites were calculated. This was done
because the non-white population of many counties, each non-white racial group has very
different smoking habits. Thus, a rate that groups all non-whites provides little meaningful
information.

   For lung cancer, the age-adjusted death rates demonstrate the magnitude and timing of
the lung cancer epidemic in the counties of Washington State (Tables 1,2). During the
period 1950 to 1959, there were eight counties without a single female lung cancer death
during the decade. Between 1980 and 1987, no county was spared a  lung cancer death.
Lung cancer death rates have doubled or tripled for women for the past 2 decades, whereas
for men, in many counties, the rates have stabilized. Since 1950, 30,133 white men and
10,513 white women have died of lung cancer. Approximately 4,000 non-whites have died
of lung cancer during this period.

-------
                                 TABLE  1

                  CANCER OF THE LUNG, TRACHEA,  BRONCHUS
                           (ICD8 162, 163, 165)

                      WASHINGTON STATE  (White-Male)
                         AGE ADJUSTED DEATH RATES
                        (# Deaths per 100,000  Pop.)
COUNTY
ADAMS
ASOTIN
BENTON
CHELAN
CLALLAM
CLARK
COLUMBIA
COWLITZ
DOUGLAS
FERRY
FRANKLIN
GARFIELD
GRANT
GRAYS HARBOR
ISLAND
JEFFERSON
KING
KITSAP
KITTITAS
KLICKITAT
LEWIS
LINCOLN
MASON
OKANOGAN
PACIFIC
PEND OREILLE
PIERCE
SAN JUAN
SKAGIT
SKAMANIA
SNOHOMISH
SPOKANE
STEVENS
THURSTON
WAHKIAKUM
WALLA WALLA
WHATCOM
WHITMAN
YAKIMA
WASHINGTON
U.S.
1950-59      1960-69       1970-79      1980-87
RATE(DEATHS) RATE(DEATHS)  RATE(DEATHS)  RATE(DEATHS)
27.8 ( 9)
17.4 ( 11)
25.5 ( 47)
28.7 ( 57)
20.4 ( 29)
23.5 (101)
37.3 ( 11)
23.3 ( 64)
22.5 ( 12)
32.4 ( 5)
25.9 ( 17)
4.7 ( 1)
17.3 ( 19)
24.3 ( 79)
29.7 ( 21)
22.8 ( 12)
31.8 (1234
27.2 (107)
18.0 ( 21)
22.0 ( 15)
28.5 ( 72)
12.4 ( 8)
18.6 ( 16)
14.3 ( 19)
30.7 ( 28)
24.7 ( 11)
27.7 (376)
19.7 ( 4)
20.6 ( 55)
25.7 ( 8)
19.6 (139)
23.5 (297)
21.2 ( 22)
23.6 ( 59)
22.1 ( 5)
24.1 ( 53)
23.7 ( 89)
19.1 ( 26)
26.3 (172)
26.4 (3331)
29.6
25.8 ( 12)
37.9 ( 29)
40.9 (100)
64.8 (147)
39.2 ( 70)
41.9 (211)
39.1 ( 12)
43.1 (122)
49.3 ( 36)
40.5 ( 5)
42.2 ( 38)
18.0 ( 3)
43.0 ( 61)
46.2 (148)
21.2 ( 21)
37.8 ( 24)
47.7 (2026)
45.8 (203)
34.4 ( 38)
36.5 ( 26)
40.3 (109)
17.2 ( 11)
39.7 ( 41)
36.3 ( 50)
48.4 ( 53)
24.3 ( 10)
45.0 (665)
44.3 ( 13)
28.2 ( 81)
56.8 ( 17)
41.6 (365)
47.3 (639)
32.5 ( 34)
42.0 (129)
55.6 ( 12)
38.3 ( 88)
35.6 (141)
27.4 ( 37)
48.7 (345)
44.1 (6174)
46.8
62.7 ( 31)
62.8 ( 53)
58.4 (184)
71.6 (182)
57.1 (145)
64.1 (420)
49.1 ( 15)
65.8 (218)
67.3 ( 61)
64.8 ( 13)
60.7 ( 68)
40.0 ( 7)
50.6 ( 99)
64.8 (220)
55.9 ( 94)
50.0 ( 49)
63.5 (2974)
70.4 (381)
50.1 ( 58)
63.0 ( 49)
56.4 (162)
48.9 ( 32)
62.6 ( 95)
69.7 (107)
73.1 ( 88)
49.0 ( 23)
69.2 (1180)
22.0 ( 12)
59.4 (194)
43.8 ( 15)
61.5 (701)
60.9 (885)
44.3 ( 52)
66.7 (263)
53.7 ( 11)
51.7 (124)
52.7 (228)
47.5 ( 66)
67.6 (518)
62.6 (10097)
64.0
64.3 ( 30)
55.1 ( 41)
77.0 (217)
75.3 (181)
69.2 (181)
79.5 (523)
34.6 ( 9)
72.6 (220)
73.2 ( 65)
32.8 ( 6)
71.0 ( 76)
84.6 ( 13)
73.7 (147)
74.5 (211)
74.8 (145)
87.6 ( 88)
63.6 (2768)
67.1 (373)
66.9 ( 73)
67.6 ( 49)
67.7 (174)
46.4 ( 28)
75.5 (122)
69.4 ( 95)
85.2 ( 85)
93.4 ( 37)
75.4 (1248)
49.6 ( 25)
70.9 (219)
74.2 ( 21)
67.8 (802)
67.1 (889)
74.5 ( 81)
70.3 (326)
57.0 ( 10)
54.3 (117)
57.6 (245)
44.2 ( 54)
73.2 (537)
68.2 (10531)


-------
                                 TABLE 2
COUNTY
ADAMS
ASOTIN
BENTON
CHELAN
CLALLAM
CLARK
COLUMBIA
COWLITZ
DOUGLAS
FERRY
FRANKLIN
GARFIELD
GRANT
GRAYS HARBOR
ISLAND
JEFFERSON
KING
KITSAP
KITTITAS
KLICKITAT
LEWIS
LINCOLN
MASON
OKANOGAN
PACIFIC
PEND OREILLE
PIERCE
SAN JUAN
SKAGIT
SKAMANIA
SNOHOMISH
SPOKANE
STEVENS
THURSTON
WAHKIAKUM
WALLA WALLA
WHATCOM
WHITMAN
YAKIMA
WASHINGTON
U.S.
   CANCER OF THE LUNG, TRACHEA, BRONCHUS
            (ICD8  162,  163,  165)

      WASHINGTON STATE (White-Female)
          AGE ADJUSTED DEATH RATES
         (# Deaths Per 100,000 Pop.)

1950-59      1960-69      1970-79      1980-87
RATE(DEATHS)  RATE(DEATHS) RATE(DEATHS) RATE(DEATHS)

2
3
4
2
3

8
1

3

2
6
3
16
5
4
6
2
6

4
2
2
6
5

4

3
4
3
6

4
3
1
5
4
5
0
.1
.4
.8
.4
.5
0
.6
.3
0
.8
0
.4
.2
.2
.3
.5
.5
.5
.0
.8
0
.5
.9
.8
.6
.2
0
.5
0
.6
.4
.6
.8
0
.3
.5
.4
.4
.8
.1
( 0)
( 2)
( 6)
( 10)
( 3)
( 14)
( 0)
( 17)
( 1)
( 0)
( 2)
( 0)
( 2)
( 16)
( 2)
( 7)
(224)
( 17)
( 7)
( 1)
( 17)
( 0)
( 3)
( 3)
( 2)
( 2)
( 70)
( 0)
( 12)
( 0)
( 24)
( 37)
( 3)
( 17)
( 0)
( 9)
( 14)
( 2)
( 34)
( 600)


7
6
6
10
8
3
8
2
0
6

3
10
5
12
8
8
4
2
8
6
10
7
9
5
7
2
6
11
6
6
3
5
12
6
5
6
6
7

0
.5
.1
.3
.5
.4
.6
.7
.9

.0
0
.9
.1
.9
.3
.5
.5
.3
.9
.2
.8
.4
.1
.3
.7
.8
.3
.3
.9
.0
.4
.1
.7
.6
.0
.2
.0
.6
.5

( 0)
( 7)
( 16)
( 15)
( 18)
( 48)
( 1)
( 27)
( 2)
( 0)
( 5)
( 0)
( 6)
( 32)
( 6)
( 7)
(447)
( 41)
( 5)
( 2)
( 23)
( 4)
( 10)
( 9)
( 8)
( 2)
(131)
( 1)
( 19)
( 3)
( 59)
(102)
( 3)
( 21)
( 2)
( 14)
( 24)
( 9)
( 49)
(1178)

11.
14.
15.
16.
22.
16.
15.
16.
9.
18.
15.
3.
16.
16.
17.
14.
18.
16.
10.
10.
16.
8.
17.
15.
16.
7.
18.
12.
13.
10.
16.
15.
7.
18.
5.
13.
13.
13.
15.
16.
7.
2
9
9
1
6
4
1
1
9
4
1
3
9
6
1
1
2
2
4
4
5
9
7
5
3
1
3
7
2
7
6
6
0
4
0
4
6
9
1
7
6
( 6)
( 15)
( 61)
( 44)
( 58)
(130)
( 5)
( 61)
( 10)
( 3)
( 19)
( 1)
( 34)
( 61)
( 30)
( 13)
(1119)
(104)
( 13)
( 8)
( 54)
( 6)
( 27)
( 24)
( 21)
( 3)
(380)
( 5)
( 49)
( 3)
(226)
(279)
( 9)
( 98)
( 1)
( 39)
( 69)
( 21)
(132)
(3241)

17
29
31
26
38
31
29
33
22
21
20
19
29
39
31
35
28
29
23
45
35
36
43
25
52
37
29
18
28
36
32
29
32
28
13
15
25
19
21
29
15
.4
.4
.2
.9
.7
.0
.4
.2
.5
.6
.9
.4
.1
.6
.7
.5
.4
.0
.2
.7
.6
.2
.7
.8
.2
.5
.0
.9
.0
.5
.4
.5
.4
.7
.9
.6
.6
.1
.9
.2
.3
( 10)
( 27)
(110)
( 77)
(113)
(242)
( 7)
(121)
( 22)
( 4)
( 25)
( 3)
( 61)
(128)
( 70)
( 33)
(1637)
(195)
( 27)
( 36)
(105)
( 19)
( 74)
( 33)
( 60)
( 15)
(599)
( 8)
(102)
( ID
(454)
(488)
( 36)
(166)
( 3)
( 39)
(127)
( 25)
(182)
(5494)


-------
                           REFERENCES

1.  Puskin, J.S., Nelson, C.B.: EPA's Perspective on Risks from
    Residential Radon Exposure. JAPCA,  39 (7):915-920, 1989.

2.  Whittemore, A.S., McMillan, A.: 1983. Lung Cancer Mortality
    Among U.S. Uranium Miners: A reappraisal. J.Natl. Cancer
    Inst. 71  (3):489-499, 1983.

3.  Waxweiler, R.J., Roscoe,  R.J., Archer, V.E., Thun, M.J.,
    Wagoner, J.K., Lundin, F.E.: Mortality Follow-up Through 1977
    of the White Underground Uranium Miners Cohort Examined by
    the US PHS. In International Conference, Radiation Hazards in
    Mining pp 823-830, 1981.

4.  Harting, F.H., Hesse, W.:Der Lungenkrebs, Die Bergkrankheit
    in Den Schneeberger Gruben. Vjschr. Gerichtl, Med. Offentl.
    Gesundheitswesen. 31:102-132, 313-337, 1979.

5.  Peller, S.: Lung Cancer Among Mine Workers in Joachimsthal.
    Human Biol 11:130-143, 1939.

6.  Tirmarche, M., Brenot, J., Piechowski, J., Chameaud, J.,
    Pradel J. The Present State of an Epidemiological Study of
    Uranium Miners in France  in Proceedings of the International
    Conference, Occupa-tion Radiation Safety in Mining Vol 1,
    Canadian Nuclear Assn, Toronto, Ontario, Canada  1985.

7.  Howe, G.R., Nair, R.C., Newcombe, H.B., Miller A.B., Frost,
    S.E., Abbatt, J.D. Lung Cancer Mortality 1950-1980  in
    Relation to Radon Daughter Exposure  in a Cohort  of Workers at
    the Eldorado  Beaverlodge  Uranium Mine. J. Natl.  Cancer Inst.
    77(2):357-362.

8.  Morrison, H.I., Semenciw, R.M., Mao,  Y., Corkill, D.A., Dory,
    A.B., de Villiers, A.J.,  Stocker H.,  Wigle, D.T.: Lung Cancer
    Mortality and Radiation Exposure Among the Newfoundland
    Fluorspar Miners  in Occupation Radiation Safety  in  Mining.
    Stocker, H. ed. Canadian  Nuclear Assn.,  1985.

9.  Edling, C., Axelson, 0.:  Quantitative Aspects of Radon
    Daughter  Exposure in Underground Miners. Br. J.  Ind. Med.
    40:182-187.

10. Fox,  A.J, Goldblatt, P. Kinlen L.J.:  A Study of  the Mortality
    of Cornish Tin Miners. Br. J. Ind. Med.  38:378.

11. Solli, M., Andersen, A.,  Straden, E., Langand, S.:  Cancer
    Incidence Among Workers Exposed to Radon and Thoron Daughters
    at a  Niobium  Mine. Scan J. Work Environ. Health  11:7-13,1985.

12. Wang,  X., Huang,  X.: Radon and Miners Lung  Cancer.  Zhonghua
    Fangshe Yixue yu  Fanghu Zazhi  4:10-14,  1984.

-------
13. Lundin, F.D., Wagoner, J.K., Archer, V.E.: Radon Daughter
    Exposures and Respiratory Cancer, Quantitative and Temporal
    Aspects. Joint Monograph No. 1. Washington, D.C.: U.S.PHS.

14. Archer, V.E., Wagoner, J.K., Lundin, F.E.: Uranium Mining and
    Cigarette Smoking Effects on Man. J. Occup. Med. 15:204-211.

15. Lundin, F.D., Archer, V.E., Wagoner, J.K.: An
    Exposure-Time-Response Model for Lung Cancer Mortality in
    Uranium Miners. In Proceeding of the Work Group at the 2nd
    Conference of the Society of Industrial and Applied
    Mathematics, Breslow, N.E., and Whittemore, A., eds, 1979.

16. Samet, J.M., Kutvirt, D.M., Waxweiler, R.J., Key, C.R.:
    Uranium Mining and Lung Cancer in Navajo Men. N.Eng.J.Med.
    310:1481-1484, 1984.

17. Cross, F.T,  Palmer, R.F., Filippy, R.E. Dagle, G.E., Stuart,
    B.O.: Carcinogenic Effects  of-Radon Daughters, Uranium Ore
    Dust and Cigarette Smoke in Beagle Dogs. Health Phys.,
    42:33-52, 1982.

18. Committee on the Biological Effects of Ionizing Radiations,
    National Research Council.  Health Risks of Radon and Other
    Internally Deposited Alpha-emitters, BEIR IV. Washington DC:
    National Academy Press, 1988:602.

19. Axelson, O., Edling, C., Kling, H.: Lung Cancer and
    Residency-a  case-Referent Study on the Possible Impact of
    Exposure to  Radon and its Daughters in Dwellings. Scand J.
    Work Environ Health 5:10-15, 1979.

20. Axelson, O., Anderson, K.,  Desai, G., et al.: Indoor Radon
    Exposure and Active and Passive Smoking in Relation to the
    Occurrence of Lung Cancer.  Scand J Work Environ Health,
    14:286-292,  1988.

21. Svensson, C., Eklung, G., Pershagen, G., Indoor Exposure to
    Radon from the Ground And Bronchial Cancer in Women.   Int
    Arch Occup Environ Health 59:123-131, 1987.

22. Simpson, S.  G.: Lung Cancer and Housing Characteristics.
    Arch Environ Health 38:248-251, 1983.

23. Lees, R. E., Steele, R., Roberts, J.H.:  A Case-Control Study
    of Lung Cancer Relative to  Domestic Radon Exposure. Int J
    Epidemiol 16:7-12, 1987.

24. New Jersey Department of Health. A Case-Control Study  of
    Radon and Lung Cancer Among New Jersey Women.  Division of
    Epidemiology and Disease Control, August, 1989.

25. Svensson, C., Pershagen, G., Klominek, J.: Lung Cancer in

-------
    Women and Type of Dwelling in Relation to Radon Exposure.
    Cancer Res 49:1861-1865, 1989.

26. Blot, W. J. , Zhao-Yi Xu, Boice, J.D., Jr., Dong-Zhe Zhao,
    Stone, B.J., Jie Sun, Li-Bing Jing, Fraumeni, J.F., Jr.:
    Indoor Radon and Lung Cancer in China. Journal of the
    National Cancer Institute.  Vol.82, NO.12. 1025-1030, June
    20, 1990.

27. Turk, B. H., et.al: Characterizing the Occurrence, Sources
    and Variability of Radon in Pacific Northwest Homes. J Air
    Waste Manage Assoc. Vol.40(4):498-506, 1990.

28. Samet, J. M., Nero, A. V.: Indoor Radon and Lung Cancer.  New
    England Journal of Medicine. Vol.320, No. 9:591-594.

29. Duval, J. S. Otton, J. K. and Jones, W. J.: Radium
    Distribution Map and Radon Potential in the Bonneville Power
    Administration Service Area. Dept. of the Interior, U. S.
    Geological Survey, 1989.

30. United States Environmental Protection Agency.  "Radon Week
    Report." October, 1990.

31. Bonneville Power Administration.  Radon Monitoring Results
    From BPA's Residential Conservation Program, January, 1990.

32. Marcinowski, Frank.  Analysis of the Relationship of Short
    Term Measurements to Annual Measurements in Support of the
    Citizen's Guide Revision.  Draft, 1990.

33. Committee on Interagency Radiation Research and Policy
    Coordination:  Federal Programs on Indoor Radon. April 1988.

34. Personal Communication with Fred Cross. March 12, 1991.

-------
                                                                  IIIP-2
              WASHINGTON STATE* S INNOVATIVE GRANT:
                    SCHOOL RADON ACTION MANUAL
                    by:   Linda B.  Chapman,  R. S.
                         Washington State Department of Health
                         Olympia,  WA 98504
                            ABSTRACT

In 1990 and 1991, the Environmental Protection Agency awarded the
Washington State Department of Health money from the State Indoor
Radon Grants Program to fund an innovative project titled
"Community Support Radon Action Team for Schools".   The goal of
the project was to establish cooperative and cost-effective
approaches school administrators could use to assess and mitigate
radon exposure in schools.

     During the first year a team of federal,  state and local
experts from a number of different fields completed the first
draft of the School Radon Action Manual.   During the second year
a full scale field test of the School Radon Action Manual was
conducted in a large suburban school district in Western
Washington.

During the field test, a team of consultants assisted school
district personnel in utilizing the manual and dealing with
community concerns aroused by the testing procedure.  At the same
time, the team members evaluated the Manual' s effectiveness.  The
results of the field test,  and the resulting modifications of the
Manual, will be discussed.

-------
                                                                          IIIP-3
Title:  Teaming Up on Local Radon Issues

Author: Robert Leker, State of North Carolina

      This paper was not received in time to be included in the preprints, and
the abstract was not available.  Please check your registration packet for a
complete copy of the paper.

-------
     Session V Posters
Radon Measurement Methods

-------
                                                                  VP-1
A DECISION-THEORETTr. MODEL FOR EVALUATING RADON TF.ST PROr.F.nTTRF.S
           BASED ON MULTIPLE SHORT-TERM MEASUREMENTS
           By: Harry Chmelynski
               S.  Cohen &  Associates
               1355 Beverly  Road
               McLean VA 22101
                           ABSTRACT
Test procedures based on multiple short-term measurements offer
an opportunity to reduce the uncertainty inherent in a single
short-term radon test result.  These procedures were examined by
the EPA in a recent revision of the Citizen's Guide to Radon.

Test performance is estimated by the percentage of homes
correctly classified over  (or under) 4 pCi/L.  Performance
calculations rely on exogenous estimates of errors due to
temporal and spatial variations in radon concentrations and
measurement error.  The joint  (multivariate) distribution of
short-term test results and long-term levels is constructed,
reflecting the serial correlation of short-term test results and
the population distribution of long-term levels in U.S. homes
from the most recent EPA survey.  Various integrals of the joint
distribution yield expected misclassification rates, costs and
benefits for a wide variety of test procedures, both simultaneous
and sequential.

-------
                                                                 VP-2
           OPF.RATTONAT. EVALUATION OF THF. RADON ALERT
                   rONTINUOUS RADON MONITOR
           By:  Emilio Braganza III  and Richard A. Levy
               U.  S. Environmental  Protection Agency
               Office of Radiation  Programs - LVF
               P.  0. Box 98517
               Las Vegas, NV 89193
                           ABSTRACT
     The U.S. EPA Office of Radiation Programs (ORP) Las Vegas
Facility  (LVF)  is the principal laboratory for evaluating new
radon/radon decay product measurement devices and developing or
modifying measurement protocols.  The devices are evaluated for
environmental sensitivity, precision, and bias, and a detailed
assessment is prepared utilizing appropriate criteria for each
instrument based on its ability to measure radon or radon decay
products.
     The Radon  Alert Instrument is a new micro-processor
controlled, diffusion type, radon gas monitoring (RGM) device
from Monitor Technologies, Ltd.  The RGM has undergone a
comprehensive evaluation at the LVF Radon Laboratory.  The
devices were exposed to radon-222 ranging from 4 pCi/L to 29
pCi/L under the following environmental conditions:  relative
humidity from 35% RH to 65% RH and temperatures from 13 »C to
29°C.  Statistical procedures were used to characterize
individual device performance, environmental sensitivity,
precision, and  bias.  Device suitability for the EPA "Indoor
Radon and Radon Decay Product Measurement Protocols" and device
eligibility for the Radon Measurement Proficiency program
Application Device Checklist was also determined.

-------
                                                                                VP-3
              A New Design for Alpha Track Detectors

                Raymond H. Johnson,  Certified Health Physicist
                              Key Technology, Inc.
                      P.O. Box 562, Jonestown, Pa  17038

      Most alpha track detectors have two sources of uncertainty that may adversely
affect the  quality  of measurement results.   First,  they may collect tracks  before
deployment for radon measurements.  When CR-39 plastic is originally produced, it is
protected from collecting background  tracks (from radon and  radon  decay products in
the air)  by a thin plastic membrane on the surface of the CR-39. When the chips of
CR-39 are mounted in alpha track devices, the protective membrane is removed.  The
entire device is  then enclosed in a metallic/mylar bag  to prevent entry  of radon  and
collection of further background tracks.  The success of this approach depends on how
well radon is excluded by  the bag.  Since most bags  will leak radon  if the partial
pressure of radon is  high  enough, or if given enough time,  background  tracks will
accumulate and may limit the shelf life of the detector.

      Secondly, once the   "radon proof  bag is opened  to  deploy the alpha track
detector, tracks begin collecting and continue collecting until the device is analyzed in
a laboratory.  That is to say,  mere is no  way to turn off the  process once the bag is
opened.   A couple of alpha track devices have provisions for placing  tape over the
opening in the detector, but this approach seals in the radon at the existing ambient
level.   The  sealed-in radon  continues to produce  alpha  tracks until  the device is
disassembled in the lab. For exposures of several months in homes at low radon levels,
the additional tracks collected by the device in transit to the lab may not be significant.
However, alpha track devices  are usually calibrated and performance tested by E.P.A.
at high levels of radon for short exposures to produce high integrated exposure levels in
pCi-days/liter. For these kinds of exposure conditions, the additional tracks collected
in transit  could be a substantial fraction of the total tracks.

      A  new device  was  designed  to correct for  both deficiencies of alpha track
detectors.  The new device provides a method for starting and  stopping the collection of
alpha tracks to define a desired exposure interval. Since the original  CR-39 is protected
by a thin membrane, that protection is maintained within the alpha track device until the
point of deployment for radon  measurement. To stop the exposure, the CR-39 is once
again covered to prevent further collection of alpha tracks.   The new  device is  also
designed such that the covering material, for stopping the collection of alpha tracks, is
protected from  the plating out of radon  decay products.  This is  to.assure that the
covering material itself will not add more tracks to the CR-39.

      This new design allows for precise exposure intervals and can be used effectively
for both short- and long-term testing  from 3 days to a year.   Since background tracks
are maintained at a consistently low level, the lower limit of detection for the new design
is about 5 pCi-days/Liter.  The upper limit of detection is about 10,000 pCi-days/Liter.
Furthermore, since the CR-39  is protected from collecting tracks while the devices are
in storage, the shelf-life of the new devices is indefinite.

-------
                                                    VP-4
Measurements of Indoor Thoron  Levels and

            Disequilibrium Factors
           Yanxia Li, Stephen D. Schery
              Physics Department
   New Mexico Institute of Mining and Technology
              Socorro, NM 87801
                   Brad Turk
                105 E. Marcy St.
                   Rm 109
              Santa Fe, NM 87501

-------
                                 Abstract








      The chemical and physical properties of radon and thoron are similar. However,




because of thoron's short half-life (55.6 s),  its transport mechanism, distribution in




the indoor air space, and the behavior of its decay products can be different from




those of radon. We took measurements in nine houses in the state  of New Mexico




to investigate diurnal thoron variations and disequilibrium of thoron progeny.  The




indoor thoron levels vary significantly with time with a maximum concentration usually



occurring  in the afternoon.  The equilibrium factor for thoron under normal indoor




conditions ranged from 0.013 to 0.084.  Predictions from a conventional well mixed



model gave values for the equilibrium factor that were a little smaller,  perhaps




because thoron wasn't well mixed in  the rooms.

-------
                                Introduction




      Our previous research has shown that in seven New Mexico houses, indoor



thoron was mainly from soil adjacent to the house, and that radon mitigation by



subfloor  depressurization  was  also  successful  for  reducing  indoor  thoron



concentrations [1]. Since health effects are more directly attributable to the alpha



particle decay of progeny, it's also important to study the behavior of thoron progeny



in indoor spaces.



      Thoron progeny in the indoor and outdoor atmosphere come  from decay of



thoron  gas, they  normally are  not  transported directly from the soil.  They are



removed by radioactive decay, by deposition on surfaces, and washout.  For indoor



air spaces, there can be an additional process of removing thoron progeny due to air



exchange between indoor and outdoor air.



      The average ground level thoron concentration over continental areas (about 2-



8 Bq  m"3 [2]) is  comparable  with that of radon.   In contrast,  indoor thoron



concentrations are usually lower than indoor radon concentrations.   The most



important contribution to the potential alpha energy concentration from progeny of



thoron, PAEC(Tn),  is 212Pb, which has a half-life of 10.6 hr, much longer than the rest



of the short-lived progeny. In the seven houses initially studied in our project [11, the



average radon concentration was about 30 times  higher than the average thoron



concentration, but  the potential alpha energy concentration for radon was only a few



times higher than that for thoron. This fact indicates that if the concentrations of



radon and thoron are the same in a certain area, the working level (WL) or PAEC(Tn)

-------
 of thoron progeny is likely to be greater than that of radon progeny. The equilibrium




 factor,  F, indicates the extent of disequilibrium between a radon isotope and  its




 progeny.  The more effective the progeny removal processes, mentioned above, the




 more the equilibrium factor is reduced.  For conditions of complete equilibrium (F = 1),




 1 WL corresponds to 3700 Bq nrV3 of radon, but to only 273.8 Bq rrr3 for thoron (1



 mWL = 20.8 nJ nY3 PAEC).




      A number of measured values of the equilibrium factor F for radon have been



 reported in the literature.  Typical values of F for indoor radon are in the range of 0.3 -




 0.6 [2]. The outdoor values for F for radon tend to be higher due to less likelihood



 of contact with surfaces. There are fewer reports for thoron  progeny and  their F



 values.  Schery [3] estimated an average indoor value for F in the range of 0.015 to



 0.04 based on the limited data and model available at the time  of his paper.  In this




 paper, we will report our most recent measurements of indoor thoron progeny and the



 equilibrium factor F.




      In addition to the -seven houses mentioned above, two more houses with




 identification DRJH and DRSS were tested in New Mexico. House DRJH is an adobe




 house with both crawlspace and basement.  House  DRSS is a single family wood




 frame house with crawlspace.  During  certain periods of testing,  blowers were




 installed to temporarily mitigate the  house by sub-floor depressurization.  Average



 pressure drops were about 3 Pa. A new thoron detector operating on the principle of




 coincidence between decay of 220Rn and 216Po was  used  in houses  DRJH and DRSS




to obtain radon and thoron concentrations. Progeny concentrations were measured

-------
 with the Scintrex model WLM-30.  As with the other seven houses, the soils at these

 two additional houses had uranium and thorium concentrations slightly above, but

 close to, the average concentrations for american soils.
                                Experiments


 Diurnal pattern of indoor thoron concentration

       The variation of indoor thoron concentration with time depends not only on the

 exhalation rate from the sources, but also on variation of ventilation conditions and

 air exchange between rooms. These are caused by meteorological conditions (wind,

 barometric pressure, temperature) and human activities such as opening windows and

 doors. Fig. 1 shows the mean diurnal indoor thoron concentrations for eight tested

 houses. House TI42 is excluded because the  experimental data are  insufficient to

 represent the  diurnal thoron  variations and  there  is  no continuous  progeny
                                                     *
 measurements available.  The experimental data were obtained during Jan. 1991 to

 May 1991 and  in March 1992.  The data shown in fig. 1 are obtained by averaging

 several days (the  number of days for each house is listed in table 1, column 7) of 2-

 hour interval (or 1 hour for DRSS and DRJH) semi-continuous measurements with the

 mitigation systems turned off. In house DRSS, a fan was used in the room where the

 measurements  took place to achieve better air  mixing.  In all the  other houses,

 measurements were made under the existing  air mixing conditions.  The average

sampling height for thoron gas is about 1 m above the floor and 0.1 m for progeny

-------
(about 1  m for house DRSS and PE21).




      The observed changes of thoron concentration during a day varied by a factor




of 2-10.  The maxima of thoron concentrations tend to occur in the afternoon and the



minima in the early morning.  The research on thoron emanation rate from soil has




shown that  the variation of thoron emanation is often characterized by a morning



minimum  with  the maximum occurring  during the afternoon due to  barometric




pressure  drop and temperature gradient [4, 5]. The increasing emanation rate could




possibly cause the  increase in indoor thoron concentrations, but observed emanation




rate  increases tend to  be smaller [4, 5],  and the houses will tend to insulate the



underlying soil from temperature change.  The potential alpha energy concentration




should increase when the parent gas entry  is increased. Hence, after adjusting for the




appropriate time delay, PAEC(Tn) can  be used  to  indicate the changes in thoron




sources in the afternoon, assuming deposition rate and ventilation have not changed




significantly.   For  the above eight  houses, the evidence from the PAEC(Tn)




measurements suggests the afternoon maxima of thoron concentration could not be




fully  explained by an increase in entry rate.   Another  possible explanation is that




thoron is  not well mixed in indoor space due  to its short half-life, and the extent of




mixing can vary with time. If this hypothesis  is correct, thoron concentration will be




higher near the sources.  For the  houses  where soil is considered  to be the major




source of indoor thoron, the highest thoron concentration would normally be found




near the floor. Therefore, a concentration gradient could exist between the floor level




and the measurement height of 1  m. In the  afternoon, if the air mixing process is

-------
faster due to convective processes or human activity, it may cause a better mix of



thoron. Therefore, the detector which is set 1 m from the floor may be able to detect




more decays than at any other time. The thoron data for house DRSS support this



hypothesis about mixing,  since this house had the greatest mixing and the diurnal




variation of thoron showed the least change, only about a factor of 2.




       As discussed in the literature [3], over the range of normal ventilation rates for




houses, due to its short half-life, the indoor thoron concentration is unlikely to  be




affected directly  by  changes in ventilation.   So ventilation changes are an  unlikely



explanation of the afternoon maximum.








Disequilibrium Between Thoron and Thoron Progeny




       The  investigation of disequilibrium between thoron and thoron progeny in indoor




space needs the  simultaneous measurements of both thoron and thoron progeny




concentrations. The instruments we used to measure progeny were described in  an




earlier paper about indoor thoron sources [1 ].  Because the WLM-30 wasn't available




during the  period  of testing house TI42, the measurements of the equilibrium factor




in this house are excluded. There is a considerable delay in the response of the WLM-




30  to thoron progeny, and a  correction for this  delay has to be made in  making



estimates of  F values.




      Thoron progeny are more sensitive to the changes in ventilation rate and aerosol




concentration caused by various house operation  and human  activities.   So the



equilibrium factor F of thoron is not necessarily stable during the day and night. The

-------
 radon mitigation systems installed in the houses did not have significant effects on



 either ventilation rates or aerosol concentrations,  and it is thus not surprising that




 there were no changes of the  equilibrium factor F observed due to the operation of



 the mitigation systems.




      Table I  lists the mean of measured concentration data  for each house.  The




 concentrations of thoron (C) and progeny PAEC(Tn) are the daily averages obtained




 from  measurements on the days (column 7) when the mitigation systems were off.




 The equilibrium factor Fm is deduced from the means of C and PAEC(Tn).  The average




 equilibrium factors are low in comparison with typical F values for radon. Variation




 in aerosol concentration is one important factor that can cause variation in F values.




 An increased aerosol concentration tends to increase the fraction of attached progeny




 in the air and cause a decrease in progeny plateout  rate.  The  result is that the




 progeny concentration in the air is increased if other parameters are kept unchanged,



 and the F value will increase.




      Grab sample measurements of aerosol concentrations were made at different




 locations of the houses using a TSI model 8510 with the impactor of a size range of




 0.01 to 3.5 /;m.  A moderate correlation between Fm values and measured aerosol




 concentrations is shown in fig. 2.  The correlation was probably not stronger because:




 1) the aerosol concentrations are from the grab samples which may not be equivalent




to the average over the  period when the  measurements were taken,  and 2) F is




influenced by many factors, for example, ventilation rate and ratio R of surface area




 (including walls, floor, ceiling, and furniture) to the  volume of the house. The value
                                      8

-------
of R changes from house to house. Houses with bigger R values would be expected



to have smaller F values due to the increased probability of progeny deposition on the



surfaces. Simultaneous measurements of aerosol concentration and ventilation rate



would provide a more accurate interpretation of variation in the equilibrium factor.








                    Comparison with Model Predictions




       The prediction of model calculations is very helpful in interpreting the influence



of the  various parameters on indoor radioactivity concentrations.  The well-mixed



room model [6] with first order sink and source terms is used in this paper to compare



predicted equilibrium factor (Fc) with the experimental data. The model calculates the



influence of ventilation rate, aerosol concentration, the physical process of plateout



 (deposition), and attachment to aerosol on the equilibrium factor.  The steady-state



concentrations of the unattached C(j and attached C" decay products are
                                                                         (1)
 and
 where the index j = 0, 1, 2, and 3 are for 220Rn, 216Po, 212Pb, and 212Bi, respectively.



 The superscript (a) is for the attached progeny, while the superscript (f) designates



 free (or unattached) progeny.  The total activity concentration of jth decay product Cj




 is

-------
The concentration C is in units of Bq m"3. The decay constant h, ventilation rate u,




deposition  rate q,  and attachment  rate X  are  in  hr1.  The  recoil fraction r is




dimensionless and is nonzero for 216Po only.  The attached outdoor concentration is




Cj°a.   Equations (1) and  (2) assume the unattached progeny concentrations from




outdoors are negligible. The attachment rate X is a function of aerosol concentration




Z (cm'3)  and can be  written as  X=@Z, where  /? in  cm3 hr1,  is the attachment




coefficient.




      To simplify the  calculation  a little further,  assume the activity concentration




outdoors for all isotopes  is negligible, C°' = Q. This assumption is valid for thoron




because the ventilation rate which determines the source entry rate from outside is




much smaller than its  decay constant. The assumption for progeny,  although often




true, is only an approximation.  For thoron progeny,




                          PAEC(T/})=69.1C2 + 6.56C3                     (4)
where PAEC(Tn) is in units of nJ m"3.  The equilibrium factor F of thoron is defined as,
 where C0 is the concentration of thoron.



      The comparison between the measured and calculated F values, Fm and Fc,




 respectively, are shown in table I.  The following parameter values were used in the



 calculations: qa = 0.2 hr1, qf = 40 hr1, 0 = 0.005 cm3 hr1, r, =0.5, and u =0.75 hr1 [3].
                                      10

-------
 Since the aerosol measurements were in units of mass concentration, //g m  , a



conversion of 9//g rrf3 corresponding to 20,000 particles cm3, was applied to derive



the attachment rate 0.  This approximate conversion factor  was obtained for our



study conditions by comparing the TSI 8510 side-by-side  with a condensation nuclei



counter  (Environmental One  model Rich 200).  The F values from the model



calculation are a little lower than the measured for most of the cases.  The measured



F values may be higher than the true values because of the underestimate of thoron



concentrations due to thoron's inhomogeneous distribution. However, the parameters



 q", qf, 0 and u were fixed at values taken from the literature. The value u was verified



 as reasonable for our houses, but the remaining parameters can vary, and are not well



 known for thoron, so could explain some of the discrepancies.








                                Conclusions








       The experimental data have shown that  indoor thoron concentrations vary with



 time.  The daily variation can be a factor of  2-10 depending on the  meteorological



 conditions.  The maxima tended to occur in the late afternoon possibly due to better



 mixing of indoor air during that period.  The working level of thoron progeny was



 comparable with that of radon progeny even though the gas concentration of thoron



 was much lower. The study of the thoron  equilibrium  factor indicates that in the



 indoor air of our houses, the disequilibrium between thoron and its progeny is much



 greater than for radon,  with an average F value generally below 0.1.
                                      11

-------
      Although thoron and its progeny in these houses were not high enough to be



a health hazard, they can be a health concern in other areas.  This study provides



information, such as F values, that could be useful for estimating dose in other



situations.  If we assume that the primary source of indoor thoron is the soil around



a building and  that other variables are equal, then  the indoor thoron concentration



should approximately scale with the thorium content of the soil. To better understand



the behavior of thoron  and thoron progeny in indoor spaces,  future studies would



need to obtain more detailed information on the ratio R, ventilation rate, aerosol



concentration, and outdoor activity concentrations.








                                References




1.    Li, Yanxia; Schery, Steve; Turk, Brad Soil as a Source of Indoor 220Rn.  Health



      Phys. Vol. 62, No. 5, May  1992: 453-457.



2.    UNSCEAR Ionizing Radiation: Sources and Biological Effects.  1982.



3.    Schery,  Stephen D.  Radon  Isotopes and Their Progeny  in  the  Indoor



      Environment.  Encyclopedia of  Environmental  Control Technology,  Vol.  2,



      Chapter 23, Gulf Publishing, Houston, Tx, 1989.



4.    Druilhet, A.; Guedalia, D; Fontan, J.;  Laurent, J.L.  Study  of  Radon 220



      Emanation Deduced from Measurement of Vertical Profile in the Atmosphere.




      J. Geo.  Res. 77, 1972: 6508-6514.



 5.    Guedialia, D; Laurent, J.L.; Fontan, J; Blanc, D.; Druilhet, A. A Study of Radon



      220 Emanation from soil.  J. Geo. Res.  75, 1970:  357-369.






                                     12

-------
6.    Porstendofer, J. Behavior of Radon Daughter Products in Indoor Air. Had. Prot.




      Dosim.  7, 1984: 107-113.
                                    13

-------
  40
co
6
cr
PQ
 o
•l-H
-P
 cd
 QJ
 O
 Ł
 o
 o

 CJ
 o
30
   10
    0
     0:00
                6:00        12:00
                         time (MST)
18:00
24:00
      Figure 1  Diurnal pattern of indoor thoron concentration

      in the houses.

    Houses  with crawlspaces:  * DRJH,  *  DRSS,  • CE11,  *  PE21.

    Houses  with slab-on-grade:  o  AL02,  * AL04, *  SF31,  °  TI41

-------
    0.10
    0.08
o
cd
0.06
0)
    0.04
    0.02
    0.00
         0
             5      10      15     20     25      30

                aerosol  concentration  (/xg  m~3)
35
            Figure 2   The  correlation  between  aerosol  concentration

                     and  measured  equilibrium  factor. The solid line

                     is fitted with F=0.017 + 0.002*aerosol concentration.

-------
Table I   Equilibrium factors,  and thoron and thoron progeny data for periods with
mitigation  systems  off.
HOUSE
AL02
AL04
CE11
PE21
SF31
TI41
DRJH
DRSS
MEAN THORON
CONCENTRATION
C (Bq nr3)
5.7
(3.3-10.4)
1.4
(0.4-2.3)
8.2
(2.2-21.1)
10.7
(5.2-19.1)
6.2
(2.1-8.1)
4.9
(0.4-10.8)
13.5
(8.0-22.9)
26.0
(14.7-30.9)
MEAN
PAEC(Tn)
(mWL)
1.1
(1.0-1.5)
0.1
(0.1-0.2)
1.9
(1.7-2.1)
2.3
(0.3-3.7)
1.0
(0.7-1.4)
1.5
(1.2-2.2)
4.1
(3.8-4.4)
1.2
(0.9-1.6)
AEROSOL
CONCENTRATION
(Mg HT3)
9
9
18
12
—
33
30
6
MEASURED
F VALUE
Fm
0.053
0.020
0.064
0.059
0.044
0.084
0.083
0.013
MODEL
PREDICTION
FC
0.044
0.044
0.051
0.047
0.045*
0.055
0.055
0.038
PERIOD
(days)
5
1
8
2
2
1
4
4
    * The aerosol concentration of  10 ng  nr3  is used  in  the  calculation.

-------
                                                                                    VP-5
        COMPARISON OF CONTINUOUS AND OCCUPANCY TIME RADON
        MEASUREMENTS IN SCHOOLS USING PROGRAMMABLE E-PERMS

                   by:   Marvin H. Haapala, Craig A. Dewitt, Robert W. Power, and
                         Robert A. Fjeld
                         Clemson University
                         Clemson, SC 29634-0919
                                   ABSTRACT
       A study of indoor radon in a county-wide school district in northwestern South
Carolina was conducted. The objectives of the survey were (1) to compare the results of
continuous (24 hour) and occupancy time (12 hour) measurements and (2) to add to the base
of data on radon in schools.  Continuous measurements were made in about 100 schools and
almost 3000 rooms.  Simultaneous occupancy measurements were made in over 600 of the
rooms.  The study was performed using portable, battery operated, electromechanically
actuated electret radon monitors that permit measurements to be made over pre-specified
periods of time.  Continuous measurements were for a three day period of time while school
was in session.  Occupancy measurements were from 6:00 am until 6:00 pm during the same
three day period.

       The arithmetic mean continuous concentration was 2.4 pCi L' (89 Bq m3) and fifteen
percent of the measurements exceeded 4 pCi L' (148 Bq m'3).  Fifty-nine percent of the
schools had at least one room with a concentration in excess of 4 pCi L' (148 Bq m3).  For
cumulative probabilities above 25%, the distribution was approximately lognormal with a
geometric mean of 1.8 pCi L1 (67 Bq m3) and  a geometric standard deviation of 2.2.  A
qualitative comparison of the occupancy and continuous measurements revealed no obvious
systematic differences.  Overall, the mean difference between occupancy and continuous
concentrations was not statistically significant.  However, in  16 percent of the rooms, the
occupancy concentration was significantly larger than the continuous  concentration and in 18
percent of the rooms the continuous concentration was significantly larger  than the occupancy
concentration.

       The work described in this paper was funded by the U.S. Environmental Protection
Agency's State Indoor Radon Grant for South Carolina. The contents do not necessarily
reflect the views of the agency and no official endorsement shall be inferred.

-------
                                  INTRODUCTION
      Since students and staff spend anywhere from six to eight hours a day in school, the
potential exists for them to receive significant radon exposures during the school day.  The
Environmental Protection Agency tested 130 schools in 16 states,  focusing on areas with a
potential for elevated concentrations.  The EPA study included a total of 3000 classrooms.
Nineteen percent of the classrooms had screening measurements in excess of 4 pCi L1 (148
Bq nr3), and three percent of the classrooms had concentrations in excess of 20 pCi L'1 (740
Bq m3). Fifty-four percent of the schools had at least one classroom with a screening
measurement in excess of 4 pCi L'1 (148 Bq m3).

      Radon concentrations  in schools can show considerable variability with time (1),
possibly more  so than in houses.  This is because ventilation rates and HVAC system
operation during occupied periods during the day can be quite different from those at night.
As a consequence,  the average radon  concentration during a continuous measurement may
differ from the average concentration  during periods of school occupancy.  In a study
involving schools in Iowa, Maryland, and North Carolina, Wiggers gt al. (2) found average
concentrations to vary widely during occupied and  unoccupied periods.  In  some buildings,
concentrations were greater during occupied periods and in  others concentrations were
greater during unoccupied periods.

       Presented in this paper are results of a study in a county-wide school district in South
Carolina.  The objectives of the project were twofold.  The first was to compare the results
of continuous and occupancy measurements and the second  was to add to the base of data
associated with radon in schools.

                                   METHODOLOGY
       The study was performed in the Greenville County School District in northwestern
 South Carolina.  Previously, between 15% and 25% of the houses in Greenville County have
 had radon screening measurements in excess of 4 pCi L! (148 Bq m3)(3,4). The district has
 97 schools with 2963  classrooms and other frequently used rooms on or below grade.  The
 schools are spread over approximately 800 square miles. The survey began in late January
 1992 and ended in late May 1992, with approximately 160 to 180 rooms being tested each
 week.

       Continuous measurements were made in each of the 2963 rooms.  The measurement
 period was from 6 am Tuesday until 6 am Friday for a total of 72 hours.  Twenty percent of
 the 2963 rooms were  randomly selected for an occupancy measurement.  Occupancy
 measurements were made in 605 of the rooms.  The occupancy measurement period was
 from 6 am to 6 pm on Tuesday, Wednesday, and Thursday for a total of  36 hours.

-------
      The measurements were made with S-type E-PERMS1 (electret-passive environmental
radon monitors)(5).  Each E-PERM was modified to permit  intermittent on/off operation of
the monitor with a programmable, battery operated, electronic timer originally intended to
operate a water flow control valve as shown in Figure la. The timer was altered to attach to
the E-PERM and to operate the plunger that turns the monitor on or off.  Each monitor had a
bar-coded label which included a school identification number, school name, room number,
and type of measurement.  The monitors  were placed in small, ventilated, locked metal boxes
to prevent tampering.  Rooms designated  for occupancy measurements had both a continuous
and an occupancy monitor in each box  as shown in Figure Ib.  The boxes were delivered to
the schools on Mondays for distribution to the classrooms. They were retrieved on Fridays
and returned to the lab where the electret voltage was read and entered into a database along
with the information on the bar code by means of a computerized reader.2  Processing time
for the 200 to 230 measurements per week  (includes continuous, occupancy, duplicates, and
blanks) was two to three hours.
a.  Programmable E-PERM

Figure 1. Measurement Apparatus
b. Continuous and occupancy monitors in
    metal box
       The test program was conducted in accordance with the quality assurance guidelines
of EPA's Radon Measurement Proficiency Program (RMP)(6) and included 230 duplicate
continuous measurements, 60 duplicate occupancy measurements, and  120 control
measurements (blanks).  A total of 4002 measurement were made during the course of the
study.  At 4 pCi L1 (148 Bq m3) the measurement uncertainty was  ± 0.34 pCi L1 (12.6 Bq
m3) for continuous measurements and  ± 0.53 pCi L' (19.6 Bq m3) for occupancy
       Rad-Elec, Inc., Frederick, MD

    :  Empac Incorporated, New Haven, CT

-------
measurements.  The occupancy measurement lower limit of detection (operationally defined
as the concentration at which the relative error is ± 50%) was estimated to be 0.95 pCi L1
(35 Bq m3).

                                      RESULTS
       Of the 97 buildings included in the survey, 57 had at least one room with a
concentration in excess of 4 pCi L'1 (148 Bq nr3). However, most of the elevated readings
were found in a few schools.  The majority of these schools were located in the northern part
of the county, which extends into the foothills of the Blue Ridge mountains.

       Presented in Fig.  2 is a cumulative log-probability plot of the continuous
measurements.  The median concentration was 1.7 pCi L'1 (63 Bq nr3) and the maximum
concentration was  21.4 pCi  L1 (792 Bq  m3).  Approximately 15% of the measurements were
in excess of 4 pCi L1 (148 Bq m3). Data above the 25th percentile are almost linear on the
log-probability plot. Their distribution can be approximated as  lognomal with a geometric
mean of 1.7 pCi L'1 (63 Bq  m3)  and a geometric standard deviation of 2.2.  These results
are compared with data obtained in house surveys for Greenville County in Table 1.  The
results obtained here for schools are similar to those obtained previously for houses.
                                   Radon Cono«crtrattoo in pCVU
                                                      9   7  9 91'
 Figure 2.  Cumulative log-probability plot of the radon concentrations in the 2992 rooms.

-------
     TABLE 1.  COMPARISON OF RESULTS OF THIS SURVEY OF GREENVILLE
             COUNTY SCHOOLS AND PREVIOUS STUDIES OF HOUSES.
Survey
Schools
Houses (3)
Houses (4)
Number of
Observations
2954
22
89
Arithmetic
Mean
2.4
2.8
3.1
Geometric
Mean
1.8
2.1
1.5
Percentage
> 4 pCi L ( Bq m )
15
23
15
       Simultaneous continuous and occupancy measurements were made in 605 rooms. A
 graphical comparison of the data is presented in Fig. 3.
                   20 -
                      0        4        8      12       16      20

                             CONTINUOUS CONCENTRATION (pCl/L)

Figure 3.  Comparison of occupancy and continuous measurements in 605 rooms.

The ordinate is the occupancy concentration and the abscissa is the continuous concentration.
The data are almost uniformly scattered about the diagonal, and there is no indication of a
systematic difference between the two  measurements.  However, there were a substantial
number of rooms for which there was  a difference between the occupancy and continuous
concentrations. Presented in Table 2 is the number of rooms in which the occupancy and

-------
 continuous concentrations differed for various concentration ranges.  The decision rule for
 this table was that the concentration difference exceed 0.8 pCi L1 (30 Bq m3).  This is the
 sum of the uncertainty estimates for continuous and occupancy measurements  which were
 0.3 pCi L-' and 0.5 pCi L'  (11 Bq nr3 and 19  Bq nr3), respectively. Overall,  the occupancy
 concentration exceeded the continuous concentration in 16 percent of the rooms and the
 continuous concentration exceeded the occupancy concentration in 18 percent of the rooms.


          TABLE 2.  COMPARISONS OF OCCUPANCY AND CONTINUOUS
                                CONCENTRATIONS
Concentration
Range
0 - 1 pCi L '
1 - 2 pCi L '
2 - 4 pCi L '
> 4 pCi L '
No.
Rooms
190
292
259
100
Occupancy
> Continuous
48
43
47
32
Occupancy
< Continuous
18
52
73
26
      Of particular interest are the instances in which occupancy concentrations exceeded
continuous concentrations.  This is because radon exposures would be underestimated if
based upon the continuous measurement.  The extremes are worth noting. In three rooms
the continuous concentration was between 1 pCi L' (37 Bq m3) and 2 pCi L'1 (74 Bq nr3) and
the occupancy concentration was over 6 pCi L1 (222 Bq nr3), in one room the continuous
concentration was 12 pCi L'1 (444 Bq nr3) and the occupancy concentration was 18 pCi L'
(666 Bq nr3), and in ten rooms the continuous concentration was  less than 2 pCi L1 (74 Bq
nr3) and the occupancy concentration was over 3 pCi L1 (111  Bq nr3).

      Presented in Table 3 are summary statistics for the continuous and occupancy
measurements.  The arithmetic means were equal, and the percentage of measurements in
excess of 4 pCi  L1  (148 Bq m3) of the two distributions were similar.
    TABLE 3.  SUMMARY STATISTICS FOR OCCUPANCY AND CONTINUOUS
                                 MEASUREMENTS
Measurement
Type
Occupancy
Continuous
Arithmetic
Mean
2.4
2.4
Geometric
Mean
2.4
1.7
Geometric
Standard Dev
2.4
2.2
Percentage
>4pCiL-' (148Bqnr3)
14
13

-------
         Paired difference tests were performed to determine if there was a statistically
  significant difference between the occupancy and continuous concentrations. Since the data
  more closely followed a lognormal distribution than a normal  distribution, the tests were
  applied to the natural logarithm of the concentrations. These tests were applied to different
  concentration ranges (0 - 1, 1 - 2, 2 - 4, >4 pCi L1 (148 Bq  m3)) and different
  measurement periods (January-May, January-March, April-May).  Overall, the difference
  between occupancy and continuous measurements was not statistically significant (p = 0 25)
  For some  of the measurement periods and concentration ranges, the mean differences were  '
  significant and for others they were not.  For the cases in which the differences were
  statistically significant, sometimes the occupancy concentration was larger and sometimes the
  continuous concentration was larger.  The largest difference was less than ten percent  Thus
  the statistical analyses support the qualitative observation that there were no systematic
  differences between continuous and occupancy concentrations.

                           SUMMARY AND CONCLUSIONS


        Overall, radon levels in the schools were consistent with those in houses  in the
  county. The arithmetic mean continuous concentration was 2 4 pCi L' (89 Bq m3)  the

  ^••T nlfS, ^i^ ^ and 15 t*™" °f thŁ r°°ms had «»«ntotions in excess
  or 4 pci L  (148 Bq m).  For cumulative probabilities above 25%, the distribution of
  continuous  concentrations was approximately lognormal with a geometric mean of 1  7 pCi L'
  (63 Bq m ) and a geometric standard deviation of 2.3.

        Qualitatively, there were no obvious systematic differences between continuous and
 occupancy concentrations.  This observation was supported by statistical tests performed on
 the mean difference between the two measurements.  However, when  considered on a room
 by room basis, significant differences were observed in 34% of the rooms.  The occupancy
 concentration was greater that the continuous concentration in 16 percent of the rooms and
 the continuous concentration was greater than the occupancy concentration in  18 percent of
 the rooms.  In a few rooms, the occupancy concentration was substantially greater than the
 continuous concentration.  The implication is that radon exposure in these rooms  would be
 underestimated significantly if based upon a continuous measurement.

                                    REFERENCES


 1.     Environmental Protection Agency.  Radon Measurements in Schools, An Interim
       Report.  EPA-520/1-89-010. U.S. Environmental Protection Agency, Office of
       Radiation Programs,  Washington, DC, 1989.

2.     Wiggers, K.D., Bullers, T.D.,  Zoske, P.A., Leovic, K.W., and Saum D  W
       Electret ion chambers for radon measurements  in schools during occupied and

-------
       unoccupied periods.  Proceedings of The 1990 International Symposium on Radon and
       Radon Reduction Technology.  February 19-23, 1990, Atlanta, GA.

3.     Fjeld, R.A., Jones, M.D., and Bivens, N.  Screening survey of indoor radon in South
       Carolina.  Health Phys. 59(2):  217, 1990.

4.     Fjeld, R.A. and Shealy, H.  State Indoor Radon Grant, Progress Report to April 15,
       1991.  Submitted to U. S. Environmental Protection Agency, April, 1991.

5.     Kotrappa, P., Dempsey, J.C., Ramsey, R.W. and Stieff, L.R.  A practical E-PERM
       (electret passive environmental  radon monitor) system for indoor K2Rn measurement
       Health Phys. 58(4): 461, 1990.

6.     Environmental Protection Agency.  Radon Measurement Proficiency (RMP) Program,
       Handbook. EPA 520/1-91-006. U.S. Environmental Protection Agency,  Office of
       Radiation  Programs, Washington, DC,  1991.

-------
                                                                               VP-6
                   INDOOR RADON  IN NEW YORK STATE SCHOOLS

                              PRELIMINARY  REPORT

                   Susan VanOrt*. Charles  Kunz+, Laurence  Keefe*,
                   William Condon*,  Kirk Fisher+ and  Karim Rimawi*

                  *New York State Department  of Health
                   Bureau of Environmental Radiation  Protection
                   2 University  Place
                   Albany, New York   12203

                  +New York State Department  of Health
                   Wadsworth Center  for Laboratories  and  Research
                   Empire State  Plaza
                   Albany, New York   12201-0509


                           ABSTRACT
     New York State is participating  in  a  project  to study radon in schools
funded in part through a grant from the  EPA.   The  program began in the spring
of 1991 and is scheduled to run for three  years.   Candidate schools are
selected from areas in which existing information  suggests there may be a high
risk for indoor radon.  These schools are  invited  to participate in an indoor
radon survey that includes short term screening  measurements,  one month to
three month follow-up measurements  in rooms  with greater than  or equal to
4 pCi/1, and one year measurements.   Follow-up measurements will be made if a
school is mitigated.  Twenty-two schools are being surveyed in the first year
(1991/1992) and forty in the second.   The  soils  under and around about one-
third of the schools surveyed will  be characterized for indoor radon potential
through soil gas measurements and examined for correlation with indoor radon
concentrations.

     The information presented in this paper details initial results of short-
term screening measurements made in twenty-two schools and soil
characterization in three of these  schools.

     This paper has been reviewed in  accordance  with the United States
Environmental Protection Agency's peer and administrative review policies and
approved for presentation and publication.

     Mention of trade names or commercial  products in this document does not
constitute endorsement or recommendation for their use.

-------
                             BACKGROUND


     Indoor radon measurements in single family houses throughout New York
State have revealed many areas with geometric mean radon concentrations
greater than 4 pCi/1.  Consequently, the New York State Department of Health
(NYS DON) has become increasingly concerned with the radon-related health risk
to school-age children in the areas with high radon potential.  The purpose of
this study is to screen indoor radon concentrations in school buildings in
areas of the state known to have elevated levels of indoor radon.  The study's
main objectives are to 1) obtain information regarding the extent and magnitude
of the radon risk in New York State schools, 2) learn more about how large
buildings behave with regard to radon entry and distribution, and 3) provide
instruction about testing procedures and temporary remediation techniques to
school officials so that they will be able to conduct radon tests on their own
in other school buildings.


                            SCHOOL SELECTION


     The two basic criteria used to select candidate schools are that the
schools should be located in high risk areas for above average indoor radon
and that the schools are distributed around the state to provide information to
as many school districts and areas as possible.  For the past several years,
the NYS DOH has been studying the geologic factors affecting indoor radon
throughout the state and at the same time the Department has been
accumulating a large database on indoor air radon concentrations in
residences across the state.  The geologic data and the extensive database of
indoor radon screening measurements in homes are used to select schools from
areas with potential for above-average indoor radon.  Cities, towns and
villages containing schools are matched with the cities, towns and villages
with the highest levels of indoor radon.  Schools in areas with high levels of
residential  radon are then located on topographical and surficial geology maps
to obtain information on geologic factors such as gravelly, permeable soils
that often correlate with above-average indoor radon.   Schools that are
located in areas with above-average indoor radon and/or areas with surficial
geologic factors that correlate with above-average indoor radon are considered
as candidate schools.  Generally, only one school per school  district is
selected so that schools are distributed throughout the state while still
targeting high-risk areas.

     In the spring of 1991,  53 candidate schools were selected and asked to
volunteer for participation in the school survey project.   Twenty schools
volunteered and participated in the project.  Two additional  schools not
included in the 53 candidate schools also volunteered and were measured in the
first year.   Several schools declined participation in the first year but
asked to be considered for the second year of the project.

-------
      Candidate schools are sent a letter explaining the indoor radon testing
 program and requesting their participation.   Those wishing to participate in
 the program are asked to fill  out a  questionnaire  of general  information about
 the school building and to return copies of  floor  plans.   This information is
 used in the selection process  and to develop a  measurement strategy for the
 selected school buildings.

      Once a school building is selected,  the school  is  notified  and requested
 to designate an individual  from the  school as the  primary  contact    We  prefer
 that this individual  take  an active  role  in  the  program from  initial  screening
 tests to post remediation  testing.   This  person  will  thus,  become  knowledgeable
 enough to carry out or supervise  testing  in  other  school buildings  in the
 district.                                                        a

      Figure 1  shows the  average basement  screening  radon concentrations  for
 each County in  the state and the  number of schools  surveyed in each  County
 during  the  first year  of the project.  The twenty-two schools  participating
 in year one of  the project  are  located in seventeen Counties   Nine  of  the
 schools were  located  in Counties with average basement  screening concentrations
 above 8 pCi/1,  twelve  in Counties averaging between 4 to 7.9 PCi/l and one in
 a County with between  2 to  3.9 pCi/1.

     When testing  is about  to commence,  a box is shipped to the school that
 includes detectors, detector deployment  instructions, data sheets and a floor
 plan that is marked with room location and quality assurance detectors    The
 detector deployment instructions explain  where to place the detectors 'how to
 record  information on the data  sheets, how long  to test for and where'to  ship
 the detectors at the end of the test  period.   In general,  this distribution by
mail  has been successful  in 1991/1992 and we  plan to continue  with this  format
 tor rear Z.

-------
                           MEASUREMENT PROTOCOLS


     All schools have long-term (one year) alpha track detectors1 and
short-term (three day) E-PERMs2 placed in every room at and below grade and
a sampling of rooms on second and third floors.  E-PERMs are utilized for
short-term (three day) and follow-up measurements (one and three months).
Table 1 details the follow-up measurement protocols.  Additional detectors
are also provided to the school for placing duplicates in about 10% of the
rooms for quality control.   An in depth quality assurance plan for E-PERM,
alpha track detectors, charcoal and continous monitoring units is in place.

     If radon in a routinely occupied room measures greater than 100 pCi/1,
Health Department personnel visit the school to collect grab samples, conduct
follow-up inspections to identify obvious entry points, deploy additional
short-term electrets and possibly collect some long-term electrets.   Based on
these results, the school will be advised regarding possible temporary
remediation,  further testing and eligibility for inclusion in the New York
State Energy Office school  diagnostics program.

     If schools participating in the measurement program undergo any
remediation to reduce indoor radon concentrations, the Health Department will
supply short and long-term electrets for post-remediation measurements
following each phase of the remediation process.
1MAn ATD is a small piece of plastic or film enclosed in a container with
 a filter-covered opening.  Radon diffuses through the filter into the
 container and alpha particles emitted by the radon decay and its products
 strike the detector and produce submicroscopic damage called alpha
 tracks."(1)

2An E-PERM (Electret-Passive Environmental Radon Monitor) System.  "The E-PERM
 Chamber is a precision, volumetric flask made of conductive plastic which
 provides the ion collection chamber for electret-ion-chamber measurement of
 radon.   When paired with the electret, these components comprise a working
 E-PERM radon measurement unit...Each E-PERM Electret is an electrically
 charged wafer of Teflon (Dupont) which has been treated to hold a stable
 electrostatic potential."(2)

-------
                           SCREENING RESULTS


     Screening results for the twenty-two schools  tested in  New York State
during the 91-92 school year are shown in Table  2.

     In twelve schools, which each had more than 20% of the  rooms tested with
greater than or equal to 4 pCi/1,  the percentage of rooms with greater than
or equal to 4 pCi/1 was about the same for the basement (46%)  and ground level
(49%).  Although only a few rooms were measured  on floors above the ground
level, 35% of the rooms tested on the second  floor in these  twelve schools
were measured with greater than or equal  to 4 pCi/1 indoor radon.   These
results indicate that for schools with above  average indoor  radon the
percentage of rooms with greater than or  equal to  4 pCi/1, is  about the same
for basement and grade levels.  The high  percent of rooms with concentrations
greater than or equal to 4 pCi/1 for above grade levels could  have resulted
from selecting rooms which have high air  mixing  with lower levels (See Table
3).  The average basement, first floor (grade level) and second floor radon
concentrations for all twenty-two schools are 7.1, 3.4, and  2.6 pCi/1
respectively.  Figure 3 represents the percentages of rooms  with radon levels
greater than or equal to 4 pCi/1 by level.  The  percentages  by floor are as
follows; basement 42%, first floor 39% and second floor 53%.  Testing
conducted on the second floor represents  only a  sampling of  the total rooms
on that floor.

     Measurements conducted in all twenty-two schools indicate that 345 of 1116
rooms had levels greater than or equal to 4 and less than 20 pCi/1 of indoor
radon and 11 had levels greater than or equal to 20 pCi/1 (See Table 2).  In
many cases the rooms with greater than or equal  to 20 pCi/1  were storage areas
or basement rooms with little ventilation and low occupancy.  We observed a
similar trend in residential indoor concentrations in areas with gravelly soils
in New York State, that is, a high percentage of homes with radon levels near
20 pCi/1 but few homes with greater than  40 pCi/1.

     Figure 2 compares radon concentrations by floor for all twenty-two
schools.  Seven of twenty-two schools did not have basement readings.
Therefore the total number of schools for first floor readings in each
category is higher.  The graph also demonstrates the decrease  in the number
with radon levels at higher concentrations.

     Due to the high percentage of rooms  with concentrations greater than
or equal to 4 pCi/1 on second and third floors,  we propose to  revise the
measurement protocol for Year 2 and will  place alpha track detectors in all
rooms on these floors.

-------
                          SOIL-RADON POTENTIAL


     Soil characteristics relating to the potential for radon infiltration
into buildings will be measured at about one-third of the schools
participating in the survey.  Soil measurements include surficial gamma,
radionuclide concentrations (Ra-226, U-238, Th-232), soil-gas radon
concentrations, and the permeability of the soils for gas flow (3).
Measurements are made at various  locations around the perimeter of the
schools and at various distances  from the school.  When accessible,
measurements are also made in the crawlspaces of the schools.  It is of
interest to determine if the same soil parameters that cause above average
indoor radon in single-family homes have the same effect on larger buildings.
This information in combination with building design will aid in further
identification of schools with high radon potential.

     Initial visits and soil measurements have been made at three schools.
During the summer and fall of 1992 additional field trips will be made to
complete the soil characterization at these schools.  Although incomplete, the
results from the initial visits are presented in Table 4.

     The soil Ra-226 concentrations listed in table 4 are the averages for 3
to 6 samples collected at a depth of 60 cm from under and around each school.
The mean Ra-226 concentration for soils in the U.S is 1.0 pCi/g (4).
Therefore the soils under and around school ONO-1 in Onondaga County are near
the mean, whereas the soils at COL-1 in Columbia County and WAS-1 in Washington
County have below average concentrations of Ra-226.

     Soil-gas samples were collected at a distance of 0.5 m from the schools
foundation, from crawlspaces and  from locations more than 10 m from the
school.  Since air being drawn through the soil near the foundation and
possibly through the soil in the  crawlspaces can significantly dilute the
soil-gas Rn-222 concentration, only samples collected at locations more than
10 m from the school were used to determine the soil-gas Rn-222 concentration.
Most of the measurements used were from samples collected at a depth of 120
cm, however, some of the measurements used were from depths between 45 to
90 cm,  due to a high water table  or rocks preventing deeper sampling.  The
values listed in Table 4 represent soil gas measurements from 5
locations at the school  in Onondaga County and only one location for the
schools in Columbia and Washington Counties.   Therefore the values listed in
Table 4 cannot be considered as very representative for the soils around the
schools.   Even so,  the values listed for the three schools are typical for
soils with average to somewhat below average Ra-226.

-------
     Since rubble and fill are often found near a school's foundation and in
the crawl spaces only permeability measurements taken more than 10 m from the
school are listed in table 4.  Although the values listed in table 2 are
averages for only a few measurements they are useful for an approximate
characterization of the permeability for the soils around the schools.   School
ONO-1 in Onondaga County is located in outwash gravel  along a creek.
Measurements were made at five locations more than 10 m from the school with
an average permeability for gas flow of 2.9 x 10-6 cm .   The soils are quite
gravelly and at one location a high water content in the soil reduced the
permeability by three orders of magnitude.  School COL-1 in Columbia County is
located about 400 m from a creek on a small plateau.  The permeability
measured at greater than 10 m from the school was 1.8 x 10-7 cm .   The soils
contain some stone but are predominantly silty sand and appeared to be well
drained.  At school WAS-1 in Washington County the permeability of the soils
for gas flow was measured at 2 x 10-9 cm .  The soils were silty with some clay
and were nearly saturated with water at depths greater than one meter.

     The most important soil characteristic correlating with indoor radon
concentrations at the three schools studied appears to be the permeability of
the soils for gas flow.   At School  ONO-1 in Onondaga County where 20% of the
rooms measured had Rn-222 concentrations greater than or equal to 4 pCi/1, the
soils are gravelly and quite permeable.   At School COL-1 in Columbia County
where 6% of the rooms measured had Rn-222 concentrations greater than or equal
to 4 pCi/1 the permeability was moderate and the soils were well drained.  At
School WAS-1 in Washington County the permeability of the soils was low, the
water table was high and only 1% of the rooms were greater than or equal to
4 pCi/1.

     A further indication that highly permeable gravelly soils correlate with
above average indoor radon in New York State schools is obtained by
considering the location of the schools with the highest indoor levels.  The
four schools with the greatest percentage of rooms with greater than or equal
to 4 pCi/1 are located in Tioga, Steuben and Cortland Counties (Table 2).
These three Counties are among the seven Counties with the greatest average
residential indoor Rn-222 concentrations in New York State.  These three
Counties are located in the south central region of the State.  The most
recent glacier formed a large morain just north of these Counties (Valley
Heads Morain) and when the glaciers melted large quantities of water flowed
down through the valleys in these Counties carrying debris from the morains
into the valleys, forming deep gravel deposits in many areas.  These highly
permeable gravel deposits are the predominant factor resulting in above
average residential indoor radon in this region of the State (5).   It appears
that the gravelly soils are also resulting in above average Rn-222 in the
schools in this region.   One of the schools in Steuben County has been visited
and although soil measurements have not yet been made, the soil under and
around the school was observed to be very gravelly.

-------
                             CONCLUSIONS
     The first year of our project has been quite successful.   We have
targeted high risk areas of New York State based on our residential  data and
have found schools in those areas to also have elevated levels.   In  some cases
we have found that high concentrations exist at more levels in a school  and
therefore we will be implementing more detector distribution at these levels
in Year 2.

     The distribution by mail has worked quite well and we will  continue it
into Year 2 with some slight modifications in instructional materials for ease
of reading.  Additionally, we will be discussing the floor plans in  greater
detail in Year 2 to prevent duplication of our efforts.

     Since EPERM and alpha track measurements (charcoal if deemed necessary)
and continuous radon monitoring units are being utilized for this study we
cannot overemphasize the need for a strong quality assurance program as an
integral part of this type of study.

     In the Fall of 1992, we will be receiving results of our year long alpha
track measurements and additional soil characterizations, which will complete
our assessment of these twenty-two schools.

-------
                           REFERENCES
1. "Radon Measurements in Schools, An Interim Report."  EPA-520/1-89-010,
   U.S. Environmental Protection Agency, Washington,  D.C., March 1989. p A-6.

2. Rad Elec Inc.," E-PERM Components Electret and Ion Chamber."

3. Kunz, C., "Influence of Surficial Soil and Bedrock on Indoor Radon in NYS
   Homes.", NYSERDA Report 89-14, October 1989.

4. Myrick, I.E., Bervin, B.A.  and Haywood, F.F., "Determination of
   Concentrations of Selected Radionuclides in Surface Soil in the U.S.",
   Health Physics, Vol. 45, No.  3, pp 631-642 (1983).

5. Kunz, C., Laymon, C. and Parker, C.,  "Gravelly Soils and Indoor Radon",
   Proceedings EPA 1988 Symposium on Radon and Radon Reduction Technology,
   Denver, CO, Oct 1988.

-------
    AVERAGE RADON
           (Basement
BY COUNTY
Only)
                   A*orifajr6. 1001
               J  O.I   19
                 -40   79
    39
   15 «*
FIGURE 1.     NUMBER OF SCHOOLS PARTICIPATING IN YEAR ONE

-------
    3
    Ł
    cc
    a
14-
* *>
12-
4 4%
10-
-
-
-
•
^




.





1




1 — n — 1
1
1
1

1







I
\
1 — i i r'


1

j



^
I
I
i
I
1

-------
      60
      50
    UJ
    g  30

    UJ
    0.


       20
       10
                   1	r
              BASEMENT
  FIRST

FLOOR
SECOND
FIGURE 3.    PERCENTAGE OF ROOMS > 4pd/l

-------
TABLE 1.
FOLLOW-UP MEASUREMENTS
SCREENING LEVEL


< 4 pCi/1


> 4 < 20 pCi/1


> 20 pCi/1
                               FOLLOW-UP MEASUREMENT PERIOD
                                    Three month EPERM
                                    One month EPERM
* These follow-up measurements are made in
  addition to one year alpha track detectors

-------
TABLE 2.   RADON LEVELS IN SCHOOLS - SCREENING RESULTS
SCHOOL/   AVG RN        # OF ROOMS  # OF ROOMS    # OF ROOMS  % OF ROOMS
COUNTY    CONC (pCi/1)    TESTED   .  > 4 < 20       ^20         > 4
AL?-1              5.6     90           38              4          46 7
ALBANY
BRO-1              4.5     54           22              0          40.7
BROOME
CHE-1              5.0     36           14               1          41.7
CHENANGO
CHE-2              6.4     44           16               1          38.6
CHENANGO
CHE-3              2.6     37            7              0          18.9
CHENANGO
COL-1              2.3     33            2              0           6.1
COLUMBIA
COR-1              5.8     51           30              1          60.8
CORTLAND
DEL-1              2.0     51            7              0          13.7
DELAWARE
DUT-1              3.6     29            8              0          27.6
DUTCHESS
DUT-2
DUTCHESS           2.2     53            5              0           9.4

-------
TABLE 2 continued

SCHOOL/     AVG RN      # OF ROOMS   #  OF ROOMS        #  OF  ROOMS   % OF ROOMS
COUNTY      CONC (pCi/1)   TESTED   > 4 <  20           > 20         ^4
LEW-1
LEWIS
LIV-1
LIVINGSTON
ONO-1
ONONDAGA
ONO-2
ONONDAGA
REN-1
RENSSELAER
SCH-1
SCHOHARIE
STE-1
STEUBEN
TIO-1
TIOGA
TIO-2
TIOGA
ULS-1
ULSTER
WAS-1
WASHINGTON
1.3 34 0 0 0.0
3.7 69 32 0 46.4
2.4 56 11 0 19.6
1.4 56 30 5.4
3.5 48 20 0 41.7
1.0 54 1 0 1.9
8.1 52 37 1 73.1
6.2 76 54 1 72.4
6.1 49 26 2 57.1
3.7 29 8 0 27.6
1.2 84 1 0 1.2

-------
TABLE 2 continued
SCHOOL/
COUNTY
WAY-1
WAYNE
ALL
SCHOOLS
AVG RN
CONC (pCi/1)
1.4

AVG RN
CONC (pCi/1)
# OF ROOMS
TESTED
31

# OF ROOMS
TESTED
# OF ROOMS # OF ROOMS °A
> 4 < 20 > 20
3 0

# OF ROOMS # OF ROOMS
> 4 < 20 > 20
', OF ROOMS
> 4
9.7

% OF ROOMS
> 4
22
3.7
1116
345
11
32

-------
TABLE 3.    SCREENING RESULTS
LEVEL                  # OF SCHOOLS WITH % OF ROOMS > 4 pCi/1

               OVER 20%          OVER 40%      OVER 60%
0
1
2
8
12
6
6
9
4
3
4
4
 FOR  THE  12  SCHOOLS WITH OVER 20% OF ROOMS ^ 4 pCi/1, THE  FOLLOWING
                        IS TRUE;

         LEVEL                   % OF ROOMS > 4 pCi/1
          0                             46

          1                             49

          2                             35

-------
TABLE 4.   SOIL RADON POTENTIAL DATA
SCHOOL       SOIL       SOIL-GAS
(COUNTY)     RA-226     RN-222
             (pCi/g)    (pCi/1)
                    PERMEABILITY
                          2
                       (cm )
                    ROOMS
                    >  4  pCi/1
ONO-1       0.96
ONONDAGA
COL-1       0.68
COLUMBIA
WAS-1       0.73
WASHINGTON
569
589
580
                                                     -6
2.9 x 10
                                                      -6
0.18 x 10
                                                       -6
0.002 x 10
20

-------
         Session VI Posters
Transport and Entry Dynamics of Radon

-------
                                                                      VIP-1
SIMPI.TFTED MODELING OF THE EFFECT OF GUFPLY VENTILATION ON INDOOK
                      -KADON                                      "
                    by   David S
                         Infiltec
                         Falls Church, VA 22041

                         Mark Modei-a
                         Lawrence Berkeley Laboratory
                         Berkeley, CA

                         Kelly Leovir.
                         US EPA ORD
                         RTF, NC 27711
                             ABSTRACT


     This paper  investigates  a simplified model of the effect of
supply  ventilation on  indoor radon  concenhrat ions.    The model
assumes  a  slab-on-gradc  structure,  a single  well-mixed  zone,  a
ventilation  fan  inducing an  air flow into  the building, winter
ctack effect weather conditions/  and a constant, radon concentration
under the slab.  The radon mitigation effect  of supply ventilation
is separated into a dilution  effert  and  a change in radon entry
rate due to prcoaurization of the buildinq  shell.   The modftl is
used  to assess  f.he  radon  mitigation  impact  of  increasing  the
ventilation  rale  to meet  ASHKAti  standard  62-1989  recommended
ventilation  levelc.   This  model was  developed to  aid  in  t-.he
understanding  the results of  1991  EPA reecarch  in  Virginia and
Maryland schools into the use of HVAC  system modifications as a
radon mitigation technique.

-------
                                                                   VIP-2
           DETERMINATION OF MINIMUM COVER THICKNESS FOR
              URANIUM MILL TAILINGS DISPOSAL CELLS

               by:  D.W. Andrews and J.P. Ambrose
                    CWM Federal Environmental Services, Inc.
                    Albuquerque, NM  87106
                             ABSTRACT

     The Uranium Mill Tailings Remedial Action  (UMTRA) Project was
initiated  by the  UMTRC Act  passed  by congress  in 1978  and is
expected to extend into the  late 1990's.  UMTRA is  involved  in mill
tailings remediation activities throughout the continental United
States.  The remediation project involves  the stabilization of the
tailings material into massive  disposal  cells.   These disposal
cells have been designed to endure environmental forces for 1,000
years.  Once all the tailings are deposited into the disposal cell
a  radon barrier  is applied.   Prior  to  placement of  the radon
barrier cover on the disposal cell, soil samples are collected at
0.6 meters intervals to a depth of 6.1 meters.  These samples are
used  to  determine  the  radon  emanation  fraction,  Radium-226
concentration,  Thorium-230  concentration, and  moisture content.
These measurements along with other field  measurements are  used in
the  RAECOM  radon  flux  computer modeling  program to  determine
minimum cell cover thickness.  Additional flux measurements are
performed to verify compliance with Subpart T, NESHAP regulations.

                          INTRODUCTION
     Public Law 95-604 known as the Uranium Mill Tailings Radiation
Control Act  (UMTRCA)  was  passed  by congress in November of 1978.
The Act required the remediation  of residual radioactive materials
at some former uranium mill tailings processing sites to aleviate
a  potential  long  term  health hazard.    The  UMTRCA  charged  the
Secretary of Energy  with  the  remediation of the designated sites
and the Administrator of the Environmental Protection  Agency to
establish  the  standards  for  clean-up.   The Nuclear Regulatory
Commission is required to approve the long-term surveillance and
monitoring plans and certify the disposal sites.

-------
     The  Department  of  Energy was  to  enter  into cooperative
agreements with the  states  and  indian tribes to  acquire the lands
and  perform  the  remedial   actions.    Remedial  actions  involve
demolition  and decommissioning  of existing mill  facilities and
consolidation  of  all associated  residual radioactive materials,
including windblown  materials,  into a designed disposal cell.

     The  Environmental  Protection Agency  established the design
standards for  tailings  disposal cells.  The disposal  cell must be
designed  in  a manner which  prevents  the  background radon levels
from increasing by more than 0.5 pCi/1 or limit the average cell
radon flux to  20 pCi/m2-s.   The  disposal cell shall  be designed to
maintain the  above standards for  1000 years and in no case less
than 200 years.  Demonstration of  compliance is  not required.

     To achieve the  long-term disposal requirements the cells are
constructed of only  natural materials.   The  radon barrier cover
layers are usually constructed of high clay content  soils to limit
radon diffusion and  to  inhibit  infiltration of moisture into the
tailings material.   Inhibiting the infiltration  of water provides
some  frost  protection  of the  tailings  embankment  and  prevents
excessive  leaching  and groundwater  contamination.   The  radon
barrier is then covered with high permeable sand layers to allow
lateral flow of water.   The side slopes  are covered by layers of
rock with low fracture potential to armor the cell against erosion.
Vegetation  is  often  used  on top  slopes  for  erosion protection
although root development can produce  conduits  for radon transport
to the surface.

     The thickness of the radon barrier is not only determined by
the radon diffusion characteristics of the cover  material but also
by other design criteria such as frost and erosion protection.  The
combination of the  above design criteria  is used  to achieve the
1,000 year  disposal  requirement.   The  radon  barrier  thickness
determination  methods presented below are used  to  determine the
minimum thickness allowed for controlling  radon emissions into the
ambient air.

            METHOD FOR  COVER THICKNESS DETERMINATION


     In achieving  the design standard prescribed  in the  federal
regulations for uranium  disposal cells, the UMTRA Project utilizes
a  radon  flux  computer  modeling  program  called RAECOM.    Given
specific soil parameters and 226Ra concentrations  for each layer of
material  associated   with  the  disposal   embankment,  the  RAECOM
program calculates radon flux using a one-dimensional steady-state
diffusion equation (1).

-------
PARAMETERS

     The  radon  barrier  cover  thickness design  for a  long term
disposal  cell  is  dependent on  soil  density,  specific gravity,
moisture,  thickness,  M6Ra  concentration,   222Rn  diffusion,  and
emanation  coefficients  (2).    For  calculation  purposes  these
parameters  can  be measured  directly   at  the  specific  site  or
reference values can  be  utilized.   If  reference values are used,
the NRC  requires justification and demonstration of conservatism
for their use at  a specific site.  Generally,  the UMTRA Project
uses only measured parameters  in the flux calculations.

     The parameters used in the RAECOM computer program are soil
layer  thickness, porosity, moisture content, 226Ra  concentration,
density, emanation coefficient, diffusion coefficient,  and ambient
air radon concentration.  The modeling  system  allows  the use of
bottom flux which  is  the radon flux entering the lowest layer of
the model.  The  bottom flux is usually ignored since twenty feet of
tailings  is  considered  an infinite  thickness  for  calculation
purposes  (1).

     In  calculating cover  thicknesses,  easily measured parameters
such as soil moisture content,  soil density,  soil  thickness, radon
emanation coefficient,  and 226Ra concentration are obtained  as the
embankment is constructed.   These  parameters are  determined using
field or on-site laboratory measurements. Diffusion coefficients,
which  are more difficult to measure, are obtained by sending soil
samples  to  geotechnical  laboratories  for analysis.  Ambient air
radon concentration is measured in  the vicinity around the disposal
site.

     As  the tailings  embankment and radon barrier cover are being
constructed, soil density measurements  using the standard sand cone
method1  or  the nuclear density  gauge  method2 are performed to
demonstrate the quality  of embankment construction.   These density
measurements are also used for the RAECOM calculations.

     Porosity is calculated using soil  density and specific gravity
measurement  information.    The  following  equation  is  used to
calculate porosity:

          P - d, /  (s, *  dj

          where:          P  =  Porosity
                          d, = Density of Soil

'ASTM Standard Test Method D-1556, Density and Unit Weight of  Soil
  in Place by  the Sand-Cone Method (3).
2ASTM Standard Test Method  D-2922, Density of Soil  and  Soil-
  Aggregate  in Place by Nuclear Methods (Shallow Depth) (3).

-------
                         s, =  Specific  Gravity  of  Soil
                         dw = Density of Water

     Percent moisture content of tailings material is measured in
the field using the oven-dry method1.  Percent moisture content for
cover materials are calculated using a long-term moisture content
equation  or  measuring the  moisture  equilibrium  potential  for a
particular soil using the 15 bar moisture method2.

     Radon emanation  coefficients for the  tailings  material are
obtained at the on-site laboratory by analyzing samples collected
at 0.6 meter  intervals  in the disposal  cell  soil column.   These
samples are sealed,  stored to achieve equilibrium,  analyzed, opened
and  reanalyzed after  a  venting  period.   The  results of  this
analysis  are  used  to  calculate the emanation  fraction.  The 226Ra
concentration of the tailings material samples is obtained from the
first analysis prior to venting.

     As  disposal  embankments  are constructed soil  samples are
collected at  0.6 meter  interval  elevations from established grid
locations. Soil sample collection is initiated at an  elevation 5.5
meters below the final design elevation of  the  top of the tailings
embankment.   Sample material  is consolidated over a  0.6  meter
depth,  homogenized and  then  split into  two  samples.   Sample
collection continues as new layers  of  tailings are placed in the
embankment until the final elevation is  attained  at  each location
which forms a sample matrix of the top  6.1  meters  of  the completed
disposal cell.  If portions of the tailings embankment exist when
sample collection  is  initiated borehole samples are collected at
required elevations below the existing surface.

     After the collection process, soil sample cans are  sealed and
leak tested as soon as possible to minimize soil moisture loss and
to prevent radon gas leakage.  Of the two samples collected at each
matrix grid point,  one is  analyzed for  230Th and the other is stored
for at least  20 days to allow 222Rn  to achieve equilibrium with 226Ra.
After  the  storage  period  the  sample  is   analyzed  by  gamma
spectrometry  for 2I4Bi activity.   The sample is then opened, oven
dried and allowed to vent for 4 to  24 hours.   Bismuth-214 was
chosen for analysis since it produces  an easily measurable gamma
peak and reaches equilibrium with 222Rn  in a relatively short time.
'ASTM Standard Test Method D-2216, Laboratory Determination of Water
 (Moisture) Content of Soil and Rock  (3).
2Two capillary moisture methods are used depending on soil particle
 size:  ASTM Standard Test Method D-2325, Capillary-Moisture
 Relationships for Coarse- and Medium-Textured Soils by Porous-
 Plate Apparatus or ASTM  Standard Test Method D-3152, Capillary-
 Moisture Relationships for Fine-Textured Soils by  Pressure-
 Membrane Apparatus  (3).

-------
The  sample  is then  resealed  and immediately reanalyzed  for 2l4Bi
activity.  The emanation coefficient  is then calculated using  the
following formula:

          EC = (RA, - RA2) / RA,

          where:        RA,  = Wet equilibrated 214Bi activity
                        RA2  = Dry unequilibrated 214Bi  activity
                        EC  = Emanation coefficient

Emanation Coefficients for the radon barrier material  are measured
at an off -site laboratory using  the same method described  above.
     The ^^a concentrations of tailings material are  obtained  by
dividing the RA, activity from the above equation by the dry weight
of the sample.  Radon  barrier  226Ra concentrations are  obtained  at
an off-site  laboratory when samples are analyzed to determine the
emanation coefficients.

     The 226Ra concentrations at a 1,000 years  must be  estimated
before the RAECOM calculations are performed.   The existing 230Th
concentration is  used  in the above estimated 226Ra  concentration.
The  1,000  year  226Ra concentration  is determined  by  using the
following equation:
                      * 0.65)  + (TH0 * 0.35)

          where:    RA10oo =     Estimated   226Ra   Concentration   in
                               1,000 years
                    RAo    =   Initial 226Ra Concentration
                    TH0    =   Initial 230Th Concentration

The  equation  given above  is  a  reduced  version  of  a  secular
equilibrium equation used to decay parent /progeny combinations (8) .

     Diffusion coefficients are obtained at an  off-site laboratory
by  conducting transient  diffusion measurements  (9) .    Radon  is
allowed  to diffuse  though a compacted  soil  column.   Activity
measurements  are  performed at the opposite  end  of the column  at
different time intervals until the they reach an equilibrium point.

