United States      Office of Research and Development
           Environmental Protection  Research Triangle Park NC 27711    September 1992
           Agency        Office of Air and Radiation
                     Washington DC 20001
vvEPA     The  1992 International
           Symposium on Radon
           and  Radon  Reduction
           Technology:

           Volume 3.   Preprints
           Session VII: Radon Reduction
             Methods

           Session VIII: Radon Occurrence in
             the Natural Environment

           Session IX: Radon Surveys

           Session X: Radon in Schools
             and Large Buildings
           September 22-25,1992
           Sheraton Park Place Hotel
           Minneapolis, Minnesota

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

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

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

                 and

    U. S. Environmental Protection Agency
       Office of Radiation Programs

                 and

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

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


                              Oral  Papers


Session I:  Radon-Related Health Studies

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

Influence of Meteorological Factors on the Radon Concentration
   in Norwegian Dwellings
      Terje Strand and N. B0hmer, Norwegian National Institute
       of Radiation Hygiene	VIIIP-1

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

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

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

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

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

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

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

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

Radon in Houses Around the Plomin Coal Fired Power Plant
      N. Lokobauer, Z. Franic,  A. Bauman, and D. Horvat,
      University of Zagreb, Croatia	  IXP-5

A Radon Survey at Some Radioactive Sites in India
      Jaspal Singh,  L. Singh, S. Singh, and H. Virk, Guru Nanak
      Dev University, India	  IXP-6

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

-------
                                                                     VIM
 DURABILITY OF  SUB-SLAB  DEPRESSURIZATION RADON MITIGATION  SYSTEMS
                        IN FLORIDA HOUSES

                   C. E. Roessler and R.  Varas
         Dept.  of Environmental Engineering Sciences and
     D. E. Hintenlang, Dept. of Nuclear Engineering Sciences
          University of Florida,  Gainesville, FL  32611
                            ABSTRACT

     Durability of sub-slab depressurization radon mitigation
systems installed in demonstration projects in nine North Florida
slab-on-grade houses is being studied.  Installations include
several combinations of single suction point and two suction
point, single fan systems.  All the systems were operating
without any reported equipment failures after post-installation
periods ranging from 22 to 37 months.   In terms of indoor radon
concentrations during the six months of follow-up (Fall-Winter
1991-92),  eight of the nine systems were effective in maintaining
levels below 4 pCi/L in houses that had pre-mitigation values in
the range of 9 to 25 pCi/L and seven of the systems were as
effective as or more effective than when initially installed.  In
one house that had a pre-mitigation concentration of 30 pCi/L,
the six-month follow-up concentration was 4.4 pCi/L as compared
to the early post-mitigation concentration of 2.5 pCi/L.  Evi-
dence of a sub-slab "short-circuit" at this house is being
investigated.

          This work was funded in part by Florida State
          University System, Board of Regents Radon
          Research Program Contract 089037, by the
          Florida Department of Community Affairs Radon
          Research Program, and by U.S. Environmental
          Protection Agency Assistance Agreements CR
          814925 and CR 817367.

          This paper has been reviewed in accordance
          with the U.S. Environmental Protection Agen-
          cy's peer and administrative review policies
          and approved for presentation and publica-
          tion.

-------
 DURABILITY OF SUB-SLAB DEPRESSURIZATION RADON MITIGATION SYSTEMS
                        IN FLORIDA HOUSES
                           INTRODUCTION

     This is a preliminary report of a follow-up and durability
study of sub-slab depressurization  (SSD) radon1 mitigation sys-
tems installed in nine existing North Florida houses during 1989-
90.

     In North Central Florida, the housing stock is primarily
slab-on-grade and the sub-slab medium typically consists of
native soil and sand. SSD systems were installed in nine houses
between May 1989 and August 1990 as part of a demonstration-
research project to evaluate radon mitigation techniques for this
region.

     Several questions were posed concerning durability for
systems operating under Florida conditions.  Would continued
operation impact the sub-slab environment in a manner that
affects the continued effectiveness of the system?  If there were
effects on the sub-slab environment, would these have structural
effects on the building?  Since system flow rates are typically
low because of the low permeability of the sub-slab medium/ would
continued performance of the fans be compromised by the low flow
and by the high temperature in Florida installations?

     Early system performance was followed through December 1990
(for post- installation periods ranging from 4+ to 19+ months).
A follow-up study of 1-year duration was initiated September
1991; at the time of this writing, two calendar quarters of data
were available.
       BACKGROUND:  SUMMARY OF HOUSES  AND MITIGATION SYSTEMS

HOUSES

    •The study involved nine slab-on-grade houses, six in
Gainesville (Alachua County) and three  in Ocala (Marion County).
The general location is shown in Figure 1; characteristics of the
study houses are summarized in Table 1.  Seven had simple
rectangular floor plans with slab areas on the order of 150 to
200 m2 (1600 to 2100 ft2)  and two had more complex, L-shaped
designs with slab areas on the order of 200 to 210 m2 (2100 to
2200 ft2) .   Several combinations  of slab and wall  type  are
         this report,  the terms "radon"  and "Rn" are used to
designate the radon-222 isotope.

-------
                      Tallahassee
Gainesville  (Alachua Co.)
         Ocala (Marion  Co.)
                                              Jacksonville
                                                         Miami
     Figure 1.   Location of Study Houses.

-------
                                    TABLE 1. FEATURES OF STUDY HOUSES & RADON MITIGATION SYSTEMS.
House Number
House Characteristics:
Age (yrs)
Slab footprint
Size, sq m
Size, sq ft
Type
Wall Construction
Air Handler Location
Return Air Route
Sub-slab Communication
Pre-mitigation Rn, pCi/L:
Subslab
Indoor
Rn Mitigation System Characteristics:
Installation Date:
System A: Fan
Pit #1 diameter, m
Pit #2 diameter, m
System B: Fan
Pit diameter, m
Continuous Data Acquisition
Post-mitigation Indoor Rn, pCi/L

O-l

14
Rectang.
167
1800
F
CB
UTIL
ATTIC
Poor

3880
20

May '89
R-150
0.92
0.92
—
—
no
1.5

O-2

17
Rectang.
164
1760
S
FR
GARAGE
ATTTC+TW
Good

3940
11

May '89
R-150
0.92
0.82
—
—
no
2.1
270
G-l

9
Rectang.
164
1760
s,
FR/BV+CB
GARAGE
TW
Poor

2820
12

Jul '89
R-150
0.92
0.84
—
—
yes
3.5
7.641
G-2

14
Rectang.
194
2087
F
CB/VN
GARAGE
ATTIC
Poor

4270
26

Nov '89
R-150
0.92
0.52
—
—
yes
2.5

G-3

19
Rectang.
158
1700
F
CB
CLOSET
TW(Door)
Good

8600
9

Oct '89
T-l
0.92
—
—
—
yes
1.4

G-4

20
Rectang.
181
1950
F
CB/BV
CLOSET
ATTIC
Fair

3260
20

May '90
R-150
0.61
0.66
	
—
yes
2.0

O-3

14
Rectang.
149
1608
P
CB
ATTIC
ATTIC
Fair

9260
30

Aug '90
R-150
0.71
0.71
	
	
no
2.0

G-5

9
L-shaped
195
2100
S
FR+STV
GARAGE
ATTIC+TW
Fair

4300
25

Jul '90 *
R-150
0.86
0.71
EXP
0.56
no
2.5

G-6

13
L-shaped
203
2188
F
CB
GARAGE
TW
Good

3400
10

Ju! '90 *
R-150
0.61
0.81

...
no
2.5

       Slab type:
       Wall construction:
       Return air:
       Sub-slab communication:
       Pressure Extension:

       Fan Type:
O = Ocala, (Marion Co.), G = Gainesville, (Alachua Co.).
F = floating, S = sealed stem-wall, P = partially sealed stem-wall
FR = wood frame, CB = concrete block, BV = brick veneer, STV = stone veneer, VN = vinyl siding
TW = through wall with no ducting, ATTIC = ducts running through attic
Poor = pressure extension of 0-0.3 m (0-1 ft), Fair = 0.3-3 m (1-10 ft), Good = greater than 3 m (10 ft).
The farthest distance at which a pressure differential of at least 0.25mm of water can be measured when a pressure
of 500 Pa is applied at a suction hole.
R-150=Fantech model R-150, 270 cfm (4.47 cubic m/min), 2150 rpm, 1/20 hp (37.3 J/s);
T-l = R.B. Kanalflakt Turbo model, 158 cfm (4.47 cubic m/min), 2800 rpm, 1/40 hp (18.6 J/s); EXP = Experimental.
* G-5 and G-6 systems turned on in July'90, became fully effective in October '90.

-------
represented.  All had central heating/air conditioning systems
and several return air configurations are represented.

     Diagnostic testing prior to mitigation system design
included sub-slab communication measurements.  Effective
permeabilities of the sub-slab media, determined in-situ from
observations of flow vs suction pressure, were on the order of
10"11 to 10"10 m2.  Pressure extension was defined as the farthest
distance at which  a pressure differential of at least 0.25 Pa
could be measured when a pressure of 500 Pa was applied at a
suction hole.  Communication values were equally divided among
poor (pressure extension of 0-0.3m or  0-1 ft), fair
(pressure extension 0.3 -3m or 1-10 ft) and good (pressure
extension >3 m or >10 ft).   Thus in a number of houses, the SSD
systems had to deal with communication distances considerably
less than half the major dimension of the house.

     Sub-slab radon concentrations as measured by grab samples
during diagnostic testing ranged from 2800 to 9300 pCi/L.
Average pre-mitigation indoor radon concentrations ranged from 9
to 30 pCi/L.

MITIGATION SYSTEMS

     The systems were installed between May 1989 and August 1990;
the installation schedule is diagramed in Figure 2.  Character-
istics of the SSD systems as installed are summarized in Table 1.
Systems with two suction points were installed in eight of the
houses.  Only a single suction point was installed in house
Gainesville-3 — it was of smaller dimensions  (slab area = 160 m2
or 1700 ft2) ,  had good sub-slab communication,  and had a con-
venient location for a central suction point.  A single-point,
single-fan system was installed in addition to a two-point,
single-fan system in house Gainesville-5 — this house had a
complex floor plan with several slab rectangles and only fair
sub-slab communication.

     Each suction point consisted of a roughly hemispherical pit
approximately 0.5 to 1.0 m in diameter.  PVC pipes 10 cm (4 in.)
in diameter were installed through the slabs and 5 cm  (2 in.) PVC
pipes were routed to the fans which were located in the attic
with discharge vertically through the roof.

INITIAL SYSTEM PERFORMANCE

     Initial system performance is summarized  in Table 2 (also
see Figure 3).  The systems were generally effective in reducing
indoor radon concentrations from pre-mitigation levels in the
range of 10-30 pCi/L to post-mitigation values of <4 pCi/L.
Levels were reduced to values on the order of  2 pCi/L or less in
four houses, (see post-mitig. column in Table  2).