OPTIMIZATION

The radon barrier thickness is optimized by utilizing  the RAECOM
radon  flux   program   and  parameters  representing  worst   case
conditions.   Optimization  is carried  out by varying the  radon
barrier thickness to achieve  a resultant radon flux of just less
than 20  pCi/m2-s.   The optimized radon  barrier  thickness is the
minimum thickness allowed in  the disposal  cell design.

-------
                     RADON FLUX MEASUREMENTS
     Subpart T of the National Emission Standards for Hazardous Air
Pollutants  (NESHAPs)1,  requires owners  of uranium mill  tailings
piles to demonstrate compliance with radon  flux emission standards.
Under these standards the UMTRA Project is required to dispose of
the tailings  such  that the measured  flux on  the surface of the
tailings  embankment does not  exceed  20  pCi/m2-s.    The   flux
measurements  are  performed under  the prescribed conditions and
according to the regulatory method.1

     Disposal embankment design  is  based on a worst case situation
over a thousand year period.  The radon flux measurement cannot be
used to demonstrate the effectiveness of  the  1,000  year  disposal
embankment design.   However,  the  radon  flux  measurement can be
utilized as an indicator of design  effectiveness.
           0.5
         Ifl
         Ł
         \
         r-(
        (J
         a
           0.4-
           0.3
0.2-
           o. i-
              0    10   20   30  40  50   60   70  80  90  100
                         LOCATION  NUMBER
                                 AUERAGE
    Figure 1.   Radon Flux Measurements for  Lowman,  Idaho Site
'The NESHAPs are part of the requirements of the Clean Air Act.

-------
     Figure 1 illustrates the results  of  radon flux measurements
from the completed Lowman, Idaho UMTRA  disposal cell.  The highest
flux measurement was  just above  0.4 pCi/m2-s  and  the average was
approximately 0.05 pCi/m2-s.  Flux measurements are typically well
below the  standard because the cell  is designed for  worst case
conditions.  Moreover, the radon barrier  may  be thicker than the
optimized thickness due to other design requirements  such as frost
and erosion protection.  This was the case for the Lowman disposal
cell.  The Lowman worst case situation  produced an optimized radon
barrier thickness of only 13 cm, whereas the actual minimum design
thickness was 46 cm (10).

     The work described  in this paper  was not funded by the U.S.
Environmental Protection Agency and therefore the contents do not
necessarily  reflect  the views  of the agency  and no official
endorsement should be  inferred.

-------
                            REFERENCES


1.   Rogers, V.C.,  Nielson,  K.K., Kalkwarf, D.R. Radon Attenuation
     Handbook  for  Uranium Mill Tailing Cover  Design.   NUREG/CR-
     3533.  U.S. Nuclear Regulatory Commission, Washington, D.C.,
     1984.

2.   U.S.  Department  of  Energy,  Technical   Approach  Document,
     Uranium Mill Tailings Remedial Action Project, Albuquerque, NM
     87108, May 1986.

3.   American Society for Testing and Materials, 1992 Annual Book
     of  ASTM  Standards,  Section  4,   Construction,  Volume 04.08,
     Philadelphia,  Pennsylvania, 1992.

4.   Turner, J.E.,   Atoms, Radiation,  and  Radiation Protection.
     Pergammon Press Inc., Maxwell House,  Fairview Park, Elmsford,
     New York,  1986.  p. 65.

5.   Nielson, K.K., Rich, D.C., Rogers, V.C.,  and Kalkwarf, D.R.,
     Comparison  of  Radon  Diffusion  Coefficient   Measured  by
     Transient  Diffusion  and  Steady-State  Laboratory  Methods.
     NUREG/CR-2875.      U.S.    Nuclear   Regulatory   Commission,
     Washington, D.C.,  1982.

6.   Lowman Idaho,  Final  Design for  Construction  -  Calculations,
     Prepared  by  Morrison  Knudsen  Corporation,   Environmental
     Services Group, for the  Department of Energy,  Uranium Mill
     Tailing Remedial  Action  Project,  Albuquerque, New  Mexico,
     1991.

-------
                                                                   VIP-3
     A MATHEMATICAL MODEL DESCRIBING RADON ENTRY AIDED BY AN

        EASY PATH OF MIGRATION ALONG UNDERGROUND TUNNELS
     Mosley, R. B.
     USEPA
     Air and Energy Engineering Research Laboratory
     Research Triangle Park, NC  27711
                             ABSTRACT
     Most  houses are connected  to the soil  through underground
tunnels of higher permeability materials surrounding such things as
sewer lines, drain lines, water lines, electrical entrances, etc.
In some cases, these buried lines may provide effective migration
paths  for radon to  approach or  enter the  houses.   These easy
transport paths are likely to be most important when the house is
surrounded by relatively low permeability  soil such  as dense clay.
A model  is presented to evaluate the  enhanced  migration  rate of
radon as a result of these transport tunnels.

-------
                                                                    VIP-4
             RADON  DIFFUSION STUDIES  IN  SOIL AND WATER
             By: Manwinder  Singh
                Surinder Singh
                H.  S. Virk
                Department of Physics
                Guru Nanak Dev University
                Amritsar 143  005,  India
                             ABSTRACT


     In soil, the contribution comes by way of diffusion and
transport of radon and thoron from a distant source and radium and
thorium present in the immediate vicinity.  In the case of natural
waters, radon is contributed by migration and dissolved radium^
content in the water.  In the current investigation, a simulation
study has been carried out in the laboratory on the behavior of
radon diffusion through soil and water.  The diffusion rate of
radon through soil has been studied by measuring the rate of
buildup with time at varying thicknesses and porosities of the
soil.  The diffusion coefficient, D, and mean diffusion lengths
range from 0.0027-0.0072 cm2/sec and 65-82 cm for porosities of 30
to 50 percent, respectively.  The radon flux follows the
exponential fall with the increase in overburden on the source.
The experimentally-observed values at different thicknesses are
lower than the theoretically-calculated values.  The radon
activity may be suppressed partly due to the diffusion leakage and
partly because of the back diffusion.

-------
                                                                    VIP-5
               Stack Effect and Radon Infiltration

                        by: Craig DeWitt
             Agricultural and Biological Engineering
                        Clemson University
                     Clemson,  SC  29634-0357
                          (803) 656-4041
                       (803) 656-0338  (fax)
                            ABSTRACT


     Infiltration of outside air as a result of pressure
gradients caused by thermal buoyancy,  or the "stack effect",
plays a factor in bringing radon gas into a home.  In some
instances, reducing infiltration can actually increase the radon
levels or potential radon problems in a house.  By taking
advantage of the principles behind the stack effect, reducing
infiltration can be an effective radon mitigation technique and
significantly reduce the forces drawing radon into a house.

     Sealing openings in the ceiling or upper locations of a
building will lower the neutral pressure level within a building.
Lower neutral pressure levels result in reducing infiltration
rates as well as lower pressure difference across infiltration
openings.  Depressurization fields of SSD and ASD systems will
also increase in strength.

     This paper has been reviewed in accordance with the U.S.
Environmental Protection Agency's peer and administrative review
policies and approved for presentation and publication.

                           INTRODUCTION


     Radon emanation from naturally occurring soils into the soil
air, which can subsequently be transported into a building is a
source of indoor air pollution.  Two conditions must exist for
radon-laden soil-air entry into a building: 1) there must be
openings in the building which couple the soil-air with indoor
air and 2) there must be a driving force that results in a flow
of the radon-laden soil-air.  Major strategies for reducing or
preventing high levels of radon in structures include 1) dilution
through increased ventilation, 2) sealing entry pathways into the
structure, 3) building pressurization, 4) soil depressurization,
and 5) sub-slab or sub-membrane depressurization  (1).

-------
     The dilution through ventilation theory is based on a simple
model; if the radon entry rate Qrn, is assumed to be constant and
set equal to the removal rate, we have;
               Qrn = V  * Crn

where V is the ventilation rate and Crn  is the radon
concentration.  Based on this relationship, any decrease in radon
concentration must involve a proportional increase in ventilation
rate.  To halve the radon concentration would require doubling
the ventilation rate. Problems encountered with this strategy
include 1) knowing the existing ventilation rate, and 2) doubling
or quadrupling the ventilation rate is about the maximum
acceptable level in terms of energy use and human comfort.

     The other radon reduction strategies listed above attempt to
reduce the radon entry rate.  This radon entry rate is a function
of the source strength, leakage area, and pressure difference
across the leakage area as the driving force, as related by:


               Qrn=f(Sm,  ELA,  AP)                               (2)

where Spn  is the source strength,  ELA  is the equivalent  leakage
area, and AP is the pressure difference.  Since changing the
source strength is not usually a  viable option, the only feasible
options are to attempt to alter leakage areas or the pressure
differences.  Soil, sub-slab, sub-membrane depressurization and
house pressurization strategies attempt to reduce  (or even
reverse) the pressure  differences across the leakage area or
radon entry sites.

     Crack sealing as  a mitigation/prevention strategy attempts
to reduce the leakage  area of the radon entry sites.  This
technique has had inconsistent results probably due to the
inability to effectively locate and seal all entry points.  It
has been observed that even small openings are sufficient to
allow unacceptable radon levels  (2) .  Crack sealing has become a
part of other mitigation strategies though, by helping to extend
the pressure fields developed in  the  soil by sealing re-
entrainment paths and  preventing  "short-circuiting" through slabs
and other building shell membranes  (3) .

     Both soil depressurization and sealing strategies attempt to
reduce the inflow of radon through manipulation of leakage areas
(cracks) and pressure  differences.  These strategies though, do
not make the best use  of the physics  behind the relationship
shown in equation 2.   This paper  will discuss the physics
involved and resultant, more efficient  mitigation techniques.

-------
                         THE  STACK EFFECT
FLOW EQUATIONS
From Bernoulli!'s equation, we can get a theoretical expression
for mass flow rate of a fluid through an opening:
                      M' =  p  * A * J2 * AP/p                    (3)
where M'  is the mass flow rate, p is the density, A is the
leakage area, AP is the pressure difference across the opening.
For complex structures, the dependency of the pressure difference
is more complicated (2).  Therefore an empirical power law
function is used:
                     M' = K  * v/p *v/5" *A * AP"
                                      e
                        = K * A0 * APn                           (4)
The exponent of the flow equation, n, varies theoretically from
0.5 for fully developed turbulent flow to 1.0 for laminar flow
(4).  A  represents a free area equivalent to the real crack.  K
is an airflow coefficient, defined as the flow rate at a pressure
difference of 1.

Leakage area

     Finding and sealing cracks in unfinished basement slabs may
be straight forward and relatively easy.  Locating cracks in
slabs of finished basements, slab-on-grade foundations or in
floors over crawl space foundations can be next to impossible due
to finish floor materials, wall partitions, plumbing  fixtures and
appliances.  As such, effectively eliminating or reducing leakage
area may be impossible.  As well, new cracks may develop with
time.

-------
Pressure differences

     The pressure differences responsible for flow of soil-air
into a building arise from 1) wind pressures on the building, 2)
forced ventilation of the building from kitchen and bath-type
exhaust fans, 3) chimneys and flues from combustion appliances
such as furnaces and water heaters, 4) fireplaces and woodstoves,
and 5) air leakage due to the buoyancy, or the stack effect.

     As wind hits a building, pressures develop as kinetic energy
is reduced.  Pressures on the windward side increase while
pressures on the downwind side decrease.  The interior of the
building will reach a pressure somewhere in between, depending
upon the location of openings.  Openings concentrated on the
windward side will tend to pressurize the building, while
openings on the downwind side will tend to depressurize the
building.

     Mechanical exhausts and chimney stacks exhaust air from the
building lowering the pressure within the building.

STACK EFFECT

     Another physical phenomena that influences the pressure
differences in and around a building is the effect of buoyancy,
or stack effect.  This is due to the density differences between
inside and outside air or between  two zones within a building,
such as downstairs and upstairs.

     Warm air is less dense  (lighter) than cold air.  This warmer
air will tend to "float" on the colder air.  The result is a
pressure difference between the inside air and the outside air,
and a flow through any available opening.  If openings exist at
different elevations within the building, warmer  (less dense) air
will  force its way out the upper openings to be replaced by
cooler  (more dense) air through the lower openings.

      The density of air is mainly  a function of temperature  and
moisture content.  As temperature  increases, density decreases.
Similarly, as the moisture content increases, density decreases.
Many  researchers and lay people relate the density differences  to
just  temperature.  For a typical winter condition with outside
air at  0°F, 50% relative humidity  (RH), and inside air at 70°F
and 50% RH, the change in temperature  accounts for about 90% of
the density difference while  the change in moisture content
accounts for about 10%.  At high temperatures, moisture content
can result in a significant density change.  For  example,  inside
air at  75°F/60% RH is less dense than outside air at 80°F/20% RH.
At higher temperatures, the  effects of moisture content are  more
pronounced.

-------
     The pressure difference between across an opening in a
building is proportional to the distance of the opening from a
neutral pressure level (NPL) and the density difference between
inside air and outside air. The pressure difference across an
opening due to the stack effect is:
             = ( P0-P,-) * g *  (H-h) / 2
(5)
where p is density, o indicates outside, i indicates inside, g is
gravitational force and h is the distance from the NPL, H.

THE NPL

     The NPL is the elevation at which the pressure inside the
building is the same as outside the building.  Above this NPL,
inside air is at a higher pressure than outside; air will tend to
leak (or be forced) out of the building.  Below the NPL, inside
air is at a lower pressure than outside; outside air will be
drawn into the house.  The further the opening from the NPL, the
larger the force pushing or pulling the air  (see Figure 1).   The
larger the temperature difference, the larger the force pushing
or pulling the air.
                     Neutral  Pressure Level
Figure 1. The NPL in a typical building with uniformly
distributed openings.  The length of the arrow indicates the
pressure difference between inside and outside.  The direction of
the arrows indicate airflow.

-------
NPL Location

     The location of the neutral pressure level is a function of
the location of the openings and their distribution throughout
the building.  The NPL is derived from a solution to the
continuity equation of volume (4):
Ł/„
                                H-h
where H is the elevation of the NPL, h is the elevation of
opening j, A is the area of opening j.


     For a simple enclosure with two openings having areas A1
(lower) at elevation h,, and A2  (upper) at elevation h2, the NPL
H, is given by:
                                 A1+A2
                                                               (6)
     From equation  6,  if A, = A2,  the NPL will  be  at  a  level
midway between h, and h2.   Where A, « A2 (openings are
concentrated  in  the upper  part  of the building),  the NPL will be
close to the  top as in Figure 2.   Where A, » A2  (openings are
concentrated  in  the lower  part  of the building),  the NPL will
also be located  close  to the bottom  of  the building, as in Figure
3.

     Buildings,  as  we  construct them, have leakage areas through
the side walls also.   But  as builders utilize  better doors and
windows and infiltration barrier  house  wraps,  the leakage through
walls will decrease significantly.   At  the same time,  many
weatherization techniques  reduce  the leakage rates through doors
and windows.  These buildings have effectively become  better
chimneys, with a greater potential for  drawing radon-laden soil
gas into the  building. Openings  in  the floor  and ceiling will
become more critical.

-------
Figure 2. The NPL in a building, with most low openings sealed,
is located near the top of the building resulting in a lower
pressure inside than outside throughout the building.  The
remaining low openings are at an even greater pressure
difference, resulting in more flow through the remaining
openings.
                     Neutral Pressure  Level
Figure  3. Sealing  openings at the  top  of  the  building causes the
NPL to  be lowered,  resulting in  a  decrease  in infiltration
through low openings.

-------
                        THE NPL AND RADON
     Since locations below the NPL are under a lesser pressure
than outside, air will try to enter the house through these
openings.  The farther from the NPL, the greater the pressure
difference, or force pushing the air into the house.  Conversely,
the higher above the NPL, the greater the force pushing air out
of the house.

     At the same time, the location of the NPL depends on the
distribution of the openings.  With uniformly distributed
openings, the NPL will be close to the center of the building.
But with openings concentrated at the top of the house, the NPL
will also be located close to the top of the house.  Openings
concentrated low in the house result in a NPL low in the house.

     Cavallo et. al.  (6) found that opening windows in a basement
(to increase ventilation) also reduces the rate of radon entry.
This is a result of increasing the ratio of openings below the
NPL to those above the NPL.  They effectively lowered the NPL and
decreased the pressure differences across the infiltration
openings.

     One primary radon mitigation or prevention technique is to
seal cracks and openings in the floor of a basement or ground
floor in an attempt to reduce radon entry.  This technique has
had mixed results, and in some cases even resulted in increased
levels of radon in the living space.  By sealing openings in the
floor (below the NPL), the remaining openings are concentrated
higher in the house.  This pushes the NPL up, creating an even
greater pressure difference across any openings in the floor
which did not get sealed.  A greater pressure difference then
potentially causes more  (not less) radon-laden air through the
cracks into the house.

SEAL THE CEILING FIRST

     Conversely, if cracks and openings in the ceiling or
locations above the NPL  are sealed, the NPL moves downward.  The
result is a lesser pressure difference  (or driving  force) across
the lower openings, resulting in less radon-laden airflow through
these openings.

     Figure  4 shows typical  infiltration/exfiltration  locations
in a typical house.   Exfiltration sites are typically  above the
NPL and  are  listed  in Table  1.  Infiltration  sites, typically
located  below the NPL,  are listed in Table 2.  Sealing
exfiltration sites will  result in a lowering  of the NPL and
resultant  lowering  of the pressure  differences across
infiltration openings.

-------
Figure 4.  Typical leakage sites in a residential structure.
Under Stack Effect pressures, sites located high in the house
will have air leaking out, while low sites will have air leaking
in.

-------
TABLE 1.  EXFILTRATION SITES AS SHOWN IN FIGURE 1.  SEALING OR
 REDUCING FLOW THROUGH THESE OPENINGS  WILL  REDUCE  THE  PRESSURE
           DIFFERENCE ACROSS INFILTRATION OPENINGS.


      Number       Exfiltration Site
          1        Ceiling/wall/beam joints
          2        Ceiling mounted light fixtures
          3        Recessed light fixtures
          4        Bathroom exhaust vents
          5        Whole house fans
          6        Attic access hatchways/stairways
          7        Inside electric/plumbing penetrations
          8        Flue/chimney penetrations
          9        Wall/ceiling perimeter joints
         10        Windows and doors
         11        Kitchen exhaust vents, range hoods
         12        Fireplace/stove dampers & doors
         13        Wall/floor baseboard joints
         14        Wall mounted heaters
         15        Leaking supply ducts
         16        Combustion furnaces/water heaters
 TABLE 2. INFILTRATION SITES AS SHOWN IN FIGURE 1.  SEALING OR
  REDUCING FLOW THROUGH THESE OPENINGS MAY ONLY INCREASE FLOW
               THROUGH OTHER INFILTRATION SITES.
      Number       Infiltration Site
         10        Windows and doors
         13        Wall/floor baseboard joint
         17        Exterior wall penetrations
         18        Interior crawl space hatches
         19        Floor cracks and joints
         20        HVAC registers/returns
         21        Band joist/mud sill joint
         22        Plumbing penetrations
         23        Electric/tele/TV penetrations
         24        Cracks/pores in concrete walls
         25        Floor/wall joints
         26        Sump openings
         27        French drains
         29        Cracks in slab floors
         30        Columns/stairs penetrating  slab

-------
Other Benefits of Lowering the NPL

     Lowering the NPL not only directly reduces the forces
drawing radon into a building, but will also increase the
efficiency of sub-slab depressurization systems. The suction
under the slab does not have to be as large to prevent radon
entry.  This would also allow further extension of an adequate
suction field under the slab (as measured relative to the
basement).

     Lowering the NPL also reduces infiltration into the house in
general.  Reducing this infiltration will have an effect on the
energy use to heat this infiltrating air.

     Exfiltration will also decrease with a lowered NPL.  In the
past, houses were almost exclusively heated with combustion
appliances, resulting in a NPL at or above the ceiling.  The high
NPL causes infiltration through every opening in the house.  As
such, cold dry air outside was drawn in through the building
shell.  With the use of heating systems other than combustion
appliances such as electric strip heaters and heat pumps, the NPL
has come down to somewhere around the midlevel of the house.
Exfiltration of warm moist air has occurred.  This exfiltration
has subsequently led to condensation and moisture problems in
walls and ceilings.  Lowering the NPL by sealing openings above
the NPL would reduce this exfiltration and potentially reduce
these moisture problems.

Words of caution

     Sealing any openings in a building may reduce the overall
air change rates, resulting in indoor air quality problems.  This
situation has to be evaluated more thoroughly to prevent
problems.

     Reducing air change rates will almost certainly result in
increased moisture content in the indoor air.  This will affect
the density of the indoor air and as such result in a larger
stack effect pressure between inside and outside.

     Powered ventilation systems such as kitchen exhaust fans, as
well as chimneys for furnaces and fireplaces, tend to raise the
NPL.  In some instances, so much air is vented that the NPL is
above the ceiling, placing the entire house under a vacuum.  In
these cases, adding an outside air supply directly to the
appliance is necessary.

     Wind forces on the exterior of the building also affect the
inside/outside pressure relationships.   Large opening upwind

-------
relative to downwind lower the NPL.  Larger openings downwind
tend to raise the NPL.

                           CONCLUSIONS


     Two conditions must exist for radon-laden soil-air entry
into a biulding: 1) an opening and 2) a pressure difference.
Sealing openings in the floor in an attempt do reduce radon entry
increases the pressure difference across any unsealed openings in
the floor.  Conversely, sealing openings in the ceiling or upper
locations of a building will lower the neutral pressure level
within the building.  This results in a lower infiltration rate
as well as a lower pressure difference across infiltration
openings.  Less radon-laden air will be drawn in through openings
in floors or below-grade walls.  Depressurization fields of SSD
and ASD systems will increase in strength relative to the
basement.
                            REFERENCES

(1)  U.S. EPA & New York State Energy Office, Reducing Radon In
     Structures/Training Manual.

(2)  Brennan T, (1990) Evaluation of Radon Resistant New
     Construction Techniques. Preprints of the 1990 International
     Symposium on Radon and Radon Reduction Technology, Vol. 5,
     p.l.

(3)  Nuess, R.M. and Prill, R. J. (1991) Radon Control - Towards
     a Systems Approach. Preprints of the 1991 International
     Symposium on Radon and Radon Reduction Technology: Vol. 4,
     VII-4.

(4)  Albright, L.D. Environmental Control for Animals and Plants.
     ASAE Textbook #4, August 1990,  Pamela DeVore-Hansen, Ed. pp.
     319-345.

(5)  Blomsterberg, A. K.  and D.J. Harrje, (1978),  Approaches to
     Evaluation of Air Infiltration Energy Losses in Buildings,
     ASHRAE Transactions, Vol 85, Pt 1, PP. 797-815.

(6)  Cavallo, A.,  K. Gadsby and T.A. Reddy, Natural Basement
     Ventilation as a Radon Mitigation Technique. Preprints of
     the 1991 International Symposium on Radon and Radon
     Reduction Technology, Vol. 2, IV-6.

(7)  House Tightening Manual For Homeowners and Weatherization
     Contractors.  Bonneville Power Administration,  October 1985,
     DOE/BP/13301-1.

-------
                                                                  VIP-6
      RELATIVE EFFECTIVENESS OF SUB-SLAB PRESSURIZATION AND
 DEPRESSURIZATION SYSTEMS FOR  INDOOR RADON MITIGATION;  STUDIES
         WITH AN EXPERIMENTALLY VERIFIED NUMERICAL MODEL

           by: Ashok Gadgil, Yves Bonnefous,  and Bill Fisk
               Indoor Environment Program
               Lawrence Berkeley Laboratory
               University of California at Berkeley
               Berkeley, CA  94720
                             ABSTRACT

     The performance of sub-slab-ventilation (SSV) systems has
been studied with a numerical model, which was earlier
successfully compared with experiment.  The parameters explored
in this study are the permeabilities of the soil and the sub-slab
gravel, the magnitude of pressurization (or depressurization)
applied by the SSV system, and the mode of SSV application (i.e.
pressurization (SSP) or depressurization  (SSD)).

     The mechanisms contributing to the successful performance of
SSP and SSD systems are identified.  The numerical modeling
demonstrates that placement  of a sub-slab gravel layer
substantially improves the SSV system performance.  Except in the
case of highly permeable soils, SSD systems are predicted to
perform better than SSP systems.  This prediction is consistent
with anecdotal experience.   The numerical model is used to
elucidate the reasons for this difference in performance.


1.   Background

     The health risk from indoor exposure to radon progeny is the
largest of the health risks  arising from  indoor pollutants (1).
SSV systems are commonly used to reduce elevated  levels of indoor
radon  in houses with basements  (2).  The mode of  operation of
sub-slab pressurization  (SSP) and  sub-slab depressurization  (SSD)
systems can be qualitatively described as follows.

     SSP systems  are used to ventilate the soil in the sub-slab
region.  This reduces the radon concentration  in  the soil-gas in
the sub-slab region and  is  intended to reduce the rate of radon
entry  into the basement.  The air  entering the basement from the
sub-slab region in  presence  of  SSP system operation  is primarily
fresh  air  injected  into  the  sub-slab  region by the SSP system,
carrying along with it a small  amount of  radon picked up during
the sub-slab transit  (Figure 1).

-------
      SSD systems invert the pressure-gradient across the basement
 slab,  which normally drives entry of radon bearing soil gas into
 the basement.   Owing to this inversion of the pressure gradient,
 basement air is sucked into the sub-slab region through cracks
 and joints in  the basement slab.   This air,  and also radon-
 bearing soil gas,  are sucked into the SSD system pipe and
 expelled to the outdoors by the SSD system (Figure 2).

      Experiments and field measurements of SSD and SSP system
 performance provide additional  details which can be summarized as
 follows.

 1.    In a majority of cases SSD systems perform better than SSP
      systems in reducing radon  entry (2).   SSP systems were found
      to perform better in one field study in Spokane with
      basement  houses built on highly permeable soil (3).

 2.    A good connection of the SSV system to  the sub-slab  region
      is important  for good SSV  system performance.   Factors
      ensuring  such connection are:   (a)  the  presence of a highly
      permeable gravel layer beneath the basement slab,  (b)
      absence of a  partitioning  of the sub-slab gravel  layer with
      internal  footers (this can be accomplished e.g.  by having
      either discontinuous internal footers,  or embedding  shot
      lengths of pipes transverse  to the internal footers),  for
     good pressure communication  throughout  the sub-slab  gravel
      layer (4-6).

 3.   ssy  performance is  improved  by the presence of a pit at the
     point of  penetration of the  SSV pipes through  the  basement
     slab (7) .

 4.   Sealing of the  cracks  and  joints in the basement slab
     improves  SSV  system  performance  by reducing short-circuiting
     of the  extension of  pressure  imposed at the SSV system pit,
     throughout the  sub-slab region (7).

 5.   A  large value  for the  ratio  of the permeability of the sub-
     slab  gravel to  the permeability  of the soil  increases SSV
     system  performance  (7).

     Most  of the findings enumerated  above have  not been
rigorously proven by  field  studies.   Becuase the uncontrolled
factors in field studies, these findings are only tentative.  A
numerical modeling study  of  SSV systems  was undertaken to obtain
a quantitative  understanding  of how these factors influence the
system performance.

-------
2.   Model Description

     Models of radon entry have been recently reviewed by Gadgil
(8).  The Darcy law, relating pressure gradient to the bulk
velocity in a porous medium, has been commonly used in a number
of numerical and analytical studies of soil gas entry into
basements.  However, this law is not a valid description of the
flow when the Reynolds number (based on the gravel diameter) of
the flow becomes large.  The literature suggests an upper limit
of 1 to 70 for the Reynolds number beyond which the Darcy law no
longer applies.  Since the Reynolds number of the flow in sub-
slab gravel greatly exceeds this upper limit during SSV
operation, a model based on non-Darcy flow must be used to study
the system performance.  In a previous study, Gadsby et al.  (9)
used an analytical model of SSV system performance based on the
exponential form of the non-Darcy flow.  However, the conclusions
of their study are limited by the assumptions that there are no
cracks in the basement slab  (so there is no flow between the
basement and the gravel layer), and that the boundary between the
sub-slab gravel and the soil is also impermeable.

     For this study, a numerical model of SSV system operation
was written.  The model, "Non-Darcy STAR"  (Non-Darcy Steady-state
Transport of Air and Radon), is a three dimensional finite-
difference model on a rectilinear coordinate system, based on the
SIMPLE  (Semi Implicit Method for Pressure Linked Equations)
algorithm developed by Patankar  (10).  The model incorporates the
Darcy-Forchheimer  (11) equation for flow through permeable media,
and the equation of continuity:
           Vp = -—(1 +


           V.V = 0                                      (2)

 where p is the disturbance pressure (i.e.  the change in soil-gas
 pressure owing to the depressurization of  the basement and the
 pressure applied by the SSV system) ,  ji is  the dynamic  viscosity
 of soil-gas,  K is the permeability  of  the  porous  medium (i.e.
 soil or gravel),  c is the Forchheimer term,  and V is the bulk
 velocity of soil-gas.  The soil block is assumed isothermal for
 the present study (although it would be straight forward to
 incorporate the effects of buoyancy changes  on soil gas flow in
 the model).  The computational domain is bounded from below, at  a
 depth of 10 m below the basement slab, by a  no-flow boundary, and
 from three sides by no-flow vertical surfaces at 10 meters
 horizontal distance from the basement walls.  The fourth vertical
 surface represents the plane of symmetry of  the problem and
 vertically bisects the basement and an internal footer (Fig. 3).

-------
 This fourth vertical surface is also a no-flow boundary owing to
 symmetry.  The computational domain is .bounded from above by the
 basement slab, the wall and footers (with a joint defined between
 the slab/wall and the footers),  and the soil-surface outside the
 basement.  The slab is impermeable except at the joint defined
 along its entire periphery.  The soil-surface outside the
 basement is defined to be at zero disturbance pressure.

      Once the velocity field for the soil-gas is calculated,  the
 radon concentration field is calculated with the following
 equation:


           V.(^ VcRn)-V.VcRn+S(S-XRn) =0                (3)

 where Ł  is the bulk diffusivity  of radon  in the  porous medium,
 cRn is tne concentration of radon in the  soil-gas,  S is the
 release  rate of radon into the  soil gas per unit volume of the
 porous medium, /.Rn  is  the  inverse  of the  time-constant of  radon
 decay, and e is the porosity of the  medium.  The model  is
 described in detail by Gadgil,  et al.  (12).   The solution
 approach used  for solving  eg.  (3)  is similar to  that used  by
 Loureiro (13).  The rate of radon entry into the basement  is
 calculated in  the model  by integrating  over the  crack area, the
 product  of soil-gas entry  rate and the  local radon  concentration
 in  the soil-gas.   Finally,  the radon concentration  in the
 building is evaluated  using the  following two  simplifying
 assumptions:   (1)  The  air  in the  building is perfectly and
 instantaneously mixed  and  (2) air-exchange rate  in  the building
 is  obtained by summing in  quadrature (14)  a  fixed typical  air
 exchange rate  and the  air  exchange rate induced  by  operation of
 the SSV  system.   To simplify the  model  and make  results easier to
 understand,  we have neglected diffusive transport of radon
 through  the concrete and the radon in outdoor  air.


 3.    Description  of  Parametric Study

      The parametric  study was defined on  a  "typical"  single
 family residence  with  a basement;  with  an  axis of symmetry; the
 latter permitted  us the modeling of  only  one of the  two  symmetric
parts of the building  and the soil block  surrounding  it.  The
 floor plan  and a  cross-section of  the modeled basement  are shown
 in Figure  3.   The layout and dimensions are based on  common
single family  houses, but are not  based on a statistical survey.

     The cracks in the basement slab are represented with a
single equivalent crack of width 1 mm,  located at the joint of
the slab with the wall and footer, and has an L shape in cross-
section.   In the model, the exterior surface of the basement

-------
walls is assumed to be displaced outwards to line up with the
footer edge (see Figure 3, basement section).  This allows some
reduction in the number of computational control-volumes required
in the model, and a corresponding reduction in computational
time.  The basement walls are assumed perfectly impermeable, so
is the top surface of the soil outside the basement that is
covered by the slab of the garage.  The permeabilities and
Forshheimer terms for three kinds of gravels are those measured
in the laboratory (12).  These three gravel types represent the
range of types used in housing construction in the state of
Washington  (12).  The backfill region is assumed to have the same
permeability as the soil, and is assumed to be in firm contact
with the exterior surface of the basement wall.   (The influence
of gaps between the footer and soil, the permeability of the
backfill, and the cracks  in the basement walls on the SSV
performance is under study).  A pit  (25 cm radius) is present
where the SSV pipe penetrages the basement floor  (see Figure 3)
 (12)   The pressures imposed with the SSV system  at this pit are
either 60 Pa  (the lower limit of SSV system after e.g. system
degradation with age), or 250 Pa  (the upper limit for systems
 installed in houses with  a sub-slab gravel layer).  The basement
 is assumed  to be under a  fixed depressurization of -10 Pa.  The
model does  not account for an upper  limit on the  flow-rate
 capacity of the SSV fan.

      In this  study, we assume a  radon concentration  in the  deep
 soil-gas of 58,830  Bg/m3, equal  to three times the smallest deep
 soil-gas radon concentration measured by Turk  et  al.  (16)  in the
 region of Spokane,  Washington.   The  sub-slab gravel  layer  is
 assumed to  release  no  radon  into the  soil-gas.  The  building
 volume  is taken  to  be  122.5 m3,  and  has  a  fixed value  of  air
 exchange  rate (in the  absence of SSV system  operation) of  0.4
 ACH.

      The  choice  of  the degree of resolution  of the  computational
 domain  into control volumes  is  determined  by the  trade-off
 between  computational  effort  and residual  errors  owing to
 inadequate  resolution of regions with large  gradients  of velocity
 and pressure.   The  computational domain in most  of  the parametric
 runs had  37,260  control  volumes.  We estimate that  the error in
 radon entry rate resulting from inadequate grid  resolution is
 less than 20%.


 4.    Results

      The performance of SSV systems is evaluated in this study
 with two criteria:   (1)  a good extension of pressure field in the
 sub-slab region, and (2) the degree of reduction of indoor radon
 concentration achieved in the building described in the previous
 sections.

-------
a
 4.1.  Indoor Radon Concentrations in Absence of SSV

      In absence of SSV,  the model predictions of indoor radon
 concentrations for different combinations of soil and sub-slab
 gravel,^are shown in Figure 4.   We note that our model
 predictions show a substantial  enhancement in indoor radon
 concentration resulting  from placement of a highly permeable  sub-
 slab  gravel layer.  This effect has been noted also by Revzan et
 al.  (17),  using a different numerical  algorithm and a distinct
 computer model to solve  equations (1-3).   Placement of an
 impermeable plastic membrane below the sub-slab gravel
 substantially reduces this detrimental increase in radon entry
 (9),  and is being studied in more detail.


 4.2.  SSD System Performance

      Figures 5 and 6 show model  predictions for SSD system
 operation  with a depressurization at the  pit of -60 Pa and -250
 Pa respectively.   Each figure displays several  curves,  each for
 different  permeability of sub-slab gravel.   The horizontal axes
 display  soil permeability.   The  maximum sub-slab pressure  at  30
 cm from  the wall,  a  measure of  sub-slab pressure extension, is
 plotted  along the horizontal axes.  This  measure is chosen so
 that  pressure extension  can be  experimentally assessed without
 being influenced by  small-scale  variations  in the width of the
 wall-slab-footer crack.   The indoor radon concentration after SSD
 operation,  expressed as  percentage of  the indoor radon
 concentration before SSD system  operation,  is shown on the
 figures  in  parentheses near each simulation  result.

      The detailed results  from these simulations  of soil-gas
velocity and soil-gas  radon concentration fields  provide
additional  insights  into  SSD operation.  SSD system operation may
be satisfactory  owing  to  either  of two mechanisms.

1.    If the  extension  of the  depressurization in  sub-slab  gravel
      layer  is  such that the  sub-slab gravel layer  is at a  lower
     pressure  than the basement  at all points, then the direction
     of pressure-gradient across the cracks in the basement slab
     is reversed, and the system functions in the desired  fashion
      (Fig. 2).

2.   If_the  flow drawn by the SSD system is large, the sub-slab
     soil adjacent to the gravel layer is "ventilated" by the
     fresh outside air entering the outdoors soil surface and
     traveling to the SSD pit.  In this case, even if the
     pressure field extension in the sub-slab region in
     inadequate to reverse the pressure-gradient across all
     cracks in the basement slab, the large reduction in radon-

-------
     concentration in the soil gas entering the basement can
     prpvide an effective system performance.

     In Figures 5 and 6, the simulation results show elimination
of radon entry through the basement cracks for the results
displayed below the horizontal line representing -10 Pa (the
pressure in the basement), owing to mechanism (1) above.  The
large reduction in indoor radon concentrations, above the line at
-10 Pa, and particularly at large soil permeabilities (near the
top right hand corner of each figure), are a consequence of the
mechanism (2) described above.

     Results presented in Figures 5 and 6 support the following
conclusions regarding SSD system operation:

1.   Placing of a sub-slab gravel layer is necessary for best
     performance of an SSD system (note the curve in both figures
     for simulation results without sub-slab gravel).  On the
     other hand, an increase in the applied depressurization from
     -60 Pa to -250 Pa allows satisfactory reduction in indoor
     radon concentrations for all cases studied in the
     simulations, even in the absence of the gravel layer, by
     ventilation of the sub-slab soil.

2.   An increase in sub-slab gravel permeability from 2 x 10~8 m2
     to 3 x 10~7 m2 at least triples the minimum subslab
     depressurization predicted at a distance of 30 cm from all
     the walls.  This improvement in the sub-slab pressure field
     extension provides a large margin of safety for assured good
     performance of the SSD system.  In addition, the increase in
     permeability of the sub-slab gravel permits a reduction in
     the applied depressurization.  This could be significant in
     some cases when system components degrade over the years
     (e.g. caulking around the SSV pipe penetration in the slab
     may develop cracks), or can be translated into savings in
     first cost by installing a lower power fan in the system.

3.   For ratios of soil to gravel permeabilities larger than
     1000, it is possible to interpret the results as being
     equivalent to the soil being completely impermeable—
     practically all the air drawn by the SSD system originates
     from inside the basement and enters the SSD pipe by crossing
     the sub-slab gravel layer after passing through the cracks
     in the slab.  For these situations, good SSD system
     performance is achievable with quite low values of applied
     depressurization (theoretically, even for an applied
     depressurization of — 10.1 Pa for a basement depressurized
     to -10 Pa).  This model prediction encourages further
     exploration of applicability of passive SSD systems  (based
     on depressurization of the SSD pipe by buoyancy forces).

-------
4.3  SSP System Performance

     Figures 7 and 8 show predicted performance of SSP system for
a range of soil and gravel permeabilities.  The axes and
presentation symbols have the same interpretations as described
for Figures 5 and 6.  Figure 7 is for an applied pressurization
of +60 Pa, and Fig. 8 for that of +250 Pa.

     In absence of a gravel layer, an SSP system can not provide
a complete elimination of convective radon entry from soil.  For^
lower pressurization (+60 Pa) and low permeability soils (K = 10
11 m2), the operation of an SSP system is predicted to lead to an
increase in the indoor radon concentration; the dilution of the
radon concentration in the soil gas entering the basement does
not compensate for the net increase in entry under these
conditions.  The placement of a sub-slab gravel layer is crucial
for successful operation of an SSP system.

     The simulations were undertaken with the  (commonly
justifiable) assumption that the radon emanation from the sub-
slab gravel layer is negligible.  The detailed simulation results
of the pressure, velocity and radon concentration lead to the
following understanding of the several mechanisms influencing
effective SSP system performance.

1.   A good extension of pressure field ensures that no region of
     the sub-slab gravel layer is under absolute depressurization
      (p < 0 Pa).  So long as this condition is satisfied, radon
     can not enter the sub-slab gravel layer by convection.  This
     condition  is not satisfied for the gravel of low
     permeability  (k = 2 x 10~8 m2 in our study) at an applied
     pressure of +60 Pa  (see Fig. 7).

2)   At the same time, the system can have a good performance
     even in the presence of a poor extension  of the pressure
     field, if  the  flow of air injected in the soil by the SSP
     system is  large  (as is the case for very  permeable soils).
     This large flow essentially washes away the radon in the
     near-basement  soil and the concentration  of radon in the
     soil gas entering the gravel slab  (owing  the poor pressure
     field extension) is reduced.

3)   The soil must  be sufficiently permeable that the flow
     between the pressurized sub-slab gravel layer and the sub-
     gravel soil must not be too  small. If this flow is
     adequately large, the air passing  from the gravel into the
     soil pushes away the radon that may  otherwise enter the sub-
     slab gravel layer from the sub-gravel soil by diffusion
     across the large concentration gradient.  In case of highly
     impermeable soils  (e.g. k <  lO"11 m2), the flow of air
     across the interface of soil and gravel is close to zero

-------
     (the soil can be practically considered impermeable),
     allowing radon to diffuse into the air passing along the
     gravel layer from the adjacent soil.  This radon then enters
     the basement from the sub-slab gravel along with the air
     flow that occurs through the cracks during all SSP system
     operation.

4)   Enhancement of the pressure imposed by the SSP system at the
     pit can reduce the diffusion of radon from the sub-gravel
     soil into the sub-slab gravel only at the cost of an
     increase in the air exchange rate in the building (by
     dilution).  This can have a significant energy cost.  In
     conclusion, for a proper functioning of the SSP system,
     highly permeable soil and gravel are required.


5.   Discussion

     Our parametric simulation study of a typical house shows
that generally an SSD system is predicted to perform better than
an SSP system, except in the case of extremely permeable soils (k
> 2 x 10~8 m2).  For the combinations of highly permeable gravels
(k > 10~7 m2) and  permeable soils (10~10 m^ < k < 10~9 m2) the
two systems have equivalent performance. This is summarized in
Table 1.

     A permeability in excess of 2 x 10~8 m2 is a unusually high
value for soil permeability. The simulations show that only in
such cases the SSP system performance is better than the SSD
system performance. However, ongoing research (18) shows that the
numerical models under-predict a soil-gas flow entering basements
(in absence of an SSV system) by approximately a factor of eight.
The phenomenon of the "washing" of the radon from near-building
soil is based on the large rate of air-flow from the SSP system
into the soil. Therefore, we suspect that SSP system performance
may exceed SSD system performance at lower soil permeabilities
than predicted by Non-Darcy STAR.

     Except for cases of the SSP system performance in the
absence of a gravel layer and for soil permeabilities less than
or equal to 10~10 m2, an increase in the value of the imposed
pressure at the SSV pit from 60 to 250 Pa allows a satisfactory
reduction in the indoor radon concentration.  However, this
increase in the imposed pressure can translate into a large
increase in the air exchange rate of the house, with the
associated energy cost for heating (or cooling) the house.   As an
example, for an SSP system functioning at an imposed pressure of
250 Pa in presence of a highly permeable gravel (k = 3 x 10~7 m2)
and a soil with low permeability (k < 10"*1 m2), the rate of air
exchange increases from 0.4 ACH assumed in the absence of SSP to
1.95 ACH in the presence of SSP operation. The energy costs

-------
 associated  with  this  increase  is  generally  significant.  For the
 system  operating in highly permeable  soils, the  flow  associated
 with  a  depressurization  of -250 Pa  exceeds  the maximum practical
 flow  rates  of  SSV fans  (about  0.2 m3/s).

      A  plastic membrane  below  the gravel  layer can be beneficial
 to the  performance of the  two  systems,  and  would permit a better
 extension of the pressure  field  (for  both the SSP and SSD) and
 also  eliminate the diffusive entry  of radon from sub-gravel soil
 into  the sub-slab gravel during SSP system  operation.

      Besides having a generally poorer performance than the SSD
 system, the SSP  systems  have two  major inconveniences: 1) the
 flow  of air entering  the house induced by the system  operation
 can entrain other soil-resident pollutants  such as termicides,
 and 2)  the  air injected  by the system is  at outdoor temperature,
 which,  in severe winters,  could lead  to freezing of sub-slab
 pipes.  An SSP  system  based on  injecting house (i.e. indoor) air
 is not  a viable  option owing to the continuous increase in indoor
 radon concentration as the radon  diffuses into the air during its
 passage through  the sub-slab gravel.


 6.    Conclusion

     A  numerical  study of  SSV  systems has been carried out using
 a model "Non-Darcy STAR".  This study  has  led to the
 identification of mechanisms contributing to the success of these
 systems: (1) for SSD  systems,  inversion of  the direction of
 pressure gradient across the basement slab  and the reduction in
 the radon concentration  in the soil and (2) for SSP system
 pressurization of the whole of the  sub-slab gravel layer
 eliminating convective entry from sub-gravel soil into the sub-
 slab gravel, reduction in  the  radon concentration  in the soil,
 and air flow across the  interface of  soil and the sub-slab gravel
 suppressing the  diffusion  of radon  into the gravel layer.

     A  layer of  highly permeable  gravel is  necessary  for best
 performance of the SSV systems. Selection of a highly permeable
 gravel  (k = 3  x  10~7 m2) permits  considerable improvement in the
 SSV system performance and also provides  a  margin of  safety in
 case of system deterioration.  In  certain  cases (e.g.  installation
 of SSD  system  in  soils with permeabilities  of less than or equal
 to 10"11 m2),  it  might allow successful operation of passive
 systems.

The SSD systems  are generally  better  at reducing indoor radon
concentrations than SSP  systems with  the  exception of
 installations  in  soils of  extremely high  permeabilities (k > 10~8
m2) .

-------
7.   Table Caption
Table 1:  This table shows which system (SSD or SSP) has the
          better performance in radon reduction for various
          specific combinations of gravel and soil
          permeabilities, as predicted by the model Non-Darcy
          STAR.

-------
8.
List of Figures
Figure 1: Conceptual representation of the functioning of a sub-
          slab pressurization  (SSP) system.

Figure 2: Conceptual representation of the functioning of a sub-
          slab depressurization (SSD) system.

Figure 3: The plan and section of the "typical" house modeled for
          the parametric simulations with Non-Darcy STAR. Owing
          to the assumed plane of symmetry of the house, only
          half of the house is shown in the plan and the section.
          The modeled space includes the soil block surrounding
          the house extending 10 meters to all sides of the
          house, and 12.5 meters below the outside soil surface.

Figure 4: Indoor concentration of radon in an idealized perfectly
          mixed house of volume 122.5 m3,  and with an air
          exchange rate of 0.4 ACH. The assumed concentration of
          radon in the deep soil is 58830 Bq m~3.

Figure 5: Predicted SSD system performance for an applied
          pressure at the pit of -60 Pa. The pressure in the
          basement is assumed to be -10 Pa.  Percentage reduction
          in the indoor radon concentration resulting from SSD
          operation is shown in parenthesis for each of the
          simulation points on the figure.

Figure 6: Predicted SSD system performance for an applied
          pressure at the pit of -250 Pa.  The pressure in the
          basement is assumed to be -10 Pa.  Percentage reduction
          in the indoor radon concentration resulting from SSD
          operation is shown in parenthesis for each of the
          simulation points on the figure.

Figure 7  Predicted SSP system performance for an applied
          pressure at the pit of +60 Pa. The pressure in the
          basement is assumed to be -10 Pa.  Percentage reduction
          in the indoor radon concentration resulting from SSD
          operation is shown in parenthesis for each of the
          simulation points on the figure.

Figure 8: Predicted SSP system performance for an applied
          pressure at the pit of +250 Pa.  The pressure in the
          basement is assumed to be -10 Pa.  Percentage reduction
          in the indoor radon concentration resulting from SSD
          operation is shown in parenthesis for each of the
          simulation points on the figure.

-------
9.   References


(1)  Nero, A.V. "Controlling Indoor Air Pollution", Scientific
     American, Volume 258, No. 5, May 1988.

(2)  Henschel, B. "Radon Reduction Techniques for Existing
     Detached Houses — Technical Guidance, third edition", U.S.
     EPA  Office of Research and Development, Washington, D.C.
     20460, First Draft, January 1992, Documents in preparation.

(3}  Turk, B.H.; Prill, R.J.;  Fisk, W.J.;  Grimsrud, D.T; Moed,
(    B A.  and Sextro, R.G. "Radon and Remedial Action  in Spokane
     River Valley Homes.   Volume 1:  Experimental Desl9nTandA^ta
     Analysis".  Lawrence  Berkeley Laboratory Report, LBL-23430,
     December  1987.

(4)  EPA  "Radon Reduction  Techniques  for Detached Houses,
     Technical Guidance  (second  edition)", U.S. EPA,  Report
     EPA/625/5-87/019,  1987.

(5)  WSBCC  [Washington State  Building  Code Council]  "Washington
     State Ventilation and Indoor  Air  Quality  Code   Chapter 5
     Radon Resistive Construction  Standards".   Published by the
     State of Washington,  1990.

 (6)  Nuess,  M.  "Northwest Residential  Radon Standard Volume 1:
      Project Report".   Bonneville  Power Administration Report,
      Portland,  OR,  1989.

 (7)   Bonnefous,  Y.C.;  Gadgil,  A.J.;  Fisk, W.J.; Prill,  R.J.;
      Nematollahi,  A.  "Field Study & Numerical Simulation of
      Subslab Ventilation Systems." to appear in Environmental
      Science and Technology,  1992.

 (8)   Gadgil, A.J.  "Models of Radon Entry:  A Review".  Accepted
      for publication in Rad.  Prot. Dosimetry,  1992.  Also
      Lawrence Berkeley Laboratory Report  LBL-31252, Berkeley, CA,
      August 1991.

  (9)   Gadsby, K.J.; Reddy, T.A.; Anderson,  D.F.; Gafgen, R; Craig,
      A B  "The Effect of  Subslab Aggregate Size on Pressure  Field
      Extension.  In:  Proceedings of the  1991 International
      Symposium on Radon and Radon Reduction Technology, vol.  4,
      April 2-5, Philadelphia, PA.  Published by the U.S. EPA.

  (10) Forchheimer, P.H.  Z. Ver. Dtsch.  Ing. 45, pp.  1782-1788,
      1901.

  (11) Patankar, S. Numerical Heat  Transfer and  Fluid  Flow.  McGraw-
      Hill Book Company, New York, 1980.

-------
(12)  Gadgil,  A.J.;  Bonnefous, Y.C.; Fisk,  W.J.; Prill, R.J.;
     Nematollahi,  A.   "Influence of Subslab Aggregate
     Permeability on SSV Performance".   Lawrence Berkeley
     Laboratory Report,  LBL-31160,  Berkeley, CA  1991.

(13)  Loureiro,  C.O.;  Abriola, L.M.; Martin, J.E.; Sextro, R.G.
     "Three Dimensional  Simulation of Radon Transport into Houses
     Under Constant Negative Pressure",  Environmental Science and
     Technology,  Vol. 24, pp. 1338-1348, 1990.

(14)  Sherman,  M.H.  "Superposition in Infiltration Modeling",
     Lawrence Berkeley Laboratory Report,  LBL-29116, Berkeley,
     CA.   Submitted to Indoor Air,  1992.

(15)  Palmiter,  L.;  and Brown, I.  "Northwest Residential
     Infiltration Survey:  Analysis and Results".  Ecotope, 2812
     East Madison,  Seattle, WA  1989.

(16)  Turk,  B.H.;  Harrison,  J.; and Sextro,  R.G. "Characterizing
     the  Occurrence,  Source, and Variability of Radon in Pacific
     Northwest Home".  J. Air Waste Manage. Assoc.  40, pp. 498-
     506,  1990.

(17)  Revzan,  K.L.;  Fisk,  W.J.; and Gadgil,  A.J. "Modeling Radon
     Entry into Houses with Basements:   Model Description and
     Verification".   Indoor Air, Vol. 1, No. 2, pp. 173-189,
     1991.

(18)  Garbesi,  K.;  Sextro, R.G.;  Fisk, W.J.; Modera M.P.; and
     Revzan,  K.L.  "Soil-Gas Entry into an Experimental Basement:
     Model-Measurement Comparisons and Seasonal Effects".
     Lawrence Berkeley Laboratory Report,  LBL-31873, Berkeley,
     CA.   Submitted to Environmental Science and Technology,
     1992.

-------
Perm.
         m2
                   2x10
                           1 x 10
                                   "9
                                          1x10
                                            -10
                                                 <1 x10
No gravel
                    SSP
                             SSD
                                           SSD
                                                   SSD
    2x 10
         "8
                 SSP
SSD
                                           SSD
                                                   SSD
     1x10-7
                 SSP
                                                      SSD
    3x10 '7
                  SSP
                                                       SSD

-------
        gravel
Figure  1.

-------
        gravel
Figure  2.

-------
Plane of
symmetry
Figure  3.
                                            Modified Wall location for
                                            easier computer description
                                                       \
                                            basement
                                      / x x y x /
y\pOOOOQt^
      pit f
                    i^:
                    i*
                    X X
                    rr
                    i\K
                                                        J$]
                                       Cross section A-A

-------
     10000
cr

EL

c
o
+3
CO
v.
*-•
C
CO
o
c
o
o

c
o
T3
CO
 O
 o
•o
 c
1000  :
 100
        10
                                   EPA guideline
, Mr\ rtrov/ol
INU yiclVUI
— Gravel No. 1 , k = 2
Gravel No. 2, k = 1
- - Gravel No. 3, k = 3

x 10'8 m2
x 10"7 m2
x 10'7 m2
           10
             ,-11
                         10
                           ,-10
10
  ,-9
10-
                                   Soil permeability [m2]
              Figure 4.

-------
(0
=     0.0
CO
CO

E
o


E
o
o
CO
+"»
CO
0)

3
CO
(0
0)

Q.

c
3

E
"5
03
•10.0
•20.0
•30.0
-40.0
-50.0   -J-
                                   (42)
                                  (33)
                   Basement pressure
-•- No gravel

-+-Gravel No 1, k = 2 x 10"8 m2

-*-Gravel No 2, k = 1 x 10'7 m2

"•"Gravel No 3, k = 3 x 10"7 m2
                                    (0)
                                     (0)
              Impermeable
                membrane
                              Soil permeability
                                                                  10
                                                                    •11
            Figure 5.

-------
03
Q.
I
CO
   0.0

 -20.0

 -40.0

 -60.0

 -80.0
 2

 O
O
CO
to  -100.0
 o
 ^  -120.0 1
 to
 8  -140.0 :

 g  -160.0 j

•|  -180.0
CO
-200.0 +-
                  Basement pressure
                -No gravel
                'Gravel No 1,k = 2x 1(T8m2
                'Gravel No 2, k= 1 x lO^m2
                'Gravel No 3, k = 3 x 10"7 m2
(2
             Impermeable
              membrane
                                   10
                                     n
                                               (0)
                                                         I i i i ni|	1 i i i mi
                                                     10'
                            Soil permeability [m2]
            Figure 6.

-------
CO
=   50.0  t~
03

I   40.0
E
o
*   30.0
o
o
CO
13
0)
3
(0
(0
Q)
Q.
E
3
E
!E  -10.0
    20.0
     10.0
      0.0
                                       (54)
 No gravel
 Gravel No 1,k = 2x 10'8 m2
-Gravel No2,k= 1 x 10"7 m2
 Gravel No 3, k = 3x 10-7m2
             Impermeable
              membrane
                       1 O'11           1 O'9

             Soil permeability  [m2]
                                                                    10
                                                                      -7
           Figure 7.

-------
CO
0.
i
(0


o



o

o
CO
+•»
CO
CO
CD

o.

E
3
E
160.0



140.0



120.0



100.0



 80.0



 60.0



 40.0



 20.0



  0.0
•No gravel

-Gravel No 1, k = 2x 10"8 m2

-Gravel No 2, k=1 x 10'7m2

•Gravel No 3, k = 3x 10'7m2
                                   (42)
                             (46)   (14)
              Impermeable
               membrane
                      10'11           1 O'9


            Soil permeability [m2]
(4) (0)
                                                                10
                                                                  r7
             Figure  8.

-------
   Session VII Posters
Radon Reduction Methods

-------
                                                                    VIIP-1
 RADON MITIGATION SYSTEMS — A LIABILITY IN COLD CLIMATE HOMES1?

          by:  Kenneth D.  Wiggers,  Ph.D.
               American Radon Services,  Ltd.
               Ames,  IA  50010
                             ABSTRACT
     Homeowner observations  and our studies have shown that there
is  not enough  air  available  in  many  relatively  air-tight  and
"average" homes to allow safe operation of natural draft combustion
appliances  (furnaces, water  heaters,  and woodburners)  and other
air-consuming appliances such as clothes dryers, exhaust fans and
radon mitigation system  fans.  Radon mitigation  system fans remove
an unpredictable amount  of conditioned air from homes.  A certain
amount of available household air becomes  dedicated to the radon
system; the air consumed by the radon system is no longer available
for  other  uses.     It   is   recommended  that  all  active  soil
depressurization (ASD) and submembrane  depressurization (SMD) radon
mitigation systems  installed in cold climate housing with natural
draft appliances be recalled.  The recall would  consist of using  a
blower door to determine the amount of air available in the homes
at different house pressure differentials.   It is recommended that
a blower door be used to  measure the  air  leakage  characteristics of
a home before  and after the  installation  of all radon mitigation
systems in cold climates.
   'Presented at The 1992 International Symposium on Radon and Radon Reduction Technology, September
22-25, 1992, Minneapolis, MN

-------
                           INTRODUCTION

     This study was prompted in part due to homeowner observations
(or perception) that radon mitigation systems caused natural draft
furnaces  and water  heaters  to backdraft.   The  homeowners  have
become aware of the backdrafting by installing tissue paper on the
draft diverters of their furnaces and by the observation that their
thermally-sensitive backdraft indicators had turned black.  After
each radon mitigation system is commissioned, backdraft indicators
are  installed on  natural  draft  furnaces and  water  heaters  and
passive carbon monoxide detectors are installed in a highly visible
place (usually on the refrigerator door).

     The potential for radon mitigation system-induced backdrafting
is discussed with the homeowner during the on-site  design phase
visit and in the subsequent "System Design for Radon Reduction"
report.   The backdraft indicators  give the  homeowner a method of
detecting  backdrafting.    One  young couple  reported that  their
backdraft  indicator on the  water heater indicated  backdrafting
(flue gas  spillage)  and questioned whether  their recent lethargy
and  short-term   memory   loss   could  be  associated  with  the
backdrafting  of the  water  heater  (and  hence the radon mitigation
system).

     The  accompanying  question  was  "Are  the  health  effects
permanent?"  In this example, the radon mitigation system consumed
(dedicated) 37 in2 of the 149 in2 premitigation available equivalent
leakage  area (ELA)  and 44 cfm of  the 129 cfm  premitigation air
available  at one pascal pressure  differential.   The cfm  of air
available at one pascal pressure dropped from a premitigation value
of  129  cfm  to  85  cfm   postmitigation.    A  furnace  requires
approximately 70 cfm air for proper combustion and drafting.  The
"numbers"  show there  is   not  much  tolerance   (forgiveness)  and
therefore installers of air-consuming appliances must address the
issue  "Is there adequate  air  for this additional  air-consuming
appliance?"

                       MATERIALS & METHODS

Approximately 25 homes that had radon mitigation systems installed
one  to  two years  ago were blower door  tested to determine the
amount of equivalent leakage area and cfm of air available at one
pascal pressure differential.  The blower door testing (BDT) was
conducted  with the  radon  mitigation  system  fan shut off  and
terminus of the PVC pipe taped shut (assumed pre-radon mitigation
conditions).    The  BDTs   were also  conducted  with  the  radon
mitigation system in operation.  All homes were retested for radon.

BDT  data have  been collected  on all  radon mitigation  systems
installed in  1992.  The equivalent  leakage area  (ELA) and the "cfm
available at  1 pascal" appear to be the most sensitive indices to
use  for  determining the effect of a radon mitigation  system on
household air availability.  The ELA sums all the leakage area in

-------
the house  into  one number.   The  "cfm available at  1  pascal"  is
calculated by the blower door computer program and represents the
cubic feet per  minute of air available for household  use  at one
pascal pressure differential between the inside and outside of the
house.  The  "n" value is  the slope  of the line drawn through the
air leakage data points.

     The "leakage ratio"  is  the ELA divided by the total 100s of
ft2 of surface area  (four  sides plus top and bottom). The ACH50 is
the air changes at 50 pascal pressure differential.  The cfm (air
flow) at any pressure differential (measured in  pascals)  can be
calculated by the following equation:   air flow =  (cfm at 1 pascal
pressure differential)(pascal").

                       RESULTS & DISCUSSION

     The  variability  of  air  leakage characteristics of  three
typical Iowa homes is illustrated in Table 1.   The values shown are
typical of the variability encountered in Iowa housing stock.

Table 1.  Air Leakage Characteristics of Three Iowa Homes  (Pre-
          mitigation)
City Address
Sioux City
Clive
Rock Rapids
cfm air
available at
one pascal
401
228
47
"n" value
0.554
0.652
0.561
ELA (in2)
422
243
62
     The  uncertainty of the  amount of available  air  in housing
compounded with  the  uncertainty of the amount of conditioned air
removed by an  ASD or SMD system creates an uncertainty about the
adequacy  of  combustion air for  proper  drafting of natural draft
appliances and hence  a  concern about the safety of the occupants.
Table  2  shows the  "pre"  and  "post"   radon  mitigation  cfm air
available and  ELA.

-------
Table 2.  Pre- and Post Radon Mitigation Blower Door Data
City
Sioux City
Clive
Rock
Rapids
Premiti-
gation
cfm air
available
401
228
47
Postmiti-
gation cfm
air
available
(pre - post)
294 (107)
172 (56)
35 (12)
Premiti-
gation
ELA (in2)
422
243
62
Postmiti-
gation ELA
(in2)
(pre - post)
294 (128)
212 (31)
53 (9)
     The numbers  enclosed in parentheses in the  third and fifth
column  show,  respectively, the cfm  air available at  one pascal
pressure differential and the ELA dedicated (consumeed by) to the
radon  system.   The  reduction amount  (value  in  parentheses)  in
available air and ELA  is  dedicated to the radon reduction system
and is no longer  available for use by natural draft water heaters
and furnaces.  Kitchen and bathroom exhaust fans, clothes dryers,
and other powered exhaust systems  further reduce the amount of air
available for natural draft water heaters and furnaces.  The amount
of air removed from the  house by  ASD and SMD  radon mitigation
systems is unpredictable  and  thus  each house  must be blower door
tested before and after the installation of ASD and SMD systems to
determine (best guess,  there are no standards)  if there is adequate
air in the home.

HOUSE ILLUSTRATIONS

Representative Case Study  — Marion. Iowa

     A typical example of the problem follows.  The owners of this
house  were  the first  to  call  us  to  tell  us that  "their radon
mitigation system made their furnace backdraft."   An alpha track
detector showed a radon concentration of 5.2 pCi/L in an ordinary
Marion, Iowa  home.   An ASD  system was  installed into a subslab
drainage tile using  4"  ID PVC pipe and an in-line fan capable of
moving 270 cfm at zero inches static pressure.

     The homeowner disconnected the fan for the family's ASD radon
mitigation system shortly after installation because it made the
furnace  "run more"  and because it  made the  furnace backdraft.
Blower door testing under various house conditions was conducted to
determine the effect of  the  ASD system  (4"  vs  2" PVC)  and the
effect of the clothes dryer on air leakage parameters.

     The  house in  the "radon  fan off"  condition theoretically
represents the  house before the radon  system was installed; the
electricity  to the  radon fan is  disconnected and duct  tape is
placed over the discharge end of the radon pipe.  Table 3 shows the

-------
air  leakage characteristics  associated  with two  air-consuming
appliances, a clothes dryer and  a  radon system with two sizes of
PVC pipe.

Table 3.  Air leakage characteristics — house in Marion.
House
condition
Radon fan
off
Radon fan
on (4H
PVC)
Radon fan
on (4"
PVC) +
dryer on
Radon fan
off +
dryer on
Radon fan
on (2")
Radon fan
on (2") +
dryer on
ELA,
(in2)
133
105
89
113
127
104
cfm
available
at 1
pascal
130
80
61
99
122
86
"n" value
0.543
0.653
0.700
0.590
0.551
0.614
Leakage
ratio
(inVlOO
ft2)
3.8
3.0
2.5
3.2
3.6
3.0
ACH50
5.0
4.8
4.3
4.6
4.8
4.4
     The house had  an apparent 133 in2  of  ELA and 130 cfm of air
available to it  at  1 pascal pressure differential in the assumed
pre-existing condition (before the  radon mitigation  system was
installed).  Operation of  the radon fan reduces the apparent ELA by
28  in2  (133-105)  and the  cfm available  at  1  pascal  pressure
differential by  50  cfm (130-80).   The  installation  of the radon
mitigation  system  reduced the availability of  household air and
thus has  the potential to contribute to the  backdrafting of the
furnace.

     A 4/2" PVC bushing adaptor was installed on the effluent end
of the 4" PVC to reduce effluent  flow (the fan was not changed).
The ELA and the  cfm available  at  1 pascal  were respectively, 127
in2 and  122  cfm,  after the 2" adaptor was installed.  Reduction of
the 4" ID PVC to  2"  ID PVC reduced  the dedicated ELA from 28  in2 to
6 in2  and  reduced the dedicated air available at 1 pascal pressure
differential from 50  to 8 cfm.
     The  dryer in  this  house  and  other houses  tested consumed
approximately  20 in2 of the total ELA.  It is noteworthy that the

-------
ACH50 fluctuated  little  in response to demands  on the household
air.    It  seems  that  "cfm available  at  1  pascal"  pressure
differential  is the  most sensitive  index to  use  to determine
adequacy of  air in the home.   The radon test data  are  shown in
Table 4.

Table 4.  Pre- and Postmitigation  radon data —  four vs  two inch
          PVC.
Location
Basement
family room
Basement
utility room
Main floor
family room
Main floor
master bedroom
Main floor
children's
bedroom
Pre-
mitigation,
pCi/L
4.5
5.2
4.8
5.8
4.5
Post-
mitigation,
pCi/L (4" PVC)
0.4
0.4
0.4
0.5
0.5
Post-
mitigation,
pCi/L (2" PVC)
< 0.3
< 0.4
< 0.3
< 0.3
< 0.3
     The 2" PVC was as effective in inhibiting radon entry as was
four inch PVC.  The 2" PVC is preferable to the 4" PVC because
the likelihood of radon mitigation system-induced backdrafting of
the natural draft appliances and heating/cooling penalty is less
than when a 4" PVC pipe is used.

Representative Case Study — Newton, Iowa

     The homeowner installed his own drain tile depressurization
(DTD) system.  The homeowner was aware that the clothes, dryer
caused the furnace to backdraft before the radon system was
installed.  The awareness of the backdrafting problem was
discovered when Mrs. Homeowner was nearly overcome by furnace
fumes while drying clothes.  The problem has been partially
solved by opening a window next to the clothes dryer when
operating the clothes dryer.

     The first 32 day screening test conducted in March 1989 gave
a radon concentration of 75.5 pCi/L.  Short-term tests conducted
in March 1990 and December 1990 showed radon concentrations of
17.3 and 23.8 pCi/L, respectively.  Postmitigation radon
concentrations were as follows (in pCi/L): crawl way, 0.6; family
room, < 0.5; living room, < 0.5; large bedroom, < 0.4; and < 0.5
in the small bedroom.

-------
     Mr. Homeowner installed a drain tile depressurization (DTD)
radon mitigation system and discovered that he had exacerbated
his existing backdrafting problems.  Mr. Homeowner found that it
was necessary to leave the window open all the time to avoid
backdrafting.  (The furnace, water heater, and dryer are in an 8
by 15 foot utility room.)  The homeowner has also noted that
weather conditions affect the backdrafting response to operation
of the radon mitigation system and the clothes dryer; under
certain weather conditions it is not necessary to have a window
open to prevent backdrafting.  Table 5 presents the air leakage
characteristics of the Newton home.

Table 5.  Air leakage characteristics — house in Newton.
House
condition
Radon fan
off
Radon fan
on (4"
PVC)
Radon fan
off (4"
PVC) +
dryer on
Radon fan
on (4"
PVC) +
dryer on
Radon fan
on (2"
PVC)
Leakage
area,
(in2)
82

55

47

34

63

cfm
available
at 1
pascal
67

27

22

14

35

"n" value
0.624

0.849

0.861

0.937

0.795

Leakage
ratio
(inVft2)
2.2

1.4

1.2

0.9

1.7

ACH50
3.3

3.2

2.8

2.3

3.4

     There is not adequate air leakage area in the home to
provide air for the clothes dryer, radon system, furnace, water
heater, nor probably adequate ventilation to provide "fresh" air
for the occupants.  The homeowner is looking for a permanent
solution to the backdrafting problems.  The homeowner wants a
solution that anyone can live with; they are concerned that
someone housesitting while they are on vacation may inadvertently
kill themselves by drying the wash in the clothes dryer!

Representative Case Study — Luana Iowa

     Two short-term tests conducted over a two-year period showed
a radon concentration of 9.4 and 9.6 pCi/L for the two respective
testing periods.

-------
     A one-hole SSD system was installed in this home.  Pre-
mitigation soil gas communication testing showed good soil gas
communication to all corners of the slab from the point chosen
for the SSD hole (under the flue chase).  Two inch ID PVC pipe
was used in part because of the relatively small air leakage of
the home and because a 2" PVC pipe would fit up the flue chase
alongside the 3" ID Class B metal flue for the natural draft
water heater.  A inline fan capable of moving 150 cfm of air at
zero inches static pressure was used.  The homeowners have a
sealed combustion furnace.  Table 6 shows the air leakage
characteristics of the home.

Table 6.  Air leakage characteristics — house in Luana.
House
condition
Radon fan
off
Radon fan
on (2M
PVC)
Radon fan
off +
dryer on
Radon fan
off +
kitchen
fan on
Radon fan
off +
kitchen
fan +
dryer on
Leakage
area,
(in2)
118

113

105


97



85




cfm
available
at 1
pascal
111

102

92


73



57




"n" value
0.557

0.577

0.593


0.655



0.704




Leakage
ratio
(in2/ ft2)
3.3

3.2

3.0


2.8



2.4




ACH50
4.5

4.5

4.3


4.4



4.1




     There were two problems with the installation:  (1) it was
discovered by the homeowner that the water heater would backdraft
when the kitchen exhaust fan was operated (if the door to the
basement was open) and  (2) the radon concentration did not go
below 2 pCi/L as promised in our contract.  Table 7  shows the
pre- and postmitigation radon testing.  We believe that the
backdrafting was a pre-existing condition discovered by our
backdraft indicator and the increased awareness we created.  It
is also clearly evident that the radon mitigation system
exacerbated (consumes 5 in2 of the leakage area) a likely pre-
existing condition.  The kitchen exhaust fan consumes 21 in2 of
the 118 in2 available leakage area in the house.

-------
Table 7.  Pre- and Postmitigation radon data — Luana home.
Location
Main floor
master bedroom
Main floor
northeast
bedroom
Main floor
living room
Basement north
side
Basement south
side
Pre-
mitigation,
pCi/L
4.6
5.3
not available
11.1
12.3
Post-
mitigation,
pCi/L (2" PVC)
1.4
1.6
1.4
3.9
4.5
     The radon concentration did not decrease to below 1 pCi/L as
we usually find  (the subslab communication is excellent to all
four corners).   The residual radon must be entering through the
block walls.  We see two choices as methods to reduce the radon
concentration to below  1 pCi/L:  (1) install wall suctions or  (2)
install a forced air ventilation (pressurization) system (FAVS).

     There are some additional considerations.  The children  in
the home have allergies and window condensation appears in cold
weather.  We think the  FAVS approach would be most appropriate
because it will  provide needed fresh air, should solve the radon
problem, and should prevent the natural draft water heater from
backdrafting.  The FAVS may obviate the need for the SSD system.

                            DISCUSSION

     Many homes  are time bombs in the  sense of having inadequate
ventilation; radon mitigation systems  shorten the fuse by
dedicating a specific amount of  a home's  leakage area to radon
mitigation.  The Spandex concept of  "one  size fits all" may work
for socks but does not  work when installing ASD and SMD radon
mitigation systems in houses.  There is too much variability  in
the housing stock to assume that a conventional ASD radon
mitigation system consisting of  four inch ID PVC and an inline
centrifugal  fan  capable of moving 270  cfm of air at zero inches
static pressure  can be  used safely in  all houses.

     Coyne1 states that "it is absolutely critical that anyone
seeking to repair ducts be aware of, and  consider, the
interactions among components of the house.  Combustion appliance
venting,  indoor  air quality, moisture, comfort, and occupant
behavior must all be  factored in."   "Proper training  is

-------
essential, and repair work must proceed with caution to avoid
'solutions' causing problems of their own."

     Coyne's recommendations apply directly to radon mitigation.
Radon mitigation is a science with too many unknown parameters.
We must as radon mitigators, appliance installers  (clothes dryers
and exhaust fans), heating contractors, plumbers, builders,
consultants, researchers, etc. treat the house as a "system" and
not as an entity with disconnected functional units.  The health
and safety of home dwellers are at stake.

     Coyne recommends that "pressure differentials and spillage
should be tested before and after sealing ducts. . .  to leave
the home safe and healthy."  Radon mitigators must do likewise.
The air leakage of a home should be measured before and after
installing a radon mitigation system.  The reduction in air
leakage area and cfm of air available at various pressure
differentials should be determined.  Research needs to be
conducted to develop standards for the amount (cfm) of air
requisite for proper operation of the air-consuming appliances
and requisite for a healthy occupant environment at different
pressure differentials.

     Some of the Environmental Protection Agency's Radon
Contractor's Proficiency training suggests using smoke with all
exhaust fans running to determine if proper drafting occurs.
Passive combustion air in the form of ductwork from the outdoors
to near the combustion appliances is suggested as a method of
providing air.  This training gives a false sense of security.

     The CMHC3 reports  that "the provision of  additional  supply
air is not likely to be effective as a remedy for pressure-
induced spillage of combustion products if the supply air is
introduced unaided through a building envelope opening of any
size likely to be considered practical."  Some homeowners
recognize that environmental conditions affect drafting.   A smoke
test at a point in time is not a reliable predictor of adequate
air for drafting of natural draft appliances.

                         RECOMMENDATIONS

     All radon mitigation systems installed in Iowa (and other
cold climates) homes should be "recalled" and evaluated;  i.e.,
air leakage area and cfm air available should be measured with a
blower door and a determination (best estimate)  made if the homes
have adequate air available for the air-consuming appliances.
Air leakage characteristics should be determined before and after
installing radon mitigation systems in cold climate homes.

-------
     The Canada Mortgage and Housing Corporation2  (CMHC)
suggests that "the most appropriate material for indicating a
temperature rise caused by spillage from a gas furnace was found
to be temperature sensitive labels or dots."  Backdraft
indicators should be installed on all natural draft gas furnaces
and water heaters.  Electronic carbon monoxide detectors should
be installed in "suspect" homes.

     Due to the complexity of installing radon mitigation
systems, do-it-yourselfers should not be encouraged.

     Relatively fool-proof procedures need to be developed to
allow a systematic approach to the installation of a safe radon
mitigation system in cold climates.

          The work described in this paper was not
          funded by the U.S. Environmental Protection
          Agency and therefore the contents do not
          necessarily reflect the views of the Agency
          and no official endorsement should be
          inferred.
                            REFERENCES
     Coyne, B. "A Million Miles of Ducts: Duct Sealing Update,"
     Home Energy. 9 (2):14-20 (1992).

     Residential Combustion Venting Failure - A Systems Approach
     Country-Wide Survey; Development and Testing of Spillage
     Detectors. Canada Mortgage and Housing Corporation, Jan.
     1987, p 11.

     Residential Combustion Venting Failure - A Systems Approach
     Summary Report. Canada Mortgage and Housing Corporation,
     July 1987, p 117.

-------
                                                                   VIIP-2
                     WHY WE LIKE DIAGNOSTICS

       By:  John W. Anderson,  Jr.  and  Jack  C.  Bartholomew,  Jr.
            Partners  in
            Quality Conservation
            E.  5805 Sharp,  #A8
            Spokane,  Washington 99212
                            ABSTRACT

We think building diagnostics are valuable .when bidding,
designing, and installing radon mitigation systems.

Diagnostics start with the homeowner interview and mapping the
plan of the house in contact with the soil, air flow study with a
smoke gun, pressure measurements, radon sniffing, and sub-slab
diagnostics, and end with system pressure documentation, testing,
and customer satisfaction.

These techniques enable us to design and install radon systems
that are not oversized and require less energy to operate, and
with less than a 1% call-back.

The work described in this paper was not funded by the U.S.
Environmental Protection Agency and therefore the contents do not
necessarily reflect the views of the Agency and no official
endorsement should be inferred.

-------
                     WHY WE LIKE DIAGNOSTICS


              RADON SYSTEM INSTALLATION DIAGNOSTICS


  Some people swear by  them.

  Some people swear at  them.

  Some people say what?

  And some do them only if  there are problems with an installed
system,  which is mostly the case since  90% of radon systems are
installed with no diagnostics before or after except for the
short-term radon tests.

  Some contractors have said  they only  have  to return to modify
10% of their systems without using diagnostics.   We like our
return rate of less than 1% with diagnostics  and it helps
customer satisfaction, referrals, and the, "You want it  done
when?",  time line.

  Diagnostics help design  and install  systems that will work  the
first time with few or no call backs.

This type of building investigation is  50% science and 50% art.
It is the sum of observation, measurement, and intuition which
comes from familiarity with diagnostics and radon system
installation performance.


                       DIAGNOSTICS BEGINS


  This process  begins with the homeowner  interview and  is
followed by a thorough  inspection of the house.

  We look for:

   1.   Type  of  furnace  system

   2.   Type  of  ventilation system

   3 .   Type  of  foundation

   4.    Slab  floor  characteristics

-------
   5.   Crawlspace access  (if it has one) and headroom

   6.   Attic access and headroom

   7.   Stack and mechanically induced pressure differences

   8.   Air movement using a smoke gun

   9.   Possible entry points by radon sniffing

   10.  A  route out of the building for the radon vent pipe

   This inspection takes about 1 to 1 1/2 hours.

   Diagnostics performed once the job begins are:

   1.   Sub-slab sniffing  through test holes drilled in the floor

   2.   Sub-slab communication testing

   These two procedures add about 1/2 to 1 hour to  the system
installation time and help determine radon sources under the
slab, and sub-slab pressure reach.   Most of these holes  should be
drilled to test the installed system performance anyway.


                   WHY DO WE LIKE DIAGNOSTICS?


A HOMEOWNER INTERVIEW


   Meeting the homeowner helps answer questions they may have
about radon and entry control,  uncovers concerns about radon  and
system routing,  may direct the investigation,  and establishes a
time-line for the system installation.

   Asking  for floor plans  and construction pictures (if
available), can help locate hidden utility chases for pipe
routing or attic bypasses, can indicate if there may be gravel
beneath the slab, and can speed the foundation plan measurement
process especially in complex houses.   Unfortunately,  very few
homeowners have these plans or pictures,  but it's worth it to
ask.

-------
A THOROUGH INSPECTION AND MAPPING OF THE HOUSE
  This  cannot be  done  sight-unseen.

  A  careful  eye and  knowledge  of  construction practices helps
locate possible entry points and pipe routes out of the building.
It has uncovered details that other contractors have missed such
as dirt floored areas under stairs and furnace return plenums
that pull air from under slabs.  Finding details that others miss
really helps in landing the job, and may uncover details that
make the mitigation system easier to install, less expensive for
the homeowner,  or may cause you to bid high and someone else gets
the job.


TYPE OF FURNACE SYSTEM


  Noting the type of furnace  system...gas,  oil,  electric,  solid
fuel; atmospheric burner, draft induced, or sealed combustion?

  Will  backdrafting  be a concern,  or  is  it  already a  concern?

  Is there  a need for  make-up  or  dedicated  combustion air.   What
about draft  inducing atmospheric gas burners for furnaces and
water heaters?

  Is the system ducted,  baseboard,  or radiant  heat in the  slab
floor or ceiling?

  How will  the heating system type effect  the  radon system
installation?
TYPE OF VENTILATION SYSTEM
   What  type of ventilation system is  there (bath fans,  down
draft range vent, exhaust only, or balanced)?

   How will  its operation effect  negative or positive pressures
in the building?  It is  important to test mechanically induced
pressure differences with major exhaust  fans on and  off (200
CFM+).

   Some  of  the newer /  tighter houses  can be depressurized to 50
pa (0.2" WC) with just 200 CFM of air exhausted from the
building.

-------
 TYPE  OF  FOUNDATION
   The type of foundation...concrete,  block,  rubble,  or wood?
What  is special about each of these foundation types and how will
it effect  the  installed system?

   How much of the foundation is accessible?
SLAB FLOOR CHARACTERISTICS
   What condition is the slab in?   Are there cracks,  holes,
carpet, or vinyl.   Is the cold joint between the slab and
foundation accessible?
CRAWLSPACE ACCESS
   Accessibility of crawlspaces  and use,  and how may the
mitigation strategy effect the house?
   1.    Pipes.   Are they insulated?  Can  they be freeze
             protected?
   2.    Ducts.   Are they sealed  and insulated?  Can they  be
             sealed and  insulated?
   3.    Furnace.   Provisions  are  needed for  access.
   4.    Storage.   If stuff is stored in this space,  then  a
             durable membrane is  needed.
   5.    Insulation.   Is  there any of this itchy stuff?  Is it  at
             the perimeter or under the floor?

   We  prefer  using sealed sub-membrane  ventilation  because it  has
a  lower energy penalty than sucking on the crawlspace  (especially
true if there are uninsulated heat ducts,pipes and floors),
however we have treated several inaccessible crawlspace houses
with unprotected pipes,  by pressurizing the crawlspaces with
house air.
ATTIC ACCESS
   Accessibility of attics.   You want  the  fan  where?   We  try  to
route pipes through closets to keep as much of the system as
possible in the heated envelope,  and place fans as near an attic
access as possible because there are electricians who are 6'5"
and 250 pounds who don't do well in attics.  Also, anything
mechanical will eventually need replacement and should be easy to
service.

-------
MEASUREMENT OF STACK AND MECHANICALLY INDUCED PRESSURE
  DIFFERENCES
   Do  these  forces  contribute  to  radon  entry  and  what  will  be
their effect on a mitigation system?

   What  effect  does  furnace  fan operation  have?   What  effect do
exhaust fans have?

   Are these significant  and need to be dealt with?

   Fifteen pascals  positive  has been measured under  slabs when
the wind blows on some houses.  We know of a 9 story building in
which expansion joints were sealed with gun grade urethane and
the material kept bubbling  from the slab joints.

   Does  a forced  air furnace create  negative  pressure  in  the
basement and how much?  Some house basements have been
depressurized by 10 pascals (O.OA inch WC) when the furnace fan
runs, which is enough to overwhelm most sub-slab suction systems.
Commercial  buildings can be much worse.

   The bottom line  is  these  pressure forces are the  major  reason
radon enters a building and can,  in some cases,  be great enough
to counteract the effects of a pressure field under a slab.  We
think they need to be measured so an effective sub-slab
ventilation system is installed the first time.   Well, 99% of the
time.
SMOKE GUN ANALYSIS
   Watching how air  behaves  with mechanical  systems on and  off
helps to tell which way air is flowing, and if pressure
differences are being created.  Does air flow from under the slab
into the basement through cracks and pilot holes?  What effect do
mechanical systems have on air flow and pressure differences?

   A small, teflon bottle  filled with  fiberglass  and 5 cc of
titanium tetrachloride, with a tight fitting cap can  be really
easy to carry in a pocket.
LOCATION OF POSSIBLE ENTRY POINTS
   Knowing what's hot and what's  not helps in directing the
mitigation approach.

-------
   In a crawlspace / basement combination house, sniffing can
 determine if  the  pollution  is  coming  from  the  crawlspace,  the
 basement, or  both.  We have  found  crawlspace / basement  houses
 where  the crawlspace  is not  the problem  and have  successfully
 lowered radon levels  by applying sub-slab  suction  to  the basement
 slab and tightening the wall between  the basement  and the
 crawlspace.   Conversely, we  have found crawlspace  / basement
 houses that only  needed the  crawlspace treated.  A stepped
 approach can  save the homeowner big bucks.

   In marginally polluted houses,  sniffing may indicate that the
 sealing of significant entry points may  control radon levels.
 This has been found with untrapped floor drains that  have geysers
 of air into the house 3 feet high and test with a  sniffer at 300
 pCi/1.

   Radon  sniffing can  confirm elevated radon  levels.


 SUB-SLAB COMMUNICATION TESTING (SLABS.,.ALWAYS)


   It is  valuable locating sumps,  number  needed, and  to tell if a
 sump is large  enough  (would a  larger sump, trench, another  hole,
 or stitching  work better?)  Sometimes the preferred location for
 a sump has no  reach or maybe under just  half the floor where an
 alternate location 10 feet away shows reach everywhere
 (difference in hole locations has been as small as 3  feet on the
 same side of  a footing).

   Sub-slab  communication  testing  is  simple to  do  during
 installation,  and maybe adds 1/2 to 1 hour to  the  installation
 time.  We call the suction point the main test hole (1 1/4" to
 fit the vacuum) and the test holes around the perimeter  (1/2"
 holes) the pilot  holes.  There is the initial  testing through the
 main test hole(s) to judge pressure reach and  sump location, and
 then several  tests of the sub-slab communication while digging
 the sump helps to see if it is large enough.   The pilot holes at
 the perimeter are needed to check performance once the system is
 installed anyway.

   Sub-slab  communication  testing can  help  size system fans. The
measurement of sump pressure 1 cubit from the main test hole (the
 distance from your elbow to your finger  tips  or as far as can be
easily dug out) shows sump pressure needed to get the same
 results at the pilot holes.  Measurement of air flow  shows how
much air needs to be moved.  The combination  of sump pressure and
CFM indicates what size fan is needed (F-100,  F-150,   F-160,
F-175,  etc.) once the static losses in the pipe are accounted
 for.

-------
  Even  in houses  that  for  some  reason  didn't  let us  drill all
the holes wanted,  and didn't submit to good communication,  the
result is large sumps that obtain a lot more reach than if the
system were installed by eyeballing it and using some rules-of-
thumb for pit size.  Uh...how's that look George... George says it
looks good so it must be good.   Yup... George is my friend...!
will follow him anywhere...
SUB-SLAB SNIFFING


   Sniffing  is  often  a  useful  diagnostic procedure  for  locating
sources under the floor and directing the focus of the sub-slab
ventilation system toward the areas of highest radon
concentrations.  Sub-slab sniffing must be done before
communication tests to sample the soil "as is" because once
negative pressure has been applied to the soil under the slab,
the distribution of radon changes.

   Here's  a  question  as  in  a  familiar  test.

   Sub-slab  sniffs  of 150  to  300  pCi/1  everywhere  except  for  a
corner at 6000 pCi/1.

   Where would  the  system  pressure field most  definitely  have to
extend to?
        1.    The  fireplace
        2.    Next to  the floor drain
        3.    The  corner  with  6000 pCi/1
        4.    The  clothes hamper

   There is  no  clue if  sources haven't  been  sniffed out.


WE RARELY USE OUR DOOR FAN IN RADON WORK
   It  is  helpful  to  have  the  ability  to  test  building  tightness.
A door fan test can asses the viability of ventilation strategies
and how much air will be needed, and can locate bypasses (ie
mechanical chases that contribute to stack effect.

   We  have  used a  door fan  to simulate winter stack effect  on  a
warm summer day to help determine entry points and sources.