-------
         EARLY MONITORING PERIOD
                           FOLLOW-UP PERIOD
                            Fall Win. Sor. Sur
        system not effective until Oct. '90
        system not effective until Oct. "90
          1989
1990
1991
1992
Figure 2.  Radon Mitigation System Installation and Follow-up Time Line.

-------
TABLE 2. INDOOR RADON CONCENTRATION IN STUDY HOUSES.
House
No.

G-l
G-2
G-3
G-4
G-5*
G-6*
O-l
O-2
O-3
House
Shape

Rectang.
Rectang.
Rectang.
Rectang.
L-shaped
L-shaped
Rectang.
Rectang.
Rectang.
Number
of
Suction
Points

2
2
1
2
3
2
2
2
2
Date
System
Began
Operation

Jul-89
Nov-89
Oct-89
May-90
Jul-90
Jul-90
May-89
May-89
Aug-90
Months of
Operation
as of
Jun-92

35
31
32
25
23
23
37
37
22
Radon Concentrations in pCi/L
Pre-
Mitig.

12
26
9
20
25
10
20
11
30
Post-
Mitig.

3.5
2.5
1.4
2.0
2.5
2.5
1.5
2.1
2.0
Fall-91
80-day
A-T

2.2
2.0
1.6
2.5
4.3
3.3
0.6 #
1.8 @
6.3
Wint-92
90-day
A-T

1.6
1.3
0.6
1.4
2.1
1.8
0.6 #
1.2
2.5
Fall-
Winter
Average

1.9
1.7
1.1
2.0
3.2
2.6
0.6
1.5
4.4
Icey to House No.: G = Gainesville, (Alachua Co.), O = Ocala, (Marion Co.)
* G-5 & G-6 systems turned on in July; installation fully effective in October 1990.
# O-l Indoor readings for Fall '91 and Winter '92 are from a single 170-day reading.
@ O-2 : Monitored for the last 35 days of the Fall '91 period.

-------
                       APPROACH AND METHODS

EARLY PERFORMANCE MONITORING  (1989-90)

     Following the phased-in installation and tuning of the
mitigation systems between May 1989 and August 1990, performance
was followed through the end of 1990  (See Figure 2).  Indoor
radon was monitored by continuous2 and/or integrating3 radon
monitors, and house visits were conducted to observe system
operation, measure system pressures and flows, service radon
monitoring equipment, and receive and respond to homeowner
questions and problems.  Post-installation monitoring was per-
formed according to the following general schedule:

   • Stage 1 (in service <6 months) - Continuous and/or
     integrating radon monitoring; house visits on a bi-weekly
     basis.
   • Stage 2 (in service 6 to 12 months) - Houses without data
     loggers were visited on a monthly schedule. For houses with
     data loggers, data acquisition was continued, data were
     reviewed,  and visits were performed as necessary.
   • Stage 3 (in service >12 months) - Visits on an approximately
     6-month schedule to inspect the systems, measure pressures
     and flows, and deploy radon monitors for week-long
     measurements.

FOLLOW-UP STUDY (1991-92)

     The durability follow-up phase, initiated in September 1991
for a 1-yr duration, involved  (1)  indoor radon monitoring to
determine continued effectiveness in radon mitigation and (2)
quarterly house visits to observe system operation, measure
system pressures and flows, observe house conditions, and receive
and respond to homeowner questions and comments.

     Radon monitoring is being performed, with passive integrating
(alpha track) detectors4 deployed  for  four consecutive calendar
     2Five houses were instrumented with continuous data
acquisition systems (Model 21X Micrologger®, Campbell Scientific,
Inc., Logan, UT); radon detectors were At Ease® Model 1021 radon
monitors  (Sun Nuclear Corporation, Melbourne, PL).  In the four
houses without continuous data acquisition systems, At Ease®
Model 1023 radon monitors were deployed as stand-alone monitors.

     3Electret ionization chambers (E-PERM®,  Rad Elec Inc.,
Fredrick, MD).

     4Rad-trak®,  Tech/Ops Landauer,  Inc.,  Glenwood, IL.

-------
quarters.  Two detectors are deployed in each house,  one in each
of two separate rooms in the living area.


                             RESULTS

EARLY PERFORMANCE

     This early post-installation phase of durability monitoring
from the time of installation until the end of 1990 involved
durations of 4 to 19 months for the various houses.  During this
limited period, the following were observed:

  1. With the transient exceptions noted below, the systems
     settled to relatively constant performance after an initial
     adjustment/stabilization period and retained their
     effectiveness in maintaining reduced indoor radon
     concentrations.
  2. In one house, failure of caulking of a leakage crack near a
     suction point resulted in "short-circuit" flows.  This was
     remediated by re-caulking with a more durable material.
  3. With the advent of cool weather in 1989, condensation from
     moist exhausted sub-slab air occurred in three systems that
     had undrained low points in horizontal attic runs.  This
     resulted in an audible gurgling noise and compromised
     performance  (reduced flow and fluctuations in indoor radon
     concentrations). This was alleviated in these systems by
     installing condensate drain lines from the "traps" and was
     prevented in subsequent installations by ensuring self-
     draining through avoidance of low points where possible
     and/or the installation of drain lines.
  4. No fan failures were observed.  Any effect of low flow on
     fan life was not expressed in this limited sample during the
     available observation period.
  5. No structural effects were observed.
  6. With the exception of the "gurgling" associated with the
     water condensation before installation of traps and drains,
     there were no homeowner complaints of noise or other
     annoyances.

FOLLOW-UP STUDY (1991-92)

     This 1-yr monitoring phase began in September 1991 after the
systems had been in operation for periods ranging from 13 to 28
months. At that time, all the systems were operating and there
were no homeowner complaints.

     As of June 1992, house visits had been performed for the
Fall, Winter, and Spring quarters and the systems had been in
operation for periods ranging from 22 to 37 months.  Indoor radon
data were available for two quarterly monitoring periods.  Radon

-------
data are presented and compared to pre-mitigation and early post-
mitigation concentrations in Table 2 and Figure 3.  The post-
mitigation and follow-up study concentrations are presented on an
expanded scale in Figure 4.

     A striking feature of these data is the fact that all the
indoor radon concentrations for the Fall quarter were greater
than those for the winter quarter  — Fall/Winter ratios ranged
from 1.4 to 2.7 and averaged 1.9.  This is the inverse of the
seasonal effect seen in unmitigated houses in two previous
Gainesville studies (1,2) and in other Florida cities (2).
Measurements of pressure and flow did not indicate any apparent
system differences between the two quarters.  Comparison of exit
air radon concentrations did not indicate any consistent differ-
ences between quarters that could be associated with the Fall-
Winter indoor radon differences.  To date, no satisfactory
explanation has been proposed.

     In all houses, the follow-up indoor radon concentrations (6-
month average) remained significantly less than the pre-mitiga-
tion values.  In most houses (8 of 9),  the concentrations were <4
pCi/L.  In 7 of the 9 houses, the concentrations were comparable
to or lower than the early post-mitigation reference concentra-
tions; this suggests that these systems are at least as effective
as when initially installed.

     In the two exceptions,  both mitigated in the second half of
1990, the Fall quarter concentrations were >4 pCi/L and
noticeably higher than the early post-mitigation values.  How-
ever, these follow-up/reference comparisons should be interpreted
with reservation.  The early post-mitigation reference values for
1990 may be biased by seasonal effects — especially regarding
Gainesville-5, Gainesville-6, and Ocala-3 for which the post-
mitigation observation period was confined to Fall (or late Fall)
and December 1990.

     Ocala-3, a rectangular house with a 2-point, single-fan
system, was the only house with a Fall-Winter average exceeding 4
pCi/L.  The average of 4.4 pCi/L, based on a Fall concentration
of 6.3 pCi/L and a Winter value of 2.5 pCi/L, was greater than
the 2.0 pCi/L early reference value from observations during
September-December 1990.  At this house, the post-mitigation flow
rate at one of the suction points was higher than typical for
these systems and about 20 times that at the other suction point;
this suggested some "short-circuiting".  At the follow-up, the
flow imbalance was even greater and the radon concentration in
the exit air was less than 100 pCi/L.  This is strong circum-
stantial evidence for intensified "short-circuiting" with
excessive dilution by house air drawn through a slab leakage
and/or ambient air drawn through a pathway to the foundation
perimeter.  This is being investigated.

-------
4 pd/L max.
              O1    G3   O2  G2   G1   G4   G6  G5   03

       Pre-Mitig.
Post-Mitig.
Fall '91
Winter'91-'92
Figure'3.  Pre-mitigation and Post-mitigation Indoor Radon Concentrations.

-------
          4pCi/L EPA suggested maximum concentration
          Ay:*V^«OT:-.V/;Xl«ySy-lY/_V:ViffiSS»S5^
                                                                  PRE-MITIGATED

                                                                  CONCENTRATIONS
             O1    G3   O2   G2   G1   G4   G6   G5   O3
Post-Mitigation I    I Fall '91
                                                      Winter'91-'92
Figure 4.  Post-mitigation and Follow-up Indoor Radon Concentrations.

-------
     In the remaining house, Gainesville-5, the Fall 1992 value
was slightly greater than 4 pCi/L and the Fall-Winter average of
3.2 pCi/L was slightly greater than the late 1990 early reference
value of 2.5 pCi/L.  These comparisons may or may not be sig-
nificant — the early post-mitigation value was based on limited
observations during late Fall and December of 1990 and the
follow-up study phase was characterized by a 1991-92 Fall/Winter
seasonal ratio of >1.0 when a value of <1.0 would have been
expected from previous experience.

     System flows and pressures and other observations will be
presented in a forthcoming report.


                     SUMMARY AND CONCLUSIONS

     Durability studies are being performed on several configura-
tions of SSD systems installed during 1989 and 1990 in nine North'
Florida slab-on-grade houses.  Early observations were made from
the respective times of installation to the end of 1990 and a
follow-up study of 1-yr duration was initiated in the Fall of
1991.  Follow-up information to date includes indoor radon
monitoring  (continuous integrating sampling) data for the Fall
1991 and Winter 1991-92 quarters and other observations into June
1992, and represents post-installation periods ranging from 22 to
37 months for the various houses.  The study demonstrated that:

   1. All the systems remained in operation without any reported
     failures.

   2. All of the systems continued to provide significant
     reductions from the pre-mitigation concentrations.

   3. Eight  of the  nine systems  (89%) were  effective in main-
     taining indoor radon concentrations below 4 pCi/L in houses
     that had pre-mitigation values  in the range of 9 to 25
     pCi/L.

   4. In terms of indoor radon concentrations, seven of the nine
     systems  (78%) were as  effective as or more  effective than
     when  initially  installed.

   5. In the eighth house  (Gainesville-5),  the system was
     effective  in  that the  Fall-Winter average concentration  was
     3.2 pCi/L  as  compared  to the pre-mitigation value of  30
     pCi/L.  However, comparison  to  early  effectiveness  (2.5
     pCi/L) was inconclusive  since this comparison may be  con-
     founded by year-to-year  differences  in  seasonal effects.
   6. The ninth  system  (Ocala-3),  a two-suction point system,
     showed evidence of a  sub-slab "short  circuit"  that  com-
     promised performance.  This house  had  a  Fall-Winter  6-month

-------
     average indoor radon concentration of 4.4 pCi/L as compared
     to the initial post-mitigation reference value of 2.0 pCi/L.
     This case is being further investigated.