   We  proposed  taking  care  of radon entry  in  an older
inaccessible crawlspace house by pressurizing the crawlspace with
house air.  We didn't feel easy about exhausting air from the
crawlspace because of unprotected water pipes.  A door fan test
of the crawlspace told how much air was needed to dump into the

-------
crawlspace to pressurize it to 5 pascals, and helped locate air
leaks that could be sealed to reduce the amount of air needed
from the house.  The house was then door fan tested to see what
effect 150 cubic feet of air per minute exhausted from the house
would have.  The effect was found to be about the same as "C" or
the flow through the house at 1 pascal pressure (in this case "C'
was 150 CFM).   Through two severe winters the occupants have
remarked how much more comfortable their home is.

   The  door fan has  been used  to  test  the pipe  tunnels  under  a
school for the amount of air needed to be exhausted to create a
negative 25 pascals in the tunnels.  The radon mitigation system
fan was sized with the help of this information.  Asbestos was
not a problem but definitely would be if present.


THE "GEE WHIZ" ASPECT OF DIAGNOSTIC EQUIPMENT


   It  can  help  sell  the job.   What,  more  flashing  lights  and
buzzers?  Neat!  It shows you care enough to analyze radon entry
and behavior to assure control of radon  levels.
                           CASE  STUDIES
A HOUSE

   The  house  has  an  indoor  lap  pool  (12 X  76),  basement,  and  two
crawlspaces.   When we were first called by the homeowner, the
concrete deck around the pool had not been poured.

   Upstairs radon was measured  at  24  and 31 pCi/1.   The pool  room
was 193 pCi/1.

   Four inch  ADS  perf was  looped around the pool  and stubbed  up
for an easy exit from the house.   The pool deck was poured, slab
joints were sealed, and the house was retested in seven places.


   Of  the  seven  test locations  the 2  crawlspaces  had the  lowest
readings of 0.7 and 8.0 pCi/1.  Basement readings were 19, 27,
and 38 pCi/1, and upstairs were 9 and 12 pCi/1.

   The  beauty of  this installation was  that the crawlspaces were
investigated to see if they were a source, and were not mitigated
because they did not have to be.   Que the happy homeowner and
another good reference!

-------
ENGLAND


  We were  asked by  a  company  in  England  to  travel  across  the big
pond and visit their country for three weeks in May '91 to teach
them how to do diagnostics and install radon systems.

  Familiarity with  diagnostics  really saved our  cookies or  is  it
crumpets?  There we were.  The same language was spoken,  but
somehow it was different.  Cars were driving by on the left side
of the road, but there were no drivers, just passengers.   And the
people we met were very courteous.

  They usually build  their homes with masonry  exterior and
interior walls,  and each wall has its own foundation and footing.
Each room is effectively cordoned off from  the others by
foundations.  Five homes were mitigated.

  Two  were a  combination of basement, suspended  timber floors,
and slabs on grade, two were slabs on grade, and one was a
suspended timber floor with and  a filled-in basement combination.
Ages of the houses was 25  to 300 years.

  Sub-slab and  crawlspace communication testing  proved helpful
in locating suction points.  A combination  basement, timber
floor, and basement house  was mitigated with a central suction
point  that accessed four rooms.  This  location was found through
sub-slab communication testing that required 4 different main
test holes.

   Sub-slab communication testing helped judge  pressure field
extension and locate  cold  joint  leaks.   Cold joint short circuits
in the 2 slab on grade houses needed  to  be  sealed  before
communication was achieved.  The sub-slab communication testing
helped que us to the  slab/foundation  cold joint.   One  house was
mitigated with one  centrally  located  sump and  the  other was fixed
with 2 sumps, one at  each  end of the  house.

   A radon system was installed in a house with a basement floor,
two slabs  on  grade, and  a  suspended timber  floor.  Monitors were
left and results were obtained a few  days later.   The  basement
was still  hot.  Sniffing found  sources  at the  basement foundation
walls.
 SCHOOLS

 Stevens

   We described Stevens Elementary School in our bid to the
 school district  as  a  mitigator's  nightmare... It  was.

-------
   This  building  has  a  footprint  of  about  12,000  square  feet, was
built around the turn of the century with rubble foundation
walls, and half the basement was dirt floored.  Finished basement
areas were used as classrooms.

   Initial  sniffing for radon  sources  showed  the  foundation walls
not to be a source, except in room 11.  A couple of pipe tunnels
were sources, and so were the dirt floors.

   Room  11  was  door fan tested to see  what pressures  and CFMs
were needed for flow reversal through the cracks in the
foundation walls.  We then pressurized room 11 to 50 pascals
which helped us locate leaks in the rubble foundation and seal
them to limit radon entry.

   The dirt floored areas  were prepared  for sub-slab  ventilation
with 4 inch ADS perf and then capped with 4 inches of concrete.
Have you ever tried moving concrete trucks through a playground
full of kids?  Score that day...Cement Trucks 0, Little Kids 6.

The sub-slab ventilation systems were completed and the accesses
to the pipe tunnels were sealed.

   All rooms in the basement were then tested.   Radon had been
controlled in all rooms except for 2 adjoining rooms, numbers 10
and 12 which were now between 4 and 16 pCi/1.

   More  diagnostics!  Where was  the  source?   Now with radon
levels much lower throughout the basement, the foundation walls
in these 2 rooms showed up hot.

   The sub-slab ventilation system for these  2 rooms  was reversed
to pressurize under the floor and reduced radon levels below 4
pCi/1, but it also pumped a lot of moisture into the rooms.  The
system was again modified to depressurize under the slab
(original configuration).

   The 2 rooms  were door fan tested  to check  tightness,  entry
point flow reversal pressure,  and cubic feet of air per minute
needed to do it.  The door fan testing also revealed a couple of
mechanical chases that led to the attic.

   The rooms were tightened to reduce  the  amount of air  needed  to
be vented into them to reverse entry point flow.  The exterior
foundation walls were not suited for crack sealing because of
crumbling mortar.

   A 300 CFM fan was  mounted in the  transom over the  door to
pressurize rooms 10 and 12 with air from  the hallway.

   Subsequent testing has  shown levels stay down as long as  the

-------
door to the rooms is closed at night, indicating that 8 to 12
hours of room pressurization a day is enough to help control
radon entry through the foundation walls in these 2 rooms.

Clover Park

   One western  Washington  school  district  tested all  their
buildings and discovered most rooms OK with just a few of them
coming back over A pCi/1.   Upon inspection we found a close
association with inadequate fresh air and radon.  We checked the
damper operation of the unit ventilators and found most not
working, and in those that were working we found blocked air
intakes.  Carbon dioxide was also found to be around 2000 ppm.
Increasing the amount of outside air delivered to the rooms would
have the double impact of reducing radon levels and providing
enough fresh air to keep the students and teachers awake for a
better learning environment.

Greenacres

   We diagnosed and  prepared a  mitigation  plan  for  a  slab-on-
grade elementary school built in 1979.  Politics of the situation
were that, if during Christmas vacation, we were not able to
lower radon levels substantially, the school would have to be
closed.

   There  were 5 HVAC systems with return ductwork under  the  slab
and radon test results from each room showed a strong correlation
to each HVAC system.

   Radon  sniffing was  done in the return ducts  at each  air
handler.  Radon sniffs of the main HVAC system showed the highest
levels and the area served by this system tested the highest in
the building.   The return ductwork had to go.

   Could  the sub-slab  return ductwork be used as a  sub-slab
ventilation network to control radon entry?

   The next step involved  drilling  15 test  holes through the  slab
to test pressure field extension from the return ductwork.  The
return grills were plugged, and the return fan was carefully
brought up to 1.5 inches WC hoping the ductwork would not
collapse.   7000 CFM was measured being drawn through the return
air ductwork.   Most of the test holes registered between -2.0 and
-30.0 pascals.

   7000  total CFM  (measured)
   1200  floor grill  leakage  (measured)
   -700  above grade  return duct leakage  (assumed)
   5100  total CFM needed to  create  same  pressure field

-------
   Since  5100 CFM produced a  large and consistent pressure change
under the slab, we figured one half that amount would still
control radon entry.

   There  was a  plan!  The return  ductwork under  the slab would be
abandoned, modified to be used as the subslab ventilation system,
and return ductwork installed in the ceiling of the building.

-------
                                                                  VIIP-3
        An Approach To Computer-Assisted Radon Mitigation

                               by

             Hormoz  Zarefar,  Pah Chen,  Patricia  Byrne

                    Portland State University
                Mechanical Engineering Department
                           P.O. Box  751
                        Portland, OR 97207

                               and

                         Charles Eastwood

                 Bonneville Power Administration
                        P.O.  Box 3621-RMRD
                        Portland, OR 97208



                             ABSTRACT
     A computer aided  approach  to  radon mitigation is discussed.
The proposed methodology involves the development of a knowledge-
based system as well as procedural computational tools to aid the
mitigators in the selection  and design of mitigation systems for
existing homes.   The computer  software  (RnX)  requests pertinent
data  about the  house, processes  this data,  and  recommends  an
appropriate mitigation method.   For sub-slab suction techniques,
the system assists  in determining  the number  of  needed suction
points and performs a  fan selection  and cost estimation.   The
development of this prototype software involved the incorporation
of  knowledge  in  existing literature,  communication  with  field
experts,  and the  creation  of rules for representing, converting,
and applying data in order to provide recommendations to the end
user.

     The software offers many tools to facilitate various aspects
of  radon  mitigation  such as  record  keeping,  compliance  with
established  EPA  protocol,   consistent  fan   sizing  and  cost
estimation.

-------
                           INTRODUCTION


     Radon mitigation can be considered a multi-faceted, knowledge-
intensive problem solving process.  Generally, the mitigators rely
on their  prior on-the-job experience, rules of  thumb and common
sense to design and install a mitigation system.   At the same time,
they must follow current EPA guidelines on the design of mitigation
systems for existing homes.

     The  current  project  formalizes  some  of  the  aspects  of
mitigation method selection as well as provides the mitigators with
the methodology used to  select  the  required mitigation method,
sizes the fan  (if  needed)  and  provides a detailed cost estimate.
This paper  describes the  development, goals,   significance,  and
characteristics  of the prototype  hybrid  knowledge-based expert
system  for indoor radon  mitigation.   Some  features   of  user-
interaction and the capabilities of the system including the use of
the system as a potential tool for training novice mitigators are
discussed.

KNOWLEDGE-BASED  (EXPERT)  SYSTEMS

     Knowledge-based  (expert)  systems are computer programs which
result  from  incorporating  and  codifying heuristic  and expert
knowledge.  Knowledge-based systems applications  have been reported
in various  fields  including  diagnostics, monitoring,  planning,
trouble shooting, and design.   There have been several attempts in
the application of knowledge-based systems to building design (1,
2), and efforts in applying expert system technology to assist in
radon mitigation have also been reported.  An initial attempt was
made by Mosley  in 1987  (3) ,  and  a  demonstration  system on  a
Macintosh computer was developed  by  Brambley  in  1990   (4).   In
addition, an interactive system was developed for a" Macintosh by
Brennan  in  1990 (5).  The demonstration  system illustrated  the
usefulness of user-directed point-to-point hypertext when working
with large amounts of textual information.  The interactive system
was designed to assist in the training of mitigation contractors.
These ventures  demonstrated  the  capability of  expert systems in
dealing with radon mitigation.

     A knowledge-based expert system is typically comprised of four
main components;  an  inference engine,  knowledge base,  working
knowledge, and user interface.  The relation between these elements
is shown in Figure 1.  The  inference engine implements the problem

-------
solving  strategy.    The  knowledge base  contains  the  knowledge
pertaining to the solution of the problem.  The working  knowledge
represents data relevant to the problem  at  hand.   This data is
extracted  from  the  user through  the  user  interface,  which may
utilize a variety of techniques for querying the user (6).
    User
  Interface
Inference
 Engine
                                             Knowledge
                                                 Base
                                              Working
                                             Knowledge
       Figure 1.  Major components  of  an expert system.

     The heuristic knowledge that characterizes the field of  radon
mitigation  is  generally  a  surface  (experiential)   knowledge.
Surface knowledge is  heuristic, experience-based information that
is the result of successfully solving a large number  of  similar
problems.  On the  other  hand,  there exists a body of  knowledge
which are well defined and are based on scientific investigations
and mathematical and physical  sciences.  They embody the segment of
the radon  mitigation  task  which includes  ducting analysis,  fan
sizing,  and  costing.   These are deep knowledge  which, if well
defined  and  formalized,  can  be programmed  via  algorithmic or
procedural  routines   (7) .    We  define  a   hybrid  system as  an
integrated  computer   program   containing  knowledge-based   and
algorithmic subprograms (8).

     Expert systems  are  suitable programming environments  when
experts are available in the subject, but are in short supply, thus
dissemination of knowledge becomes a crucial task.   The knowledge
base must be fairly narrow and should not vary  (fluctuate) . Expert
systems are also suitable where the task takes  a significant amount
to do manually and  there  are  many factors involved in  a decision;
a poor decision  will  make a significant difference in  the outcome
when the competitors are performing the job consistently.   These
classes of programs can handle incomplete and/or uncertain data by
making reasonable assumptions (similar  to an expert) and proceed to
come up with an  answer.

-------
     There are several  advantages associated with the knowledge-
based (expert)  systems.  Knowledge-based systems allow experts to
dedicate more  time  to tasks  that  require human  ingenuity  and
creativity.  The programs can be employed in training novices while
simultaneously  providing  a  collective  pool  for  sharing  and
propagating  knowledge.     Expert   systems  also   permit   the
standardization of processes and techniques.

             RNX - A RADON MITIGATION EXPERT SYSTEM
     Considering the potentials of knowledge-based development and
the issues of  concern  in  radon mitigation,  a hybrid knowledge-
based/algorithmic  approach was selected.   The computer program,
which is  regarded  as  an  advisory system, is  a  modular program
composed of several segments.  A modular structure was adopted to
limit the size  of the separate  components, and to  facilitate error
detection and correction,  and  potential expansion of the system.
The modules, which  include a  house  investigation  summary,  fan
selection, and  cost estimation,  are activated  hierarchically, as
shown in Figure 2.
House investigation
     information
                               HOUSE INVESTIGATION
                                  SUMMARY MODULE
    Ducting information.
      Number of points
                    H
                                           i
FAN SELECTION
                                           i

                                    COST ANALYSIS
                                                    I
       Figure 2.  A representation of the system modules.

     The knowledge-base was constructed by employing a  rule-based
architecture,  in which the knowledge is stored as a series of IF-
THEN statements.   This  particular architecture  was deemed most

-------
appropriate after careful observation of structure of information
presented in  the  EPA publications and analyzing  the sequence of
events and decisions which  lead  to  the  selection  of a particular
radon mitigation scheme.

     The computer program  was developed using  the  expert system
development  environment LEVELS/OBJECT  running under  Microsoft
Windows Graphical  User Interface.   This particular combination
enhances the presentation of information to the  end user and makes
it  possible   to   incorporate  graphical  as   well   as  textual
information.  Each module is composed of its own demons, methods,
classes, attributes,  and instances (9).  Demons and methods are the
IF-THEN rules of the system.   Demons are  the heuristic rules and
are different from methods.   Methods  are purely  procedural, like a
simple computer code.   Variable names and their  related properties
are grouped  in  classes and attributes.   A class  encompasses the
attributes and  instances,  which  are a group  of related objects.
For example,  the  class "House" may  include  an  attribute  such as
"foundation  type."    An   instance   of  foundation  type  may  be
"crawlspace."  So, to specify the foundation type of the house in
an object oriented  fashion,  it would be stated as:   "foundation
type OF house IS crawlspace."

     Some of the modules use the  same classes and attributes.  The
values of these attributes are obtained from the  user by  one of the
modules, and  passed  between the  modules  during  execution.   The
method used to pass the  data was to write the  information to a text
file which in turn is accessed by a subsequent module.

     The hybrid  nature of  the entire system  is  evident  in its
integration of  heuristic and procedural methods.   The heuristic
portions of the rule base pertain mostly to the mitigation method
selection.    The   procedural  methods  are  exclusively  for  the
numerical  computations  associated   with  ducting  analysis,  fan
selection, and cost estimation.

PROGRAM STRUCTURE

     RnX begins by performing data acquisition.  The first module,
which  is  the  House  Investigation  Summary  module,   extracts the
characteristics of  the house.   This module  was  designed  to be
analogous to  the house investigation summaries  that mitigators
often use to  gather pertinent  data  about  a  particular house  (10,
11, 12).  The user is  prompted to either  create a new file or to
edit an existing  file.   This user-named file contains  the input
data.   If the user  has opted to edit an existing file, the data
from that file is loaded into the system and all previous entries
are  displayed  and  may be  changed.   Once data  gathering  is
concluded, the  new  or  newly-edited file is  saved.   Finally, the
module  assesses the  input  data  and recommends  one  or  several
mitigation methods.   This   is  followed  by an explanation  of the

-------
 logic that was  followed in  determining the recommendation.    A
 sample explanation screen is shown in Figure 3.   If the recommended
 mitigation method(s)  is  or includes sub-slab suction,  the module
 proceeds  to  recommend the number  of  slab penetrations  for the
 ventilation system.   This number is based on data pertaining to the
 degree of  sub-slab  communication,  the area of  the slab,  and the
 presence of  footings that partition the  sub-slab area  and could
 interfere with the pressure field  extension.  At the conclusion  of
 this module,  the  user may choose to exit the  system,  or proceed
 with either a fan selection or a cost estimation.
                       RnX- House Investigation Summary
 File
 Explanation for the recommended mitigation method

 The highest measured radon concentration was specified as 40 pCi/L. The type of foundation
 that was specified included a crawlspace.  The highest measured radon level is less than or
 equal to 40 pCi/L and does not necessitate an active ventilation mitigation method, so
 passive (natural) ventilation of the crawlspace is recommended. Also, since there are not any
 utilities present in the crawlspace, there is little danger of freezing. The type of foundation
 specified included a basement and/or a slab. It has been specified that there are not exterior
 or interior footing drains present so draintile methods are not applicable. The sub slab
 communication was not very good, or was not available, but the foundation walls are mostly
 poured concrete, so block wall ventilation will not be applicable. Because of this, sub slab
 communication is recommended.
       Figure 3.  An example of the  explanation  facility.

     The Fan Selection module  performs a ducting analysis and fan
selection for sub-slab suction techniques.  This  module can analyze
ducting  systems with up to three branches.  For a great majority of
U.S.   houses,   three  branches  (or  three   suction   points)  are
sufficient for reducing the  radon  level.   At  the  time  of this
writing, the  module supported duct diameters  of  3,  4,  and 6 inches,
and modifications were being effected to  enable  it to also support
diameters of  2  and  5  inches.    This module  requests  the data
necessary to  perform the calculations for the ducting analysis and
subsequent fan  sizing.    The  user  is  presented  with  a  screen
displaying the  total air flow of  the system along with  the total
system  friction  loss  (Figure  4) .     In  addition  to  these  fan
requirements,  the  user  is   shown  the maximum  velocity that  is
achieved in the duct, and for multiple branch systems, the user is
also shown the  ratio of  the branch  friction losses.    The optimum
value  of the  ratio  should be around 1.0 to provide equal pressure

-------
drop to  the last pipe junction point.   Otherwise,  less flow from
the branch with the  higher pressure  drop will render that branch
ineffective in reducing the  radon level.  To achieve the ratio of
1.0, one can change the pipe  size, fitting type, or add pipe length
or a flow  damper to balance  the pipe network.
                         RnX- Fan Selection Module
 file
   An appropriate fan will meet the following flow and pressure requirements:

               95 cfm         0.332 WC   (    82 Pa)
     For the given diameter(s) and flow rate,
     the maximum velocity in the duct is:
               Ratio of pressure drops
611  fpm
            Two branch system, branch 1 to branch 2 : Not Applicable

           Three branch system, branch 1 to branch 2 : 1.08

   Three branch system, intermediate branch to branch 3 : 0.75
                   More on
                 pressure drop
                     ratio
        More on
       maximum
        velocity
 Next
screen
       Figure  4.   A screen from the  Fan Selection Module.

     A small database of brand name fans has been incorporated into
this module.   The system determines the particular models of the
brand name  fans that are capable of satisfying the fan requirements
that were presented  earlier.   These models,  with their respective
purchase costs  and  power consumption ratings,  are  displayed  in
order for the  user to select one  for use with the cost estimation.
At the conclusion of this module, a temporary file containing the
data concerning the  diameter,  length,  and  number of fittings used
in the  ducting system as  well as the  fan  data  is  written to the
disk.  This is considered a temporary file  because subsequent runs
of  the  module will  overwrite the  file.    Lastly,  the  user  may
proceed  with a cost  estimation.

-------
     There   are   currently  three  modules   that   perform  cost
estimations.  These  modules  are  customized  to the tasks required
for either  block  wall  suction,  sub-membrane suction,  or sub-slab
suction.  If the user is entering the cost estimation directly from
the fan selection process, the cost estimation reflects the tasks
that are needed for  sub-slab suction.  The tasks are itemized and
the user can enter the amount of  time each task will consume along
with the number of times that the task must be performed.  This is
used to determine the  number of  hours that  are spent  on the job.
The labor rate is entered  by the user.  The cost estimation also
covers material costs.   The module also determines  the material
costs for the ducting system by accessing the data file generated
by the  fan  selection module.  The  total  cost for labor  and the
total material cost  is considered to be a  one-time installation
cost.  The annual  cost to the home owner includes energy costs and
maintenance.  Energy  cost is determined by the power consumption of
the fan, regional cost per kilowatt hour,  and the number of hours
that the fan  is in operation (usually year-round).   The cost for
maintenance varies regionally.  The user  can  incorporate a discount
and/or an overhead  factor.   The  module is  equipped with  default
values for the various costs, and all of them may be changed.

     All of the modules  incorporate  some form of error checking.
The system  checks  for unusual entries and  for  conflicting data.
For example, if the user  specifies that the house  is 200 years old,
or that the ceiling of the basement is 88  feet tall,  these may be
considered unusual entries.  The  user is notified of the situation
and may either verify or correct the value.  If conflicting entries
occur, the  user must make a  correction or  the program will not
proceed.  In some  cases the user  may not have all of the data that
the system requests.   For example, sub-slab  communication testing
may not have been performed for some reason or another.  If that is
the case, the user  selects  the  response  "none observable  or no
testing performed."  By  assuming  the worst  case,  the  system will
produce a recommended solution that is conservative.

     Once the House Investigation Summary module has been executed
for  a particular  case,  subsequent modules can  be  repeatedly
executed without invoking the previous module.   This  arrangement
makes it possible  for the user to  reactivate individual modules as
many times as is desired.  An  additional and completely independent
module is the Fan Tutorial, which is similar to the Fan Selection
module.  This module provides on-line assistance  and  defines the
notation that is used for the Fan Selection  module.

     RnX recently has  been reviewed by  members of  the Northwest
chapter of  AARST.  After the review comments were  received,  the
system was  modified  to eliminate  difficulties  and  incorporate
features which had been overlooked during development.  In addition
to  debugging,  the  user-friendliness  and   the  error-detection
capability of the interface were increased.

-------
           ADVANTAGES  OF A  COMPUTER-ASSISTED APPROACH


     One of the most attractive  features of a computerized process
is the  capability to  create  and  store systematic records.   The
records are  easily updated  and have  other  potential  uses.   As
previously mentioned, RnX has the capability to create and maintain
a file which contains the house investigation summary data.  This
file is created in the  initial data  run,  and may be  updated by
employing the editing feature.

     Another  important  feature is  that computerized  assistance
lends consistency and reliability.  The mitigation method selection
process uses pertinent  information only. The recommended method is
followed by an explanation which furnishes the user with the logic
followed  in  determining  the  recommended  mitigation method.   An
explanation  facility reveals  that the solutions  are  deduced in a
systematic fashion, and shows  which data were  used to arrive at the
solution.   It may also  enhance the experts' own ability to suggest
consistent solution methods.    In addition,  significant  attempts
have been made to comply  with present EPA protocol.

     The benefits in terms of computational  efficiency are obvious.
The  numerical nature  of the fan  selection  and cost  estimation
modules are  especially suited to coding.  The additional features
of the  fan selection module  (velocity  and pressure  drop checks)
provide a pre-design refinement  capability.  The advantage here is
that the  necessary calculations  for  alternative designs  may be
carried out in rapid succession until a suitable design  is reached.
When entering the cost estimation directly from the fan selection
module, the  user  is  spared  the  task  of  recording  some  of the
material  cost data, since this is  transferred via the temporary
file that  is  created by  the fan selection module.

                         ADVANTAGES OF RNX


     As   the  preceding  discussions   illustrate,   the  proposed
knowledge-based  advisory system  (RnX)  can be beneficial  to the
mitigators in terms of record keeping, observance of  established
EPA protocol, consistent fan  selection, and cost estimation.

     The  ability  to  store  the interactive  house investigation
sessions  allows  the  mitigators  to keep,  update  and  store the
information  as  business records as  well as allowing for in-depth
analysis such as observing trends in radon concentration levels and
the ability  to  infer suitable mitigation methods for similar  data
files.  Furthermore, the advisory system has the potential to be
interfaced with a database.   This  would facilitate data management
and  would make it  possible for mitigators as  well as health and
regulatory agencies to keep a database of cases for a particular

-------
region.  The information contained in such a database may be of use
to planners and researchers involved in radon related studies.

     The advisory system provides a form of quality assurance for
mitigators  just entering  the  field.   The  advisory  system in its
final form will have incorporated the suggestions of  field experts,
and will reflect their collective opinion.  Therefore, the novice
user will have the assurance of  the expert  opinion,  and the client
will be reassured that the recommended mitigation method complies
with established protocol and professional practice.

     The Fan Selection module  alleviates the amount  of computation
that must  be  performed during  the design of a  ducting  system.
Since the module also informs  the user of potential  noise problems
and imbalances, it  encourages fine-tuning of the ducting system
before the actual  installation.  This may help to manage the amount
of  time  that is  spent  on  installation,  and  may  reduce  the
possibility  of the  need  for follow-up  corrections  to  ducting
systems.    Another  benefit of  the module is  that  it  provides
consistent  results  for similar cases  and will  prevent  possible
oversizing of the fan.  An oversized fan consumes more energy and
raises operating cost.

     The  Cost Estimation  module  is  an  efficient way   for  the
mitigator to provide a quick estimate to the client.  It will also
assure the client that the quoted price is justified.   The module
is  flexible enough  to  include  unforeseen costs  involved  in the
installation and maintenance of  a mitigation system.  When used in
conjunction with the Fan Selection module,  the  cost estimations for
several alternative designs may be presented to the client.  This
would  also be a  benefit  to the client  who  is  deciding  between
several possible ducting configurations.

     Aside from assisting  established professional mitigators, the
advisory system  illustrates the potential for  a knowledge-based
system  to  serve  as  an  interactive  training   tool  for  novice
mitigators.  A modified version of the  software can be developed
with the eventual  goal of training novice mitigators. A successful
training tool  should  have a  simple  and  highly visual method of
communicating  the knowledge to  the novice, and  a graphical  user
interface  such as the one employed  in  the advisory  system  is a
fitting  representation   of  this  technique.     An   additional
consideration is that the  effectiveness of  a computerized training
tool is reasonably dependent upon the availability of the computers
needed to  run  the software.   Current trends  in  the PC market are
evidence   that  advanced   technology   is   becoming  increasingly
available at a lower cost.   In addition to  analyzing single family
residences, the present system  could be  expanded to  incorporate
other building types such as schools and commercial offices.

-------
                             SUMMARY
     The feasibility of applying expert systems methodology to the
problem of  indoor  radon mitigation has been  illustrated  by this
work.  The prototype RnX hybrid advisory system addresses various
facets of the  radon mitigation problem,  from the  selection of a
mitigation  method   to  the  determination  of  necessary  building
materials and  cost estimation.   Efficient modification  and the
implementation of experts' opinions is facilitated by the modular
structure.   To date,  the  system  has received  favorable  reviews
after demonstrations at several conferences and meetings.   At the
time of this writing,  the remaining tasks for  the completion of
project includes the addition of recommendations  submitted by radon
mitigation experts  who  have been contracted to  review the software.


     The work  described  in this paper was  not funded by the U.S.
Environmental  Protection Agency and therefore the contents do not
necessarily  reflect the views of the  Agency  and no  official
endorsement  should be  inferred.

                         ACKNOWLEDGEMENTS


     The authors acknowledge  the  Bonneville  Power Administration
for  its funding of this work under research contract IAG-07967.

     We wish to thank the members of the  Northwest Chapter of AARST
for  their review of RnX.

-------
                            REFERENCES


1.   Hitchcock, R.J.  "Knowledge-Based System Design Guide Tools."
          To be included in ASHRAE Transactions, V. 97  Pt. 2, 1991.


2.   Mayer, R., Degelman, L.O.,  Su, C.J.,  Keen, A., Griffith, P.,
          Huang,  J.,  Brown,  D.,  Kim,   Y.S.    "A Knowledge-Aided
          Design  System For Energy-Efficient Buildings."   To be
          included in ASHRAE Transactions, V.97, Pt. 2, 1991.

3.   Mosley, R.B.  Personal communication, 1990.

4.   Brambley, M.R., Hanlon, R.L., Parker, G.B.  "Expert Systems:
          A New Approach To Radon Mitigation Training And Quality
          Assurance." Proceedings, Indoor Air Conference, Toronto,
          Canada, August 1990, pp 483-487.

5.   Brennan,  T. ,   and  Gillette,  L.M.      "Interactive   House
          Investigation and Radon  Diagnostics  Computer Program."
          Proceedings of the 1990  International  Symposium on Radon
          and Radon Reduction Technology,  January 1990.

6.   Luger, George  F. ,  and Stubblefield,  William  A.   Artificial
          Intelligence  and  the Design  of  Expert Systems.   The
          Benjamin/Cummings Publishing Company, Inc.,  1989.

7.   Dym,  Clive  L.,  and  Levitt,  Raymond.  E.    Knowledge-based
          Systems in Engineering.   McGraw-Hill, Inc, 1991.

8.   Zarefar, H.  "An Approach to Mechanical Design Using a Network
          of  Interactive  Hybrid Expert Systems."   Ph.D.  Thesis,
          University of Texas at Arlington, 1986.

9.   LEVELS OBJECT User's Guide and Reference Guide.  Information
          Builders, Inc., 1990.

10.  Mosley,  R.B.,   and Henschel,  D.B.    Application of  Radon
          Reduction  Methods.    EPA/625/5-88/024,  Air  and  Energy
          Engineering Research Laboratory,  Research Triangle Park,
          NC, August 1988.

11.  Brennan, T., and Galbraith,  S.  Practical Radon  Control for
          Homes.  Cutter Information Corp., Arlington, MA,  1988.

12.  Henschel,  D.B.     Radon  Reduction Techniques for  Detached
          Houses, Technical Guidance (Second Edition).  U.S. EPA,
          EPA/625/5-87/019, Research  Triangle  Park,  NC,  January
          1988.

-------
                                                                                             VIIP-4
                        RADON CONTROL: FIELD DEMONSTRATIONS

                     DIAGNOSTIC AND MITIGATION TECHNIQUES USED
                         IN TWENTY-SIX RADON FIELD WORKSHOPS

                     By:    Craig E. Kneeland and Mark R. Watson,
                            New York State Energy Office
                            Two Rockefeller Plaza
                            Albany, NY 12223

                            Wade Evans
                            Evanshire Company, Ltd.
                            Rome, NY  13440

                            Terry Brennan
                            Camroden Associates
                            Oriskany, NY 13424


                                         ABSTRACT

       The New York State Energy Office conducts two-day radon field workshops as follow-up training
for attendees of its three-day "Reducing Indoor Radon" workshop.  These workshops also meet the
training requirements for the EPA's Radon Contractor Proficiency Program.  Field workshop sites are
single family homes whose owners  have participated  in an energy conservation program.  During the
workshops students perform a variety of diagnostic tests, and design and install mitigation systems. To
date, twenty-six workshops have been conducted.  Pre-mitigation test results from charcoal canisters
placed in the basements of the houses used in this program range from 20 to 522 pCi/l, with an average
of 101  pCi/l.   Data  collected  during radon diagnostics and  mitigation installation include: house
construction features, sub-slab radon concentrations; soil communication tests; and basement ambient
air radon concentrations before, during and  after mitigation. The mitigation techniques used in these
houses include  sub-slab  depressurization  (23  of the  26  systems  installed),  sub-membrane
depressurization  (in  three  houses,  in  conjunction   with  sub-slab  depressurization),  sub-slab
pressurization/dilution  (3 of the systems) and various sealing measures (all of the systems).  Post-
mitigation radon measurements are made using a charcoal canister (2-5 days), continuous radon monitor
(7 days) and an alpha track detector (one year). The average post-mitigation radon measurements were
1.6 pCi/l for charcoal canisters (CC) and 1.4 pCi/l with a continuous radon monitor (CRM).  Alpha track
detector results for the houses with depressurization systems average 0.6 pCi/l. Two of the three houses
with sub-slab pressurization/ dilution systems have been problematic.  Pre-mitigation (CC) measurements
for these  houses  were  163  pCi/l and 298 pCi/l.   Short-term post-mitigation  measurements were
satisfactory: in one house the readings were  3.3 pCi/l  (CRM) and 1.8 pCi/l (CC); in the other house the
readings were 1.5 pCi/l (CRM) and 0.8 pCi/l (CC).  Alpha track detector  measurements, however, were
19.0 pCi/l and 22.2 pCi/l.  A case study of the workshop and post-mitigation activities in one of these
difficult to mitigate houses follows a discussion of test procedures and data collected in all workshop sites.

       The work described in this paper was not funded by the U.S. Environmental Protection Agency
and therefore the contents do not necessarily reflect the views of the Agency and no official endorsement
should be inferred.

-------
                                       BACKGROUND

       The New York State Energy Office (NYSEO) has been involved in radon training and education
since 1983.  At that time radon was a major component in the indoor air quality unit of the agency's
•Better Builders' workshops.  In 1986 NYSEO received an EPA grant to develop the original "Reducing
Radon in Structures' training manual and three-day radon workshop. Then, in 1987, NYSEO was given
a mandate by the New York State Legislature to provide training for radon assessment specialists (those
who perform radon diagnostics and design radon mitigation systems), utilizing Petroleum Overcharge
Restitutionary (POOR) funds. NYSEO was chosen for these tasks because of its extensive experience with
radon and other training programs for builders, developers, heating contractors, architects, engineers and
others.

       With POOR monies, the NYSEO completely revised the original radon training manual and slides,
and developed a videotape on radon diagnostics and mitigation. In addition to providing radon-related
instruction, these workshops supplied a vehicle for addressing the perceived connection between tight"
houses and high radon levels.

                                     RADON PROGRAM

       The foundation of the NYSEO radon  training program is the revised  and updated three-day
workshop entitled 'Reducing Indoor Radon'. The training program also includes two one-day workshop
series which focus on radon-resistant new construction and radon and real estate issues, and a two-day
hands-on mitigation field workshop.

       The radon field workshops are conducted as follow-up training for attendees of the three-day
workshop.  The goal of the field workshops is to reinforce the diagnostic and  mitigation information
presented during the classroom instruction of the three-day course.  This is accomplished by providing
hands-on training in radon diagnostics and mitigation system design and installation.

       NYSEO field workshops are conducted in single family homes whose owners  have participated
in an energy conservation program. The training is conducted by personnel with extensive experience
in radon services, including testing, diagnostics and mitigation.  NYSEO began offering these training
sessions in January, 1989; in January of 1991 the EPA approved these workshops for the continuing
education requirement of the Radon Contractor Proficiency Program (RCPP).

                                RADON FIELD WORKSHOPS

                                 PRE-WORKSHOP ACTIVITY

       The night before each training session begins a continuous radon monitor (CRM) is set up in the
house to take hourly measurements.  The CRM is used to record radon levels immediately prior to
mitigation, monitor students'  exposure during the workshop, and  measure the effectiveness  of the
mitigation system after it has been installed.

DAY ONE

       The goal for the students on the first day of this course is to design an appropriate and effective
mitigation system.  To this end, the first part  of the morning of day one is spent completing  a house
investigation summary form. This form is used  to guide students through the appropriate steps for a
thorough radon assessment. Students are directed toward identifying potential entry routes, thermal by-
passes, and sources of negative pressure, while paying close attention to house-specific construction

-------
features. After this careful visual inspection of the house, students conduct a homeowner interview in
order to record the testing history of the house, its unobservable construction features and to determine
what the owners' future plans are for the house. This helps to determine the type of mitigation system
selected and,  if necessary,  the pipe routing and fan location.  The remainder of the  morning is spent
performing a battery of diagnostic tests to determine the feasibility of certain mitigation approaches.
These tests can include grab samples, sniffer measurements, sub-slab and block wall communication
tests and whole house and basement fan door tests.

       The results of the diagnostic survey are discussed by students in a two hour post-lunch meeting.
Students are left alone to  develop a detailed mitigation plan which is expected to include a cost estimate
for labor and materials needed to complete the installation. Students are also expected to list appropriate
post-mitigation tests which can be conducted both during and after the workshop.  The plans and
specifications  are reviewed by instructors.  Differences between students' and  instructors'  plans are
discussed to ensure that students understand all the factors that experienced contractors consider when
designing effective mitigation systems.  If time allows and the mitigation strategy warrants it, students
return to the house to begin the installation of the mitigation system.

DAY TWO

       The goal of day two is to complete the installation of the selected mitigation system in accordance
with EPA Mitigation Guidelines.  When this is finished, tests are conducted to measure the effectiveness
of this system. These post-mitigation tests include sub-slab/block wall communication, and air flow and
radon measurements in the ventilation stack. To complete the installation, all components of the system,
including monitoring devices, are labeled. The operating principles of the mitigation and monitoring
systems are explained to the homeowner, and key components of  the installation are identified.

        Post-mitigation measurements are used to determine the success of the installed system.  A short-
term charcoal canister and a long-term alpha track detector are left with the homeowner to set out 1-2
days after the workshop ends. The CRM is also left on site for a week after the workshop.  Before leaving
the site, the class restores the house to its 'as found1 condition.

        Along with their letters of attendance, students receive the  post-mitigation test results from both
the CRM and charcoal canister. SEO staff record all testing, diagnostic and post-mitigation results, and
write detailed  reports for  each workshop.

        It is expected that students will leave this course with:

               a greater understanding of diagnostic procedures
               the  ability to select an effective, energy efficient mitigation  plan using diagnostic test
               results
               the ability to design a mitigation system which will reduce radon levels and fit in with the
               homeowner's present and future lifestyle
               specific information on appropriate materials, equipment and installation methods that
               assure high quality work
               an awareness of procedures available to maintain worker safety
               resources for answering questions that arise after the course.

WORKSHOP  RESULTS

        To date the  New York State Energy Office has conducted 26 radon field workshops locations
throughout the State. These training sessions, which are limited to ten students and one representative

-------
from the New York State Department of Health per house, were attended by a total of 249 people.  Eighty
of these attendees are Radon Contractor Proficiency Program (RCPP) listed for New York State. This
number represents 64% of the New York State RCPP contractors. Other trainees are providing services
in states throughout the nation, including Maine, New Hampshire, Vermont, Massachusetts, Rhode Island,
Connecticut, New Jersey, Pennsylvania, Ohio, Virginia, Iowa, and California

PRE-MITIGATION MEASUREMENTS

        There are a number of pre-mitigation charcoal canister results for twenty-five of the field workshop
houses (one house has only pre-mitigation ATD results). Using the highest basement measurement for
each house, the results range from 20.0 to 522.1 pCi/l, with an average reading of 101 pCi/l. These tests
were conducted by the homeowners, using detectors received through the New York State Department
of Health free detector program.  Most of the tests (81 %) were conducted during the heating season
(October to March), with the detectors exposed for three to four days.

       Pre-mitigation alpha track results are available for twenty-two of the twenty-six houses used in this
project. Eight of these testing devices were exposed in the basement, ten on the first floor, and four in
an unspecified location.  The duration of these tests ranges from two to twelve months, and all but two
results include at least two months' exposure during the heating season. These measurements range
from 4.0 to 222.7 pCi/l, averaging 40 pCi/l.

SITE SELECTION

       A number of factors were taken into  account when selecting houses for use in this project.  In
addition to elevated radon levels and participation in an energy conservation program, selection criteria
included house construction features and the two-day time frame. The first house chosen had poured
foundation walls, floor cracks, a french drain,  and an open sump pit.  As our experience with workshops
increased, we began to select houses with more features, such as crawlspaces and difficult pipe routes,
that may have needed to be addressed.

       Of the twenty-six houses used for this project, fifteen had block wall foundations and three more
had interior block walls.  The block tops were sealed in seven of the homes with block wall foundations
and one of the homes with an interior block  wall.  In the remaining houses, the tops of the walls were
inaccessible. Ten of the project sites had crawlspaces, seven with dirt or stone floors. There were eleven
houses with french drains, and nine with sump holes.

MITIGATION SYSTEMS

       Twenty-three  of  the houses mitigated have active sub-slab depressurization (ASD)  systems.
Among these houses, there are six one-point  systems,  twelve two-point systems, three three-point
systems, and two four-point systems. Active sub-membrane depressurization was used in conjunction
with ASD in two houses  having two-point systems and one house with a single point system.
The other three houses involved in this program have two-point active sub-slab pressurization systems.

       The average cost of materials only (labor is supplied by students) for these mitigation systems,
including electrical inspection, fire stop(s) and, in one case, a licensed electrician, was $625.  The cost
of the systems ranged from $389 to $949.

       It should be noted that the nature of this project sometimes dictated a more extensive mitigation
system than may have been necessary to reduce radon levels to or below 4 pCi/l.  At times this decision
was made for instructional purposes, since the primary goal of the workshops is to teach by providing

-------
hands-on experience.  This sometimes caused in-depth discussions with students about the necessity
of sealing floor cracks or block tops, for example. Another consideration was liability, which led to the
installation of new sump pumps in each house with an existing pump, and numerous self-sealing floor
drains. We also realized that it was less expensive to install what may have been extra materials during
the workshop, rather than returning later.

POST-MITIGATION RESULTS

       Extensive post-mitigation testing was conducted in every house used in this project. Continuous
radon monitor measurements are available for each house. These measurements, which were made in
the basement on an hourly basis, from the time the mitigation system fans were turned on and continued
for at least seven days, averaged 1.4 pCi/l.  The test results ranged from 0.4 pCi/l (in three houses) to 3.9
pCi/l. The distribution  (in pCi/l) was: 0.1 to 1.0 (11 houses), 1.1 to 2.0 (8 houses), 2.1 and 3.0 (4 houses),
and 3.1 to 3.9 (3 houses).  Each of the houses with CRM results over 3.0 pCi/l is being studied further.


       Post-mitigation charcoal canister  results, all from  basement readings,  are available  for 23
workshop sites. Of these results, 15 are below 1  pCi/l, 4 between 1 and 2 pCi/l, and the remaining four
are between 2 and 3 pCi/l. The range of results is from 0.2 to 2.9 pCi/l, with an average of 1.6 pCi/l.  In
all but one house the charcoal canister was exposed near the CRM in order to compare the two results.
In the houses for which we have data from both devices, the two measurements usually (81% of the time)
varied by less than one pCi/l.

       Currently, we  have post-mitigation alpha track detector results for seventeen houses.  The
homeowners were instructed to set the ATD's in the same location as their pre-mitigation alpha tracks (if
they had one).  The actual location of these long-term testing devices was  either in the basement (6
houses), first floor  (5 houses) or unspecified (6 houses).  The average of these results in houses with
active sub-slab depressurization systems is 0.6 pCi/l, with a range of 0.1 to 1.5 pCi/l. In each of these
houses the ATD results are less than or equal to the short term measurements.  We are satisfied that
these houses have been successfully mitigated.

       Of the three houses where sub-slab pressurization was used, however, the alpha track results are
acceptable in only one instance (1.5 pCi/l).  The  short term test  results  in all three houses were
satisfactory, but the long term results in two of them, 19.0 and 22.2 pCi/l, were not.  The remainder  of this
paper will discuss the work done in one of these houses.

                                        CASE STUDY

HOUSE DESCRIPTION

       Field workshop #13 was conducted in a four year old raised ranch located on a hilltop. The
house has a block wall foundation with a walk-out basement and an attached slab-on-grade garage. The
basement walls, framed out with 2 x 4's, are insulated and paneled. The basement floor is unfinished
concrete, except for vinyl tile in the bathroom. The house has approximately 3,000 square feet of surface
in contact with the soil (1,500 s.f. basement floor, 480 s.f. garage floor, 1,020 s.f. basement walls). The
house has its own well, cistern and septic system. According to the homeowners the well driller went
through 85 feet of gravel before hitting rock.

-------
 PRE-MITIGATION MEASUREMENTS

        This house was first tested for radon in August, 1988.  A charcoal canister in the basement from
 the 18th to the 22nd measured 49.4 pCi/l.  An alpha track detector left in the living room (first floor) from
 August 18 to November 19, 1988 measured 85.5 pCi/l.

        On March 7, 1989 NYS Department of Health Radiological Health Specialists visited this house
 to conduct on-site testing. Grab samples were taken in the ambient air and  at suspected radon entry
 points.  The  working level grab samples  in the basement air  (0.3 WL) and living room air (0.2 WL)
 compare with radon grab samples of 88.2 pCi/l (basement) and 61.2 (first floor) for an equilibrium ratio
 of approximately 0.3. The radon concentration at floor cracks in the basement ranged from 119.2 pCi/l
 to 208.7 pCi/l, while the sump hole measured 504.9 pCi/l.  The well water radon content was 343 pCi/l.

        After  the homeowner attempted mitigation by sealing the floor cracks and open sump pit, the
 house was retested. Charcoal canisters exposed in the basement and living room from January 10-12,
 1990, measured 298.3 pCi/l and 164.2 pCi/l. Further testing, conducted March 6-8,1990, measured 165.4
 pCi/l in the basement and 93.4 pCi/l in the  living room.  A water test (3/28/90) found 165 pCi/l in the well
 water.

 PRE-WORKSHOP ACTIVITY

        The night before the workshop started, a continuous radon monitor (Pylon AB-5 with a passive
 radon detector) was set up in the basement to print out hourly readings. The windows and doors, which
 had been open because of the warm weather,  were closed at the start of the monitoring (6:30 PM,
 8/15/90). When students and instructors arrived on site at approximately 8:15 the next morning the radon
 level, which had climbed steadily through the night, was 13.6 pCi/l. Before students began their radon
 assessment of the house the instructors opened the basement windows and set up a fan to blow fresh
 air into the basement.

 DIAGNOSTICS

        Students began the radon assessment process by conducting a visual inspection of the house.
 Exhaust appliances they found were an electric dryer vented to the outside and a wind-driven turbine on
 the roof.  Observed thermal by-passes were limited to kitchen soffits and recessed  lighting fixtures.
 Students noted that space heating is provided by electrical thermal storage units; the domestic hot water
 is also electrically heated.

        Potential  entry routes  discovered by students included  utility entrances and exits through the
 block walls, open block tops, plumbing penetrations through the slab, and hollow support columns. After
 the visual inspection the students conducted a homeowner interview.  The homeowner  verified the
 inspection results, provided photos of the house  under construction, explained his future plans to finish
 off the basement,  and gave the students the radon testing history of his house.

       After completing the visual inspection, students took sniffer measurements in the sub-slab soil and
 block walls. These measurements were made to locate areas with high radon concentrations.  Pilot holes
 (1/41 diameter) were drilled in five locations in the basement floor (see floor plan).  Sub-slab radon
 measurements at  FB (514.6 pCi/l), FD (489.6 pCi/l), FE (446.7 pCi/l), and FF (564.5 pCi/l) were similar.
At FC the radon level dropped to 131.2 pCi/l.  A  cistern is located in the corner of the house where FC
was drilled. It was theorized that the backfill in this area was not compacted, thereby allowing radon easy
access to the surface.  Samples in the walls above FC and FA were approximately 15 pCi/l each.  Two
jack posts were measured, with one containing 23 pCi/l and the other 5 pCi/l.

-------
       The final diagnostic procedure was sub-slab communication testing.  Before conducting this test
the students, using a micromanometer, measured the pressure differential between the basement and the
sub-slab soil under normal conditions.  With the exception of W2 (+0.001' WC), in the common wall
between the basement and garage, there was no measurable pressure differential at any of the pilot
holes.

       Next, a  1 1/4" hole was made at FA so that a vacuum cleaner hose could be inserted. Another
pipe was attached to the vacuum exhaust port and vented outside through a basement window.  Then
the vacuum was turned on and pressure differentials were measured at each pilot hole. At FB, 20 feet
from FA, the measurement was -0.007" WC, at FC, 30 feet from FA, it was -0.003" WC, and at FF, 20 feet
from FA, the pressure differential measurement was -0.010" WC. There was no measurable difference at
FD, 55 feet from FA, FE, 40 feet from FA, W1, 30 feet from FA or W2, 3 feet from FA.

       At this point the diagnostic portion of the workshop was completed.  After extensive discussion
of the diagnostic test results, students and instructors decided that a two-point sub-slab depressurization
system would be used to mitigate the house.

SYSTEM INSTALLATION

       Students  and instructors returned to the house, bringing in the tools and materials needed to
begin the installation of this system.  The students used a core drill to make holes for the suction points
at the mid-point of the long walls (see Floor Plan).  The optimum location for the suction points would
have been at the mid-point of the short walls, but the homeowners' future plans for the basement
precluded this option. The holes were located approximately six inches from  the floor-wall intersection
to take advantage of the settling which often occurs around footings.  Five to ten gallons of sub-slab soil
were removed at each suction point in order to extend the pressure field and reduce the resistance to
airflow.

       The students cut risers to length from 4" schedule 20 PVC, attached  elbows to the top end of
each, and inserted them in the suction points. The pipe from suction point #1 (see  Floor Plan) was run
directly to the garage rim joist in a joist bay.  Suction point #2 was connected to this manifold with a tee"
and  a 45° elbow.  Where the exhaust pipe penetrated the garage wall, the students installed a firestop,
consisting of intumescent wrap and a restricting collar.  From the rim joist the pipe was run up the rear
wall, across the garage between the trusses and into the  attic.  There an in-line centrifugal fan rated at
270 cfm in free air was connected to the pipe with rubber couplings.  The pipe above the fan penetrated
the roof near the ridge, was cut off approximately twelve inches above the roof line, and given a rain cap.
Students installed a roof boot to make this penetration weather-tight. They also painted the exposed
portion of the pipe to match the shingles and retard deterioration due to exposure to ultra-violet rays.

        The students insulated the pipe from the fan to the rim joist, to reduce noise and condensation.
To monitor the system an electronic pressure gauge was installed on the back wall of the garage. Service
switches were mounted adjacent to the gauge and fan.

        Most  of the sealing this house  needed had already been done by the homeowner.  He had
chiseled out and caulked the floor cracks and filled the sump hole with concrete. This concrete shrank
a little, so the students used a grinder to widen the crack, put gun-grade polyurethane in the bottom of
the crack and filled the remainder with pourable polyurethane caulk. The suction point holes were sealed
with mortar and pourable polyurethane around the pipe and the edge of the mortar patch. Students used
expanding foam to seal the hollow support posts.

-------
        Before the fan was permanently installed in the garage attic it was temporarily mounted to the end
 of the pipe extending from the rim joist. This was done so that the students could check the extension
 of the pressure field and determine whether a third suction point would be needed. At FA, 24 feet from
 the second suction point, the micromanometer measured -0.041" WC; at FB, approximately 27 feet from
 each suction point the reading was -0.062* WC, at FD, 25 feet from the first suction point, the reading was
 -0.007' WC; at FE, 25 feet from the first suction point, the pressure differential was measured at -0.010"
 WC; and at FF the measurement was -0.066" WC.

 POST-MITIGATION TESTING

        The pressure measurements seemed to indicate that a third suction point would not be needed,
 so the system installation was completed. This involved an additional 25 feet of pipe and four fittings and
 moving the fan to the attic of the garage. The communication test was then repeated with the following
 (in '  WC)  results:  FA:  -0.037, FB: -0.053, FC: -0.048 (FC was inaccessible  during the previous
 communication test), FD: -0.006, FE: -0.008, and FF: -0.054. The pressure field developed by the system
 did not  extend into the walls during any of these tests.

        The continuous radon monitor which had been set up before  the workshop  and  was making
 hourly readings during it, was left in the basement for one week after the conclusion  of the workshop.
 The average measurement during that week was 1.52 pCi/l. A charcoal canister exposed in the basement
 for four days during the same week measured 0.8 pCi/l.

        An  alpha track detector placed in the basement at the conclusion of the workshop was due to
 be returned for analysis in August, 1991. The homeowner decided to mail the ATD to the laboratory in
 March because he wanted winter-only post-mitigation data. This alpha track detector, which was exposed
 from 8/24/90 to 3/7/91, measured  22.2 pCi/l.  When the homeowner received this information, he called
 the Energy Office to report it.

 FOLLOW-UP PROCEDURES

        Although the homeowners' waiver, which is signed by all participating homeowners, expressly
 states that post-mitigation results are not guaranteed, the Energy Office responded to the homeowner's
 request for assistance.

        The homeowner had conducted a short-term test with a charcoal canister from May 10 to May
 15,1991, which measured 0.3 pCi/l. This reinforced the initial assumption that the ATD had been analyzed
 incorrectly.  To confirm the  long-term  test results, a continuous radon monitor was installed  in the
 basement of the house for five days (5/16-21/91). This monitor was set  up by an employee of the State
 Health Department who  works in the  area and also happened to be an  attendee of the workshop
 conducted in this house.  The hourly readings during this time frame ranged from 0.1  to 6.0 pCi/l, with
 an average of 2.3 pCi/l.

       The top graph in  Figure 1 shows the post-mitigation CRM measurements in this house when the
 sub-slab soil was being depressurized. Line A is the data collected immediately after the system was
 installed (August 1990). The data in line B was gathered after the ATD results were returned (May 1991).

       On  May 21,1991 a contractor and a New York State Energy Office representative, who were the
 instructors  of the field  workshop conducted in this house, visited the house  to  check the  CRM
 measurements and assess the operation of the system.  After examining the radon measurements made
earlier in the week, we realized that the ATD results could not be dismissed as inaccurate, and began to
think about what options were available. The two alternatives considered were adding a second fan to

-------
the system (in series), and reversing the fan so that it would 'pressurize' the sub-slab soil.  Following a
lengthy discussion, we decided to reverse the fan, continue monitoring the radon level in the house, and
then determine whether additional work would be required to bring the radon levels below 4 pCi/l.  The
decision  to try pressurization was based on several  factors, including high soil permeability and the
energy cost of adding a second fan.

       The measurements made with the fan reversed, from 5/21 to 6/1/91, averaged 0.9 pCi/l,  with
minimal variation (0.4 to 1.4 pCi/l). These results are plotted out on line A of the bottom graph in Figure
1.  In order to determine whether these test results were influenced by the weather we reviewed local
data This review showed that the weather patterns during each mode of this test (depressurization and
pressurization)  were similar, including the amount of rainfall.  In addition, the homeowners were on
vacation, so closed  house conditions were maintained for the duration of the testing.

       Follow-up CRM measurements,  from  1/6-13/92, showed much more  variation than the  May
measurements (1.5  to 6.0 pCi/l), but they  averaged 3.1 pCi/l (see line B in the bottom graph in Figure
1). This average is encouraging, and we believe that the house has been satisfactorily mitigated. Again,
ATD results will be used as the final determinant.  Because of a misunderstanding with the homeowner,
this testing did not begin until January 1992, so the results are not currently available.

       If the long-term test results are unsatisfactory, the options being considered include adding a third
'drop' (floor penetration), converting the system back to sub-slab depressurization, and stacking fans.
Short-term testing will be conducted after each option is completed to determine its effectiveness. Long-
term testing will follow satisfactory short-term measurements.

-------
Table 1     Field Workshop #13

Pre-mitigation Radon Measurements
Communication Tests
Location
first floor
bftsofnont
firat floor
first floor
WON "WAIST
Date Tested
r 8/88-11/19/88
1/10-12/90
1/10-12/90
3/6-8/90
3/28/90
Method
AID
CC
CC
CC
Liq.Sctnt
Results (pCi/L)
85.5
298.3
1642
93.4
165
Post-mitigation Radon Measurements
Locstion
basemont
paMinont
basement
basement
basement
i Date Tasted
8/17-24/90
8/20-24/90
8/24/90-3/7/91
5/21-6/2791
1/6-13/92
Method
CRM
CC
ATD
CRM
CRM
Results (pCi/L)
1.5
0.8
222
0.9
3.1
Location
FB
TO
FD
Pro-mitigation
dPTW.C.)
(Vacuum)
-0.007
-0.003
+0.000
I1 —
FE | -0.000
..
FF
W1
W2




-0.010
+0.000
+0.001




Foot front
Suction Point
27
48
63
51
25
50
4




•SnHfef*
pCi/L
515
131
490
447
564
15
15




Poet-mitigation
dP fW.C.)
(Fan)
-0.037
-0.053
-0.006
-0.006
-0.054
-0.001
+0.000




                                              4T
              • FD
                                        ISink
                                                             Suction Point «1
                 • FE

                            24'
                    Field Workshop #13
                    Drawing Not To Seal*
                                                    Suction
                                                    Point «2
                                 Windows
                                 (dowdOff)
                                                                     Storage
                                                                               FC* fl

                                                                                  W1
                                                              Support Poets
                                                                                FB«
                                                                            FF«
                                                                  Garage

-------
FIGURE 1
          FIELD WORKSHOP #13 POST-MITIGATION DATA
                 SUB-SLAB DEPRESSURIZATION (SSD)
                              HOURS

          . 8/17-24/90; immediately after installation; avg: 1.5 pCi/1

           5/16-21/91; post-mitigation follow-up; avg: 2.3 pCi/1
    5 -
          FIELD WORKSHOP #13 POST-MITIGATION DATA
                   SUB-SLAB PRESSURIZATION (SSP)
                               HOURS

          5/21-6/2/91; immediately after fan reversed; avg: 0.9 pCi/1

          1/6-13/92; post-mitigation follow-up; avg: 3.1 pCi/1

-------
                             Field Workshop Installation Summary
 Site           System Description

 #1 Waterford: 1/89

               single point SSD
               trench drain
               sump hole
               floor cracks
               utility penetrations
Measurements
Device  Location
#2 Olean: 1/89
              two point SSP
              floor cracks
              utility penetrations
#3 Poughkeepsie: 2/89
              two point SSD
              floor cracks
              block tops
              utility penetrations
#4 Manlius: 2/89
              one point SSD
              french drain
              floor cracks
              utility penetrations
              sump hole
#5 Binghamton: 3/89
              one point SSD
              floor & wall cracks
#6 Painted Post: 3/89
              two point SSP
              floor cracks
              support columns
              utility penetrations
Pre-mitigation: 78.0 pCi/l CC basement
Pre-mitigation: 15.9 pCi/l AID 1st floor
Post-mitigation: 0.8 pCi/l CRM basement
Post-mitigation: 0.5 pCi/l CC basement
Post-mitigation: 0.3 pCi/l AID 1st floor
Pre-mitigation: 522.1 pCi/1  CC  basement
Pre-mitigation: 12.7 pCi/l AID 1st floor
Post-mitigation: 2.7 pCi/l CRM basement
Post-mitigation: 2.9 pCi/l CC basement
Post-mitigation: 1.5 pCi/l AID 1st floor
Pre-mitigation: 43.0 pCi/l CC  basement
Pre-mitigation:  4.0 pCi/l AID 1st floor
Post-mitigation: 0.8 pCi/l CRM basement
Post-mitigation: 1.1pCi/ICC  basement
Post-mitigation: 0.8 pCi/l AID unspecified
Pre-mitigation: 143.0 pCi/1 CC  basement
Post-mitigation: 0.4 pCi/1 CRM basement
Post-mitigation: 0.7 pCi/1 CC basement
Post-mitigation: 0.2 pCi/1 AID  1st floor
Pre-mitigation:  66.1  pCi/1 CC basement
Pre-mitigation:  13.1  pCi/l ATD unspecified
Post-mitigation:  2.4  pCi/1 CRM basement
Post-mitigation:  2.1  pCi/1 CC basement
Post-mitigation:  1.3  pCi/l ATD 1st floor
Pre-mitigation: 163.0 pCi/l CC  basement
Post-mitigation:  3.3 pCi/l CRM 1st floor
Post-mitigation:  1.8 pCi/l CC basement
Post-mitigation: 19.0 pCi/l ATD basement

-------
#7 Voorheesvllle: 6/89
              two point SSD
              two point SMD
              french drain
              block tops
              utility penetrations
#8 E. Aurora: 6/89
              one point SSD
              one point BWD
              block tops
              floor cracks
              utility penetrations
#9 Manlius: 6/89
              two point SSD
              one point SMD
              french drain
              floor cracks
              sump hole
              utility penetrations
#10Rexford: 11/89
              three point SSD
              french drain
              floor cracks
              floor drain
#11 Goshen: 11/89
              two point SSD
              perimeter crack
              floor & wall cracks
              support columns
              utility penetrations
     Willlamsville: 3/90
three point SSD
block tops
sump hole
plywood deck suction
                            Pre-mitigation: 58.0 pCi/l CC  basement
                            Pre-mitigation: 4.9 pCi/l AID 1st floor
                            Post-mitigation: 0.4 pCi/1 CRM basement
                            Post-mitigation: 0.5 pCi/1  AID unspecified
                            Pre-mitigation: 196.8 pCi/l CC  basement
                            Pre-mitigation: 222.7 pCi/l AID basement
                            Post-mitigation: 0.8 pCi/l CRM basement
                            Post-mitigation: 0.7 pCi/l CC basement
                            Post-mitigation: 0.5 pCi/l AID unspecified
                            Pre-mitigation: 70.8 pCi/l CC basement
                            Pre-mitigation: 66.2 pCi/l AID basement
                            Post-mitigation: 0.6 pCi/l CRM basement
                            Post-mitigation: 1.8pCi/ICC basement
                            Post-mitigation: 0.6 pCi/l ATD 1st floor
                            Pre-mitigation: 56.9 pCi/l CC  basement
                            Pre-mitigation: 21.2 pCi/l ATD basement
                            Post-mitigation: 0.6 pCi/l CRM basement
                            Post-mitigation: 0.7 pCi/l CC  basement
                            Post-mitigation: 0.2 pCi/l ATD unspecified
                            Pre-mitigation: 37.9 pCi/l CC  basement
                            Pre-mitigation: 38.8 pCi/l ATD basement
                            Post-mitigation: 1.2 pCi/l CRM basement
                            Post-mitigation: 1.1pCi/ICC  basement
                            Post-mitigation: 0.4 pCi/l AT 1st floor
                                           Pre-mitigation: 39.4 pCi/l CC basement
                                           Pre-mitigation:  7.5 pCi/l ATD unspecified
                                           Post-mitigation: 2.3 pCi/1  CRM basement
                                           Post-mitigation: 2.3 pCi/l CC basement
                                           Post-mitigation: 1.2 pCi/l ATD basement

-------
#13 Naples: 8/90
              two point SSD
              floor cracks
              support columns
              two point SSP
#14 Washingtonville: 8/90

              two point SSD
              block tops
              floor cracks
#15 Syracuse: 10/90

              two point SSD
              french drain
              floor drain
#16 Voorheesville: 10/90

              two point SSD
              french drain
              floor cracks
              sump hole

#17Cortland: 12/90

              two point SSD
              block tops
              french drain
              floor cracks
     Oneonta: 5/91

              three point SSD
              block tops
              support columns
              floor cracks
#19 Marilla: 5/91
              two point SSD
              floor cracks
              support columns
              sump hole
Pre-mitigation: 298.3 pCi/l CC basement
Pre-mitigation:  85.5 pCi/l AID 1st floor
Post-mitigation:  1.5 pCi/l CRM basement
Post-mitigation:  0.8 pCi/i CC  basement
Post-mitigation: 22.2 pCi/l AID basement

Post-mitigation:  0.9 pCi/l CRM basement
Post-mitigation:  3.1 pCi/l CRM basement
Pre-mitigation:  52.5 pCi/l CC basement
Post-mitigation:  1.1 pCi/l CRM basement
Post-mitigation:  0.8 pCi/l CC basement
Post-mitigation:  0.4 pCi/l ATD unspecified
Pre-mitigation:  31.8 pCi/l CC basement
Pre-mitigation:  12.0 pCi/l ATD unspecified
Post-mitigation: < 1.4 pCi/l CRM basement
Post-mitigation:  0.2 pCi/l CC basement
Post-mitigation:  0.2 pCi/l ATD basement
Pre-mitigation:  11.7 pCi/l ATD basement
Post-mitigation:  0.7 pCi/l CRM basement
Post-mitigation:  0.8 pCi/l CC  basement
Post-mitigation:  0.3 pCi/l ATD basement
Pre-mitigation: 73.8 pCi/l CC basement
Pre-mitigation: 37.2 pCi/l ATD basement
Post-mitigation: 0.5 pCi/l CRM basement
Post-mitigation: 0.4 pCi/l CC basement
Post-mitigation: 0.1  pCi/l ATD unspecified
Pre-mitigation: 152.4 pCi/l CC basement
Pre-mitigation:  42.8 pCi/l ATD basement
Post-mitigation:  1.3 pCi/l CRM basement
Post-mitigation:  2.9 pCi/l CC  basement
Pre-mitigation:  33.3 pCi/l CC basement
Pre-mitigation:  32.3 pCi/l ATD unspecified
Post-mitigation:  2.8 pCi/l CRM basement
Post-mitigation:  0.3 pCi/l CC basement

-------
#20 Fayetteville: 6/91
              two point SSD
              floor cracks
              sump hole
              utility entrances
Pre-mitigation: 105.5 pCi/l CC basement
Pre-mitigation: 102.5 pCi/l AID basement
Post-mitigation: 3.3 pCi/l CRM basement
Post-mitigation: 0.9 pCi/l CC basement
NOTE: POST-MITIGATION CRM MEASUREMENTS AT FW 19 & 20 MAY BE INACCURATE BECAUSE
OF EQUIPMENT ERROR.
#21  Mountainville: 6/91
              four point SSD
              floor cracks
              block tops
              perimeter crack
#22   Manlius: 7/91

              one point SSD
              french drain
              floor cracks
              sump hole
              EPDM in crawlspace

#23   Loudonville: 7/91

              two point SSD
              french drain
              floor cracks
              sump hole
              floor drains
Pre-mitigation:  51.9pCi/ICC  basement
Post-mitigation:  1.1 pCi/l CRM basement
Post-mitigation:  0.6 pCi/l CC  basement
Pre-mitigation:  20.0 pCi/l CC  basement
Pre-mitigation:  24.1 pCi/l AID 1st floor
Post-mitigation:  1.2 pCi/l CRM basement
Post-mitigation:  0.7 pCi/l CC  basement
Pre-mitigation:  57.9 pCi/l CC  basement
Pre-mitigation:  4.1 pCi/l AID 1 st floor
Post-mitigation: 0.4 pCi/l CRM basement
Post-mitigation: 0.4 pCi/l CC  basement
#24   Florida: 7/91
              one point SSD
              two point SMD (2 crawls)
              floor cracks, patching
              floor drain
Pre-mitigation:  43.7 pCi/l CC  basement
Pre-mitigation:  21.1 pCi/l AID unspecified
Post-mitigation:  0.6 pCi/l CRM basement
#25   Monroe: 12/91
              four point SSD
              floor cracks
#26   Schenectady: 12/91

              two point SSD
              french drain
              block tops
              floor drains(2)
              well cover
Pre-mitigation:  43.8 pCi/l CC  basement
Pre-mitigation:   8.7 pCi/l ATD 1st floor
Post-mitigation:  1.7 pCi/l CRM basement
Post-mitigation:  1.6pCi/ICC  basement
Pre-mitigation:  74.8 pCi/l CC  basement
Post-mitigation: 61.0 pCi/l ATD basement
Post-mitigation:  3.9 pCi/l CRM basement

-------
                                                                  VIIP-5
        RADON MITIGATION AT  SUPERFUND REMEDIAL ACTION  SITES
                    FIELD EXPERIENCE AND RESULTS

          by: Mr. Jean-Claude  Dehmel, CHP
              S. Cohen  & Associates, Inc.
              McLean, VA 22101

              Mr. Ronald F.  Simon
              R.F.  Simon Company,  Inc.
              Barto, PA 19504

              Mr. Eugene Fisher
              U.S.  Environmental Protection Agency
              Office of Radiation  Programs
              Washington, D.C. 20460

      The Environmental Protection Agency  (EPA)  has initiated a
radon  mitigation project  in  Montclair,  West  Orange,  and  Glen
Ridge,  NJ.  In these  communities,  numerous properties have  been
contaminated with radium tailings  which were initially introduced
around homes as backfill and used  as construction materials. This
practice has  long  since been  discontinued, but  many residential
properties  remain  contaminated.  These homes have  been listed in
the  National  Priority  List  under  the CERCLA and  Superfund  acts
and  are  currently  being  remediated by   the  EPA's  Region  II
Offices. In 1983, the EPA mitigated  a number  of  homes with radon
levels  above  4  pCi/L.  The  mitigation  technology  relied  on
forced-air ventilation  systems. The old systems required frequent
repairs, were  noisy,  exhibited temperature fluctuations causing
excessive humidity  and  condensation, and many of the systems were
also not  successful in reducing  ambient radon levels below the
EPA guideline.

      In this project,  old mitigation  systems were removed in 28
homes and replaced  with depressurization  systems  designed to fit
the construction features of each  home. The effectiveness of each
mitigation system was determined by conducting field inspections,
pre-operational tests,  and by using continuous radon monitors and
alpha  track detectors. Test  results  revealed   that  all  radon
levels  were well  below the  EPA  guideline.  Reduction  in  radon
levels  ranged  from  99%  to  45%.  Reduction  gains  were  more
significant  for those  homes  that  had   initially  higher  radon
concentrations  (above  20  pCi/L).  For  those  which  had  only
marginally elevated levels  (above  4 pCi/L), the  resulting gains
were significantly lower,  ranging  from 45 to 83%.

-------
       The work  described in this  paper was  funded  by the  U.S.
 Environmental  Protection  Agency.   This  paper,  however,   was
 exempted from Agency  and administrative  review. Therefore,  the
 contents do not necessarily reflect  the views of the Agency  and
 no official  endorsement should  be inferred.  The work  described in
 this paper was  conducted under  a  cooperative agreement  between
 the Office of Radiation Programs and  Region  II's Emergency  and
 Remedial Response  Division.

                       PURPOSE AND SCOPE

      Old mitigation   systems  were  removed   in  28 homes   and
 replaced with   state-of-the-art   depressurization  systems   (1).
 Radon  diagnostic techniques and  mitigation strategies developed
 by the  EPA  in its  House Evaluation Program  (HEP) were applied  in
 the planning stages.  The new systems were assembled  on site  and
 designed to  fit the  construction characteristics  of each home.
 Each assembled  system was subjected  to a complete assessment  of
 its performance,   including   pre-   and  post-mitigation  radon
 testing.  The monitoring  results  were   used  to  evaluate   and
 document the effectiveness  of  each system in  reducing  indoor
 radon levels. Finally, the old ventilation systems were  removed
 and appropriate  restoration work  was conducted  under  a negotiated
 agreement with each homeowner.

      In general,   diagnostic  testing  procedures,  selection  of
 remediation  system designs,  and  system installation  relied  on
 techniques  and  methods which have proven to  be  successful.  The
 field work was divided  into five  major  phases:

 1)   Pre-mitigation activities,
 2)   Installation of mitigation system,
 3)   Evaluation of mitigation systems,
 4)   Mitigation activity closure,  and
 5)   Removal of old ventilation systems  and restoration (R&R)  of
     affected  areas.

      Because of the presence of  soils  contaminated with radium,
the work was performed under the umbrella of site specific health
and  safety plan. A QA/QC plan was also  developed for  the purpose
of ensuring that all radon monitoring methods and techniques used
in  the   course   of  the  project  complied with EPA  guidelines.
Descriptions  of  field  and  project management activities  were
described in  three documents,  a  Work  Plan,  Health  and  Safety
Plan, and a Quality Assurance Plan (2, 3, 4).

-------
        Following   the   completion   of   each   mitigation   and
restoration, a  completion report was  prepared and  submitted to
the  EPA Region  II  and EPA-ORP  offices.  Each  completion  report
contains information and documentation regarding the negotiations
held  with  the  homeowner,  diagnostic  and  pre-mitigation  test
results,  mitigation  system  design  plans  and  specifications,
information  about   the   installation  activities   of   the  new
mitigation  system  and  removal   of  the  old  ventilation  unit,
summary  of  all  health   and  safety  survey  results,  and  data
regarding the placement  of 3- and 12-month alpha track detectors
(ATD).

The  project  was  initiated  on   April  27,  1990  following  the
issuance  of the  work  assignment. The  full mobilization  of  the
field equipment and personnel was  completed in early  June.  All
field work  and  support activities were also started at the same
time. All project activities were closed on September 30, 1991.

                      PROJECT ACTIVITIES

PRE-MITIGATION ACTIVITIES

Homeowner Participation

      Homeowner  participation varied significantly.  Some  owners
were  quite  cooperative  in  signing a  release  form  and starting
the diagnostic and mitigation work. For homes equipped with older
pressurization  and  ventilation systems,  some owners acknowledged
their  frustration  and  inconvenience   caused  by  the  need  for
frequent  service and repairs. System breakdowns  often involved
lengthy delays before repairs could be made resulting in elevated
radon levels.  For  some  residents,  the resulting  elevated radon
levels were of  specific  concern.  Some  homeowners were interested
in  selling  their homes  in  the near  future,  and  therefore were
looking  for documentation that  radon was  no  longer  a concern.
Other  homeowners  were  very  skeptical  and  uncooperative  and
responded only  after persistent  inquiries.  In a  few instances,
participation was obtained only after the intervention of the EPA
Region II's representative. Some  homeowners seemed frustrated by
years of  surveys, testing,  and disruptions, and  expressed doubt
that  any  final  resolution to  the problem  would ever  be  found.
Others,  particularly  elderly  owners,  verbalized disbelief  and
confusion as  to  the  nature  and  extent  of the  potential  health
risks associated with radium contaminated  soils  (elevated radon
and  gamma  exposure levels)  given that  they had lived  there  for
several decades.

-------
Other  homeowners  expressed anger and  fear  given that four homes
remained vacant for so many years only to be recently demolished.

       Initially,  much effort and time were devoted in obtaining
early  homeowner  approvals.  Less  emphasis was  devoted  to less
cooperative  homeowners  or to  homes with  relatively  lower radon
concentrations. Homes which  had just been mitigated were used as
referrals  to hesitant  homeowners.   Neighboring  homes  were also
used,  with prior approval,  as model  installations and hesitant
homeowners  were  invited  to  inspect  finished  installations  and
talk  with their  owners.  This  approach  was  used  to  enlist  the
participation of  the  remaining homeowners.

Diagnostic Work

       Diagnostics and pre-mitigation measurements were conducted
in all homes prior to the installation of any mitigation system.
Before any work  was  performed,  a signed  agreement was obtained
from  each  homeowner. In some  instances,  diagnostic work  was
initiated  based  on  prior EPA  agreements  obtained during earlier
site visits. Discussions  were  held  with  each homeowner regarding
the extent  and  types of  diagnostic  activities  and intentions to
replace  the old  mitigation  system  with  a new  one.  Discussions
also addressed the significance of earlier radon test results and
those  obtained   during   this  round   of  diagnostic  testing.  A
walk-through was  also conducted to  go over system installation,
pipe routing in living areas, system operations, operating costs,
maintenance needs, electrical  hook-up, etc.. Initial  health  and
safety surveys  were   normally  performed  in conjunction with  all
diagnostic activities.

Pre-Mitiaation Radon  Measurements

      Radon measurements  were  made  prior to  the installation of
any new mitigation system. A  continuous  radon monitor was placed
in each home and  left running for a period of  approximately  one
week.   During such tests, the  older radon  mitigation  system  was
rendered inoperable.  These  measurements  were made  in  the lowest
"livable"  area  (generally  the  basement)  or the  lowest  ground
floor, in  the case of split-level homes. The tests were  started
at  the  conclusion  of  the  diagnostic  activities when  it  was
possible to schedule  the  mitigation  for  the following week. When
scheduling   would  not   permit   this   type    of   arrangement,
pre-mitigation tests  were started at  a later date and scheduled
to end just before the start of the mitigation work.

-------
 This approach was  used to ensure  that pre- and  post-mitigation
 measurements were performed under similar seasonal conditions.  At
 the onset of the mitigation work,  the old system was  re-activated
 to reduce ambient  radon levels for  health and safety  purposes,
 especially in some homes where radon  concentrations were at times
 greater than 100 pCi/L.

       Pre-mitigation measurements  were  performed under  closed-
 house conditions whenever  possible,  given that  some tests  were
 performed during  summer months.  Since  none  of  the  homes had
 central air-conditioning and  only a small  percentage  of  homes
 possessed room  air-conditioners,  summer  time  measurements  were
 not  always  performed  under  true   "closed-house"  conditions.
 Instead,  attempts were made to  have  homeowners maintain  "closed
 basement"  conditions,  as  best  as   they could.  Only  one  pre-
 mitigation measurement was made with an existing  pressurization
 and ventilation system  left  running,  since  the owner  expressed
 concern for  health  reasons   about  rising  radon   concentrations
 during the conduct  of the testing period.

 Radon Monitoring Methods

       All  radon measurements associated  with   pre-mitigation,
 follow-up,  and post-mitigation testing were  performed  utilizing
 Femto-Tech continuous radon monitors, Model R210F. The continuous
 radon monitors  were calibrated  by  the  manufacturer  every six
 months.  A total  of ten monitors were deployed  during the  course
 of  the project.  As  monitors were removed  from homes (particularly
 from  those homes with  elevated  radon  levels),  these units were
 returned   to  the  field  office   (where  radon  concentrations
 approaching  outdoor ambient had been  documented) until sufficient
 flushing had occurred. Pre-mitigation measurements involved using
 both  the monitor and  its data-logger, which automatically records
 hourly  readings  and running average radon concentrations.

      Radon  measurement  results  were  routinely  scrutinized to
 identify  the  possibility  of  power   interruptions  which  would
 re-set   the   data-logger.  Duplicate  measurements  were  also
performed  in  several  tests  and while  instruments  were  placed in
 storage at the field  office. One  monitor failure occurred during
the   conduct   of  a  final  post-mitigation  measurement.   This
measurement  was  later repeated using another monitor.  A second
monitor  failure  occurred while performing  an  interim evaluation
of  an   installation.  A  third   failure  was  discovered  while
performing duplicate radon measurements at the field office.

-------
      Diagnostic  radon measurements  were conducted using  Pylon
AB5 units. The  Pylons were used to characterize  radon  levels in
walls, under basement  slabs,  in  crawl spaces,  in exhaust stacks,
etc.  The  measurements  were performed  only for  the purpose of
assessing relative rather than absolute radon concentrations. The
instrument  were  initially  calibrated  by  the  manufacturer  or
EPA-ORP, and were  later re-calibrated at the U.S.  Department of
Energy's  facility  located  in New  York City.   Pylon measurement
results  were  not  used  to  assess  the  effectiveness  of  the
mitigation systems.

      Short-  and  long-term  radon measurements  were  performed
using Radtrak alpha  track detectors  (ATDs) supplied and read by
Tech/Ops  Landauer.  ATD measurement  results were used  to assess
the  effectiveness  and to  document  the  installation   of  radon
mitigation  systems.  A  group of ATDs were spiked  at  the  EPA's
NAREL facility  and sent to  the  supplier for processing. In a few
instances, radon measurements were also  performed  using E-PERMs
for comparative purposes.  As before, E-PERM measurement results
were not used to assess the final effectiveness of the mitigation
systems.

MITIGATION ACTIVITIES

System Designs  and Descriptions

      Radon mitigation  systems were designed after reviewing  data
collected  during diagnostic  activities  and  files  from earlier
characterizations. The formal designs and  floor  plans  were  then
presented  to  the homeowners  (and  the EPA Region II and the ORP
Work Assignment Manager) for  review and approval. At times, minor
modifications were made to the original system designs  to satisfy
specific  homeowners   requests.  Mitigation  system  designs  and
installations varied very  little  from house to  house. Although
the  work plan  allowed flexibility  in selecting  any   mitigation
approach,   standard   sub-slab   depressurization   systems    were
designed  and  installed in all homes. These systems involved one
or  more   floor taps   depending   upon  the  extent  of  sub-slab
communication.  Areas with  elevated gamma  radiation levels  were
generally  avoided  when locating  floor  taps and  fan boxes.  When
required,  crawl spaces were also depressurized either by tapping
through  knee-walls  or directly  down  and  through crawl  space
floors.  Table 1 summarizes  some of  the  important  system  design
features.

-------
TABLE 1.  MITIGATION SYSTEMS INSTALLATION VARIABLES
Installation Variables*
Home Fan No. of Amount Finish Crawl Basem't Work
Address LOG. Taps"1" Sealing Basem't Space Access Condition
1 Alan box
2 Alan attic
13 Amelia box
15 Amelia box
18 Amelia attic
145 Carteret box
149 Carteret box
151 Carteret box
153 Carteret box
26 Fremont attic
30 Fremont attic
32 Fremont attic
34 Fremont box
2 James box
6 James box
8 James box
21 Lorraine attic
26 Lorraine attic
53 Nishuane attic
56 Nishuane attic
62 Nishuane box
64 Nishuane attic
66 Nishuane box
26 Virginia box
28 Virginia attic
30 Virginia attic
31 Virginia roof
35 Virginia attic
3
1
2
1
3
4(lw)
2
3 (1cm)
3(2cs)
3(lcm)
2
3(2cs)
3
1
1
2
4(lcs)
3
2(lsog)
3(lw)
3(lcs)
4(lcs)
4(lcm)
3
3(lw)
3
1
3
avg.
less
less
less
avg.
avg.
avg.
less
less
more
avg.
more+
avg.
less
less
avg.
avg.
avg.
less
less
less
avg.
avg.
avg.
less
less
less
more
no
no
no
yes
yes
no
yes
no
no
no
no
no
no
semi
yes
yes
no
no
semi
no
yes
no
no
no
no
no
yes
no
none
none
none
none
large
small
none
small
small
small
small
small
small
none
none
none
small
none
none
none
avg.
avg.
small
none
none
none
none
small
med.
good
poor
good
good
avg.
avg.
avg.
avg.
poor
poor
poor
poor
good
good
good
avg.
avg.
good
avg.
avg.
avg.
avg.
poor
poor
good
poor
poor
avg.
good
poor
good
avg.
poor
avg.
poor
poor
poor
avg.
avg.
avg.
avg.
good
good
avg.
avg.
avg.
poor
avg.
avg.
avg.
avg.
poor
avg.
poor
avg.
 * Definition of variables: Fan - where the fan was located;
   Taps - total number of locations tapped for each SSD system;
   Sealing - approximate amount of sealing that was involved;
   Finish - whether basements were finished or not; Crawl -
   presence of a crawl space; Access - ease or difficulty in
   accessing various basement areas; and Condition - ease or
   difficulty in working in basement areas due to the
   availability of open spaces.

-------
TABLE 1.  MITIGATION SYSTEMS INSTALLATION VARIABLES, CONT'D

 + Key to system tap designations: "lw" - denotes that one of the
   taps was a wall tap; "1cm" - denotes that one of the taps was
   installed to depressurize a crawl space with a membrane cover;
   "Ics" - denotes that one of the taps was installed to
   depressurize a crawl space over an existing floor (generally
   involved coring through knee-walls); and "Isog" - denotes that
   one of the taps was installed to depressurize a slab on grade
   area (generally involved coring through knee-walls).


      Various  types  of equipment,  supplies,  and  materials  were
used equally  in both mitigation and  restoration work.  Materials
used for most of the 28 installations were fairly consistent with
a few exceptions. Every job required the following items:

o An in-line exhaust fan
o Pressure flow indicator and audio-visual alarms  (as negotiated)
o PVC piping, joints, dampers, and fittings
o Sealants, caulks, and PVC cements
o Electrical switches, wiring, and wiring boxes
o Exhaust caps and roof flashing
o Miscellaneous hardware and supplies
o Maintenance and warranty package and operating instructions

Extra materials were  at times used when  installing  fans outside
or in attic areas; including:

o PVC fan box  (outdoor installations only)
o PVC down-spouts with fittings (used to direct exhaust above
  roof levels for outdoor installations only)
o Sound and thermal insulatory material for pipes and fan boxes

      The  placement  of  exhaust  fans can  be  divided   into  two
groups,  conventional installation and outdoor  fan  box.  In 14 out
of 28  homes,   installations  involved  locating  the fan  in  attic
spaces.  For one home with an attached garage (there were only two
such homes), the fan was located  in the  garage attic.  In another
home, a roof  fan was  installed,  however.  Homes where  fans  were
located   in   attic   spaces   required   the   construction   and
installation of  "box-ins"  in closets  and/or  rooms on  the  first
and  second  floors  for  aesthetic  reasons.  The   "box-ins"  were
finished  to  match  or  approximate  their  surroundings,  since
replacement  moldings  were   sometimes  difficult  to  obtain  and
accurately match those typically found in such older homes.

-------
      Because of the difficulty  in  routing a manifold up through
the first and second floors, the remaining installations involved
locating the fan outside near a  basement  wall.  Most of the homes
were of "Dutch Colonial" designs with basements, first and second
floors, and unoccupied  attic spaces that  did not  facilitate the
direct routing  and installation  of exhaust manifolds.  The  fans
were encased  in insulated  PVC  boxes for  weather  protection and
noise control. The top  of  the fan box  was located  at grade level
to permit  access for maintenance and servicing.  Exhaust  stacks
were  made  of  PVC  down-spouts   rising   from  the  fan  box  and
discharging above the roof line.

      Sealing  and caulking were  performed with   each  sub-slab
depressurization   system   installation.    Sealing   and  caulking
enhance the effectiveness of the sub-slab depressurization system
by  minimizing  basement air losses while  maximizing  negative
pressures under the slab.  The amount of  sealing varied from home
to home.  A few  homes  required  very little sealing  because new
floors had been  recently   installed or  there  were  only  a  few
joints, penetrations, cracks, etc..  Conversely,  in one home, the
entire basement  floor had  to be  re-finished in  order to isolate
and depressurize the sub-slab areas.

System Installations

      Labor time  and material expenses balanced one  another out
when compared  to the efforts required to install a  fan  either
indoors or outdoors. Box-in materials were similar  in cost to the
materials  used  to  manufacture  the fan  box.  Labor  hours  were
approximately the same  to  build  "box-ins" as opposed to building
the fan  box and excavating  the  fan box  hole.  Typically,  health
and safety  implications were more  significant  in  excavating fan
box  holes,  as  there   always  was  the  risk of   finding  radium
contaminated  soils.  Table  1   summarizes  some   of  the  most
significant variables found  to  impact  work productivity. For the
28 mitigated  homes,  installation times  ranged between one and
eight  days.  This  effort  does  not   include  pre-site  preparation
work, follow-up activities, and time spent off-site in support of
installation work or health and safety functions.

      Work productivity increased by about  30%  after completing
the first  few  installations, as work  procedures and  skills were
being fine tuned.  Scheduling problems  also affected productivity
as  some  homeowners  were  confused  about  making arrangements  to
have someone at the home for several consecutive days.

-------
Follow-up visits to support system modifications or touch-up work
also tended to  increase the overall level of effort. As mentioned
earlier, the level of effort associated with sealing and caulking
varied significantly among homes.

      Health  and  safety  requirements  generally  reduced  work
productivity  between 30  and  50  percent.  This  was particularly
true when  coring and excavating  floor taps or  fan box holes in
areas suspected of containing radium contaminated soils. All work
areas  had  to  be  evaluated  for  each  home  since  radiological
conditions  varied  from house  to house. Homes  with contaminated
soils in  immediate  work areas were of concern  since  soils could
be re-suspended or entrained in otherwise non-contaminated areas.
In  such  instances,  cautionary  measures  were  implemented  to
complete the  work  as efficiently as was possible  while ensuring
that all health and safety requirements were met. In seven homes,
the basement  floors were  found  to be  covered  with lead shielding
and  carpeting.   In  five  other  homes, sand  bags  were used  as
shielding  to cover  exposed  soils in crawl  spaces.   For  these
homes, the mitigation phase took much  longer  since the shielding
had to  be first removed  to drill  floor  or wall taps  and later
modified to permit  its  re-installation.  Additional  time was also
spent conducting radiation surveys to  verify the integrity of the
shielding at the completion of each installation.