  7. During the first cold period following some of the 1989
     installations, unintended trapping of moisture condensation
     in horizontal attic runs compromised performance and drains
     had to be installed in these systems.  Lessons learned were:
        • Moisture condensation in the SSD system during cool
          weather is a potential problem, even in Florida.
        • New installations should be designed to be self-
          draining through avoidance of low points ("traps")
          where possible and/or by installation of drain lines.
        • Maintenance should include inspections for these and
          other inadvertent effects.

  8. The Fall 1991 indoor radon concentration was greater than
     the Winter 1991-92 value in all of these houses; the average
     ratio was 1.9.  This is a reversal of the "conventional
     wisdom" about seasonal effects in unmitigated Florida
     houses.
                           REFERENCES

 1.  Roessler, C.E., Revell,  J.W. and Wen, M.J. Temporal Patterns
     of Indoor Radon in North Central Florida and Comparison of
     Short-term Monitoring to Long-term Averages.  In; Proceed-
     ings: The 1990 International Symposium on Radon and Radon
     Reduction Technology: Volume 1.  EPA-600/9-91-026b (NTIS
     PB91-23444).   U.S. Environmental Protection Agency, Research
     Triangle Park, North Carolina,  1991.  P. 3-131.

2.    McDonough, S., Williamson, A. and Sanchez, D.C.  Correlation
     Between Short- and Long-term Indoor Radon Concentrations in
     Florida Houses.  In; Proceedings:  The 1991 International
     Symposium on Radon and Radon Reduction Technology: Volume 3.
     (EPA-600/9-91-037C (NTIS PB92-115377) U.S. Environmental
     Protection Agency, Research Triangle Park, North Carolina
     1991.  p. PNL3-21.

-------
                                                                   VI1-2
       A NOVEL BASEMENT PRESSURIZATION-ENERGY CONSERVATION
             SYSTEM FOR RESIDENTIAL RADON MITIGATION
            by:  K. J. Renken and S.  J.  Konopacki
                 University of Wisconsin-Milwaukee
                 Radon Research Laboratory
                 Department of Mechanical Engineering
                 P.O. Box 784
                 Milwaukee, Wisconsin 53201
                 U.S.A.
                            ABSTRACT


     Basement pressurization is one method of radon mitigation
which has vast potential.  It has been theorized that
pressurization could provide significant reduction in radon
levels for a house with a "tight" basement.  The largest drawback
of the technique is the energy penalty associated with increased
air infiltration during the heating season.

     This paper presents the results of an investigation on a
system that produces radon reduction by basement pressurization
while conserving heating fuel and providing improved air quality
as well as human comfort.  A secondary heat exchanger that
increases the efficiency of a conventional residential furnace is
modified to provide fresh air heat exchange and pressure
regulation.  The system has subsequently been named "The Radon
Buster".  A Wisconsin home with an initial average radon level of
33.5 pCi/L was chosen for the retrofit.  Following the
installation of the unit, the average radon concentration
diminished to 5.7 pCi/L and the heating bill reduced enough to
operate the system at an estimated zero cost.

     Measurements of radon reduction levels, fuel usage, and
environmental factors that affect radon migration are documented.
A state-of-the-art radon data acquisition system with
accompanying instrumentation are also described.

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

-------
                           INTRODUCTION


     The method of basement pressurization has received very
little attention in the radon mitigation literature.  The method
prevents radon soil gas from entering a building by providing a
positive pressure shield which causes air to flow through cracks
and holes in the basement out into the soil.  The method
simultaneously reduces radon concentrations by diluting the radon
laden air with fresh air supplied by a blower.  Long term data
has not been published to show the effectiveness and reliability
of the technique, its relationship with energy costs or its
lengthy structural effects.

     A review of pertinent investigations on the subject reveals
only a handful of papers that can be associated with this study.
One of these is a study of natural basement ventilation by
Cavallo et al. (1).  They have shown the method to be effective
as a radon reduction measure when the basement is slightly
pressurized and highly diluted.  However, they discovered that
during the heating and cooling seasons a severe energy penalty
was assessed since the outdoor air must be either heated or
cooled to the indoor air temperature.

     The Heat Recovery Ventilator (HRV) method is another proven
technique that reduces radon levels by combining basement
pressurization with dilution.  Heat Recovery Ventilation employs
an air-to-air heat exchanger to recover energy from stale
radon-laden indoor air with fresh outdoor air.  The reduction
will solely be attributed to dilution if the inlet and outlet air
streams are of equal flow rate.  Basement pressurization can be
realized if the inlet air stream is at a higher flow rate than
the outlet air stream.  However, this method imposes a demanding
energy penalty during the heating and cooling seasons.  A study
utilizing a balanced HRV by Nazaroff et al. (2) concluded that
radon reduction below the EPA guidelines could be reached with
this method.  Another study employing a balanced HRV by Holub et
al. (3) determined the same method could reduce radon levels by
about a factor of six.

     Brennan et al. (4) achieved radon reduction through basement
pressurization by sealing the return ductwork and creating an
opening in the supply ductwork.  Their results showed that the
method was as effective as soil depressurization at controlling
indoor radon levels.

     The objectives of our research were (1) to develop an
effective system of radon mitigation using basement
pressurization with the additional benefit of heating energy
conservation and (2) to characterize the environmental effects
(e.g.  pressure differentials, soil and ambient temperatures,

-------
precipitation, barometric pressure, wind velocity and direction,
humidity, etc.) that influence indoor radon entry and egression.
The final product of our research efforts has been termed "The
Radon Buster."

                   DESCRIPTION OP RADON BUSTER


     A secondary heat exchanger is the main component of the
Radon Buster design.  In general, the purpose of a secondary heat
exchanger is to improve the heating efficiency of industrial
boilers, unit heaters and residential conventional furnaces by
increasing the heating capacity of the existing heating
equipment, thus reducing fuel consumption.

     In our design, the secondary heat exchanger is arranged in
counter flow orientation with unmixed air and flue gas streams.
The flue gas is distributed through a copper header (20" long, 3"
O.D.) into 23 sinusoidal copper tubes (5' straight length, 7/8"
O.D.), recombined in another header and exhausted to the outdoors
with an exhaust blower.  The external dimensions of the heat
exchanger are 40" x 20" x 10".  A fan is required to move the
flue outdoors because it is no longer hot enough to rise through
the entire length of the chimney, therefore requiring the chimney
to be capped off.  The return air moves through the heat
exchanger and is preheated before entering the furnace whdre it
is then heated to a much higher temperature than the normal feed
air.  This results in an elevated supply air temperature causing
a reduction in furnace operation, which decreases fuel
consumption and heating costs.  A backdraft damper is placed on
the outlet end of the flue pipe to prevent outdoor air from
entering.

     The conventional furnace/secondary heat exchanger system
operates in the following manner.  The home thermostat calls the
furnace for heat and closes an electrical path to the exhaust
blower.  The blower switches on and creates a vacuum within the
secondary heat exchanger and subsequently closes the vacuum
switch.  When the vacuum switch closes it allows the electronic
ignition to ignite the pilot and the furnace operates normally.
The vacuum switch is a safety feature which will not allow the
furnace to operate if the exhaust blower fails.  As the
thermostat opens and calls for the furnace to cease operation the
furnace blower continues to circulate air through the house until
the flue inlet header in the secondary heat exchanger cools to
135°F.  At this point a thermostat on the flue inlet header opens
and cuts power off from the furnace blower.

     A Radon Buster is a modified secondary heat exchanger with
the same internal and external dimensions which is attached to
the furnace in the same way.  The major difference is that the

-------
outdoor air is the supply air instead of the return air.  Cold
outdoor air moves through the Radon Buster and is preheated then
deposited into the basement.  The thermostat on the flue inlet
header is electrically connected to the air side blower on the
heat exchanger instead of the furnace blower.  A check valve is
situated at the air side outlet obstructing the flow of air to
the outdoors while maintaining positive basement pressure.  The
Radon Buster is illustrated schematically in Figure 1.

                 DESCRIPTION OF  EXPERIMENTAL  SITE
     The experimental site was the residence of Peter and Judy
Roidt located in Hartland, Wisconsin (35 miles west of Milwaukee,
Wisconsin).   The house is a 35 year old conventional single floor
ranch style over a full basement, slightly shielded by trees and
other houses on three sides and open in front.  The basement is
adequately sealed from the upstairs living area by a tight
plywood sub-floor, and consists of a poured concrete floor,
hollow-block walls, two sump wells, and four tightly sealed well
windows.  A water drainage system exists along the east end of
the basement where a thin slice of the slab was removed and weep
holes were drilled.  The weep holes and the sump wells are
assumed to be the major entry routes for the indoor radon gas.
Installed in the basement was a Thermopride gas fired furnace
with electronic ignition and a heating capacity of 80,000 Btu/hr.
The upstairs storm windows and doors as well as the basement
windows and stairway were tightly sealed by adding weather
stripping and industrial silicone.  This measure preserved the
positive pressure created by the Radon Buster.  The house volume
was measured to be approximately 19,200 ft3.   A whole  house
blower door test determined the natural ventilation rate to be
0.30 ACHnat and the effective leakage area to be  122 in2 (Lawrence
Berkeley Laboratory Air Infiltration Model).

                         INSTRUMENTATION
     The test site was continuously monitored for radon
concentration, differential pressures, and indoor/outdoor
environmental parameters under normal living conditions from
October 1991 to April 1992.  Figure 2 shows schematically the
locations of the various sensors that were used during our
measurements.

     A Hewlett-Packard Vectra QS/16S computer interfaced with a
Hewlett Packard 75000 Series 3 Mainframe provided the means for
full automation of the data acquisition.  The mainframe was
fitted with three Hewlett-Packard plug-in modules which supplied
the interface for the instrumentation.  The first was a 5 1/2

-------
        31 Flue Pipe (Golv. Steel
Supply   —]
Air Duct
                 Gas-Fired
                 Furnoce
                 80,000
                                                                                   Exhaust
                                                                                   Blowr
 Control
 Ponel

Vocuun —
Safety
Sir itch
                                                              Heat Exchanger
                                                               Iforn Air
                                                               Outlet
•a
•n
                                                                                     Condensate •
                                                                                     Drain
                                                                 • Return
                                                                  Air Duct
                                  3'  Flue Pipe (PVC)—j
                                  (to Outside)
                                                                                                                                 Inlet
                                                                                                                                 Blo*er
                                                                                                                                 300 cfn
                                    8' Cold Air Inlet—'
                                    (Golv, Steel)
                                    (Iron Outside)
                                                               Figure  1.   Rodon  Buster

-------
  Enclosure
10
                                                                 Wind Breok
                                                         2,5


Upper Level
1 7












/
J






\

\






(













Ro
Bu


Fu







do
st


rn







n
er
1 	 I

oce

















Conputer ond
Instrunentotion
-Bosenent . . c K 7 0 n
1, 4, b, o, /, 8, 11


Supply Air Duct 1, 12



\
Bosenent B
1 - Tenperoture
2 - Soil Tenperoture
3 - Boronetric Pressure
4 - Relative Hunidity
                                                         5
          5 -  SoiI / fiosenent Pressure
          6 -  Outdoor / Bosenent Pressure
          7 -  Upstairs / Basenent Pressure
          8 -  Basenent / Basenent Pressure
 9 - Precipitation
10 - Kind Velocity  / Direction
11 - Rodon Concentration
12 - Airflow
                                                 Figure 2.  Location of Instrunentotion

-------
digit multimeter which connected a 16-channel thermocouple relay
multiplexer used to interface continuous voltage and current
output sensors to the mainframe.  The last plug-in module was a
3-channel universal counter employed to connect the voltage pulse
output sensors.

     The menu-driven data acquisition software package "Labtech
Notebook" performed all the necessary data collection and
internal conversions of output to true engineering units for each
sensor and executed calculations between channels.  Readings were
taken at 5 minute intervals which allowed for two continuous
weeks of data collection before the computer memory was filled.
Excursions to the test site were taken every two weeks to
down-load the data and check on equipment performance.

     Radon concentrations in the basement were measured by a
calibrated Pylon AB-5 continuous radon monitor which used a 300
ml scintillation cell.  The unit was connected to the 3-channel
universal counter card of the data acquisition mainframe.  In an
effort to verify the readings of the Pylon unit for indoor radon
level, several Tech/Ops Landauer, Inc. Long Term Alpha Track
Radon Gas Monitors were logistically placed in the basement and
the upstairs.

     Outdoor weather conditions were recorded in an effort to
determine the parameters affecting radon entry.  Outdoor
temperature was measured with a Type-T thermocouple probe
enclosed in an environmental chamber that shielded various
sensors from radiation, wind, rain, snow, insects, animals, and
intrusion.

     The soil temperature was measured in two locations: on the
east side of the house at distances of about 1' and 20' from the
base of the house.  The soil probe was fashioned from a metal
protection tube fastened to an aluminum protection head and
inserted 18" into the ground.

     Outdoor relative humidity was measured with a current output
relative humidity transmitter (General Eastern) which was also
placed in the environmental chamber.

     A tipping bucket rain gauge  (Climatronics Corp.) was
enclosed within a 4' high circular windbreak.  The presence of
the windbreak reduced eddy currents and turbulence around the
gauge providing for a more accurate precipitation measurement.
The rain gauge was mounted level, in the center of the windbreak
and located in a clear area away from trees and buildings.  A
thermostatically controlled heater mounted on the unit prevented
it from freezing and allowed snow fall to be measured as well as
rain.

-------
     An in-line wind speed and direction transducer (Earth and
Atmospheric Sciences, Inc.) was mounted upon a 20' stand away
from obstructions and connected to the data acquisition system.

     Barometric pressure was measured with a low differential
pressure transmitter (Earth and Atmospheric Sciences,  Inc.).  The
transmitter was located in the basement with the high port
residing in the environmental chamber connected by means of 1/4"
copper refrigeration tubing with the other port sealed.

     During pre-mitigation and post-mitigation data collection,
conditions within the house were monitored to observe what affect
the operation of the Radon Buster had on them.  Relative humidity
in the basement was measured with a relative humidity transmitter
of the same type as the outdoor sensor.  Temperature was measured
in both the basement away  from the Radon Buster and in the
upstairs living area.

     One of the functions  of the Radon Buster is to pressurize
the basement to impede radon gas entry.  Differential pressure
measurements were made to  examine the change in pressure with the
addition of the mitigation system.  Pressure differential was
measured with very  low differential pressure transducers  (Setra
Systems, Inc.) with  a range of +/- 0.25 inches H2O (+/- 62 Pa)
between the basement near  the Radon Buster  (low port) with
respect to other positions (high port).

     Pressure measurements were collected between the basement
and the soil using  soil probes.  Each soil probe was constructed
from a well point fitted to 1/4" copper refrigeration tubing with
a compression fitting, which in turn was sealed to 1/4"
polyethylene tubing and connected to the high port of a
transducer.  These  transducer units were placed 4' in the ground
in four positions around the perimeter of the house: west,  south,
southeast, and east (north side of test site had  a one car
garage).  The outdoor/basement pressure differential process  line
ran from the high side of  a transducer to the environmental
chamber with 1/4" polyethylene tubing.  Differential pressures
were also measured  between the basement and the upstairs  living
area as well as between the Radon Buster area and the  far side of
the basement away from the unit.

     The Radon Buster was  instrumented with four  thermocouples
which  were  inserted into  the  inlet and outlet sides of both the
air and flue gas  streams.   A  hot wire  anemometer  transducer (TSI,
Inc.)  was  employed  to measure  air velocity  of the air  inlet
stream which was  converted to  a volumetric  air  flow rate.

    ' Finally,  a Gas Alert Plus Alarm was  installed to  check for
carbon monoxide  emissions from the  furnace.

-------
                      DISCUSSION OF RESULTS

PRE-MITIGATION RESULTS

     Radon screening and follow-up measurements revealed an
average pre-mitigation radon concentration in the basement of the
test site of 33.5 pCi/L as shown in Table 1.
       TABLE  1.   PRE-MITIGATION AVERAGE RADON CONCENTRATION

Detector Type   	Date	        Location    Radon FpCi/LI
Charcoal
3/16/91
- 3/18/91
Basement
Family Room
35.0
22.0
Alpha Track    11/12/91 - 1/18/92
Pylon Monitor   9/11/91 - 9/12/91
               10/18/91 - 1/17/92
Basement        32.7
Family Room     21.9

Basement        33.5
Basement        33.6
     Radon concentration was measured continuously starting
October 18, 1991 with the Pylon monitor.  Figure 3 displays the
extremes in radon levels for the period of November 22 through
December 2, 1991.  The radon levels range from 10 pCi/L to 60
pCi/L with an arithmetic mean of 33.6 pCi/L.

     Differential pressure measurements between the basement and
the surrounding soil indicated the average soil pressure to be
about 3 Pa higher than that of the basement as evidenced in
Figure 4.  This magnitude of pressure difference is indicative of
the natural state of basement depressurization.

     Environmental parameters were also studied for their effect
on radon entry.  During our pre-mitigation measurements, the
following environmental effects displayed no direct correlation
with radon entry:  indoor/outdoor relative humidity, soil
temperature, and wind direction.  On the other hand, the
following parameters did show noticeable relationships with
indoor radon levels:  precipitation, barometric pressure, and
natural ventilation rate.

     Precipitation events greater than 1/4" during the period of
December 20, 1991 through January 2, 1992 were analyzed.  For 12
independent events where precipitation was a dominant variable,
we found that the radon concentration increased by 10% - 40% with
the most common value being 30%.  These events demonstrated a
direct relationship between precipitation and an increase in
radon concentration which supports the work of Montague and

-------
       0    1
                          Time [Day]
      Figure 3.   Pre-Mitigation Radon Concentration
                 (11/22/91 - 12/2/91)
  CO
     -2
                        345
                          Time [Day]
8
Figure 4.  Pre-Mitigation Soil / Basement Differential
           Pressure  (12/4/91 - 12/11/91)

-------
Belanger (5).   Also observed is a delay of up to four hours
between the initial rise in radon concentration and the onset of
precipitation.  A plot illustrating the relationship between
radon concentration and precipitation is shown in Figure 5.

     An inverse relationship between barometric pressure and
radon concentration is observed in Figure 6.  A fall in
barometric pressure is usually attributed to precipitation, but
for this particular event no precipitation was measured.
Therefore, it is hypothesized that barometric pressure was the
primary driving force for this event.  Barometric pressure will
normally communicate through a building relatively quickly.  A
decrease in barometric pressure will then allow soil gas to
migrate into buildings and the atmosphere, thus increasing indoor
radon levels.   Whereas an increase in barometric pressure will
retard soil gas movement decreasing the radon concentration.  A
similar inverse correlation between the indoor radon
concentration and the rate of change of barometric pressure was
also observed in the basement of a house by Hernandez et al. (6).

     The natural ventilation rate of a building is a function of
indoor/outdoor temperature differential and wind velocity.  The
relationship between the natural ventilation rate and indoor
radon concentration is highlighted in Figure 7.  On this
particular day, there was no measurable amount of precipitation.
It has been reported by Nazaroff and Nero (7) that an increase in
the natural ventilation rate will increase the depressurization
of the building which will subsequently increase the rate of
radon entry.  However, in the majority of our measurements the
natural ventilation rate was not easily identified as the major
transport mechanism.  It did not correlate well with the degree
of radon entry confirming the work of Nero et al.  (8) and Doyle
et al. (9).

POST-MITIGATION RESULTS


     The Radon Buster was installed on January 18, 1992.  The
airflow rate through the unit was measured at 300 cfm  (0.94 ACH).

     A 10 day period from February 19 through February  28,  1992
was examined for post-mitigation effects.  Figure 8 displays
basement radon concentrations for that time period.
Unmistakably, the radon concentration had taken a drastic
reduction from pre-mitigation levels.  The post-mitigation  levels
ranged from 1 pCi/L to 24 pCi/L with an arithmetic mean of  5.7
pCi/L.  This resulted in an average radon reduction of  83%.  It
is evident that the fluctuations in radon are more systematic
than that of pre-mitigation levels.  The reason is that the
cyclic operation of the Radon Buster is the dominant radon  entry
control variable: causing radon egression during the on mode and

-------
                                                    c
                                                    o
                                                    •*—'
                                                    CD
                                                    •^
                                                    'g.
                                                    'o
                                                    CD

                                                    ol
                                                fO.5
                        12/21/91
 Figure 5,
Pre-Mitigation Radon Concentration vs.
Precipitation (12/21/91)
                                                r31.9
O
Q.
c
O

2
•*-

CD
O
C
o
O
c
o
-a
nj
cc
                                                 31.6
                        12/27/91
 Figure 6.
Pre-Mitigation Radon Concentration vs
Barometric Pressure  (12/27/91)

-------
                                                120
-r  40-
o
Q.
I	1

c
o

75
-*—*

0>
O
C
o
o



I
CO
CE
Figure 7
                                                70
                       12/23/91

            Pre-Mitigation Radon Concentration vs.

            Natural  Ventilation Rate  (12/23/91)
     0
          1
                   3456

                        Time [Day]
8
  Figure 8.   Post-Mitigation Radon Concentration

              (2/19/92 - 2/28/92)

-------
radon migration while in the off mode.

     Figures 9 and 10 present radon concentration as a function
of Radon Buster operation on a daily basis  (the spiked dotted
line represents Radon Buster operation).  It is obvious from
these figures that radon concentration  is reduced when the system
is operating.  Each day in the 10 day time period was examined
for minimum and maximum radon levels while the Radon Buster was
in the on mode and off mode, respectively.  The minimum values
(on mode) revealed a range in radon concentration of 1 pCi/L to 4
pCi/L with an arithmetic mean of 1.8 pCi/L and a standard
deviation of 0.7 pCi/L.  This indicates that if the Radon Buster
operates continuously the radon concentration would remain around
1 pCi/L.  The maximum values (off mode) were in a range of 6
pCi/L to 24 pCi/L with an arithmetic mean of 12.8 pCi/L and a
standard deviation of 4.4 pCi/L.  On a percentage basis, a radon
reduction of 97% is feasible with continuous operation of the
Radon Buster.