      Lack of access was  also  found to have a significant impact
on  productivity.  Basements  with  doors  and  windows,  however,
allowed  easy  entry  and offered  convenient ways  with which  to
ventilate all  work areas.  However,  other basements with  narrow
doors, steep and narrow stairs,  low head-room,  etc. proved to be
difficult  in  staging  the  work  and  required  more  elaborate
ventilation  schemes.   Some basements   were  virtually  empty  or
otherwise afforded ample space to store tools and equipment.

      On the  other  hand,  some basements were packed with  boxes,
clothes,  furniture,  tools, and  other assorted items  requiring
frequent moving and re-location.  Most basements,   however,  were
found to  be between  these two  extremes.  Similar  problems  were
encountered in  "living-spaces".  Homes  with crawl  spaces  usually
required more time  depending  upon the  size of the  space  and  the
work that  needed to be performed.  Although most basements  were
unfinished,   a  few  were  clearly  finished  and  were  used  as
"lived-in" areas, while a few others  were  "semi-finished".  Work
in  finished  spaces  always  required  extra   precautions  and,
therefore, more time and effort.

-------
Table 1 summarizes some of the major installation variables which
had an impact on productivity and project resources. It should be
noted that not one single variable can be used as an indicator of
the ease or difficulty encountered during any of the mitigations.
Finally, the  level of effort  is  difficult to break-out  by work
phases,  since  in  many  situations,  mitigation  work  occurred
concurrently  with removal  and restoration  activities.   In such
situations, time  expenditures evened  out  between both types  of
activities since some work could equally be defined as mitigation
or restoration.

      All  mitigation activities  were performed  by the  project
team,  with  the  exception   of electrical  hook-ups  to  circuit
breaker  panels and  utility  poles.  A  licensed electrician  was
sub-contracted to  perform  such tasks and to  secure  the  required
permits. All  new  sub-slab depressurization  systems  were wired to
dedicated electrical meters. Finally, in re-flooring the basement
of one  home,   a  plumbing company  was sub-contracted to  re-hook
sink drains to the waste line.

      Radon measurements were  made  following  the installation of
each  new  mitigation  system.  Continuous  radon  monitors  were
deployed in  each  home for  a period of approximately two weeks.
During  such  tests,  the  older  radon ventilation  systems  were
rendered inoperable.  The measurements were  made  in the lowest
"livable"  level   (usually  the basement,  except for  split-level
designs) within a relatively short-time after the mitigation work
had  been   completed,   generally  within  24  hours  after  the
depressurization system became operational.

Evaluation of Mitigation System Effectiveness

      The  effectiveness  of  each  mitigation  system  installation
was  based   on evaluating  pre- and  post-mitigation  measurement
results. Short-term measurements  were  most often  utilized  to
perform  this  function,  but  long-term  radon  measurements  are
better   indicators  in  confirming   actual  reductions.   Post-
mitigation  measurements   were   performed   under   closed-house
conditions whenever possible. Since none of the homes had  central
air-conditioners   and   only  a   few  homes   possessed   room
air-conditioners,  summer time measurements were for the most part
not  performed  under true   "closed-house"  conditions.   Instead,
attempts were  made to have  residents  maintain  "closed  basement"
conditions. Successful  mitigations were considered  to be  those
with measurements that were below the EPA guideline of 4  pCi/L.

-------
Table 2  presents  pre- and post-mitigation test  results with the
associated  reductions.  The results reveal that  all  radon levels
were  well below  the EPA  guideline.   Reduction  in  radon levels
ranged from as high as 99% to as low as 45%.  Typically, reduction
gains are more significant  for  those homes which  had initially
higher radon concentrations  (above 20 pCi/L). For homes which had
only marginally higher initial radon levels,  when compared to the
EPA guideline,  the gains were significantly  lower,  ranging from
45 to 83%.

Removal and Restoration  (R&R) Activities

      The  R&R activities  usually  started with  an  interview  to
discuss  and  list  those  tasks  needed  to restore the basement  to
its original  conditions,  i.e.,  prior to the  installation of the
old  ventilation  system.  Soliciting   the  cooperation  of  these
homeowners  into the  removal and  restoration  (R&R)  program was
generally  much  easier  since   the   project   team  had  already
established some credibility during the mitigation work.  The EPA
Region II office  had also  informed  homeowners  that  they  could
retain the  equipment providing  that  they  assume all  energy and
maintenance costs. Few homeowners chose this option when  informed
of  the  typical  energy  bills  associated with  the  operation  of
these older systems.  A few others,  however, retained basement
window air-conditioning units.  A few homeowners,  were not pleased
with  losing   their  older  ventilation  systems   despite   the
advantages  that the  new mitigation  systems  offered.  Typically,
these owners  used the  old units to  cool  and heat  the basement
areas, thereby  off-setting heating and cooling costs in the rest
of  the  house.  Fortunately,  this situation  was  an  exception  as
most homeowners were pleased with the removal of the old systems.

      The  R&R work  was scheduled after such negotiations  were
finalized  in  an agreement.  Generally,  most  R&R activities  took
one day  followed  by  a  few  short  return trips to  conduct  minor
touch-ups, e.g., application of  a  second coat  of paint or stucco
and dry-wall  finishing. Some  R&R  activities, however,  required
several  days  (up  to  three), when  more  extensive  agreements had
been reached  with  homeowners.  Wall openings left by the removal
of the old  systems and duct work were  replaced  with new windows
and/or filled in.  In some instances,  new windows and screens were
custom made  in order  to match  existing  ones.  All  windows  were
made operable  or  fixed  shut as  was  specified by  the homeowner.
Window fans that  were part  of the old system were  also removed
and the windows were restored.

-------
TABLE 2.  PRE- AND POST-MITIGATION RADON MEASUREMENT RESULTS
House
1 Alan
2 Alan
13 Amelia
15 Amelia
18 Amelia
145 Carteret
149 Carteret
151 Carteret
153 Carteret
26 Fremont
3 0 Fremont
3 2 Fremont
34 Fremont
2 James
6 James
8 James
21 Lorraine
26 Lorraine
53 Nishuane
56 Nishuane
62 Nishuane
64 Nishuane
66 Nishuane
26 Virginia
28 Virginia
30 Virginia
31 Virginia
37 Virginia
I.D.I
211
212
112
113
221
312
313
314
315
141
142
143
144
241
242
243
321
322
163
164
165
166
167
173
174
175
176
178
Short-Term*
Pre Post
144.8
2.8
18.7
8.0
7.3
3.0
27.7
25.8
22.6
5.4
30.4
25.3
11.8
32.6
13.7
1.8
17.7
3.1
13.7
25.4
1.1
145.4
4.0
11.1
7.0
12.9
15.3
136.3
1.3
1.0
0.5
2.4
3.6
1.4
1.2
1.8
0.5
0.9
1.2
1.3
2.2
2.1
0.6
0.6
1.3
0.6
0.7
0.5
0.6
2.4
0.7
0.8
2.1
0.5
0.9
2.3
Percent
Reduction
99%
64%
97%
70%
51%
53%
96%
93%
98%
83%
96%
95%
81%
94%
96%
67%
93%
81%
95%
98%
45%
98%
83%
93%
70%
96%
94%
98%
Long-Term Post"*"
ATD E-PERM
0.8
0.3
ND&
ND
ND
ND
ND
ND
0.6
ND
ND
0.8
ND
ND
<0.2
ND
<0.2
<0.2
0.6
<0.2
ND
0.8
ND
ND
1.1
ND
ND
ND
0.7
0.4
-
-
-
-
-
-
-
-
0.7
-
-
<0.4
-
0.4
0.7
3.0
-
-
0.4
-
0.8
    All pre- and post-mitigation tests were performed in the
    "lowest livable" area (basement) using continuous radon
    monitors (Femto Tech R210F); pre- and post-tests were
    performed within several weeks of each other and under closed
    house conditions as best as possible. All radon results
    are expressed in pCi/L.  See text for detail.
    Long-term test results listed above are for basements or
    lower floors under "normal house conditions". E-PERMs were
    deployed only in nine homes. See text for detail.
    ND signifies "no data" as the ATDs were still deployed at
    the completion of field activities.

-------
Only conventional health  and safety precautions were implemented
during  R&R  activities,   since   no  intrusive  work  was  being
performed in radium contaminated soils.

      Construction materials  typically included concrete blocks,
mortar,  windows or  glass  panes,   finish  trims,  and paint  and
caulking  compounds.  Several  homes  required the  installation  of
wooden storm and/or screen windows,  which  often were custom made
since replacements were difficult to  find  for such older homes.
In other instances, the overall level  of effort was significantly
higher,  especially  for  those homes   requiring  more  extensive
restoration work and those  involving the use of sub-contractors.

      Finally,  discarded   equipment   and  debris  were  disposed
through  the  local  waste  carting  company.  Refrigerant  gases
[chlorofluorocarbons  (CFCs)]  contained in  the old  HVAC systems
were  drained  and  collected  for  proper  disposal.  Each  unit
contained up to five pounds of Freon gas.  Arrangements were made
with a specialized firm for the collection and proper disposal of
refrigerant gases. Such efforts were made  to abide by the intent
of the recently enacted EPA Clean Air  Act Amendments of 1990.

Health And Safety (H&S) Activities Summary

      Health  and  safety  monitoring  results   revealed  that  by
adopting  simple protective measures,  personnel  exposures  were
maintained  well  below   occupational   standards  and,   in  some
instances, at the threshold of measurement detection limits. Some
of  the  applied protective measures  included  working  in  well
ventilated areas, judicious uses  of local  exhaust ventilation at
the  source  of  contaminants,  application   of  dust  suppression
techniques, use of the functional  sections  of mitigation systems
to minimize  radon exposures  and  resuspended  particulates  while
completing  its  installation,  use  of  containment  methods  to
minimize  the  spread of  contaminants,  and  restricting  personnel
traffic out  of the work  areas. The use of  monitoring equipment
was shown to be helpful  in detecting trends in ambient radiation
exposure rates and radon levels. Routine surveillance of all work
activities has  also  allowed the timely detection of potentially
problematic situations.

      A review of the  survey results  revealed that all exposures
were well within occupational  radiation protection standards and
OSHA criteria.  Survey  measurement results varied depending upon
pre-existing conditions  and type  of mitigation work.  Typically,
average  radon  levels  varied  from  0.4  to  32.5  pCi/L;  radiation
exposure rates  ranged  from 6 to 460 uR/h;  surface contamination

-------
were below  the detection  limits of  17  dpm/100  cm2;  long-lived
radionuclides concentrations  were  <6.8 x  10~13  uCi/mL;  asbestos
fiber  concentrations varied  from  <0.002  to  0.016  fibers/cm3;
total  suspended  airborne particulate  concentrations  varied from
<0.01 to 0.65 mg/m3; and organic vapor concentrations ranged from
<0.36  to  13 ppm-TWA for compounds  typically  found  in  caulking
compounds and PVC cements.

      In  support  of the field work,  several  soil  samples were
analyzed  for   the  presence  of  U-238,   Ra-226,   and  Th-232.
Contaminated soils and  tailings  originated from  the extraction
and  purification  process  of  uranium ores  to  produce  radium
luminous  paints.  The  analyses were performed  by  GeLi  gamma
spectroscopy. The maximum soil concentrations were noted to be 27
pCi/g  for U-238,  123 pCi/g  for Ra-226, and 5.9  pCi/g for Th-232.
Background soil concentrations for these nuclides in Northern New
Jersey are typically <1.0 pCi/g  (5).

                           CONCLUSIONS

      The  project   revealed   that  the   installation  of  radon
mitigation  systems  in  homes  with  contaminated soils  can  be
effectively   performed,   even   when   the   distribution   and
concentrations of  radium are  not well known.  The installation of
standard  mitigation systems,   and,  at times,  variant designs to
meet specific  needs, proved to be  successful  in achieving radon
reductions  as  high as 99% and  in  meeting  the  EPA guideline of  4
pCi/L  for  all  28  mitigated  homes.  When  compared  to  other
traditional  mitigation  activities,   this  project  involved more
extensive   efforts,   since   it   was  required   to   remove  old
ventilation systems  and restore affected areas to their original
conditions.  In  some  cases,   the  work  was  further  delayed  or
complicated by demands made by  the  homeowner.  In all instances,
negotiated  agreements  had to  be reached with each homeowner. The
experience  gained during this project may prove  to  be useful in
other  similar  mitigations, since  there  are still  a few hundred
homes  with  elevated soil radium  concentrations  and radon  levels
in Northern New Jersey.  Results  from health and safety monitoring
activities  revealed that by adopting simple protective measures,
personnel   exposures were  maintained well  below   occupational
standards  and,  in  some  cases,  at  the threshold  of detection.
Excluding the  presence  of  radium  contaminated  soils, the  health
and  safety  monitoring  methods used  in  this  project  could be
applied  during  the installation  of  mitigation systems  under
conventional conditions.

-------
                           REFERENCES


1.   Environmental Protection Agency, Office of Radiation
     Programs, Work Assignment No. 1-39, House Evaluation Program
     Applied to Superfund Sites, S. Cohen & Associates, Inc., EPA
     Contract No. 68D90170, April 27, 1990, amended October 26,
     1990.

2.   Work Plan, House Evaluation Program Applied to Superfund
     Sites - Montclair, West Orange, and Glen Ridge, New Jersey,
     Prepared by S. Cohen & Associates, Inc. for the U.S.
     Environmental Protection Agency, Office of Radiation
     Programs, Work Assignment No. 1-39, Contract 68D90170, May
     1990, revised November 1990.

3.   Health and Safety Plan, House Evaluation Program Applied to
     Superfund Sites - Montclair, West Orange, and Glen Ridge,
     New Jersey, Prepared by S. Cohen & Associates, Inc. for the
     U.S. Environmental Protection Agency, Office of Radiation
     Programs, Work Assignment No. 1-39, Contract 68D90170, May
     1990.

4.   Quality Assurance Plan, House Evaluation Program Applied to
     Superfund Sites - Montclair, West Orange, and Glen Ridge,
     New Jersey, Prepared by S. Cohen & Associates, Inc. for the
     U.S. Environmental Protection Agency, Office of Radiation
     Programs, Work Assignment No. 1-39, Contract 68D90170, Rev.
     1, August 1990.

5.   Camp Dresser & McKee, Inc. Supplemental Feasibility Study
     for the Montclair/West Orange and Glen Ridge Radium Sites,
     Vol. 4, prepared for the U.S. Environmental Protection
     Agency, Region II, Edison, NJ, April 3, 1989.

-------
                                                               VIIP-6
       Dose and Risk Projection for Use of Sub-Slab Radon
          Reduction Systems Under Realistic Parameters

                          Larry Jensen
              U.S. Environmental Protection Agency
                     Chicago,  Illinois 60604

                          Felix Rogers
                   Centers for Disease Control
                        At1anta, Georgia

                          Charles Miller
                   Centers for Disease Control
                        Atlanta, Georgia

Structures found to have elevated indoor radon levels that require
reduction are often fitted with sub-slab depressurization systems
that remove the radon soil gas from beneath the foundation before
it can enter the home.   Such systems generally discharge the radon
soil  gas  to  the atmosphere,  relying  upon  mixing  to  reduce
concentrations to levels indistinguishable from that of the ambient
air.  However, if a community where a significant fraction of the
residences that exceeded the U.S. Environmental Protection Agency's
radon  action  level   were to  install  these  systems  then  the
collective radon  emission to  the atmosphere potentially might be
consequential.  To estimate the impact of large-scale installation
of sub-slab depressurization  systems  a  hypothetical  scenario was
modeled using  realistic housing densities  and  the consequences,
measured as increased radon concentration  in the community and its
surroundings, were projected.
             1992 International Symposium on Radon &
                    Radon Reduction Technology

                      Minneapolis,  Minnesota
                      September 22-25, 1992

-------
                        PROBLEM STATEMENT

One standard system to mitigate homes that exceed radon guidelines
is sub-slab depressurization.  Piping is inserted through the floor
slab, into a sealed sump or into drainage tile to draw the soil gas
out and vent  it  to the atmosphere through  a chimney (He88).   The
standard assumption is that the  atmosphere  will  rapidly  mix and
dilute the radon, making it  indistinguishable from ambient levels.
This method might  work well with  one or a  few homes, but what if
entire communities  installed these systems, thereby  introducing a
substantial volume of radon to the atmosphere through large numbers
of stacks?   The  purpose of this  paper  is  to test the hypothesis
that community-wide utilization of soil gas venting will not lead
to an  appreciable  increase in the  ambient  radon   levels nor  a
significant increase  in ambient risk.

                         HOUSING  DENSITY

In order to be realistic, census data giving a breakdown of housing
units in 4  suburban and 3 central city  census  tracts  in  a major
midwestern city,  along with the  area of the  tract,  were  used to
derive single family  housing densities  (Br92).   This density was
used to construct  a square,  area  emission  source  with a uniform,
rectangular, housing array.   Single story housing densities ranged
from about 250 structures per square kilometer in suburban areas to
about 1575  structures per square kilometer  in  the  central city,
with areas from 4 to  1.4 square kilometers, respectively.

                             ANALYSIS

The Atmospheric  Turbulence  and  Diffusion Laboratory (ATDL)  model
(Ha72)  was used to calculate the concentration of radon within and
in the area surrounding the  hypothetical community.   The community
(source) was  a  2000 meter square with  multiple emission points,
rectangularly arrayed,  equal  to  the  housing  density.   Receptors
were placed at the center of  2000 meter square grids surrounding
the area source.  No  sources were in the receptor grids.

The transit time of the radon from the soil  to the atmosphere is so
rapid that it was assumed that  only radon, no  decay products,  were
emitted  from the  emission points.    Drawdown  in  the soil  was
believed  not to  occur to  any  significant degree based  upon
discussions with  a major  mitigator-investigator (Si92)and an EPA
reseracher (He92) and, therefore,  was not considered.  Radioactive
decay during diffusion from the stack was  included.  Wind speeds
were set at 2 and 5 meters per  second with 2 m/s representing calm
conditions.  An actual wind rose  from a neighboring community was
used.  Calculations were made for each of the seven census tracts
at two wind speeds with the  highest and  lowest results reported in
this paper.

-------
                        RADON CONCENTRATION

A  normalized value  for the radon  emission  rate was  calculated.
Assuming  the low end of the range  for  radon  soil gas  is  about 100
picocuries  per  liter (pCi/L) (NCRP87)  and a common value  for fan
volumes is  about 200 cubic feet per minute  (cfm)  (He88),  then an
order of  magnitude emission rate is about 104 picocurie per second
 (pCi/sec) per house.  The emission rate range, over background, for
the  communities used was from  2.4  to  14.9 picocuries per square
meter per second (pCi/m2-sec).  This  compares with natural  soil
emission  rates  of 0.0054 to 1.89 pCi/m2-sec  (UNSCEAR88).

                              RESULTS

Ambient  concentrations,  above  background,  for  the  highest  and
lowest cases are shown  in  Figures  1  and 2.   With a  5 meter  per
second  (m/s)  wind speed the concentration is  about 57 picocuries
per  cubic  meter  (pCi/m3,   0.057  pCi/L)  at  the  community  and
diminishes  rapidly  from the  source.   Under calm conditions,  the
community concentration is  about  860  pCi/m3  (0.86 pCi/L).   For
other conditions, results  can  be  scaled.    For example,  with  an
elevated  soil radon concentration of 10,000 pCi/L (10'  pCi/m3,  100
times higher than the base calculation),  an effective  fan speed of
50 cfm (1/5 the base calculation),  and  a  radon  reduction  factor of
1/3  (from soil  to stack)  (Si92) the net  result is to multiply the
base  result  by 6.    This  could  lead to  an  elevated community
concentration of about 860  (6)  = 5160 pCi/m3  (5.16 pCi/L) for a
wind speed  of 2  m/s.

                            CONCLUSIONS

The  result  of  these calculations  indicate  that,  for low range
assumptions,  community  radon concentrations  would  be  elevated
slightly.    In radon  problem areas  where  these systems would  most
likely be installed, the resulting levels  could be  elevated  to
higher concentrations  based upon  local conditions and mitigation
system characteristics.  Inclusion of occupancy times, an  outdoor
to indoor reduction factor and an equilibrium factor would  lead to
a more  precise  determination of  the  community  impact.    Further
modification  of the estimation process  would be to  include  all
communities  of  the  city area.   This  would raise  the projected
ambient concentration.  Overall, it  would appear that installation
of  these  systems  in  radon  problem   areas  could  reintroduce
sufficient  radon to  the ambient air  so  that  the utility  of  the
mitigation system would be  offset.
The work  described  in this  paper was  not  funded  by  the  U.S.
Environmental Protection Agency and therefore the contents do not
necessarily  reflect the  views  of the  Agency  and no  official
endorsement should be inferred.

-------
                           BIBLIOGRAPHY

Br92      Brugman, B.L., 1992, Personnel correspondence

Ha72      Hanna, S.R. ,  1972, Description of ATDL Computer Model for
          Dispersion from  Multiple  Sources,  National Ocenaic and
          Atmospheric Administration, Oak Ridge, TN  (April)

He88      Henschel, D.B.,  1988,  Radon Reduction  Techniques for
          Detached  Houses,  Technical Guidance,   2nd   ed. ,  U.S.
          Environmental   Protection   Agency,   EPA/625/5-87/019,
          Research Triangle Park, NC  (January)

He92      Henschel, D.B.; Personnel communication; June 1992.

NCRP87    National   Council    on   Radiation   Protection   and
          Measurements, 1987,  Exposure of the  Population in the
          United  States   and   Canada  from  Natural  Background
          Radiation, NCRP Report No. 94,  Bethesda, MD (December)

Si92      Simon, R.; Personal communication; June  1992.

UNSCEAR88 United Nations Scientific Committee on the Effects of
          Atomic Radiation. Sources, Effects and Risks of Ionizing
          Radiation. United Nations, NY; 1988

-------
0.79
0
0.24
0.17
0.17
0
0.20
0
1.08
0.33
0.23
0.24
0.27
0
0.87
1.19
1.91
1.40
0.48
0.65
0.48
0.5
0.68
4.26
60.46
5.83
1.44
1.06
0.4
0.54
0.76
3.25
1.62
1.21
0.89
0
0.43
0.54
0.60
0.70
0.92
0
0.32
0
0.40
0.44
0.52
0
0.67
Figure 1: Lowest case: Radon concentrations (pCi/m3)
surrounding community for a wind speed of 5 m/s. Blocks are
2000 meters, center to center.

-------
15.11
0
4.53
3.17
3.32
0
3.78
0
20.64
6.19
4.33
4.54
5.16
0
16.62
22.70
36.38
26.56
9.10
12.38
9.16
9.52
13.00
81.13
1151.08
110.97
27.45
20.09
7.55
10.32
14.55
61.85
30.93
23.11
16.92
0
8.25
10.32
11.35
13.41
17.54
0
6.04
0
7.55
8.31
9.82
0
12.84
Figure 2: Highest case: Radon concentrations (pCi/m )
surrounding community for a wind speed of 2 m/s (calm
conditions). Blocks are 2000 meters, center to center.

-------
           Session VIII Posters
Radon Occurrence in the Natural Environment

-------
                                                                                  VIIIP-1
         INFLUENCE OF METEOROLOGICAL FACTORS ON THE RADON
               CONCENTRATION IN NORWEGIAN DWELLINGS
                       Terje Strand and Nils H. Bohmer
                    National Institute of Radiation Hygiene
                        P.O.Box 55, N-1345 Osterfie
                               NORWAY
                              ABSTRACT
Owing to variations in  different meteorological factors, ventilation conditions and
human behaviour, the radon level in indoor air can vary by more than an order of
magnitude over a few days period. In Norwegian houses, the radon concentration is
usually much higher In  the winter season, when the temperatures are well below
freezing point, than in the summer. In our etudy, short-term and long-term variations
in the radon concentration, and influencing factors, were investigated. Over a period
of one year an extensive measurement program were undertaken in four typical single
family houses with elevated levels of radon (between 1500 - 4000 Bq/mJ). In one of
these houses, the level  in  the summer was found to be about twice the level In the
winter and the  variation pattern was almost  opposite to the other three houses.
Between February 1987 and March 1989. measurements of radon in Indoor air were
made in a total of 7500 randomly selected dwellings from all parts of Norway. These
data were correlated with meteorological data from different parts of the country for
the some period. The results of these studies are reported in this  paper.

-------
                                                                                   VIIIP-2
            SOIL RADON POTENTIAL MAPPING AND VALIDATION
                            FOR CENTRAL FLORIDA

            by:    K. K. Nielson and V. C. Rogers
                   Rogers and Associates Engineering Corporation
                   Salt Lake City, Utah 84110-0330

                   R. B. Brown and W. G. Harris
                   University of Florida Soil Science Department
                   Gainesville, Florida 32611

                   J. K. Otton
                   U.S. Geological Survey, MS-939
                   Denver, Colorado 80225
                                   ABSTRACT

      Maps of soil radon potentials are being developed to provide a possible geographic
basis for implementing radon-protective building construction standards in Florida.  The
maps are being developed from soil properties, independent of institutional boundaries or of
particular present radon limits.  The radon potentials are defined from the calculated rates
of radon entry into a hypothetical house modeled over soil profiles for each map polygon. Two
approaches are tested in prototype radon maps developed for Alachua  County.  The first
approach defines soil profiles and associated transport properties (air permeability, radon
diffusion, moisture) from county soil survey data for 65 map units occurring in 15,000 map
polygons.   Radon source strengths are based  on 323  radium  and  radon emanation
measurements, averaged by geologic unit, from archived samples collected at reference pedon
sites in the soil survey. The second approach defines soil profiles and transport properties
from the state-wide Statsgo soil maps and data, which include 30 map units occurring in
approximately 200  polygons  in Alachua County.  The second  approach uses NURE
aeroradiometric data and geologic classifications to define individual radon source strengths
for each map polygon.

      The two approaches were compared with ground-truth data that included surface
radon fluxes, soil-gas radon concentrations, and indoor radon data.  The mapped soil radon
potentials were related most precisely to the radon fluxes, followed by  the soil gas radon
concentrations.  Indoor  radon data exhibited more variation, as expected, due to house
variability. A mapping precision of ± 1 tier in a four-tier approach appears possible.  The
higher resolution of the first mapping approach did not demonstrate  any advantage in
mapping precision when comparing the maps with ground truth data.  The second approach
is proposed for mapping broader areas of the state because it utilizes data and samples that
are already available throughout most of the state.

-------
       This project is funded by the U.S. Environmental Protection Agency (EPA) and the
Florida Department  of Community Affairs  (DCA).  This  paper has been reviewed in
accordance with the U.S. Environmental Protection Agency's peer and administrative review
policies and approved for presentation and publication.
                                 INTRODUCTION


BACKGROUND

      Radon (222Rn) gas is formed by decay of radium (226Ra), which occurs naturally in
virtually all soils at varying concentrations. Radon formed in the top few meters of soil can
enter indoor environments  through pores and cracks in building foundations. If sufficient
radon enters a building and if its dilution by outdoor air is small, it can accumulate to levels
that pose significant risks of lung cancer with chronic exposure. Indoor radon levels in the
U.S. average about 1 pCi L"1, with approximately 1 to 3% of homes exceeding 8 pCi I/1 (1).
The U.S. Environmental Protection Agency (EPA) attributes 5,000 to 20,000 lung cancer
deaths per year to indoor radon exposure, and suggests that levels averaging 4 pCi L"1 or
higher warrant remedial action (2).

      The Florida Department of Community Affairs (DCA) is developing radon-protective
building standards for new construction that will help reduce public health risks from indoor
radon (3,4). The standards are to be integrated into the state-wide uniform building code,
but may add  an incremental cost for  constructing new buildings when certain radon-
protective measures are required.  To minimize economic burdens and still provide the
intended health protection, requirements for extra-cost radon protective measures can be
based on the potential of the building site for causing elevated indoor  radon accumulation.
 Although  elevated indoor radon occurrences are  highly  variable, regional trends and
geographic clustering (1,5-6) suggest the possibility of defining geographic criteria for certain
radon-protective construction requirements.

      State-wide mapping of radon potentials in Florida has been proposed  as one means
of estimating  regional needs for radon-protective construction  features (7).  The radon
potential maps assume that soils are the primary source of indoor radon, a condition that is
generally acknowledged  (2). The specific relationships between indoor  radon levels and soil
properties are complicated and often unclear, however, despite a sound theoretical basis and
a large body of empirical evidence (8-26).

      Soil radon potentials primarily depend on soil radium concentrations, radon emanation
fractions, soil moisture, air permeability, diffusivity, and density. The exact relations to these
parameters are obscured by the frequent variations in indoor pressures and house ventilation
rates.   Other, invariant  house parameters  such  as floor  properties and foundation
construction details also affect the relation between indoor radon and  soil radon potential.
Despite the dependance  of indoor radon on static and dynamic house properties as well as
soil properties, it is possible to partition the effects of soil properties for mapping soil radon
potentials.

-------
      Numerous radon maps have been compiled previously, as reviewed in a DCA-EPA
radon mapping workshop and related feasibility studies (7,27).   The maps have mainly
presented empirical correlations of indoor radon measurements or related parameters with
various institutional units such as state, county, or township boundaries, ZIP Code areas, or
occasionally geologic or physiographic regions. They most commonly present multi-tiered
geographic classifications of areas correlated with indoor radon concentrations.  Numerical
radon indices and  other,  surrogate parameters related to radon potential also have been
mapped, including  aeroradiometric  gamma activity, uranium mineralization zones, and
surface outcrop areas of geological formations with elevated radon potential. Although these
approaches indicate where elevated radon has been observed, they tend to be indirect or
imprecise predictors of indoor radon for new construction, and they are difficult to relate to
the needs for or results of using radon-protective construction features.  Maps aimed at
optimizing testing programs or locating areas of highest observed indoor radon are already
available for Florida (28).
OBJECTIVES

       The mapping approach  for implementing radon-protective building standards in
Florida differs from previous mapping efforts. The new maps concentrate on radon source
potentials of soils to satisfy the basic objectives for the DCA  radon-protective building
standards. The radon mapping objectives include:

       •  Identify as precisely  as possible regions that require radon-protective
          building features to attain prescribed indoor radon concentrations.

       •  Avoid political and institutional boundaries that are  unrelated to radon
          potential.

       •  Avoid restrictive association to a particular radon standard (i.e., 4 pCiL ).

       •  Minimize uncertainties related to variations in time, house design, and
          occupancy.

       To attain these objectives, an approach was developed to separate soil radon  source
 parameters from other parameters affecting indoor radon concentrations. A numerical radon
 transport and entry model was used to compute radon source  potentials for  soil profales
 occurring throughout the regions to be mapped. The soil profiles are  defined from Soil
 Conservation Service (SCS) soil survey data.   Maps were developed by attributing the
 calculated radon potentials to the areas represented by each soil  profile. An initial, detailed
 mapping approach was compared with an alternative approach that provided more complete
 coverage throughout Florida.   Field measurements of radon  fluxes  and soil gas radon
 concentrations were conducted to evaluate the mapping approaches. Indoor radon data from
 a prior land-based survey (28) and a University of Florida survey (29) also were compared
 with the maps of calculated soil radon potentials. This paper presents an interim report on
 the prototype mapping effort, which presently is in progress.

-------
                                    THEORY
PARTITIONING OF RADON SOURCE PARAMETERS

      Indoor radon distributions are subject to variations in source parameters, to variations
in house (and occupancy) parameters, and to time variations of these parameters. Although
indoor radon concentrations are the ultimate concern in this mapping program, it is their
geographic variability, independent of house and time variations, that is  of particular
interest.   The large uncertainties normally associated with populations of  indoor radon
measurements therefore can be reduced by considering only the geographic distribution of
radon source parameters. Time variations are eliminated by using only invariant or long-
term average parameters in developing the maps. House variations similarly are eliminated
by using average or typical parameters to represent a constant, reference  house for all
geographic locations. The reference house, a slab-on-grade single-family dwelling, is defined
and modeled for radon entry calculations using soil profile properties at each source location.
The source variations throughout the county or state thus are characterized from calculated
radon entry rates, independent of actual house variations. Once the distribution of radon
source potentials is determined, the additional variations due to house variability also can
be assessed by statistical comparisons with indoor radon data. Actual house parameters are
only required to relate the potentials to particular indoor radon concentrations.  This study
addresses only the radon source characterization.
RADON ALGORITHM

      The algorithm to compute radon entry into the reference house on each soil profile is
developed  using  the  unified-theory  representation of multi-region, multi-phase radon
generation and transport by both advection and diffusion (30). The algorithm is implemented
in 2-dimensional numerical calculations by the RAETRAD model (RAdon Emanation and
TRAnsport into Dwellings,  31).   The steady-state  radon balance equation  solved by
RAETRAD for the map calculations is :

                V.faDV(Cb/fs)  - V.(K/uX(yfs) VP  -  XCb +  RpXE  = 0              (1)

where

      V      =   gradient operator
      fa     =   p(l-S+SkH)
      p     =   soil porosity (dimensionless: cm  pore space per cm bulk space)
      S     =   soil water saturation fraction (dimensionless)
      kH    =   222Rn   distribution   coefficient   (water/air)  from   Henry's   Law
                 (dimensionless)
      D     =   diffusion coefficient for 222Rn in soil pores (cm2 s"1)
      Cb    =   fsCa = 222Rn concentration in bulk soil space (pCi cm"3)
      C     =   222Rn concentration in air-filled pore space (pCi cm"3)

-------
      f.     =   p(l-S+SkH)+pka
      p     =   soil bulk density (g cm"J, dry basis)
      ka    =   k^extf-bS)                                   .
      k °    =   dry-surface adsorption coefficient for    Rn (cm  g )
      b*    =   adsorption-moisture correlation constant (g cm )
      K    =   bulk soil air permeability (cm )
      ji     =   dynamic viscosity of air (Pa s)
      VP    =   air pressure gradient (Pa cm" )
      X     =   222Rn decay constant (2.1xlO"6 s"1)
      R    =   soil 226Ra concentration (pCi g"1)
      E    =   total 222Rn emanation coefficient (air + water) (dimensionless).

      This equation applies to gas-phase advective transport of radon, and to combined gas-
phase and liquid-phase diffusive transport of radon. The combined-phase diffusive transport
is characterized by appropriate moisture- and porosity-dependent values of the pore-average
diffusion coefficient, D (32,33).  This approach is important to correctly characterize radon
diffusion in unsaturated soil pores that may have small intermittent water blockages, but
that still may transmit significant radon  flux (32,34).  Liquid-phase advective transport of
radon is not addressed because it typically is negligible. The radon fluxes between different
soil layers and at the top surface are calculated as

                           F =  -D fa  VCa + (K/u) VP Ca                         (2)

where

      F     =  bulk flux of 222Rn (pCi cm"2 s"1).

       A 2-dimensional form of equation (1) is used in modeling the reference house with
elhptical-cylindrical symmetry over the specific soil profiles that are being mapped (Figure
1)  Two-dimensional modeling has been found previously to adequately represent houses for
calculating indoor radon entry. Soil horizons defined by county soil survey date are used to
define the vertical soil profiles. Radial uniformity of the soils is assumed. The house is
modeled to have a foundation crack near its perimeter for permitting advective transport of
radon by pressure-driven flow. It also permits radon transport through the foundation slab;
however this transport is dominated almost completely by diffusion because of the low air
permeability of concrete. To approximate rectangular house geometry, a skewing factor is
 applied to the radial gradient term that results from the 2-dimensional gradient operator in
 Equation (1).

-------
                     Concrete
                       Floor
Pressure-Driven
  Air Flows;
Advective Radon
  Transport
                                           Radon Gas Diffusion
                                                          5 m
                         House Center Line
                        (symmetry assumed)
                                                        RAE-103688
      Figure 1.   Cylindrically-symmetric  house  and  foundation   soil
                   profiles used to model radon entry rates.
                          RADON MAPPING METHODS
MAPPING APPROACHES

      Two radon mapping approaches (A and B) were devised and tested for the Florida
radon mapping program.  Approach A was developed first (35,36), using a digital, high-
resolution (1:24,000) county soil survey map, intersected with geology map units, to define
individual  radon map polygons.   A total of 65 map units were defined to  occur in
approximately 15,000 polygons in the Alachua County map.  Radon source potentials were
calculated for the area represented by each polygon with the numerical model using existing
soil  physical  and  hydrologic  data to define the  radon  transport  properties.   New
measurements of soil radium concentrations and radon emanation coefficients, averaged by
geologic unit, were used to define radon source strengths of the soils.

      Approach B was developed afterward to utilize lower-resolution digital Statsgo  soil
maps (37) that are available with broader coverage over nearly all of Florida. Radon map
polygons for Approach B were defined from the intersection of the Statsgo polygons with
geology map polygons. Thirty map units were defined to occur in approximately 200 polygons
for the Alachua  County  map.   Radon source potentials for  Approach B used NURE
aeroradiometric  data (equivalent uranium),  averaged  over  each radon map polygon, to
estimate average (equilibrium) radium concentrations.  Radon emanation coefficients were
estimated as a function of radium concentration from the new measurements.

-------
      For both approaches, radon source potentials were calculated as the rate of radon
entry into the hypothetical reference house that was modeled on the soil profiles of each map
unit. The radon entry rates were expressed in long-term units (mCi y" ) to emphasize their
long-term average nature. The radon potentials then were grouped into four tiers of similar
numerical values for color-coded display of the soil radon potentials associated with each map
polygon.  The soil radon potentials can be related to indoor radon concentrations for the
reference house using a simple ventilation-rate model,

                                 C =  114Q/(VhXh)                              (3)

where

      C      =  average indoor radon concentration (pCi L" )
       114   =  unit conversion (pCi L"1 h"1 per mCi m^ y"1)
       Q     =  average radon source potential (mCi y" )
       Vh    =  house volume (m3)
       Xh    =  house ventilation rate (h" ).

       The reference house used to represent Florida housing with Approach A was described
previously (35,36).  The reference  house used with Approach B was nearly identical, and
consisted of a 28 ft. x 54 ft. rectangular slab-on-grade house with nominal dimensions and
characteristics as summarized in Table 1.  Its volume was based on that of a median U.S.
family dwelling (38), and is similar to that of typical Florida houses (39).  A nominal 8 ft
(2.4m) ceiling height was used to estimate its area, which also is similar to  other estimates
of Florida floor slab areas (39). Its ventilation rate corresponds  to the nominal median U.S.
house  ventilation  rate (38), although even lower values sometimes  have been used  to
represent Florida houses (39). The floor crack location is chosen near the slab perimeter to
approximate a slab/footing shrinkage crack. The stem wall footing is assumed to be 3 ft (91
cm) deep, penetrating 2 ft (61 cm) into the natural terrain. A  1 ft (30 cm) layer of fill soil
beneath the floor slab is comprised of material identical to the surrounding surface soil. The
indoor pressure is typical of that used previously to generically model indoor thermal and
wind-induced pressures in U.S. houses (40), and to represent Florida housing in particular
(39).  Concrete slab  permeabilities and diffusion coefficients were estimated from data
measured on Florida floor slabs (41).


    TABLE 1. NOMINAL VALUES OF PARAMETERS USED IN RADON ENTRY
                 CALCULATIONS FOR THE REFERENCE HOUSE
    House Area             143 m2         Fill Soil Thickness              30 cm
    House Length/Width     1.9 (ratio)      Indoor Pressure                -2.4 Pa
    House Volume          350 m3         Concrete Slab Thickness         10 cm
    House Ventilation Rate   0.5 h-1         Concrete Slab Porosity          0.22
    Floor Crack Width       0.5cm         Concrete Slab 226Ra*Emanation   0.07 pCi g
    Floor Crack Location     slab perimeter  Exterior Footing Depth          61 cm ^    ^
    Crack Area Fraction     0.002          Concrete Air Permeability        1x10'   cm
                                        Concrete Rn Diffusion Coeff.      8x10"* cnT s"

-------
SOIL PHYSICAL PARAMETERS

      Layers of the soil profile for each map unit were defined for modeling throughout the
top 5 m. Soils in the top 2.0 to 2.5 m were characterized explicitly from SCS soil survey data,
which typically represented the various occurrences of the A, E, B, and C soil horizons and
their subdivisions in about six layers.  An additional layer was defined beneath the SCS-
characterized layers to represent the remaining deeper soils to  a depth of 5 m.  This layer
was defined either as an extension of the lowest layer from the SCS-characterized horizons
or as a layer of the Hawthorne formation. The selection of these deeper layers was made
according to a geologic map of surface occurrences of the Hawthorne Formation in Alachua
County.  Horizontal uniformity was assumed in the distributions of the radon source and
transport parameters in the vicinity of the reference house.

      The parameters used to define each soil layer included the soil density, porosity, water
content,  radon diffusion  coefficient, air  permeability, radium concentration, and radon
emanation coefficient.  Most of the physical parameters were defined or derived from data
in the Alachua County soil survey report (42) or from more detailed data files maintained by
the University of Florida Soil Science Department.   Soil densities (bulk, dry-weight basis)
were taken directly from the soil survey report for each horizon in each soil map unit. Soil
porosities were calculated from soil density and specific gravity as

                                    P  = 1 - P/Pg                                 (4)

where
                                                      o
      pg      =   soil specific gravity (nominally 2.7 g cm ).

      Soil water contents were estimated from soil water drainage data in the Alachua
County soil survey (42).  For Approach A, the water contents were estimated as drained,
field-capacity values corresponding to -0.1 bar matric potential for sands, and to -0.33 bar
matric potential for clays (36).  Later theoretical water balance calculations with the
FEMWATER code (43,44) indicated that sub-slab matric potentials are well-approximated by
the position above the water table for wet climates and shallow water tables as in  most of
Florida.  Therefore the water contents for Approach B were interpolated from the soil survey
water drainage data at matric potentials corresponding to the height above the water table.
Water table depths were computed as  time-weighted averages from the high water table
limits and durations reported in the soil survey (42), and deeper limits for other seasons  as
documented with Approach A (36). Soil water contents in volume percent and weight percent
units in the soil survey report were converted to fractions of saturation as:

                             100 S = My/p =  pMw/p                          (5)

where

             =  soil water content (volume percent)
      Mw   =  soil water content (dry  weight percent).

-------
      Soil radon diffusion coefficients were estimated from the water contents and porosities
of the soils using a predictive correlation that is based on 1073 laboratory measurements of
radon diffusion in recompacted soils at moistures ranging from dryness to saturation (33).
The soil textures ranged from sandy gravels to fine clays, and their densities covered the
range of most of the Florida soil densities.  The correlation exhibited a geometric standard
deviation (GSD) between measured and calculated values of 2.0, and had the form

                            D =  Dj) exp(-6Sp - 6S14p)                          (6)

where

      D      =   diffusion coefficient for 222Rn in soil pores (cm2 s'1)
     D0      =   diffusion coefficient for 222Rn in air (l.lxlO'1 cm2 s'1).

      Soil air permeabilities were estimated similarly from the water contents, porosities,
and grain diameters of the soils using a predictive correlation that was based on more than
a hundred in-situ field measurements of soil air permeability, including measurements in
Florida (33).  This correlation exhibited a GSD between measured and calculated values of
2.3, and had the form

                          K = 104 (p/500)2 d4* exp(-12S4)                       (7)

where

      K      =   bulk soil air permeability (cm )
      d      =   arithmetic mean soil particle diameter, excluding >#4 mesh (m).


SOIL RADIOLOGICAL PARAMETERS

      Soil radium concentrations and radon emanation coefficients used with Approach A
were  obtained  from new  226Ra measurements on 323 samples  and  radon  emanation
measurements on 131 samples. Most of the emanation measurements were made on samples
with 226Ra concentrations  exceeding 1 pCi g"1, since these provided the best measurement
precision. More than 280 of the soil samples were obtained from an archive  of samples
collected at the reference pedon sites by SCS during the original Alachua County soil survey.
The remaining samples were obtained from new borings in May 1991 by the U.S. Geological
Survey and and Florida Geological Survey at selected supplementary sites in Alachua
County.  Individual results of  these analyses were reported previously (36).  Radium
concentrations exhibited a geometric mean of 0.8 pCi g"1 and a geometric standard deviation
of 3.3. Radon emanation coefficients had an arithmetic mean and standard deviation of 0.48
± 0.16.

      Soil radium concentrations in Approach B were estimated for the top 2.5 m layer from
the mean equivalent uranium concentration as averaged for each Statsgo map unit from
NURE aeroradiometric data (45).  Equilibrium was assumed between the  uranium and
radium activities. A weak correlation observed between the measured radium concentrations

-------
and emanation coefficients (Figure 2) was used to estimate emanation coefficients from the
radium concentrations for Approach B.  The relationship from Figure 2 was approximated
as
                     E = min(0.15R + 0.20, 0.55), R<8 pCi g'1

                     E = 0.50,                   R>8 pCi g -1
               (8)
Radon Emanation Coefficient
D p p p p -
D M ** b> b» c


(
* 4
•
7
•
i
:


& i
>m
*:
B
i
4 ^
% '



• 1


•

1
m •



i i
•
•
<



•
• i
•
•


..










•





•:

                0123456      10       20      30

                           Radium-226 Concentration (pCi/g)
40
                                                            RAE-103638
      Figure 2.   Relationship  between   measured  radon   emanation
                  coefficients and radium concentrations.
                VALIDATION MEASUREMENTS AND RESULTS

      Mapped soil radon potentials were compared with two sets of empirical ground-truth
measurements and two sets of indoor radon measurements to estimate the degree of success
in mapping particular land areas in Alachua County and to compare mapping Approaches
A and B.   The comparisons were made  by locating each measurement location on the
Approach A or Approach B maps  from their latitude/longitude coordinates,  and then
associating  the color-tier of the map location with the numerical measurement. Each of the
four map color tiers was assigned to an integer (l=blue, 2=green, 3=orange, 4=red), which
was used to estimate mapping precision. The measurement data and map tiers  from each
set were then sorted by measurement, and the number of measurements in each map tier
was assigned successively to the sorted data set.  The resulting tier classification errors then
were computed by subtracting for each measurement the sorted tier assignment from the
actual tier definition.  This approach defined the maximum tier classification precision for
each data set. The results of these analyses were plotted as frequency histograms, and the
standard deviation of each set of tier classification errors was computed to estimate overall
precision of the approach.

-------
      To account for potential imprecision in map polygon line positions, a 500 meter fringe
was imposed in ranked order on the polygons of alternative versions of the Approach A and
Approach B maps. The fringe was applied first to extend the size of the polygons of greatest
radon potential (red), followed next by orange, and then green. The fringe could expand the
size of a polygon only  into an area occupied by a lower-order color. In this way, points of
elevated potential were included in a higher tier if they were within 500 m of its boundary.

      The first ground-truth data set was collected by USGS personnel in May 1991, and
consisted of 66 soil gas radon measurements at 1 m depth at different locations throughout
Alachua County. Additional measurements also were collected at other depths, but only the
1-m measurements were analyzed here to assure consistent  comparisons  among different
locations.  The detailed methods and results of this study are reported separately (46).

      The second ground-truth data set was collected in April 1992 by Rogers & Associates
personnel, and consisted of 81 surface radon flux measurements.  The measurements were
made in triplicate at 10 m spacing at each of 27 sites throughout Alachua County. Sampling
was conducted over a 24h period with small charcoal canister radon samplers, and analyses
were  conducted on the  sealed, retrieved samplers by spectrometric gamma assays.  The
procedure for these measurements and a comparison of the procedure with EPA method 115
(47) has been reported previously (48).

       The first set of indoor radon measurements used for comparisons with the radon
mapping tiers was taken from the state-wide land-based radon survey conducted in Florida
in 1986 (28).  The radon measurements utilized charcoal canister radon samplers, and were
located in 102  areas throughout Alachua  County.   The  second set of indoor radon
measurements was conducted by University of Florida (29), and utilized alpha track radon
detectors in 35 homes throughout Alachua County.

       The precisions of  the tier  classifications for mapping Approaches A  and B are
illustrated in Figure 3 for the 81 individual radon flux measurements.  Conservative
(positive) errors resulted from including low radon measurements in a high radon-potential
tier, and non-conservative (negative) errors resulted from including high radon measurements
in a low radon-potential tier. As expected, the 500 m fringe made the classifications slightly
 more conservative for both mapping approaches. Although Approach B exhibited more data
 points with correct classifications, it also had slightly more non-conservative errors at the -2
 level Numerical summaries of the data in Figure 3 (means and standard deviations of each
 distribution) are presented in Table 2. They indicate that Approaches A and B have similar
 precisions for the basic map definitions, but that the 500 m fringe improves the precision for
 Approach B and degrades it for Approach A.

-------
      TABLE 2. MEANS AND STANDARD DEVIATIONS OF MAP TIER
                CLASSD7ICATION ERROR DISTRD3UTIONS









a
b
c
d
Data Set
RnFluxa
RnFluxa
Rn Flux meansa
Rn Flux means8
Soil Gas Radonb
Indoor Radonc
Indoor Radond
Indoor Radon
This study.
Otton et al., (46).
Nagda et al., (28).
Roessler, (29).
Approach
A
B
A
B
B
B
A
B

n
81
81
27
27
66
102
35
35

No Fringe
0.00 ± 1.21
0.00 ± 1.22
0.00 ± 1.11
0.00 ± 1.11
0.00 ± 1.37
0.00 ± 1.22
0.00 ± 1.39
0.00 ± 1.73