     Differential pressure measurements of the soil with respect
to the basement for this same time period are displayed in Figure
11.  Highlighted in this figure is the positive basement pressure
with respect to the surrounding soil (as indicated by the data
points which are less than 0).   This can be credited to Radon
Buster operation as seen in Figure 12 where Radon Buster
operation is plotted versus differential pressure.  The
mitigation system is blowing in 300 cfm of outdoor air creating a
positive pressure field within the basement.  Typically, while
the Radon Buster is on, the outdoor and soil pressures are about
5 Pa lower than the basement pressure and the upstairs is about 4
Pa lower.  The differential pressure measurement across the
basement where the low port was near the Radon Buster and the
high side was in another isolated basement room indicated a
negligible pressure difference.

     A concern of this mitigation system was the delivery of cold
air to the basement through the Radon Buster.  For this 10 day
period the outdoor temperature ranged from 23°F to 57°F and the
resultant basement temperature was 57°F to 67°F.

     Another concern was raising the indoor relative humidity to
a level where mold and rot would begin to develop on the interior
of the building shell.  This fear was dispelled since the
basement relative humidity remained stable at about 45%
throughout the entire experiment.

     The heat exchanger sensible heat effectiveness was about
0.75.   The latent heat effectiveness,  and therefore the total
heat effectiveness is unknown since water content in the flue gas
was -not measured,  but it was higher than the sensible
effectiveness since a steady flow of condensate drained from the

-------
                            2/19/92


   Figure  9.   Post-Mitigation Radon Concentration vs.

               Radon Buster Operation  (2/19/92)
    O
    Q.

    C
    O

    "
    c
    0)
    O
    c
    O
    O
    c
    O
    T5
    cn
    cc
Figure 10,
               2/20/92


Post-Mitigation Radon  Concentration vs.

Radon Buster Operation (2/20/92)

-------
         6-r
              1
                                 8
10
                  234567

                            Time [Day]

Figure 11.   Post-Mitigation Soil / Basement  Differential
             Pressure (2/19/92 - 2/28/92)
     CO
     D_
     w
     CD


     CD
     W
     CD
     m
       0
Figure 12
                                           O


                                           o


                                           CD
                                           O)
                                           u
                                           m
                                           c
                                           o
                                           -o
                                           CO
                                           DC
                2/21/92


Post-Mitigation  Soil /  Basement Differential
Pressure vs.  Radon  Buster Operation  (2/21/92)

-------
flue gas side.

     The initial investment of this radon mitigation system is
about $2000.  The system is estimated to cost  nothing to operate,
since the cost of electricity to run the fans  is  offset by a
decrease in heating fuel consumption.  This  is illustrated in the
following example with an estimate based on  the 1991/1992 heating
season in the Milwaukee, Wisconsin area.

Assumptions:  Heating Degree Days  (HDD) = 6600°F/days
              Natural Gas Cost  (Fcosl)  = 0.55 $/Therm
              Electricity Cost  (Ecost)  = 0.0677  $/KW/hr
              Hours of Operation  (HO) = 1500 hr/yr

     Table 2 presents the Roidt's pre-mitigation  natural gas
consumption in fuel usage per heating degree day  (Fpre) .


         TABLE 2.   PRE-MITIGATION NATURAL GAS  CONSUMPTION
                     £>
                   Date	Fpre	

                 11/18/91     0.1018
                 12/19/91     0.1033
                  1/23/92     0.1034


     The mean value  (Fprem)  for this 3 month period is 0.1028
Therms/°F day.  The post-mitigation natural  gas consumption (F^,,)
is 0.0982 Therms/°F day.  The net reduction  in heating fuel
consumption (F,^,)  is  given  by,

            Fred =  Fpre.m - Fpojl                                    (1)

which is 0.0046 Therms/°F day.
        The cost reduction  (CR)  from waste heat recovery is given
by,

            CR = F^,  Fcost HDD                                    (2)

which equals  $17.00.

     The energy penalty comes from fan  operation.  On the air
side a Dayton split capacitor blower with  a  free  air capacity of
488 cfm at 1580 rpm and 157 watts  (Pair)  was used,  and on the flue
gas side an ITT Jabsco  electric  blower  with  a  free air delivery
of 150 cfm at 2240 rpm  and  78 watts  (Pnue) •   It is assumed that
half the power consumed by  the  air side blower is recovered in
the form of heat from the motor  and  air imparted  to the fan
blades, which will warm the incoming air  stream.   Hence, the

-------
 factor  of  0.5  in  the  following  equation describing fan operating
 cost  (CF).

            CF =  (0.5 Pair  +  Pnue)  Ecost HO /  1000                   (3)

 The factor of  1000 converts KW  to  Watts.  The cost of operating
 the fans is $16.00.

      Based on  this experimental site  it costs $16 per heating
 season  to operate the fans  and  $17  is recovered by the heat
 exchanger, therefore there  is no estimated net cost to operate
 the Radon Buster.  With a subsurface  ventilation system there
 exists  no payback mechanism and the customer must forever pay for
 fan operation.

                            CONCLUSIONS


      The Radon Buster has proven itself an effective method of
 radon mitigation  with the additional  benefits of energy
 conservation and  an increased indoor  air quality.  It is one of
 the first radon mitigation  systems to boast of these attributes.

      The unit reduced radon concentration levels by a combination
 of basement pressurization  and dilution.  Radon gas
 concentrations dropped from an average pre-mitigation level of
 33.5  pCi/L to  an  average post-mitigation level of 5.7 pCi/L
 during  intermittent operation.  It is believed that if the system
 is run  continuously, the radon  levels would remain at around 1
 pCi/L which is well below the EPA  standard of 4 pCi/L.

      Fuel consumption decreased even  though cold outside air was
 blown into the basement.  This reduction in fuel costs was enough
 to compensate for the energy cost  from operating the fans.

      Moreover, an additional benefit  of the Radon Buster which
was reported to us by the occupants of the test site was that
their home was more comfortable in cold weather now than prior to
the installation of the Radon Buster.  This is attributed to the
device's ability to hold a  positive indoor pressure which
prevents the infiltration of local cold air draft discomforts and
 its introduction of more oxygen (due  to the outdoor fresh air
 intake)  within the living environment.

                        ACKNOWLEDGEMENTS


      Financial support for  this research provided by the U.S.
Environmental Protection Agency through the 1991 State Indoor
Radon Grant (SIRG) No. K1995011-02 is greatly appreciated.  The

-------
authors would like to thank Peter and Judy Roidt for the use  of
their home, Robert C. Brauer for the fabrication of the Radon
Buster, Dr. Conrad Weiffenbach with the Wisconsin Department  of
Health and Social Services for his ideas and suggestions,  Paul
Tellier with the Waukesha County Department of Health  for  helping
us locate a test site, and Tech/Ops Landauer, Inc. of  Glenwood,
Illinois as well as Gas Alert Plus of Berkeley, Illinois for
their equipment donations.

                           REFERENCES


1.  Cavallo, A., Gadsby, K. and Reddy, T.A.  Natural basement
    ventilation as a radon mitigation technique.   In.:
    Proceedings of The 1991 International  Symposium  on Radon  and
    Radon Reduction Technology, Philadelphia, Pennsylvania, Vol.
    2 Preprints, 1991.

2.  Nazaroff, W.W., Boegel, M.L., Hollowell, C.D.  and  Roseme,
    G.D.  The use  of mechanical ventilation with heat  recovery
    for controlling radon  and radon-daughter concentrations in
    houses.  Atmospheric Environment.   15:  263,  1981.

3.  Holub, R.F., Borak, T.B., Droullard,  R.F. and  Inkret,  W.C.
    Radon-222 and  222Rn progeny concentrations measured in  an
    energy-efficient house equipped  with a heat exchanger.
    Health Physics.  49: 267,  1985.

4.  Brennan, T., Brodhead,  W.  and Dyess,  T.M.   Preliminary
    results  of  HVAC  system modifications to control  indoor radon
    concentrations.   In;   Proceedings of The  1991  International
    Symposium on Radon  and Radon  Reduction Technology,
    Philadelphia,  Pennsylvania, Vol.  4  Preprints,  1991.

5.  Montague, A. and Belanger,  W.E.   Effects  of humidity and
    rainfall on radon  levels  in a residential  dwelling.   In:
    Proceedings of The  1991 International Symposium on Radon and
    Radon Reduction Technology,  Philadelphia,  Pennsylvania, Vol.
    3  Preprints,  1991.

6.  Hernandez,  T.L.,  Ring, J.W.,  and Sachs,  H.   The variation of
    basement radon concentration  with barometric pressure.
    Health Phvsics.   46:  440,  1984.

7.  Nazaroff,  W.W. and Nero,  A.V.  Soil as a source of indoor
    radon:  generation,  migration, and entry.   In:  Radon and  Its
    Decay products in Indoor Air.  John Wiley & Sons, New York.
    p.  57.

-------
8.  Nero, A.V.,  Boegel, M.L., Hollowell, C.D., Ingersoll, J.G.
    and Nazaroff, W.W.  Radon concentrations and infiltration
    rates measured in conventional and energy-efficient houses.
    Health Physics.  45: 401, 1983.

9.  Doyle, S.M., Nazaroff, W.W. and Nero, A.V.  Time-averaged
    indoor Rn concentrations and infiltration rates sampled in
    four U.S. cities.  Health Physics.  47: 579, 1984.

-------
                                                                                  VII-3
   THE ENERGY PENALTY OF SUB-SLAB DEPRESSURIZATION RADON MITIGATION SYSTEMS

            by:    David. L Bohac and Timothy S.  Dunsworth
                  Center for Energy and the Urban Environment
                  Minneapolis,  MN  55403

                  Lester S. Shen
                  Underground Space Center
                  University of Minnesota
                  Minneapolis,  MN  55455

                  Christopher J. Damm
                  Department of Mechanical Engineering
                  University of Minnesota
                  Minneapolis,  MN  55455
                                   ABSTRACT

      The purpose of this paper was to quantify the energy penalty incurred by
sub-slab depressurization radon mitigation systems in a cold, heating-
dominated climate.  Two sets of houses were studied.  The first set of houses
consisted of houses with previously installed mitigation systems which had
been in operation for at least one year.  For the second set, mitigation
systems were installed in five houses during the fall of 1990 and their fuel
usages were monitored over the subsequent heating season.  The average pre-
mitigation consumption of the 11 houses was 1210 ccf.  Since the average
percent increase in total gas usage for the combined results was about 6%, the
average gas usage penalty was 76 ccf per year or $38 per year, assuming a cost
of 50$ per ccf of natural gas.  With an average annual electrical cost of $37
per year, the average total energy cost was $75.