Fringe
0.59 ± 1.43
0.07 ± 1.16
0.59 ± 1.42
0.07 ± 1.03
0.12 ± 1.23
0.29 ± 1.27
—
~~~

00
to
I 50
^ 45
« 40
(0
| 35
x 30
u.
c 25
o
"8 20
a:
"o 15
S 10
1 s
Z n
n s 81 measurements



Non-Conservative
m


j
P

l_ '•
fc :
Hi!

/
" ^
/
^
s
j
r*
$ C3 A, No Fringe
\ 0 A, 500m Fringe
^ • B, No Fringe
\ 0 B.SOOm


:ringe

# Conservative
^ _— i- •s^


s
t
f
s
/ R^_
? I/
/ ', #
t ',
f : S
/ '.
/ '. S


1
^^





jp
^
1







/
,/
P'
Number of Radon Flux Measurements
3 O1 O Ul O

n = 27 measurement
locations
|: /
Non-Conservative ;j


'', ED A, No Fringe
0 A, 500m Fringe
• B, No Fringe
E2 B, 500m Fringe
Conservati
ve
/
/
(
-

      -3-2-10123
            Tier Classification Error
                             RAE- 1040W

Figure 3.  Map tier precisions from
          comparisons with individual
          radon flux measurements.
       -3-2-101    2
             Tier Classification Error
   3

RAE- 1040U
Figure 4.  Map tier precisions from
          comparisons with means of
          radon flux measurements.

-------
      Identical analyses were performed for the means of the radon flux data at each of the
27 sites to test the effects of local heterogeneity and the possible benefits of replicate
sampling.  Figure 4 summarizes the analyses with averaged radon flux data, which exhibit
slightly better precision in all cases. The numerical summaries in Table 2 indicate standard
deviations of 1.11 compared to about 1.21 for the individual measurements.  Including the
500 m fringe for  the averaged data made little  difference for the Approach A  map, but
reduced the standard deviation for averaged data with Approach B to 1.03 compared to 1.16
for the individual measurements.

      The precision of the tier classifications for  the USGS soil  gas  radon data is
summarized in Figure 5 for Approach B. The 500 m fringe again caused a slight conservative
shift in the distribution. The numerical summary in Table 2 suggests larger tier  errors for
the soil gas radon measurements than for the radon flux measurements. This is consistent
with theoretical calculations,  which suggest that  radon fluxes provide  a more  precise
indicator of soil radon potential (as computed here with the RAETRAD model) than soil gas
radon concentrations.

      The precisions of tier classifications for the Geomet indoor radon data with  Approach
B are summarized in Figure 6.  As expected,  they exhibit greater tier errors due  to the
contributions of house variability, along with variations in soil radon potential. The 500 m
fringe again causes a slightly conservative shift; however the numerical summaries in Table
2 indicate that the fringe slightly reduces the mapping precision.
 Ł
 3
 (0
 CO
 -8
 (0
 QC
 (0
 m
 O
 "5
 CO
       n - 66 measurements
          afl m depth
  c
  
-------
      The precisions of the tier classifications for the University of Florida indoor radon data
with Approaches A and B are summarized in Figure 7. They also exhibit greater tier errors
then for radon fluxes and soil gas radon due to the contributions of house variability. The
numerical summaries of these indoor radon data (Table 2) indicate greater variation than for
the Geomet data set, particularly for Approach B. This may result in part from the smaller
number of houses in this set or from location differences between the two sets.
                             -3-2-10    1     2    3
                                   Tier Classification Error _._ ,„„.
       Figure 7.   Map tier precisions from comparisons with University of
                   Florida indoor radon measurements.
                                   DISCUSSION

       The present analyses suggest that the lower-resolution Approach B for mapping soil
radon potentials does not significantly degrade mapping precision compared to Approach A.
This is probably due in part to the improved method for defining soil radium concentrations
from  area-specific aeroradiometric data instead of relying on soil classifications to define
radiometric similarities among  soils  throughout  the  county.  It is still recognized  that
localized anomalies probably occur within many map polygons, and that these will be largely
undetectable using the NURE aeroradiometric data (6-mile flight-line spacing, (45)). Based
on the limited ground-truth comparisons made to date, however, mapping Approach B may
provide maps with a tier classification uncertainty of approximately one tier if a four-tier
classification system is used.

       The development of soil radon potential maps from soil (Statsgo) data, Geologic maps,
NURE radiometric data, and soil radon emanation measurements is presently in progress for
twelve counties in central Florida. The maps will utilize Approach B for consistent coverage
of parts of the state where digitized high-resolution soil maps are unavailable. The maps will
be displayed as soil-defined polygons with a colored-tier system to illustrate general areas of
high  and low soil radon potential. For more complete interpretation, numerical values of the

-------
calculated soil radon potentials will be displayed or provided for each map polygon.  These
can be used to estimate indoor radon levels for the reference house, or for other specific house
conditions if more detailed house parameters are available. For the reference house, the
indoor radon concentration is estimated from the simple ventilation-rate model in equation
(3), which gives a radon concentration of 4 pCi L"1 for a soil radon potential of about 6 mCi
y*1.  Statistical estimates  of variability within each polygon will be available from the
standard deviations of the NURE-based radium estimates. These may potentially provide
a basis for estimating confidence intervals on the radon potentials for the map polygons.
                                 REFERENCES

1.    Nero, A. V., Schwehr, M.B.,Nazaroff,W.W.,andRevzan, K.L. Distribution of Airborne
      Radon-222 Concentrations in U.S. Homes. Science 234:992-997; 1986.

2.    EPA. A Citizen's Guide to Radon. Washington D.C.: U.S. Environmental Protection
      Agency and U.S. Department of Health and Human Services; report OPA-86-004;
      1986.

3.    Sanchez, D.C., Dixon, R, and Williamson, A.D. The Florida Radon Research Program:
      Systematic Development of a Basis for Statewide Standards, proceedings of the 1990
      International Symposium on Radon and Radon Reduction Technology. Atlanta, GA:
      February 1990.

4.    SBCCI.  Florida Code for Radon-Resistant Construction and Mitigation," Southern
      Building Code Congress International, Inc., Rev. 1, January 1990.

5.    Cohen, B.L.  A National  Survey of 222Rn in U.S. Homes and Correlating Factors.
      Health Physics. 51:175-183; 1986.

6.    Peake, R.T., Gundersen, L.C.S., Ronca-Batista, M., James, J., and Schumann, R.R.
      Determination of Radon Geologic Provinces in the United States, proceedings of the
      1990 International Symposium on Radon and Radon Reduction Technology. Atlanta,
      GA: February 1990.

7.    Nielson,  K.K. and Rogers, V.C.  Proceedings of the Florida Workshop on Radon
      Potential Mapping.  Salt Lake City, UT:  Rogers & Associates Engineering Corp.;
      report RAE-8945-4; 1990.

8.    Brookins, D.G.  Indoor and Soil Rn Measurements in the Albuquerque, New Mexico
      Area.  Health Physics 51:529-533; 1986.

9.    Peake, R.T. and Hess, C.T.   Radon and Geology: Some  Observations," in Trace
      Substances in Environmental Health XXI, D. D. Hemphill, ed., pp. 186-194, Columbia:
      Univ. of Missouri, 1987.

-------
10.    Sextro, R.G., Moed, B.A., Nazaroff, W.W., Revzan, K.L., and Nero, A.V. Investigations
      of Soil as a Source of Indoor Radon. In: Radon and Its Decay Products.  P. K. Hopke
      ed., Washington, DC: Am. Chem. Soc., ACS Symposium Series 331, pp. 10-29, 1967

11.    Gundersen, L.C.S., Reimer, G.M., and Agard, S.S.  Correlation Between Geology,
      Radon in Soil Gas, and Indoor Radon in the Reading Prong,  in Geologic Causes of
      Natural Radionuclide Anomalies.  M. A. Marikos and R. H. Hansman, eds., Missouri
      Dept. of Nat. Res. Spec. Publ. 4, 91-102, 1988.

12.    Gundersen, L.C.S., Reimer, G.M.,  Wiggs, C.R., and Rice, C.A.  Radon Potential of
      Rocks and Soils in Montgomery County, Maryland. U.S. Geological Survey map MF-
      2043; 1988.

13.    Nazaroff,  W.W., and Nero, A.V.   Radon  and Its  Decay Products in Indoor Air.
      Chapters 1 and 12; New York: Wiley & Sons; 1988.

14.    Nazaroff, W.W.,  Moed, B.A., and Sextro, R.G.  Soil as a Source of Indoor Radon;
      Generation, Migration and Entry.  In: Radon and Its Decay Products in Indoor Air,
      W.W. Nazaroff and A.V. Nero, eds., New York: Wiley & Sons, p. 60-65; 1988.

15.    Otton,  J.K, Schumann, R.R., Owen, D.E., Thurman, N. and Duval, J.S.  Radon
      Potential of Rocks and Soils in Fairfax County, Virginia. U.S. Geological Survey map
      MF-2047, 1988.

16.    Buchli, R. and Burkart, W. Influence of Subsoil Geology and Construction Technique
      on Indoor Air 222Rn Levels in 80 Houses of the Central Swiss Alps.  Health Physics
      56:423-429; 1989.

17.    Duval, J.S., Otton, J.K, and Jones, W.J. Estimation of Radon Potential in the Pacific
      Northwest Using Geological Data.  Portland: Bonneville Power Administration, 1989.

18.    Kunz,  C.,  Laymon,  C.A.,  and Parker, C.   Gravelly  Soils and  Indoor Radon,
      proceedings of the 1988 Symposium on Radon and Radon Reduction Technology, Vol.
      1, EPA600/9-89/006a (NTIS PB89-167480),  paper. 5-75; 1989.

19.    Muessig, K.W. Correlation of Airborne Radiometric Data and Geologic Sources with
      Elevated Indoor Radon in New Jersey,  proceedings of the 1988 Symposium on Radon
      and Radon Reduction Technology, Vol. 1, EPA60019-89/006a (NTIS PB89-167480),
      paper 5-1; 1989.

20.    Reimer, G.M. and Gundersen, L.C.S.  A Direct Correlation Among Indoor Rn, Soil Gas
      Rn and Geology in the Reading Prong Near Boyertown, Pennsylvania. Health Physics
      57:155-160; 1989.

21.    Smith, D.L. and Hansen, J.K  Distribution of Potentially Elevated Radon Levels in
      Florida Based on Surficial Geology.  Southeastern Geology 30:49-58; 1989.

-------
22.    Yokel, F.Y.  Site Characterization for Radon Source Potential.  Washington DC:
      National Institute of Standards and Technology report NIST-IR 89-4106; 1989.

23.    Gregg, L.T.  and Coker, G.  Geologic Controls on Radon Occurrence in Georgia,
      proceedings  of the 1990 International Symposium on Radon and Radon Reduction
      Technology.  Atlanta, GA: February 1990.

24.    Laymon, C.A.  and Kunz, C.  Geologic Factors  and House Construction Practices
      Affecting Indoor Radon in Onondaga County, New York, proceedings of the 1990
      International Symposium on Radon and Radon Reduction Technology. Atlanta, GA:
      February 1990.

25.    Otton, J.K.  and Duval, J.S.   Geologic Controls on Indoor Radon in  the Pacific
      Northwest, proceedings of the 1990 International Symposium on Radon and Radon
      Reduction Technology. Atlanta, GA:  February 1990.

26.    Schumann, R.R., Peake, R.T., Schmidt, K.M., and Owen, D.E. Correlations of Soil-Gas
      and Indoor  Radon with Geology in Glacially Derived Soils of the  Northern Great
      Plains, proceedings of the  1990  International  Symposium  on Radon  and Radon
      Reduction Technology. Atlanta, GA:  February 1990.

27.   Nielson, K.K.  and  Rogers, V.C.   Feasibility  and Approach  for Mapping Radon
      Potentials in Florida. Salt  Lake City, UT: Rogers & Associates Engineering Corp.;
      report RAE-8945/3-1; 1990.

28.   Nagda, N.L., Koontz, M.D.,  Fortmann, R.C., Schoenborn, W.A., and Mehegan, L. L..
      Florida Statewide Radiation Study.   Germantown, MD: Geomet Technologies  Inc.
      report IE-1808; 1987.

29.   Roessler, C.E., University of Florida  Environmental Engineering Department, GRU
      data, private communication, 1992.

30.   Rogers, V.C. and Nielson,  K.K.  Multiphase Radon Generation and Transport in
      Porous Materials.  Health Physics 60:807-815; 1991.

31.   Rogers, V.C. and Nielson, K.K.  Benchmark and Application of the RAETRAD Model,
      proceedings of the 1990 International Symposium on Radon and Radon Reduction
      Technology. Atlanta, GA: February  1990.

32.    Rogers,  V.C., Nielson, K.K, and Merrell,  G.B.   Radon Generation,  Adsorption,
       Absorption, and Transport in Porous Media. Washington D.C.: U.S. Department of
       Energy report DOE IERI60664-1; 1989.

 33.    Rogers, V.C. and Nielson, K.K. Correlations for Predicting Air Permeabilities and
       222Rn Diffusion Coefficients of Soils.  Health Physics 61:225-230; 1991.

-------
34.    Nielson, K.K., Rogers, V.C., and Gee, G.W. Diffusion of Radon Through Soils: A Pore
      Distribution Model. Soil Science Society of America Journal 48:482-487; 1984.

35.    Nielson KK., Rogers, V.C., Brown, R.B., Harris, W.G., and Otton, J.K., Prototype
      Mapping of Radon Potentials in Florida, Rockville, MD: The 1991 Annual AARST
      National Fall Conference: Conference Preprints, Vol. 1, 145-162, 1991.

36.    Nielson, K.K. and Rogers, V.C., Development of a Prototype Map of the Soil Radon
      Potentials  in Alachua County  Florida.   Salt Lake City:  Rogers  & Associates
      Engineering Corp.  report RAE-9127/3-1, October 1991.

37.    SCS, State Soil Geographic Data Base (Statsgo) Data Users Guide, Lincoln, Nebraska:
      National  Soil Survey  Center,  Soil  Conservation  Service, U.S. Department  of
      Agriculture, draft report, 84pp, 1991.

38.    Nazaroff, W.W., Doyle, S.M., Nero, A.V., and Sextro, R.G. Radon Entry via Potable
      Water, pp. 131-157. In: Radon and Its Decay Products in Indoor Air, W. W. Nazaroff
      and A. V. Nero, eds., New York: Wiley & Sons; 1988.

39.    Acres. Measurement of Crack and Opening Contribution to Radon Entry (Feasibility
      Study).   Vol. Ill of Radon Entry  Through Cracks  in  Slabs-on-Grade,  Acres
      International Corp., report P09314; 1990.

40.    Nazaroff, W.W., Lewis, S.R., Doyle, S.M., Moed, B.A., and Nero, A.V. Experiments on
      Pollutant Transport from Soil into Residential Basements by Pressure-Driven Airflow.
      Environmental Science and Technology 21:459-466; 1987.

41.    Nielson, KK. and Rogers, V.C. Radon Entry Into Dwellings Through Concrete Floors.
       in Proceedings: The  1991 International Symposium on Radon and Radon Reduction
      Technology.  Philadelphia, PA: paper V-3; EPA-600/9-91-37c, April 1991.

42.    Thomas, B.P., Cummings, E, and Wittstruck, W.H. Soil Survey of Alachua County,
      Florida.  Gainesville, FL: U.S. Department of Agriculture, Soil Conservation Service;
      1985.

43.    Yeh, G.T., FEMWATER: A Finite Element Model of Water Flow through Saturated-
      Unsaturated Porous Media - First Revision, Oak Ridge, Tennessee: U.S. Department
      of Energy report ORNL-5567IR1, 1987.

44.    Sullivan, T.M.,  Kempf,  C.R.,  Suen,  C.J.,  and  Mughabghab, S.M., Low-Level
      Radioactive Waste Source Term Model Development and Testing, Washington DC:
      U.S. Nuclear Regulatory Commission report NUREG/CR-5204, 1988.

45.   EG&G Geometries, Aerial Gamma Ray and Magnetic Survey Gainesville and Daytona
      Beach Quadrangles Florida.  Final  Report.  Grand Junction, Colorado:   U.S.
      Department  of Energy report GJBX-101, 1981.

-------
46.    Otton, J.K., Asher-Bolinder, S., Reimer, G.M., and Henry, M., Prototype Mapping of
      Soil Radon Potential in Alachua County Florida: Geologic and Radiometric Support
      Investigations, Denver, Colorado: Draft  USGS report to  Florida Department of
      Community Affairs, December 1991.

47.    EPA. Monitoring for Radon-222 Emissions, Method 115, Appendix B of National
      Emission Standards for Hazardous Air Pollutants; Radionuclides, 40 CFR Part 61,
      Federal Register 54:51654-51715,1989.

48.    Rogers,  V.C., Nielson,  K.K.,  Sandquist,  G.M., and Rich,  D.C., Radon Flux
      Measurement and Computational Methodologies, Albuquerque, NM: U.S. Department
      of Energy report UMTRA-DOE/AL-2700-201, 1984.

-------
                                                                                VIIIP-3
   CORRELATION OF INDOOR RADON SCREENI.NG_lffiASUREMENTS WITH SURFICIAL GEOLOGY
                     USING GEOGRAPHIC INFORMATION SYSTEM?


      by: Charles Schwenker,  Jia-Yeong Ku*,
          Charles Layman   and Charles Kunz

          N.Y.S. Department of Health
          Wadsworth Center for Laboratories and Research
          Laboratory of Inorganic and Nuclear Chemistry
          Empire State Plaza
          Albany, N.Y. 12201-0509

         *N.Y.S. Department of Health
          Bureau of Environmental Radiation Protection
          2 University Place
          Albany, New York
         ft
          Present Address
          National Aeronautics and Space Administration
          Marshall Space Flight Center
          Huntsville, Al 35812
                                   ABSTRACT


     The State of New York has a growing database of over 50,000 homes for
which radon levels have been measured.  These data include information on home
construction, type of heating, and location by address.  We have looked at
correlations with many home variables but have not been able to do a thorough
analysis of radon-level relations to surficial geology, which requires
accurate spatial relation of home location with surficial geology maps.  We
are doing this by using a Geographic Information System mapping program (CIS),
to which we will add boundary maps of surficial geology, derived by digitizing
existing maps.  The indoor radon measurement data for Albany County have been
linked to latitude and longitude coordinates using the U.S. Bureau of the
Census Topologically Interpreted Geographic Encoding and Referencing (TIGER)
database.  The surficial geology boundary map has also been prepared for
Albany County and relationships of indoor radon to surficial geology are
discussed.   Comparisons are made for mapping indoor data using county-wide
averages, zip code averages and individual measurement mapping.

-------
                                INTRODUCTION
      The State of New York has a growing database of over 50,000 homes for
which radon levels have been measured.  These data include information on home
construction, type of heating, and location by address.  Mapping the average
or geometric mean indoor-radon concentrations for the counties or by zip code
is useful, providing information on the regional distribution of indoor-radon
concentration.  More information is obtained when each measurement is
specifically located and mapped.

      Computer-based Geographic Information Systems (CIS) are being developed
and utilized to map various spatially located information such as roads,
buildings, and geologic features.  Identifying the latitude and longitude
coordinates for the individual radon measurements would make it possible to
map the indoor-radon data utilizing a Geographic Information System.

      The geologic features that affect indoor-radon concentrations often vary
over smaller spatial dimensions than county, zip code, township, or other
politically defined areas.  Therefore it is sometimes difficult to associate
indoor-radon concentrations with environmental factors such as surficial
geology.  When the individual radon measurements are located by latitude and
longitude, the distribution of indoor concentrations can be correlated with
areas of any size or shape.  Surficial geology maps have been drawn for New
York State at a scale of 1:250,000.  These maps are being digitized into a GIS
so that the surficial geology can be readily correlated with spatially located
radon data.  This will help facilitate the identification of surficial geology
associated with above-average indoor radon.  The most direct method for
determining the potential for above-average indoor radon is to measure indoor
concentrations in a number of homes in the area in question.  Finding a
correlation of high radon levels with a county or zip code is often not of any
value as a predictor of what to look for in another county or zip code.
However, if correlations can be developed with surficial geology for certain
soil types, this could serve as an indicator of a high probability that homes
in similar unmeasured areas will also have above-average radon.  There are
other situations in which spatially located radon measurements can be useful.
For example, when the individual data points within a county or zip code area
are mapped, it may be possible to identify high and low risk areas within the
county or zip code area.  Counties with average or below average
concentrations of indoor radon may in fact contain areas with above-average
indoor radon.  The radon potential of a specific location, such as a school or
building site, can be estimated by averaging the indoor measurements within a
specified radius of the location and by examining the radon potential for the
geology at that location.  When the individual radon measurements are mapped,
the spatial distribution of the data will identify those areas where most of
the measurements have been made so that proper interpretation and weighting
can be applied to the area-wide averages.  Clearly, the utility of indoor-
radon data is considerably increased when they are specifically located by a
coordinate system.

-------
                       LOCATING RADON MEASUREMENT DATA


      The key element in this project is to find an efficient means of
assigning coordinates to large numbers of indoor radon measurements.  When
dealing with less than a few hundred radon measurements, it may be possible to
use street maps to locate each measurement.  For a large database such as the
more than 50,000 homes measured by the New York State Department of Health, it
is not feasible.  The Albany County database selected for this study consisted
of 1080 homes.   This database included 150 homes which were P.O. box or rural
delivery, leaving 930 homes with radon measurements and street addresses.  At
this point it is not possible to match a P.O. box or rural delivery address
with latitude and longitude coordinates using the TIGER database.  This
hinders our ability to match coordinates with addresses in rural areas.  Each
radon measurement is associated with a name, address, home construction, and
home operation information.  Radon readings at different levels of the house
are available for many homes.  Only the readings in the basement are used in
this study.

      We are investigating two approaches to link latitude and longitude
coordinates with the addresses for the indoor radon measurement data.  One
approach is to use the US Bureau of the Census Topologically Integrated
Geographic Encoding and Referencing (TIGER) database.  The second is to use
the New York State Department of Equalization and Assessment (NYS DEA) tax
assessment database.  The tax assessment database contains detailed
information about each household, but not every locality's tax assessment
database has latitude and longitude information in it.  The geographical
location of the radon measurement is assigned by matching the common link
between the radon data and the TIGER and NYS DEA data, which is the name and
address field.

      We were able to locate spatial coordinates for 73% of the 930 radon
measurements in Albany County that had street addresses.  The radon
measurements that we were unable to locate include homes in newly developed
areas.  Other reasons for not matching measurements with coordinates are being
investigated.
                    INDOOR RADON DISTRIBUTION BY COUNTY
      The average basement screening indoor-radon concentrations for each
county in New York State are shown on figure 1.  Various regional differences
are evident  and have been discussed elsewhere(l).  The counties with the
highest average indoor radon are located in the southwestern part of the
state,  generally in an arc south of the Finger Lakes.  Gravelly glacial out-
wash deposits forming highly permeable soils in the valleys of this region is
the principal factor resulting in above-average indoor radon (2).  The
counties with the lowest average indoor-radon concentrations are located in
the Adirondack Mountains and on Long Island.  The soil in both these areas has
below-average radium and is of moderate permeability.  Regional differences
for indoor-radon concentrations can usually be understood from the general

-------
characteristics of the regional geology.  However, mapping and interpretation
at this scale requires sufficient field data at specific locations to
determine the radon potential of the surficial geology in the area.
Additional information is obtained by looking more closely at the distribution
of the indoor radon measurement data.

                               ZIP CODE MAPPING
      The average basement screening concentration in Albany County is 3.8
pCi/L which is less than the statewide average of 5.5 pCi/L.  The geometric
mean for Albany County is 1.7 pCi/L and the statewide geometric mean is 2.5
pCi/L.  Average basement screening data distributed by zip code are shown for
Albany County in figure 2.  A total of 836 screening measurements were used to
determine the zip code averages.  Less than 10 measurements were made in 8 of
the 25 zip code areas; however, in the more populous areas sufficient
measurements were made to obtain more reliable averages.  Geometric means were
only calculated for zip codes with 10 or more measurements.  Most of the
measurements were made in and around the city of Albany in the northeastern
part of the county, with few measurements in the rural areas.

      The surficial geology for Albany county is shown in figure 3.  The map
for Albany County has been digitized into a CIS from surficial geology maps
prepared by the NYS Geological Survey.  The NYS Geological Survey is in the
process of digitizing the surficial geology maps for the entire state.
Although the surficial geology in Albany county is varied and complex,
particularly in the area around the city of Albany, it is possible to obtain
some correlation between zip codes, average indoor concentrations, and the
surficial geology.  In the northeastern region of the county, there are fairly
large areas of sand dunes and lacustrine (lake) sand   (figure 3). Several soil
measurements relating to indoor radon potential have been made in this
area(3).  The sandy soil in the dunes has below-average Ra-226 (0.5pCi/g),
below-average soil gas Rn-222 (250pCi/L at a depth of 1.2m) and moderate
permeability resulting in below-average potential for indoor radon.  The
average concentration of Ra-226 in soils in the U.S. is 1.0pCi/g(4) and
average soil-gas Rn-222 at a depth of 1.2 m in New York State is about
700pCi/L (3).  The several zip code areas located in the region of sandy soils
have geometric-mean indoor-radon concentrations of less than 1.0 pCi/L which
is below the County mean of 1.7 pCi/L and well below the statewide mean of 2.5
pCi/L.  The zip code areas in the sandy region are quite populous and the
number of measurements in these zip codes is sufficient for reasonably
reliable average values.

      The geometric mean indoor radon concentrations for several of the zip
code areas are above the county mean.  Two of these areas will be examined in
more detail.

-------
                        INDIVIDUAL MEASUREMENT MAPPING


      There were 1080 screening measurements  for Albany County at the time the
paper was being prepared.  From this database of 1080 screening measurements,
680 homes were matched with  latitude and  longitude coordinates and are plotted
in figure 4.  The size of  the dot representing each measurement is
proportional to the indoor radon concentration.  The measurement density is
greatest in and around the city of Albany.  Few screening measurements were
made in the rural areas of the county;  in addition few coordinate matches were
made in the rural zip codes, resulting  in few or no data located by
coordinates for the rural  areas. Some regions with above-average levels can be
seen from the distribution of measurements.

      Zip code 12211 and the surrounding  area are shown in figure 5.  The
measurement data are superimposed on the  surficial geology.  It can be seen
that most of the higher levels are located  in or near the kame area while the
measurements in the dunes  are below average.  It should be noted that the
boundaries for the different surficial  geologies are not as sharp as pictured.
The New York State Geological Survey Surficial Geology maps were drawn to
conform to USGS maps, in which some terrain features are not accurately
located.  This means that  surficial geology areas are not always plotted to
their exact locations.  Plotting and transcribing errors are magnified when an
area is expanded for closer examination.  In  addition, there is normally a
gradation from one surficial geology to another while the maps show sharp
boundaries.  These factors can combine  to give errors of up to 500 meters in
actual surficial geology boundary locations.  The soils in the kame area were
measured for Ra-226 concentration, soil gas Rn-222 concentration and
permeability for gas flow at the location indicated by the circle of figure 5.
The soil Ra-226 concentration was measured at 0.7 pCi/g, which is below the
U.S.   average of 1.0 pCi/g.  The soil-gas Rn-222 concentration at a depth of
1.2m was 600 pCi/L which is about average for soils in New York State; the
permeability for gas flow was measured  at 1.7xlO-7cm"2 which  is moderate.
These soil characteristics suggest average indoor radon.  However, in the U.S.
Geological Survey topographical maps it was noted that a gravel pit is located
in this kame deposit area  (figure 5), indicating that the soils in portions of
this  area can be gravelly and with higher permeabilities than observed at the
measurement location.  More soil measurements are required to resolve this
question but it appears that the above-average indoor radon in this kame area
results from gravelly soils and fairly high permeabilities for gas flow.

      The zip code area with the highest average indoor radon in Albany County
is zip code 12186 (figure 2).  The northeastern part of this zip code area is
shown in figure 6 with the individual home measurements superimposed on the
surficial geology.   Before the indoor radon screening program in New York
State was initiated, a home with above-average indoor radon was identified in
this  area and a special study was conducted,  which included measurements in
about 80 homes.   Subsequently,  when the screening program became available,  a
number of residents in this area requested measurements, resulting in a
disproportionate number of measurements from  an area identified as an above-

-------
average area.  Both the screening and special study measurements are shown in
figure 6.  A number of soil measurements have been made in this area of above-
average indoor radon.  The soil Ra-226 averaged l.OpCi/g, the soil-gas Rn-222
675 pCi/L and the permeability for gas flow averaged 7xlO-6cm"2(2).  The soils
in this rather small area of 1 to 2 square miles are very gravelly and are
highly permeable.  The gravels are over 10m deep for most of the area.
Although the soils have average levels of Ra-226 and soil-gas Rn-222,  the high
permeability of the deep gravel deposits is the soil characteristic resulting
in above-average indoor radon.  The measurements shown in figure 6 indicate
that, although the indoor levels are considerably above average in the area
bounded by the two railroad tracks, the levels are higher in the region closer
to the point where the two tracks cross.

      There can be considerable variation of the characteristics relating to
indoor-radon potential such as soil Ra-226, soil-gas Rn-222, and permeability
within a specific surficial unit.  Variability is greater for some units such
as lacustrine delta, which is characterized by coarse to fine gravel and sand
and kame deposits, which includes kames, eskers, kame terraces, and kame
deltas, with coarse to fine gravel and sand.  Within classifications such as
lacustrine delta and kame deposits the permeability for gas flow can be highly
variable depending on the proportion of coarse and fine particles.  In
surficial units such as dunes, lacustrine sand, and lacustrine silt and clay
the permeability is generally moderate to low and average to below-average
indoor radon is expected.

      All of the measurements in Albany County falling within the boundaries
of a particular surficial soil unit such as dunes were combined to determine
the geometric mean indoor radon for that particular surficial geology (table
1).  There are sufficient data for six surficial units to calculate geometric
mean screening levels.  The geometric mean screening levels for New York State
and Albany County are 2.5 and 1.7pCi/L respectively. The geometric mean levels
for the dunes, lacustrine sand and lacustrine silt and clay range between 0.9
and 1.1 pCi/L, which is considerably below the mean for both State and County.
The geometric means for the lacustrine delta and kame deposits are 3.6 and 3.2
pCi/L, which are above the State and County means.  The 17 measurements used
to calculate the mean for the lacustrine delta are from the screening
measurements and do not include the special study data.  However 13 of the 17
measurements were located in and around the area identified as an above-
average area.  Clearly, the data for the lacustrine delta are heavily weighted
toward a gravelly region and cannot be interpreted as an accurate average for
this surficial unit.

                                 CONCLUSIONS
      This pilot study was limited to Albany County as a demonstration
project.  As the process of locating measurement data with a coordinate system
is extended throughout the state, our understanding of the relationship
between indoor radon and surficial geology will be improved.  A CIS allows for
overlaying additional spatial information for correlation such as topography,

-------
NURE data and more detailed soil information from the U.S. Geological Survey
Soil Conservation Service maps.  Clearly, the value and utility of the indoor
radon data are increased when they are spatially located and analyzed with a
CIS.  In the future thought should be given to facilitate coordinate location
of measurement data as they are acquired.

      Finally care must be taken in presenting and using the data to maintain
measurement confidentiality.  The scale and street information should be such
that specific homes cannot be identified.

      The work described in this paper was not funded by the U.S.
Environmental Protection Agency and therefore the contents do not necessarily
reflect the views of the Agency and no official endorsement should be
inferred.

                                ACKNOWLEDGEMENT
      We would like to acknowledge discussions with Bob MacHaver and Andy
Silfer in 1989, who first suggested the use of the tax assessment database to
locate home measurement data.
                                  REFERENCES
1.   Layman, C., Kunz, C. and Keefe, L., Indoor Radon in New York State:
     distribution, sources and controls, New York State Department of Health,
     Technical Report. Nov 1990

2.   Kunz, C., Layman, C. and Parker, C., Gravelly Soils and Indoor Radon:
     Proceedings EPA 1988 Symposium on Radon and Radon Reduction Technology,
     Denver, CO. Oct 1988.

3.   Kunz, C., Influence of Surficial Soil and Bedrock of Indoor Radon in New
     York State Homes, New York State Energy Research and Development
     Authority Report 89-14. Oct 1989.

4.   Myrick, T.E., Bevin, B.A., and Haywood, F.F., Determination of
     Concentrations of Selected Radionuclides in Surface Soil in the U.S.,
     Health Physics, Vol.45, No.3,  pp631-642 (1983).

-------
             (Basement  Screenings  Only)
                      As of October 1. 1991
                      0.1 - 1.9
                      4.0 - 7.9
                           Illlllllll 2.0 -  3.9
                                 8.0 - 15.9
Figure 1.
           Headings are measured In pCl/1
County averages from about 50,000 basement screening
measurements in New York State.

-------
                     A bony  County
 Geometric  Mean  Indoor  Radon Concentration  by Zip  Code
Radon pCi/l
 0 to 0.49
                                                                         0.5 to 0.99
                                                                         1 to 1.49
                                                                         1.5 to 1.99
                                                                        2 to 2.49
                                                                        2.5 to 4
                                                                        Missing
                                                                          Miles

                                                                                    10
Figure 2.  zip Code averages for Albany County.  Averages for zip codes with less than 10 measurements are not included.

-------
                          Albany   County
                           Surficial  Geology
                                                                           Map Legend
                                                                           Bedrock
                                                                                     Dunes
                                                                                     Kame
                                                                                     Kame moraine
                                                                                     Lacustrine Delta
                                                                                     Lacustrine sand
                                                                                     Lacustrine silt and clay
                                                                                     Outwash sand and Gravel
                                                                                     Swamp deposits
                                                                                     Recent deposits
                                                                                      ill
                                                                                     Till-moraine
                                                                                        Miles
                                                                                          E
                                                                                          5
                                                                                           i (!
Figure 3.
Surficial geology for Albany County.  Adapted from surficial geology maps prepared by the ms Geological
Survey, State Education Department.

-------
         Albany  County  -  Zip  Codes
    Radon Measurements     Matched Data Points
Map Legend
                                                                   Zip Codes
                                                               •   Datapoints
                                                                  Radon pCi/l
                                                                    31.2
                                                                    62.5
                                                                    125.0
                                                                    Miles


Figure 4. Radon screening measurements for Albany County. The size of the dot is proportional to the indoor radon
       concentration.

-------
         Albany  County  -  Zip   12211
Surficial  Geology  and  Indoor  Radon  Concentrations
     Map Legend

   Bedrock

   Dunes

   Kam e

   Kome moraine

   Lacustrine Delta

few! Lacustrine sand

   Lacustrine silt and clay

   Outwash sand and Gravel

   swamp deposits

   Recent deposits

   Till

   Till moraine

   Gravel pits

   Radon in Soil test site

   Datopoints
                                                                         Radon pCi/l

                                                                             7.5


                                                                            15.0


                                                                            30.0
                                                                            Miles
                                                                         o
         0.5
Figure 5. Surficial geology and indoor radon screening measurements for zip code 12211 and surrounding area in
        Albany County.

-------
           Albany   County  -   Zip   12186
 Surficial  Geology  and  Indoor  Radon  Concentrations
     Map  Legend
   Bedrock
   Dunes
   Kame
   Kame moraine
   Lacustrine Delta
   Lacustrine sand
   Lacustrine silt and clay
   Outwash  sand and Gravel
   Swamp deposits
   Recent deposits
   Till
   Till-morame
— Highways
**-* Railroads
X  Schools
•  Datapoints
                                                                                Radon pCi/l
                                                                                  31.2
                                                                                  62.5
                                                                                  125.0

                                                                                   Mi 6S
                                                                                     0.5
Figure 6.  Surficial geology and indoor radon measurements for the northeastern part of zip code 12186 in Albany County,
         Indoor radon measurements  include both screening and special study measurements.

-------
               TABLE 1.  GEOMETRIC MEAN INDOOR RADON LEVELS FOR
                       SEVERAL DEPOSITS  IN ALBANY  COUNTY
    Surficial
     Geology
  No. of
Measurements
Geometric
  Mean
 (pCi/L)
  Min.
(pCi/L)
  Max
(pCi/ L)
Dunes                         213
Lacustrine Sand               258
Lacustrine Silt & Clay         53
Lacustrine Delta               17
Recent Deposits                36
Kame Deposits                  35

Surficial Deposit Explanation*
                   0.9
                   0.9
                   1.1
                   3.6
                   1.5
                   3.2
                 0.1
                 0.1
               30
               34
               10
               58
               16
               74
Dunes:  Fine to medium sands, well sorted, stratified, non-calcareous,
unconsolidated, generally wind-reworked lake sediments, permeable, well
drained, thickness variable (1-10 meters).

Lacustrine sand:  Sand deposits associated with large bodies of water,
generally a near-shore deposit or near a sand source, well sorted, stratified,
generally quartz sand, thickness variable (2-20 meters).

Lacustrine silt and clay:  Generally laminated silt and clay, deposited in
proglacial lakes, generally calcareous, potential land instability, thickness
variable (up to 100 meters).

Lacustrine delta:  Coarse to fine gravel and sand, stratified, generally well
sorted, deposited at a lake shoreline, thickness variable (3-15 meters).

Recent deposits:  Generally confined to floodplains within a valley, oxidized,
non-calcareous, fine sand to gravel, in larger valleys may be overlain by
silt, subject to frequent flooding, thickness 1-10 meters.

Kame deposits:  Includes kames, eskers, kame terraces, kame deltas, coarse to
fine gravel and/or sand, deposition adjacent to ice, lateral variability in
sorting, coarseness and thickness, locally firmly cemented with calcareous
cement, thickness variable (10-30 meters).
* Taken from surficial geologic maps of New York, University of the State of
New York, State Education Department.

-------
                                                                                   VIIIP-4
                   Analysis of Indoor Radon in New Mexico Using
                       Geographic Information Systems (GIS)
                                 Richard A. Dulaney
                     Lockheed Engineering and Sciences Company
                            Environmental Programs  Office
                             1050 E. Flamingo Road #120
                               Las Vegas, NV  89119
                                   (702)798-3158
                                   FTS: 545-3158
                                    ABSTRACT

Geographic Information Systems (GIS) are a powerful computer based information management
tool for spatial,  or geographic, data.  This technology,  which blends  traditional  computer
cartography with state-of-the-art database management, is being employed to analyze indoor
radon in New Mexico.  The project, being undertaken in  support of EPA Region 6, looks at
both state-wide and local patterns of indoor radon.  The geographic distribution and spatial
relationships of over  1000 indoor radon samples, taken from the EPA/State random survey
conducted in 1989, are examined.  The correlation of these radon samples with geology, soils,
aerial radiometric and population data is explored using GIS spatial overlay techniques.

-------
                                                                                          VIIIP-5
     A RADON "PIPE" (?) IN THE BREVARD FAULT ZONE NEAR ATLANTA. GEORGIA

                          L.T. Gregg, P.G., and John Costello, P.G.
                               Atlanta Testing & Engineering
                                11420 Johns Creek Parkway
                                  Duluth, Georgia  30136
                                       ABSTRACT

       The Brevard fault zone (BFZ) extends from the Smith River allochthon in southern Virginia
southwestward to the Fall Line unconformity in central Alabama. The BFZ contains a sequence of
retrograded (low-grade) metamorphic rocks that has been affected by major thrust and lateral-slip
faults that moved under both ductile and  brittle conditions  through the Late  Paleozoic Era.
Compression-related movement is believed to have ceased following the Alleghanian orogeny (ca. 300
ma).  The adjacent Palmetto Granite (ca. 325 ma) in Southwest Atlanta is sheared  by Brevard
faulting, and the BFZ is cross-cut by Mesozoic extensional fractures and dikes constraining the latest
movement. Bedrock lithologies include mylonite, phyllonite, and cataclasite. Regional strike of the
BFZ in the Atlanta area is N45°E and regional dip of foliation is, in general, 20 to 60 degrees SE.
The dip angle on individual fault planes is generally about 15 to 20 degrees. BFZ rocks have mostly
weathered by chemical processes, to a locally thick saprolite residuum, typically up to 50 to 75 feet
or more.

       In March, 1990, a soil-gas radon survey was conducted at a new building site located in the
BFZ southwest of Atlanta.  Radon concentrations up to 13,050 picocuries per liter (pCi/1) were
measured at original grade in near-surface soil. After site development, which involved undercutting
the original grade up  to 15 feet,  another soil-gas survey  in October,  1990  showed  radon
concentrations up to 1,130 pCi/1.  By separately contouring the measured concentrations from both
surveys and overlaying the resulting maps, a preliminary geologic model was developed.  The data
suggest a very localized upward-spreading "pipe" through the  soil/saprolite residuum.   Geologic
controls on this "pipe" probably include northeast-trending foliation and/or (most probably)  brittle
faults and joints which may be cross-cut by northwest-trending Mesozoic extensional fractures. The
intersection of these permeable zones may cause or control the  "pipe".

-------
                                    INTRODUCTION

       During the period March-October, 1990, a rather curious near-surface geological feature was
mapped in subsurface soils of the Brevard fault zone (BFZ) at a new building site some 13 miles
southwest of Atlanta, Georgia.   We have  named  this  feature a radon "pipe" because of certain
morphological (not diagenetic) similarities to kimberlite pipes in southern Africa and elsewhere. The
"pipe" was delineated by two soil-gas radon surveys. From what is known about the geology of the
BFZ in this area, we have hypothesized a set of geologic controls on occurrence and orientation of
the "pipe".

                      GEOLOGY OF THE BREVARD FAULT ZONE

APPALACHIAN REGION

       The BFZ is a conspicuous NE-SW-trending linear feature on satellite images and small-scale
topographic  and  geologic  maps of the southern  Appalachian  crystalline  core.   At least  31
interpretations of Brevard zone geology (1) have been advanced since the pioneer mapping of Keith
(2) identified the Brevard  Schist, named after exposures of dark schistose rock near Brevard, NC.
Since that time, most geologists have  interpreted various types of faulting to explain  BFZ rock
textures (e.g., mylonite, phyllonite, etc.), discontinuity and repetition of mappable internal rock units,
and the overall linearity of the belt.  Conventional Brevard zone characteristics are summarized by
Hatcher (3):

              The zone is a linear fault zone throughout most of its extent, but it appears
              to splay toward its ends.  Some geologists (4; 5) have questioned the existence
              of faulting along segments of the zone.

              Foliation dips moderately over much of its length.

              The zone is stratigraphically controlled at least along part of its extent.

              Local slices of comparatively unmetamorphosed platform-type carbonate rocks
              occur within the zone.

              The zone occupies several rock units along strike,  but the same lithologies
               (metagraywacke, schist, amphibolite, aluminous schist, and quartzite) that
               comprise several major, locally named lithostratigraphic sequences  (Sandy
               Springs Group,  Tallulah  Falls  Formation,  Ashe  and   Alligator Back
               Formations, and  Lynchburg Formation) lie northwest of the zone, except in
               the Grandfather  Mountain, NC area.

               The BFZ has been affected by at least one early ductile deformation and at
               least one later brittle deformation.  Most likely, more than one of each type
               of event has occurred.

        Bobyarchick and others (1) summarize BFZ movement history throughout the Late Paleozoic
 Alleghanian orogeny as beginning with a ductile, dextral (right lateral) shear zone. Structures related

-------
 to this shearing crosscut Early to Middle Paleozoic fold-generated fabrics.  Later Alleghanian thrust
 faulting, accompanied by retrograde metamorphism, reactivated the deep part of the older ductile
 shear  zone to emplace the Blue  Ridge-Inner Piedmont  crystalline thrust  sheet.  The latest
 Alleghanian continent-continent collision reactivated the Blue Ridge-Inner Piedmont thrust and
 generated a high-angle, brittle thrust splay off of the Blue Ridge-Inner Piedmont thrust fault that
 extended upward through the older ductile fabrics generated during dextral shearing.  This last thrust
 event emplaced the platform carbonate slices and brecciated older mylonitic rocks.

 PIEDMONT PROVINCE OF GEORGIA

       The BFZ in northern Georgia (including  the Atlanta area) separates the eastern Blue
 Ridge/northern Piedmont  Geologic Province from the Inner Piedmont Geologic Province.  In the
 Atlanta region, the BFZ is bounded to the north by the Sandy Springs Group (6) and to the south
 by the Atlanta Group  (7).  Throughout most of the Atlanta region, the Chattahoochee River is
 entrenched into the BFZ and adjacent strike ridges in the Sandy Springs Group (8).

 METROPOLITAN ATLANTA AREA

       Typical BFZ rocks in the Atlanta region include protomylonite, mylonite, blastomylonite,
 button schist, and phyllonite (generally in order of decreasing grain size), produced by compressive
 deformation of various metasedimentary and igneous parent rocks under ductile conditions.  Older
 mylonitic textures are overprinted by later brittle fracturing and brecciation, enhancing rock porosity
 and permeability.  Local evidence of normal faulting (most likely also brittle) in the BFZ is reported
 by McConnell and Abrams (9).

 WEATHERING OF IGNEOUS/METAMORPHIC ROCKS

       Brevard fault zone rocks throughout the Atlanta region are more or less chemically weathered
 in natural surficial exposures. The weathered mantle above bedrock typically exhibits a compositional
 and textural spectrum ranging from shallow, bioturbated residual soils, through an intermediate layer
 of saprolite (completely weathered bedrock that  retains original rock textures and structures), to a
 deep, partially weathered zone of interlayered saprolite, partially weathered rock, and bedrock. The
 thickness of this weathered  mantle is dependent on bedrock mineralogy, concentration  of joints
 and/or other open, water-bearing structures, and  local erosion rates.

            PREVIOUS RADON STUDIES IN APPALACHIAN  SHEAR  ZONES

       A recent and highly valuable compendium of radon studies  in Appalachian  shear zones is
 presented in Section 2 (pp. 39-64) of (10).  In particular, Gundersen reports  on measured soil gas
 radon concentrations in shear zones at or near Boyertown, PA, Brookneal, VA,  Glen Gardner, NJ,
 and Montgomery County,  MD, and Agard  and  Gundersen discuss detailed soil geology and
geochemistry in the  Boyertown-Easton, PA area.  In these studies, Gundersen and her co-workers
showed very good correlation between mapped shear zones and high measured values (> 2,000
picocuries per liter, pCi/1) of radon in soil gas. The geology of these shear zones  consists basically
of mylonites emplaced  or  occurring  in hornblende and  biotite gneisses,  granodiorite, hornblende
granite, amphibolite, and various schists, gneisses, mafic rocks and ultramafic rocks. In  these mylonite

-------
zones, Gundersen and her co-workers measured radon in soil gas up to 4,000 pCi/1, although the
average values were somewhat lower.

       In addition,  Gundersen and Gregg collaborated in 1989-90 (unpublished) on  a  series  of
informal, small-scale measurements of radon in soil-gas in the BFZ on the west side of Atlanta, GA,
along 1-285 near its  crossing of the Chattahoochee River.  This limited work, in which  Gundersen
took the  lead and Gregg and others served primarily as observers and consultants, showed isolated
readings  (1989) as high as 35-40,000 pCi/1; these were not replicated in the 1990 revisitation to the
site, but readings as high as 10,000 pCi/1 (Gundersen) and 8,500 pCi/1 (Gregg) were recorded by two
side-by-side teams with separate instrumentation.  These limited data are nowhere near diagnostic,
but we feel they indicate the potential for high radon concentrations in soil (saprolite) overlying BFZ
rocks.

                               SOIL-GAS RADON SURVEY

SITE LOCATION, CHARACTERISTICS, AND DEVELOPMENT

       The  site is located  about 13 miles southwest  of downtown Atlanta, GA.  Original site
topography,  as taken from visual observations and the  U.S.G.S. 7.5-minute topographic quadrangle
map, showed site drainage to be moderately sloping to the southwest. At the time of the first survey
(March, 1990), the site was ungraded and generally devoid of surface vegetation. Surface soils were
observed to  be primarily clayey sandy silts and silty sands derived from phyllonite.  As  part of the
geotechnical investigation, 5 borings were drilled on original topography by hollow-stem auger  to
depths up to 15 feet.  Subsurface soils (saprolite) encountered were primarily micaceous silty clays.

       Site development called for excavation to final grade up to  15 feet, to provide a slab-on-grade
building  footprint of approximately 160,000 square feet.  The planned  structure was  a one-story
building with no basement or below-grade structure. Grading was completed in August, 1990.

INSTRUMENTATION AND  SURVEY LAYOUTS

       Because the BFZ is known to locally have high concentrations of both Radon-222 (Rn-222)
and Radon-220 (Rn-220) in near-surface soil, the client requested a soil-gas radon survey. The initial
survey was  conducted in  March,  1990,  utilizing an EDA RD-200  Radon  Detector (alpha
scintillometer) and associated  peripheral equipment.  Soil-gas was sampled  and  a-counted at  14
stations,  at a depth of 14 to 18 inches below  ground surface. Sample station locations  were taken
from architect's drawings of building column locations,  and were established by pace and compass as
well as by visual observations of topography.

       Because of the high concentrations of both radon  isotopes (discussed below) that were
measured in the initial survey, a second survey was recommended.  The client agreed, and the second
survey was conducted in October, 1990 at final grade. This soil gas sampling was done using the same
procedures and protocols as the first survey. The locations were selected to bracket the two locations
(i.e., sampling stations) which reported the highest radon concentrations  in the initial survey.

SURVEY RESULTS

-------
       The results of the first survey are shown in Table 1 and Figure 1, while the results of the
second survey are shown in Table 1 and Figure 2. Note the change in contour interval from Figure
1 to Figure 2. The highest Rn-222 value measured in the first survey was 13,050 pCi/1 (Station 12),
while the highest value measured in the second survey (Station 15,  a short distance northwest  of
Station 12), was 1,130 pCi/1. The reduction in highest measured value, over 15±  feet of vertical cut,
was about 11.5:1. The Rn-222 isorad contours in Figures 1 and 2 have been drawn on the assumption
of linear gradients between sampling stations, which may not be correct.

       The instrumentation used in these surveys is designed to primarily measure concentrations of
Rn-222, but it also measures Rn-220 concentrations. A linear regression analysis of Rn-222 (x) vs.
Rn-220 (y)  values showed the following correlation coefficients (r/r2):

              1st Survey:  r =      0.998;  r2  =   0.9%
              2nd Survey:  r =      0.999;  r2  =   0.998

       The conclusion from  the measured data and the r/r2 values is that Rn-222  and Rn-220
concentrations are highly correlated.  From that conclusion, we draw a further conclusion, viz. that
both uranium and thorium are present in anomalous amounts in the BFZ bedrock/saprolite/soil at
this site.  This latter conclusion is generally borne out by the results of a reconnaissance study (11)
of the Tyrone Granite (Phase) some 14 miles south of the site, where anomalous values of both
uranium  and  thorium were reported using gamma scintillometry. The exact geologic relationship
between  the Tyrone Granite (Phase) and the BFZ is  murky at best. Current thinking (7) is that the
Tyrone Phase is related to the Ben Hill/Palmetto Granites, which are locally exposed in outcrop and
are hypothesized to be cupolas of a deep batholith emplaced up to 25 ma(?) before  the last
compressional movement along the BFZ (ca. 300 ma).  The Ben Hill/Palmetto  granites have been
mapped over  a large areal extent on the south side of the BFZ.  In any event, we believe there is a
relationship, whether syngenetic or epigenetic, between the anomalous radionuclide concentrations
in BFZ rock/saprolite/soil and those to the south in Tyrone rock/saprolite/soil.

                      INTERPRETATION AND GEOLOGIC MODEL

       As is well-known to practitioners, exact geologic interpretation of soil gas radon data is often
problematic at best and uncertain at worst (10; 12).  The  geologic,  geochemical, and geophysical
controls on radon movement through saprolite and soil are only now becoming somewhat understood,
or at least enumerated (13; 14).

       With this apologia, we offer the following model for the site geology and radon concentrations
measured at this site:

       •      Dominant lithology is  low-grade  (retrograde metamorphosed to  greenschist
              facies) phyllonite and minor mylonitized granitic rocks.

       •      Strike of mylonitic foliation is approximately N45°E and dip is 20 -  60°SE.

       •      Cross-cutting Mesozoic  (Triassic?) extensional joints and fractures  strike
              generally NW; dip is generally steeply inclined to vertical.

-------
TABLE 1. MEASUREMENT OF RADON IN SOIL-GAS;
BFZ SITE, SOUTHWEST ATLANTA, GEORGIA
Sampling
Station
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
March, 1990 Survey
Rn-222.
pCi/l
120
BDL
500
10
BDL
BDL
BDL
180
1,020
BDL
9,030
13,050
BDL
40
—
—
—
—
...
Rn-220.
pCi/l
—
BDL
130
10
BDL
BDL
20
80
160
30
6,330
8,660
BDL
BDL
—
—
—
—
—
October, 1990 Survey
Rn-222.
pCi/l
—
—
—
—
—
„.
—
—
—
...
—
—
—
—
1,130
1000
BDL
40
60
Rn-220,
pCi/l
—
—
—
—
—
—
—
—
—
—
—
—
—
—
200
260
BDL
BDL
BDL
Notes:
1. pCi/l = picocurieslliter
2. BDL = Below Detection Limit
3. Measured values rounded to nearest 10 pCi/l

-------
         LU
         LT\
         -3-
                   BUILDING FOOTPRINT
                  CONTOUR  INTERVAL =  2000  pCi/1

                  7•  SOIL GAS SAMPLING STATION

                 —I t CROSS SECTION
                   A'         105          210 FT.
FIGURE 1.   RADON-222 ISORADS IN SOIL GAS AT ORIGINAL SURFACE

-------
       LT\
       -3-
              19
               10.
                                     6.
                    BUILDING  FOOTPRINT
               CONTOUR INTERVAL =200  pCi/1

              ?•  SOIL GAS SAMPLING STATION
                  CROSS SECTION
          A     A1
              9
105
210 FT.
FIGURE 2.  RADON-222 ISORADS IN SOIL GAS AT FINAL GRADE

-------
              Strike  of the short axis of  the Rn-222 anomaly shown in Figure 1  is
              approximately E - W; strike of the long axis is approximately N39°E. The two
              axes are obviously not orthogonal. There is also a discernible tertiary (?) axis
              striking  approximately  N45°W, somewhat elongated  to  the NW and
              foreshortened (?) to the SE.

       .       Strike  of the long  axis of the Rn-222  anomaly shown  in Figure 2  is
              approximately N8°E. The short axis strike appears to be approximately the
              same as in Figure 1, viz. E - W.

       •       Figure 3 presents a N - S cross-section through the anomalies contoured  in
              Figures 1 and 2.  We interpret Figure 3  to represent an upward-spreading
              "pipe" that widens  as the original surface  is approached.  Geologic controls,
              we believe, are northeast-trending mylonitic foliation coupled with similarly
              oriented brittle faults  and  joints that have been cross-cut by northwest-
              trending Mesozoic extensional fractures and joints.  These structural features
              extend from bedrock upward into the saprolite.  The intersection of these two
              "permeable" zones (foliation plus structural features) may cause or control the
              "Pipe-"

              Bedrock depth is uncertain, due to the lack of bedrock outcrops near the site
              and the limitations of the geotechnical engineering survey that was conducted;
              however, we suspect that bedrock (geologically, unweathered rock) may occur
              at about 20 to 30 feet  below final grade, which was the grade at which the
              second radon survey was conducted.

                                     CONCLUSIONS

       While we have advanced a structurally-oriented model (relict foliation, fractures,  and joints
in saprolite) for the occurrence and overall lateral geometry of this radon "pipe", we do not have a
plausible hypothesis  for its vertical (apparent upward spreading) geometry.   To develop  such a
hypothesis would have required an extremely detailed study of soil and saprolite physical and chemical
properties, as well as measurement of radon concentrations in soil gas at several different  elevations
(i.e., between original grade  and final  grade) and on a closely-spaced lateral grid pattern.  This was
not possible within the constraints of the client-funded work reported in this paper.

       A detailed micro-model  of radon movement in Piedmont soil/saprolite  would require
consideration and measurement of a large  number of parameters.  As is  well known, once radon is
"liberated" (i.e., emanates) from its parent radium source,  it will migrate by diffusion and convection
to zones of lower concentration  (diffusion)  and pressure (convection), e.g., vertically toward  the
surface. As  Gregg and Coker (14) have pointed out, radon concentration and migration in saprolite
are influenced by the lithology of the parent rock, the amount and degree of jointing and fracturing
and interconnection, the degree of water saturation, permeability  and porosity, thickness, zonation
(whether the saprolite is structured or massive), and the distribution and extent of nanopores (pores
less  than one micron in width).   In surface  and near-surface soil, the principal  influences  are
thickness, zonation (A, B, and C  zones), moisture content (8 to 15 percent has been suggested as
optimum for radon emanation), permeability and porosity of the  soil, and finally the temperature

-------
       r~
       A
15±FT
                              NE
                 	   ISORAD CONTOUR (pCi/1)
                          0       55 FEET
          VERTICAL/HORIZONTAL EXAGGERATION » 3-7
A'
                                                          GRADE FOR
                                                          FIGURE 1.
                                                          GRADE FOR
                                                          FIGURE 2.
             FIGURE 3-  CROSS SECTION OF RADON "PIPE"

-------
 gradient from the surface, which determines the water vapor pressure of the soil. Finally, there are
 meteorologic and topographic effects on radon migration that can be enumerated but are not at all
 well understood.

        The "upward-spreading pipe" model suggested above has ramifications for other studies of
 radon in soil gas.  Except for a few client-funded studies such as this one, there is rarely if ever the
 opportunity to revisit a site and measure radon at different depths (i.e., elevations below original
 grade). Yet these revisitations and different sampling depths may be critical in trying to understand
 the subsurface migration of radon.

        The U.S.G.S. (10) has proposed (and uses) sampling soil gas for radon at a nominal depth of
 70 to 75 cm. below ground surface. In the Piedmont Province, this depth presumably puts one into
 the C soil horizon, or lowermost B and uppermost C horizons. We suggest that the relationship
 between radon concentrations in the C horizon (or any soil  horizon)  and those at  final  grade
 (Finished  Floor Elevation, or FFE) should be site-specific,  and  that  site-specific  (i.e.,  FFE)
 measurements should be taken before actual construction begins, so that (if necessary) the architect
 can incorporate either passive or active radon mitigation features into the final design.  As is well
 known, saprolite is not soil and it is not rock, yet it has some physical and chemical characteristics
 of both soil and rock. It needs to be intensively studied in its own right (14) for its contribution to
 and influence upon the vertical and lateral migration of radon from source (radium atoms) to surface.

        The work described in this paper was not funded by the U.S. Environmental Protection
 Agency, and therefore the contents do not necessarily reflect the views of the Agency and no official
 endorsement should be inferred.

                                       REFERENCES

 1.     Bobyarchick, A.R., Edelman, S.H., and Horton, J.W., The role of dextral strike-slip in
       the displacement  history of the Brevard zone.  In Secor, D.T., ed., Southeastern geological
       excursions:  Geological Society of America Southeastern Section guidebook,  1988. p. 53.

2.     Keith, A., Description of the Mount Mitchell quadrangle.  U.S. Geological Survey Atlas  Folio
       124, Washington, D.C.,  1905.  10 pp.

3.     Hatcher, R.D., Jr., Tectonics of the western Piedmont and  Blue Ridge, southern
       Appalachians: review and speculation. American Journal of Science, 278:276.  1978.

4.     Crawford, TJ. and Medlin, J.H., The western Georgia Piedmont between the Cartersville and
       Brevard fault zones.  American Journal of Science. 273:712.  1973.

5.     Hurst, V. J., Geology of the southern Blue Ridge belt. American Journal of Science  273-643
       1973.

6.     Higgins, M.W. and McConnell, K.I., The Sandy Springs Group and related rocks of the
       Georgia Piedmont:  nomenclature and stratigraphy.  In Platt, P.A., ed., Short contributions
       to the geology of Georgia.  Georgia Geologic Survey Bulletin 93, Atlanta, Georgia, 1978.
       p. 50.

-------
7.      Higgins, M.W. and Atkins, R.L., The stratigraphy of the Piedmont southeast of the Brevard
       zone in the Atlanta, Georgia area. In Wigley, P.B., ed., Latest thinking on the stratigraphy
       of selected areas in Georgia.  Georgia Geologic Survey Information Circular 54-A, Atlanta,
       Georgia, 1981.  p. 3.

8.      Higgins, M.W., Geology of the Brevard lineament near Atlanta, Georgia. Georgia Geologic
       Survey Bulletin 77, Atlanta, Georgia, 1966.  49 pp.

9.      McConnell, K.I. and Abrams, C.E., Geology of the greater Atlanta region. Georgia Geologic
       Survey Bulletin 96, Atlanta, Georgia, 1984. 127 pp.

10.    Gundersen, L.C.S. and Wanty, R.B., eds., 1991, Field studies of radon in rocks, soils, and
       water.  U.S. Geological Survey Bulletin 1971, Washington, D.C., 1991.  334 pp.

11.    McConnell, K.I., and Costello, J.O., Uranium evaluation of graphitic phyllites and other
       selected rocks in the Georgia Piedmont and  Blue Ridge.  Georgia Geologic Survey Open -
       File Report 80-5, Atlanta, Georgia, 1980.  127 pp.

12.    Gregg, L.T. and Holmes, J.J., 1990, Radon detection  and measurement in soil and
       groundwater.  In Ward, S.H., ed., Geotechnical and Environmental Geophysics.  Vol. I:
       Review and Tutorial.  Society of Exploration Geophysicists, Tulsa, Oklahoma, 1990. p. 251.

13.    Tanner, A.B., Methods of characterization of ground for assessment of indoor radon
       potential at a site.  In Gundersen, L.C.S., and Wanty, R.B., eds.. Field studies of radon in
       rocks, soils, and water. U.S. Geological Survey Bulletin 1971, Washington, D.C., 1991. p.l.

14.    Gregg, L.T., and Coker, G., Geologic controls on radon occurrence in  Georgia.  In Bearce,
       D.T. and Neilson, M.J., eds., Case studies in applied geology in the Southeastern United
       States.  Georgia Geologic Survey Bulletin 122, Atlanta, Georgia, 1990.  p.40.

-------
Session IX Posters
  Radon Surveys

-------
                                                                             IXP-1
             SUMMARY OF REGIONAL ESTIMATES OF INDOOR SCREENING

                           MEASUREMENTS OF 222pn

         by:  Barbara Alexander, Nathaniel Rodman, and S.B. White
              Research Triangle Institute
              Research Triangle Park, N.C.  27709

              Jeffrey Phillips
              U.S. Environmental Protection Agency
              Office of Radiation Programs
              Washington, D.C.  20460
                                 ABSTRACT
    As part of an EPA/State cooperative program,  a random sample of 54,851
houses from 38 of the 48 conterminous states has  been screened for 222pn
over a six year period.  The number of houses in  each state was
sufficiently large to allow meaningful estimates  to be derived for
geographic areas within each state.  Summary statistics (e.g., the
arithmetic mean and the percentage of houses exceeding 4 pCi/L) and
associated 95% confidence intervals have been computed for 225 geographic
regions within the 38 states.  This paper summarizes the 225 regional
estimates in tabular and graphical form.  The results can then be used to
identify "hot spots" within the 38-state area.  For example, there are 21
regions within seven states where more than 50 percent of the target
population houses are estimated to have screening measurements greater than
4 pCi/L.

    This paper has been reviewed in accordance with the U.S. Environmental
Protection Agency's peer and administrative review policies and approved
for presentation and publication.

-------
                               INTRODUCTION
    In response to the growing concern about potential health risks associ-
ated with indoor radon exposure, the U.S. Environmental Protection Agency
(EPA) began a program to provide assistance to states to measure radon
concentrations in homes.  Since the winter of 1986-87, EPA has assisted 38
of the 48 conterminous states in conducting probability-based surveys of
indoor 222Rn.  Short-term screening measurements were made in the lowest
livable level of over 54,000 randomly selected houses during winter heating
seasons.  The 38 state radon surveys included in this paper were carried
out during six years of the program as listed below.

              Year 1, 1986-87 heating season, eight surveys:
               Alabama  (AL)              Rhode Island  (RI)
               Kansas  (KS)               Tennessee  (TN)
               Kentucky  (KY)             Wisconsin  (WI)
               Michigan  (MI)             Wyoming  (WY)

              Year 2, 1987-88 heating season, seven surveys:
               Arizona  (AZ)              Missouri  (MO)
               Indiana  (IN)              North Dakota  (ND)
               Massachusetts  (MA)         Pennsylvania  (PA)
               Minnesota  (MN)

              Year 3, 1988-89 heating season, seven surveys:
               Georgia  (GA)              Ohio  (OH)
               Iowa  (IA)                  Vermont  (VT)
               Maine  (ME)                 West Virginia  (WV)
               New Mexico  (NM)

              Year 4, 1989-90 heating season, eight surveys:
               California  (CA)            Nevada  (NV)
               Idaho  (ID)                 North Carolina   (NC)
               Louisiana  (LA)             Oklahoma  (OK)
               Nebraska  (NE)             South Carolina   (SC)

               Year 5,  1990-91  heating season,  six surveys:
               Arkansas  (AR)             Mississippi   (MS)
               Illinois  (IL)             Texas  (TX)
               Maryland  (MD)             Washington  (WA)

               Year 6,  1991-92  heating season,  two surveys:
               Montana  (MT)              Virginia  (VA)

-------
    The goals of the state radon surveys were twofold.  Some measure of the
distribution of radon levels among residences was desired for major
geographic areas within each state and for each state as a whole.  In
addition, it was hoped that each state radon survey would be able to
identify areas of potentially high residential radon concentration ("hot
spots") in the state, enabling the state to focus its attention on areas
where indoor radon concentration might pose a significant health threat.


                                  METHODS
    Two-day deployment of open-face charcoal canisters was used by 22
states during the first three years of the state radon survey assistance
program (1).  During these years a diffusion barrier charcoal canister was
developed specifically to be less sensitive to the effects of humidity and
air flow than the open-face canister (2).  Two-day deployment of barrier
canisters was used by eight states in Year 4 of the program.  The exposure
period for the barrier canister was increased from two days to seven days
in Years 5 and 6.  The performance of the charcoal canisters was monitored
periodically through the use of unexposed canisters, canisters exposed to
known levels of 222Rn, and collocated canisters.

    To ensure the discovery of elevated radon concentrations within a home,
the charcoal canister was exposed under closed-house conditions during the
winter and was placed on the lowest livable level.  Thus, the estimates of
indoor radon concentration provided by the survey reflect a worse-case
scenario and maximize the likelihood of identifying residences with high
radon concentrations.  This method was used in the state surveys because it
is the one EPA recommends to homeowners for determining whether additional
tests are needed (3).

    For each state,  a probability sample of listed residential telephone
numbers was selected from a sampling frame constructed from telephone
directories for all  communities in the state.  Probability sampling permits
the extrapolation of survey results to the sampled population and enables
the calculation of measures of precision for the estimates.  Because one of
the goals of each survey was to characterize the distribution of
residential radon levels for the state as a whole, the use of probability
sampling was imperative.  In addition, probability-based surveys permit
valid comparisons of results from one state with those from another.

    The target population for the surveys consisted of owner-occupied
residences with a permanent foundation, at least one floor at or below
ground level, and a  telephone number published in the latest directory.
(Mobile homes with permanent foundations and airtight panels/skirts were
included beginning in Year 3.)  The statistical estimates generated from
the survey data apply to a target population of over 28.6 million homes
from 38 states.

-------
    Each state was divided into strata based on population density,
potential for high 222pn, and geographic areas for which separate
statistical estimates were desired.  For convenience in selecting the
sample of telephone numbers, county boundaries were used to delineate the
geographic reporting regions.  Each stratum was then sampled at a different
rate to ensure a wide dispersement of the sample across the state and to
enhance the chances of finding areas with elevated radon.

    Interviewers from each state placed telephone calls to random
subsamples of the residential telephone listings.  The interviewer first
screened for survey eligibility, which required that the dwelling qualify
as a member of the target population.  Eligible households agreeing to
participate in the survey were provided with a charcoal canister and
instructions for placing it on the lowest livable level of their home.
Participants were instructed to return the canister to the EPA analysis
laboratory after exposing the canister under closed-house conditions for a
designated time period.

    Because telephone numbers in different strata were selected at
different sampling rates, it was necessary to assign sampling weights that
counterbalanced the unequal selection probabilities.  The weights assigned
were the inverse of the sample selection probabilities.  An additional
weight adjustment was made to compensate for nonresponse so population
aggregates could be easily estimated from the sample data.  All data
analyses were carried out using properly weighted data that reflected the
full complexity of the sample design.  This permitted the generation of
unbiased statistical estimates (4).
                          RESULTS AND DISCUSSION
    Based on program needs and available resources,  each state determined
how many houses would be tested.  The actual number of houses that provided
valid test data are shown in Table 1 and ranged from 376 in Rhode Island to
2,680 in Texas.  In 30 of the 38 states, more than 1000 houses were tested.
The number of houses tested in each state was large enough to provide
reliable estimates of 222Rn concentrations for subpopulations of each state
(e.g., groups of counties formed by political or geologic boundaries).

    During the course of the surveys, state personnel were asked to
identify geographic regions for which estimates of 222Rn levels were
important.  These regions were composed of one or more counties (not
necessarily contiguous) with a combined sample size of at least 100 valid
screening measurements.  Table 1 shows the number of geographic regions
defined for each state which ranged from one region in Rhode Island to 13
regions in Texas.  The sample size in the regions ranged from 47 to 1,215
measurements, and only 4 of the 225 regions had fewer than 75 screening
measurements.  Summary statistics (e.g., means, percentiles and
proportions) and associated 95% confidence intervals were computed for the

-------
TABLE 1.  RANGE OF STATE RADON REGIONAL ESTIMATES FOR 38 STATES
STATE
AL
AR
AZ
CA
GA
IA
ID
IL
IN
KS
KY
LA
MA
MD
ME
MI
MN
MO
MS
MT
NC
ND
NE
NM
NV
OH
OK
PA
RI
SC
TN
TX
VA
VT
WA
WI
WV
WY
OVERALL
SURVEY
YEAR
1
5
2
4
3
3
4
5
2
1
1
4
2
5
3
1
2
2
5
6
4
2
4
3
4
3
4
2
1
4
1
5
6
3
5
1
3
1

NO. OF
HOUSES
TESTED
1,180
1,535
1,507
1,885
1,534
1,381
1,266
1,450
1,914
2,009
879
1,314
1,659
1,126
839
1,989
919
1,859
960
833
1,290
1,596
2,027
1,885
1,562
1,734
1,637
2,389
376
1,089
1,773
2,680
1,156
710
1,935
1,191
1,006
777
54,851
RANGE OF REGIONAL ESTIMATES
NO. OF
REGIONS
8
6
3
9
3
9
7
3
5
6
6
4
11
4
7
4
5
6
6
3
5
6
5
4
8
4
7
10
1
4
11
13
5
5
4
10
3
5
225
ARITHMETIC
MEAN (PCI/L)
0.7- 4.4
0.5- 1.8
1.1- 1.7
0.6- 1.5
0.8- 2.1
7.4-10.3
2.0- 5.1
2.0- 4.6
2.8- 4.2
1.1- 4.8
1.1- 4.5
0.3- 0.6
1.7- 4.6
0.7- 5.5
2.3- 5.6
1.5- 5.9
3.0- 6.3
1.9- 3.8
0.4- 1.2
4.1- 8.3
0.4- 3.4
4.8- 8.9
3.0- 7.1
2.2- 4.7
1.1- 3.2
3.2- 7.0
0.6- 1.6
2.3-17.8
3.2- 3.2
0.5- 1.5
1.0- 5.1
0.4- 3.4
0.6- 3.8
2.0- 3.2
0.4- 7.7
2.6- 4.8
1.9- 4.7
2.5- 5.4
0.3-17.8
% OF HOUSES
> 4 PCI/L
0.3-25.1
0.0- 9.1
1.2- 7.9
0.5- 5.5
0.3-10.3
61.4-86.9
8.9-30.5
9.6-35.3
18.5-36.2
3.4-41.2
2.2-34.5
0.0- 1.6
2.8-37.5
4.4-38.3
16.6-40.9
4.8-44.7
17.5-62.3
11.1-29.6
0.9- 3.8
36.4-52.5
0.6-17.9
46.1-72.4
25.5-62.9
11.2-41.6
3.8-24.7
18.3-48.4
1.1- 7.5
11.3-74.0
20.6-20.6
0.0- 7.0
0.7-29.9
0.0-22.4
0.7-27.5
10.8-19.8
1.3-45.0
14.4-44.3
10.7-31.2
12.6-51.0
0.0-86.9

-------
 225 geographic regions within the 38 states.  This data base was then
 sorted in descending order of the various parameter estimates to examine
 patterns in the regional estimates.

     As noted earlier, all households in a given state were not selected
 with equal  probability.  To counterbalance these unequal selection
 probabilities, appropriate sampling weights were calculated and used in all
 data analyses to assure unbiased statistical estimates.


 ARITHMETIC  MEAN


     Table 1 shows the range of regional  estimates for two of the parameters
 computed for each of the 38 states.   Previous papers  utilizing similar
 parameter estimates at the state level  (5,6) have shown that 222Rn
 concentrations vary widely from state to state.   It is evident from Table 1
 that radon  levels vary even more from region to  region.  For instance
 estimates of the  arithmetic mean (AM)  for the 225 regions range from 6.3
 pCi/L in Louisiana to 17.8 pCi/L in  Pennsylvania.   Twenty-three of the 38
 states  have at least one region where the estimated AM is greater than 4
 pCi/L and all  the regions in  three  states (Iowa,  Montana,  and North Dakota)
 have AMs above 4  pCi/L.   Figure 1 graphically identifies regions  with
 II  ^  ^ than  2 pC1/L'  2>  AMs between 2 and 4  pCi/L,  and  3)  AMs  greater
 than 4  pCi/L.   The lower two  groups  contain  83 and  82  regions respectively
 and there are  60  regions with AMs greater than 4  PCi/L.   In  keeping with
 the goal  of identifying  areas with elevated  levels  of  radon  or "hot spots"
 the 60  regions  with  AMs  greater than  4 pCi/L are  examined  in  greater detail
 in  Table  2  and  Figure 2.   Table 2 gives,  for each of the 60  regions,  the
 number  of houses  tested,  summary statistics  (i.e.,  AM  and  percentage of
 houses  exceeding  4 pCi/L)  and associated  confidence intervals,  and  a
 listing of  the  counties  included in  the  region.   The regions  in Table 2 are
 listed  in descending  order  according to  the  arithmetic  mean.   Figure 2
 shows, graphically, the  60  regions broken  down into three  groups  according
 to  the  level of the AM.   There  are 36 regions  with  AMs  between  4  and
 6 pCi/L,  12  regions with AMs  between 6 and 8  pCi/L, and  12 regions  with AMs
 greater than 8  pCi/L.  The  12  regions with the highest mean concentrations
 consist of  seven regions  in Iowa, three regions in  Pennsylvania,  and  one
 region each  in  Montana and  North Dakota.


 PERCENTAGE OF HOUSES  EXCEEDING 4 PCI/L


    Figures 1 and 2 focus on a region's arithmetic mean which is only one
 point on the distribution of 222Rn screening measurements.  Another point
 of particular interest on this distribution is the percentage of houses
within a region that have screening measurements exceeding 4 pCi/L.   For
 the 225 regions, this percentage varied from 0% in Arkansas to 86.9%  in
 Iowa.  Twenty-one of the 38 states have at least one region where more than
30-s of the houses have screening measurements exceeding 4 pCi/L.  More than

-------
      Arithmetic Mean in pCi/L
AM<=2              2
-------
Arithmetic Mean in pCi/L
4
-------
TABLE Z.  STATE RADON REGIONAL ESTIMATES AND CONFIDENCE INTERVALS FOR REGIONS HITH ARITHMETIC MEAN VALUES > 4 PCI/L
STATE
PA
PA
IA
IA
IA
IA
NO
IA
IA
MT
PA
IA
NO
IA
NA
IA
REGION
2
3
6
Z
1
3
4
9
8
2
7
7
1
5
1
4
NO. OF
HOUSES
TESTED
258
270
179
160
144
128
470
133
143
266
207
169
423
187
708
138
ARITHMETIC MEAN
ESTIMATED
PCI/L
17.8
10.8
10.3
10.2
9.8
9.1
8.9
8.8
8.7
8.3
8.3
8.1
7.8
7.7
7.7
7.4
9SX CONF.
INTERVAL
12.3-23.4
8.4-13.2
8.8-11.7
9.1-11.3
8.3-11.2
7.9-10.2
8.1- 9.6
7.1-10.5
6.6-10.8
6.0-10.6
5.2-11.3
6.6- 9.6
6.8- 8.7
6.3- 9.0
6.3- 9.2
6.4- 8.3
X OF HOUSES
ESTIMATED
PERCENT
74.0
51.3
81.0
82.1
86.9
71.7
72.4
61.7
67.6
52.5
40.4
61.4
62.4
62.5
45.0
67.5
> 4 PCI/L
95X CONF.
INTERVAL
68.9-79.0
49.0-53.7
75.1-87.0
71.8-92.3
82.4-91.4
63.6-79.8
68.6-76.2
55.5-68.0
52.1-83.1
44.0-61.0
37.0-43.9
55.3-67.4
56.7-68.2
57.3-67.6
43.1-47.0
63.6-71.4
LIST OF COUNTIES INCLUDED IN THE REGION
CUMBERLAND, DAUPHIN, LANCASTER, LEBANON, PERRY, YORK.
ADAMS, BEDFORD, BLAIR, CARBON, CENTRE, CLINTON, COLUMBIA,
FRANKLIN, FULTON, HUNTINGDON, JUNIATA, LYCOMING, MIFFLIN,
MONTOUR, NORTHUMBERLAND, SCHUYLKILL, SNYDER, SULLIVAN,
UNION.
BOONE, DALLAS, GREENE, GRUNDY, HAMILTON, HARDIN, JASPER,
MARSHALL, POLK, STORY, MEBSTER.
AUDUBON, CALHOUN, CARROLL, CRAWFORD, GUTHRIE, HARRISON,
IDA, MONONA, SAC, SHELBY, HOODBURY.
BUENA VISTA, CHEROKEE, CLAY, DICKINSON, EMMET, LYON,
O'BRIEN, OSCEOLA, PALO ALTO, PLYMOUTH, POCAHONTAS, SIOUX.
ADAIR, ADAMS, CASS, FREMONT, MILLS, MONTGOMERY, PAGE,
POTTAHATTAMIE , RINGGOLD, TAYLOR, UNION.
BARNES, CASS, GRAND FORKS, HIGHLAND, SARGENT, STEELE,
TRAILL.
DAVIS, DES MOINES, HENRY, JEFFERSON, KEOKUK, LEE, LOUISA,
MUSCATINE, VAN BUREN, HAPELLO, WASHINGTON.
APPANOOSE, CLARKE, DECATUR, LUCAS, MADISON, MAHASKA,
MARION, MONROE, WARREN, WAYNE.
BEAVERHEAD, BROADHATER, DEER LODGE, GALLATIN, GRANITE,
JEFFERSON, JUDITH BASIN, LEWIS AND CLARK, MADISON,
MEAGHER, PARK, POWELL, RAVALLI, SILVER BOH, SWEET GRASS,
HHEATLAND, YELLOWSTONE NAT PARK.
BRADFORD, MONROE, PIKE, POTTER, SUSQUEHANNA, TIOGA, HAYNE,
WYOMING.
8ENTON, CEDAR, CLINTON, IOWA, JACKSON, JOHNSON, JONES,
LINN, POWESHIEK, SCOTT, TAMA.
ADAMS, BILLINGS, BOWMAN, GRANT, HETTINGER, MORTON, OLIVER,
PEMBINA, SLOPE, STARK.
ALLAMAKEE, BLACK HAWK, BREMER, BUCHANAN, CHICKASAW,
CLAYTON, DELAWARE, DUBUQUE, FAYETTE, HOWARD, WINNESHIEK.
DOUGLAS, FERRY, GRANT, LINCOLN, OKANOGAN, PEND OREILLE,
SPOKANE, STEVENS.
BUTLER, CERRO GORDO, FLOYD, FRANKLIN, HANCOCK, HUMBOLDT,
                                                                  KOSSUTH, MITCHELL, WINNEBAGO, WORTH, WRIGHT.

-------
        TABLE 2.  STATE RADON REGIONAL ESTIMATES AND  CONFIDENCE INTERVALS FOR REGIONS HITH ARITHMETIC MEAN VALUES > 4 PCI/L
STATE
         NO.  OF
         HOUSES
REGION	TESTEP
                             ARITHMETIC MEAN
                          ESTIMATED   95X CONF.
                            PCI/L     INTERVAL
                              '/. OF HOUSES > 4 PCI/L
                              ESTIMATED   95X CONF.
                               PERCENT    INTERVAL
                                            LIST OF COUNTIES INCLUDED IN THE REGION
  PA


  PA

  NE


  OH




  ND


  NO


  MN



  NE




  MI

  MN
  ME

  MT



  PA


  MO

  HY
   2

   5
           268


           273

           358


           428




           210


           121
142

190
   4

   B
211

264



279


316

135
7.3


7.3

7.1


7.0




6.9


6.6


6.3



6.1




5.9

5.7
5.6

5.6



5.6


5.5

5.4
6.0- 8.6


6.5- 8.1

6.0- 8.2


5.6- 8.4




4.8- 9.0


5.6- 7.7


j.i- 6.6



5.6- 6.6




4.6- 7.1

5.2- 6.1
4.5- 6.8

3.8- 7.4



3.0- 8.2


4.6- 6.4

4.0- 6.8
                                                     44.2
                                 48.4
                                             39.9

                                             36.4



                                             27.5
                                                              39.3-49.1
                                 43.9     SB.1-49.7

                                 57.9     55.0-60.8
                                          46.0-50.7
60.8     54.3-67.2


65.3     57.5-73.1


62.3     54.4-70.3



62.9     58.9-66.8




44.7     38.8-50.6

55.4     48.7-62.1
         34.2-45.7

         30.7-42.0



         17.0-38.0
38.3     32.8-43.8

51.0     39.0-63.0
BERKS, BUCKS, CHESTER, DELAWARE, LEHIGH, MONTGOMERY,
NORTHAMPTON.

BEAVER, BUTLER, LAMRENCE.

ANTELOPE, BOONE, BOYD, BURT, CEDAR, CUMING, DAKOTA, DIXON,
HOLT, KNOX, MADISON, PIERCE, STANTON, THURSTON, WAYNE.

ATHENS, BELMONT, FAIRFIELD, FRANKLIN, GALLIA, GUERNSEY,
HOCKING, JACKSON, LAMRENCE, LICKING, MEIGS, MONROE,
MORGAN, MUSKINGUM, NOBLE, PERRY, PICKAWAY, PIKE, ROSS,
SCIOTO, VINTON, WASHINGTON.

BENSON, CAVALIER, DICKEY, EDDY, FOSTER, GRIGGS, LA MOURE,
NELSON, RAMSEY, RANSOM, STUTSMAN, WALSH, WELLS.

DUNN, EMMONS, GOLDEN VALLEY, MCKENZIE, MCLEAN, MERCER,
SIOUX.

BLUE EARTH, BROWN, DODGE, FARIBAULT, FILLMORE, FREEBORN,
GOODHUE, HOUSTON, LE SUEUR, MOWER, OLMSTED, RICE, STEELE,
WABASHA, WASECA, WINONA.

BUTLER, CASS, COLFAX, DODGE, DOUGLAS, FILLMORE, GAGE,
JEFFERSON, JOHNSON, LANCASTER, NEMAHA, OTOE, PAWNEE,
PLATTE, POLK, RICHARDSON, SALINE, SARPY, SAUNDERS, SEWARD,
THAYER, WASHINGTON, YORK.

HILLSDALE, LENAWEE, WASHTENAW.

AITKIN, BENTON, BIG STONE, CARLTON, CARVER, CHIPPEWA,
CHISAGO, COTTONWOOD, DOUGLAS, GRANT, ISANTI, JACKSON,
KANABEC, KANDIYOHI, LAC QUI PARLE, LINCOLN, LYON, MARTIN,
MCLEOD, MEEKER, MILLE LACS, MURRAY, NICOLLET, NOBLES,
PINE, PIPESTONE, POPE, REDWOOD, RENVILLE, ROCK, SCOTT,
SHERBURNE, SIBLEY, STEVENS, SWIFT, TRAVERSE, WATONWAN,
WRIGHT, YELLOW MEDICINE.

CUMBERLAND, YORK.

CASCADE, CHOUTEAU, FLATHEAD, GLACIER, HILL, LAKE, LIBERTY,
LINCOLN, MINERAL, MISSOULA, PONOERA, SANDERS, TETON,
TOOLE.

CAMERON, CLARION, CLEARFIELD, CRAWFORD, ELK, ERIE, FOREST,
JEFFERSON, MCKEAN, MERCER, VENANGO, WARREN.

ALLEGANY, FREDERICK, GARRETT, WASHINGTON.

LINCOLN, NIOBRARA, SHERIDAN, WESTON.

-------
TABLE Z.  STATE RADON REGIONAL ESTIMATES AND CONFIDENCE INTERVALS FOR REGIONS HITH  ARITHMETIC  MEAN VALUES  > 4 PCI/L
STATE
ID
NO
TN
KS
ME
NO
HI
ME
NM
OH
PA
MV
IL
MA
MN
REGION
1
3
4
3
7
5
1
3
2
1
9
1
2
5
ft
t.
NO. OF
HOUSES
TESTED
66
131
245
195
10Z
241
101
95
422
445
261
211
539
219
160
ARITHMETIC MEAN
ESTIMATED
PCI/L
5.1
5.1
5.1
4.6
4.8
4.8
4.8
4.7
4.7
4.7
4.7
4.7
4.6
4.6
4.6
95X CONF.
INTERVAL
3.4- 6.9
4.7- 5.6
3.9- 6.3
4.7- 5.0
4.0- 5.7
4.2- 5.3
4.1- 5.5
1.9- 7.5
4.3- 5.1
4.1- 5.3
3.7- 5.6
3.2- 6.2
3.7- 5.5
3.9- 5.2
3.9- 5.2
X OF HOUSES
ESTIMATED
PERCENT
30.5
46.1
29.9
38.9
40.9
47.4
34.9
33.7
41.6
36.7
27.4
31.2
35.3
37.5
41.7
> 4 PCI/L
95X CONF.
INTERVAL
27.4-33.5
37.7-54.6
24.3-35.6
33.3-44.5
31.9-49.9
40.4-54.4
24.6-45.2
27.2-40.1
36.9-46.3
33.3-40.0
23.3-31.6
23.4-39.0
29.2-41.3
34.0-41.0
31.9-51.5
LIST OF COUNTIES INCLUDED IN THE REGION
BENEWAH, BONNE R, BOUNDARY, KOOTENAI, SHOSHONE.
BOTTINEAU, KIDDER, LOGAN, MCHENRY, PIERCE, ROLETTE,
TOWER.
DAVIDSON.
CLAY, CLOUD, DICKINSON, ELLSWORTH, GEARY, JEHELL, LINCOLN,
MITCHELL, MORRIS, OSBORNE, OTTAWA, REPUBLIC, RILEY,
RUSSELL, SALINE, SMITH, WASHINGTON.
AROOSTOOK.
BURKE, BURLEIGH, DIVIDE, MCINTOSH, MOUNTRAIL, RENVILLE,
SHERIDAN, WARD, WILLIAMS.
MARATHON, PORTAGE.
FRANKLIN, OXFORD, SOMERSET.
COLFAX, HARDING, MORA, RIO ARRIBA, SAN MIGUEL, SANTA FE,
TAOS, UNION.
ALLEN, AUGLAIZE, CRAWFORD, DEFIANCE, DELAWARE, ERIE,
FULTON, HANCOCK, HARDIN, HENRY, HURON, KNOX, LUCAS,
MARION, MERCER, MORROW, OTTAWA, PAULDING, PUTNAM,
RICHLAND, SANDUSKY, SENECA, UNION, VAN WERT, WILLIAMS,
WOOD, WYANDOT.
ALLEGHENY.
BERKELEY, GRANT, GREENBRIER, HAMPSHIRE, HARDY, JEFFERSON,
MERCER, MINERAL, MONROE, MORGAN, PENDLETON, POCAHONTAS,
RANDOLPH, SUMMERS.
ADAMS, BROWN, CASS, CHAMPAIGN, CHRISTIAN, CLARK, COLES,
CUMBERLAND, DE HITT, DOUGLAS, EDGAR, FORD, FULTON,
HANCOCK, HENDERSON, HENRY, KNOX, LOGAN, MACON, MASON,
MCDONOUGH, MCLEAN, MENARD, MERCER, MORGAN, MOULTRIE,
PEORIA, PIATT, PIKE, ROCK ISLAND, SANGAMON, SCHUYLER,
SCOTT, SHELBY, STARK, TAZEWELL, VERMILION, WARREN,
WOOD FORD.
WORCESTER.
BECKER, BELTRAMI, CASS, CLAY, CLEARWATER, CROW WING,
                                                                  HUBBARD,  ITASCA,  KITTSON,  KOOCHICHING,  LAKE OF THE WOODS,
                                                                  MAHNOMEN, MARSHALL,  MORRISON,  NORMAN, OTTER TAIL,
                                                                  PENNINGTON,  POLK, RED  LAKE,  ROSEAU,  STEARNS, TODD, WADENA,
                                                                  WILKIN.

-------
        TABLE 2.   STATE RADON REGIONAL ESTIMATES  AND CONFIDENCE INTERVALS FOR REGIONS HITH ARITHMETIC MEAN VALUES > 4 PCI/L
                 NO.  OF
                 HOUSES
STATE   REGION   TESTED
                   ARITHMETIC MEAN
                ESTIMATED    95X  CONF.
                  PCI/L      INTERVAL
                               '/. OF HOUSES > 4  PCI/L
                               ESTIMATED   9SX  CONF.
                                PERCENT    INTERVAL
                                                                         LIST OF COUNTIES INCLUDED IN THE REGION
  NY

  KY



  HI

  AL


  KS




  TN


  IN
  NE



  PA


  ID


  IN



  MA

  MT
 4

 4



 7

 6
11
 3

 5



 5


 5


 2



 6

 3
108

143



118

156


190




ZOO


448
298

463



249


 70


456



391

303
4.6      3.1- 6.0      35.9     25.9-45.9   ALBANY,  CONVERSE, GOSHEN.

4.5      3.8-5.2      34.5     28.0-41.0   ANDERSON,  BOURBON, BOYLE, BRACKEN, CLARK, FAYETTE,
                                            FRANKLIN,  HARRISON, MADISON, MERCER, MONTGOMERY, NICHOLAS,
                                            ROBERTSON, SCOTT, SHELBY, WOODFORD.
4.5

4.4


4.4




4.3


4.2
4.2

4.2



4.2


4.1


4.1



4.1

4.1
3.6- 5.4

0.0-10.1


3.8- 4.9




3.7- 5.0


3.5- 4.9
3.8- 4.6

4.0- 4.5



3.6- 4.8


2.4- 5.7


3.7- 4.6



3.8- 4.5

3.7- 4.6
44.3     38.8-49.9

 7.3      2.8-11.8
                                           39.8
                                           29.6
                                           33.7
                                                   34.7-44.9
                                                    24.0-35.3
                                                    27.4-40.0
40.1     33.<»-*6.9

40.8     36.3-45.3



29.4     23.7-35.2


22.9     14.6-31.2


36.2     30.5-41.9



25.2     20.3-30.0

37.6     34.3-40.9
KENOSHA, RACINE, NALNORTH, HAUKESHA.

AUTAUGA, CALHOUN, CHILTON, CLAY, CLEBURNE, COOSA,  ELMORE,
RANDOLPH, TALLADEGA, TALLAPOOSA.

CLARK, COMANCHE, EDHARDS, FINNEY, FORD,  GRANT,  GRAY,
GREELEY, HAMILTON, HASKELL, HODGEMAN, KEARNY, KIOHA,  LANE,
MEAOE, MORTON, NESS, PAWNEE, RUSH, SCOTT, SEMARD,  STANTON,
STEVENS, WICHITA.

CARTER, GREENE, HANCOCK, HAHKINS, JOHNSON, SULLIVAN,
UNICOI, WASHINGTON.

BLACKFORD, BOONE, BROWN, CLINTON, DEARBORN, DECATUR,
DELAWARE, FAYETTE, FOUNTAIN, FRANKLIN, GRANT, HAMILTON,
HANCOCK, HENDRICKS, HENRY, HOWARD, JAY,  JOHNSON, MADISON,
MARION, MONTGOMERY, MORGAN, OHIO, OMEN,  PUTNAM, RANDOLPH,
RIPLEY, RUSH, SHELBY, SWITZERLAND, TIPTON, UNION,  MARREN,
WAYNE.

ANOKA, DAKOTA, HENNEPIN, RAMSEY, WASHINGTON.

ADAMS, BUFFALO, CLAY, FRANKLIN, GARFIELD, GREELEY, HALL,
HAMILTON, HOWARD, KEARNEY, MERRICK, NANCE, NUCKOLLS,
SHERMAN, VALLEY, WEBSTER, WHEELER.

ARMSTRONG, CAMBRIA, FAYETTE, GREENE,  INDIANA, SOMERSET,
WASHINGTON, WESTMORELAND.

BLAINE, CAMAS, CASSIA, GOODING, JEROME,  LINCOLN, MINIDOKA,
TWIN FALLS.

ADAMS, ALLEN, CASS, DE KALB, ELKHART, FULTON, HUNTINGTON,
KOSCIUSKO, LAGRANGE, MARSHALL, MIAMI, NOBLE, STEUBEN,
WABASH, WELLS, WHITLEY.

MIDDLESEX.

BIG HORN, BLAINE, CARBON, CARTER, CUSTER, DANIELS, DAWSON,
FALLON, FERGUS, GARFIELD, GOLDEN VALLEY, MCCONE,
MUSSELSHELL, PETROLEUM, PHILLIPS, POWDER RIVER, PRAIRIE,
RICHLAND, ROOSEVELT, ROSEBUD, SHERIDAN,  STILLWATER,
TREASURE, VALLEY, WIBAUX, YELLOWSTONE.

-------
one-half of the houses in 21 regions (from seven states) have screening
measurements greater than 4 pCi/L.  Figure 3 graphically identifies regions
where the estimated percentage of houses with screening measurements
greater than 4 pCi/L is 1) less than 15%, 2) between 15 and 30%, and
3) greater than 30%.  Of the 225 regions, 103 fall into the lowest group,
66 regions fall into the middle group,  and 56 regions have at least 30% of
the houses with screening measurement greater than 4 pCi/L.  In order to
make a more meaningful identification of "hot spots", the 56 regions in the
upper group are examined in more detail.  Figure 4 shows the top 56 regions
broken down into three groups.  There are 35 regions where 30-50% of the
houses have screening measurements greater than 4 pCi/L, 15 regions where
50-70% of the houses have screening measurements greater than 4 pCi/L,  and
there are six regions where more than 70% of the houses have screening
measurements exceeding 4 pCi/L.  These six regions have estimated AMs
greater than 8.8 pCi/L and appear among the first seven regions listed in
Table 2.
                                CONCLUSIONS
    Over a six year period, 38 of the 48 conterminous states completed
statistically-designed surveys to characterize the distribution of indoor
radon.  These surveys have produced screening measurements in over 54,000
randomly selected houses.  An important element of each state survey was
the inclusion of enough test houses to permit the distribution of radon to
be characterized within geographic regions within each state.  As a result,
the 38 states were divided into 225 regions.

    This paper examined the screening measurements in these 225 regions for
the purposes of identifying specific locations or patterns of locations of
elevated radon levels.  A listing is provided of the 60 regions (and the
counties making up the regions) having arithmetic means exceeding 4 pCi/L.
The first 24 regions listed are particularly important because they make up
the top two groups highlighted in Figure 2  (i.e., arithmetic means over
6 pCi/L).  Collectively, these 24 regions may be characterized as having

    1)   2.8 million houses which had a positive chance of being tested,

    2)   an arithmetic mean (of short-term measurements) of 8.9 pCi/L,

    3)   a geometric mean (of short-term measurements) of 4.9 pCi/L,

    4)   a median (of short-term measurements) of 5.1 pCi/L, which
         translates to 1.4 million houses exceeding 5.1 pCi/L,

    5)   1.6 million houses (57.3%) exceeding 4 pCi/L, and

    6)   240,000 houses  (8.6%) exceeding 20 pCi/L.

-------
      Percent >4 pCi/L            PCT<15         •Mii 15<-PCT<30
                                    White areas not tested
30< "Pa-
3.  Distribution of Percentage of Houses with Screening Measurements >4 pCi/L in 225 Regions

-------
     Percent >4 pCi/L
30<>PCT<50            50<=PCT<70    MBBM 70<-PCT
            White areas not in top regions or not tested
Figure 4. Distribution of Percentage of Houses with Screening Measurements >4 pCi/L in the Top 56 Regions

-------
                                REFERENCES
1.  Gray, D.J. and Windham, S.T.  EERF standard operating procedures for
    radon-222 measurement using charcoal canisters.  EPA-520/5-87-005, U.S.
    Environmental Protection Agency, Washington, DC, 1987.

2.  Gray, D.J. and Windham, S.T.  NAREL standard operating procedures for
    radon-222 measurement using diffusion barrier charcoal canisters.  EPA-
    520-5/90-032, U.S. Environmental Protection Agency, Washington, DC,
    1990.

3.  U.S. Environmental Protection Agency, U.S. Department of Health and
    Human Services, and U.S. Public Health Service.  A citizen's guide to
    radon (second edition):  the guide to protecting yourself and your
    family from radon.  EPA 402-K92-001, U.S. Government Printing Office,
    Washington, DC, 1992.

4.  Shah, B.V.  Software for survey data analysis.  Am. Stat. 38:68-69,
    1984.

5.  White, S.B.,  Bergsten,  J.W., Alexander,  B.V.,  and Ronca-Battista, M.
    Multi-state surveys of indoor 222Rn.  Health Phys. 57(6):891-896, 1989.

6.  White, S.B. Bergsten, J.W., Alexander, B.V., Rodman, N.F.,  and
    Phillips, J.L.  Indoor 222Rn concentrations in a probability sample of
    43,000 houses across 30 states.  Health  Phys.  62(1):41-50,  1992.

-------
                                                                    IXP-2
                   Texas Residential Radon Survey

     Charles Johnson,  Ph.D., Gilbert Ramirez, Dr.P.H., and
     Terry Browning
     Southwest Texas  State  University
     San Marcos, Texas  78666

     Gary Smith, Ph.D.,  Paul Breaux, and  Vonya Boykin
     Texas Department  of Health
     Austin, Texas 78756-3189

                             ABSTRACT

     Exposure to elevated levels of indoor radon in residential
structures has been suggested by many researchers to pose a public
health risk and be related  to a potential increase in the
incidence of lung cancer. To thoroughly examine this issue, the
Texas Department of Health,  with the help of an EPA grant,
commissioned a statewide survey of indoor residential radon to
determine the extent of the problem in Texas, and to identify
potential "hot spots." This report examines the radon measurements
from over 2900 randomly selected Texas homes measured during the
winter of 1991.  Texas homes, when viewed  on a statewide basis,
have a relatively low level of radon, averaging 1.2 pCi/1  (pico
curies per liter).  Such levels are not a  major public health
concern,  as it would be extremely costly  and difficult to achieve
lower average residential levels on a statewide basis. This Texas
average is within the national norms, where U.S.  homes have been
reported to have average indoor radon levels between 1.0 and 2.0
pCi/1.  However,  when examined on a county basis,  several areas of
Texas are identified where  local geology  is suspected of
contributing to elevated levels of indoor radon.  The Panhandle
area of Texas,  especially those counties  clustered in a band
through its center, is shown to have the  highest potential for
indoor radon. This area of  the state is the only area to report
any sizable number of homes with radon over 20 pCi/1.
Correspondingly,  it's also  the area of the state with the greatest
number of homes measuring over 4 pCi/1. Other areas of the state
with a potential (based on  geology)  for elevated radon levels
include the Big Bend area (also based on  survey results), Llano
Uplift, and the uranium mining areas in South Texas.

     Since indoor radon in  Texas is a localized problem, efforts
to educate citizens about the potential dangers of radon can be
focused most effectively in those counties with elevated radon
potential.  For the most part,  the areas of Texas where radon
levels are highest are also areas of lower population density,
minimizing the public health risk and maximizing the potential to
find and correct any threats to the public health.

                   History  of the Radon Question

     For many years, radiation scientists and epidemiologists have
noted a strong correlation  between exposure to elevated

-------
concentrations of radon  (including radon decay products) and
increased risk for lung cancer among underground uranium miners
 (5,6,12).  More recently, studies by other scientists have  found
higher than expected radon concentrations in homes in various
parts of the United States  (1,2,4,7,8,11). This has led public
health specialists to a concern that radon exposure within  our
homes may be a harmful health risk comparable to that experienced
by many underground miners. The EPA has suggested that indoor
radon is the most serious environmental carcinogen which the EPA
must address for the general public (9). While cigarette smoking
is recognized as the principle agent responsible for lung cancer,
accounting for about 85% of the lung cancer deaths in the United
States; exposure to radon gas has been suggested as a major agent
involved in the remaining fifteen percent (3). In response  to the
public concern about the potential harmful effects of radon, the
Environmental Protection Agency (EPA)  began a campaign in 1986 to
determine the average radon exposure for homes within the United
States and to encourage all citizens to test their homes for
indoor radon.

     In 1986, the Texas Department of Health  (TDH) began
distribution of indoor radon information in response to requests
from Texas citizens.  In June 1989, the Governor of the State of
Texas designated the Texas Department of Health, Bureau of
Radiation Control (BRC)  the lead agency for evaluation and  further
analysis of the potential for indoor radon in Texas.  Following
that designation,  the BRC applied for additional funding from the
EPA to conduct further analysis and testing of indoor radon in
Texas homes.  An EPA grant was awarded to TDH in April,  1990, in
part to fund a statewide survey of indoor radon. The results from
that statewide survey of indoor radon are the subject of this
report.  The survey was designed to address two questions:

(1) What is the average radon concentration in Texas homes,  and

(2) Are there any Texas regions of higher radon potential ("hot
spots")?

                      The Texas Survey Design

     The overall objective of the Texas state survey of indoor