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

-------
   THE ENERGY PENALTY OF SUB-SLAB DEPRESSURIZATION RADON MITIGATION SYSTEMS
                                 INTRODUCTION

      This project measures the impact of the operation of radon mitigation
systems on residential energy use.  The project studied single family,
detached houses located in the Twin Cities Metropolitan area, a cold, heating-
dominated climate of 8,050 heating degree days (°F-day, base 65°F).  The
mitigation systems were limited to active soil depressurization (ASD) systems.
The systems consisted of sub-slab depressurization or a combination of sub-
slab depressurization with either concrete block wall suction or crawl space
floor depressurization.  ASD radon mitigation systems can affect the building
energy performance in several ways.  Firstly, there is the inherent electrical
energy cost resulting from the continuous operation of the mitigation system
exhaust fan.  Concomitant effects of the system include increased building air
infiltration and substructure heat loss.  The augmented infiltration is a
result of house depressurization caused by air flow from the lower level of
the house into the mitigation system.  The system also draws outdoor air
through the soil surrounding the substructure.  During the heating season,
this results in reduced soil temperatures and increased substructure heat
loss.  The mitigation system may also reduce soil moisture throughout the year
which can impact the moisture and latent heat effects in the house.

      Previous investigations to quantify the associated energy penalty of
radon mitigation systems have been reported in the literature (1,2,3,4).
These studies have estimated the mitigation energy penalty by measuring the
exhaust flow and extrapolating the added heating costs.  With typical
mitigation flows, this corresponds to a heating penalty of $16 to $125 per
year depending on local climate and fuel costs, with an additional fan
operation penalty of comparable size plus a cooling penalty in cases with
central air conditioning.  Because the Twin Cities area has a heating-
dominated climate (8050 heating degree days versus 662 cooling degree days in
a typical year) with a low penetration of central air conditioning, the
emphasis will be on heating and fan operation penalties.

      To ascertain the magnitude of the energy penalty and moisture effects,
two sets of houses were studied.  The first set consisted of houses with
previously installed mitigation systems that had been in operation for at
least one year.  Changes in energy use were determined through a pre/post-
mitigation fuel bill analysis employing the Princeton Scorekeeping Method
(PRISM).  For the second set,  five houses were selected from a preliminary
sample of houses that had performed long-term alpha track detector (ATD) tests
and had measured radon levels greater than 5 picoCuries per liter (pCi/1)  in
the lower level of the house.   Mitigation systems were installed in these

-------
houses during the fall of 1990.  Fuel use was monitored over the subsequent
heating season as the ASD systems were alternately enabled and disabled.
                                 METHODOLOGY

      This project was divided into two separate analyses:  an energy analysis
of previously mitigated houses and detailed monitoring of five houses which
were mitigated as part of the project.  The analysis of the previously
mitigated houses allows the comparison of results from two methods of direct
impact measurement, as well as increases the sample size without a
corresponding increase in monitoring costs.  From the fuel bill analysis of
the first set of houses, the impact of mitigation systems on whole house
natural gas use could be determined.  This use includes space and water
heating use and for some houses, cooking use and clothes drying.  The detailed
monitoring of the second set of houses permitted closer scrutiny of the impact
of the ASD system on electrical and natural gas space heating consumption.
Site visits were conducted on both sets of houses to measure mitigation system
air flows, entrained air flows from the living area, and house and system
moisture levels in an effort to better quantify the heat loss mechanisms
induced by the ASD systems.  Results of the air flow measurements will not be
described here but are included in the project final report (5).  The
following discussion describes in detail the methodology used for the
investigation of the two sets of houses.

ENERGY ANALYSIS OF THE PREVIOUSLY MITIGATED HOUSES

      An analysis of previously mitigated homes was initiated to determine the
magnitude of the whole house energy penalty incurred by the mitigation
systems.  A fuel bill analysis comparing the post-mitigation energy use with
the pre-mitigation use provides an indication of changes in the energy
consumption that may have been affected in part by the mitigation system.
Since whole-house natural gas use data is analyzed for these houses, any
changes in consumption due to occupancy effects, building structure, and the
heating system will act to mask the mitigation system effects.

Identification of Candidate Houses

      To find houses in the Twin Cities (TC) area that had previously received
radon mitigation work, the eight contractors in the TC area who had passed the
EPA Radon Contractors Proficiency Program  (RCPP) exam were contacted to get
access to likely candidates.  It was found that the majority of mitigation
jobs performed in the TC area resulted from relocations and that relatively
few established homeowners opted for mitigation work.  This made it difficult
to find houses with a year of pre-mitigation data and a year of post-
mitigation data from the same occupants.  Nevertheless, three of the
contractors were able to provide 14 houses for analysis.  Six of these houses
fit the selection criteria and were included in the study.

-------
 Evaluation  of  Energy Use  Using  PRISM

      Natural  gas  utility bill  records of the pre- and post-mitigation energy
 use were obtained  for each house.  Evaluation of the energy use of the houses
 was performed  using  the Princeton Scorekeeping Method (PRISM)  (6).  Using the
 PRISM computer program is the standard practice for quantifying the energy
 usage in residential buildings.  The program calculates a weather-normalized
 annual energy  consumption which is a measure of the fuel consumption of a
 house under average  weather conditions.  The method is based on a model that
 assumes that below a certain house-specific reference temperature, the fuel
 consumption per day  of a  house  varies linearly with the daily average outdoor
 temperature.   Above  this  reference temperature, the daily consumption is taken
 as a constant.  In reality, non-heating use often shows seasonal variation
 that is imperfectly  correlated  with degree days, but this simplified model has
 been found  to  work extremely well for analyzing natural gas use in heating-
 dominated regions  of the  country.  Various studies have also found that
 outside temperature  is generally an adequate explanatory variable without the
 addition of other  weather variables, such as wind, humidity, or insolation.

      PRISM uses a statistical  regression procedure which uses the actual
 meter readings of  a  house and the local daily average outdoor temperatures.
 The model calculates the  annual fuel use corresponding to long-term average
 local daily outdoor  temperatures.  This index is known as the Normalized
 Annual Consumption (NAG).   Comparing the NAC value of the post-mitigation year
 with the pre-mitigation year should provide a measure of the increased energy
 use resulting  from the mitigation system, adjusted for differences in weather
 conditions  of  the  two  years being compared.  In general, PRISM has been found
 to provide  a reliable  index of consumption and has been used to measure energy
 savings in  individual  house retrofits as well as long-term trends in
 residential energy use (6).

 House Screening and  Selection

      Several criteria were placed on the houses that could be in the project.
The restrictions were  instituted to ensure that the relevant energy data was
easily accessible,  to  simplify the analysis with monthly and bimonthly data,
to eliminate the effect of the change in residents on energy use,  and to
provide reliable estimates of energy use.  The criteria  were:

      1.     stable occupancy - The occupants of the houses had to have resided
            in the houses for a full year before the system was installed and
            a full year after with no changes in household size.
      2.     natural gas or electrically heated homes.
      3.     no supplemental space heating.
      4.     good fuel use records.
      5.     no retrofits  to the building structure or the heating system
            during this two year period.
      6.     a record of pre-/post-mitigation radon levels to establish that a
            radon problem had existed and that it  had been satisfactorily
            remedied.

-------
      For the evaluation, two additional criteria affected the eligibility of
the houses in the study.  A house remained in the study if the PRISM analysis
of the pre- and post-mitigation data resulted in an R2 greater than or equal
to 0.95 and a relative standard error of 5% or less (7).

      Six of the 14 houses provided by the contractors fit the selection
criteria and were included in the study.  Of the eight houses that were not
included, five used fuels other than natural gas.  Two of the homeowners did
not respond to correspondence and one resident did not have an entire year of
data from before the mitigation system installation.

Site Visits

      Each of the six houses was visited at least twice by project staff to
inspect the operation of the systems and to determine the physical
characteristics of the houses.  The work performed during these visits
included measurement of the dry and wet bulb temperatures in the below-grade
portion of the building and in the exhaust system, air velocity measurements
in the mitigation system exhaust, measurement of the below-grade area
dimensions, recording of the fan specifications, and interviews with the
homeowner on pre- and post-mitigation radon levels and on any general changes
in the house dealing with energy use and perceived changes in interior
humidity levels.

MONITORING OF THE FIVE TWIN CITIES AREA HOUSES

      While the first set of houses gives a measure of the general change in
whole house energy use after installation of the radon mitigation systems, a
more detailed monitoring of the mitigation systems is necessary to gain a
better understanding of the energy loss mechanisms of the systems.  For this
portion of the project, five houses located in the Twin Cities area were
monitored during the 1990-1991 heating season after an active mitigation
system was installed.

House Selection

      The selection process for the five houses to be mitigated and monitored
in this study was begun in March 1990.  Likely candidates for the study were
obtained from a pool of households that had voluntarily performed long-term
monitoring using alpha track detectors (ATD)  through participation in one of
the following programs: the Minnesota Department of Health survey and
monitoring programs sponsored by the American Lung Association of Minnesota,
Group Health,  and CEUE.  Those participants that were located in the Twin
Cities area and had ATD results greater than 5 pCi/1 in either the first floor
or lowest living level were informed by their sponsoring agency that they
qualified for inclusion in the EPA Innovative Grant study.  Interested parties
were instructed to contact the University of Minnesota.

-------
 House Screenino
       From the selection process, 21 houses were obtained for the initial
 screening.  A survey of these houses showed that an ATD test had typically
 been performed for at least three months on each of the houses, with many
 houses performing a one year ATD.  All the houses used natural gas as their
 primary heat source.  A fuel bill analysis using PRISM was performed on each
 house to determine the fitness of the model in analyzing the 1989-90 pre-
 mitigation heating season.  Based on these results, site visits and homeowner
 interviews were performed on 11 of the houses.  These visits were mostly
 performed during July,  with one house done in early September.  Two to eight
 day E-Perm tests were performed under closed house conditions, when possible,
 to verify the ATD results.

 Radon Diagnostics and Mitigation

       Radon mitigation contractors performed diagnostics on eight of the
 houses.   Four of these houses were selected for inclusion in the study and
 contracts were signed with the participants.   Radon mitigation using active
 soil depressurization was performed on these houses.   Work on three of the
 houses was completed during July and August and the fourth house was completed
 in the beginning of  January.    A fifth house was located using contacts from
 one of the contractors  and was added to the study in  October.   This house also
 received active soil depressurization.

       Each of the homeowners  was concerned  about the  high level of  radon in
 their homes  and had  an  interest  in installing a radon mitigation system prior
 to volunteering for  participation in the study.   Two  of  the  homeowners had,  in
 fact,  previously  received  diagnostic testing  and bids from contractors to
 perform  mitigation services.   Under  the services of this  project, the
 mitigation systems were installed at no charge to the homeowner in  exchange
 for their  cooperation in  the  study.