radon is to respond to the public's questions and concerns about
indoor radon exposure.  The potential risk associated with long
term exposure to elevated radon concentrations is an increased
risk for lung cancer.  A statewide screening survey was designed to
define; (1)  the statewide average indoor radon concentration in
homes,  and (2)  the regional average indoor radon concentrations
(to identify "hot  spots")  by county and other geographical regions
where appropriate.

The Survey Sample

     To simplify survey procedures, only owner-occupied houses,

-------
selected at random, were surveyed  in  this study. Rental houses
were excluded because of the need  to  simplify survey procedures
and avoid the problems of gaining  permission from house owners to
conduct radon measurements. In addition, high-rise structures,
apartments, and group quarters were excluded from the study in
order to create a uniform sampling population. These exclusions
did not materially affect the statistical basis of the survey, and
had the advantage of making a complex undertaking feasible in a
reasonable amount of time.

Survey Method

     Indoor radon measurements were made with charcoal canisters
(CO supplied by the EPA. Measurements were made under
closed-house conditions using EPA  screening survey protocols. In
addition, long term alpha track detectors (ATD)  are being used to
measure twelve-month radon levels  for a subset of the survey
sample. Both the canisters and ATD detectors are returned to the
EPA, which is responsible for the  analysis of radon levels.

Regional Sampling Plan

     A regional sampling plan, based  largely on geological
potential for indoor radon, was produced to aid in the random
allocation of radon detectors, and to help provide a basis for
interpretation of survey results.  Survey staff examined available
geological and population data for the state of Texas and produced
a map which grouped all Texas counties with respect to their
potential for indoor radon. Contiguous counties with a similar
potential for radon were grouped into regions. Large metropolitan
areas were designated as regions to control their dominance of the
sample due to larger populations.  Figure 1 shows the regional
sampling areas developed in consultation between the EPA,  Research
Triangle Institute, and the Texas  researchers.

     Telephone lists were generated randomly, with all residents
within a defined region having an  equal chance of being chosen.
With the total number of homes needed for the survey known, it
then became a matter of choosing a proportional sampling plan to
insure that the large metropolitan areas did not dominate the
survey. Each wave of fifty names contained residences from
throughout the state in proportion to the percentages for the
regions designated by the sampling plan. The large metropolitan
areas were sampled at a lower percentage to insure that rural
areas would have adequate numbers  in  the survey. Even with 13
sampling regions (more than any other state), we may have had to
include some sparsely populated  counties  in a  region with  more
populous counties;  leading to the probability of some counties not
being sampled.  Financial constraints  prevented more regions from
being used in the sampling plan.

     Home owners were contacted by telephone in each region on a
random basis at a frequency determined by the county's radon

-------
potential and population density. The results of the indoor radon
survey were analyzed and reported at the county level in order to
better understand identified "hot spots."

                  Figure  1.  Regional  Sampling  Plan
 1  -  Southwest Texas
 2  -  El  Paso
 3    Big Bend              10
 4  -  West Texas  Shales     11
 5  -  North  Texas          12
 6  -  Dallas/ Fort  Worth    13
 7  -  East Texas
  Llano Uplift
 • Central Texas (Austin - San Antonio)
- Tertiary Sands Cresent
  Harris County (Houston)
- Gulf Coast
- Texas Panhandle

-------
Sample Size and Allocation of Detectors

     Survey staff initially placed over 4000 charcoal canisters
throughout the state for the measurement of indoor radon,  and over
300 Alpha Track Detector's (ATD's) for the determination of radon
levels for a twelve-month period. Those homes receiving the twelve-
month ATD measurement devices will also receive a charcoal
canister for each climatic season, or a total of four charcoal
canisters for the year. The routine placement of charcoal
canisters was accomplished through the use of "waves" or lists of
50 telephone numbers (with names and addresses)  supplied by the
Research Triangle Institute under contract to the EPA. In order to
avoid introducing statistical bias, all telephone numbers in each
opened wave were called in a search for working numbers and
eligible participants.  Initial refusals to take part in the study
were called again, at different times in hopes of better
explaining the importance of the study and gaining the homeowner's
cooperation. Up to six tries were made to convince home owners of
the importance of participation in the survey.

     Although the placement of detectors was accomplished by using
a random telephone list, the actual number of canisters placed in
any given geographical region depended on the regional potential
for radon and the population density. To insure samples in rural
areas of the state and identify these rural "hot spots," it was
necessary to sample at a higher density in these rural regions
which may have a higher geological potential for radon, but a
lower population density; and at a lower frequency in densely
populated metropolitan areas. Because of this objective to
identify rural "hot spots," the statewide sample cannot be
considered a true proportional sample of the state, nor truly
representative of the state as a whole. Therefore, since the rural
areas of the state were sampled in greater numbers dis-
proportionately to the metropolitan areas, the statewide
percentages cannot simply be multiplied to determine the number of
homes in the state with elevated indoor radon. Percentages
measured within individual counties, however, can be used as
determinants of the potential radon problem for that county, since
all residents within a specific county had an equal chance of
being chosen for the survey. Readers should be cautioned however,
that counties with fewer than five measurements are still
tentative at best.

Quality Assurance/Quality Control

     Precision  for radon analysis  was  established by placement and
analysis of duplicate samples in  survey staff member's homes.
Accuracy estimates were provided by  the EPA's Montgomery, Alabama
laboratory as part of their routine  QA/QC program. Blank samples
were submitted for radon spiking and analysis at the rate of 2
percent of the total canisters placed, and were selected from
throughout all canister shipments.

-------
Staffing for the Survey

     The Texas Department of Health contracted with health
researchers in the Department of Health Administration at
Southwest Texas State University. Working from a central location
on the SWT campus near Austin (the location of the Department of
Health headquarters), part-time graduate and under-graduate
students operated the telephone banks,  which were the heart of the
survey. On-site supervision was provided by SWT faculty; and
project oversight was the responsibility of TDK.

     SWT health researchers presented a proposal based on the
placement of up to a maximum of 4500 CC's and 450 ATD's. TDH staff
evaluated the proposal and negotiated an interagency contract with
SWT. Personnel from both SWT and TDH received appropriate radon
survey training from the EPA and Research Triangle Institute. EPA
guidelines for a random statewide survey were followed. Potential
survey participants were first contacted by mail, which was
followed by telephone interviews to confirm eligibility. SWT staff
mailed out radon detectors to eligible participants and confirmed
their use. Under TDH agreement,  SWT received copies of all
analytical reports from EPA (TDK received duplicates), encoded
them in a computer database,  and mailed results to all survey
participants.

Standard Forms

     Standard form letters were used to announce the house survey
to potential participants and to return radon analytical results
to participating home owners.  EPA-supplied questionnaires were
used to gather demographic and other relevant information from
eligible participants. These data were entered into machine
readable form by the staff of the Research Triangle Institute,
under contract to the EPA, and are due to be returned to the SWT
health researchers for further analysis.

Data Management and Analysis

     Texas Department of Health worked with SWT and the EPA to
develop an information storage and management system for all radon
related data. Specifically,  relevant questionnaire responses and
radon analytical results were stored in a computer database.
Radon analyses were done by EPA's National Air and Radiation
Environmental Laboratory in Alabama and finished data sets were
returned to TDH and SWT.
             Results  and  Discussion  of  the Texas  Survey

     Charcoal canisters were mailed to over 4000 homeowners
throughout Texas. Of these canisters,  2692 valid winter
measurements were returned. The attrition in the sample was due to
a variety of reasons, including some homeowners who delayed

-------
conducting the tests until past the winter season, home owners
deciding to not take part after agreeing to the test, or canisters
being lost in the mail. Results from those homeowners who delayed
their home measurements and results delayed for other reasons will
be addressed in a future report.

Statewide Results

     When examined statewide, Texas has a low level  of indoor
radon in homes, with an arithmetic mean of 1.2 pCi/1 reported.
The percent of Texas homes tested during this survey with a radon
level above 4.0 pCi/1  (the threshold of concern by EPA definition)
is 4.7 percent (Table  1). Furthermore, 0.2 percent of the Texas
homes tested in this survey have a radon level above 20 pCi/1.
These results should be examined carefully because of the decision
not to use exact proportional sampling.

     Disproportionate  sampling was used to insure that less
densely populated areas of the state  (such as West Texas) were
over sampled; so that  potential hot spots would not  be missed.
Many of these potential hot spots were in rural areas and called
for disproportionate sampling because of the low population
density and widely separated areas between towns or  homes. Because
of the sampling strategy used, the survey findings,  as reported in
this report, cannot be generalized to make a statement about all
homes in Texas.  These preliminary findings are useful, however,
to help identify which geographical regions in the state have a
higher potential and concern for indoor radon.

Arith.
Mean
pCi/1
1.2

Geo.
Mean
oCi/1
0.6
Table

Median
pCi/1
0.7
1. Texas Statewide Results
75th
Percentile
DCl/1
1 .4
90th
Percentile
pCi/1
2.7

% Houses %Houses
>4 oCi/1 >20 DCi/1
4.7 0.2
County Results

      When statewide data were  analyzed on a  county  basis,the
geological potential  for radon becomes  more  evident.The sampling
plan  used  in the survey  was a balance  between geological potential
for radon  and population density  factors.In  many cases, counties
in Texas were grouped into larger regions  identified by their pre-
survey geological potential for radon,  and weighted dis-
proportionately according to their population density.Within these
defined larger groups of counties, all  homes had an equal chance
of being  selected for participation  in the survey.In some cases,
more  heavily populated counties within these regions overshadowed

-------
the less populated counties, causing some counties to have low
numbers or no homes participating in the survey. Due to the large
number of Texas counties, it was necessary to group them into
these larger regions for sampling cost considerations. In those
counties where measurements were not made as part of the random
sample reported in this report, additional canisters were placed
on a volunteer basis, and will be reported and discussed in a
future report.

     In general, the Texas counties which have higher potential
for residential radon are found in the West Texas Panhandle
region; the Big Bend area; the Llano Uplift area; and inland from
the coastal bend in South Texas where underground formations of
ancient Mesozoic beach sands, rich in uranium, can be found. All
the counties where higher levels of radon were found have geology
which supports their higher potential. Valid radon measurements
were collected from 229 out of the 254 Texas counties.

     Counties where a calculated average level of radon exceeds
the 4.0 pCi/1 threshold level of concern as identified by the
Environmental Protection Agency include Carson  (1 of 4
measurements above 4.0 pCi/1), Hale (7 of 14 measurements above
4.0 pCi/1),  Randall  (7 of 20 measurements above 4.0 pCi/1),
Sherman (4 of 4 measurements above 4.0 pCi/1), and Swisher  (2 of 5
measurements above 4.0 pCi/1). All of these counties are found in
the Central Panhandle region of Texas.

     Examination of the Texas counties map  (Figure 2) for
percentage of homes with radon measurements above 4.0 pCi/1
clearly shows a greater potential for elevated indoor radon in the
Texas Panhandle region.

     The  counties of Jeff Davis  (3 of 16 measurements above 4.0
pCi/1), Presidio (8 of 44 measurements above 4.0 pCi/1), and
Brewster  (10 of 60 measurements above 4.0 pCi/1) are all found in
the Texas Big Bend region, and have subsurface geology which
support a higher potential for indoor radon. The counties of Mason
(2 of 19 measurements above 4.0 pCi/1) and Llano  (7 of 47
measurements above 4.0 pCi/1) are both in the Llano Uplift region,
and also have local geology supportive of radon production.

-------
       Figure 2 Texas Counties   Percent of Homes > 4.0 pCi/1
   SSSS
   ssss
   s/ss
Less than 10 %

10.1 to 20 %

20.1 to 50 %

More than 50 %

One or no valid measurements
     Other counties showing elevated potential for indoor radon
have only a few measurements,  but they are reported here since they
have surface and subsurface geology which could provide a source
for radon, and there is the likelihood that some homes in these
counties have elevated radon levels. Because of these geological
sources and the likelihood that some homes in these counties have
elevated radon levels, the counties are shown on the map. However,
because of the small number of measurements, they should be taken
as inconclusive without further and more numerous measurements.
These counties include the Coastal Bend and South Texas counties of
La Salle  (1 of 2} measurements above 4.0 pCi/1), Karnes  (I of 3

-------
measurements above  4.0 pCi/1), Victoria (1 of 8 measurements  above
       Figure 3. Texas Counties   Percent of  Homes > 20 pCi/1
       .
     SSfS.
     SSSS*
     ssss.
     ssss
Less than 10 %


10.1 to 20 %


20.1 to 50 %


One or no valid measurements

-------
4.0 pCi/1), and Lavaca  (1 of 8 measurements above 4.0 pCi/1). In
each of these counties, only one measurement was found above 4.0
pCi/1. Further volunteer placements will be reported in a future
report in an effort to better define the extent of indoor radon for
these counties.

     Three additional Texas counties of Bastrop  (1 of 10
measurements above 4.0 pCi/1), Brown  (2 of 6 measurements above 4.0
pCi/1), and Taylor (3 of 26 measurements above 4.0 pCi/1) had at
least 10 percent of their measurements above 4.0 pCi/1. Examination
of Texas county map for the percentage of homes over 20 pCi/1
(Figure 3) shows only two counties with potentially more than 10
percent of the homes at this level. Hale  (2 of 14 measurements
above 20 pCi/1) and Carson  (1 of 4 measurements above 20 pCi/1) are
both in the earlier identified Texas Panhandle region of higher
radon potential.

                   Conclusions  and  Recommendations

     The measurements taken in this survey followed EPA guidelines
for closed home testing during winter months, thus leading to a
higher than normal accumulation of indoor radon. More
realistically, the true annual radon  levels and occupant exposures
for these Texas homes would be lower  than measured. The numbers
reported here represent the higher radon potential for these homes,
not necessarily the radon levels experienced by the home owners on
an annual basis. It should be remembered  that the EPA guidelines
for health risk are based on a 70-year lifetime exposure to  an
average radon concentration, and a 75 percent house occupancy rate.
Less exposure would result  in a lower estimate of lung cancer risk.
Given  that the survey protocol for the data reported in  this report
required that the  canisters be placed during the winter months  (in
closed-house conditions), and that in all likelihood average indoor
radon  levels on an annual basis would be  up to 25 percent less, as
reported by Rood,  et al  (10), the annual  risk levels for the homes
measured in this survey are  less than the initial data would
suggest.

      For this preliminary report, geology is the only  factor
examined in relation to  indoor radon  levels. A future  report will
examine other  factors  such  as type of dwelling or presence of a
crawl  space.  The  Panhandle  High Plains  area of  Texas  has the
highest potential  for  indoor  radon, with  some counties such  as
Hale,  Swisher, Randall, Carson, and Sherman reporting  at least  25
percent  (or higher) of  the  homes with indoor radon levels above 4.0
pCi/1. These  findings  are consistent  with surface and  subsurface
Permian shales with higher  potential  for  uranium.

      Another part  of the  state with elevated potential  for indoor
radon is  the  Big Bend  area.   The three  counties  of Brewster, Jeff
Davis, and Presidio all had at  least  15  percent  of the homes
measured with  radon  levels  above 4.0  pCi/1. Radon  in these counties
most  likely  results  from uranium in Tertiary  igneous intrusions and

-------
shales from Paleozoic periods.  The Davis Mountains are noted for
their Tertiary basaltic lavas and volcanic mud flows. Numerous
faults and fractures are present in this region,  having contributed
to the upward movement of molten rock and volcanic vents. These
faults and fractures act as conduits for upward migration of deeper
sources of radon.

     Other parts of the state with elevated potential for indoor
radon include counties above a crescent of Tertiary sands which
parallels the Texas Gulf coast. This crescent extends from deep in
South Texas over toward the Texas-Louisiana border. Fortunately,
most of the counties did not exhibit elevated measurements for
indoor radon. Counties above these sands, which did show some
elevated potential, include La Salle and Karnes.  Karnes county has
been the site of commercial uranium mines for years. For a very
limited number of measurements, other counties in this region show
a small potential for elevated indoor radon. Lavaca and Victoria,
both, report one out of eight canisters above 4.0 pCi/1.

     Overall, Texas does not have as great a problem with elevated
indoor radon as other states in the Midwest and Northeast United
States. The fact that some areas of Texas report  indoor radon
levels higher than other areas is justification for further study
and analysis. The goals of this survey were to establish a
statewide radon average and to identify areas of  the state which
may have potential "hot spots." Those goals have  been accomplished.
Further research and surveys should now concentrate in the areas of
the state where elevated indoor radon was reported.

                             References

I.Cohen,  B.  L.,  and R.  S. Shaw, "Mean radon levels in US homes by
states and counties," Health Physics,  1991,  60,  243-259.

2.Cothern,  C. R.. 199G. "Indoor air radon,"  Review of Environmental
Contaminants and Toxicology,  111,  1-60.

3.Council on Scientific Affairs,  American Medical Association,
"Health  Effects of Radon Exposure," Archives of  Internal Medicine,
1991, 151,  4,  674-677.

4.Ganas,  Michael J.,  "Radon Contamination in Dwellings." In
International Journal of Environmental Studies,   1989,  Vol.  32,  pp.
247-260.

S.Harley,  N.  H.,  and J.  H.  Kariey,  "Potential lung cancer risk from
indoor radon exposure," CA,  1990,  40,  5,  265-275.

6.National Research Council.  1988.  Health Risks of Radon and Other
Internally Deposited Alpha-Emitters,  BEIR IV. National Academy
Press, Washington,
D. C.

-------
7.National Research Council. 1991. Comparative Dosimetry of Radon
In Mines and Homes, National Academy Press, Washington, D. C.

S.Nero,  A. V., M. B. Schwehr, W. W. Nazaroff, and K. L. Revzan.
1986. "Distribution of Airborn Radon-222 Concentrations in U.S.
Homes,"   Science, 234, 992-997.

9.Puskin, Jerome S., and Christopher B. Nelson. 1989.  "EPA's
Perspective on Risks from Residential Radon Exposure," JAPCA, 39,
7,  915-920.

10.Rood, A. S., J. L. George, M. D. Pearson, and G. H. Langner,
Jr.,  "Year-to-year variations in annual average indoor 222Rn
concentrations," Health Physics, 1991, 61, 3, 409-413.

11.U.S.  Environmental Protection Agency. 1988. Radon Reduction
Techniques for Detached Houses.. Office of Research and
Development, Washington, D.C., EPA/625/5-87/019, January  1988.

12.Woodward, A., D. Roder,  A. J. McMichael,  P. Crough, and A.
Mylvaganam, "Radon daughter  exposures at the Radium Hill  uranium
mine and  lung cancer rates  among  former workers, 1952-87," Cancer
Causes Control,  1991, 2, 1,  213-220.

-------
                                                                     IXP-3

                              ABSTRACT

                 Radon Survey of Oregon Public Schools

                                  by

                      R.D. Paris and G.L. Toombs
                        Oregon Health Division
To  assist in addressing the concerns of potential elevated radon levels in
public schools nationwide, the Oregon Health Division conducted a limited
study to determine the average radon levels in the schools in Oregon.

Thirty-one  schools were selected at random out of a  population of 1,190
statewide to  participate in this study during the 1990-1991  school year.
Long-term alpha track detectors were placed in each of the ground-floor and
basement classrooms to obtain the  average  radon levels this entire nine-
month school year.

The results of this  study showed that the mean radon concentrations in
Oregon public schools statewide was 1.1 pCi/liter.  This compares to a mean
of 1.2 pCi/liter for indoor radon in homes in the state.  Two of the schools
surveyed had many rooms above the 4 pCi/liter EPA guidelines.

Follow-up with these schools and their options for lowering the levels in the
elevated rooms are discussed.

-------
                                                                             IXP-4
                     QUALITY ASSURANCE  IN RADON SURVEYS

           by: William M. Yeager,  Robert M.  Lucas,  and Jane W.  Bergsten
               Research Triangle Institute
               Research Triangle Park,  NC  27709

               Frank Marcinowski,  and Jeffrey Phillips
               Office of Radiation Programs
               U.S.  Environmental  Protection Agency
               Washington,  DC  20460
                                  ABSTRACT

    This paper describes a variety of quality assurance  (QA) activities that
were performed throughout the planning and execution of  three U.S.
Environmental Protection Agency  (EPA)-sponsored radon surveys -- the National
Residential Radon Survey (NRRS), the National School Radon Survey  (NSRS),  and
the State Residential Radon Surveys  (SRRS) — to assure  that data would be of
known quality and would satisfy survey objectives.  These activities included
the preparation of comprehensive QA plans, training programs for interviewers,
performance evaluations of radon detectors, technical systems audits of
detector processing facilities and of Research Triangle  Institute  (RTI) data
entry, and audits of data quality.  The use of spiked, blank, and duplicate
detectors is described.  As a result of these activities, several problems
were identified and resolved, and the performance of the detectors over the
range of concentrations encountered in the surveys was characterized.

    The final precision and bias of the radon measurements will be of interest
to those who wish to utilize the results of the NRRS, the NSRS, and the SRRS.

    This paper has been reviewed in accordance with the U.S. Environmental
Protection Agency's peer and administrative review policies and approved for
presentation and publication.

-------
                               1.0 INTRODUCTION
     This paper describes a variety of quality assurance activities that were
 performed throughout the planning and execution of three U.S. Environmental
 Protection Agency (EPA)-sponsored radon surveys.  These activities were
 designed to provide confidence that the surveys would satisfy identified
 goals.  Integrating quality assurance into all phases of these projects
 allowed us to recognize  and resolve problems as well as to document and defend
 the quality of the data  produced.

 1.1 QUALITY ASSURANCE/QUALITY  CONTROL

     Quality assurance (QA)  activities include

     •   Defining  appropriate measures  of quality,
     •   Setting objectives  for  those measures,
     •   Specifying operating procedures capable  of  meeting  those objectives,
     •   Assessing actual  performance,
     •   Improving performance,  and
     •   Documenting
           what  was planned,
           what  was actually done,
           how it  was done, and
           how well it was done.

     Quality  control  (QC) activities include

     •   Verifying that the specified procedures are being followed, and
     •   Monitoring  the measures of quality to confirm that  the QA objectives
         are being met.

1.2  IMPORTANCE OF  QA

    Many people  believe that because  they are competent and committed to doing
good work, they  do not need a  formal  QA program.  Consequently, they may
resent  the time  and resources  required to develop and implement a QA program
for  their work.  A good QA program, however, is an investment that will pay
off in  several ways.  It is a useful  complement to project planning and test
design.  The earlier in the course of a project that one asks, -What can go
wrong with this  procedure?  How can I recognize a problem  if it occurs?  How
could I change the procedure to eliminate this possibility or make it easier
to recognize?",  the more likely the project will produce useful data.  QA
makes it easier  to recognize and resolve the problems that inevitably occur.
QA also provides documentation about the procedures used and the quality of
the data produced  that is valuable for defending the results.  This documen-
tation may be helpful for training new staff, clarifying planning with the
client, etc.

-------
 1.3  MEASUREMENT METHODS

     Two  types  of  radon  detectors were used in these surveys.   Short-term
 measurements  (2 to 7  days)  were made with EPA-supplied charcoal  canisters,
 which were  processed  by EPA laboratories.  Long-term measurements (5  to 12
 months)  were made with  alpha track detectors (ATDs),  which were  purchased from
 and  processed  by  commercial laboratories.

 1.4  EPA-SPONSORED RADON SURVEYS

 1.4.1 State Residential Radon  Surveys (SRRS)

     The  SRRS measured radon in  42 states  and 6  Indian lands between 1986 and
 1992.  These surveys  had two goals:  1) to estimate the statewide frequency
 distribution of residential screening measurements, and 2)  to  determine high
 risk areas  within states.   These were really 48 separate surveys.  The  EPA
 provided the radon detectors, the reading of those detectors,  technical
 assistance  in  the areas of  geology and physics,  and assistance with the survey
 research methodology  through a  contractor.  Research Triangle Institute  (RTI).

     RTI  designed  and  selected the sample  of houses,  developed  the  forms and
 data collection procedures,  trained the interviewers  (who were provided and
 paid by  the states),  processed  the survey forms,  calculated the  sampling
 weights, and analyzed the data.   In addition to providing the  telephone
 interviewers,  the states solicited the cooperation of households selected into
 the  sample, distributed related literature  and  the radon detectors to partici-
 pants, and  reported the results of the measurements to the participants.  The
 states mailed  each participant  a charcoal canister and instructions for
 placing  it  according  to the EPA protocols for screening measurements.   The
 canister was placed on  the  lowest livable level  of the house during the
 heating  season.   During the first four years of  the SRRS,  canisters were
 deployed for two  days.   During  the last two years,  canisters were  deployed for
 seven days.  After the  first year of the  survey,  approximately 10  percent of
 the  participating houses were selected to receive alpha track  detectors (ATDs)
 in addition to the charcoal canister;  one ATD was placed on each level  of the
 house and remained there for 12  months.

    The  target population of  the SRRS  was owner-occupied houses  with 1) a
 permanent foundation, 2) at least one  floor at or below ground level  (this
 includes houses over  a  crawlspace),  and 3)  a telephone  number  published in the
 latest directory.    This  survey  did not include rental units, mobile homes,  or
 upper-floor condominiums.

 1.4.2  National Residential Radon  Survey  (NRRS)

    The NRRS measured annual average radon  concentrations  in a probability
 sample of residences around the  U.S.   The goal of  the survey was to estimate
the  frequency distribution  of annual average radon  concentrations  in occupied
housing units.   Two explicit precision constraints  were  defined  as part of the
 study objectives:   1) the national  estimated fraction of  residences with
concentrations >  10 pCi/L should have  a standard  error Ł 50%,   if the fraction

-------
were > 0.005, and 2) for an EPA Region, the estimated fraction of residences
with concentrations > 4 pCi/L should have a standard error < 50%, if the
fraction were > 0.07.  A probability sample of 11,400 homes was selected.  An
RTI field interviewer visited each home to determine eligibility and solicit
participation.  Housing construction, heating, ventilating, and air
conditioning characteristics were determined during on-site interviews with
the residents.  ATDs were placed in 7118 households (84% of the 8444 survey-
eligible households) in the summer of 1989 by the residents, with guidance
from the RTI field interviewers.  Detectors were retrieved from 5694 of these
homes (89% of the 6419 households that remained eligible during the entire
monitoring period) after approximately 12 months.  Only detectors exposed for
11 to 13 months were used in the statistical analysis.

    The target population of the NRRS was housing units (including rental
units) and their permanent residents in all 50 states and the District of
Columbia.  To avoid including vacation homes, a housing unit had to be 'lived
in' at least nine months during the year to be survey-eligible.  Also excluded
from the NRRS were all places of residence on military bases and all institu-
tional residences (e.g., prisons, nursing homes).  Note that this survey
included housing units without ground contact, such as upper-floor apartments.

1.4.3  National School Radon Survey (NSRS)

    The NSRS measured radon concentrations in a probability sample of schools
around the U.S.  The goals of the survey were to determine 1) the frequency
distribution of radon measurements in schools nationwide,  as well as in
schools located in EPA-identified high risk areas; 2)  the relationship between
short-term and long-term measurements in schools; 3) the relationship between
ground-floor and upper-floor radon measurements; and 4) the correlation
between radon levels and construction or ventilation characteristics.  Short-
term radon concentrations were measured in 928 schools using charcoal canis-
ters provided and processed by EPA.  Canisters were shipped to participating
schools,  placed by a school contact person, exposed for 7 days, and shipped
via overnight express to an EPA laboratory for processing.  Additional long-
term measurements were made in 101 of these schools using ATDs that were
placed by RTI field interviewers in December 1990 and retrieved in late May
and early June, 1991.

    The target population for the NSRS consisted of all public schools located
in any of the 50 states or the District of Columbia, and with one or more of
grades K through 12.  For each survey-eligible school, all school buildings
containing one or more survey-eligible rooms were included in the survey
population.  Most occupied school rooms (e.g., classrooms, offices, cafe-
terias,  gymnasiums)  were considered survey-eligible.  Three types of rooms,
however,  were specifically excluded: 1) "wet' areas (e.g., lavatories,
showers,  kitchens),  2) areas used strictly for passage (i.e., hallways), and
3) areas used strictly for storage.  Survey-eligible rooms were further
classified according to whether or not they had ground contact.  Ground-
contact rooms either had a portion of the floor or a wall directly contacting
the ground, or were separated from the ground by a crawlspace that was com-
pletely enclosed.  Both canisters and ATDs were placed in all survey-eligible

-------
rooms of the 101 "canister-ATD" schools.   In  the 827  "canister-only" schools,
canisters were only placed in survery-eligible  rooms  with ground contact.

1.5 ORGANIZATION OF THIS PAPER

    This paper discusses the QA activities  for  these  three radon surveys.
Section 2.0 reviews the importance of planning  for quality assurance early in
the project.  Section 3.0 discusses a variety of activities that supported
data collection.  Section 4.0 defines data  quality indicators for these
surveys.  Section 5.0 describes the QA audits performed during the surveys.
Section 6.0 presents the data quality indicators for  the radon measurements.
Section 7.0 summarizes our QA experience with these surveys.
                     2.0 QUALITY ASSURANCE PROJECT PLANS
    Quality assurance should be integrated  into all phases of a project,
including the initial planning.  Recognizing  this, the U.S. EPA requires a
Quality Assurance Project Plan  (QAPjP)  for  all projects that make environ-
mental measurements.  This planning document  identifies the QA requirements of
a particular project, describes how those requirements will be satisfied, and
specifies how the quality will be documented.  It should also describe the
planning, monitoring, assessing, and documenting of the quality of data from
the project.  A QAPjP was prepared for  each of these surveys by the QA
coordinator and senior project staff. '2f3

2.1 PLANNING

    The QAPjPs included a description of the  objectives, design, and schedule
for each survey.  Scheduling can have a critical effect on the quality of
radon surveys.  Two of these surveys required that measurements be made during
the heating season.  This meant that schedules could not slip a month or two
without severely compromising the goals of  the surveys.  Key activities, the
QA components of those activities, the  personnel responsible, and the lines of
authority among those personnel were identified.  The QAPjPs also defined the
target populations and described the sample designs.

    Goals for data quality were specified.  This involved identifying critical
measurements and documentation  (e.g., items on the questionnaires), specifying
critical concentrations, describing the measurement methods, defining indica-
tors of data quality (e.g., completeness, precision, bias) and specifying
goals for those indicators.  For each survey, the goals for data quality
indicators were related to the overall  goals  of that specific survey.
Procedures for collecting and documenting both questionnaires and radon
measurements were described.   An important part of any large survey is
tracking the questionnaires, samples/detectors, and associated documents in
the field, office, laboratory, and archive.   Procedures for documentation and
data processing were described.  This included the editing of the question-
naires, statistical analyses, and reporting of the measures of quality.
Laboratory procedures for processing the radon detectors were also described.

-------
 2.2 MONITORING

     The QAPjPs described procedures and frequency for calibration of the radon
 detectors by the processing laboratories.  QC checks performed by RTI were
 also described.  These checks included verifying that the specified procedures
 for sampling,  documenting, and detector processing were followed.  An impor-
 tant part of monitoring data quality was designing and implementing sets of
 spiked,  blank, and collocated (duplicate) radon detectors to collect data on
 the quality of the radon measurements.  This is discussed in section 4.0.

 2.3 ASSESSING

     Each QAPjP described how to assess the quality of data and included
 procedures to calculate the data quality indicators.  Three types of audits
 were described:  1)  technical systems audits,  2)  performance evaluation
 audits,  and 3) audits of data quality.  These will be discussed in more detail
 in section 5.0.

 2.4 DOCUMENTING

     The  QAPjPs described how data quality would be documented.   Written
 reports  on all QA audits were prepared.   Circumstances that would trigger
 corrective action were specified,  and  QA reports to management were outlined.
 These QAPjPs  also included the QA plans  of the detector processing labora-
 tories as  appendices.
                        3.0  SUPPORT OF DATA COLLECTION
 3.1 DETECTOR PLACEMENT GUIDELINES

    Detector placement guidelines were developed  for  each survey.  These were
 consistent with the appropriate EPA guidance,4 but  tailored to the objectives
 of the survey.  In the SRRS, one charcoal canister  was  placed on the  lowest
 livable level of each participating house.  In the  NRRS,  an ATD was placed on
 each level of the home with a minimum of two ATDs at  different locations in
 single-level homes.  In  the NSRS, one radon detector  was  place in each  survey-
 eligible room with ground contact.  Detectors were  not  placed where they would
 be subject to drafts (e.g., near windows, doors,  air  vents,  or in hallways) or
 to high levels of moisture  (e.g., kitchens, bathrooms,  school locker  rooms).
 They were to be exposed  to air that people would  breathe,  and not placed in
 cupboards, closets, or drawers.

 3.2 TRAINING OF FIELD PERSONNEL

    Field personnel (RTI interviewers,  SRRS state coordinators, SRRS  state
 telephone interviewers) were trained to use the survey  instruments and
procedures for each survey.  This training included basic  information about
 radon and the radon detectors used in that survey,  as well  as a review  of the
detector placement guidelines.

-------
     For the SRRS,  RTI developed survey forms and data collection  procedures.
 In addition to presenting a detailed description of the SRRS procedures at the
 annual orientation program for the state coordinators,  RTI conducted in-state
 training sessions  for the telephone interviewers and clerks that  were hired by
 each state.  RTI also edited the completed survey forms as they were received
 from the states and provided immediate feedback to the state coordinators  so
 that any deviation from prescribed data collection procedures could be
 corrected immediately.

     The NRRS involved administration of a complex questionnaire to  capture
 housing unit construction,  as well as heating,  ventilation,  and air condi-
 tioning characteristics;  occupant demographic data; and the amount  of time
 each occupant spent at  different levels of their home.   Radon detectors were
 also placed in each home using a uniform protocol.  More than 150 field staff
 were thoroughly trained and tested in the proper administration of  the ques-
 tionnaire,  radon detector placement protocols,  and other survey procedures.
 The field staff were provided with training materials and tested  before the
 training sessions  to identify any specific areas that needed emphasis.   Upon
 completion of a 2-day training session,  the field staff were tested again.
 Only field staff who tested satisfactorily were allowed to work on  the survey.

     Radon detectors were  deployed by school or  school district personnel in
 about 90% of the NSRS sample schools and by trained RTI interviewers in the
 remaining 10%.   A  training  videotape was prepared by SC&A,  Inc.,  to demon-
 strate how to deploy the  charcoal canisters in  school rooms.   Detailed written
 instructions were  prepared  to supplement the videotape,  and both  were sent  to
 participating school districts and to the RTI interviewers.   Most of the field
 interviewers for the NSRS had worked on the NRRS the previous year.   RTI pre-
 pared detailed written  instructions for placing ATDs in school rooms.   In
 addition,  RTI survey specialists conducted interviewer  training sessions by
 telephone.

 3.3  PANEL MAINTENANCE

     Panel maintenance involved recontacting survey participants with long-term
 (ATD)  radon detectors periodically during the exposure  period.  They were
 asked if  there  had been any problem with the detector (e.g.,  it disappeared,
 it  fell down and the dog  chewed it)  or if their survey-eligible status  had
 changed  (e.g.,  if  they were planning to move).   In addition  to providing
 information about  the detectors  and the participants, panel maintenance
 reinforces  the  participants'  impression of the  importance  of  the  survey.

    Starting  in year  2 of the  SRRS,  in most of  the state surveys a  10%
 subsample of  homes  was designated to receive ATDs  in addition to  charcoal
canisters.   In  these  subsamples,  two to  four ATDs  were  placed in  each
participating home  for a  1-year  period.   States were encouraged to  recontact
these participants  periodically  during the year, but little panel maintenance
actually occurred  in  the year  2  surveys.   Beginning in year 3, seasonal short-
term charcoal canister measurements,  as  well  as 1-year ATD measurements, were
made  in the homes  in  the ATD subsamples.   Both  ATDs  and canisters were
deployed  in the winter and  additional  canisters were sent  to each home  during

-------
 the spring, summer, and fall. Thus, homes in the ATD subsample were contacted
 approximately every three months, ending one year after the initial ATD and
 canister deployment with a request to return the exposed ATDs.

     In the NRRS,  every participant was recontacted at roughly 3-month
 intervals.  For each recontact,  the participant was mailed a letter that
 contained a postage-paid card,  which they returned to RTI verifying the status
 of their detectors.  Participants who did not return these postcards were
 contacted by telephone.

     For the NSRS,  field interviewers placed ATDs during December 1990 in a
 sample of 101 schools and retrieved them 5 to 6 months later.  Although no
 special panel maintenance program was developed, about 4 to 6 weeks after
 placing the ATDs  the interviewers returned to the schools to deploy charcoal
 canisters for a 7-day period.  The interviewers were instructed to check the
 status of each ATD and to replace any that had been damaged or removed.
                         4.0 DATA QUALITY INDICATORS


     Current  EPA guidance5  requires  that  projects  address the five data quality
 indicators  (DQIs):  precision,  accuracy (including both systematic and random
 error),  completeness,  representativeness, and comparability.  The first three
 DQIs are quantitative;  the last  two,  qualitative.  The QAPjPs specified goals
 for the  quantitative DQIs,  which were related to  the specific objectives of
 each survey.

     Precision  is the random error of  the measurement  process and can  be deter-
 mined by replicate  measurements.  The precision of field measurements is best
 assessed using collocated  (duplicate)  detectors.   In both the NRRS and the
 NSRS, 5% of  the homes  or schools were designated  to receive collocated
 detectors.   Two detectors  were placed side-by-side at each measurement
 location.  The reported values of each collocated pair were compared  to
 determine the  precision of the detectors for  several  ranges of radon
 concentrations (see Section 6.0).   During the first year of the SRRS,  a
 subsample of participants  received  two detectors  to be collocated.
 Unfortunately,  many of  the participants  did not appreciate the purpose of
 collocated detectors,  so they  placed  them in  different locations or not
 simultaneously.  In subsequent years  of  the SRRS,  only people working on the
 survey, who had attended training sessions where  the  importance of  collocated
measurements could  be explained, received two canisters.

    Bias  is the  systematic  error of the measurement process  and can be  deter-
mined by  blind analysis of  spiked detectors.   For each survey,  RTI  arranged
 for some  detectors  purchased for that  survey  to be reserved for quality
control.  Some  of these received known exposures  at EPA radon chambers  and
were then mixed  in  with field-exposed  detectors and sent  to the processing
 laboratories.  The  chamber  exposures were designed to span the range  of
exposures expected  in the  field.  In each survey,   measurements near 4 pCi/L
were considered  especially  important,  so some detectors were exposed  at this

-------
  level.   The reported values for the spiked detectors were divided  by  the
  monitored values in the radon chamber to calculate their recovery; an ideal
  detector would have a recovery of  1.0.

      Accuracy is a  measure  of  the total  error,  both systematic and  random, in a
  measurement.   When the QAPjPs for  these surveys were prepared,  the EPA's
  Radon Measurement  Proficiency Program was  using the mean absolute  relative
  error (MARE)  to assess the accuracy of  commercially available radon
  detectors.0  RTI specified an objective for  the MARE in  each QAPJP.

      The  method detection limit  (MDL)  is  the  smallest  value that can be
  reliably distinguished from zero.   It is related to the  standard deviation of
  detectors  exposed  at  or near  zero  concentration.   In  each survey,  unexposed
  (blank)  detectors  were mixed  with  the field-exposed detectors sent to the
  processing laboratories.   The  standard  deviation of the  collocated detectors
  in the lowest  exposure range was another indicator  of the MDL.

     Completeness is a measure of the number of valid data points obtained
 versus the number  that were planned to meet the  established statistical level
 of confidence  for  the  project.  The precision of  a  survey estimate depends on
 the completeness of the survey.  For example, one of the objectives of the
 NRRS was to estimate the fraction of U.S. homes with annual average radon
 concentrations > 10 pCi/L with a relative standard error of < 0.50.  RTI
 projected that a final sample size  of 5000 homes would be necessary to meet
 this objective.  RTI then projected the fraction of addresses that  would
 produce survey-eligible homes, the  fraction of these homes that  would be
 willing to participate in the NRRS, and the fraction of residents  that would
 move during the course of the survey.  These estimates led to an original
 sample size of 11,400 homes.  The actual fractions were carefully monitored
 during tbe time when field interviewers  were contacting prospective
 participants and during panel  maintenance.

     Representativeness is  the  degree to  which data accurately and precisely
 represent the characteristic of a population.   Each of these  surveys used
 measurements in carefully  designed  samples  of the target  population to satisfy
 the  objectives of the survey.   It is important to keep the  specific objectives
 of the survey and the definition of the  target population in mind when consi-
 dering the results  of a survey.  For example,  only the NRRS was  designed to
 provide data on the time people spent  where radon was  monitored.

    Comparability is  the confidence  with which one data set can be compared
 with  another.   One  must consider the measurement  method and the objectives of
 the measurement  program, as  well as  documented precision  and bias,  when
 comparing data  collected by  different  surveys  or  programs.  For example,
although both  the SRRS  and  the NRRS  have measured radon concentrations in U.S.
homes, their results are not directly comparable  because  they had different
objectives, target  populations, and measurement methods.

-------
                                 5.0 QA AUDITS


     A QA audit is a formal review of facilities, procedures, and documen-
 tation.   Ideally, the audit is performed by personnel organizationally
 independent of the project management to minimize the risk of any conflict of
 interest.  Because the QAPjP is the basis of a QA audit, it is important that
 the plan describe all important aspects of data collection and processing and
 that it be accurate and current.  Three types of audits were performed by RTI
 during these surveys: technical systems audits (TSAs), performance evaluation
 audits (PEAs), and audits of data quality (ADQs).  After each audit, a written
 report describing the audit objectives, procedures,  and findings was prepared.

     A TSA is a qualitative on-site evaluation of a measurement system.  Its
 objective is to assess and document all facilities,  equipment, recordkeeping,
 data validation,  calibration procedures,  and QC procedures.  Any undocumented
 or unauthorized deviations from the QAPjP should be noted in the audit and
 included in the audit report.  Although a TSA does not provide quantitative
 information,  the information collected during these audits helps the auditor
 quickly determine whether or not the quality of data is likely to be adequate
 for its intended use.

     RTI conducted five TSAs for these surveys.   Two  commercial ATD processing
 facilities  were audited,  as well as EPA's primary canister processing faci-
 lity.   In addition,  RTI's data  processing and data management  for the NRRS and
 the NSRS were audited. Written reports for  each audit were prepared and the
 majority of the auditor's recommendations were implemented by  project
 management.

     A variety of  problems or potential  problems were  identified by these TSAs.
 For example,  calibration  control charts for  canister  counting  stations could
 not be  readily produced,  although this  had been  recommended in a  previous
 audit.  When  these charts were  finally  produced, they indicated a problem with
 one of  the  counting  stations.   Also,  ATD  processing laboratories  should rou-
 tinely  calibrate  at  exposures > 4  pCi-y/L, since this is a  critical  value in
 mitigation decisions.  Calibrations at high  exposures also  help to separate
 the effects of  the background correction  from the response  of  the sensitive
 material  in the detector.    For a  large project, data processing  and
 management system should  be  developed before data collection begins.   This
 allows data editing  to begin as soon  as the  first data is available,  thereby
 providing feedback to  correct some problem patterns before all  the data have
 been collected.

    A PEA is a quantitative evaluation of a measurement system.  This usually
 involves analysis of prepared samples or spiked detectors.  Blank and collo-
 cated (duplicate) samples/detectors may also be used.  As described above, RTI
 evaluated the performance of all the radon detectors used in these surveys
using blank, spiked,  and collocated detectors.  In most cases,  these QC detec-
 tors were mixed with field-exposed detectors and processed 'blindly by the
 laboratories.  These results were tabulated and compared with the goals for
DQIs specified in the QAPjPs.   Formal PEA reports were prepared for each

-------
 survey.  In some cases, the laboratories identified and corrected problems
 with calibration or processing.  In other cases, the values of the field-
 exposed detectors reported by the laboratories were adjusted based on analyses
 of the spiked and blank detectors.

     An ADQ involves the assessment of the methods used to collect,  interpret,
 and report the information required to characterize data quality.  RTI
 performed an ADQ for each of these surveys.  These audits were overall reviews
 of the survey implementation,  detector performance, and data processing.  They
 involved detailed reviews of the recording and transfer of raw data,  data
 calculations,  documentation of procedures,  and the selection and discussion of
 appropriate DQIs.  Written reports were prepared for each audit.


                                 6.0 RESULTS

 6.1 MEASUREMENT ERROR

     Measurement error was determined from the spiked,  blank,  and collocated
 detectors described in section 4.0.   Table  1 shows the results of the spiked
 ATDs from the  NRRS,  the NSRS,  and  year 4 of the SRRS.   Recovery is  the average
 reported exposure divided by the monitored  exposure +_ the standard  deviation
 of the  reported exposures divided  by the monitored exposure.   For each of
 these surveys,  the DQI goals for ATDs  were  for recovery to be between 0.85  and
 1.15 (i.e.,  |bias|  < 15%)  and  the  MARE < 0.25.   Note that the NSRS  ATDs with
 monitored exposures  > 0.9 pCi-y/L  had  recoveries around 0.80  +_ 0.15.   A linear
 regression of  the monitored exposures  on the reported exposures gave  a formula
 to adjust the  reported exposures of  the field-exposed detectors to  correct  for
 this bias:

     Spikes:            Monitored  =  (1.30 ± 0.04)*Reported -  (29.4  ±  8.6).
     Field-exposures:   Adjusted   =  (1.30)*Reported - 29.4

 Figure  1  is a plot of  the reported recovery  versus  the monitored exposure;
 Figure  2  shows  the adjusted recovery versus  the  monitored exposure.

     Although Table 1 shows  the effect  of a similar  adjustment  on  ATDs  from the
 SRRS, the  reported values were within  the QA  objectives specified in  the  QAPjP
 and  the values  reported to  participants were  not  adjusted.

    Table 2 shows the  results of the spiked charcoal canisters  from the NSRS
 and year 5 of the SRRS.  The goals for DQIs of spiked  charcoal  canisters were
 the  same as for ATDs:   recovery was to be between 0.85  and  1.15 and the MARE <
 0.25.   These diffusion barrier canisters were simultaneously  exposed  for one
week in EPA radon chambers.  The NSRS  canisters were shipped via  overnight
express to the processing laboratories.  The  SRRS canisters were  shipped via
overnight express to the state survey  coordinators, who then sent the  canis-
ters to the processing  laboratory via  U.S. Mail.  This  table includes  only
those SRRS canisters that were processed within 7 days  of exposure.  The NSRS
canisters were counted  the next day after exposure  and  have better precision.

-------
             TABLE  1.  RECOVERY OF  SPIKED ATDs
Monitored
Exposure
(pCi-y/L)
4
10
1.
2.
4.
8.
1.
2.
4.
8.
0.
0.
1.
4.
0.
0.
1.
4.


2
1
3
5
2
1
3
5
4
9
7
1
4
9
7
1
Survey
NRRS
NRRS
SRRS4
SRRS4
SRRS4
SRRS4
SRRS4*
SRRS4*
SRRS4*
SRRS4*
NSRS
NSRS
NSRS
NSRS
NSRS*
NSRS*
NSRS*
NSRS*
N
302
286
68
68
68
68
68
68
68
68
36
36
36
36
36
36
36
36
Recovery
Mean+Std
0
0
1
0
0
0
1
0
0
0
0
0
0
0
1
0
0
1
.974
.925
.08
.98
.95
.93
.04
.99
.99
.97
.95
.77
.79
.81
.05
.91
.99
.04
+
±
_f
^+
•»•
±
+
+^
•f
±
+
+
^
jf
±
+_
_+
+_
.Dev.
0
0
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
.093
.070
13
14
07
07
14
15
08
07
32
17
14
12
40
21
17
14
MARE
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
.074
.084
.12
.08
.07
.08
.12
.09
.06
.07
.27
.25
.21
.19
.31
.19
.13
.12
* The reported values were adjusted based on a linear
regression of the monitored on the reported values.
N = number of detectors

-------
I
4J
Vl
a
05
1
u
                                                    new
  Ficfure 1.
          Monitored Exposure (pCi-y/L)

Reported recovery of  spiked ATDs  in the NSRS.
 -O

 o
 4->
 •H
 c
  a
 f->
 a
                                                 HI    new
                                                 O    old
                        Monitored Exposure (pCi-y/L)

 Figure 2.  Adjusted recovery  of spiked ATDs in the NSRS.

-------
      TABLE 2.  RECOVERY OF SPIKED DIFFUSION BARRIER CHARCOAL CANISTERS
Monitored


Concentration
(pCi/L)
2.7
3.9
4.1
6.2
8.6
2.7
3.9
4.1
6.2
8.6
2.7
3.9
4.1
6.2
8.6
Survey
SRRS5
SRRS5
SRRS5
SRRS5
SRRS5
NSRS
NSRS
NSRS
NSRS
NSRS
NSRS
NSRS
NSRS
NSRS
NSRS
N
10
24
5
19
20
16
16
16
16
16
16
16
16
16
16

Recovery
Mean+Std.Dev.
0.94 + 0.08
1.02 i 0.11
0.91 + 0.07
0.87 i 0.06
0.94 +. 0.09
0.96 + 0.06
1.10 ± 0.07
1.02 i 0.06
1.02 + 0.05
0.96 + 0.04
0.87 i 0.11
1.09 + 0.09
0.96 i 0.09
0.98 i 0.07
0.97 + 0.07


MARE
0.09
0.08
0.09
0.13
0.09
0.06
0.10
0.06
0.05
0.05
0.15
0.11
0.08
0.06
0.06

Processing
Laboratory
NAREL




NAREL




LVF




      N = number of detectors
    Each survey mixed some blank  (unexposed) detectors in with the field-
exposed detectors.  In the NRRS,  two blank ATDs were included in each etch
batch.  RTI asked the processing  laboratory to report their calculated
exposure for each detector, no matter how low, although their normal procedure
was not to report a specific value when the calculated exposure was < 30 pCi-
d/L (0.082 pCi-y/L).  The NRRS QAPjP specified that all blank ATDs should have
reported exposures Ł 0.16 pCi-y/L.  The average reported exposure of the 158
processing blanks was 0.11 +_ .13  pCi-y/L; 59% had reported exposures > 0.082
pCi-y/L and 23% were > 0.16 pCi-y/L.  This suggests that the laboratory may
have underestimated the background correction of the NRRS ATDs.  The effect on
12-month exposures was not considered significant, however.  RTI also distri-
buted blanks to field supervisors around the U.S. to evaluate the potential
for exposure of detectors during  shipping.  Eight field blanks were mailed to
each of the 11 field supervisors  in September 1990, shortly after most of the
field-exposed detectors had been  retrieved.  The supervisors were instructed
to open the foil bags,  apply the  adhesive foil seals, and return these
detectors immediately to RTI.  There was no significant difference in the
reported values for any location; their average reported exposure was 0.15 +,
.12 pCi-y/L.

    In the NSRS,  two blank detectors were sent to each school with instruc-
tions that they were to be stored, unopened, while the other detectors were
being exposed.   After the other detectors had been retrieved, the blank

-------
detectors were opened, immediately sealed, and packaged with the field-exposed
detectors for shipment to the processing  laboratory.  The NSRS QAPjP specified
that all blank ATDs should have reported  values  <  45 pCi-d/L  (0.12 pCi-y/L)
and all blank canisters should have reported activities < 0.5 pCi/L.  The
average adjusted exposure of the 188  field blank ATDs was 0.06 i .10 pCi-y/L;
4.3% were > 0.12 pCi-y/L.  Only 0.6%  of the 1485 charcoal canisters had
reported activities > 0.5 pCi/L; their average activity was -0.05 +. .18 pCi/L.

    There were no true field blanks in the SRRS; that is, all detectors sent
to houses were exposed.  Some unexposed ATDs, however, were mixed with spiked
ATDs and field-exposed ATDs when they were sent  to the processing laboratory.
Also, 2% of the canisters sent to the states for distribution to participants
were mailed unexposed to the EPA processing laboratory.  In year 4 of the
SRRS, 40% of the 68 blank ATDs had reported exposures > 0.16 pCi-y/L, but only
9% of the adjusted exposures were this large.  During year 5 of the SRRS, all
of the 166 blank canisters had reported activities < 0.5 pCi/L, which was the
goal specified in the QAPjP.  The average activity was -0.12 ± .27 pCi/L.

    Precision of the radon detectors  was  assessed  using collocated, field-
exposed detectors.  The mean value (M^) and difference  (D^) of each collocated
pair (A^, B^) were determined:
The mean value was used to group pairs into exposure  ranges.  The pooled mean
(Mp) ,  standard deviation  (SDp) , and coefficient of variation  (CVp) were
calculated for each range:

    Mp = (!, M^/N

    SDp={2; Di2/(2N)]1/2

    CVp= SDp/Mp= t(2N)I(Ai-Bi)2]1/2/[Z(Ai-(-Bi)]


The pooled standard deviation  can be used to calculate confidence intervals.

    Table 3 shows the precision of the ATDs used in the NRRS, the NSRS, and
year 4 of the SRRS.  The  goal  in each survey was for  the CV of pairs exposed
near 4 pCi-y/L to be < 0.20.   Since the NSRS detectors were exposed for only 5
months,  there are few pairs  in the higher exposure ranges.  Their precision,
however, did not satisfy  the goal specified in the NSRS QAPjP.  A review of
individual pairs with poor precision showed that their precision varied by
lot.  Two lots of ATDs were  used for the NSRS:  one lot of 5000 was purchased
in the fall of 1990, and  another lot of 500 ATDs, left over from the NRRS, was
used to supplement these  detectors .  The NRRS detectors were  the same model as
the ATDs purchased for the NSRS, but were purchased in the spring of 1989.
These detectors were distributed to RTI field interviewers around the U.S.
during the summer of 1989, but were not placed in homes.  The field
interviewers returned the unused ATDs to RTI where they were  stored in a low
(but not zero) radon environment.  It is possible that some of the radon-proof
foil bags received minor  damage during shipment in the summer of 1989 and that

-------
 these  detectors  accumulated some exposure during  storage  for over two years.
 When the precision of the NSRS ATDs was calculated by  lot, the performance of
 the 'new" ATDs was much better than that of the 'old*  ones.  Less than 10% of
 the field-exposed ATDs were left over from the NRRS, but  42% of the spiked
 ATDs were "old."  The linear regression of monitored on reported values for
 the spiked ATDs  was repeated for each lot and the lot-specific regression
 coefficients were used to adjust the values reported to participating schools.
                   TABLE 3.  PRECISION OF COLLOCATED ATDs
Average Exposure
(pCi-v/L)
0.00
0.25
0.75
1.50
3.00
0.00
0.25
0.75
1.50
0.00
0.25
0.75
1.50
- 0.25
- 0.75
- 1.50
- 3.00
- 6.00
- 0.25
- 0.75
- 1.50
- 3.00
- 0.25
- 0.75
- 1.50
- 3.00
N
44
47
32
28
17
139
78
68
30
113
60
20
3
Mp ± SD
(pCi-y/L)
0.11 i 0.08
0.46 _+ 0.19
1.17 i 0.21
2.06 + 0.13
4.22 i 0.26
0.08 + 0.08
0.42 ± 0.14
0.98 HK 0.35
2.22 ± 0.49
0.07 i 0.07
0.45 ± 0.10
1.00 ± 0.18
2.01 ± 0.22
CV- Survey
0.754 NRRS
0.336
0.177
0.063
0.060
0.96 NSRS All*
0.33
0.36
0.22
0.96 NSRS New*
0.22
0.18
0.11
         N = number of pairs
         *  Includes some ATDs left over from the NRRS
         ** Includes only ATDs purchased for the NSRS
    Table 4 shows the precision of the diffusion barrier charcoal canisters
exposed for one week in the NSRS and year 5 of the SRRS.  The last row shows
the precision for canisters exposed near 4 pCi/L and processed the next day
after exposure.  The goal in each of these surveys was for canisters exposed
at or above 4 pCi/L to have a CV < 0.10.

    As described in section 4.0, the goal was to make 12-month measurements in
at least 5000 homes.  RTI estimated the fractions of homes contacted that
would be eligible and willing to participate in the NRRS, and that would
actually return detectors.  RTI also estimated the fraction of those detectors
that would be usable (i.e., undamaged with good placement information).  Based
on those estimated fractions,  RTI determined that 11,400 homes should be
contacted.  Table 5 shows the completeness goals and actual results for the
NRRS.  Only 74% of the housing units contacted were eligible for the survey
(continuously occupied for at least 9 months during the year with no plans to

-------
            TABLE  4.   PRECISION OF COLLOCATED CHARCOAL  CANISTERS
Average Cone .
(pCi/L)
0.00
0.25
0.75
1.50
3.00
0.20
0.25
0.75
1.50
3.00
6.00
3.50
- 0.25
- 0.75
- 1.50
- 3.00
- 6.00
- 0.25
- 0.75
- 1.50
- 3.00
- 6.00
-
- 4.50
N
9
14
12
8
1
107
665
343
239
64
24
30
Mp ± SDp
(PCi/L)
0.11 ± 0.13
0.50 i 0.09
1.00 +. 0.32
2.13 + 0.65
3.43 + 0.42
0.23 + 0.16
0.47 i 0.15
1.07 +. 0.21
2.09 i 0.37
4.12 + 0.33
9.77 ± 0.46
3.90 i 0.35

ŁVp_

0.19
0.32
0.31
0.12
-. ___
0.32
0.20
0.18
0.08
0.05
0.09

Survey
SRRS5




NSRS





NSRS
            N = number of pairs
move within 12 months) and about 10% of these did not remain eligible
throughout the 12-month monitoring period.  Hence, only  68% of the sample
housing units were eligible for the NRRS.  More of these units than
anticipated were willing to participate in the survey and returned usable
detectors, however, so the final results included data from 5700 homes.  With
this higher response rate, the potential for bias due to missing data was
substantially reduced.  Recontacting the participants every 3 months (panel
maintenance) was an important factor in the high response rate.
                     TABLE 5.  COMPLETENESS OF THE NRRS
Fraction
Eligible
Willing
Returning detectors
Usable detectors
Overall
Estimate
0.79
0.80
0.69
0.98
0.43
Actual
0.74
0.84
0.80
0.999
0.49
    The SRRS QAPjP specified a cumulative usable rate of at least 40% for
charcoal canisters.  Eligibility requirements for the SRRS were stricter than
for the NRRS, because the house had to be owner-occupied with a permanent
foundations and ground contact.  The estimates listed in Table 6 below were
based on experience during years 2 and 3 of the SRRS.

-------
                TABLE 6.   COMPLETENESS OF YEAR 4 OF THE SRRS

                Fraction	   Estimate   Actual
                Eligible                    0.63      0.59
                Willing                     0.88      0.90
                Returning usable detectors  0.82      0.78
                Overall                     0.45      0.41
    The NSRS Q.APJP specified three completeness objectives: 1)  that 80% of the
schools contacted would agree to participate and would actually place
detectors, 2) that 80% of the canisters placed in schools would be returned
and would produce usable readings, and 3)  that 70% of the ATDs placed in
schools would be returned and would produce usable readings.  A higher
attrition rate was anticipated for the ATDs because they were placed for
several months rather than one week.  Actual attrition during placement was
much less than anticipated for both types  of detectors {see Table 7).


                     TABLE 7.  COMPLETENESS OF THE NSRS
Canisters
Fraction
Goal
Eligible and
Returning
Overall
willing
usable


detectors

0
0
0
.80
.80
.64
Actual
0
0
0
.78
.91
.71
ATDs
Goal
0
0
0
.80
.70
.56
Actual
Not
yet
avail
6.2 SAMPLING ERROR

    For each of these surveys,  sampling weights that accounted for the unequal
probabilities of selection were calculated so that unbiased population
estimates could be generated.  Because 100% response rates were not obtained
in any of the surveys, the sampling weights of the participants were increased
to account for sample cases that failed to yield usable information; the
resulting analysis weights were used in analyzing the survey data.

    Weighted data were used in the analysis for each of these radon surveys.
Sampling errors of survey estimates of means, proportions, and totals were
computed using SUDAAN, a computer software program developed by RTI for
analyzing survey data with complex error structures.  Thus, the estimated
sampling error for each of the survey estimates took into consideration the
full complexity of the sample design.

    This procedure implicitly included the random measurement errors
(measurement precision) in the calculation of the sampling errors.  The SUDAAN
software, however, does not routinely incorporate systematic errors  (such as
measurement bias) in estimating these errors.  Statistical techniques do exist

-------
to adjust the parameter estimates for measurement bias,  but they greatly
increase the complexity of the analysis.  The QA data allowed us to estimate
the magnitude of the measurement bias and judge whether the bias was
sufficiently large to warrant efforts for more complex analysis.   In the
NRRS,  the measurement bias was small enough that the more complex analysis was
not necessary.  In the SRRS and the NSRS, some of the ATD values reported by
the processing laboratory were adjusted to reduce the bias using the
regression procedure described in section 6.1.

6.3 UNCERTAINTY IN THE FREQUENCY DISTRIBUTIONS

    Estimating the frequency distributions is a special case of estimating the
proportions alluded to in section 6.2.  The frequency distribution was
estimated by grouping the radon measurements into intervals and estimating the
proportion (or percent) of housing units with radon measurements in each
interval.  The standard error for the percent in each interval was estimated
using our SUDAAN software.

    In this special case, the measurement precision may bias the estimates of
the percent in each interval.  Radon measurements near the interval boundaries
may be assigned to the incorrect interval because of either systematic or
random measurement error.  Bias arises when misclassifications due to
measurement errors do not "balance out."  Because the QA procedures controlled
both the measurement bias and precision to satisfactorily low levels, the
effect of measurement error on the uncertainty in the frequency distributions
was judged not to be significant.

    The sample design, its implementation, and the QA procedures were
instrumental in achieving the primary objective of the NRRS: an accurate
estimate of the frequency distribution.  The relative standard error of the
fraction of homes with radon levels greater than 10 pCi/L was 28%, well below
the goal of 50% established during the design of the NRRS.
                                 7.0 SUMMARY


    Quality assurance activities were  incorporated in all phases of three EPA-
sponsored radon surveys.   Project planning  included preparation of
comprehensive QA project plans based on  the specific objectives of each
survey.  These project plans  included  goals for data quality and specified
operating procedures capable  of meeting  those  goals.  QA audits were  conducted
during each survey to verify  that these  procedures were followed.  The
performance of the radon detectors  was monitored with spiked, blank,  and
collocated detectors.  The QC detectors  were processed  "blindly' along with
the field-exposed detectors.  This  helped us to identify and resolve  some
calibration and processing problems.   In addition to detector performance, the
QA/QC activities documented the procedures  for collecting and processing
questionnaire data.

-------
     The precision of  ATDs  exposed to at least 3 but not over 6 pCi/L for 12
 months in the NRRS was  about  6%.   The precision of the diffusion barrier
 charcoal canisters exposed to at  least 3 but not over 6 pCi/L for one week in
 the NSRS was about 8%.   The precision of charcoal canisters with similar
 exposures in the SRRS was  slightly worse (about 12%),  probably because they
 were returned via U.S.  Mail rather than by overnight express.


                                8.0 REFERENCES


 1.   Yeager,  William M.,  et  al.  State Radon Assessment  Program:  Quality
     Assurance Project Plan.   RTI/4658/16F-03F,  Research Triangle Institute,
     Research Triangle Park, NC, December 12,  1990.

 2.   Daum, Keith  A., et  al.  National  Residential Radon  Survey: Final  Quality
     Assurance Project Plan.   RTI/4240/10-01F, Research Triangle Institute,
     Research Triangle Park, NC, September 6,  1989.

 3.   Yeager,  William M.,  et  al.  National  School  Radon Survey:  Quality Assurance
     Project  Plan.   RTI/4658/24F-01,  Research  Triangle  Institute,  Research
     Triangle Park,  NC, January  30,  1991.

 4.   Office of Radiation  Programs,  U.S.  Environmental Protection Agency.
     Indoor Radon and Radon  Decay  Product  Measurement Protocols.   EPA 520-1/89-
     009, U.S.  Environmental Protection Agency,  Washington,  DC,  March 1989.

 5.   Quality  Assurance Management  Staff, U.S.  Environmental  Protection Agency-
     Interim  Guidelines and  Specifications for Preparing  Quality Assurance
     Project  Plans.  QAMS/005-80,  U.S. Environmental Protection  Agency,
    Washington, DC, 1980.

 6.  Gearo, J.R.,  et al. The Growth of the National Radon Measurement
     Proficiency  (RMP)  Program.  In:  Proceedings of the 1988 Symposium on Radon
    and Radon Reduction Technology, Vol.  2, EPA-600/9-89-006b,  U.S.
    Environmental Protection Agency. Research Triangle Institute, Research
    Triangle  Park, NC, March 1989.

7.  National Council on Radiation Protection and Measurements.  A Handbook of
    Radioactivity Measurements Procedures.  NCRP Report 58, 1985.

-------
                                                              IXP-5
   RADON IN HOUSES AROUND THE PLOMIN COAL FIRED POWER PLANT

     Lokobauer N.. Franie Z., Bauman A. and Horvat D.
     Institute  for  Medical  Research  and Occupational  Health.
     University of Zagreb
     Zagreb,  Croatia

                           ABSTRACT

     The paper presents the  results of investigations  of  radon
activity concentrations in the old houses,  assumed  to  be  built
from materials containing slag and ash of coal combusted in  the
Plomin Coal Fired Power Plant. Measurements were performed using
solid state nuclear track detectors (Kodak  Pathe  LR-115 films,
type III mounted inside open and filter cups. Detectors were
installed during  three months at  15 locations in  30 houses.  In
order to  determine radon  emitting  source  (building material,
soil,  radioactive  airborne  particulates)   as   precisely   as
possible, films were also installed in certain number of houses
built more recently.  The radon activity  concentrations inside
houses  varied between  370  Bqm~3 (old  houses)   and  16  Bqm~3
(recently built houses).

-------
                          INTRODUCTION

     Radon emitted from building materials which usually contain
more  uranium/radium than soil,  may cause a  high  indoor radon
concentration,  particularly  in winter,  when  there  is a greater
tendency  to  seal houses  to  conserve heat.  In the  area  of the
impact  of the Plomin Coal Fired Power  Plant  (CFPP),  it  was of
special importance to investigate the concentration of radon, and
possible  health risk to the  population in hoxises assumed to be
built from ash  and slag.

     Plomin CFPP.  located on the Istrian Peninsula in Croatia,
uses coal containing elevated uranium series concentrations. The
impact of the Plomin  CFPP on the environment has been studied for
years  by  the  Department for  Radiation Protection  of  this
Institute.  In  the  eighties  uranium   activity  concentrations
measured  in the coal was  10 to  15 times higher than the average
world values.

     Ash and slag remaining after coal combustion contain uranium
and  its decay products  which cause redistribution  of natural
radioactivity  from the soil   to the locations at which  it may
produce a significantly  impact  to the environment and health of
the  inhabitants.  In the  vicinity  of Plomin  CFPP the  need for
radon measurement was noted since ash and slag were used in the
past as building materials for  the  construction of houses.
Another possible source  of radon emission for local population
could be cement produced in a  highly developed industry based on
regional mineral resources.

                      MATERIAL AND  METHODS

     Radon activity concentrations were measured by Kodak Pathe
LR-115 films, type II, mounted inside open and filter cups. The
detectors were  exposed during  three  months   (December 1991  to
February 1992) at thirty houses. The track densities were counted
under a microscope,   after  etching  the films under  precisely
controlled conditions (2.5 N  NaOH,  60 °C, 90 min). Bare detector
was  calibrated  in  National  Radiological  Protection  Board,
Chilton, Didicot, and detector sensitivity coefficient was 10.3
BQ m~3 d per tracks cmr2.

-------
                    RESULTS AND DISCUSSION

     Detectors for radon activity measurement were installed in
old and recently built houses on selected  locations in the radius
of  60  km around  the  Plomin  CFPP.  The  values  of  mean  radon
activity concentrations  measured  in winter  in   thirty  houses
around the plant are given in Table 1.
TABLE 1.   THE  MEAN RADON ACTIVITY  CONCENTRATIONS* IN  HOUSES
AROUND THE PLOMIN COAL FIRED POWER PLANT (DEC.  1991 -FEB.  1992)

                 Activity concentration  (Bqm~3)
Location      Old houses         More recently built houses
1
2
3
4
5
6
7
8
9
10
11
12
13
14*
15*
209
270
77
90
160
105
122
102
370
152
179
107
132
118
61
+
+
+
+
±
±
+
±
±
±
±
+
±
±
+
14
16
9
9
13
10
11
10
19
12
13
10
11
11
8
35
23
16
48
60
66
66
64
43
25
79
39
48
16
22
+
+
±
±
±
+
±
±
±
±
±
+
+
+
+
6
5
4
7
8
8
8
8
7
5
9
6
7
4
5
     * Activities reported as ± one sigma error.
     * Location of cement industry


     The data show that radon activity concentrations were higher
in  the  old  houses  which  points to  conclusion  that  recently
constructed  houses  were built  from material  containing  lower
uranium  series radionuclide  concentrations.  On  basis of  the
obtained values  it  can be noted that  impact of Plomin CFPP as
well as nearby cement  industry  could be neglected with respect
to  radon  concentrations in  the  air.  Building  material  was
therefore major radon emitting  source  in old houses.

-------
     In 12 old houses radon activity  exceeded  the  value  of 100
Bq m~3 and in 3 old houses radon  activity  was  even higher than
200  Bq m~3.   According  to  recommendations  of  International
Commission on Radiological  Protection this is  the  level which
requires protective measures (2).

     Investigation of track densities (tracks cm-2) of the bare
and filtered detectors installed  in  selected houses around the
plant  gave  equilibrium  factor F  of radon  and its  daughters
ranging from 0.2 to 0.8.

                          CONCLUSION

     Investigations carried out in the area of the impact of the
Plomin CFPP including old houses built mostly from ash and slag
of the  coal  combusted in  the  plant  showed  that  indoor  radon
concentrations may  cause risk to human health.  In order  to
protect people inhabiting this area  it should  be  necessary  to
continue  the  investigations   and assess  the  effective  dose
equivalent, to the population.

               The work described  in  this  pap>er was not  funded
               by the U.S. Environmental Protection Agency and
               therefore  the contents  do not necessarily reflect
               the  views  of   the  Agency  and   no   official
               endorsement should be inferred.

                          REFERENCES

1.   Marovic G.  Enhanced Natural  Radioactivity around  a  Coal
     Fired Power  Plant.  M.Sc.  Thesis.  Technological  Faculty,
     University of Zagreb,  1985. (In Croatian)

2.   International  Commission  on  Radiological   Protection,
     Publication 39,  Pergamon Press 1984.

-------
                                                                    IXP-6
         A  RADON  SURVEY AT SOME RADIOACTIVE SITES IN INDIA
            By:  Jaspal  Singh,  Lakhwant  Singh,  Surinder  Singh,
                 and  H.  S.  Virk
                 Department of  Physics
                 Guru Nanak Dev University
                 Amritsar-143005,  India
                             ABSTRACT
     From studies carried out all over the world, there is
scientific agreement between the incidence of excess lung cancer
among underground miners and exposure to radon and radon
daughters.   The epidemiological studies also support the fact that
there are increased incidences of bronchial carcinomas among
individuals exposed to relatively high levels of radon and its
daughter products.  Increased incidences of lung cancer have also
been found in homes even at moderate values of radon activity.
The current study has been carried out for a complete year at two
sites in Himachal Pradesh, India.  In one of the areas under
study,  the average radon activity has been found to be 26.41 pCi/L
in summer and 31.73 pCi/L in winter, giving an annual exposure of
47.09 mSv.   In the second area, these levels were found to be
16.54 pCi/L and 18.52 pCi/L, respectively, which gives an annual
exposure of 28.39 mSv.  The study is being conducted for the
second year.   The complete data collected for the two years will
be presented in the paper.

-------
                                                                                       IXP-7
           Islandwide Survey of Radon And Gamma Radiation
                         Levels in Taiwanese  Homes
            Ching-Jiang Chen, Chi-Wah Tung and Yu-Ming  Lin

          Taiwan  Radiation Monitoring Center, Atomic Energy Council
              823 Cherng-Ching Road, Kaohsiung, Taiwan 833, ROC
                                    ABSTRACT


     An islandwide survey of Taiwanese homes was conducted to determine the average annual effective
dose equivalents to Taiwan population from exposure to radon and i radiation. The radon concentration
was measured using cellulose nitrate films (CN film) as the solid state nuclear track detectors, while  7
radiation dose was measured using thermoluminescent dosimeter (TLD). The CN films were put in a
plastic cup with a filter on the top and mailed to approximately 250 randomly distributed householders and
replaced for every two months.  The TLD were sent to 42 selected homes for 3 months exposure.  The
                                                                -3
average Rn concentration in Taiwanese homes measured over 1.5 year is 10 Bqm   while the average 7
dose rate is 0.12 u Gy h  .   Using appropriate conversion  factors,  the annual  average effective  dose
equivalents to  the Taiwan population  were  determined to be 0.5 mSv and 0.59 mSv for Rn and 7
radiation exposure respectively.

-------
1.  INTRODUCTION

    There are  many papers deal with radon recently.   According to the  UNSCEAR report (1) the annual per
caput  dose from natural radiation in the world is 2.4 mSv.  Fourty-six percent of it comes from radon (  Rn ->
"> 14
  Po) and 33% of it comes from external gamma radiation which include cosmic rays.  In a lifetime of human
beings about 80% were spent indoors, so indoor radon and  indoor gamma dose are the most important items in
assessing population doses.  From the UNSCEAR 1988 report, there are 27 countries that performed large scale
indoor radon survey including 96,469 houses.    Most  of the  data show a log-normal distribution with medians
ranged from 10-140 Bqm .  Most of the data came from  developed countries  located in  the middle and high
latitude of northern hemisphere.   However  these data cannot stand  for the world  average.  Some of the newly
published data show somewhat different from the UNSCEAR report.   New large  scale and nationwide survey and
reassessment are still undertaken (2,3,4,5).
    Tai%van island is  located in lower latitudes of northern hemisphere (22'  -25 °  N), along the  west coast of
Pacific ocean, which is an area with oceanic subtropical  climate.   Both  temperature and  humidity are high in
Taiwan.   Geologically,  sedimentary  rocks formed  of shale and  sandstone dominate  in the whole island.  Its
natural radioactivity is  quite normal as shown in table 1.   More than 90% of the  houses in Taiwan are build with
concrete  and bricks.    Most of the  building materials  are made of domestically produced river sand, cement,
gravels and bricks made of clay.  Their floors are made of concrete, tile, sheet vinyl, marvel, ground aggregate
and granite.  Because the temperature is warm and tlie windows are large and opened frequently, the ventilation
 rate is normally high in Taiwan.  Owing to the  promotion of living standards, the air-conditioning facilities are
 popular and there are  many skyscrapers with central air-conditioning systems  in city areas.  Islandwido survey of
 indoor radon level has been an urgent project for  Taiwan Radiaiton Monitoring Center.

 2. EXPERIMENTAL DESIGN
 Sampling
     One thousand householders with profession as science teachers in junior high schools were randomly selected
 and invited to  participate in the survey, and 250  householders responded positively.  A questionnaire including 9
 questions about the basic information of each house was sent to  each participants.  The format of questionnaire
 is shown in Table 2.   Part  of the participants terminated  the participation  during the  survey.   About  212
 householders participate in this study from the beginning to the end.
 Indoor radon measurement
       Two cellulose  nitrate (CN) films in a plastic cup was mailed to each participant  together with a letter
 describing how to install  the measuring device.   In  order to avoid the pollution  of aerosol  in the air, a whatman
 GFC glass  fiber filter was attached  to the cap of the cup.  Thirty  small  holes were drilled on  the cap to let the
 air diffuse through the filter. Figure 1 shows the  structure of the measuring  cup.    Two pieces of CN film cut  into
 30 x  45 mm were adhered on  the inner wall of the cup.
 Indoor 7 dose measurement
      The pen-type TLDs  of Panasonic UD-200S CaSO4:Tm were \ised to  measure  the  cumulate  indoor t  dose.
 Fourty two houses were measured  for three months  exposure.   The TLDs were put  in  the bottom of radon
 measurement cup. All the dosimeters were mailed through post office.

-------
building material and structure of the building may have positive effect on the radon level.
     Figure 6 shows indoor radon level.  According to the measured result, it is found  that the indoor Rn level
                                                              -3                      -3
follows log-normal distribution.   There are 30% between 3-6 Bqm , 28% between 6-9 Bqm and 20% between  9-
       -3                                          -3                                           -3
12 Bqm .  Hence 79% houses are between 3-12 Bqm .  As a whole, the geometry mean is 8.5 Bqm    and  the
geometry standard deviation is 0.59.
Seasonal variation
      Figure 7  shows the  seasonal variation of indoor radon.   Data were collected and analyzed  in  every two
months.  The mean concentration fluctuates between 8.4-11.1 Bqm   with no  apparent  seasonal  variation.    In
medium and high latitude  countries, winter is  cold and window is often closed.   Radon concentration becomes
higher due to poor ventilation.   In the summer, radon concentration becomes lower in good ventilation rate.   In
Taiwan, the seasonal variation of indoor Rn is not significant.
Relationship between indoor radon concentration and gamma dose rate.
                                                                                                -i
     Results of TLD measurement shows that indoor gamma dose rate are between 0.066-0.189  u. Gyh   and the
average is 0.121 u Gyh  (including contribution of cosmic  rays).   The result is  high  as compared with other
countries.   It  is because  that more concrete  and brick is used as the  building materials, which contains more
natural radionuclides.   Figure 8 shows the correlation between  indoor gamma  dose rate and indoor radon activity.
The regression curve based on 42 pairs  of data is y = 0.000562X+0.113032, and the correlation coefficient,  i  is
equal to 0.3.  The  result shows that the indoor radon concentration is  in  correlation with gamma dose rate  at
0.95 confidence  level.
Correlation of indoor radon and building factors
      In order  to  explore the  factors  that  might contribute  significally to  the  level  of Rn concentrations  in
Taiwanese houses, the following information was sought  through the questionnaires^!) building  style; (2) main
building materials;  (3) floor materials; (4) wall materials;  (5)  in which year the house  was built.   Most of the
householders had forgotten the year of construction, so this factor could not be analyzed.  The other factors were
analyzed as shown in Table 4.
       The first row in Table 4 listed 4 types of house including the  number of house and  the average radon
concentrations for each type of house.   There  are no correlation between radon concentration and type of house.
Average radon  concentration of each type of house is very close to the gross average.
     The second row in Table  4 shows 4 types of main  building materials. There are  no wooden  house in this
study.  Wooden house is hardly seen in  Taiwan  nowadays.   The others  include 4 houses with an average of 19.4
Bqm  which is  double that of gross average.  These houses are the old style and were built with local  soil and
bricks.   Their  flooring were just natural soil that  might  possess higher radon exhalation  rate.   This is  an
interesting finding.
     The third  row shows the effect of 7 types of floorings.  The  items  of wooden floor and others show higher
radon  concentration than the other 5  types  of flooring.   The fourth row shows the effect of wall surface
materials.There  are  no much difference between the  five types of wall materials.
Kinman district
     Kinman isle is close  to the seashore of Mainland China which is famous for its miltary factility.   There are
granite  rocks  on  this isle.  Natural radioactivity  in  granite  is 3-4 times higher than Taiwan  rocks.  RaJon

-------
Outdoor radon measurement
    In order to measure the outdoor radon level, 145 participants was selected all over Taiwan.  The measuring
device is the same as indoor and the cup was put in balcony or under eaves to get rid of rainfalls and sunshine. In
order to enhance  the efficiency of CN films, the exposure period was extended to 3 months.

3. Experimental
Measurement of radon
     Measuring methods of radon in the air can be  catagorized  into active and passive methods.   For long term
observation of radon concentration in air,  the passive  alpha track method was selected in  this survey.  CN film
method is a popular method in large-scale  indoor  radon  survey.  The CN film used is the  LR-115  Type I  ,
strippable which is a product of Kodak Pathe, France .  The optimized etching condition  is 60 °C   for 70 min.
under 2.5N (10% by weight) Na(OH) solution.  After chemical etching, the  films were rinsed in clean water for
10  min.   and then dried  in air.   The stripped films were counted by a spark counter.   Figure 2 shows the
etching device, and Fig.3   shows the structure of spark counter.   To  enlarge the holes on the CN films, the films
were presparked  three  times  under 600V  HV.  Six times spark counting under 450V can get the a  tracks on
each film in a good statistics.   Figure 4 shows the pulse of spark counter from an oscilloscope.
Calibration of CN film
     For each batch of films to be etched, 10 calibration  films  were inserted during chemical etching to get the
individual  response factors.  The  calibration films were exposed  in a radon chamber of 3.0m x 1.8m x 2.4m with
radon concentration mainteined at around  1000 Bqm .  The radon concentration  in the  chamber was monitored
                                                                      -3          -2      -3
by  a scintillation  cell.   The response factor of CN films is  1.02 ± 0.12 x 10 track . cm /Bqm .h  obtained  from
 10  exposed films.
Measurement of  indoor t dose
      Indoor i  dose was  measured with Panasonic UD-200S TLD and UD-512 TLD reader.   Its feasible dose
range is from 0.1 mR  to 20 R.   While in  reading, the TLDs were heated with hot nitrogen for 10 s.   from room
temperature to 430 °C .  The CaSO4:Tm was sealed  in glass tube and shielded with copper cap to reduce the low
                                                          g
energy response.  The system was calibrated with a  3.7 x 10 Bq radium-226  source  which is close to  the
characteristics of natural radiation field.  Its fading  effect is negligible.

4.  Results and Discussion
 Analysis of the results
        The results of radon  survey in Taiwan  during Dec. 1989 through June 1991 is shown in Table 2.    The
 Taiwan island is divided into 24  districts and data  are collected every two months.  Each district contained 6-29
 homes in the survey.   Buildings  in granite  rocks are  different from others and were  considerd  as a special
 district.Indoor to outdoor  concentration ratios and indoor/outdoor mean concentrations are also listed in Table 2.
 The indoor mean concentration is 9.93 + 4.08 Bqm  while the outdoor is 4.0 + 3.2 Bqm  .
       Figure 5 shows the indoor and outdoor radon concentrations  and the mean ratio is 2.5.  Nantou County
 (district-lO)has higher indoor and outdoor radon levels as compared with the other counties.  The Nantou  county
 is  far from the coastal line and surrounded by  mountains where radon is easier to accumulate in air  and leads to
 higher level.    In addition,  Nantou county has many  brick-made  houses that  were included in this survey. The

-------
concentration in underground buildings of granite is 36-64 Bqm  with an average of 51 Bqm .   Its 5 times higher
than the average of Taiwanese houses. The last row of Table 2 shows that the radon levels in houses of Kinman
are also higher than that of Taiwan.
Comparison with other countries
     The New York State data show that there are 70% homes between 0-37 Bqm  and  17.6% between  37-74
Bqm . Table 4  lists  the data of some countries  in the world.  Radon level in Taiwan is low as compared with
  ,          .  (4)
other countries  .
Dose assessment
other countries . The result is close to Japan and far below New York State of U.S.A.
     There are many dose assessment models for radon-induced lung dose( 1,7,8). According to the dose model of
James, the  dose conversion factor of indoor radon is 0.05 mSvy per Bqm   for adults.   The calculated per caput
dose for Taiwan population is 0.5 mSv/y and the collective dose is  10  manSv/y.
     The average indoor  7 absorb dose rate is 0.12  u. Gy/h.  Suppose that 80% of the time  is spent indoor, the
induced per caput dose is  0.85 mGy/y.   By applying ihe conversion factor 0.7 for absorbed dose to effective dose
equivalent (l),the induced  average effective dose equivalent to Taiwanese  population is 0.59 mSv/y. The collective
dose in Taiwan is 1.19 X 10 man • Sv/y.

5. Conclusions
       Indoor radon level  is strongly dependent on  the ventilation rate  and radon exhalation rate. Radium-226
content is  moderate in concrete of Taiwanese house and the exhalation rate from concrete is  not high as
compared to that of soil.   The warm temperature makes  the ventilation rate in Taiwanese house higher than that
in temperate region.  So the radon level in Taiwan is lower as comapared with that in temperate region countries.
  From the previous discussion, we conclude:
(1) The average radon level based on the survey of 250 Taiwanese  homes  is 9.9 ± 4.1 Bqm" which  is much lower
   than the action level of U.S.  EPA.  Even the maximum indoor Rn concentration  is still  much lower than 150
   Bqm .   The outdoor Rn level is 4.0 ± 3.2 Bqm .
(2) The ratio of indoor to outdoor radon level is 2.5.
(3) Both Nantou  County  and Kinman  County possess  higher  radon level because of their geographical and
   geological characteristics.
(4) The traditional Taiwanese houses  have  higher indoor Rn level than modern concrete houses.
(5) Indoor  t  dose  rate varied from 0.066 to  0.189  u. Gy/h with an average of 0.121  u  Gy/h which is higher
   than temperate countries.  The annual per caput dose from i  radiation is 0.59 mSv.
(6) The correlation  coefficient  of  indoor  y   dose rate  and Rn level  shows  a significant correlation at 0.95
   confident level.
(7) Indoor Rn  level  shows  no seasonal variaion in subtropical Taiwan.
(8) Building type and wall  materials  show no significant  correlation with indoor Rn  level.  Traditional building
   materials showed higher indoor radon level.
(9) Building in granite rocks shows high radon level even in good air ventilation system.
(10) The effective dose equivalent due to indoor radon is about 0.5 mSv/y in Taiwan.

-------
                                            REFERENCES
1. United Nations Scientific Committee on the effect of Atomic Radiation. Sources, effects and risks of ionizing
  radiaiton.  New York, United Nations,  1988.
2. Cohen, B.L. and Gromicko, N. Radon-222 levels in low income households. Health Phys. 56:349, 1989.
                                                                                222
3. Langroo, M.K., Wise,  K.N., Duggleby, J.C. and Kotler, L.H.  A nationwide survey of  Rn  and  7   radiation
    levels in Australian  homes, Health Phys. 61:753, 1991.
                                        222
4. Rahlenbeck S.I. and Stolwijk, A.   Indoor   Rn level in  New York state, North Carolina, and South Carolina.
    Health Phys.  61:879, 1991.
5. Cohen,  B.L. and Shah, R.S.    Radon  levels in United States  homes by states and counties.   Health Phys.
    60:243, 1991.
6. Tommasion, L. Etch track techniques in radiation dosimetry, Radia. Protect. Dosim. 17:135, 1986.
7. 1CRP. Lung  cancer  risk from  indoor exposures  to radon  daughters.   Oxford: Pergamon Press: ICRP
  Publication No.50, 1987.
8. James, A.C. Lung dosimetry In: Nazaroff, W.W., Nero, A.V., eds.  Radon and its decay products in air, New
  York Wiley, 1988. P.259.

-------
Fig. 1. Passive measuring device for environmental radon. CN films were adhered on the inner wall of the CUD.
                             Fig. 2. Chemical etching device for CN films.

-------
CN film
           Positive electrode Negative e|ectrode.  °
Heavy material






 Thick cloth





     Aluminum mylar
                                                                                    Base
                           Fig. 3. Structure of the electrode of spark counter.
                                  Fig. 4. Pulses of spark counter.

-------
                                         (  1989.1E - 1991.06 )
           15
c
cr
&
                 i   2  3  4  5  6   7   8  9  10 11 12  13  W 15 16  17 13 19 20 21  22 23
                                                Districts


  Fig. 5. Indoor and outdoor radon levels of 23 districts in Taiwan.   The average indoor/outdoor ratio is 2.5.




i"
PERCENT







30
28
26
24
20
18
16
14
12
10
S
6
4
2
0
30.29













.07
~|

\
\
\
\
\
\
\
\
\
\
\
s
\

\
\

\
28.2
\
\
\
\
\
\
\
\
\
\
\
\

\
\

\
2


20.33
\
\
\
\
\
\
\
\

\
\

\





7.05
VI
4.56
\ r\i
\ \
^ *^ u 83D 83% ^ "^" 0 83
\ \ Fvi 1X1 IXI n f\l An ° ° Pvl ° p^aAo ° ° rU
                                    Rn ACTIVITY (Bq nr
                  Fig. 6. Distribution of radon concentrations in Taiwanese homes.

-------
E

$
<
 c


W
O


UJ
  15  -r


  14  -


  13  -


  12  -


  11  -


  10  -


  9  -


  8  -


  7  -


  6  -


  5  -


  4  -


  3  -


  2  -


  1  -


  0  --
11.06
                      10.53
                      9.14-
                                            9.73    9.6
                               9.6
                                                                          9.27
                             8.39
                     89.12   90.02  90.04   90.06  90.08   90.10   90.12   91.02


                                          YEAR AND MONTH




                  Fig. 7. Time variation of indoor radon level in Taiwan.
O

 3
 O
 Q
 o
0.19


0.18


0.17


0.16


0.15


0.14


0.13


0.12


0.11


 0.1


0.09


0.08


0.07


0.06
                                  20                   40


                      Fig. 8. Correlation of indoor radon and gamma dose.


                             r  =0.3 for 42 houses.
                                                                     60

-------
Table 1. The mean natural radionuclide concentrations and their ratios in rocks
Radionuclide concentrator
Type of rocks
Igneous
Andesite
Basalt
Dunite
Gabbro
sedimentaries
Conglomerate
Sandstone
Shale
Limestone
Metamorphics
Slate
Green schist
Black schist
Phyllite
Gneiss
«.,

1.11
1.35
0.13
0.05

0.88
1.75
2.70
0.26

2.80
1.07
2.96
3.13
1.90
Th(ppm)

4.18
4.86
0.10
0.15

4.09
7.48
12.40
1.26

13.50
1.62
13.40
16.30
9.70
U(ppm)

1.20
1.27
0.10
0.07

0.87
1.53
2.51
1.36

2.62
0.43
2.47
2.81
1.85

Th(ppm)/

3.77
3.60
0.67
2.75

4.65
4.27
4.58
4.85

4.84
2.14
4.51
5.19
5.09
Raitos
U(ppm)/

1.08
0.94
0.82
1.25

1.15
0.87
0.93
5.23

0.94
0.62
0.83
0.90
0.97


Th(ppm)/
U(ppm)

3.
3.
0.
2.

4.
4.
4.
0.

5.
4.
5.
5.
5.

48
82
82
20

05
89
92
93

17
22
40
79
23
*K: 308 (Bq kg"')/K%
232Th:  4.07(Bq k
238
  ;U :12.62(Bq kg-'j/ppm

-------
                       Table 2  Questionnaire for indoor radon survey




1. Your name:



2. Your address:



   Phone No.:



3. Type of house:(l) Appartments (2) Bungalow (3) Detached (4) Others



4. Main building materials:(l) Concrete (2) Brick with plaster  (3) Wooden (4) Wood and bricks




  (5) Others



5. Flooring  materials:(l) Concrete  (2) Sheet  vinyl (3)  Marble (4) Wooden  (5) Tile (6) Grind




  aggregates (7) Others



6. Wall materials:(l) Tile (2)  Paint (3) Wall paper (4) Wooden  (5) Others




7. Construction year:



8. Measured room:(l) Living  room (2) Bed room



9. Dimension of measured room:	m long,	m wide	m height.

-------
 Table 3  Indoor and outdoor radon concentration in 24 districts of Taiwan
district
01 Keelung
02 Hualien *
03 Taitung *
04 Ham *
05 Taipei
06 Taipei *
07 Taoyuan *
08 Hsinchu *
09 Miaoli
10 Nantau *
1 1 Taichung *
12 Taichung
13 Clianghua *
14 Yuanlin *
15 Chioyi
16 Tainan *
17 Tainan
18 Kaohsiung *
19 Fungshan
20 Kaohsiung
21 Pingtung *
22 Penghu *
23 Kinman *
24 Kinman #
indoor Rn Cone.
range (Bqm"3)
5.7 - 9.9
5.5 - 12.6
5.4 - 12.8
6.2 - 13.9
4.7 - 19.0
4.4 - 24.1
5.1 - 13.6
6.3 - 17.3
7.2 - 14.5
9.8 - 22.0
8.1 - 15.3
5.7 - 20.4
4.7 - 11.8
4.6 - 18.0
5.0 - 23.8
5.8 - 14.9
8.3 - 21.9
4.7 - 10.5
6.3 - 10.1
6.4 - 22.1
5.5 - 17.8
5.3 - 12.4
9.0 - 24.8
36.3 - 63.5
mean indoor Rn
cone. (Bqiri3)
8.4 ± 1.5
8.2 ± 2.1
8.9 ± 3.1
8.0 ± 2.9
10.6 ± 4.3
10.2 + 5.6
9.0 ± 2.5
10.1 ± 3.0
9.9 ± 2.3
14.4 ± 3.8
10.4 ± 2.6
10.8 ± 5.4
7.8 ± 3.0
8.0 ± 4.1
9.8 ± 6.4
8.2 ± 3.6
13.1 ± 5.0
8.3 ± 2.5
8.5 ± 1.7
10.6 ± 3.8
7.2 ± 3.9
8.3 ± 2.3
14.4 ± 4.0
50.8 ± 13.4
outdoor Rn cone.
( Bqm-3)
3.3 ± 1.8
4.3 ± 1.4
*** ***
*** ***
3.0 ± 1.7
5.5 ± 1.5
5.0 ± 2.7
4.3 ± 4.9
3.4 + 1.5
5.5 ± 1.5
5.4 ± 4.9
4.0 ± 4.2
2.2 ± 1.8
3.8 ± 4.5
1.8 ± 1.3
* *# _L. * * *
3.1 ± 2.3
4.4 ± 4.9
5.4 ± 5.3
3.1 + 3.9
5.1 ± 3.0
4.3 ± 3.7
3.5 ± 3.6
3.5 ± 3.6
Remark :  1. Indoor Rn mean cone.  9.9 ± 4 Bqm"3 .
         2. Outdoor Rn mean cone. 4.0 ± 3 Bqm'3 .
         3. * county.
         4. # under ground passage .

-------
Table 4. Dependence of indoor radon level on building characteristics
Type of building Main building materials
Appartments Concrete
10.8 (58) 9.9 (159)
Bungalow Brick with plaster
9.4 (33) 8.9 (42)
Detached Wooden
9.3 (90) — (0)
Others Others
10.6 (31) 19.4 (4)






Flooring materials Wall materials
Concrete Tile
10.3 (29) 10.6(10)
Sheet vinyl Paint
9.2 (20) 9.5 (158)
Marble Wall paper
8.6 (14) 10.4 (19)
Wooden Wooden
14.5 (8) 10.7 (6)
Tile Others
9.5 (74) 12.1 (19)
Grind aggegates
9.6 (62)
Others
16.9 (5)

-------
Table 5. Radon concentration obtained in some
          recent larg-scale indoor surveys
Country
or area
Argentina
Belgium

Denmark



Germany



Italy

Ireland

Japan



Norway

United

Taiwan

Number of Type of Date of
dwellings sampling completion
112 Radon, passive 1985
79 Radon, passive, 1984
one-year exposure
400 Radon, passive 1935
3-month exposure
in summer and
in winter
5970 Radon,passive 1984
3-month exposure.
1 in bedroom
1 in living room
1000 Radon, passive, 1984
3-12 month exposures
736 Radon passive 1987
6-month exposure
250 Radon, electrostatic 1988
integrating, 2 months;
total 1.5 year
exposure
1500 Radon, passive 1985
1-week exposure
552 various 1934

250 Radon, passive 1991

Average vaule
(Bq/m3)
37 (median)
41 (median)


50 (median)


40 (median)
49 (mean)


43 (median)

37 (median)


10 (mean)


90 (mean)

35 (median)
61 (mean)
10 (mean)
8.5 (median)

-------
         Session X Posters
Radon in Schools and Large Buildings

-------
                                                                  XP-1
         SOLAR FRESH AIR VENTILATION FOR RADON REDUCTION
               by:  Monty Holmes
                    Intermountain Radon Service
                    P. O. Box 3
                    Salida, CO  81201

                    Kelly W. Leovic
                    Air and Energy Engineering Research Laboratory
                    U.S. Environmental Protection Agency
                    Research Triangle Park, NC  27711
                            ABSTRACT

     This paper  discusses the construction  and evaluation  of  a
solar fresh air ventilation system installed in two classrooms of
a Ranches de Taos, New Mexico, school.  The project was initiated
because school radon  mitigation  research has shown  the  need for
improved indoor air quality in addition  to reduction of  elevated
radon levels.    Additionally, there are certain types of school
buildings  where  subslab  approaches  to   radon   reduction  are
impractical or expensive.

     A datalogger recorded continuous radon concentration, carbon
dioxide concentration, room temperature,  and subslab-to-classroom
differential pressures in two  ventilated  and one control classroom
during  the  February  to  April 1992  testing period.   While the
building was occupied the solar ventilation system did not provide
much  additional   radon  reduction  because  occupants'  frequently
opened windows in the school.  However, the classrooms  with the
ventilators did have slightly  lower carbon  dioxide levels compared
to the control classroom, indicating some benefit of ventilation.

     To evaluate the effect the ventilation system on radon levels