 House Monitoring

       Immediately after the mitigation  systems were installed,  the  contractors
 placed continuous  radon monitors  (CRM's)  in the  houses to  document  the
 successful operation  of the systems.  After this  post-installation radon
 monitoring was completed, the  systems were turned off until the  house
 monitoring began.  Consequently,  the mitigation  systems were not in operation
 during the fall of 1990.  Beginning  the  first week  of January  1991, the
 mitigation systems were activated  for two weeks.  Mitigation fan electric
 usage and the whole house gas consumption were recorded for this period.  The
 natural gas usages were obtained  from visual reading of the house gas meters
while the electric use of the mitigation  system  fans were obtained from
electric meters connected to the  fans.  After this period, the systems were
deactivated for two weeks.  Two week alternating mode experiments' were
performed in this manner for the remainder of the heating system (end of
June).  The period of two week duration was chosen to provide an adequate
sample of experimental data and to ensure that the heat loss and moisture
transfer characteristics had reached stability.  Discrete time relative

-------
humidity measurements using a sling psychrometer were made in the basement and
living space of each house during the bi-weekly visits.

      Radon monitoring was begun on December 3, 1990 with three E-PERM
monitors placed in each house: one in the basement,  one in the living area,
and one in the attic.  These were collected and replaced every two weeks,
corresponding to the operating mode of the mitigation system.  An EPA-
certified radon measurement contractor provided the E-PERM monitors and the E-
PERMs were measured at the contractor's office.  The monitors were transported
between the sites by project staff and were placed in the houses by staff
according to approved protocols.  Blank and replicate measurements were
performed following the guidelines and protocols established by the Minnesota
QA/QC Plan approved by EPA.
                                    RESULTS

FUEL BILL ANALYSIS OF THE PREVIOUSLY MITIGATED HOUSES

      The results of the PRISM analysis of the six previously mitigated houses
are shown in Table 1.  All of the houses had mitigation systems installed
during 1989 with the exception of house 11 which had its system installed
during the fall of 1988.  Four of the houses have full basements while two
houses are a combination full basement with an adjoining crawl space.  The
average floor area for the four basement houses is 1219 ft2.  For the
combination basement/crawl space foundations, the average basement floor area
is 600 ft2 and the average crawl space floor area is 335 ft2.   Sub-slab
depressurization systems were installed for the basement areas and in the
crawl spaces, floor depressurization systems were installed.  The walls are of
concrete block construction in all six houses.  Air sealing of the basements
was also performed in these houses.  This included sealing of large cracks in
the foundation walls and floors and floor areas open to the earth for water
lines.

      The pre-mitigation data period spanned from 10 to 13 months and started
as early as September 1987 and ended as late as November 1989.  The specific
dates depended on when the system was installed, occupancy variations, and
robustness of the PRISM analysis.  The post-mitigation period spanned from 12
to 13 months, typically incorporating the 1989/90 heating season but also
including a portion of 1990/91 heating season if circumstances warranted.

      For each house, the NAC, the coefficient of variation (CV), and R2 are
provided for both the pre- and post-mitigation years with the values shown in
parentheses being the standard errors of the NAC.  The standard error can also
be taken to be the standard deviation of an estimated statistic, for example,
a regression parameter such as the NAC value. For the purposes of this report,
a 68% confidence interval about the model estimate is defined as ±1 standard
deviation about the estimate and a  95% confidence interval  is defined by ±2
standard deviations.  A 95% confidence interval about the estimate means that
with a 95% certainty, the true long-term (or "population") value lies within

-------
the defined interval.  95% confidence intervals will be reported when
confidence intervals are included with the results.

      The energy penalty incurred by the mitigation system is computed by the
change in the total NAG (ANAC).  Since only natural gas use was studied in
this fuel bill analysis, results are given in ccf of natural gas (which is
very nearly equal to 100,000 Btu or one therm).  The ANAC is a gauge of the
energy penalty of the mitigation system on the whole house natural gas use.
Again, the standard deviations for each value are given in parentheses and the
value given below each standard deviation is the p-value which is a measure of
statistical significance.  For instance, a p-value of 0.05 means that there is
a 95% confidence level that the calculated change in NAG is different from
zero (a two-tailed test) and does not simply fall within the error bounds of
the model estimates.

      Examination of the p-values for each ANAC shows that the calculated ANAC
for houses 12 and 16 are not statistically significant and that statistically
no impact from the ASD system could be observed on their natural gas use.  The
ANAC of house 14 has only a 90% statistical significance while the ANAC's of
the remaining houses have at least a 95% significance.  Examining the energy
use impact on this set of houses as a group, the average pre-mitigation NAC
(whole house use, including heating and base load use), the normalized whole
house energy use, for the six houses was 1206 ccf while the average post-
mitigation NAC was 1291 ccf.  The average increase in NAC was 85 ccf, for an
overall sample increase of 7.0%.  The highest increase was 24.6% and one house
had a slight decrease in usage of 2.5%.  A paired difference t-test on this
group of houses produced a p-value of about 0.03.

MONITORING OF THE FIVE TWIN CITIES AREA HOUSES

House Screening

      Table 2 shows the results of the pre-mitigation radon tests, PRISM
analysis, and house characteristics of the five houses selected for detailed
monitoring.  The long-term ATD results show that elevated radon levels
(exceeding 5 pCi/1) were measured in all five houses.  With the exception of
site 2 which shows a low radon level, the E-PERM tests for the most part
confirm the ATD results.  The short-term E-PERM tests were performed during
the summer and even though an attempt was made to maintain closed house
conditions, greater confidence is placed in the ATD results.

      The average pre-mitigation NAC for the five houses was 1251 ccf.  Three
of the houses are heated by forced air systems (houses 2, 3, and 5) while the
remaining two houses have hydronic systems (houses 1 and 4).  All five houses
have basement foundations, with sites 2 and 4 also having crawl spaces under
adjoining additions.  The average floor area of the basement-only houses is
837 ft2 while the combination foundation houses has an average basement floor
area of 835 ft2 and an average crawl space floor area of 228 ft2.   Site 4 has
a ppured concrete wall foundation while site 1 has both a stone and concrete
block foundation.  The three remaining houses have concrete block foundation
walls.  Sites 2, 4, and 5 are all one-story houses with house 2 having a walk

-------
in basement and a tuck-under garage.  Site 1 is a two story house and house 3
is a story and a half construction.  Each house had a sub-slab
depressurization system installed in the basement area.  Site 5 also had a
wall suction system installed while crawl space floor depressurization was
also used in sites 2 and 4.

Radon Levels

      To gauge system performance of the ASD systems, E-PERM monitors were
placed in the basement, first floor living space, and attic during each two-
week period of the alternating moAe experiment.  Figure 1 shows the average
radon levels for both the system-on and system-off modes for each of the
houses.  The systems were all working effectively with the apparent exception
of the system at site 2.  The other sites show an average radon reduction of
93% on all floors, which is highly significant.

      At site 2, the radon levels were reduced in the first floor when the
system was operating.  However, there were numerous elevated levels that were
measured in the basement.  A continuous monitor (CRM), that had been cross-
calibrated with several other devices, was placed in the basement adjacent to
the E-PERM.  For the two week period when the system was operating, the CRM
recorded radon levels of less than 0.5 pCi/1 when the E-PERM reading was 25.8
pCi/1.  Diagnostics of the house with the system running showed a substantial
low pressure field was being generated under the slab.  It was concluded that
the system was operating properly.  At present, no explanation can be
attributed for the spuriously high E-PERM readings.  Two E-PERM measurements
were performed in the same basement location after May 14 (and, therefore,
after the end of the heating season) and both readings were below 1 pCi/1.

      While Figure 1 illustrates how well the systems are operating, it is
also interesting to note the radon levels in the first floor living space when
the systems are off.  Sites 3 and 5 show a generally gradual attenuation in
radon levels going from the basement to the first floor and to the attic
(except for the attic in house 5) while low levels are observed in both the
first floor and attic in sites 1 and 4.  On average, for sites 3 and 5, the
first floor radon levels with the ASD off is 73% of the basement while it is
14% for sites 1 and 4.  This pattern is a result of the heating systems of
each house.  Sites 3 and 5 have forced air systems while sites 1 and 4 have
hydronic systems.  Duct leakage and air mixing from the forced air system have
resulted in the increased transfer of radon gas from the lowest level to the
living space.  Indeed, for sites 1 and 4, if the basement spaces are not to be
used as habitable space, the need for installing a mitigation system could be
questioned.

      The average attic radon levels for sites 1, 2, and 3 suggest
communication between the living space and the attic.  Sites 2 and 3 show
measurable attic radon levels with the mitigation system operating while in
site 1, attic and living space levels are comparable, at about 0.5 pCi/1 when
the system is off.  A large attic bypass was found in site 1 while appreciable
ice dams above the kneewall areas of site 3 (even with soffit and roof
venting) also indicate the need for attic bypass sealing.  It was learned from

-------
conversations with the occupants of site 2 that a previous history of ice dam
problems resulted in the installation of roof turbines by the homeowners the
previous year.  All of these observations and the measured attic radon levels
are indicative of substantial living space to attic air flows.

Natural Gas Usage

      A method for converting each energy use value to a residual was used to
compare the monitored energy use with the pre-test PRISM model.  The residual
was computed as the difference between the observed use per day in each period
and the use that would be expected based on the pre-test PRISM model and the
test period weather conditions (which can be calculated manually or by using
the PREDICT software developed by CEUE for this type of situation).  This
method also implicitly assumes that the pre-test PRISM model is not grossly
different from the (unknown) test period models.  To the extent that the two
operating modes have roughly similar distributions of weather, this assumption
can be violated somewhat without completely undermining the validity of the
resulting comparisons.  This method also has the advantage of not losing the
information of periods with no heating use (as the use per degree day method
does by excluding such points), nor is this approach liable to exaggerating
any deviations in data with small numbers of degree days (or in the worst
case, a fractional number of degree days per day).

       After rescaling the data by any of these methods, a two sample t-test
was used to compare operating modes at each site.  Single sample t-tests were
used to compare each mode to the pre-mitigation PRISM model (especially the
system-off mode which provides a measure of how consistent the test period
energy use was with the pre-test use).  The two-sample t-tests for all the
sites can be combined by using a log-likelihood method and comparing the
resultant test statistic to a Chi squared distribution.  This helps to
determine whether several statistically non-significant but fairly consistent
results add up to a statistically significant overall result.  For the
purposes of this discussion, statistical significance is defined by a p-value
less than 0.05.

      A summary of the results of the residuals approach are displayed in
Table 3.  Tests of the off-mode versus the pre-test models show no apparent
pattern of altered usage since two cases rose non-significantly, two others
decreased non-significantly, and the pooled test was also non-significant.
While this is not a rigorous test for changes in energy use patterns (since
changes in several model parameters could produce residuals in colder weather
that balance opposite residuals in milder weather), it does increase
confidence in the subsequent test results at least somewhat.  Tests of the on-
mode versus the pre-test model show one significant increase in use (site 5, p
= 0.017), two non-significant increases, and one non-significant decrease.
The pooled test provides a test statistic at the p = 0.035 level.  Two-sample
comparisons between modes have -roughly similar results, with only one
significant individual test (site 5, p = 0.020) and a marginally non-
significant pooled result at the p = 0.081 level.