without the confounding factor of the natural ventilation provided
by open windows,  the system was also evaluated during spring break
when  the  building was  unoccupied.   Results during  this period
indicate lower radon  levels in the two classrooms  with the solar
ventilators as compared with the control classroom.

     This paper  has been reviewed in accordance with the  U. S.
Environmental Protection Agency's  (EPA)  peer  and  administrative
review policies and approved for presentation and publication.

-------
                           INTRODUCTION

     School  radon mitigation  research has  shown  the need  for
improved indoor air quality  in  addition to reduction of elevated
radon levels  (1).  To address both  of  these  needs,  a solar fresh
air ventilator was developed by Monty Holmes (one of the authors)
and received an "honorable mention"  in the 1991 Innovative Radon
Mitigation  Design Competition.    The  design  was   a  fresh  air
ventilation system utilizing solar tempered outdoor air to dilute
(and possibly pressurize)  a  school classroom.   This presents an
alternative radon mitigation technique that  provides conditioned
outdoor air  to  help reduce  radon  levels  and improve  indoor air
quality at a relatively low operation cost.

     The system was designed  to  use  an electrically powered fan to
continually supply outdoor air  during  occupied  periods to reduce
radon levels and improve indoor air quality.   The system supplies
"tempered" outdoor air to the classrooms and is not intended as a
heating source.   The Environmental Protection Agency's  (EPA's) Air
and Energy Engineering Research Laboratory (AEERL)  evaluated the
solar ventilator since it  offered  a radon mitigation alternative
for buildings with the following characteristics:

     •    elevated radon levels
     •    sunny climate (for solar capability)
     •    occupied  primarily  during  daylight  hours  (to  take
          advantage of solar capability)
     •    active soil depressurization (ASD) was  not reasonably
          applicable (low permeability subslab fill)
     •    existing heating system did not supply  sufficient outdoor
          air

     Radon measurements in New  Mexico  school buildings indicated
that over 20  percent  of  the 125 schools tested had at least one
classroom with elevated radon levels.  In Taos County, every school
had rooms where weekend radon measurements following current EPA
protocols (2)  exceeded 4 pCi/L.*  Taos has a sunny climate; however,
it also has over 7,000 degree days of heating load.  As a result,
building plans from a number of Taos schools under consideration
for this project were  reviewed.  In a majority of  the buildings
investigated, slabs were poured  on  compacted  adobe or dirt with no
subslab aggregate.  It was felt that an alternative to ASD would be
necessary in these types of buildings.
* 1 pCi/L (picocurie per liter) *= 37 becquerels per cubic meter

-------
                   SCHOOL BACKGROUND INFORMATION

     A school located in Ranches de Taos,  New Mexico, was selected
 for  this  research  project because  the  school  has  moderately
 elevated  radon levels  (between  5 and  15 pCi/L)  in half  of the
 classrooms, slab-on-grade construction  on compacted adobe, and no
 conditioned  outdoor air  in most classrooms.    Selection  of an
 appropriate building where  school officials were also amenable to
 the  project  was more difficult  than  anticipated.   Consequently,
 when the school in Ranches  de Taos was  selected  for this project,
 school officials and the researchers agreed that the installation
 would be temporary for  the  purpose of evaluating the solar fresh
 air  ventilation system.  School officials  would have the option of
 keeping the system once the project was completed.

     Constructed in 1965, the school is a single-story cinder block
 building  with a   2  percent slope  built-up asphalt  roof.   The
 exterior finish  is stucco, and  the interior  walls  are painted.
 There are 15 classrooms that average about 25 occupants and about
 900 square feet* in area.  The eight perimeter classrooms are heated
 by a gas fired hot  water radiant heating system  with one thermostat
 controlling all the  classrooms.  Each  perimeter classroom has two
 operable 40 x 24 inch" single glazed awning type windows.  Except
 for the north-facing rooms,  the rooms in the school  were very warm.
 A floor plan of the school  is displayed in Figure  1.

     Classrooms 3  and 4, located on  the south- and west-facing
 sides, respectively,  were  selected for evaluation of  the solar
 ventilator.  Room 5,  on the  west  side, was selected as the control
 room.    January  1992  radon  measurements  using  2-day  charcoal
 canisters  over a  weekend  were  9.7,   11.5,  and  11.0  pCi/L in
 classrooms 3, 4, and 5, respectively.   Alpha track detector (ATD)
 measurements by the New Mexico Environment  Department from February
 to May 1991 were 3.6 pCi/L in room 3;  5.6  pCi/L in  room 4; and 2.4
 pCi/L in room 5.   Lower  average radon levels would  be expected
 during this 3.5 month period since windows and doors were probably
 open during the spring.

     Subslab radon concentrations in  classrooms  3, 4,  and 5 were
measured in February 1992  through a  0.375  inch diameter hole using
 a Pylon AB5 in a "sniff" configuration.  Classroom 3 measured about
 800 pCi/L; classroom 4 measured  about 600 pCi/L; and classroom 5
measured highest at about 1250 pCi/L.
*1  square  foot  =  0.093  square meter
**1 inch = 2.54 centimeters

-------
                SOLAR FRESH AIR VENTILATOR DESIGN

     In order to evaluate the solar ventilators in two classrooms,
two 64  square  foot  solar  collectors  were constructed  for  this
project.   A schematic of  the  ventilator design is  displayed in
Figure  2  and  discussed  in Reference  3.   Each collector warms
outdoor air as the  air  moves  through a  30  square inch* cross-
sectional area  serpentine  channel 20 feet" long.   The air moves
under a 5 mil"* black chrome selective surface absorber plate.  A
1.5 inch dead air space  between the absorber and  tempered glass
glazing decreases heat loss from the collector.

     To provide the necessary ventilation air to reduce radon, an
airflow of  150 cfm"" was  targeted.    The  volume of air in each
classroom was about 7500 cubic  feet*"".  An outdoor air supply of
150 cfm would be  slightly  greater than 1 air change per hour, at
9000 cubic feet per hour. A 265 cfm (free air)  blower was  initially
selected  for each  of the  ventilation  systems,  but  these were
replaced with 495 cfm (free air) blowers because of  the  high static
pressures in the systems.

     The collectors were built on the school site in February 1992
using the following materials:

       preassembled wooden framework
       ductboard, 1 inch thick  and 4 x 10  feet in area
       1.25 inch metal "zee" flashing
       3.625  and  1.625 inch metal stud "c"  channel
       24 x 94  inch selective surface  absorber sheets
       0.15625  x  46 x 96 inch low iron tempered  glass
       miscellaneous screws and rivets
       high temperature  silicone caulking

      Material  costs  for   these "state of  the  art"  all-metal
collectors  were about $500  dollars  each.   Costs  for additional
materials  (such as fans,  ducting, diffuser grille,  insulation,  and
wiring) totaled about $300  for  each  system.   Each  system required
about 25 person hours to  build and 25 person hours to install.   The
collectors  are  easy to build and have a 20 year life expectancy.

     The  two  collectors were  mounted  at  70  degrees  from  the
horizontal  at a south-southwest compass orientation,  adjacent to
classroom  3.   (This installation was considered temporary for the
 *1 square inch = 6.45 square centimeters
 **1  foot  =  0.305  meter
 *"l mil = 25.4 micrometers
 ****! cubic  foot per minute  (cfm)  =0.47  liter per second
 ***"*! cubic foot = 0.028 cubic meter

-------
purposes  of  evaluating  the ventilators'  potential  for  radon
reduction.   For  a permanent installation, a roof  mount  would be
preferred.)  The collector support used for this project was a 4 x
4 inch redwood beam supported 4 feet on center with 4  x 8 x 16 inch
solid cinder blocks and secured to the ground  with 36  inch steel
stakes.  Angle brackets  (12 x 12  inches) secured  the top  of the
collectors to the outside wall of classroom 3.

     Outdoor air  enters the collector  through  an  8 inch diameter
furnace filter covered with 0.25 inch grid hardware mesh.  A blower
supplies 100 percent outdoor air into a 4 x 12 x 7 inch round sheet
metal boot at the collector inlet.  The air is warmed as it moves
through a  20 foot long serpentine pathway with  a 1.5 x  20 inch
cross-sectional area under the absorber.

     The heated  air  exits the  collector via a 4 x 14  inch metal
boot connected to an elbow, and then  flows into  8 inch diameter
insulated flexible ducting.   (In  a permanent  installation,  sheet
metal  ducting would  be  preferred.)    The flexible ducting  is
insulated with 1  inch  foil-faced  duct  wrap to  provide additional
insulation and weather resistance.  The ducting from the collector
to classroom 3  is about  20  feet  long, and the ducting  from the
collector to classroom 4  is about  60 feet  long.  From the ducting,
the  air enters  the  classroom  through  a 2  x  2  foot  diffuser
installed  vertically  in  the   outside wall.    The  system  was
originally designed with  a remote bulb thermostat to  activate a
1250 watt in-line duct heater when air temperature dropped below 60
degrees Fahrenheit;* however, this additional  heat was not needed
so the heater was removed since it contributed to pressure loss.

     A timer was used to  control fan operation.  The timer was set
to operate Monday through Saturday,  7:00 a.m.  to 4:00 p.m.   On
Sunday  the system  was off  to  obtain  background  data with the
classrooms unoccupied.

     Teachers in classrooms 3, 4, and 5  were instructed to minimize
open doors and windows during the testing period.   Impromptu site
inspections  indicated that  this instruction was  not consistently
followed. This was because the existing heating,  ventilating, and
air-conditioning  (HVAC) system  in  the school often  keeps the rooms
on the south and west sides of the building too warm.

     System airflow measurements were made on April 1 and repeated
May 21, 1992.  The airflows into classroom 3 averaged 165 cfm, and
the airflows into classroom 4 averaged 130 cfm.  These airflows are
consistent with the longer  duct run to classroom  4 (60 versus 20
feet).


*°F = (9/5 °C)  + 32

-------
      Indoor-to-outdoor differential pressure measurements were made
with  a micromanometer.   Baseline measurements  made in February
averaged  -  0.008  inch water  column"  in perimeter  classrooms 1
through 8.  These measurements were repeated  in April and May with
the ventilators operating.  However, conditions were  windy at both
times,  and  it was  difficult  to  ascertain  the  effect of the
ventilator  on the indoor-to-outdoor differential  pressure.   The
effect  of  the ventilator  on  the  room-to-subslab   differential
pressure  (typically  less  sensitive to wind effects)   is  discussed
later in this paper.

      The measured airflows to the classrooms,  together with the fan
manufacturer's  performance  curve  data  (airflow   vs.  static
pressure),  indicate  that  the static pressures  are 0.94 and 0.98
inch  water  column in classrooms  3 and  4, respectively.  Static
pressures in the system are caused by:

      (1)  The  resistance  to air  movement through the  collector
          baffles;
      (2)  90 degree turns in the air pathways in the  ducting;
      (3)  Resistance  of  the  spiral duct  walls of  the flexible
          ducting;
      (4)  Resistance of the electric coil heater in the  airstream
          of the duct heater (this was subsequently removed);
      (5)  Flexible  ducting  from  collector   to diffuser   (in  a
          permanent system,  sheet metal ducting would be used); and
      (6)  The diffusing air grille in each classroom.


                  EVALUATION OF SOLAR VENTILATOR

      Classrooms 3, 4, and  5  were instrumented  with  a datalogger to
collect   continuous   radon    concentration,   carbon   dioxide
concentration, temperature,  and differential pressure  measurements
from February to May  1992.  The datalogging system used by AEERL is
described in Reference 4.

     Data collected during February and March 1992 indicated that
radon levels were frequently below 4 pCi/L in both the ventilated
rooms (classrooms 3 and 4) and the control (classroom 5) whenever
the school  was occupied.    Site inspections  and analyses  of the
continuous data (for example, only slightly higher carbon dioxide
levels in classroom  5 during occupied periods)  implied that the
windows were frequently opened during occupied  periods, interfering
with the evaluation of the solar ventilator.  As a result, the most
appropriate data for  simulating the effect of the ventilators under
typical winter  conditions were collected during spring break in
April when the building was unoccupied.   The measurements during
this week form the basis of much of the discussion below.
  1  in. H,0 = 2.5 kPa

-------
     Figure 3 shows  that background radon levels —  measured on
April 20 after the ventilators had been off for about 32 hours —
were about 5 pCi/L in classrooms 3 and 4.  Background levels during
this same period  in  classroom 5 (the control) were about  6  to 7
pCi/L.  It is  probable that the radon  levels in classrooms 3 and 4
were not able  to build back up to background levels in the 32 hours
that the ventilators were off.  Unfortunately, long-term continuous
radon data were lost.

     During the day, while the ventilators are  operating and the
building is closed, radon levels in the two ventilated classrooms,
3 and 4,  are  reduced to below 2 pCi/L.  Radon  concentrations in
classroom 5 were always above 5 pCi/L, and sometimes were above 10
pCi/L during this week.

     For comparison,  radon data for the following week — while
school was in  session — are shown in Figure 4.  Radon reduction in
classrooms  3  and 4 presumably occurred  due  to operation  of the
ventilator and by opening windows, as radon levels during the day
were well  below  1 pCi/L.   Radon reduction  in classroom  5   (the
control) was probably a result of open windows during the day. On
Saturday, May 2,  radon levels  were  still  well below  1 pCi/L in
classrooms  3  and  4  while the  ventilator was operating  and the
school unoccupied.  Levels in the control room were about 8 pCi/L
during this time.

     The radon data  in  Figures 3 and 4  often  show that, when the
ventilators are operating in classrooms 3 and 4 and the school is
not  occupied, radon  levels  in  classroom  5  increase.    This is
particularly  apparent observing the radon peaks in classroom  5 on
April 20, 21,  25, and May 2.    It is not known whether these peaks
are coincidental  or whether the ventilators in classrooms 3 and 4
actually increase radon levels  in classroom 5.

     The  indoor-to-outdoor  differential  pressure  measurements
indicated  that  there  was  very little  pressurization  in  the
ventilated  classrooms.    This  is  not  surprising  since  the
ventilators supply only about 130 and 165 cfm  of  outdoor air to
classrooms  4  and 3,  respectively.   With  these relatively low
differential  pressures,  other factors  such  as  wind  speed and
direction,  room  occupancy schedules,  and barometric pressure can
also greatly influence the indoor-to-outdoor differential pressure.

     The  continuous  classroom-to-slab pressure differentials for
the week of spring break under  closed building  conditions are shown
in  Figure  5.   These  data  indicate  that  room  3  was slightly
pressurized relative to  the  subslab  when  the  ventilators  were
operating.  The room-to-subslab pressure in  room  3 decreased on
Sunday when the ventilator was off.   Room 4, with less  outdoor air
supply,  was not consistently pressurized,  maintaining a pressure
close  to  neutral  most  of   the time.   The  control  room was
consistently  under a negative  pressure throughout the week.

-------
     Analyses of the carbon dioxide data for the week of March 23
(Figure 6)  show slightly higher  levels in classroom 5.  These data
also show that the carbon dioxide levels took longer to level off
at the end of  the day in classroom 5 than  in  classrooms  3  and 4
(with the ventilators).   This is probably  because  some dilution
occurred in all three  classrooms during the day when the windows
were frequently opened; however, classrooms  3  and 4  also had the
additional ventilation provided by the solar ventilators so that
carbon dioxide levels fell rather quickly  once  the  source (the
occupants)  was removed.  This  was because the ventilators operated
for about 1 hour after the occupants left.

     Background  carbon dioxide  levels  indicate some  variation
between the three  rooms.   Since background  carbon dioxide levels
should  be  the same  in  all  three  rooms,   these data  imply  a
calibration  problem.    The manufacturer  of  the carbon  dioxide
monitor  recommends calibrating  the instruments to  300 ppm  in
outdoor air.   If the  background  levels  for all  three  rooms are
adjusted so that  background levels are 300  ppm, these data still
show that  carbon dioxide levels are higher in classroom 5.   In
fact, spikes above the 1000 ppm guideline  frequently occurred in
classroom  5  and  only  occasionally  in  classrooms  3  and  4,
demonstrating  the added benefit of  this controlled  outdoor air
supply.

     The temperature  data  are not presented here since the room
temperatures in the classrooms  are  influenced by many factors in
addition to  the  ventilators.    These  include:  existing heating
system, south-southwest room  orientation, and opening of windows
during occupied periods.

                           CONCLUSIONS

     An objective evaluation of  the ability  of  the solar fresh air
ventilation system to  reduce indoor radon levels and improve indoor
air quality was  complicated by  occupants'  opening the windows in
both the ventilated and control classrooms.   During the occupied
periods,  radon  levels were  well below  4  pCi/L  in  all  three
classrooms  because occupants  frequently opened windows  in the
school, even during cold weather.   Although the open windows do
have a  positive effect on radon reduction, this is  not a cost-
effective long-term radon  (or thermal  comfort) control strategy.
Although  radon   levels were  consistently  low  in  all  three
classrooms,  carbon dioxide  levels showed  that  some additional
ventilation benefits were achieved in the two  classrooms with the
ventilators.

     To simulate the effect of the  ventilators  under closed school
conditions —  that one would normally expect during the winter in
a cold climate — radon levels  for all three rooms were  compared
during spring  break when the  building  was unoccupied and windows
were closed.   Results during this period  indicate  consistently

                                8

-------
lower radon levels in the two classrooms with the solar ventilators
than in the control room.

     In retrospect,  this school  was  not  an  ideal candidate  to
evaluate  the solar  fresh  air  ventilation  system  because  the
occupants   frequently   opened   the  windows  even  without   the
ventilators.  The  teachers  were asked to keep  the windows  and
classroom-to-hallway doors closed as much  as possible  during  the
testing,  but certainly  not to compromise  the   comfort  of  the
students.  Switches on  the operable windows and doors  would  have
helped to monitor this condition.  As  a long term solution to the
over-heating problem (caused by the existing heating system),  the
researchers  recommended  that  the  school  maintenance  personnel
adjust the  building HVAC  system to  better  adjust  to temperatures
throughout the building.  For example,  rooms on the south and west
sides are typically very warm,  whereas  rooms on the north side are
sometimes cold.   It was also recommended that a permanent supply of
conditioned outdoor air be delivered to all occupied rooms in the
school. This will help to address comfort, radon concentration, and
indoor air quality.

     In order to approach the ASHRAE guideline of 15 cfm per person
(about  375  cfm   per  classroom) , the air  supply  of the  existing
ventilators  would  need to be  increased.   Using  larger  diameter
sheet metal  ducting  instead  of flexible  ducting,  decreasing  duct
length, and using a larger fan would all help to  increase airflow
into the classrooms.

     The use of  single classroom solar ventilators  in a school with
several  rooms  with  elevated  radon  levels  is  not  the  best
application  of  this  system.   One  64  square foot  collector  per
classroom would  be  too expensive,  cumbersome, and aesthetically
unpleasing to use for radon control  throughout the building.  If a
school has  special construction or  site  conditions that preclude
ASD, and only a  few rooms with high radon  levels, this system might
be more practical.

     A  better application  for  this  system might  be to  supply
tempered outdoor air to the return plenum of a central HVAC system
serving several  classrooms.  The limited ability to preheat outdoor
air is a real problem  in schools in cold climates.   As a result,
outdoor air  supplies are often cut  off or restricted during very
cold weather to  avoid freezing  the heating coils and/or to reduce
energy costs. Use of the  solar collector to preheat  this air would
help to maintain a consistent  outdoor air supply throughout the
year.

-------
                           REFERENCES

1.   Leovic, K.W., A. B. Craig, and D.  W. Saum, The Influences of
     HVAC Design  and Operation on  Radon Mitigation  of  Existing
     School Buildings.  IN:  Proceedings of  ASHRAE IAQ '89.   The
     Human Equation:  Health and Comfort.  San Diego,  1989. (NTIS
     PB89-218762).

2.   Radon  Measurements in  Schools:  An  Interim  Report.    U.S.
     Environmental   Protection  Agency's  Office   of  Radiation
     Programs, EPA-520/1-89-010 (NTIS PB89-189419).  March 1989.

3.   Kornher, Steve and  Andy Zuagg.  The Complete Handbook of Solar
     Air Heating Systems.   Rodale  Press, Emmaus,  PA, 1984.

4.   Harris,  D.B.  and  B.E.  Pyle,  Data  Logging  Systems  for
     Monitoring Long-Term Radon Mitigation Experimental Programs in
     Schools  and  Other  Large Buildings.   Presented  at the  85th
     Annual AWMA Meeting,  Kansas City,  MO,  June 21-26,  1992.
                        ACKNOWLEDGEMENTS

The authors would like to express their appreciation to the school
officials  in  the  Taos  Municipal  School  System  in New  Mexico.
Appreciation is  also extended  to  EPA and Acurex  staff  for their
invaluable assistance.
                               10

-------
   C.R. #6
                    C.R. #7
                       _J
C.R. #6
                                                   Exit
                                                                          A


                                                                          I
                                                                          <5
                                                                          CD
                                                                          o
                                                                          o
<«Exit
                                                     Closed Corridor       ToGym»»»» ,'
    C.R. #5

1
    C.R. #4
                       C.R. #9
                        Library
                                    C.R.
                                                   Mech.
                                                           Sto.
                                                        Girts
                                         /N*/Vk*1
  C.R. #11
                                                         Boys


                                                     M  W Nurse
                                                     Special Ed.
                                                                 Reading Lab
                                                                   Sio.
                                                                      3
                                                                       rincipa
                                                                   Reception
                                                                                C.R. #12
 Workroom
    C.R. #3
                    C.R. #2
                                         :T
                                             Exit>»       Covered Entry
 C.R. #1
                                                                  N

                                                                    \
                       Figure  1.   Floor  plan of school,
                                           11

-------
Southwest Facing
Wall of School
             Electric  Heater
 2 x 2 foot     (optional)
 diffuser
   \_f
8 inch diameter
flexible ducting
    Classroom Windojc
                      Fan
•'>
«



,-*
\
D
VN

\
i

                                                         64  square
                                                     ^ foot  solar
                                                    "^ coll
                           ector
                       (arrows show
                        air flow)
                                                         Grade  level
       Figure 2.  Plan view of solar fresh air ventilator.
                               12

-------
20
                               Ventilator On Time,  x-axis
                               (Mon  to  Sat, 7:00  a.m. - 4:00  p.m.)
 0
 04/20/92    04/21/92   04/22/92   04/23/92   04/24/92    04/25/92    04/26/92
   Mon       Tues       Wed        Thur        Fri        Sat        Sun
                              DATE (Midnight)
                 Figure 3.    Continuous  radon levels under
                               closed/unoccupied  conditions.
                                      13

-------
   12
   10
    8
w
5
                 Ventilator  On Time,  x-axis
                 (Mon to  Sat,  7:00  a.m. -  4:00 p.m.)
    2 -
    0
    04/27/92 04/28/92  04/29/92 04/30/92  05/01/92  05/02/92 05/03/92  05/04/92  05/05/92
      Mon      Tues    Wed      Thur     Fri       Sat      Sun      Mon     Tues
                                    DATE (Midnight)
         Figure 4.  Continuous  radon levels under occupied conditions.
                                          14

-------
  0.3
  0.2
  0.1
I
W
 •-0.1
g
  -0.2
  -0.3
 -0.5
 -0.6
     Room 3-to-Subslab
Room 4—to—Subslab
                                  Room 5_to-Subshb
                          CMMI •>•
   04/20/92   04/21/92   04/22/92   04/23/92   04/24/92   04/25/92    04/26/92
     Mon       Tues      Wed        Thur       Fri        Sat        Sun
                               DATE (Midnight)
          Figure 5.  Continuous  differential pressure measurements
                       under  closed/unoccupied  conditions.
                                        15

-------
3000
                               -"*-*""•-^-——L
 03/23/92   03/24/92    03/25/92    03/26/92   03/27/92   03/23/92   03/29/92
   Won        Tues       Wed        Thur       Fri        Sat        Sun
                             DATE  (Midnight)
            Figure 6.   Continuous carbon dioxide measurements
                           under  occupied conditions.
                                       16

-------
                                                                   XP-2
          CHARACTERISTICS OF SCHOOL BUILDINGS IN THE U.S.


                         by:


               Harry J.  Chmelynski
               S.  Cohen & Associates,  Inc.
               McLean,  VA 22101

               Kelly W.  Leovic
               Air and Energy Engineering Research Laboratory
               Environmental Protection Agency
               Research Triangle Park, NC  27711


                         ABSTRACT

       A subsample of 100 schools from the Environmental
Protection Agency's  (EPA's) National School Radon Survey were
visited to obtain information on building structure, location of
utility lines, and the type of heating, ventilating, and air-
conditioning system.  Information for each school was entered
into a database to determine the relative proportions of physical
characteristics of the U.S. school building population.  The
results will be used by EPA to guide future radon mitigation
research in schools.  The building characteristics will also be
correlated with school radon levels to identify any relationships
between the physical characteristics and radon levels.

     This paper has  been reviewed in accordance with EPA's peer
and administrative review policies and approved for presentation
and publication.

-------
INTRODUCTION

     The Environmental Protection Agency's (EPA's)  Radon
Mitigation Branch (RMB) conducts research and development on
reduction of indoor radon levels.  To help guide future radon
research in schools and better focus technical guidance
documents, RMB conducted a literature search to find information
that quantifies the physical characteristics of U.S. school
buildings.  Information specific to radon mitigation research in
schools was not found in any existing reports or databases.  In
fact, according to a 1989 publication sy the Education Writers
Association (1),  "Nationally, not even a marginally adequate data
base about school facilities exists.... Several national groups
have conducted surveys of school facilities, but these tend to be
either outdated or incomplete."  As a result, RMB chose to
characterize the U.S. school building population using a sample
of schools from EPA's National School Radon Survey  (NSRS).

     The schools are a nationally representative random sample
selected for the NSRS by EPA's Office of Radiation Programs
(ORP).  To record the necessary information, a building
characteristic profile sheet was completed for each of a sample
of the schools by RMB staff engineers and selected contractors
during 1991 and 1992.

     This paper discusses the random sample selection procedures,
describes the information collected on the building
characteristic profile sheets, summarizes some of the results
recorded on the school profile sheets, compares the results with
those observed in RMB's research schools, and presents the
statistical limitations of this study.  All analyses from this
project will ultimately be summarized in an EPA report.

SAMPLE SELECTION PROCEDURES

     The NSRS consists of two independent samples:  (1) a large
sample of approximately 1,000 schools where all-ground contact
rooms were measured with charcoal canisters, and (2) a smaller
sample of 101 schools where all occupied rooms were measured with
both alpha track detectors  (ATDs) and charcoal canisters.  This
smaller sample was selected independently of the larger sample.
The schools were drawn randomly from lists of schools in 25
geographical areas called Primary Sampling Units (PSUs).  These
25 PSUs were randomly selected for the NSRS from the 125 PSUs
used previously by EPA for the National Residential Radon Survey.
ORP's use of these residential PSUs in selection of schools for
the NSRS is intended to permit comparison of residential and
school building radon concentrations in these PSUs.

-------
     The 125 PSUs used for the Residential Survey were selected
from a list of counties or county-equivalents covering the entire
U. S., except for portions of Alaska and all territories and
possessions.  This list was partitioned in 22 strata,  developed
to guarantee proportioned sample sizes in each of the 10 EPA
regions.  Within each region, counties were assigned to one of
three radon potential categories:  High, Medium,  or Low.  The
assignment of states and substate areas to radon potential
categories is summarized for each region in Table 1.  The number
of residential PSUs selected for the Residential Survey is shown
in the far right column.

     Within each of the 25 NSRS PSUs selected randomly from Table
1, approximately 5 public schools were randomly selected for
inclusion in the NSRS ATD/canister sample, resulting in a total
of 125 schools.  This small sample of schools represents a random
sample of the 78,715 U. S. public school population in 1988 (2).

     For the NSRS, radon was measured (using both ATDs and
charcoal canisters) in 101 of the 125 schools in the sample.  The
remaining schools either refused to participate or were unable to
decide to participate within the time frame allotted for
placement of the ATDs.  One of the 101 schools did not
participate in the profile, resulting in a sample of 100 schools
for our study. The locations of the 100 participating schools are
shown in Table 2.


DESCRIPTION OF PROFILE SHEETS AND DATA ENTRY

     A three-page profile sheet was developed for this project
for on-site characterization of the structure, utility
penetrations, types of heating, ventilating, and air-conditioning
(HVAC) equipment, and other building features pertinent to radon
diagnostics and  mitigation.  Because many schools have several
contiguous structures often constructed at different times and
each with its own unique characteristics, the profile sheet was
completed separately for each structure.  In a few  cases, where
the structures are not contiguous but are campus-style school
complexes, profile sheets were completed  for each distinct
structure in the school, unless all were  of the same vintage and
construction type.

     Where available, building plans were examined  to determine
structure and HVAC system  information that  is not always
available through on-site  observation.  Following inspection of
the building plans, the school was visited  to verify  information
on the  plans and to collect  any  additional  profile  sheet

-------
information that was not on the plans.   Complete sets of
construction plans were available for only 40% of the structures.
When the plans were not available, the profile sheet was
completed based on discussions with school personnel and the
judgement of the researchers.

     Distribution of the profile sheet responses into the
categories used for data analysis required reducing detailed
responses to shorter/ categorical responses for many of the
profile sheet questions.  The original responses for each school
were entered into a DBase IV file along with the shorter
categorical responses used for the statistical analyses.

     Because many of the schools have a number of distinct
structures, it is difficult to generally describe the entire
school for a given characteristic, except rarely when all
structures have the same characteristic.  For example, in a
school with two additions to the original building, two of the
three buildings might be slab-on-grade and the third building a
basement. Each of the individual buildings would be treated
separately on the profile sheet.  Therefore, no attempt is made
to calculate percentage distributions based on the number of
schools in each category.  Instead, distributions are calculated
both in terms of the number of sample structures and in terms of
structure area.

       A sample of the results is discussed in the following
section.  Statistical limitations of the study are contained in
the final section. Detailed results of the complete analysis for
this project will be included in  a final project report.


PRELIMINARY RESULTS OF  SCHOOL BUILDING PROFILE

     The sample of schools selected for this profile are
nationally representative.  However, due to the  small  sample
size,  extrapolation of  estimates  based on the sample statistics
to the national population of schools involves  some degree  of
sampling error.  The standard deviations due to  sampling  errors
for reported percentages range  from 2.5 to  5.5  percentage points
for population estimates of  5%  and 50%, respectively.   The  95%
confidence intervals for these  estimated population percentages
thus range from +/-  5%  points to  +/-  11%.   Due  to  the  large
confidence intervals,  small  differences  (less than 10  percentage
points)  in reported  population  percentages  may  not be  significant
at the commonly used 95% level  of significance.  These
statistical  limitations are  discussed in detail in the next
section.

-------
     The results presented in this paper are the actual
proportions of the school characteristics for the nationally
representative sample of 100 schools.  For most of the
characteristics, the results are presented both in terms of the
percentage of the number of structures and in terms of area.  The
discussion is grouped into structural characteristics and HVAC
system characteristics.

     Where available, comparisons from RMB's 47 research schools
are presented.  Although the RMB research schools do represent a
biased sample in that they are located in radon prone areas,
comparisons of these two samples are helpful in observing trends.

Structural Characteristics

     The schools used for this study typically contain two or
three unique structures.  The distribution of structures by year
of construction is shown in Figure 1.  Nearly half of the school
structures were built between 1950 and 1969, with about 20% built
before and 30% built after.  This distribution is consistent with
the survey conducted by the Education Writer's Association that
found that more than 50% of the schools  in use today were
constructed during the 1950s and 1960s  (1).  By comparison, 46%
of the schools  in our profile were constructed during this
period.

     Over 90% of these school structures have a conventional
classroom design, with a corridor that has classrooms on either
side.  Approximately 5% have a campus-type design, with a number
of individual buildings.

     The distribution  of school structures in terms  of total  area
is shown in Figure  2.  Approximately 45% of  these structures  are
less than  10,000  square  feet*,  probably because many of the older
buildings have  had  additions to the  original building.
Approximately one out  of eight structures  (12.3%)  have more than
50,000  square feet,  ranging to over  600,000  square feet  in  one
school  structure.

     For radon  reduction research, the  substructure of  a  school
is of  interest.   As seen in Figure 3, slab-on-grade substructures
are most prevalent,  accounting  for 72.6% by  structure and 51.6%
by area.   Crawl spaces and basements account for 10.3%  and 6.7%
of the structures,  respectively.   These results are consistent
with RMB's research schools which are 70%  slab-on-
 *  1 square  foot =  0.093 square meter

-------
             Figure 3 also shows that about 10% are combination
             s, such as slab-on-grade and crawl space in the same
 *h« r^o  <- ComParin9 the Percentage by number of structures with
 the percentage by area, there is a tendency for a crawl space to
 be constructed in conjunction with either a slab-on-grade or a
 basement.  These two categories account for approximately 8% of
 the number of structures, but almost 35% of the area.  More than
 two-thirds of all school structures consist of only one floor?

      Location of subslab footings and the presence of subslab
 aggregate are very important in designing a subslab
 depressurization (SSD)  system for radon mitigation.   As seen in
 Figure 4, gravel (which improves the SSD system effectiveness)
 was indicated on the plans for about 45% of the structures with
 information available.   Many of the structures did not indicate
 the subslab material on the plans or the plans were not
 available.  The remaining structures indicated fine-grained
 material (such as sand  or earth)  under the slab.   The location
 SS ?USbSr °* sub?lab footings is also important  in determining
 subslab barriers for SSD systems.   Figure 5 shows that over half
 of the structures have  no internal footings (typically post-and-
 beam construction,  facilitating SSD).   However, 24%  hive footings
 between classrooms and  along the corridor,  complicating a SSD
 system installation.

      Location of utility lines is  also important  since utility
 lines located under the  slab or in a  subslab tunnel  can serve as
 a  major radon entry route.   The data  in Figure 6  show that about
 Hr^oif?    *K® structures  (and area)  have overhead utility lines.
 However,  a third  (a quarter  by area)  also have utility lines  in
 either a tunnel  or  subslab.   Utility  tunnels were  present in  one-
 third of RMB's research  schools  (3),  and  tended to be more
 prevalent in  certain school  districts  than  others.

 HVAC  Svstem Characteristic

     Research  on the use of  HVAC systems  for radon reduction
 includes  a large portion of  both RMB's  and ORP's radon  research
 Jvn^S  S?  m*7^AS \resuit' i1:.is important to quantify the  various
 types  of  HVAC  systems found  in existing U.S. school buildings.

     The  distribution of types of HVAC equipment in the sample
 schools is shown in Table 3.  The categories in this table are
mutually exclusive.  Only one-third of all schools have a  single
type of equipment in all structures.  Most often,  this  is a
                                                ,
 fl         K          Radiant heat only <6*) or fan coils only
(8%) or both (2%) are present in 16% of the surveyed schools,
indicating that the other 84% of schools have either central HVAC
or unit ventilators capable of delivering conditioned outdoor

-------
air.  The remainder of the schools have various combinations of
central HVAC, unit ventilators (UVs),  fan coils (FCs),  and
radiant heat (RAD).  In some schools,  other radiant heat systems
have been abandoned (RAD-NU) for heating, but their presence must
be considered from a radon perspective.

     In Table 4, the distribution of the four basic types of HVAC
systems is tabulated by number of schools, count of structures,
and structural area.  These categories are not mutually
exclusive, due to the occurrence of combinations of HVAC systems
within a school structure.  Central HVAC is the predominant
system, occurring in 71% of the schools and 52% of the
structures, either alone or in combination with other equipment.
Radiant heat, including abandoned systems, is the second most
common system, when counted by schools (56%) or by structures
(44%).  In terms of structural area, radiant heat systems are as
prevalent as central HVAC systems.  Unit ventilators and fan coil
systems (with no ventilation capability) are less common than
central HVAC and radiant heat systems, each occurring in
approximately 30% of all structures and 40% of all schools.

     Considering the combinations of HVAC systems within a given
school, 45% of RMB's research schools have central air handling
systems; 43% have unit ventilators; 30% have radiant heat; and
11% have fan-coil units  (3).  Only radiant heat (11%) or only
fan-coil units  (6%) are present in 17% of the research schools,
indicating that the other 83% have some type of installed HVAC
system that can deliver conditioned outdoor air.

     The school profile sheets contain more detailed information
concerning the  location of  air supply and return ducts, the
location of unit ventilators and  fan coils, and the types of
radiant heating systems.  The most common location of air supply
and return ducts  in structures with central HVAC systems is  in
the ceiling or  suspended  overhead.  However, ducts located  in
corridors, basements, or  tunnels  occur more often  in larger
structures.

     The most common  location of  unit ventilators  and fan coils
is  along the outside  wall.  Radiators  are used  in  most
structures,  but baseboard systems amount  for more  structural
area.


STATISTICAL  LIMITATIONS

     Because of the random  selection  of  NSRS ATD/canister  schools
within the 25 selected residential  PSUs,  the sample  of  profiled
schools  is nationally representative.   However, extrapolation  of

-------
 the survey estimates to the national  population of  schools  must
 reflect the magnitude of sampling errors  expected for a  survey of
 this size.   Sampling errors should also be  considered when  the
 relative proportions of two response  categories are compared.

      Clustering of  the sample  schools within  the residential PSUs
 results in some loss of sampling  efficiency compared to  a truly
 random sample  of schools for this survey.   An additional loss  of
 sampling efficiency arises  due to non-response adjustments  to  the
 sampling weights, which will be made  when the final weights are
 provided by ORP.  At this time, the sampling  weights for the NSRS
 ATD/canister sample have not been determined.

      The loss  in sampling efficiency  can  be explained in terms of
 a  design factor (DF)  for the survey,  defined  as:

                             DF =  »
                                    n

 where N represents  the  actual  sample  size (100)  and n represents
 the reduced effective  sample size for this  design.   The  effective
 sample size is  defined  as the  required size for  a truly  random
 sample to generate  the  same  sampling  errors.   Because of the
 random selection of residential PSUs  and the  random selection  of
 schools within  these PSUs, we  estimate that a  worst case DF would
 not exceed  1.25.  For this assumption, the  effective sample size
 is approximately 80.

      The standard error  (SE) of an  estimated population
 percentage  (P)  is given  by:
                       SE =
                                P ( 1 - P )
                                    n
                       N
                      -=-=,  =  effective sample size
                      DF
     Knowledge of the standard errors of the estimated
percentages permits determination of approximate 95% confidence
intervals (CIs) for the reported estimates.  An estimated
population percentage P has a 95% CI extending approximately two
standard errors on either side of the estimate.  Thus, the
approximate 95% CI for a population percentage estimate P would
be the interval (P - 1.96SE, P + 1.96SE).

     The estimated 95% CIs for various estimated population
percentages are reported in Table 5 for the specified effective
sample size.  The 95% CI for an estimate of 5% extends

                                8

-------
approximately from 0 to 10%.  For a population percentage of 20%,
the 95% CI extends approximately from 10 to 30%.  Similarly, an
estimate of 80% has a 95% CI ranging approximately from 70 to
90%.

     The 95% CIs reported in Table 5 are relatively large, due to
the small effective sample size of approximately 80.  The size of
these CIs should be considered when comparisons are made between
the reported population proportions for two different response
categories.  For example, if outcome "a" is observed in A% of the
100 schools, and outcome Mb" in B% of the schools, then A + B <=
100% with the inequality applying if more than two outcomes are
possible.  To determine if A is significantly less  (or greater)
than B, the standard error of the difference  (A - B) is
determined by:
                 SE( A- B)  =  i  " - " - < A -
                                        n
     To test the hypothesis that A is greater than B  (or A is
less than B), the difference between A and B should be
significantly greater  (less) than 0.  Hence, the quantity  (A - B)
should be more than two standard errors away from 0,  indicating
that the difference is significantly positive  (A greater than B)
or significantly negative  (A less than B).  Regions where A is
significantly greater  (or  less) than B are shown in Table 6.  In
this table, the symbol «  denotes that A  is significantly less
than B, and » denotes that A  is significantly greater than B  (at
the 95% significance level) as determined by:

                    | A - B |  >  1.96 SE( A  - B )

For example, an estimate of 8% is significantly less  than an
estimate of 20%, but it is not significantly less than an
estimate of 16% because the SE of the  difference between
estimates of 8 and  16% is  about 5 percentage points.  Thus the
difference of 8 percentage points is less than 1.96 SE, and the
difference is not considered significant  at the 95% significance
level.
 CONCLUSIONS

      The school profile sheets contain many significant findings
 concerning the distribution of school building characteristics.
 The profile sheets provide evidence of the variety of building

-------
 structures  and HVAC equipment  found  in  typical  schools.   The  age
 of a  school,  number and  size of  different  structures, type  of
 substructure,  location of utility  lines, and  types  of HVAC
 equipment vary widely in the sample  schools.

      The substructure of a school  has important implications  for
 radon diagnostics  and mitigation.   Determination of substructure
 detail depends on locating building  plans, which were available
 for only half  of the structures.   Where identified, subslab
 materials were almost evenly divided between  gravel and fine-
 grained material such as earth or  sand.  Internal footings  are
 found in half  of the structures, with footings  under both
 corridor and classroom walls in  one-quarter of  the structures.
 Utility lines  may enter  the building at a wide  variety of
 locations,  including tunnels, subslab penetrations, and overhead.
 In  a  few schools, older  unused radiant  heating  systems may
 provide additional  radon entry routes.

      Commonly  encountered structural characteristics include
 slab-on-grade  with  a conventional  school building design  with a
 single floor.   Central HVAC is common,  but often combined with
 other HVAC  systems within a single school.  Where applicable,
 central HVAC ductwork is usually located in the ceiling or
 suspended overhead.   Radiant heat, using baseboard or radiator
 systems, is the second most common HVAC system.  Unit ventilators
 and fan coils  also present in many of the schools are most  often
 located along  outside walls,  but may be in the ceiling,  suspended
 overhead, along an inside wall, or on the roof.


REFERENCES

1.   Education Writers Association, Wolves at the Schoolhouse
     Door.  Washington,  DC,   1989.

2.   Quality Educational  Data (QED),  Inc.,  Denver,  CO,  1988.

3.   Leovic, K.W., A.B.  Craig,  and D.B.  Harris,  Update on Radon
     Mitigation Research  in Schools.   Presented at the 1991 AARST
     Conference, Rockville, MD,  October 1991.
                               10

-------
      TABLE 1.
        ASSIGNMENT OF RADON POTENTIAL CATEGORIES
            FOR RESIDENTIAL SURVEY
  EPA
Region
  8
 10
 Radon
 Potential
 Category
 ^—s^^^^—^.

 High
 Medium
 Low
         High
         Medium
         Low
         High
         Medium
         Low
        High
        Medium
        Low
        High
        Medium
        Low
        High
        Medium

        Low
        High
        Medium
        Low
High
Medium
Low
        High
        Medium
        Low
High
Medium
Low
State/Substate Area
ME, NH, VT
MA, CT, RI
None

Northern NJ
NY
Southern NJ
            PA, Western MD, WV, Western VA
            None
            DE, Central and Eastern VA,
            Eastern MD, DC

            Western NC, Western SC,
            Northern GA, Northern AL,
            Eastern TN
            KY, Western and Central TN
            Central and Eastern NC,
            Eastern SC, Southern GA,
            Southern AL, MS, FL
            MN, WI, IL, IN, OH
            None
            MI
            NM
            OK,  Western and Central TX,
            Northern AR
            LA,  Southern AR, Southeastern
            TX
            NE,  IA
            KS,  MO
            None
                   MT, WY, UT, CO, ND, SD
                   None
                   None
            NV
            None
            CA, AZ,  HI
                   AK, ID
                   None
                   WA, OR
 No. of
  PSUs
Selected
   =
    3
    5
    0
                                                 4
                                                 8
                                                 2
                                    15
                                     0
                                     2
                                     3
                                     7
                                   30
                                    0
                                    2
                                    2
                                    0
                                    8
                                    2
                                    0


srn






2
5
3
4
3
0
6
0
0
                            11

-------
TABLE 2.   LOCATION AND CHARACTERISTICS OF PARTICIPATING SCHOOLS
                  IN THE SCHOOL PROFILE SAMPLE
EPA
Region
1
2
3
4
5
6
7
8
9
10
State
Massachusetts
New Jersey
New York
Virginia
West Virginia
Mississippi
Tennessee
Illinois
Ohio
New Mexico
Oklahoma
Texas
Kansas
Nebraska
Utah
Arizona
California
Washington
Total
No. of
Schools
3
5
7
5
5
7
5
4
4
4
5
11
5
4
5
4
13
4
100
Type of Schools*
K-6, K-6, K-6
K-6, 7-12, K-6, K-6, K-8
7-9, K-6, K-6, P-3, K-6,
K-6, K-6
6-8, 6-8, K-6, K-6, K-6
K-6, K-6, 7-12, K-6, 7-12
6-8, P-K, 10-12, 6-8,
K-6, K-6, K-6
6-8, K-8, K-6, K-6, SP-ED
K-8, K-8, K-6, 9-12
SP-ED, 6-8, 7-9, K-6
K-6, K-6, K-6, 6-8
K-6, 7-12, K-6, 6-8, 9-12
9-12, K-6, 6-8, K-6, K-6,
6-8, K-6, K-6, K-8, 9-12,
K-6
9-12, 7-12, K-6, 7-12, K-6
7-12, 6-8, K-6, K-6
K-6, K-6, 6-8, K-6, 9-12
9-12, K-6, K-6, 6-8
K-6, K-6, K-6, 7-9, K-6,
K-6, P-K, K-6, 6-8, K-6,
9-12, K-6, K-6
K-6, K-6, K-6, 6-8

 P - Primary
 SP-ED = Special Education
                              12

-------
     TABLE 3.   DISTRIBUTION OF TYPES OF  HVAC  SYSTEMS  BY SCHOOL


Type of  System;                        No.  Schools
Central  HVAC  only       (HVAC)             13
Unit ventilators only  (UV)                 7
Fan coils only          (FC)                 8
Radiant  heat  only       (RAD)                6


UV/RAD-NU*                                     1
HVAC/UV/RAD-NU                               1
HVAC/UV/FC/RAD                               3
HVAC/UV/FC/RAD-NU                            1
HVAC/UV/RAD                                  16
HVAC/FC                                        7
HVAC/RAD                                      8
HVAC/FC/RAD                                  12
HVAC/UV                                        7
HVAC/FC/RAD-NU                               2
UV/FC                                          1
FC/RAD                                         2
UV/RAD                                         4
HVAC/UV/FC                                    1
      Total Number of  Schools             100
* NU =  not used
  TABLE 4.   TYPE OF HVAC EQUIPMENT BY NUMBER  OF SCHOOLS,  NUMBER
                  OF  STRUCTURES,  AND STRUCTURAL AREA

                                                          Count by
                  Count by            Count by                  Structural Area
Type of System         Schools  Percentage     Structures      Percentage    (square feet)    Percentage
Central HVAC          72      71.3       120        51.5        3643604       67.1
Radiant heat          57      56.4       103        44.2        3659727       67.4
Unit ventilators         43      42.6        70        30.0        1734323       31.9
Fan coils             38      37.6        61         26.2        1883987       34.7

Note:'  Unknown types are not included in analysis.
   k  Parcents add to more than 100% due to the possibility
      or more than one system for a structure.
                                     13

-------
    TABLE 5.   APPROXIMATE 95% CONFIDENCE INTERVALS FOR ESTIMATES
                 Estimated
           Population Percentage
                                                Expected
                                          95%  Confidence Interval1
P * 5%
P = 10%
P - 20%
P - 50%
or P -
or P =
or P =

95%
90%
80%

(P -
(P -
(P -
(P -
4
6
8
11
.8%,
.6%,
.8%,
.2%,
P H
P H
P H
P H
K 4.
K 6.
h 8.
h 11.
8%)
6%)
8%)
2%)
    The actual confidence intervals surrounding the estimates
     will not be symmetric except for the case  P -  50%.
      TABLE 6.  REGIONS OF SIGNIFICANT DIFFERENCE  BETWEEN TWO
             POPULATION PERCENTAGE ESTIMATES, A AND  B.

                                 PERCENTAGE B

        4  8 12 16 20 24 28 32 36 40 44 48 52  56 60 64 68 72  76 80 84 88 92 96
    4
    8
   12
   16
   20
   24
   28
   32
< 36
M 40
O 44
Ł 48
P 52
ft 56
   60
   64
   68
   72
   76
   80
   84
   88
   92
   96
g
 NOTE:   The symbol n»" denotes that percentage A  is  significantly
 greater than percentage B at the 95% significance level;   the
 symbol "«" denotes that A is significantly less  than B;   and the
 symbol "." denotes that A and B are not significantly different
 at the 95% significance level.
                                 14

-------
     100

      00 -

      60 -

      TO -

      60 -

      60

      40 -

      30 -

      20 -

      10 -

       0
                                     23.2X    23.2X
                       I3.3X
 0.4X
I900-O9   I9M-I9   1920-28
                                      1940-49  I95O-59   1960-69  I970-T9  1950-89

                                TEAR CONSTRUCTED
    Figure 1.   Distribution of structures  by year constructed,
i
I
o
i
100

 90

 60

 70

 60


 50

 40

 50

 20

 10

  0
 I5.0X
                     PIX
                          6.4X  6.B*
                                                        Uqfl»O.OB3*4m
                                              0.9X
                                                                 o.9X  0.5%
                                                                   r- - r
           0-10 10-20 2O-30 9O-40 40-50 SO-60 60-70 7O-M 6O-90 90-XXIOO-2OO330-SOO 600t
                                  ARC A. COO •<) It


             Figure  2.   Distribution  of structures  by  area.
                                          15

-------
      too
      80 -
i
           T2.6X
      30
      20
       10
                                                          0.4X     0.8*
                   or«wl     bimt   dab/crawl b»m«/ef»wl  vlcb/pm*  »Ub/»>»mt  b»ml/»1»b
1
                   crawl
b*ml   (lab/craw! b»m«/e*«wl •tab/pUr*  alab/b«ml  temt/alab

    SUBSTRUCTURE TYPE
             Figure  3.   Distribution by  substructure type
         (top %  by  number of structures; bottom %  by  area)
                                       16

-------
      K>0
10
E
I

2
           •and    earth  gravd grvf-a*nd  dirl-ll  prvt-fl  cndy-tol ckider-fl p*a pvl fil-aand
           •and   «a»tt»
                               •UBSLAB MATERIAL
              Figure  4.   Distribution  of subslab material

          (top %  by number  of structures;  bottom  % by  area)
                                       17

-------
2
oc
I
M
                           eterm           corridor
                           SUBSLAB WALL LOCATION
         Figure  5.   Distribution  of subslab wall  locations.
                                    18

-------
too
                                         0.8X   0.5%   0.8X


    crawl evrhd  b*ml  •ubslab evlng  tunnel nweti rm o* w«I *Ub crvdvrhd *w*l m»ch rrfauled
100 -|


60 -


eo -


TO -


60 -


SO -


40 -


3O -


20 -


 10 -
30.3X
    cr«w(  ovrhd  b*m(
                               iiUvbd

                         VTUTY UNE LOCATION
    Figure  6.   Distribution of utility  line locations
    (Top %  by  number  of structures;  bottom  % by area),
                                  19

-------
                                                                   XP-3
                   RADON IN SCHOOLS IN WISCONSIN
           by:   Conrad Weiffenbach and John  C.  Lorenz
                Wisconsin Dep't.  of Health and  Social  Services
                Bureau of Public  Health
                1  W.  Wilson  St.,  P.O.  Box  309
                Madison,   WI 53701-0309
                             ABSTRACT


     Three-fourths of the school districts in the state have made
three month  winter  alpha-track detector measurements.  More than
35,000 radon screening measurements have been made in classrooms in
this  continuing voluntary program.   Results in about  1% of  the
classrooms were higher  than 4 pCi/1.   Follow-up  studies using
continuous radon monitors  show that ventilation  systems reduce
radon concentrations in about half of those rooms to below 4 pCi/1
during the hours they are occupied  (8  AM  -  4 PM) .   Most of  the
other rooms  with elevated screening  levels  are underventilated.
The administration of the program,  the  value of continuous radon
monitor data, and mitigation measures taken and their effectiveness
are discussed.

     This  paper  been  reviewed  in  accordance  with  the  U.S.
Environmental Protection Agency's peer and administrative review
policies and approved for presentation and publication.

-------
                           INTRODUCTION


      The Wisconsin school radon program was established in 1989 by
the  Radiation  Protection  Unit in  the Bureau  of  Public  Health,
working   with   Wisconsin's   Cooperative   Educational   Service
Associations  and  the  Department  of  Public  Instruction.   The
Cooperative Educational Service Associations (CESAs)  serve schools
in  12 regions  into  which  the state is  divided,  offering  (among
other services)  the  coordination  of  environmental  work among
buildings and grounds personnel of participating schools, which can
include private schools.  The  Radiation Protection Unit put out a
request for bids for a large number of alpha track detectors, and
the CESAs coordinated detector purchases by the schools through the
supplier selected by the Radiation Protection Unit.   An engineer
from  the Radiation Protection Unit traveled around the state in the
Fall  of  1989 giving training sessions for  personnel of  schools
participating in the program.
                   SCREENING MEASUREMENT METHOD


     Screening measurements were made by school personnel following
procedures  recommended  by  the U.S.  EPA  (1)  for  alpha  track
detectors.   Many of the  school personnel  also received training
from the Bureau of Public  Health.  The Landauer (Terradex) Radtrack
alpha track  detectors  (ATDs) that were  used  were placed one per
occupied room,  at or below ground level, or one per 2,000 square
feet in  large rooms, for exposures of  three  winter months.   The
detectors were typically  hung from ceiling light fixtures.

     When  the results were received,  they  were entered  into a
relational data  base at  the Wisconsin Geological/Natural History
Survey.  The data base was  transferred  on  magnetic media to the
Bureau of Public  Health.
QUALITY ASSURANCE


Duplicates
     School  personnel  were  instructed  to  place  side-by-side
detector duplicates in ten percent of the rooms as quality control
measures, but  in  analysis of data we  received  from the detector
laboratory,  where two  monitors  are  listed as  having  the  same
location, it is evident that in  many  cases that they were not side
by side.   Pairs such as  (0.3,  4.5),  (0.4,  7.5), and  (0.6,  5.5)

-------
together in data from one school, and (55,  1.2);  (65, 2.5) together
in another school, with the same indicated  locations, indicate that
the information  in the data base  is not  suitable for measurement
quality  assessment  based  on  duplicates.    Multiple  detector
placements are expected in areas (gyms, etc.)  with over 2000 square
feet,  but it  seems  school personnel  made many experiments  and
unnecessary  measurements  with  multiple  detectors  in  rooms.    A
scatter   graph  of   98   measurement  pairs,  that   are  labeled
specifically  "duplicate"   in  the  data  or  known   from  personal
communications to be  duplicates,  is presented in Figure  1.   The
average coefficient of variation (standard deviation  / average) for
those 98 pairs is 0.08.   The reproducability indicated is good.


Control Blanks


     Many  detectors  labeled  as controls  in  the  data  base showed
significant radon exposures.   This is because many of the schools,
following  a procedure outlined in EPA literature for schools (1),
opened  the control blanks at  the  beginning of the  measurement
period and sealed them with the metallized stickers coming with the
detectors.  Unfortunately the stickers are not radon proof.
The  few known properly  handled control blanks  had  results below
detection  limits of 0.3 to 0.4 pCi/L.


Spikes


     Four detectors were sent  for exposures in radon  chambers, with
three  control  blanks, to  The  U.S.  EPA National  Air and Radiation
Environmental Laboratory,  in Montgomery, Alabama. The exposure was
reported  by EPA as 447  +/~ 15  pCi/L - days.  The detectors were
forwarded,  blind with three different names and a false exposure
time of 91 days  (which would  result  in an indicated 4.9 pCi/L +/~
4%),  to  Landauer  Corporation  by  the Bureau  of Public  Health.
Landauer  reported  the readings as  4.7, 4.5, 5.0,  and 4.8 pCi/L,
with less  than 0.4 and 0.3 pCi/L for the  three blanks.
The standard deviations given for these detector readings and for
most of the school measurement results are about 7%  at 4 pCi/L and
11% at 1.5 pCi/L.

     Landauer Radtrack alpha track detectors have regularly passed
EPA  Radon Measurement Proficiency  testing.  With  our duplicate,
blank,  and spike results,  we  believe the  precision, accuracy, and
reliability of these detectors  has  been satisfactory.

-------
 SCHOOL SCREENING RESULTS


      For 31,999 screening measurement results that were in the data
 base as of October 1991,  the average radon  level was 1.0 pCi/L and
 775 were greater than 4.0 pci/L.   These  included a large number of
 elevated results  for detectors placed  in  unoccupied  spaces  for
 examples storage rooms and  tunnels.   The average screening radon
 level in spaces that have been verified to be occupied  is less than
 i pci/L.   Several  thousand  results from winter of  1991-92 were
 received this spring but are not counted in this report.

      Schools with the highest screening results or for which many
 of the  rooms  exceeded  4 pci/L  were given the first  follow-up
 investigations.  Of 477  results  greater than 4.0 pCi/L that had
 been investigated by May  1992,  262 were in unoccupied spaces or
 otherwise invalid,  and 215 were occupied and called  for follow-up
 measurements.                                                    *

      Nineteen screening results were reported at greater  than 20
 pci/L.   Seventeen of these were in unoccupied spaces  or had start
 and stop date data-entry errors.  The highest results  for occupied
 S^SS *l®re:. a  locker room at 36 pci/L which  could not be confirmed
 with Radiation Protection Unit charcoal  canisters, though  it had
 been confirmed by the school with  a second  alpha track detector;
 and a classroom  at  29 pci/L, which   had had ventilation  system
 repairs  before we could  make a  follow-up  measurement.   Both  of
 these had  follow-up  measurement results  below 4  pCi/L  durina
 occupied hours.                                                 y


                  FOLLOW-UP  MEASUREMENT  METHODS



 INTENSIVELY  STUDIED  SCHOOL


      When the first  screening results arrived in  1990,  we selected
 one school with many rooms having elevated screening levels  in the
 4  and 5  pci/L  range  for  intensive study,  to  develop follow-up
 methods  and investigate  mitigation strategies.   The buildings
 supervisor for  this  school district was  very cooperative,  as was
 his replacement when he retired.  This experience was to be applied
 for  follow-up   studies and mitigation for  other schools  in the
 state.

     This school has  been studied and worked on for over two years
with  dozens  of  alpha track,  charcoal,  and  continuous  radon

-------
measurements.  The levels have now been  reduced to average below 4
pCi/L in  all rooms during occupied hours,  though  some are still
close.  We have developed considerable  experience  that has been
useful in  follow-up studies for other schools.

     The  first  follow-up procedure  that  was  studied involved
measuring  with  charcoal canisters  with  ventilation  system  on
continuously  for two  days,  and remeasuring with the ventilation
system  off  for  two  days,  in  an elaboration of  the procedure
suggested  in U.S. EPA  literature on radon measurements  for schools
(1).  Diagnostic grab  samples and charcoal canisters in  tunnels and
other non-classroom locations were used,  as was a continuous radon
monitor.   The continuous radon monitor was  found  to be the most
useful follow-up measurement tool, as  it indicated when, how much,
and how fast levels dropped when the heating/ventilation system
came on in the morning.

     It was  decided that continuous radon  monitors would be the
best way  to  do follow-up measurements if  we could get enough of
them, if they were suitably automated as data  loggers, and if the
data  could be  easily uploaded  for  analysis and  graphing  with
computers.


CONTINUOUS RADON MONITORS
     Ten continuous radon monitors (CRMs), assembled at the Bureau
of Public Health, have been used in follow-up studies in over two-
hundred  school  rooms  during  two winters.    They  are based  on
portable nuclear multichannel analyzers (Canberra Corp., model S-
10), which were  surplussed to  the Bureau of  Public  Health  by
Commonwealth   Edison   Corporation  of   Illinois,   and  diffusion
scintillation  cells, of  5.7  cm inside  diameter  and 12.5 cm long,
with eight 1.9 cm diameter holes circling each end, manufactured by
the Radiation  Protection Unit.

     Quality assurance for the data from these detectors is based
on continuing  efforts of three types.

     a.)  Calibration is done by running all ten CRMs side-by-side
in a high-radon basement along with a commercial Pylon AB5 monitor
and  charcoal  canisters.    The  Pylon CRM  has  a  flow-through
scintillation  cell,  and is calibrated with a  Pylon flow-through
radon source.  It's calibration has been verified with two Femto-
Tech 210F CRMs and an identical Pylon CRM, independantly calibrated
at other institutions.  The results from the  ten  CRMs and the Pylon
are consistent with each  other, with variations of about +/- 10% in
the raw  counts among them, for  which  calibration factors adjust
during data analysis.

-------
      b.)  Background corrections are made by running the monitors
 plus the  flow-through Pylon CRM,  calibrated  for background with
 nitrogen gas,  outdoors.   This introduces an  uncertainty of less
 than 1/4 pCi/L in the overall calibration.

      c.)   Comparisons of CRM  data with  charcoal  canister radon
 measurements are made in every school  room  in which the CRMs are
 deployed:  more than 200 schoolrooms in the past two years.

 „„ ^ Variations of up to 30% from  unity  for the ratio of the CRM
 48-hour average to the 48-hour charcoal canister result are common.
 For nine CRMs, the average ratio for from 7  to 10 classrooms each
 ™ Wlth,ln 10% of unity.  For the tenth it was  1.19.   For all ten
 CRMS and the 76 classroom measurements with the lowest precision
 deviation?116  3Verage rati° is  1'01'  +/~  0.22  (one  standard


      Several factors contribute to the differences between charcoal
 canister results and  CRM 48-hour  averages,  though it  is  seldom
 evident which are significant  in  the  data  for individual  rooms.
 The canister results  are typically  +/- 10% in precision.  The CRMs
 are +/- 10% in calibration.   For open-faced charcoal canisters, the
 desorption of radon is reported (2) to be exponential:  exp(-t/T)
 with a  24-30 hour characteristic time  T,  so  the result can depend
 strongly on the radon level in the final 12  or so hours  of the
 measurement period, while the CRMs give truer averages.  Canisters
 results are  also  susceptible to  temperature changes  and  air
 currents,  while the CRMs  are not affected by these.

      In a typical field trip, three or more days of data from each
 ot  the  ten monitors is uploaded to a laptop computer,  and they are
 transported to different schools and  started  in the next  set of
 rooms needing follow-up measurements.  No operator  or  student has
 damaged a  continuous  radon   monitor  in  over   200   classroom
 placements,  in schools of  all sorts,   all  over the  state.   The
 monitors have  been  operated  by  personnel  from  county  health
 departments,   Milwaukee  Public  Schools,  as well as the  Radiation
 Protection Unit, but school maintenance personnel are not generally
 asked to run  them.                                              *
       FOLLOW-UP RESULTS, BEFORE AND AFTER MITIGATION WORK


     A  summary  of the  results  for  follow-up measurements  is
presented in Table 1.

-------
     In more than half of the 206  occupied rooms for which follow-
up measurements have been made with continuous radon monitors, the
radon levels were found to be below 4 pCi/L during occupied hours.
At  least  20 of these  rooms  had had ventilation  adjustments and
repairs before follow-up testing was done.

     Fifty of the  78  occupied rooms  that had been confirmed with
follow-up measurements  to be over 4 pCi/L  during occupied hours
have been mitigated with work on the heating/ventilation systems,
including balancing,  adjustments, repairs,  and new construction;
bringing radon levels to  below  4  pCi/L during occupied hours, as
indicated in further CRM follow-up measurements.

     Twenty-eight  rooms,  in  7  buildings,  may  still  have radon
levels over  4  pCi/L during occupied hours,  averaged through the
school year.   For some  of  these, the schools  have  not finished
heating/ventilation system adjustments and sealing; for others the
Radiation Protection Unit has not  been able  to return for a second
set of follow-up measurements after  whatever work has been done.
For a few, mitigation via work  on the ventilation system has not
been effective enough, and study of the problem is continuing.
TYPICAL DATA FOR CLASSROOMS
     Figure 2 shows radon versus time in a classroom in which the
radon  is  greatly lowered during occupied hours  by the automatic
setback heating/ventilation system,  'in  CRM data  graphs, each data
point has been centered on the hour mark between the middle and end
of  the hour collection  interval, depending  on when the  CRM was
started.  In the evening and over night, the heat and ventilation
is  automatically set back to  be minimal,  and radon levels climb.
In the morning, the heat  and ventilation come on, driving down the
radon  levels.   Radon  in most  rooms in this building  followed a
similar  pattern,  and  many school  buildings  exhibit a  similar
pattern, though without going as high over night.

       A correction, for the decay and buildup  of radon progeny in
the diffusion scintillation cell, which is largest for rapid
changes with time like this example, is also shown in the figure.
This correction, involving  adding various  fractions  of the radon
counts for several  previous  hours to the count for a given hour, is
derived from  empirical studies of the  response  of the diffusion
scintillation cell  to step function increases and  decrease of radon
in a chamber.  When the  changes of  radon  concentration with time
are less rapid, the correction is not important.

     The radon in  the  classroom of  figure  2  averages less than 2
pCi/L during occupied hours  8 AM to 4 PM, though the full three day

-------
average  is  close to 10 pCi/L.   This  school had done repairs and
balancing for the ventilation system before the CRMs were placed in
it.  The alpha track screening measurement result for this room was
29 pCi/L.

     Figure 3 is for a classroom in which the major source of radon
was a sub-slab return air duct,  in contact with the ground.  When
the  furnace return air fan  is  on the  duct is below atmospheric
pressure.  The radon in this classroom was reduced by installing a
ceiling  duct  for the  return  air.

     Figure 4  is for a classroom for which one of the return air
paths is through a grill in  the  door to  the hall, while another
path is  through return air ducting from  the  room.  The door was
kept closed,  then open,  on alternate days.   Because the hall is
quite negative  and the door grill is the major impedance between
the room and hall, when the door is  open,  the  pressure in the room
is more negative, and the  radon  is more elevated than when the door
is kept closed.  The door  of this room was usually kept open.  Many
schools  in the state use halls  in return air  circuits.
   CHARACTERISTICS OF BUILDINGS WITH ELEVATED FOLLOW-UP LEVELS


INTENSIVELY-STUDIED SCHOOL


     The  building has a  system  of return air  tunnels  which had
radon  levels  of  about  7.5  pCi/L,  2 to  3 pCi/L  higher  than the
classrooms.  Since air is  recirculated to the furnace through these
tunnels and then distributed to all rooms, they were believed to be
important sources of radon  for the rooms.

     Among the methods used or attempted to lower the radon were:
increasing ventilation in  some rooms,  starting the heat/ventilation
cycle earlier  in day; balancing  the  ventilation system; blending
more fresh  air into the  recirculated mix  in  the coldest season,
when the dampers had been closed maximally; and a moderate amount
of sealing of openings to  soil in the  tunnels and associated return
air system.  Work that was done over  one  summer  to try to seal the
tunnels  against radon  entry had  little or  no impact  on  radon
levels.

     It was found that spring and fall radon values in the building
were lower than winter values.  This is to be expected, because the
heating/ventilation system automatically blends more fresh air into
the system when the outdoor temperature is warmer.  Averaging radon
levels over the academic year was necessary to achieve levels below
4 pCi/L for some of the rooms.

-------
     At present  all  rooms in the  intensively-studied  school are
below the 4 pCi/L guideline during  occupied hours, though some are
near  4.    The heating  bill  for  the  school  has probably  been
marginally increased.

     The characteristics of the school appear comparable to those
reported  (3) by  U.S.  EPA researchers for a  Colorado  school  with
return air tunnels which was difficult to mitigate.


OTHER SCHOOLS


     About a third of all occupied rooms with elevated screening
levels had been  converted  to offices  or small  classrooms  from
underventilated  locker   rooms  or  storage areas not  originally
designed to be occupied, or were actual locker rooms.

     Rooms occupied for  less than  a cumulative hour per day were
considered unoccupied if the radon was just a few pCi/L over 4.0.
A locker room occupied for less than fifteen minutes each hour, for
changing clothes at the beginning and end of each of four periods
per day, is classed as not occupied.

     Two  of the school  buildings with  elevated or  borderline
follow-up results have been  closed  in the normal  course of events,
not because of  radon.  Administrators of two schools with confirmed
elevated levels would like to move or replace  the  buildings, mainly
for reasons not concerning radon, but cannot get the communities to
support these plans.

     Forty-three occupied rooms with elevated screening levels are
in  school  systems  which have  refused  the  offer  of  follow-up
measurements.  Some have not returned repeated telephone contacts,
and others have simply declared they don't want the involvement of
the state.

     Schools generally have been reluctant to call  the Radiation
Protection Unit for follow-up measurements. Some have made 9-month
follow-up  measurements  with  alpha  track  detectors,  using  a
procedure  suggested  in  the  U.S.   EPA  literature  (1)  on  radon
measurements in  schools.  With these they often obtained results
similar  to the  3-month screenings.   In  general the  Radiation
Protection  Unit  contacts schools   after receiving  data  showing
elevated screening levels,  and explains about  the  CRM follow-up
measurements that are available.

  Most  schools  appreciate   Bureau  of   Public   Health  follow-up
assistance when  it is explained that it is at no charge; that the
continuous radon monitors often  show reduced radon  levels during
occupied hours; that  the Radiation  Protection Unit has accumulated

-------
experience that can help them deal with elevated levels if they are
confirmed; and that action to reduce levels is not compulsory.


                           CONCLUSIONS


     The voluntary measurement program  has given schools a common
and inexpensive screening procedure.  Although each year additional
schools make the  radon screening tests,  it doesn't seem that all
schools will have tested for radon until after testing is required.

     CRM  follow-up measurements  reveal significant reductions of
radon levels during occupied times in many schools with automatic
setback HVAC systems, and produce useful diagnostic information in
some cases.

     Quite  a  few schools have made  minor and otherwise required
adjustments to ventilation systems that reduced radon levels.  For
some  of the schools  with moderately  elevated  radon  levels, the
common mitigation strategies are unsatisfactory, and the analysis
of the problem  and search for a  means  of radon  control continues.
                       ACKNOWLEDGEMENTS


     The authors would like  to  thank John Micka,  Larry McDonnell,
Teri  Vierima,  Mike Mudrey,  and Tom Shepro,  who established the
school  radon  program,   and  all  the  school  officials  who have
graciously permitted measurements to be made in their buildings.
                            REFERENCES


 1.   U.S.  EPA.   Radon measurements in schools:  an  interim report.
 EPA 520/1-89-010.   March,  1989.

 2.   George,  A.C.,  and Weber, T.  An  improved  passive  activated C
 collector for  measuring environmental 222Rn in  indoor air.  Health
 Physics 58:  583-589,  1990.

 3.   Leovic, K.W.,  Pyle, B.E.,  Borak,  T., and Saum,  D.W.   HVAC
 complications  and control for radon reduction in school buildings.
 Presented at the 1991 International Symposium  on  Radon and Radon
 Reduction Technology,  Philadelphia,  April, 1991.

-------
TABLE 1.  RESULTS SUMMARY FOR SCREENING AND FOLLOW-UP MEASUREMENTS
Alpha track screening measurements, by Oct.  1991        31,999


Alpha track screening results > 4.0 pCi/L                 775
     No. of these investigated by May 1992:               477
          Were in unoccupied tunnels., etc.               262
          Were in occupied school rooms                   215


Rooms with charcoal follow-up measurements only             9
     (all were < 4.0 with vent. syst. on continuously)

Rooms with CRM follow-up measurements:                    206
     Initial results < 4.0 between 8 AM and 4 PM*         128
     Initial results > 4.0 between 8 AM and 4 PM           78
          Mitigated to < 4.0, 8 AM to 4 PM                 50
          Not yet mitigated to < 4,.0, 8 AM to 4 PM         28


"Some were mitigated with balancing  or repair of  ventilation
systems before follow-up measurements were made.

-------
  7-


  6-
O
Q.
                    a
                      a
               a
                     n
                  n I mi i
                               a
                     345
                       pCi/L, 1STATD
 Figure  1.   Scatter  plot of alpha  track duplicate results.
  20-
  18-

  16-

  14-

>^ 12-
o
 „- 10-
    12
        n

                         a

                                           ffl
24       12      24      12
         TIME (HOUR OF DAY)
             a UNCORRECTED
                              CORRECTED
                                          24
 Figure  2.    Effect  of ventilation system, and correction  for
 progeny delays.

-------
   35




   30




   25-




|  20-


z
oc
   15-
   10-
    5-
                      n
                      a
                    a
                       a

                        a
                  D
                                a


                                a
                                 a a

                               a  a
                                                        a


                                                       a
                                                a


                                              cP
                                                ODD,
12      24       12       24

                  TIME (HOUR OF DAY)
                                     1111111111111111111111111111
                                      12       24       12
 Figure 3.   Radon in  room with sub-slab return air duct,
  2-
                    a
                    D
      a el-,
       f~l uj

      qSp a   a


n   &   rP     a
                            a     off
                           a,   a n
                           tJ Q   CfclD
                           Q  '-' ndP o
                                                  a

                                                  a
                              a
                                                a

                                                n
                  on
                   n
                               a
    DOOR CLOSED OPEN  CLOSED   OPEN    (SAT.)  (SUN.)
  1-
  0-Uw,
   12
      12
                  12
                  iiiiininiiiiinmniiininniniiiiiniiiiininiiiiiiiiiniiinniii	in
                   12     12      12      12      12
                      TIME (HOUR OF DAY)


Figure  4.   Door effect,  return  air partially  through hall,

-------
                                                                   XP-4
            INVESTIGATION OF FOUNDATION CONSTRUCTION
             DETAILS TO FACILITATE SUBSLAB PRESSURE
               FIELD EXTENSION IN LARGE BUILDINGS

               by:   Mike Clarkin
                    Camroden Associates,  Inc.
                    RD #1,  Box 222,  East  Carter Road
                    Oriskany,  NY  13424

                    Kelly W. Leovic
                    U.S. Environmental Protection Agency
                    Air and Energy Engineering Research Laboratory
                    Research Triangle Park,  NC  27711

                    Fred McKnight
                    The H.L. Turner Group
                    Harrison,  ME  04040

                            ABSTRACT

     Pressure field extension (PFE) measurements were conducted in
four recently constructed Maine buildings to determine the effect
of foundation design on radon control in new buildings.  The goal
was to evaluate the impact of subslab materials on the strength and
extent of the subslab pressure field.  Three  schools  and one office
building, each incorporating different  foundation  designs,  were
studied.

     Results support the  following  EPA guidelines for installing
subslab depressurization systems in new schools and large buildings
constructed in radon prone  areas:
     •    install a layer of clean, coarse gravel beneath the slab;
     •    limit barriers to subslab communication or plan features
          that will extend the pressure field across unavoidable
          subslab  barriers;
     •    install  a large  suction pit and radon vent pipe; and
     •    if  elevated  radon  levels  are measured,  activate  the
          system with a fan.

     The  results   also  demonstrate  the   need  for  observing
installation of radon prevention systems before the slab is poured
in order to ascertain system design parameters.   Radon prevention
systems  in  three of the  four  buildings  were much  less effective
than  anticipated since clean, coarse gravel  had not been placed
under  the slab.

     If  gravel is not used, a  subslab  pipe network will help to
extend the subslab pressure field.  However,  the PFE developed by
a pipe network (without gravel)  is not as evenly distributed as PFE
under  a gravel bed.  The builder must also consider the additional
expense  of purchasing and installing  the pipe network.

     This  paper has  been reviewed in accordance with the U. S.
Environmental Protection Agency's peer  and administrative  review
policies and approved for presentation  and  publication.

-------
                           INTRODUCTION

     The U.S.  Environmental Protection Agency's  (EPA's)  Air and
Energy Engineering Research Laboratory (AEERL)  began research on
indoor radon in residential buildings in the mid-1980's.  In 1987,
AEERL  initiated research  on  developing  and  demonstrating radon
reduction  techniques that could  be  used  in  existing and  new
schools. AEERL has also conducted limited  research in existing and
new large buildings.

     One of  the most frequently  used and effective  methods for
radon  reduction   is  subslab   depressurization.     A  subslab
depressurization system maintains the soil gas beneath  the slab at
a lower pressure than the  air in  the building.   As a result, air
moves from the  building  into the soil rather than from the soil
into the building,  effectively keeping radon from  entering the
building by pressure driven transport.  If a fan is used to create
the negative pressure under the  slab,  the system is referred to as
"active."  A system without a fan is called "passive."  An active
system  is  much more effective  in  achieving  and maintaining  a
consistent negative pressure under the slab.

     In radon prone areas,  AEERL  recommends  installing a subslab
depressurization system during  building construction  in order to
facilitate radon mitigation if needed  (1).   The objective of this
project  was to determine  the  impact of  certain  construction
techniques   on  the  effectiveness  of  subslab  depressurization
systems.   Addressing these construction  techniques  during  the
design and  construction  stages will  result  in a  more effective
radon control system  at  a  lower cost  than installing a retrofit
system after construction (2).   Indoor  radon measurements were not
part of this investigation.

     Four recently constructed Maine  buildings  were s .lected for
this study.   Investigation  of each building included the following
steps:

     (1)  Review the  blueprints  and construction  documents.   The
     radon   mitigation system   design  was  determined from  the
     drawings,  and served as a road map for the next step.

     (2)  Thoroughly inspect the building.  Assistance of someone
     familiar with the building was critical  during this step.

     (3)  Ask maintenance personnel and  the builder for information
     concerning footings,  slab openings,  subslab material,  and
     footing drains.

     (4)  Determine the airflow and pressure characteristics of the
     subslab radon control  system using an in-line centrifugal fan
     to draw different  airflows.    The resulting  pressure  field
     extension (PFE)  across the slab was mapped.

-------
                             RESULTS
     Three schools and one office building, each located in Maine,
were selected for this study.  The major characteristics of these
buildings are summarized in Table 1, and each building is discussed
separately.

          TABLE 1.  SUMMARY OF BUILDING CHARACTERISTICS
Building
School A
School B
School C
Building D
Subs lab Aggregate
25-60% pass 0.25
inch* sieve
25-60% pass 0.25
inch sieve
0.50 to 0.75 inch
in diameter
25-60% pass 0.25
inch sieve
Radon Control System
1 perforated pipe,
1 suction point
perforated pipe
perforated pipe
perforated pipe;
footing bridge
SCHOOL A

System Description

     This  is  a  two  story slab-on-grade building.  The floor plan
consists  of a  large  rectangle and  three  pods.    The  school was
constructed  with two  separate passive  subslab depressurization
systems.  System 1 has four risers exiting through the roof,  and is
connected  to  an interior footing drain loop.  This system  covers
the  rectangle and two of the pods.   System  2  is  a single point
passive  depressurization system in  the  third pod  with a gravel-
filled  well connected to  a  roof  exhaust.   An  interior footing
between the two systems effectively separates the subslab area into
two  zones.   Figure  1 schematically  illustrates the layout  of the
school and the  two  radon control systems.

     The perimeter  loop system is connected to a storm drain in two
places.   If air is  drawn from the storm drain into the  perimeter
loop, it would be difficult to produce a low pressure field beneath
the  slab.  The  plans do not indicate use of water traps or reverse
flow values to  separate  the  storm drain  from the perimeter  loop.

     Subslab  aggregate specifications for the school required a 12
inch  layer  of  base  material,  with  95%  compaction.    Size
specifications  for  the subslab aggregate are in  Table  2.
 *  1 inch =2.54 centimeters

-------
        TABLE 2.   SCHOOL A SUBSLAB AGGREGATE SPECIFICATIONS
Sieve Size
3 inch
1 inch
0.5 inch
0.25 inch
# 40
# 200
Percent Passing
100
80-100
35-75
25-60
0-25
0-5
 The subslab  aggregate  specifications allow for  a  large percentage
 of  stones less that 0.25  inch in diameter.   Observation of the
 subslab  material  during the  investigation revealed that a  large
 percentage  of the aggregate  was,  in fact 0.25 inch or  smaller.
 This fine-grained material can  impede subslab PFE (1).

 Procedures

      The System 1 perimeter  loop was constructed of 4 inch poly
 vinyl chloride (PVC)  perforated pipe wrapped in geotextile wrap and
 buried in the subslab aggregate. This loop was connected  to solid,
 Schedule 20  PVC pipe risers.  This system was designed to provide
 passive  depressurization to a subslab area of about  48,000 square
 feet.*   PFE  measurements were made to determine the strength and
 distance of the  pressure  field, from the perimeter loop.   The
 objectives of investigation of this system were to determine if the
 perimeter loop was helping to  extend the subslab pressure  field and
 if  the interior footing was restricting PFE.

      The researchers searched for  the perimeter loop/scorm drain
 connection to  determine if there  were  large air leaks  into the
 subslab drainage layer. Unfortunately, the connection was not near
 any access holes.  Building maintenance personnel said the storm
 drain emptied  into a pond located on the  school  grounds, not to
 daylight.  Subsequent airflow  measurements did not indicate any air
 leaks into the system.

      According to available building plans, System 2  began beneath
 the  slab in  a  pit filled with 0.5  cubic yard** of crushed stone.
 Schedule 20 PVC pipe was used  for the exhaust  riser in this system.
 The  plans did  not specify the size of the  stone used in the pit,
 but  the builders' standard specifications for stone used  for this
 purpose is angular crushed  natural  stone, free from shale, organic
matter and debris; and  0.375 to  0.75 inch in diameter. This single
point system was  intended to  provide  depressurization beneath an
 area  of approximately 8800 square feet.
*  1 square foot = 0.093 square meter
** 1 cubic yard = 0.765 cubic meter

-------
     The locations for the subslab PFE measurements were selected
based on opportunity rather than systematically.  Nearly all floor
surfaces in the  school were covered with vinyl tile,  and  school
officials were hesitant to allow many  slab perforations.   As  a
result, a compromise between scientific evaluation  and suitable
drilling  locations  was  made.   Locations  chosen  for the  slab
penetrations included  under floor-mounted doorstops  and  beneath
carpets.  Figure  1  illustrates approximate  positions of all test
points.

     Measurements on the two systems included exhaust pipe airflows
and subslab-to-indoor pressure differences.  The measurements were
made in the passive mode (without a fan, as designed) and with an
in-line centrifugal  fan at two  airflow rates.  The  other  stacks
were capped during  the active tests to prevent short-circuiting.
The fan used is rated at 80 cfm* at 0 inch water column (WC).**

Results

     Figure 2 shows  subslab pressure versus distance measurements
for System  i  — the perimeter loop system — using  the  fan for
depressurization.   The subslab pressure field did not reach test
point  3,  which  was  28  feet  from  the suction point.   It  was not
possible to drill  holes in the  immediate vicinity of the footing
(which  was  approximately  18  feet  from  the  suction  point)  to
determine if the footing was restricting the pressure field.  As an
alternative, a test hole was drilled in another location,  at the
same  distance  from  the perimeter  loop  as the  interior  footing
(approximately 18  feet).   No pressure field was detected at that
test point.  These measurements indicate that the subslab aggregate
was restricting PFE, so that  it was not possible to determine the
effects of the footing.

     Figure  3  shows the results  of the  subslab  pressure versus
distance  measurements  for  System  2  — the  pit  system.   These
results also confirm that there  is fine-grained material under the
slab.  A static pressure of -1 inch WC in the suction pit does not
extend a pressure  field beyond 10 feet.  It is suspected, but not
confirmed,  that  the  exhaust  riser  did  not  rest  in a  pit as
specified,  but  terminated in the subslab  aggregate.   This would
result in a suction's  being applied to only 12.6 square inches* of
*  1 cubic foot per  minute  (cfm) =0.47  liter per second
** 1 inch water column  =  249 pascals

-------
aggregate, rather than the surface area of a pit. The importance of
the area of aggregate exposed to the suction has been documented in
previous research (2) .   This situation illustrates that inspection
of the  system installation by a qualified  radon professional is
essential during building construction.

     We had hoped to test whether the interior footing interferes
with PFE.   Unfortunately, a direct measurement of this  was not
possible since the pressure field did not extend to the footing.

     Both systems were also tested in the passive  mode; however, no
airflow or  negative  pressure was measured under the  slab.   As a
result,  if  this  system is to  be used  to consistently  exert a
negative subslab pressure, it must be operated in an active mode.


SCHOOL B

System Description

     This   school,   still  under   construction   during   this
investigation,  has  a  rather  extensive  subslab  pipe  network
connected  to  a  single exhaust  riser.    The  pipe  network  is
constructed of 4 inch diameter perforated PVC pipe.  The diameter
and number of holes per  linear foot** of pipe was  not documented by
the builder.  Figure  4 illustrates the pipe network in  a portion of
this school.

     At the time of the investigation, the riser was terminated 2
feet above the slab surface,   conversations with the site manager
revealed that there  was no intention  to run  the  riser outdoors
unless  indoor radon  measurements indicated that a radon control
system  was  required.   At the  time of the  investigation  these
measurements  had not  been made;  however, subslab  measurements
indicated a radon concentration of 1,200 pCi/L***  beneath the slab.
The  building  architect was informed  of  these  levels  and  was
considering running the  riser through the roof during construction.

     Subslab  aggregate  specifications  were  not  available,  but
conversations  with  the building  project manager  indicated  that
crushed stone  similar to  the  aggregate in School A was used.   A
polyethylene barrier was placed on  top  of  the crushed stone, a 2
inch layer of sand placed  above the  aggregate,  and the slab poured
on the sand.   Sand was  probably used to reduce the chance of the
slab's cupping as the result of unequal  drying  between the top and
bottom.  However, the sand interfered dramatically with the subslab
*   1 square inch = 6.45 square centimeters
**  1 foot = 0.305 meter
*** 1 picocurie per liter (pCi/L) = 37 becquerels per cubic meter
     (Bq/m3)

-------
PFE measurements.  In many cases, the sand would fill in around the
hole as soon as the drill bit was removed.  It took several tries
at most holes to take a pressure reading in the subslab aggregate
rather than in the sand layer.

Procedures

     An in-line centrifugal fan (rated at 80 cfm at 0 inch WC) was
placed on  the  end of the riser.   Airflow out of  the  system and
pressure differences were  measured at  the riser and  at  various
distances from the riser.

Results

     Subslab-to-indoor pressure differences are shown in Figure 5.
It was expected that  the strongest negative pressure readings would
be found near the subslab pipes.  The predicted pressure pattern is
not clear in Figure 5.  The  strength of the subslab pressure rises
and falls  in a cyclic way,  but the pipe pattern  would  have to be
different  than was shown on the building plans.    There  is a low
pressure field at almost all tested points,   with  a mean  of -0.1
inch WC.  This  would  be expected because the pipe network should be
exerting a near  uniform pressure  at  all points.   Some  of the
variation  in the measurements may  be an artifact of  the sand's
interfering with  the pressure  taps.   For example,  the  relatively
low negative pressure at 30 feet appears like an outlier for both
airflows.
SCHOOL C

System Description

     The radon control system at this school was being installed in
an addition to the  original building.   The foundation,  slab, and
walls were in at  the time of the measurements,  but the roof was not
on.  While this  might have some effect on the test measurements,
the  subslab  PFE  was  strong  enough and  extensive enough  to  be
conclusive.  The  slab is a 6700 square foot rectangle.   A 6 inch
perforated pipe  runs under the center  of the  slab through the
drainage layer.   The  drainage  layer is composed  of stone pebbles
that are 0.5 to 0.75 inch in diameter.   The frost walls are poured
concrete, and there are very few slab penetrations. Figure 6 shows
a floor plan of the building and the location  of  the test points.

-------
Procedures

     Conversations  with  the  site  manager  revealed  that  the
perforated pipe ran straight  across the width of the slab, exiting
through a footing.  It then snaked its way underground, beneath an
adjacent slab, and finally exited to daylight.  Assured by the site
manager that we had correctly identified both ends of the pipe, an
in-line centrifugal fan was  attached to one end  of  the pipe,  and
the other end was  carefully sealed with duct tape.  Test holes were
drilled through the slab, and subslab pressures were measured and
recorded.

Results

     Figure 7 shows the subslab pressure profile for this school.
Pressures at  test points beyond  10  feet measured -0.012 (±0.005)
inch WC beneath the slab. Test Point 1, located directly above the
perforated  pipe,   measured   -0.016  (±0.002)   inch  WC.    Subslab
differential pressure measurements were not made across the width
of the building because building materials were piled on one side
of  the  slab.    The  subslab flow  and pressure  characteristics
indicate a fairly  tight foundation, low leaks, and highly permeable
fill.  The pressure field does not drop even within 2 feet of the
slab  edge,  and there is  only a  25%  decrease in pressure field
strength at all other  test holes as compared to the pressure at the
pipe.  This indicates that a  layer of clean, coarse gravel beneath
the slab provides a more uniform pressure field,  compared to the
fields found  in the other buildings described in this paper.

     This type of  system, with a perforated pipe in the middle of
a layer of aggregate, behaves like an air plenum.  It is expected
that pressure down the length of the pipe would be fairly uniform
because  the  resistance  to airflow  of the perforations  is large
compared  to  the  resistance  to  airflow in  the  pipe.    This
distributes the low pressure pattern along a line  down the center.
Airflow through the aggregate bed will come largely from the slab
edges  that  are  parallel  to  the  pipe.    The  air  will  move
perpendicular to  the direction of the pipe.  If the stem wall and
slab  are fairly tight  (as  in this case), the resistance of the
cracks and holes through them will be much greater than through the
aggregate bed.

     This  can be  modeled using  design aids for  thermal  storage
stone pebble beds.  Applying the design aid to this  problem, it is
expected  that drawing 196  cfm from the  pipe  would result  in  a
differential pressure that is 0.002 inch WC less 30 feet from the
center pipe than  1 foot from the pipe.

     On the other hand,  if  196 cfm was drawn through a system with
a single suction  point, this type of uniform pressure field would
not be expected unless the leaks were so small that very low air-
flows were drawn  from the system.  This is because  the resistance
to  airflow through the  aggregate  increases  as the speed of air

                                8

-------
moving through the aggregate increases.  As air moves through the
aggregate bed  to  the central suction  point,  the cross-sectional
area of bed it travels through becomes increasingly smaller.   The
air speed must increae in order to conserve mass so the resulting
pressure  drop with  distance  from a  central  suction point  is
expected  to  be nonlinear  and greater  than for  the  plenum-type
system achieved with the central pipe system.


BUILDING D

System Description

     This building was designed to facilitate radon reduction, if
needed, using a subslab depressurization system.   Subslab aggregate
specifications are the same as those for the School A.  In addition
to the subslab aggregate,  a perimeter drain  loop  pipe network with
two passive risers was installed.  In order to facilitate subslab
PFE,  interior footings  were not  used  except   in  a  small  area
surrounding  a vault.   To extend  the  pressure  field  across the
footing,  the top  of the  footing  was recessed  4 inches  in the
doorway, and  a 4  inch layer  of  gravel  was placed in this recess.
The floor slab was then poured.  The building floor plan is shown
in Figure 8.   The  purpose of testing in Building  D was to determine
the effect of the recessed footing on PFE.  Measurements were not
made to assess the effectiveness of the drain tile system.


Procedures

     The PFE measurements were performed using two suction points,
as shown in Figures  9 and  10.  Suction was applied first to point
SP1  (Figure   9)  and  then  to SP2  (Figure  10) .    Pressures  were
measured under the slab on both sides of the opening through the
footing in order  to  determine the effect on PFE.

Results

     The baseline differential pressure under the slab was neutral,
indicating that  the  passive  system  was not  exerting  a negative
pressure  under the slab.  However, these measurements were made
during  the summer  with little  temperature differential between
indoors and outdoors.

     The opening  in  the footing  does appear to facilitate subslab
PFE.  A pressure  of - 32 inch WC at SP1 produced a pressure of -
0.240 inch at TP7 which is 6  feet away and on the same  side of the
footing  (Figure 9).   Pressure at TP3,  5 feet from SP1 but on the
other side of the footing,  was -0.02 inch WC.  Although this is not
a very strong negative pressure, it is apparent  that the negative
pressure  does reach across  the opening in  the  footing.   When
suction is applied at SP2 (Figure 10), a slight  negative subslab
pressure  is  also  observed  across the footing.

-------
     We suspect that there is some obstruction between TP3 and TP4
which is causing the quick drop in the pressure field between these
two points that are only 1 foot apart.  This drop in pressure would
not be caused by the footing because a large drop in pressure also
occurs when suction is applied  at SP2 which is on the same side of
the footing (Figure 10).

     Maintaining  a  continuous  layer  of aggregate  under a  door
opening  (which is  otherwise  surrounded by  footings)  should  be
researched in  a building  with high permeability aggregate.   The
aggregate produces  a very uniform subslab  pressure (as  seen  in
School C  results),  and any pressure drop across a  footing  could
then  be  attributed  to   the   footing  rather  than  the  subslab
aggregate.


                           CONCLUSIONS

     The  PFE  measurements in  these four Maine  buildings support
AEERL recommendations for radon prevention in the construction of
schools  and other  large  buildings  (1) .    The  recommendations,
together  with  the  supporting  data  from  this  research,  are
summarized as follows:

1) install a 4 to  6  inch layer  of subslab aggregate  (ASTM #5) - In
the four buildings researched for this project,  the building with
the best PFE was the building with clean, coarse subslab aggregate
(School C).   The  three other building had subslab  material with
more fines (see Table l); thus, the pressure field did not extend
as far or was as strong.

2) Avoid barriers to subslab communication - Although openings in
the  footings  helped  to  bridge subslab  barriers,   the   PFE  was
somewhat  reduced  (Building D).   Research  of  this construction
technique  should  be  pursued   in a  building  with  clean,  coarse
subslab aggregate.

3) Install a  suction pit  and radon vent pipe -  Previous research
(2) has shown that a suction pit will  improve  PFE by  increasing the
surface area of the applied suction.   The data from School A help
to confirm the  importance of a suction pit.

4) Attach a suction  fan to the vent pipe to activate  if needed - In
the two buildings that were designed to operate passively (School
A and  Building D),  measurements showed that subslab PFE was not
achieved  without  use  of  a suction  fan.   However,  these  two
buildings  did  not  have  clean,  coarse  subslab  aggregate,  and
measurements were made during the summer with minimal temperature
differential.

This study demonstrated   the need  for observing  installation of
radon  prevention  systems  before the  slab is poured in  order to
ascertain  system  design   parameters.   This  study  also indicated
that, in areas where gravel is  not readily available or expensive,
use  of a subslab pipe network  will  help to extend the subslab

                                10

-------
pressure field (Schools A and B) .   However, the PFE developed by a
pipe network is not as evenly distributed as PFE through a coarse,
crushed  aggregate  bed.    The  builder  must also  consider  the
additional expense of installing the pipe network.


                           REFERENCES

1.   Leovic, K.W., A.B. Craig, and D.B.  Harris, Radon Prevention in
     the  Design  and Construction of  Schools  and  Other  Large
     Buildings, Architecture/JResearch,  October 1991,  1:1, pp. 32-
     33.

2.   Craig, A.B.,  K.W.  Leovic,  and  D.  B. Harris, Design  of New
     Schools and  Other Large Buildings  Which Are  Radon Resistant
     and Easy to  Mitigate.   Presented  at the Fifth International
     Symposium  on The Natural  Radiation  Environment,  Salzburg,
     Austria, September 1991.
                                11

-------
 Connect to
 storm drain
                                         System 1
                                         Passive perimeter
                                         drainage loop
                                      System 2
                                      single point
                                      passive system
   Library with
   load-bearing
   walls
Figure 1.  School   A   subslab   depressurization   system
            layout   (not  to  scale).   SP denotes suction
            points;  TP  denotes test points.
                         12

-------
                                       TP2
                    TP3
Perimeter loop

pipe
 LU
 E
 _
 LU
 a.

 v.
 :'
 LU
 DC
 -
System 1 at 61 cfm


System 1 at 31 cfm
           -0.5
                                  10         15        20


                                        DISTANCE (ft)
               25

                 Figure 2. School A System  1  subslab pressure
                            characteristics.  TP denotes test points.
                                    13

-------
                Shz
_
E
         -0.2
         -0.4
         -0.6
         -0.8
                    TP5      TP6
                              10        15        20

                                     DISTANCE (ft)
TP4
     30
          Figure 3. School  A System 2 subslab pressure
                    characteristics. SP denotes suction
                    point;  TP denotes test  points.
                                 14

-------
       10ft-
6
   \
      Suction
      point
                 All test points
                 along this line
                     \
         Figure  4. School  B subslab pipe network  layout.
                                   15

-------
_
_

~-
.
_
r
          -0.2
          -0.4
          -0.6
          -0.8
           -1
                               It
          Airflow a\ 50 cfm


         Airflow at 26 cfm
, i	





 15       20


 DISTANCE (ft)
                                                                  :-
                                                                           -
     Figure  5.   School  B subslab pressure characteristics,
                                     16

-------
        42ft'
                       T
                        2ft
                     TP4
                        20ft




                     TP3 X


                        10ft
                        10ft
                        T
                This end capped
         4 in. drainage pipe

                   X
                     TP1  X
Fan installed  on this
end
10ft
 i
                     TP5 X
  Figure  6.  School  C test point locations.
             TP denotes test points.
                          17

-------
                  TP5      TP1    TP2    TP3
TP4
_
f
* '
I I 6.

o ~
-0.004 ~
-0.008 ~
-0.012 "
-0.016 ~

-
•


•




\
\
t




Subslab drainage pipe

\

-0.02 H
20 10

/





















•
*
•
•
0 10 20 30 40 50
DISTANCE (ft)
           Figure 7. School C subslab pressure characteristics,
                     TP denotes test points.
                                18

-------
                            Vault room with
                            footings. See Figure
                            9 for details.
Figure 8.   Building D floor plan  (not to  scale)
                                19

-------
1 1
1 1
0 1


SP2
I I 1 1 1
1 1 1 1 1
23456

TP4 TP3
-0.004 -0.02
V V

SCALE (ft)
III 1 1 1 1 1 1
7 8 9 10 11 12 13 14 15 16 17

TP2 TP1 SP1 TP7
-0.348 -1.23 -32 -0.240
S^ NOTE: All SP and TP
 Vault
 room
                                           measurements are in
                                           inches WC.
Footing
                                      TP6
                                     -0.167
      Figure 9.  Building D subslab pressure field
                 characteristics with  suction at SPl.
                 SP  denotes suction points; TP denotes
                 test points.
                            20

-------
                              SCALE (ft)
                  •1	1	h
                                 8   9   10  11   12  13  14 15   16   17
  SP2
  -49
      TP4  TP3
     -0.254 -0.026
                     TP2
TP1
-0.005
                                            SP1
 TP7
-0.003
                                        -x—1
     TP8 TP9
     -2.57 -0.362
 Vault
 room
Footing
           x
        TP5
        -0.046
                                NOTE: All SP and TP
                                measurements are in
                                inches WC. No measurement
                                at TP2 and TP6.
                                            TP6
 Figure 10.
Building D  subslab pressure field
characteristics with suction at  SP2. SP
denotes suction points; TP denotes test
points.
                               21

-------
                                                                        XP-5
Title:  Radon Measurements in the Workplace

Author:  David Grumm, U. S. EPA, Office of Radiation Programs

      This paper was not received in time to be included in the preprints, and
the abstract was not available. Please check your registration packet for a
complete copy of the paper.

-------
                                                                 XP-6


             RADON SURVEY OF OREGON PUBLIC SCHOOLS
               by:  George L. Toombs & Ray D. Paris
                    Oregon Health Division
                    Suite 705
                    800 NE Oregon #21
                    Portland, Oregon  97232
                            ABSTRACT

     To assist in addressing the concerns of potential elevated
radon levels in public schools nationwide, the Oregon Health
Division conducted a limited study to determine the average
radon levels in the schools in Oregon.

     Thirty-one schools were selected at random out of a
population of 1,190 statewide to participate in this study
during the 1990-1991 school year.  Long-term alpha track
detectors were placed in each of the ground-floor and basement
classrooms to obtain the average radon levels this entire nine-
month school year.

     The results of this study showed that the mean radon con-
centrations in Oregon public schools statewide was 1.1 pCi/lit-
er.  This compares to a mean of 1.2 pCi/liter for indoor radon
in homes in the state.  Two of the schools surveyed had many
rooms above the 4 pCi/liter EPA guidelines.

     Follow-up with these schools and their options for lower-
ing the levels in the elevated rooms are discussed.

     This paper has been reviewed in accordance with the U.S.
Environmental Protection Agency's peer and administrative
review policies and approved for presentation and publication.

-------
       OREGON  HEALTH  DIVISION  PUBLIC  SCHOOL RADON  PROJECT
               by:  George L. Toombs & Ray D. Paris
                    Oregon Health Division
                    Suite 705
                    800 NE Oregon #21
                    Portland, Oregon  97232
HISTORY
     Indoor radon surveillance in Oregon was initiated in 1984
by the Bonneville Power Administration (BPA) as part of their
regional Residential Conservation Program covering the five
Northwestern States.  Their measurements were essentially
conducted over a 3-month period, mainly during the winter
months.  It should be noted that their service districts were
the only areas covered in this study.  As of January 1, 1991, a
total of 12,871 homes had been tested.  The indoor radon value
for these homes had an arithmetic mean of 1.2 picocuries per
liter  (pCi/1) with about 4 % of the homes exceeding the
Environmental Protection Agency's (EPA) guideline value of 4
pCi/1.

     The Division initiated the Oregon Radon Project (ORP) in
March of 1988 to obtain needed information covering a 12-month
period.  The results of this 12-month project gave an
arithmetic mean for indoor radon at 1.4 pCi/1.  This compared
very closely with BPA's value of 1.2 pCi/1.  The conclusion was
that indoor radon problems in Oregon are small when compared to
other regions of the country.

     Although in Oregon, indoor radon problems are relatively
small, areas would have to be defined wh^re it could pose
problems to large numbers of people.  This prompted the
Division to develop a project to evaluate potential radon
concerns in Oregon public schools.

PUBLIC SCHOOL PROJECT

     Elevated radon concentrations in schools have been a
growing public health concern nationally.  This is because high
levels have been found in a large number of schools around the
country and children are likely to be more vulnerable to the
effects of radon.  Therefore, many states, including Oregon,
are involved in projects to assess the indoor radon situation
in schools.

     Oregon has 1190 public schools statewide.  It was beyond
the scope of this project to test every school.  Therefore, the

-------
goal of the Division's Public School Radon Study was to conduct
a limited study to provide an estimate of the average radon
level in Oregon school buildings.  Thirty-one schools were
randomly selected to be tested.

     The schools selected to participate in this study were
chosen from the Oregon School Directory 1989-1990 edition.
Each was given a number from 1 to 1190.  A random number table
was then used to determine which schools would be contacted and
asked to participate in the study.  The 31 schools that chose
to participate were locatqd statewide in 20 of the 36 counties
and are listed in Table 1.  The distribution of these
participating schools followed the population density very
closely.  The sizes ranged from a three-room country school in
the Willamette valley to a 78-room complex in the southern part
of the state.  A monitoring device was placed in every ground-
floor and basement classroom in each school.

     Alpha-track detectors were used exclusively in this study.
The design of the device, durability, ease of installation,
reporting format options of the vendor and cost were the
factors considered in selecting the vendor by using a
competitive-bid process.  The selected vendor also had to be
listed in the National Radon Measurement Proficiency Report of
EPA dated January 1990.  Radiation Safety Systems, Inc. (RSSI)
in Morton Grove, Illinois, was awarded the bid.

     The alpha-track monitoring devices received from RSSI were
delivered to each of the schools by Health Division staff.  The
purpose of the study was explained to the personnel along with
instructions on how to install the devices.  They returned to
the Division the data sheets, the numbered shipping foil
packets, and the floor plan showing the rooms where the devices
were installed.  This provided for accountability and the
assurance that the detectors were actually installed.  The
Division also inspected the installations in about 20% of the
schools after the devices were installed as part of the overall
quality assurance plan.

     Each of the 31 schools was notified, as a reminder, in
March 1991 that the devices would be picked up in May.  The
Division also provided additional informational brochures about
radon in the home and schools at this time.

     The monitoring devices were picked up by the Health
Division during the last week of May and first week of June
1991.  Any device that had the security seal broken or had
visible damage was not processed.

     Quality assurance is a vital part of any study.  The
objective is to insure the data obtained is valid.  The guality

-------
 assurance plan for this study consisted of administrative
 controls,  placement techniques,  documentation, supervision  and
 followup.   It included sending the vendor "spiked" monitoring
 devices exposed to known levels  of radon.  Blank and duplicate
 devices were also returned to the vendor for processing.
 Fictitious names and addresses of schools were used for these
 controls so the vendor could not distinguish them from the  real
 ones.   A total of 789 devices were used in this study.  Only 30
 of  the  installed devices were unable to be processed because
 they were either damaged or missing.  This is about a 96%
 success rate,  which is very high when compared to school survey
 projects of other states.

 RESULTS

     Figure 1 shows the radon concentration in pCi/1 in
 relation to the percentage of classrooms tested.   The
 arithmetic mean for the indoor radon levels in all rooms of the
 31  schools tested was 1.1  pCi/1.   The summary of results per
 pCi/1 for  each school tested,  are listed in Table 2.  This
 compares very favorably with indoor radon levels found in homes
 as  shown by the BPA and Oregon Radon Project data listed below.

                     Mean Radon      Measurement     Number of
                   Value in pCi/1       Period      Measurements

 Bonneville Power data           1.2            3 months            12 871 (homes)
 Oregon Radon Project           1.4           12 months            l' 562 (homes)
 Oregon Public Schools           l.l            9 ^nths       31 schools (689 rooms)
     Only two schools out  of  the  31  tested had any classrooms
greater than the  EPA guideline  of 4  pCi/1.   Not all rooms in
those schools exceeded the 4  pCi/1 value.   One school had 14
rooms out of 39 greater than  that value,  and the other school
had 12 out of 35  rooms exceeding  the guideline.   The highest
value observed in a classroom was 9.6 p^i/1.   When the
concentrations observed in all  rooms were  taken into
consideration in  each of these  two schools,  neither had an
average indoor radon value greater than the  4 pCi/1 guideline.


     The Health Division subsequently worked with these two
schools which had classrooms  with radon levels greater than the
EPA guideline of  4 pCi/1.   Specifically, we  wanted to define
the radon sources and the  options available  to reduce the
levels.   This required additional short-term testing as well as
continuous monitoring in one  of the  schools  to evaluate the
source and to observe how  the heating/ventilation/air
conditioning system (HVAC)   affected  the radon levels in the
classrooms.

-------
Elevated levels were observed in the utility tunnels under one
of these schools.  The rooms directly above these tunnels
generally had higher levels.  Continuous monitoring of the
classroom with the highest concentration (9.6 pCi/liter)  showed
a strong correlation between radon levels and the mode of
operation of the HVAC system.  Figure 2 shows the radon
concentrations in relation to the various operation modes of
the HVAC system.

     This school, using the above data, is planning on making
changes in ventilation and HVAC operation to minimize radon
levels in the classrooms.  Also, they are considering
appropriate mitigation techniques to lower the radon
concentrations in the tunnels to further reduce the levels in
the classrooms.

     This paper has been reviewed in accordance with the U.S.
Environmental Protection Agency's peer and administrative
review policies and approved for presentation and publication.

-------
                TABLE 1.   PARTICIPATING SCHOOLS
NAME OF SCHOOL
SURVEYED
Applegate Elem.
Ardenwald Elem.
Awbrey Park Elem.
Burns Union High
Butte Falls High
Colton Elem.
Crooked River Elem.
Dickey Prairie Elem.
Elgin Jr. High
Fairplay Elem.
Ferndale Elem.
Franklin High
Glide Jr. High
Grant Union High
Hanby Jr. High
Irish Bend Elem.
Judson Middle School
Lapine Elem.
Mabel Rush Elem.
Madras Jr. High
Molalla Primary
Mt.Vernon Elem.
North Medford High
Park Place Elem.
Parkdale Elem.
Prospect Elem.
Riley Creek Elem.
Rocky Heights Elem.
Sand Ridge Elem.
Scappoose High
Territorial Elem.
COUNTY
Lane
Clackamas
Lane
Harney
Jackson
Clackamas
Crook
Clackamas
Union
Benton
Umatilla
Multnomah
Douglas
Grant
Jackson
Benton
Marion
Deschutes
Y.amhill
Jefferson
Clackamas
Grant
Jackson
Clackamas
HoodRiver
Jackson
Curry
Umatilla
Linn
Columbia
Lane
GRADES
K-5
K-6
K-5
9-12
9-12
K-6
K-5
K-8
7-8
K-5
K-6
9-12
7-8
9-12
7-8
K-8
7-8
K-6
K-4
7-8
K-5
K-6
9-12
K-6
K-5
K-8
K-8
K-6
K-8
9-12
K-4
NO. OF
STUDENTS
126
295
397
340
63
489
595
93
180
186
241
1,325
146
235
224
23
862
479
559
373
446
109
1,374
334
278
139
304
481
72
535
127
NO. OF
ROOMS
SURVEYED
17
16
31
30
7
28
28
7
35
10
20
28
13
14
6
3
42
27
26
23
24
11
74
10
18
19
21
28
7
39
6
Total

Total Oregon Public School
Enrollment 1990-1991 (K-12):
             11,430
668
495,922

-------
              TABLE 2.   PUBLIC SCHOOL RADON PROJECT






       SUMMARY OF RESULTS IN  Picocuries per  liter  (pCi/1)
School
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
No. of Rooms
Tested
17
16
31
30
7
28
28
7
35
10
20
28
13
14
6
3
42
27
26
23
24
11
74
10
18
19
21
28
7
39
6
No. of Rooms
Over 4 . 0
0
0
0
0
0
0
0
0
12
0
0
0
0
0
0
1
0
0
0
0
0
0
0
0
0
0
0
0
0
14
0
MAXIMUM
3. 1
2.3
0.9
2.3
1.4
3.5
0.8
1.3
8.7
2. 1
1.1
0.8
1.8
2.8
2.6
5.0
2. 1
1.3
2.7
0.7
0.5
0.8
1.3
0.4
1.9
2.9
4.0
1.2
1.2
9.6
3.0
MINIMUM
0.8
0.5
0.2
0.2
0.7
1.2
0.2
0.2
0.4
0.4
0.3
0.2
0.8
0.4
0.5
0.4
0.3
0.1
0.1
0.2
0.2
0.2
0.2
0.2
0.3
0.5
0.4
0.4
0.3
1.2
0.3
Mean ± S.D.
1.6±0.7
1.6±0.6
0.7±0.2
0.910.2
1.010.3
1.910.7
0.510.2
0.910.4
3.411.7
1.110.8
0.710.2
0.410.2
1.210.3
1.010.7
1.210.9
2.012.6
0.710.4
0.410.2
1.110.6
0.4+0.1
0.310. 1
0.610.2
0.410. 1
0.310.1
1.010.4
1.510.7
0.810.8
0.610.2
0.710.3
4.012.0
1.511.1
TOTALS     668           27    MEAN  2.412.1   0.410.3     1.110.8

-------
                  FIGURE 1.
     OREGON  HEALTH DIVISION
PUBLIC  SCHOOL RADON  PROJECT
30
Q
LU
CO 25~
LU
CO
S 20-
o
O
cc
o
LL 10-
O
h-
LU 5-
0
CC
LU
Q.
n —

















^>
















           Percent of classrooms at various concentrations based
          upon 9 consecutive month measurements (Sep '90 - Jun '91)
                      n = 668
                  MEAN ± S.D. = 1.1 ± 0.8
                     MAX = 9.6
                    4.2% > 4pCI/l
                    1.9% > 5pCI/l
                     0% > 10pCI/l
 12345678
      RADON CONCENTRATION IN  pCi/L
10

-------
                         FIGURE 2.
Effect of various modes of HVAC system operation on
       radon concentrations in school classroom
                         ROOM D1
                             Off
Minimum
Outside
 Air
Exhauct Off
Supply On
Exhaust On |
Supply Off  •
                     4            1
                     VENTILATION CYCLES

-------
         Session XI Posters
Radon Prevention in New Construction

-------
                                                                  XIP-1
               MODEL STANDARDS  AND  TECHNIQUES  FOR
               CQMTROL OF RADOM IM  NEW BDILPIMflfl

                               by

                         David N. Murane
                       USEPA Radon Division

                             ABSTRACT

It is anticipated that  EPA's "Model standards and Techniques for
Control of Radon in  Nev Buildings" vill be published in final
fora in the Spring of  1992.  This tvo part paper will provide:
(1) An update on ths key provisions of the final "Standard", and
(2) Details of an outreach  program designed to inform, educate
and gain support of  State and local building officials for
adoption of the Standard in National and local building codes.

-------
                                                                  XIP-2
               COMBINED VENTILATION AND ASD SYSTEM
                    by   David Saum
                         Infiltec
                         Falls Church, VA 22041

                         Fred Sickels
                         NJ Department of Environmental Protection
                         Princeton, NJ 08625
                             ABSTRACT


     Radon mitigation in new construction is generally accomplished
by installing an active subslab depressurization  (ASD) system, but
most U.S. houses  have no  mechanical ventilation  system that will
assist  them in  meeting  the  ventilation recommendations  of the
ASHRAE  62-1989 standard.   This paper describes the performance of
a single  fan system  for houses that was designed to provide both
radon mitigation and  increased ventilation. Systems were installed
in three Maryland hous,es with radon levels between 4 and 20 pCi/L.
Performance monitoring included continuous radon and carbon dioxide
measurements.

-------
                                                                                   XIP-3
     EVALUATION  OF PASSIVE  STACK MITIGATION  IN  40 NEW HOUSES
                  By: Michael Nuess
                  Washington State  Energy Office
                  N.  1212 Washington St., #106
                  Spokane,  WA 99210-2401
                                  ABSTRACT
The Passive Stack Study is a study of the performance of passive stacks (PS) in 40 new
residential homes in Washington State. It is anticipated that the study period will be about 18
months. The total study cost will be about $200,000. It is anticipated that the Washington State
Department of Health will manage the project with support from the Washington State Energy
Office-Energy Extension  Service,  U.S. EPA, and the Bonneville Power Administration. The
study homes will be located in the  same geographical/climatic area (Spokane, Washington) in
order to:

    • reduce weather-driven variables
    • maximize the potential for appropriate candidate houses:
           • houses with  indoor radon levels encompassing the range of levels occurring in the
               state and the region
           •houses with code-required PS installed
           • houses in both low-medium and high permeability soils
           • houses with both winter and summer climate conditions
    • minimize study COSt

The study will:

    • measure and compare the effect of PS — installed according to the new Washington State
        Ventilation and  Indoor Air Quality Code (V&IAQ Code) -- on reduction of indoor
        radon during both heating season and summer conditions in the study homes
    • measure and compare the effect on indoor radon of the code-allowed partial substructure
        sealing of the houses versus complete below grade sealing
    • sufficientiy characterize the  study houses in order to:
           •enhance interpretation of the data obtained
           •enable design of the detailed follow-up investigations

Homes will be selected into a four component matrix of 40 houses. This matrix structure allows
division of the study houses into two principal categories of soil type, and two principal
categories of substructure type. Each of the four components will consist of 10 houses:

    High Perm Soil        Med Perm Soil         High Perm Soil        Med Perm Soil
    Basement             Slab on Grade          Basement             Slab on Grade

-------
Each study house will be evaluated in order to obtain information about other important
variables:

               • site factors (terrain, shielding, large surface pavements, etc.)
               • building factors (area, volume, height, geometry)
               • substructure details (gravel, pressure field extension characteristics, geometry)
               • degree of substructure sealing (visual inspection)
               • house depressurization potentials (characterization of mechanical systems
                    pressure impacts and use patterns)
               • passive stack installation details

The follow-up investigations will intensively investigate selected study houses and subsets of the
study houses in order to:

     • explain and substantiate the results obtained
     • investigate the effects of several parameters on the PS performance:
           • desirable fan capacity for activated stacks
           • pressure-difference impacts of different residential mechanical systems
           • sub slab membrane location and effectiveness
           • alterations of pipe/gravel interface
           • etc.

The study will hopefully recommend enhanced system design parameters and modifications to
code requirements related to PS systems.

-------
                                                                                               XIP-4
                      RADON REMEDIATION AND LIFE SAFETY CODE*
                                 by:  Lyle Sheneman, Architect
                                     Chem-Nuclear Geotech, Inc.
                                     Grand Junction Projects Office
                                     Grand Junction, Colorado 81503
                                          ABSTRACT

The preface of the Uniform Building Code states that the code "is dedicated to the development of better
building construction and greater safety to the public..."  It accomplishes its mission through enactment of
uniform building laws that are based on the performance and use of building materials and systems.
Professionals in the building trades are schooled in the proper use of the code and are committed to
its principles.

The purpose of this paper is to address the concerns that radon mitigation systems are, at times, designed
and installed without apparent regard to life safety systems, as required by the Uniform Building Code.
Remedial action contractors need a heightened awareness of the necessity to either be aware of the code
or to enlist the assistance of professionals in the design and installation of these systems.
                                  SCHOOL INSTALLATIONS

One of the major areas of focus for radon mitigation has been the nation's public school buildings.
Because schools are occupied by children, building code officials have subjected these buildings to
vigorous analysis of life safety concerns. Occupants of a school have the expectation that if the building
is threatened by fire or smoke, corridors are available as safe exits from the building. The corridors are
designed to resist involvement in the fire or entrainment of smoke for a period of 1 hour to allow the
occupants ample time to exit the building. All walls, ceilings, floors, and doors are constructed to be part
of this 1-hour system of protection.
   *Work performed under the auspices of the U.S. Department of Energy, DOE Contract
No. DE-AC04-86ID12584.

-------
Because large unobstructed areas are required for instruction, a school is often constructed with long-
span, open-web steel joists that commonly bear on corridor walls.  This type of construction results in a
grade beam being installed under those walls, dividing the sub-slab area into small "noncommunicating"
areas. Consequently, a radon mitigator attempting to depressurize the area below the floor slab in the
corridor must place the point of entry through the slab in the corridor and extend a vent stack through the
roof deck to evacuate the radon (see Figure 1). As this vent stack is extended through the ceiling finish, it
penetrates the protective barrier and violates its integrity, leaving a point of entry for both fire and smoke.
A chase is placed around the vent stack to finish and protect it, but this chase must be constructed with
materials that equal the construction of the corridor as outlined by  the code. The radon vent stack and
chase must  also be placed in the corridor in such a way to not obstruct the egress of persons in the
corridor (see Figure 2). The corridor width has been established by the occupant load of the building and
must not be reduced by the width of the chase.
                                  ROOFING

                                  INSULATION

                                  ROOF  DECK
VENT  STACK
WITH  CAP
                                                                   FLASHING
                                  SOIL
                               Figure 1. Radon vent stack section.

-------
                  CORRIDOR
                                                   CLASSROOM
                                                     PREFERRED CHASE
                                                     LOCATION
                                                   CLASSROOM
                                      CHASE  NOT PERMITTED
                                      IN  THIS  LOCATION
                                      1'-0"  +-  CHASE  WIDTH
                                      REQUIRED CORRIDOR
                                      WIDTH BASED UPON
                                      OCCUPANT LOADS
                           Figure 2. Radon vent stack location.
                        GENERAL BUILDING CONSTRUCTION

Material selection of exterior walls of all buildings, including schools, is governed by the code and
depends on the building's proximity to the property line, the proximity of adjacent buildings, and the
building usage. Openings in or penetrations through an exterior wall are not permitted when that wall is
within a specified distance to a property line or within the allowed distance to an adjacent building.  If a
penetration is required, it must be "protected" as defined by the code.  In many cases, a radon mitigator
will place a penetration through the wall, extend the radon vent stack through the opening, and terminate
the stack at rooftop height to vent the radon at a safe distance from the building's occupants. While this
seems to be an acceptable solution, the penetration through the wall must be accomplished in accordance
with the code.

-------
In the initial building survey for radon mitigation, all systems need to be considered. Underlying the
programmatic design criteria of a building are its systems requirements, such as ingress/egress, life
safety, structural, electrical, plumbing, and heating, ventilating, and air conditioning (HVAC).
Once these systems are understood, an evaluation is possible of the relationship between them and the
proposed radon vent system.

In many instances, the space above the ceiling finish materials and below the structural members is
commonly used as a chase for mechanical, electrical, and plumbing systems. It may also be used as a
return air plenum for the HVAC system. In such a case, all material placed in this space must meet code-
specified criteria. Exposed polyvinyl chloride (PVC) pipe, which is normally used in a radon vent
system, may not meet those criteria because of its flame spread and smoke emission characteristics.  As a
result, the radon vent chase may need to be extended through this space to provide protection to both the
space and the vent system without altering the proper function of any system.

Commercial roofing systems are normally covered by a bond or warranty. A qualified technician must
alter such a roofing system to not void the bond or warranty. Although this is not a code-covered item, it
is an important issue to the property owner.

Many localities have adopted energy codes that govern the specifications  and installation of a building's
insulation system. The workmanship employed in the installation of the radon vent system can be
critical to the proper performance of the insulation system. Disruption of the thermal barrier could
result in increased operating costs and heating/cooling loads on the HVAC system  and could compromise
the comfort level of the building's occupants.


                               RESIDENTIAL INSTALLATIONS

Noise and vibration from the operation of a fan are ongoing problems in all  active radon ventilation
systems.  Many manufacturers have developed acceptable components that, when installed correctly,
either eliminate or reduce this annoyance to acceptable levels. However,  some mitigators have developed
unique and creative ways of dealing with this problem.

 One such solution is to cut a hole in the bottom of a 5-gallon plastic bucket and place it in the attic of the
 subject property. The radon vent stack is then extended through the bucket  and the bucket is filled with
 concrete. While this would no doubt stop any vibration from extending down the vent stack, it introduces
 a few problems of its own. A 5-gallon bucket with a 4-inch-diameter pipe extended through it will
 contain slightly more than 0.5 cubic foot of concrete.  Concrete weighs 155 pounds per cubic foot  If this
 bucket of concrete were to sit directly on the gypsum board ceiling, the result could be immediately
 dramatic. If the ceiling did not fail at the time the concrete is placed in the bucket, it would, over a  period
 of time, begin to show signs of deformation because of the imposed load. While roof trusses are
 designed to be primarily top chord load-bearing members, they are capable of carrying short-term point
 loads on their bottom chords. However, similar to the gypsum board ceiling, they will show signs of
 deformation when subjected to loads imposed on their bottom chords for long periods of time.


                                    RECOMMENDATIONS

 While it is idealistic to think that all individuals involved in the alteration of buildings are well versed in
 the life safety codes, it is imperative that  all alterations be done in strict accordance with the codes. To
 accomplish this, appropriate building officials should review the documents outlining the system
 installation.  In addition, a professional architect or engineer should review and concur with the proposed

-------
system design and components.  Each system should be documented by a detailed set of working
drawings showing the system components and techniques employed in their installation.  This set of
working drawings would be necessary for obtaining any required building permits, would provide the
vehicle for the professional review, and would serve as a valuable resource for any future warranty work
that the system may require.

The work described in this paper was not funded by the U.S. Environmental Protection Agency and
therefore the contents do not necessarily reflect the views of the Agency and no official endorsement
should be inferred.

-------
                                                                    XIP-5
                   A PASSIVE STACK SYSTEM STUDY
             By: Geoffrey Hughes and Katherine Coleman
                 State of Washington Department of Health
                 Office of Toxic Substances LD-11
                 Airdustrial Center Bldg.  #4
                 P.  0. Box 47825
                 Olympia, WA 98504-7825
                             ABSTRACT
     Washington's interim radon-resistive construction standards
require the installation of passive stack systems in new homes in
eight counties.   The Washington State Department of Health and the
Washington State Energy Office are collaborating on a study on the
effectiveness of these passive stack systems.  The research
entails the selection and characterization of 40 study  houses,
followed by a series of radon tests.  Each house will be tested in
both summer and winter in a series of four two-week periods of
alternating open and closed stack conditions.  A subset of houses
will also be tested before and after the sealing of below-grade
penetrations.  This paper summarizes the passive stack system
research methodology and,provides an up-to-date progress report of
the study.

-------
Session XII Posters
  Radon in Water

-------
                                                                                XIIP-1
            RADON IN WATER MEASUREMENTS USING A COLLECTOR - BUBBLER

                 Robert E. Dansereau and Joseph A. Hutchinson*
                      New York State Department of Health
                 Laboratory of Inorganic and Nuclear Chemistry
                     P.O.  Box  509,  Albany,  N.Y.  12201-0509.

                                *Current  address
                      Lockheed Engineering  and  Science  Co.
                            975 Kelly Johnson Drive
                              Las Vegas,  NV 89119


                                   ABSTRACT


      A  Collector-Bubbler  (C-B)  device was developed  to allow  precise and
sensitive  field or  laboratory measurements of  222Radon  (Rn)  in  water.   The
device, which is used in  conjunction with  Lucas  cell  (LC)  counting, is rugged,
inexpensive and easy to use.   After a measured quantity of water  is collected
directly in the C-B,  no liquid transfer  is required.   Rn is immediately purged
into a LC, which can be measured in the field.   Rn decay and loss during sample
transport and transfer are eliminated.


                                  INTRODUCTION


      In recent years  a great  deal of concern has  focused  on indoor radon as a
serious health problem.   Although the major  source of  indoor Rn is usually the
underlying soil and rock,  drinking water is also a source, and in some cases, the
predominant one.   In order  to evaluate water supplies as a Rn source, studies
measuring  waterborne   Rn  concentrations  have  been  conducted by  the  U.S.
Environmental Protection Agency (1,2,3) and States  (4).  These measurements will
increase dramatically  if the EPA's recently proposed Drinking Water Regulations
(5) are implemented.  The  Country's 68,000  public water supplies will have to be
measured quarterly during the  first year and until they  comply with the  300 pCi/L
maximum contaminant level (MCL) and annually  thereafter.  Analytical techniques
must be capable of  measuring  this level  within a  ±30% uncertainty.   Although
these regulations do not  apply to the estimated 13 million private wells, its
expected that the new  regulation may result  in many of them being  tested.

      The two techniques  commonly used and proposed in the new regulations for
measuring Rn  in water  are  the Lucas cell (LC)  and liquid scintillation (LS)
methods (6).  Both measurements are usually performed in the laboratory since a
liquid scintillation spectrometer is required for the LS method and the  LC method
requires transfers using fragile equipment.  Sample transport and preparation for
counting allows Rn decay and possible release from  the sample.   This reduces the
measurement sensitivity and could give inaccurately low concentrations.


                                  DESCRIPTION


      The C-B device  (Fig 1)  utilizes a 20-cc  graduated nylon syringe with a
hollow plunger.  A one-cm diameter medium-porosity fritted disc is fitted into
a machined hole at the inner end of the plunger.  Luer stopcocks are attached to
both the  tip  of the syringe   (outlet)  and to  the outside end of  the plunger
(inlet).  A tube containing 2 g of activated charcoal is fitted inside the hollow
plunger, and is connected to  the  inlet stopcock.

-------
                                  PROCEDURES


COLLECTION


      The sampling procedures used in the EPA recommended methods (6) are used
for sample collections  with  the C-B.  For  laboratory  measurement  a bottle is
filled in the field following the LC  collection procedure and is shipped to the
laboratory.  For field measurements follow the LS collection procedure. With the
outlet stopcock open  and the inlet  stopcock  closed hold  the  C-B vertically,
immerse the  tip (outlet) below  the  water  surface  and draw a  measured 15-ml
aliquot of water into the C-B.  Close the outlet stopcock.


RADON TRANSFER


Position the C-B with the outlet stopcock pointing upward.  An initial purge of
Rn is accomplished by opening the inlet stopcock and withdrawing the plunger 5
cc, causing air to pass  through the charcoal tube and bubble through the sample.
(The function of the charcoal tube is discussed in the testing section).  This
also creates the necessary air space above the water to allow for sample bubbling
(Rn transfer) .   The inlet stopcock is closed and the  clamp  is secured  to prevent
plunger travel. The C-B is connected  to a transfer system (Fig 2) which consists
of a drying column attached to a fine metering valve.   An evacuated Lucas cell
is  fitted above the  valve.   (A  hand-operated  vacuum  pump is  used  for  field
measurements) .   Open both the Lucas cell and the C-B outlet stopcocks.  Start the
bubbling action (Rn transfer) by opening  the metering valve two turns.  The C-B
inlet stopcock is then opened and the metering valve  is  adjusted to maintain the
desired bubbling rate.  The Lucas cell is  closed immediately after the bubbling
ceases.


MEASUREMENT AND CALCULATION


      Allow four hours for secular equilibrium to be established with the short
lived Rn daughters.  Following the manufactures directions  count the Lucas cell
for 20-100 minutes.  Divide the net sample count-rate by the  counting efficiency,
decay factor and sample volume to obtain a value for the Rn concentration.


SENSITIVITY


      The  sensitivity of measurements  using the  C-B  is  dependent upon the
performance  of  the  counting  equipment.   LC counters usually have  an absolute
efficiency of  75%  and relative  efficiency  of 225%  (2.25 cpm/dpm for 222Rn in
equilibrium with 218Po and 214Po) .  Using  a  cell with a  background of  0.5 cpm, a
lower limit of  detection of 0.16 pCi can be achieved for a  20 minute count.  The
minimum detectable  concentration for a  15-mL sample  would be 10 pCi/L.   The
measurement uncertainty at 300 pCi/L, the proposed MCL  for  drinking water, is ±
10%.  This is well below the 30% uncertainty limit in the proposed drinking water
regulations.  This  sensitivity  is determined  for field measurements where the
time for Rn decay is the four-hour daughter ingrowth period.

-------
S TANDARCIZATION.


   226 standardizati°n of this LC/C-B method is performed by purging the Rn from
a    Radium  solution obtained  from the  National Institute of  Standards  and
Technology.   The solution is transferred  to  a  40-mL vial, which has a Teflon
lined cap.   The  vial  is completely  filled  (no voids).   After  a Rn ingrowth
period, a  15 mL  aliquot  of the  standard  is  drawn  into  the  C-B and measured
following the measurement  procedures section.  Since the  Rn is quantitatively
transferred from the C-B to the  LC, the counting efficiency for  this measurement
is given by:

                  E  =    S-B
                       AID
where:
                  E - efficiency  (cpm/dpm)
                  S = standard  count  rate  (cpm)
                  B = background  count  rate  (cpm)
                  A = activity  of standard  (dpm)
                  I = Rn  ingrowth factor
                  D = Rn  decay  factor
                                    TESTING
REMOVAL OF Rn FROM PURGE AIR
      Depending on field conditions, the ambient air used to purge Rn from the
sample into the Lucas cell could contain an appreciable quantity of radon, thus
resulting in an erroneously high Rn in water measurement.

      The effectiveness of the C-B's internal charcoal  in removing Rn from the
purge air was tested by drawing air containing 270 pCi/L  of Rn through the device
and into a Lucas Cell.  The C-B did not contain water.  Four consecutive times
the C-B was tested with  no  (<0.6 pCi/L) detectable Rn passing into the cell.  Rn
release from the charcoal was checked by passing ambient air through the system
after  each run,  then again  one  and  six  days later.   The  immediate  checks
contained no measurable Rn,  while the later checks contained less then 1% of the
Rn collected on the charcoal.  The presence of Rn  in the later checks is due to
degassing.  Therefore, compressed gas  (helium or aged air),  as used in the EPA's
LC method, is not required for C-B/LC measurements.


EFFECT OF INITIAL LC PRESSURE


      The effect of partially evacuated Lucas cells on the Rn transfer efficiency
from the sample to the cell was evaluated.  Aliquots of water containing ~ 8,000
pCi/L of Rn were measured using the C-B with Lucas cells having varying reduced
initial pressures.  The transfer was quantitative  when the initial pressure was
less than 360-mm Hg.  Therefore, a hand-operated portable vacuum pump,  capable
of reducing pressure to 125-mm Hg, can be used for  field measurements.


FIELD MEASUREMENTS


      Field measurements were made at several public water  supplies using the
C-B, a portable Rn counter 165-cc Lucas cells. Samples were also measured by the
LS method to provide reference values.  The  C-B results agreed very well with the

-------
reference values.  Additional measurements will be performed at selected sites,
identified in the NYS study (4), to evaluate the C-B/LC field method performance
fr\v a tj^ /^c* var^/r^ rt-F T>n s*r*+- 4 *r4 ^ *» 1 AYTA! o
for a wide range of Rn activity levels.
SUMMARY

      Results of our initial evaluation  of the Collector  - Bubbler  device  for
waterborne Rn measurement  indicate that its  use  improves  the accuracy of  the
EPA's LC method.  Additionally, its use facilitates field measurements.  The need
for fragile equipment  (glass radon bubblers), compressed gas, a mechanical vacuum
pump  and/or  an  elaborate water  degassing system is  eliminated.  Individuals
measuring indoor radon with a portable Lucas cell  counting  equipment  can easily
and accurately measure radon in water.

      The work described in this paper was not funded by  the U.S. Environmental
Protection Agency and therefore the contents do not necessarily reflect the  views
of the Agency and no official endorsement  should be inferred.

-------
                                  REFERENCES
1.    U.S.  Environmental Protection  Agency.   Environmental Radiation  Data:
      Report  43,  Office   of  Radiation  Programs,   Report  EPA/520/5-86-007
      (Washington, DC: USEPA), 1985a.

2.    U.S. Environmental Protection Agency.  Nationwide Occurrence of Radon and
      Other Natural  Radioactivity in  Public Water Supplies.  EPA/520/5-85-008
      (Montgomery, AL:USEPA), 1985b.

3.    U.S.  Environmental  Protection   Agency.    Distribution Tables  for  the
      National Inorganic and Radionuclides  Survey.  Results Memorandum to Arthur
      Perler, STB from Jon Longten, WSTB,  1988.

4.    New York State  Dept. of Health.  Report of Statewide Surveillance for Radon
      in  Selected Community Water  Systems,  Bureau of  Public Water  Supply
      Protection  (Albany, NY), 1990.

5.    U.S.  Environmental Protection Agency.   National  Primary  Drinking Water
      Regulations; Radionuclides.  Federal Register 56:33050;  40CFR Part a 141,
      142;  1991.

6.    Whittaker, E.L.; Akridge, J.D.;  Giovano,  J. Two Test procedures for radon
      in drinking water: Interlaboratory   collaborative study.  Las Vegas, NV:
      U.S.  EPA Environmental  Monitoring Systems  Laboratory, Office of Research
      and Development; EPA  600/2-87/082; 1989.

-------
                         TO LUCAS CELL
1
FRITTED DISC .
r
FIRST
GRADUATION
Dl 1 IM/^CD i
rLUIMvatn , '
CLAMP 	 *-C
•
OTnornri/ D
k
u
>
i 	 1

/




t



Hr
—^b. (V/t^v
^ RTOPrOCK A
^ O 1 \Ji \j\J\Ji\ f\
- RAFtRH
* DrMiiitL
SECOND
GRADUATION
1
THRFADFD AREA
•
Figure 1.  Diagrzun of the Collector-Bubbler  (charcoal tube not shown)

-------
                     LUCAS CELL
                             —STOPCOCK
        12/2 BALL JOINTv
          12/2 SOCKET^ I  LKOVAR TO PYREX SEAL
    FINE METERING VALVE
      SILICONE SEPTA-
 MAGNESIUM PERCHLORATE<
                           -16 GAUGE HYPODERMIC
                                        NEEDLE
                               -TUBING
                           -ASCARITE
               COLLECTOR - BUBBLER
Figure 2. Rn transfer system (C-B, not shown, attaches to the drying column)

-------
                                                              XIIP-2
               MEASUREMENTS OF RADON IN WJ
                            via
                  SODIUM IODIDE DETECTORS
                    Paul N. Houle,  Ph.D.
                      Dept. of Physics
                  E.  Stroudsburg University
                  E.  Stroudsburg, Pa.  18301
                              &
                     David Scholtz, M.S.
                    Prosser Laboratories
                     Rt.  115  & State Rd.
                     Effort, Pa.   18330

                          ABSTRACT

The measurement of radon in water is typically accomplished
with a liquid scintillation system or a lucas cell system.

This paper describes a technique vhich uses Marinelli
Beakers along with the typical radon in air measurement
equipment to determine radon in water concentrations.
Counting times, LLD's, and corresponding analyses are
included.

Using this technique, laboratories that already perform
radon in air measurements will be able to add radon in water
measurements with little added expense.

-------
                                                                XMP-3
    CONTINUOUS MEASUREMENT OP THK RATYW mm RNTRRTTOW TW WATER
                USING ELECTRET ION CHAMBER METHOD

                         P.K.Hopke Ph.D.,
                       Clarkson Dniversity
                      Potsdam, NT 13699-6670
                               and
                        P.Kotrappa Ph.D.,
                       5310  H Rad Elec  Inc
                          Spectrum Drive
                       Frederick, MD 21701


                             ABSTRACT

A radon concentration of  300 pCi/L has  been proposed by U.S.
Environmental Protection  Agency as a limit for dissolved radon
in water for Municipal water supplies.  There is a need for a low
cost continuous  monitor  to  ensure that a daily average does not
exceed  this  limit.  A  system has  been  designed  that uses  the
principle of electret ion chamber.  Water flows through a
container at a predetermined low  flow rate of about 0.2 liter per
minute.  Radon exhaled into  the container during the flow is
monitored continuously with  a specially designed electret ion
chamber.  Electret is removed once a day  for measurement and
inserted back after measurement.  Calibration factors
derived in comparison with liquid scintillation system are used
to convert the data into  an  average radon concentration in water.
Sensitivity and error analysis presented  in the paper
demonstrated  that this  relatively low cost  instrument  can give
satisfactory results at 300  pCi/L.

-------
                                                                   XIIP-4
PERFORMANCE TESTING THE WD200 RADON IN WATER MEASUREMENT  SYSTEM


                G. VANDRISH, L DAVIDSON
                   INSTRUSCIENCE LTD
                    OTTAWA, CANADA
                     MAY 20/1992
                      ABSTRACT

THE  SUITABILITY  OF  THE WD200  SCINTILLATION  CELL FOR THE RAPID AND
SIMPLE MEASUREMENT  OF RADON  IN  WATER AND IN AIR HAS BEEN
EXHAUSTIVELY  STUDIED BY  INTERCOMPARISON WITH NIST REFERENCE
MATERIALS.

THE  MINIMUM DETECTABLE  LEVEL, SENSITIVITY AND REPRODUCIBILITY,
HAVE BEEN EXAMINED  FOR  SEVERAL  WATER SAMPLING PROCEDURES TO
ESTABLISH THE MOST  ACCURATE,  SENSITIVE AND  RELIABLE METHOD.
PROCEDURES INCLUDE  SAMPLING  DIRECTLY FROM THE WATER SOURCE OR
DECANTING A BOTTLED SAMPLE LOCATED WITHIN THE UNIT.

STUDIES  WERE  MADE OF BACKGROUND BUILDUP, MEMBRANE DETERIORATION,
RESPONSE TIME AND STABILITY.
                     BACKGROUND

 THE TWO PRINCIPAL METHODS FOR RADON IN WATER; THE DEGASSING
 METHOD AND THE LIQUID SCINTILLATION METHOD HAVE LIMITED PRACTICAL
 UTILITY IN THE FIELD.

 DEGASSING METHODS ARE LABOUR INTENSIVE, TIME CONSUMING AND
 REQUIRE SKILL TO DEPLOY, WHILE LIQUID SCINTILLATION METHODS USE
 CHEMICALS, CAN BE EXPENSIVE AND PROVIDE RESULTS AT A LATER TIME.

 THE SYSTEM DESCRIBED IN THIS PAPER SATISFIES THE NEED FOR A
 QUICK AND ACCURATE FIELD METHOD FOR RADON IN WATER.

 THE SUITABILITY OF THE WD200 FOR MEASURING RADON IN AIR AND
 RADIUM IN WATER WAS ALSO INVESTIGATED AND PRELIMINARY RESULTS ARE
 REPORTED  IN THIS RESEARCH.

-------
               DESCRIPTION OF WD200 SYSTEM                  2

THE WD200 RADON  IN  WATER MEASUREMENT SYSTEM IS A WATER
SAMPLER/EQUILIBRATION CHAMBER AND SCINTILLATION CELL IN ONE
UNIT. THE SYSTEM IS DESIGNED TO MEASURE RADON OVER A WIDE RANGE
OF ACTIVITIES  (2 PCI/L AND HIGHER) QUICKLY AND RELIABLY WITH A
MINIMUM OF OPERATOR INVOLVEMENT OR SKILL. ALTHOUGH THE SYSTEM IS
DESIGNED TO OPERATE WITH SAMPLE VOLUMES-OF 0.1 TO 0.15L, MUCH
LARGER VOLUMES CAN  BE EMPLOYED IN SPECIALIZED APPLICATIONS (SUCH
AS RADIUM IN WATER).

TO OBTAIN A MEASUREMENT THE OPERATOR SIMPLY TAKES A WATER SAMPLE
WITH THE EQUILIBRATION CHAMBER (OR INSERTS A FILLED SAMPLE
BOTTLE), WAITS FOUR HOURS FOR THE RADON TO DEGAS PASSIVELY FROM
THE WATER SAMPLE INTO THE AIRSPACE AND DIFFUSE THROUGH THE
MEMBRANE INTO  THE SCINTILLATION CELL (WHERE IT COMES TO SECULAR
EQUILIBRIUM WITH ITS DAUGHTERS). A TEN TO THIRTY MINUTE COUNT IS
THEN OBTAINED  WITH  A SUITABLE SCALAR (RN2000) AND THE RADON IN
WATER CONCENTRATION OBTAINED BY CALCULATION WITH A SIMPLE
FORMULA.

FOR THE 0.1 TO 0.15L WATER SAMPLE VOLUMES EMPLOYED A SENSITIVITY
OF 0.2 CPM/PCI/L AND A MINIMUM DETECTABLE LEVEL OF 2 PCI/1 OF
RADON IN WATER IS POSSIBLE AT 20C.

ALTERNATIVELY, WITH THE EQUILIBRATION CHAMBER REMOVED, THE WD200
CAN BE EMPLOYED AS A PASSIVE RADON GAS  SCINTILLATION CELL FOR
RADON IN AIR MEASUREMENTS WITH A  SENSITIVITY OF 0.5 CPM/PCI/L.
THIS CALIBRATION VALUE  IS REFERENCED TO  THE USEPA.

THE RESULTS  DESCRIBED HERE ALSO DEMONSTRATE  THE POSSIBILITY OF
MEASURING  RADIUM IN WATER WITH THE SAME  SCINTILLATION  CHAMBER
AND A ONE  LITER SAMPLE  SIZE.

-------
FORMULA RELATING RADON IN WATER CONCENTRATION TO MEASURED  COUNT
RATE WITH RN-2000 RADON DETECTOR
          Rn(W)=(0.26+Va/Vw)*(C/T/S/D(t))
Rn(W)           IS RADON  IN WATER  CONCENTRATION  IN  PCI/L

Va              IS VOLUME (L)  OF AIR  (SCINTILLATION CELL  +
                AIRSPACE)

Vw              IS VOLUME (L)  OF WATER SAMPLE

C               IS TOTAL  COUNT (MIN)

T               IS COUNT  PERIOD (MIN)

S               IS SENSITIVITY OF  CELL

0.26            IS OSWALD COEFFICIENT  (20C)

D(t)            CORRECTION FOR DECAY OF RADON  FROM  TIME OF
                SAMPLING

-------
                          EXAMPLE


TYPICAL PARAMETERS  ARE

Va   =    0.320  L  (VOLUME CELL=0.17L;VOLUME EQUILIBRATION
          CHAMBER=0.279L;VOLUME WATER SAMPLE=0.13Lj

Vw   =    0.13 L (BOTTLE SIZE)

C    =    880 COUNTS IN 10 MINUTES

T    =    10 MINUTES (COUNT LONGER FOR BETTER STATISTICS)

S    =    0.52 CPM/PCI/L (EPA AND NIST TRACEABLE)

D(t) =    0.97  (4  HOURS DELAY SAMPLING TO COUNTING)

0.26 =    OSWALD COEFFICIENT FOR 20C ( +5C=0.42; +10C=0.35;
          4-30 = 0.20) TEMPERATURE MUST BE HELD CONSTANT FOR BEST
          ACCURACY

THUS:

Rn(W)  =    (0.26 + 0.32/0.13)*880/10/0.52/0. 97
           478 PCI/L
 BY  TAKING LARGER SAMPLE VOLUMES, OR COUNTING FOR LONGER PERIODS
 BETTER  SENSITIVITY CAN BE ACHIEVED.

 FOR THE SAMPLE SIZE CHOSEN  (0.13L), AN APPROXIMATE SENSITIVITY OF
 0.2 CPM/PCI/L IS POSSIBLE.  THE BACKGROUND OF THE SCINTILLATION
 CELL AND THE DETECTOR DETERMINE THE MINIMUM DETECTABLE LEVEL.
 THUS WITH A BACKGROUND OF 0 . 2 CPM  (NO CONTAMINATION); A
 CONSERVATIVE ESTIMATE OF THE MINIMUM DETECTABLE LEVEL WOULD BE 2
 PCI/L FOR THE 0.13L SAMPLE  SIZE.

-------
                    SUMMARY OF TESTING

RADON DIFFUSION AND SOLUTION IN THE PLASTIC  CONSTRUCTION
MATERIALS, LEAKAGE RATES AND LOSSES, DIFFUSION  TIMES  FROM
SOLUTION AND THROUGH THE DIFFUSION MEMBRANE  WHICH  AFFECT
MEASUREMENT TIMING AND SYSTEM UTILITY, BACKGROUND  BUILDUP,
CONTAMINATION OR MEMBRANE DETERIORATION,  SENSITIVITY,  AND
REPRODUCIBILITY WERE ALL ADDRESSED DURING THIS  INVESTIGATION.
THERE  IS NO EVIDENCE OF LEAKAGE  OF  RADON  FROM THE WD200 DURING
THE CALIBRATION EXPERIMENTS.  LEAKAGE WOULD HAVE RESULTED IN LOWER
THAN EXPECTED RADON BUILDUP AND  HAVE BEEN ESPECIALLY APPARENT
OVER THE LENGTHY  TEN DAY  BUILDUP PERIODS  EMPLOYED.
                RADON  AND  RADIUM  BACKGROUND BUILDUP IN WD200
                CONSTRUCTION JjATERIALS

SINCE THE WD200 IS  CONSTRUCTED OF ABS  AND PVC PLASTIC THERE IS
THE POSSIBILITY OF  SOME RADON PENETRATION AND/OR A RESIDUAL
BUILDUP  DURING  PROLONGED  HIGH RADON  LEVELS WHICH COULD BE
RELEASED DURING LATER EXPERIMENTS AS A "RADON BACKGROUND".

THIS POTENTIAL  PROBLEM WAS  INVESTIGATED AFTER EACH CALIBRATION
RUN IN THE  FOLLOWING  MANNER. THE SOURCE WAS REMOVED,  THE
EQUILIBRATION CHAMBER WAS FLUSHED AND  THE WD200 WAS RECLOSED
AFTER A  PERIOD  OF ONE HOUR  (*JO PERMIT  GAS PHASE RADON TO ESCAPE),
THE SYSTEM  WAS  THEN FOLLOWED FOR UP  TO 10 DAYS TO ALLOW THE
BUILDUP  OF  RADON FROM ADSORBED RADIUM  OR FROM RADON DISSOLVED IN
THE PLASTICS.

NO MEASURABLE RESIDUAL WAS  NOTED IN  ALL EXPERIMENTS.  RADON
PENETRATION AND/OR  SOLUTION WOULD HAVE ALSO RESULTED IN LOWER
THAN EXPECTED RADON LEVELS, AN EFFECT  WHICH WAS NOT OBSERVED IN
THESE STUDIES.

NO DETERIORATION NOR  CONTAMINATION OF  THE MEMBRANE WAS NOTED IN
THESE EXPERIMENTS.  THE DIFFUSION MEMBRANE CONTACTED THE WATER
SAMPLE ONLY BRIEFLY IN MOST MEASUREMENTS.

-------
                     DIFFUSION RATES

THE WD200 CELL WAS  PLACED OPEN IN A CALIBRATION CHAMBER AND THE
INGROWTH CURVES MONITORED.  SECULAR EQUILIBRIUM WAS ACHIEVED IN
THE 3-4 HOUR PERIOD TYPICAL OF RADON DAUGHTERS, INDICATING THAT
THE DIFFUSION RATE  THROUGH THE MEMBRANE WAS NEGLIGIBLE IN
COMPARISON TO THE TIME TO ACHIEVE SECULAR EQUILIBRIUM (4HRS).

THE REVERSE DIFFUSION RATE WAS ALSO STUDIED BY REMOVING THE UNIT
FROM THE CHAMBER AND LETTING THE RADON ESCAPE. NO DIFFUSION
CONTROLLED DELAY DUE TO MEMBRANE EFFECTS OR SOLUTION IN THE
PLASTIC WAS OBSERVED IN THE INGROWTH AND DECAY CURVES.

THESE EXERIMENTS WERE REPEATED WITH AN 0.12L WATER SAMPLE. THE
TIME REQUIRED TO ACHIEVE PARTITION EQUILIBRIUM WAS SMALL IN
COMPARISON TO THE TIME TO ACHIEVE SECULAR EQUILIBRIUM.

-------
                    SAMPLING PROCEDURES

THE ACCURACY OF SAMPLING DIRECTLY WITH THE WD200 EQUILIBRATION
CHAMBER OR WITH A O.13L SAMPLE BOTTLE WAS EXAMINED BY MEANS OF AN
NIST REFERENCE STANDARD. THE STANDARD WAS PLACED IN THE SAMPLE
BOTTLE, SEALED FOR ABOUT ONE DAY, THEN TRANSFERRED BY ONE OF TWO
METHODS TO THE WD200 SAMPLING CHAMBER

IN THE FIRST METHOD THE BOTTLE WAS  INSERTED GENTLY INTO THE
SAMPLER, UNCAPPED, AND THE SAMPLER  IMMEDIATELY  SEALED. THE UNIT
WAS THEN TILTED TO DECANT THE SAMPLE FROM THE BOTTLE.

     *TYPICAL RESULTS ARE 988, 970, AND 990 PCI/L FOR  A 1000
     PCI/L NIST TRACEABLE SAMPLE. IN GENERAL, RESULTS WERE WITHIN
     5% OF THE NIST TRACEABLE STANDARD.

IN THE SECOND METHOD THE SAMPLE WAS POURED FROM THE BOTTLE INTO
THE SAMPLER WHICH WAS THEN SEALED.

     *TYPICAL RESULTS ARE 733, 888, AND 920 PCI/L FOR A 1000
     PCI/L REFERENCE SYSTEM. IN GENERAL, GOOD ACCURACY AND
     REPRODUCIBILITY WERE DIFFICULT TO ACHIEVE.

EVIDENTLY THE FIRST PROCEDURE PROVIDES ACCEPTABLE RESULTS WHILE
THERE  IS SOME LOSS OF RADON DUE TO  POURING IN THE SECOND
PROCEDURE.

THE FEASIBILITY OF USING THE WD200  TO SAMPLE DIRECTLY FROM A
WATER  RESERVOIR WITHOUT THE USE OF  A SAMPLER BOTTLE IS CURRENTLY
BEING  STUDIED.

-------
                    CALIBRATION

THE WD200 SYSTEM WAS  CALIBRATED  AGAINST A 1000  PCI/L  NIST
TRACEABLE REFERENCE STANDARD.  0.120  LITERS OF THE STANDARD WAS
PLACED IN THE EQUILIBRIUM CHAMBER AND ALLOWED TO DEGAS  FOR FIVE
HOURS BEFORE SEALING.  THE INGROWTH OF RADON INTO THE  DETECTOR WAS
THEN MONITORED CONTINUOUSLY FOR  SIX TO TEN DAYS. THE  OSWALD
FACTOR EMPLOYED IN THESE CALCULATIONS WAS 0.27  (18C).

                    TABLE

INGROWTH OF RADON FROM NIST STANDARD/CONTINUOUS MEASUREMENTS

     TIME      COUNT           CALCULATED
                               SOURCE ACTIVITY

     (HR)       (MIN)             (PCI/L)

     67         72.4                 1056

     96         88.7                  996

     116        93.9                  987

     120        104.9                1019

     144        113                   987
 IN  ALL CASES,  THE CALCULATED AND MEASURED SOURCE ACTIVITIES ARE
 WITHIN 5% OF THE NIST TRACEABLE STANDARD.

 THE RESULTS DEMONSTRATE THAT AFTER A FIVE DAY  INGROWTH PERIOD A
 0  12L RADIUM IN WATER SAMPLE OF 1000 PCI/L WILL GENERATE ABOUT
 1000 COUNTS/10 MINUTES. THUS BY INCREASING THE SAMPLE SIZE TO
 ABOUT A LITER (10X), A RADIUM  IN WATER  LEVEL OF ABOUT A PCI/L CAN
 BE MEASURED AFTER A ONE DAY  INGROWTH PERIOD  (PROVIDED THAT THE
 SYSTEM IS NOT CONTAMINATED). THE ONE LITER SAMPLE  SIZE IS EASILY
 ACHIEVED WITH SLIGHT MODIFICATION OF THE  SYSTEM.

-------
               MEASUREMENT OF RADIUM WITH THE WD200

THIS TOPIC WAS INDIRECTLY EXPLORED WITH THE 1000PCI/L NIST
TRACEABLE STANDARD DURING THE CALIBRATION STUDIES. RESULTS
DEMONSTRATE THAT WITH A 130 ML SAMPLE VOLUME A COUNT RATE OF ONE
COUNT PER MINUTE PER PCI/L RESULTS AFTER A FIVE DAY INGROWTH
PERIOD. IF THE EQUILIBRATION CHAMBER IS INCREASED  IN SIZE TO
SEVERAL LITERS, A PRACTICAL MEASUREMENT COULD BE MADE AFTER A ONE
DAY INGROWTH PERIOD WITH A MINIMUM DETECTABLE LEVEL IN THE
SUBPICOCURIE RADIUM RANGE.

               MEASUREMENT OF RADON  IN AIR WITH THE WD200 SYSTEM

THE WD200 SAMPLER SYSTEM OPERATES AS A REGULAR DIFFUSION CELL
WITH A SENSITIVITY OF 0.5 CPM/PCI/L  IN AIR.
               WD200 RADON IN WATER  SCINTILLATION  CELL

-------
WD200 RADON  IN  WATER SCINTILLATION CELL
                               WINDOW
                               SCINTILLATION
                                    CELL
                               DIFFUSION
                               MEMBRANE
                               SEALING GASKET
                               SCREW THREAD
                               FASTENER
                               EQUILIBRATION
                               CHAMBER

-------
^-»•"-.-; • ;_.
•    •' '^.fcfc,
      •-*r.

-------
                                                                 XIIP-5
TEMPORAL VARIATIONS IN BEDROCK WELL WATER RADON AND RADIUM. AND
           WATER RADON'S EFFECT ON INDOOR ATR RADON

       by:  Nancy W. McHone, Geologist
            Margaret A. Thomas, Environmental Analyst III
            CT Dept. of Environmental Protection
            Natural Resources Center, 165 Capitol Ave.
            Hartford,  CT  06106   (203) 566-3540

            Alan  Siniscalchi, Radon  Program Coordinator
            CT Department  of Health  Services Radon Program
            150 Washington St.
            Hartford,  CT  06106   (203) 566-3122
                            ABSTRACT

 Monthly sampling during 1989-1990 of a high radon bedrock well
 in Connecticut revealed radon and radium variations of more
 than 94% over an 18 month period (1).  This research involves
 systematic sampling of granitic bedrock radon and radium in
 well water and provides data for evaluating single
 'representative' well water radon analyses.  Five wells are
 being sampled hourly, daily, and weekly to document temporal
 variations in well water radon and radium.  These data are
 bSing compared with water use, water chemistry, meteorological
 data? and seasonal climatic effects   A 'high ' radon w ell
 (662 000 pCi/L), three  'moderate- radon wells  (1 5,000 - 25 000
 pCi/L), and a relatively  'low' radon well  (3,500 pCi/L) are
 included in the study.  Water radon data  is compared with
 mSnihly aSha track  indoor air radon measurements to examine
 its derivative  and integrated effects on  indoor air radon.

 Generally, hourly radon measurements rose quickly during the
 earty hours of  the day  with  13%  to  48% variation overall.
 Daily radon levels varied from 24%  to 41%.  Radon and radium
 analyses" are part of weekly  testing which will continue through
         Radon variability  potentially caused by a difference  in

       Although the work described in this paper was partially
       funded by the U.S. Environmental Protection Agency  the
       contents do not necessarily reflect the views of the
       Agency and no official endorsement should be inferred.

-------
                         INTRODUCTION

     From 1985 to 1989 about 600 private and public bedrock
wells in Connecticut were tested for radon (1, 2, 3).  Ninety-
five percent of the wells tested were over the EPA proposed MCL
of 300 pCi/L for public water supplies.  It was found that
radon values varied statewide, and by rock type.  In
particular, granites and granitic gneisses contained the
largest percentage of wells with elevated radon values (2).

     Repeated radon testing was conducted on a well in the
Nonewaug granite by the CT Department of Health Services and
the U.S. Geological Survey.  Radon levels in this well ranged
from a high of 660,000 pCi/L to a low of approximately 50,000
pCi/L and these values sometimes changed from the high to the
low value in just one month.  Analyses of preliminary data from
one and a half years of nearly continuous monthly water radon
sampling indicated a possible seasonal variation.  Radon
values, from another study testing one well frequently over a
period of up to six hours, varied within that short time period
(3).  To learn more about radon variation in well water, and to
determine if such change occurs in other wells, a study was
designed to look at temporal variations of radon and radium in
well water.  Because both short term and long term variations
were to be examined, the study was designed in three phases.
Wells were to be tested hourly during phase one, daily during
phase two, and weekly during phase three.

                          STUDY DESIGN

     For the purposes of this study, a "low" well has water
radon levels around 3000 to 4000 pCi/L, a "moderate" well is
one  in the range of 15,000 to 25,000 pCi/1, and a "high" well
is over 50,000 pCi/L.  Five wells are included in the study.
One  is the previously mentioned high well in the Nonewaug
granite, three are moderate level wells in the Eastford phase
of the Canterbury gneiss, and one is a low radon level well in
the  Hope Valley Alaskite gneiss.

     The Nonewaug granite is a two-mica granite containing many
pegmatite bodies  (4).  Compositionally similar to Nonewaug is
the  Eastford gneiss, described as a muscovite-biotite-
microcline-oligoclase-quartz gneiss  (5).  The Hope Valley
gneiss ranges from albite alaskite to oligoclase quartz
monzonite in composition (6, 7).

PHASE ONE  (HOURLY TESTS)

     Table 1 illustrates the various sampling parameters  of the
three phases of the study.

-------
TABLE 1.  TESTING PARAMETERS AND WELL IDENTIFICATION
TESTING PARAMETERS
                           Water testing interval
                           Hourly     Daily      Weekly
Water radon
Water temperature
Barometric pressure

Alkalinity
x
PH
Eh
x
Conductance
x
Dissolved oxygen
x
Water use
Rainfall
Air radon  (measured monthly)
Radium
WELLS
                        Well number   Radon range (pCi/L)
Low radon well
Intermediate wells


High radon well
MS-48
ED-10
ED- 15
ED-18
BM-3
3,315
13,325
12,055
15,400
33,560
- 5,953
- 24,725
- 56,825
- 31,330
- 597,815
For the purposes of  this  study, the terms  "low",
"intermediate"  and "high"  are used merely as relative terms
to distinguish  groups  of  radon values,  and are not intended
to be  qualitive descriptions  for  these  wells.

-------
     Phase one consisted of two days of hourly tests on each of
the five wells.  Its purpose was to measure possible radon
variation during a typical "water use day".  The testing began
before the home owners arose in the morning and continued until
after they went to bed at night, yielding sixteen to eighteen
tests per day.  Each hour a water sample was collected for
radon analysis and the temperature, pH, Eh, alkalinity,
conductance, and dissolved oxygen of the water were determined.
In addition, a record was kept of the barometric pressure and
the amount of time the well pump ran since the last sample,
representing "normal" household water use.  The two days of
testing on each well were separated by five to twenty days.

PHASE TWO (DAILY TESTS)

     Radon results from the first phase of testing were
examined to determine the time of day at which radon levels
appeared to be most stable  (Figure 1).  Daily tests were then
conducted during these times of radon value stability.  The
same tests for water chemistry and physical characteristics
were performed during this phase as in phase one.  Each well
was tested at the same time of day for fourteen consecutive
days.  In order to examine the effects of rainfall on radon
levels, rain gauges were installed at each well location and
checked daily.

PHASE THREE  (WEEKLY TESTS)

     This phase consists of weekly testing of the five wells
for one year.  Phase two results were used to determine the day
of the week on which there seemed to be the least variation in
water radon levels.  Some consideration was also given to the
laboratory schedule for analysis, resulting in a Wednesday
sampling day.  During this phase the same water analyses were
performed as  in phase one. and water samples were also
collected for radium-226 and radium-228 analyses.  Also alpha
track detectors were installed to monitor the air radon in the
kitchen and basement of each home, to examine if air radon
fluctuations mimic water radon changes.  These detectors were
not placed in the high radon home  (BM-3), where air and water
are mitigated.  The detectors were changed each month and sent
to the manufacturer for analysis.  In addition, one year alpha
track detectors were left in each kitchen and basement.

SAMPLING METHODS

     Water samples for radon analysis were collected from an
inverted funnel according to the EPA method (8) after running
the water for ten minutes  (except as noted below for the daily
samples).  Ten ml samples were transferred immediately to 25 ml
glass scintillation vials containing 10 ml of scintillation

-------
                     Hourly Water Radon
 500000
               BM-3
  250000
           8  12  16 20  24


           Time of  Sample


                ED-15
   28000  -
o  24000
O.


c  20000
o
   16000  -
     4200
 >  4000
         4  8  12 16 20 24


            Time  of Sample


                MS-48
  a
  o_
     3800
     3600
            ED-10
                              O
                              a_


                              c
                              o
                              -o
                              o
                              a:
18000 -
16000
        8  12  16  20  24


        Time  of Sample


             ED-18
         8  12  16  20


         Time of Sample






           Legend
          O Day One


          • Day Two
           8   12  16  20  24


            Time of Sample


     Figure 1. Water radon levels for samples taken hourly

         on each of two days.  Note changes in scales.

-------
fluid.  Samples were delivered to the Department of Health
Services Laboratory the next morning for same day analysis on a
UTC Packard Tri-Carb 4530 Liquid Scintillation Spectrometer.
Water was collected from an outside tap at each house except at
BM-3, where water was collected directly from the pressure tank
outlet via a garden hose.  During the winter, one well was
tested by collecting water from the kitchen faucet.  For the
water chemistry tests, probes of the various instruments were
inserted through holes in the styrofoam lid of an acrylic box.
The box was connected via rubber tubing to a garden hose
attached to the outside tap.  Water was allowed to run through
the box and thus around the probes.  The water temperature was
also recorded in the box.  The probes measured dissolved oxygen
(YSI Model 57 Oxygen Meter), Eh (Markson Model 93 Eh Meter), pH
(Orion Research Model 211 pH Meter) and conductance (Markson
Conductance/TDS Meter).  Alkalinity was measured by titration
with a LaMotte Chemical Alkalinity Test Kit Model WAT-MP-DR.  A
digital time totalizer from GFA Engineering measured,  by
induction, current traveling to the well pump each time the
pump operated.

                             RESULTS

     Nine previous radon tests of the Eastford phase of the
Canterbury gneiss yielded a geometric mean of 18,000 pCi/L with
a range from 3,000 to 35,000 pCi/L (1, 2).  The ten
exploratory tests done for this study on the same geologic unit
have a geometric mean of 28,000 pCi/L with a range from 17,000
to 74,000 pCi/L.

RADON

Hourly tests

     Radon levels in the high well (BM-3)  rose at the  beginning
of the first day from a low, at 6 AM, of 435,460 pCi/L to a
high, at 9 AM, of 488,390 pCi/L (Figure l).   This high was
maintained until about 3 PM, at which time values began a
decline.  This decline continued to the end of the testing day,
when the radon was back at the morning low.   The second day of
testing, twenty days after the first, began with a 6 AM test
value of 297,860 pCi/L and rose to 318,900 pCi/L at 9  AM.
After an hour at that level there was a steady decline to a low
of 251,040 pCi/L at the end of the day at  10:45 PM, a
difference of 27% in 12 3/4 hours.

     Changes in the moderate wells during  a  day of testing were
less dramatic. Although total radon values were lower,  they
followed a similar trend of rising swiftly at the beginning of
the water use day.  ED-10 rose from a low  of 16,145 pCi/L at 5
AM to a high of 21,335 PCi/L at 9 AM (Figure 1),  an increase

-------
of 32%, before dropping back to around 19,000 pCi/L for six
hours, then increasing to around 21,000 pCi/L for the rest of
the day.  The second day of testing, eight days later, began
with a low of 16,560 pCi/L at 5 AM, climbing to a high of
21,680 pCi/L at  10 AM, before a small decline occurred followed
by a gradual rise for most of the rest of the day.  Maximum
levels reached 21,675 pCi/L, a total difference of 5120 pCi/L
for the day, a 31% increase.

     The second  moderate well, ED-15, showed a pattern similar
to ED-10's on its second day of testing, with an overall
increase of 8,435 pCi/L, or 42%  (Figure 1).  The first test
day, however, produced much less variation.  At 5 AM the radon
level was 21,375 pCi/L, rising to 22,815 pCi/L at 7 AM. It
stayed around 22,000 pCi/L until 6  PM, when it dropped suddenly
to 15,030 pCi/L.  At 9 PM the radon level rose  to 23,260
pCi/L, to 25,175 pCi/L at midnight, then dropped to 20,960
pCi/L at 1 AM.

     Moderate well ED-18 basically  rose quickly in the morning
of both days, to highs at 11 AM  and noon, then gradually
declined  (Figure 1) .  On day one the second sample was lower
than the first,  but radon values then increased, and there was
a sharp anomalous decrease at 5  PM.  Excluding this anomalous
decrease, the variations were 43% and 58% over the two days.

     The radon pattern of the low well, MS-48, was different
 (Figure 1), as there was no morning increase.  Instead, on both
days, the radon  levels decreased gradually all day from morning
highs.  Variations during each day  of testing were 16% and 5%.

     Attempts to correlate radon values with water use resulted
in statistically significant  (Student's t test, t = 1.33 or
higher  at  10% significance  level) correlations on only four of
the ten days  of  testing  (Table 2) .  Three of these were
positive  correlations, the other negative.  The low well  (MS-
48) had positive correlations on both days, the high  (BM-3) and
one moderate  well  (ED-15) had correlations on one day  each, and
the other  two moderate wells  showed no  correlations.  All wells
had significant  correlations  between radon and conductance, but
MS-48  and  ED-15  each  correlated  on  only one day.  Two  of the
other  wells had  a  negative  correlation  one day, positive the
other.  Correlations  with  other  parameters were more  variable
 (Table  2) .

Daily  tests

      During the  daily tests  an  additional question was
 addressed —  is  there a  difference  when collecting samples
between running  the water  two minutes  —  as recommended by EPA
 —  and running  it ten minutes — as is  commonly practiced.  As

-------
TABLE 2. T-TEST RESULTS FOR RADON AND OTHER PARAMETERS

HOURLY TESTS






BM-3-1 BH-3-2 ED-10-1 ED-10-2 EO-15-1 ED-15-2 ED-18-1 ED-18-2 MS-48-1 MS-48-2

Water temperature
Barometric pressure
Alkalinity
PH
Eh
Conductance
Dissolved oxygen
Water use
DAILY TESTS

Water temperature
Barometric pressure
Alkalinity
P«
Eh
Conductance
Dissolved oxygen
Water use
Rainfall
WEEKLY TESTS
Water temperature
Barometric pressure
Alkalinity
P«
Eh
Conductance
Dissolved oxygen
Water use
Rainfall
Ra-226
COMBINED TESTS
Water temperature
Barometric pressure
Alkalinity
PH
Eh
Conductance
Dissolved oxygen
t
1.87
4.39
-1.43
-1.13
1.63
-1.43
-1.62
-0.57
Btt-3
t
-3.54
2.34
-2.03
-3.36
1.81
-0.56
1.76
-2.79
0.53
2.42
1.68
6.80
-0.15
-0.36
3.23
-6.71
0.39
-0.91
2.45

4.29
3.22
6.27
6.29
-3.72
3.1*
-4.06
t
-2.19
-1.03
2.80
1.25
2.64
1.47
0.16
2.12
ED- 10
t
-0.45
-2.60
-1.19
-2.26
5.34
-2.03
0.82
4.49
0.24
-0.80
-1.08
0.55
-0.03
-0.83
1.97
2.36
1.36
-0.67


0.43
-0.32
-0.81
-1.26
3.09
0.57
-6.35
t
--
5.32
-1.08
0.85
3.85
-1.43
1.56
-0.08
ED- 15
t
1.61
-0.04
-1.32
2.32
0.27
-2.37
3.46
0.61
-1.21
-0.04
1.95
0.11
0.60
-1.20
0.22
-0.30
2.98
-0.29


4.83
0.26
5.70
-0.15
-1.92
0.12
0.23
t
1.72
0.18
-4.38
-5.04
2.57
-2.13
1.26
-0.40
ED- 18
t
0.20
-0.50
0.01
-1.36
-0.43
-1.67
0.68
1.24
0.03
0.31
-0.30
-1.07
-0.76
0.38
-1.55
-0.64
2.29
-0.69


2.18
-2.78
5.25
0.80
-4.02
-2.98
-0.29
t t
-0.67 3.04
0.59 2.81
0.44 1.65
0.56 -3.80
0.74 -2.85
1.14 -6.20
0.06 2.32
0.66 -2.22
MS-48
t
-0.21
0.34
-1.35
-0.86
-0.17
0.49
0.83
-1.45
-0.35
-0.49
-1.03
1.55
-1.52
-3.03
1.03
-2.44
0.72
-6.60


2.50
-0.83
2.39
1.05
-1.62
0.53
-2.97
t t t t
-0.63 -0.39 -1.40 1.51
4.30 -2.51 1.36 1.91
-0.98 -0.85 1.76 -2.11
0.16 -0.40 1.71 0.13
-0.11 -0.36 2.22 0.04
1.54 -2.31 1.76 0.02
-1.45 -1.07 2.87 -0.16
-0.63 0.65 2.90 1.50


Note: t = 1.33 at 10%
significance level
for hourly tests,
t = 1.34 for daily
tests, and t = 1.30
for weekly and for
combined tests





















-------
can be seen from Figure 2, ten minute radon levels are
generally slightly higher than two minute levels, an average of
about 4.3% higher.  Over the fourteen days of tests, variation
within the five wells ranged from 16% to 57% for the two minute
samples, and from 24% to 41% for the ten minute samples.  The
overall higher radon values of the ten minute samples may be
partially attributed to water sampled directly from the well,
rather than a mixture of well and pressure tank water, as may
occur when the water is run for only two minutes.  If pressure
in the tank was  low when the water was turned on, fresh water
would begin entering the tank sooner than if the initial
pressure was higher, thus delaying the onset of pumping and the
flow of fresh water  into the tank.

     The different water use patterns of the well users  (Figure
3) were expected to  influence the radon levels we would find
during the daily water  sampling.  The well pump  electrical
current, as measured by the time totalizer, was  used  as a
determination of water  use.  The correlations of radon with
water use are variable.   The high and  low wells  both  had
negative correlations  (t  =  -2.79 and -1.45, respectively, t =
1.34 at  10%  significance  level).  One  of the moderate wells had
a  positive  correlation  (t = 4.49),  while the other  two  did not
show significant correlations  (t =  0.61 and  1.24).  The
coefficients  of variation for  radon during these two  weeks of
daily tests  ranged from 0.08  for BM-3, through  0.09 for MS-48,
0.10  for  ED-10  and ED-18,  to 0.25  for  ED-15.

      Significant correlations  of  radon with  water chemistry
 indicators  is variable (Table  2).   The best  correlation is
between radon and pH - three wells had a  negative correlation,
one a  positive correlation,  and only MS-48  had  no correlation.
This  compares with only three days of  significant correlations
during the ten days of hourly testing.

Weekly tests

      The weekly tests began on 11 September 1991 and will
 continue for a year.  As of this writing forty-two weeks of
 testing are complete.  Except for the high well, radon levels
 declined gradually to mid-winter (Figure 4).  ED-15 then
 continued to decline, while ED-10, ED-18 and MS-48 began a slow
 increase.  Coefficients of variation over this time have ranged
 from 0.06 for MS-48 through 0.21 for ED-15 (which was 0.25
 during daily testing).

      The greatest variation, 0.73, occurred in the radon values
 from the high well  -  from a high of 597,820 pCi/L to a low of
 33 560 PCi/L .   in mid-November, 1991, radon in BM-3 dropped
 from 482 380 pCi/L  to  33,560 pCi/L in one week, stayed near
 this level for  four weeks, then increased again.  Another drop

-------
                              Daily Water  Radon
                    BM-3
                                                     ED-10
^  420000 ,r

<         *
O  390000 -
 Q.
 C
 o
T>
 D
o:
    360000 -
330000 -
    300000
  o
  a.
  o
  TJ
  O
          SAT MON WED  FRI SUN  TUE THU


                Day of Sample


                   ED-15
  c  30000 -
  o
 TJ
   .  20000
          SAT MON WED  FRI SUN  TUE THU
                Day of Sample


                    MS-48
       4000
O
a.


c
o
T3
O
Ct
                                       O
                                       Q.
                                   C
                                   O
                                   T>
                                   O
                                   o:
                                           SAT  MON  WED  FRI  SUN  TUE THU


                                                 Day of Sample


                                                       ED-18
                                           28000  -
                                           20000 -
                                                SAT MON  WED  FRI SUN TUE  THU
                                                  Day of Sample
                                                  Radon Samples
                                                   o  2 minute

                                                   •  10 minute
          SAT  MON WED FRI  SUN TUE  THU


                Day of Sample


  Figure 2. Daily radon levels for all wells,  showing water radon values

      for samples collected after running water  2 minutes and  10

                minutes.  Note changes in vertical scales.

-------
                 BM-3
                                                         ED-1.0
o
CL

c
o
T>
D
ac.
   450000
    400000
350000
    300000
                                0.4
       SUN TUE THU SAT MONWED FRI


            Day of week


               ED-15
 O
  CL
  c
  o
  D
  cr
                                    3
                                    o
                                 0)
                                 v>
                                 3

                                 l_
                                 a>

                                 15
                                 0.4
       SUN TUE THU SAT MONWED FRI


             Day of week


               MS-48
                                 0.8
                             0.6
                                     3
                                     O
                                     a>

                                     "5
       3500
           SUN TUE THU SAT MONWED FRI


                 Day  of week
o
a_

c
o
-o
o
a:
                                                                          O
                                                                          _c
                                                                   0.6  >-
                                                                       0.4
                                                  SUN TUE THU SAT MONWEO FRI


                                                       Day  of week



                                                         ED-18
                                            28000 -
                                                                           a>

                                                                           o
                                                  SUN TUE THU SAT MONWED FRI


                                                       Day of week
                                                    Legend
                                                     • Radon


                                                     O Water use
       Figure 3.  Daily radon and  water use in all weils.

-------
                            Weekly Water Radon
                    BM-3
                                                  ED-10
o
Q.


C
O
-D
D
CC
600000  h

500000

400000

300000

200000

100000
             10/2712/262/24 4/24 6/23


                Date of Sample


                    ED-15
 o
 CL

 c
 o
 TJ
 o
 ct:
                                   26000

                                   24000

                                o  22000
                                a.
                                   20000
                                    8000  I-

                                    6000
                                            10/27 12/26 2/24  4/24  6/23


                                               Date of sample


                                                   ED-18
             10/2712/26 2/24 4/24 6/23


                Date  of sample
                                            10/2712/262/24 4/24 6/23


                                              Date of sample
                                     MS-48
                    o
                    Q.
                    D
                    ce
                              10/2712/26 2/24  4/24  6/23


                                 Date of  sample


    Figure 4.  Weekly  radon levels for all wells.  Solid line  is second
          order  regression.  Note changes in  vertical scales.

-------
occurred in early March, from 440,670 pCi/L to 62,420 pCi/L in
one week.  This time radon stayed at this level for an
additional eight weeks before rising again to its higher level.
Then in June, over two weeks, the radon values went from
486,872 pCi/L to 45,638 and back to 439,335  pCi/L.  Because of
the huge variation found in this well, duplicate samples are
routinely collected each week.

     During phase three, correlations of radon with water
parameters were best for conductance, dissolved oxygen and
water use, all of which had three significant correlations for
the five wells  (Table  2).  However, only with water use were
the correlations all of the same sign.

RADIUM

     Except  for BM-3,  radium-226 ranged from 0 to  4.32 pCi/L
with arithmetic means  from  0.6  to 1.4 pCi/L for the individual
wells.  With  error bars of  approximately 0.7 pCi/L, there  is
little perceived variation  in the radium-226 values.  In BM-3
radium-226 ranged  from 0.1  to 70.8  pCi/L, with an  arithmetic
mean of  38.9  pCi/L  (Figure  5).  There is a good positive
correlation  between  radium-226  and  radon in this well (Table  2
and Figure 5).  Other  researchers have  found groundwater levels
of radium-226 to  be  generally  lower than average in the
northeastern United  States  (10) and that radium does not
necessarily  correlate  with  radon  (9,  11).

     Radium-228 values for  all  wells  yield arithmetic means of
between  0.5  and 0.7  pCi/L.   However,  with error bars greater
than the reported values  in most  cases,  it is  difficult to
ascertain variation in the  radium-228.

AIR RADON

     Air radon results are  shown  in Figure  6.  Alpha track
detectors from the Terradex Corporation were  placed  in  the
kitchen  on the wall near the sink and on a wall  in the  basement
of  each  test home.   Homes ED-15 and ED-18  are less than five
years  old,  BM-3 was built in 1970,  ED-10 about 1935,  and MS-48
 in  1928  (but on a foundation built around 1900).   Except  in
home  ED-15,  there is a washer and dryer in  each  basement.   All
basement walls are concrete except ED-10,  which  is built  of
unmortared fieldstones.

      Coefficients of variation for monthly alpha track  readings
 are large,  ranging from 0.34 to 1.27 since September.   Basement
 air radon at home ED-10 has an arithmetic mean of 10.8  pCi/L.
 From a value of 7.5 pCi/L for  September it rose  to 18.5 pCi/L
 for November, then declined.   Kitchen air radon  began with its
 highest value - 3.1 pCi/L - then declined.   Its  arithmetic mean

-------
                  BM-3  The "high" well
u
Q.


c
o


°
600000
500000 -
400000 -
300000 h
200000 -
100000 -
              10/27     12/26     2/24


       1991              Date of test
                                            4/24     6/23


                                             1992
    Figure 5.  Radium-226  and radon in well BM-3.

-------
                            Air radon
o
   16
o   12
Q.
     4
     0
               ED-10
        OCT  DEC   FEB  APR


            Month  of test
                               o
                               c


                               "D
                               o
12 h



 8



 4



 0
                                             ED-15
                                      OCT   DEC  FEB  APR


                                          Month of test
               ED-18
                                             MS-48
o
CL


c
o
  a:
        OCT   DEC  FEB   APR


            Month of test
                                  c
                                  0
                                  T3

                                  O
                                       OCT  DEC   FEB   APR


                                           Month of test
                              Legend
                       O Kitchen air radon


                       • Basement air radon
     Figure  6.  Air radon from four week alpha tracks.  Note

   changes in vertical scales.  BM-3 is not represented here, as

                its air radon has been mitigated.

-------
is 1.9 pCi/L.  Water radon for this well has increased overall,
although it declined from mid-December through late February
(Figure 4).

     The large increase of water radon in ED-15 during
December, 1991, occurred during a month in which kitchen air
radon declined slightly.  The ten-fold increase in basement air
radon in Ed-15 in March 1992 is an anomaly, as there were no
significant changes in water radon or in kitchen air radon.

     Home ED-18 had a steady increase in kitchen air radon from
October through March, while the water radon decreased from
September through December, then increased.  The basement air
radon generally decreased during this time.

     The low well, MS-48, began with the highest kitchen radon
(6.3 pCi/L) at a time the water radon was declining from a
high.  Except for the first month, the kitchen and basement air
radon are very similar  (Figure 6).

                          DISCUSSION

     The solubility of uranium and radium in water is partially
limited by the Eh and pH of the ground water  (9).  Radon,
however, being an inert gas, is not so affected.  Comparisons
of radon levels in well water with Eh and pH of the water gave
mixed results.  When the pH and radon results from all phases
of testing are combined for each well, only the high well, BM-
3, shows a statistically significant correlation between pH and
radon  (Table 2).  As an examination of Table 2 shows,
correlations between pH and radon for the individual phases are
quite variable.

     Statistically significant correlations between Eh and
radon occur  for all welli, when all of the data is combined,
although ED-10 has a positive correlation while the other four
are  negative.  As with  pH, the individual correlations for Eh
within the three phases of the study are variable  (Table 2).

     For the combined results, when the relationship between
dissolved oxygen and radon is examined for each well, there are
three  significant negative correlations, in wells BM-3, ED-10,
and  MS-48.   Within the  individual phases, dissolved oxygen -
radon  correlations occur in ten of the twenty tests  (Table 2).

     The  alkalinity tests showed statistically significant
positive  correlations between alkalinity and radon in all wells
except ED-10,  for the combined data.

     For  the  individual  testing phases, the largest number of
correlations between radon and water chemistry occurred with

-------
conductance, a measure of total dissolved solids (Table 2).
These were both positive and negative correlations.  When the
combined test results are examined, statistically significant
correlations between conductance and radon occur only in BM-3
and ED-18; for BM-3 the correlation is positive, for ED-18 it
is negative.

There seem to be no consistent significant correlations of pH,
Eh, dissolved oxygen, alkalinity or conductance with radon in
all five wells.  Szabo and  Zapacza  (9) reached this same
conclusion for groundwaters in the Newark Basin.

     Statistically there was correlation between water use and
radon on only four of the ten days of hourly testing, three
positive and one negative.  However, water use seemed to have a
dramatic effect on early morning radon levels in all except the
low well, MS-48, as radon levels rose quickly during the first
few hours of the water use  day  (Figure 1) .  The low well, MS-
48, however, had positive correlations between radon and water
use for hourly testing on both days.  BM-3 had a statistically
significant positive correlation between radon and water use on
one day of the hourly tests and ED-15 had a negative
correlation on one day.  The other  six days of tests produced
no significant correlations.

     On the daily tests MS-48 and BM-3 had negative
correlations between radon  and water use, while ED-10 had a
positive correlation  (Table 2).  On the weekly tests there are
significant positive correlations between water use and radon
for wells ED-10, ED-15 and  ED-18.   This variation  seems to
indicate there is no relationship between radon and water use
other than the early morning depression of water radon levels.

     The large variation  in water radon levels found in the
high well, BM-3, is not reflected  in any of the water
parameters  for which testing was done, including water use and
rainfall.  The variation may be caused by seasonal changes in
groundwater  levels which  bring  in radon-rich water at some
water levels and radon-poor water  at others.  The  well
penetrates  235 feet of granite which  is known to be highly
pegmatitic  (4) .  Some  of  the water  moving into the well must
pass through a pegmatite  rich  in uranium and/or radium bearing
minerals.  However, comparison  of  rainfall with radon does not
reveal  any  recurring pattern  of rainfall amounts before or
during  the  times of  sudden  low  radon  levels.  Measurements of
groundwater  levels  in  the  area  or water  levels  in  the well
itself  may  prove useful  in  testing  this theory, as may
examination  of  the  bedrock  mineralogy to  look for radon
sources and  transport  paths.

-------
AIR RADON

     Air radon results are summarized in Figure 6.  Both
basement and kitchen air radon levels have generally decreased
in home ED-10, while the water radon levels have gradually
increased over the September to June weekly testing period.
The nearby ED-15 home also has an overall decrease in air
radon, while the water radon has also decreased.  Home ED-18
has mixed results for air radon.  After an initial drop,
kitchen radon steadily increased while basement radon decreased
overall.  The water radon decreased from September through
December, then increased (Figure 4).

     Kitchen and basement air radon are remarkably similar to
each other at the home of the low well (MS-48).  The kitchen
radon declined rapidly at the beginning of the study, then
increased slightly in December.  Basement radon has increased
slightly throughout the study.  At the same time, water radon
decreased through September and October,  then increased (Figure
4).

     These air data are being evaluated for the water borne
radon contribution.  Of the four homes in this study in which
air radon was monitored along with water radon, it appears that
in the kitchens of homes ED-15, ED-18 and MS-48 a significant
amount of the air radon could be coming from the water.  In
those homes kitchen air radon value trends seem to follow the
trends of the water radon values.  Basement air values also
follow the water trends in homes ED-15 and MS-48, possibly due
to air flow between'the two levels of the homes.  A full
diagnosis of the homes is required in order to identify
dominant indoor air radon sources originating from water,  soil
or bedrock sources.

                 EARLY IMPLICATIONS OF THE STUDY

     This study has demonstrated that variation occurs in the
radon levels found in water wells.  Early morning well samples
are likely to give abnormally low radon results for that well.
Wells tested later the same day may yield radon results as much
as 58% higher than the early morning test.   Seasonal variations
were also found in well water radon.  Wells ED-18 and MS-48 had
lower water radon in early winter, followed by increases
beginning about mid to late December.  Water radon in well ED-
10 has increased overall throughout phase three of the study,
although with some variability.  The high well,  BM-3, had three
sudden large decreases followed by more gradual increases in
its water radon levels (Figure 4).

     These results indicate a need for multiple testing of
wells being considered for mitigation,  to determine the maximum

-------
range of radon levels  likely to occur in that well.  The high
well, BM-3, is an excellent example of the need for multiple
testing.  If it had been tested while at one of its low values,
mitigation equipment might have been designed for a threshold
of 40,000 pCi/L of radon from the water.  When the levels rose
to 600,000 pCi/L the equipment would likely be inadequate.
Because this well's second low period lasted nine weeks,
seasonal tests may avoid multiple tests being done in one of
these lows.

     Variations in radium-226 levels in BM-3 were also
significant enough to  cause the radium problem to be overlooked
if initial testing yielded a low value  (Figure 5).

     The results of this study to date indicate that a single
well water radon analysis is not necessarily representative of
the radon character of the well.  Multiple water radon tests
over several seasons are likely to be required to assess the
range of radon values.  These values can vary by as much as 94%
for an individual well.  Radon values have been found to be
most stable from mid-morning to mid-afternoon, possibly due in
part to water use and  well recharge dynamics.

                        ACKNOWLEDGMENTS

     The authors especially wish to thank Denis Healy of the
U.S. Geological Survey for his  invaluable help and
encouragement, without which this study would not have been
possible.  And we also wish to  thank the owners of the wells
used in the study for  their gracious participation and
cooperation.

-------
                          REFERENCES

1. Dupuy,C.J., Healy, D., Thomas,  M.A.,  Brown, D.R.,
   Siniscalchi, A.J. and Dembek, Z.F.  A survey of naturally
   occurring radionucleides in groundwater in selected bedrock
   aquifers in Connecticut and implications for public health
   policy.  In: Gilbert, C.E. and Calabrese, E.J., (eds.)/
   Regulating drinking water quality. Lewis Publishers,
   Chelsea, Michigan, 1992. p. 95.

2. Rothney, L.M.  A survey of radon-222 occurrence in
   Connecticut private well water; assessing geologic and
   hydrologic parameters.  M.S. Thesis,  Yale University, 1987.
   160 pp.

3. Torgersen, T., Mackie, D., and Benoit, J.  Geologic control
   of Rn-222 concentrations in groundwater and its impact on
   the health risks of indoor Rn-222.  Unpublished report,
   1988.  29 pp.

4. Gates, R.M.  The bedrock geology of the Woodbury
   Quadrangle.  State Geological and Natural History Survey of
   Connecticut, Quadrangle Report 3, 1954.  23 pp.

5. Dixon, H.R. and Pessl, F., Jr.   Geologic map of the Hampton
   Quadrangle, Windham County, Connecticut.  U.S. Geological
   Survey, Map GQ-468, 1966.

6. Snyder, G.L.  Bedrock geology of the Willimantic
   Quadrangle, Connecticut.  U.S.  Geological Survey, Map
   GQ-355, 1964.

7. Rodgers, J.  Bedrock geologic map of Connecticut.
   Connecticut Geological and Natural History Survey, 1985.
   2 sheets.

8. Radon in water sampling program.  EPA/EERF-MANUAL 78-1,
   U.S. Environmental Protection Agency, Montgomery, Alabama.
   11 pp, 1978.

9. Szabo, Z. and Zapecza, O.S.  Geologic and geochemical
   factors controlling uranium, radium-226, and radon-222 in
   groundwater, Newark Basin, New Jersey.  In:  Gundersen,
   L.C.S. and Wanty, R.B., (eds.), Field studies of radon in
   rocks, soils and waters.  U.S.  Geological Survey Bulletin
   1971: 243, 1991.

-------
10. Paulsen, R.T.   Radionuclides in ground water,  rock and
   soil, and indoor air of the northeastern United States and
   southeastern  Canada — a literature review and summary of
   data.  In: Gundersen,  L.C.S.  and Wanty, R.B. (ed.), Field
   studies of radon in rocks,  soils and waters.  U.S.
   Geological Survey Bulletin 1971: 195, 1991.

11. Szabo, Z. and  Zapecza, O.S.   Relation between natural
   radionuclide  activities and chemical constituents in ground
   water in the  Newark Basin,  New Jersey.  In; Graves, B.
   (ed.), Radon, radium,  and other radioactivity in ground
   water: hydrogeologic impact and application to indoor
   airborne contamination.  Proceedings of the NWWA Conference.
   Somerset, NJ, 1987. p. 283.
                   • U.S COVERNMEVTPWNTINCOFFICI:! 992 -6<.e -00 3/60 021

-------