-------
      The average residuals from the analyses above can be multiplied by 365
days and then added to the pre-mitigation NAG to yield estimates of the
approximate annual energy use with the ASD system on and off.   A change in
annual energy use can then be computed.  As shown in Table 4,  this calculation
yields average values of 1216 ccf with the systems off and 1278 with the
systems on, for an average  gas use penalty of 62 ccf (or 5.3  ± 10.0 percent)
with a p-value of 0.33.  At current gas prices,  this translates to an economic
penalty of $31 per year.

      An average energy penalty for the two sets of houses can be obtained by
combining the results of the two subsets.  The analysis of variance (ANOVA) of
the two subsets of houses revealed that the two sets did not differ
significantly.  Consequently, a paired t-test of all ten houses could be
performed.  The average pre-mitigation energy use for the ten  houses was 1210
ccf while the average post-mitigation use was 1286 ccf.  The paired t-test
showed a statistically significant overall energy penalty of 76 ccf or a
penalty of about $38 per year.  The standard error of the energy penalty
result was 81 ccf.  The p-value for the paired t-test was in the range
0.01
-------
The Chance in Humidity Levels Due to the Mitigation System

      In addition to radon gas, the soil around the building is also a large
source of moisture.  Typically, the air trapped in the pores of the soil is
saturated with moisture in the form of water vapor.  The same mechanisms which
work to draw radon gas into the house also bring in this water vapor.  By
reducing radon in the house, the mitigation system also acts to reduce this
important source of moisture in the building.  During the heating season, the
mitigation system can act to relieve high moisture levels in a house by
reducing the entry of moist soil air and also by drawing in cold dry outside
air from induced air infiltration.  Concurrently, the mitigation system will
have an impact on the latent heating loads of the building.

      In the absence of continuous house moisture monitoring, house moisture
levels were measured on a discrete time basis using a sling psychrometer
during the biweekly visits.  Average values are given in Table 6.  For every
house, the average relative humidity was lower in the basements and living
areas when the mitigation system was in operation.  Relative humidities ranged
from about 25% to 45% and were typically higher in the basement than the
living area in all houses but site 2 where the reverse trend was true.

      Because relative humidity is a function of both air temperature and
moisture content, a more useful measure of the impact of the mitigation system
on house moisture levels is the humidity ratio (in terms of pound mass of
moisture per pound mass of dry air).  The humidity ratio is not temperature-
dependent and provides an absolute measure of the moisture content of the air.
Table 6 also shows the average humidity ratios in each of the houses with the
system on and off.  For all five sites, the moisture levels in the basement
decreased an average of 18% when the mitigation system was operating.  In the
living space, the average humidity level went down in all the houses but site
4 during system operation.  For sites 1 through 4 the humidity ratio was
higher in the living space than the basement when the system was both on and
off.  This was also true for site 5 when the system was operating.  However,
when the system was off, the basement humidity ratio was slightly higher than
the living space.  To test the significance of the variation in humidity
levels, a multivariable regression was performed for the humidity ratio on the
measurement date, operation mode, floor, floor/mode interaction, and site.
The results show that the variation by date, operation mode, floor, and
between some houses are all significant.

      The data of Table 6 suggest that for sites 2 through 4, human activities
are probably an equal or greater source of house moisture in the living space
as the soil.  The results of sites 1 and 5 indicate that the soil is a
significant source of house moisture and that for both houses, when the system
is operating, basement moisture levels reduce dramatically.
                    CONCLUDING REMARKS AND RECOMMENDATIONS

      The measurement of the impact of the radon mitigation systems on energy
use found an average whole house gas use penalty from the combined results of

-------
the two sets of houses of 76 ccf.  This increase was 6% greater than the pre-
mitigation average use level (1210 ccf) and at a cost of 50C per ccf,
represents an average energy penalty of $38 per year.  The whole house gas use
includes space and water heating use and for some houses, cooking use and
clothes drying (but any changes are most likely due to the added space heating
load).  The test statistic for the paired t test of all ten houses was in the
range of 0.01
-------
                                  REFERENCES

1.    Harrje,  D.T.,  Hubbard,  L.M., Gadsby,  K.J.,  Bolker,  B.  and Bohac,  D.L.
      The Effect of Radon Mitigation Systems on Ventilation  in Buildings.
      ASHRAE Transactions.  95: 107, 1989.

2.    Clarkin, M.,  Brennan,  T., and Osborne, M.C.  Energy penalties associated
      with the use of a sub-slab depressurization system.  In;  The 1990
      International Symposium on Radon and  Radon Reduction Technologies.  U.S.
      Environmental Protection Agency AEERL and ORD,  Atlanta,  Georgia,  1990.

3.    Turk,  B.H.,  Harrison,  J. and Sextro,  R.G.  Performance of radon control
      systems.  Energy and Buildings.  17:  157, 1991.

4.    Henschel, D.B.  Cost analysis of soil depressurization techniques for
      indoor air reduction.   Indoor Air. 1: 337,  1991.

5.    Bohac, D.L.,  Shen, L.S., Dunsworth,  T.S., and Damm, C.J.  Radon
      Mitigation Cost Penalty Research Project.  Underground Space Center,
      University of Minnesota, Minneapolis, MN, 1991.  68 pp.

6.    Fels,  M.F.  PRISM:  An Introduction.   Energy and Buildings.  9: 5, 1986.

7.    Dunsworth, T.S. and Hewett,  M.  Data  quality considerations in using the
      'PRISM'  program.  In;   Proceedings of the Second National Conference on
      Energy Conservation Program Evaluation.  U.S. Department of Energy,
      Chicago, Illinois, 1985.

-------
11
10
:S 9
T 8
c **
0
1
m
o 6
o
c 5
o
•a
0)
a? 3
fli
< 2
1







85% -Percent
RSI Reduction
1
- i
$$
N$:
$S
>oo< ^Vs >^V< *
1.b f a 2. b






^
S^
^
f
















$ft
a











90°y
1
1
I
1
88
3.b


'o


1
I
$£
f





i
I
sVs
a



(




i



345
^
i
I
8$
*.t
<


^



'^ r^
) f a
35°/
W//////M

1
s
1
I

5. b
r

I
1
i

i
SV-,
f







'^ V ^
a
                          b: basement, f: firstfloor a: attic      * _ jnvaljd measurements
                          ^ system on  R^l system off
Figure 1.  Average Radon Levels for System On and System Off Modes of the Monitored
         Houses

-------
TABLE 1.  ENERGY USE RESULTS FOR THE PREVIOUSLY MITIGATED HOUSES
House
ID
11
12
13
14
15
16
PRB
Time Period
9/17/87-9/20/88
10/26/88-10/26/89
2/22/88-2/21/89
11/2/88-9/29/89
12/8/88-11/2/89
3/21/88-4/20/89
*of
Period*
9
12
8
11
11
7
MAC
(ccffr)
686
(47)

1255
(24)

1097
(22)

2412
(56)

889
(26)

896
(24)

R«R.
CVf%)
0.950
6.9

0.994
1.9

0.996
2.0

0.991
2.3

0.982
2.9

0.994
2.7

POST
Time Period
4/18/89-5/17/90
12/27/89-12/27/90
8/22/89- 8A22/90
11/28/89-12/3/90
1/9/90-1/10/91
3/21/90-3/21/91

*of
Periods
7
12
11
12
11
11
NAC
<)
24.6
(9.0)
[0.021
2.9
(3.1)
[0.4]
7.8
(3-4)
[0.05]
6.0
(3.7)
[0.11
10.7
(4.1)
[0.02]
-2.5
(2-8)
[0.4]
8.3
9.2
0.08

-------
    TABLE 2. RADON MEASUREMENTS, PRISM RESULTS, AND BUILDING
      FOUNDATION  CHARACTERISTICS OF THE MONITORED HOUSES
House
ID
1
2
3
4
5
ATD
Result
(pd/L)
5.7
12.5
5.0
5.7
10.1
E-Perm
Result
(pCi/L)
4.4
5.6
13
7.1
NA
#of
Periods
12
12
12
12
12
NAC
(cctfyr)
1755
(17)
1459
(25)
728
(16)
1275
(27)
1036
(28)
CVof
NAC
(*)
1.0
1.7
22
2.1
2.7
R*R
0.998
0.995
0.994
0.993
0.986
Below Grade
Tjrpe
full
basement
base/CS
full
basement
base/CS
full
basement
Constr
block
block
block
concrete
block
Space
Heat
System
hydronic
forced
air
forced
air
hydronic
forced
air
base/CS: partial basement plus crawlspace

-------
 TABLE 3. SUMMARY RESULTS OF THE RESIDUALS APPROACH COMPARING THE
              MONITORED DATA WITH THE PRISM MODEL
House
ID
1
3
4
5
Pooled

average=
sd=
t-stat=
p-value=
average=
sd=
t-stat=
p-value=
average=
sd=
t-stat=
p-value=
average=
sd=
t-stat=
p-value=
chi-sq=
p-value=
On-
Model
-0.042
0.596
-0.141
0.895
0.081
0.213
0.659
0.557
0.397
0.808
0.983
0.381
0.434
0.224
3.875
0.017
16.4877
0.035
Off-
Model
-0.331
0.664
-0.997
0.375
0.123
0.236
1.042
0.356
0.491
0.595
1.650
0.173
-0.089
0.244
-0.730
0.506
9.3424
0.314
On-
Off
0.289
0.446
0.648
0.541
-0.042
0.173
-0.242
0.818
-0.094
0.502
-0.187
0.857
0.523
0.167
3.158
0.020
14.0359
0.081
TABLE 4. ENERGY USE RESULTS FROM THE ALTERNATING MODE EXPERIMENT
                  COMPARED TO THE PRISM MODEL
House
ID
1
3
4
5
average=
sd=
t-stat=
p-value=
NAC (ccf/year)
Pre
Model
1755
728
1275
1036
1199
433
System
Off
1634
773
1454
1004
1216
397
System
On
1740
758
1420
1195
1278
413
Differences (ccf/year)
On-
Model
-15
30
145
159
79
86
Off-
Model
-121
45
179
-33
18
127


On-
Off
105
-15
-34
191
62
106
1.164
0.329

-------
TABLES. MITIGATION SYSTEM EXHAUST FAN ELECTRICAL USAGE, FLOW RATES,
                       AND STACK PRESSURES
House
ID
1
2
3
4
5
average=
Fan Power (Watts)
Rated
90
90
90
100
90
92
Measured
59
72
72
94
70
73
Mew/Rated
(%)
66
80
80
94
78
79
Mitigation Fan Flow (cftn)
Rated:
free air
270
270
270
360
270
288
Measured
10
52
15
46
40
33
Meas/Rated
<*)
4
19
6
13
15
11
Mit. Pipe
Pressure
("water)
1.6
1.2
1.5
1.4
1.1
1.4
TABLE 6. HUMIDITY RATIO AND RELATIVE HUMIDITY MEASUREMENTS IN THE
                        MONITORED HOUSES
House
ID
1
2
3
4
5
Averages
Location

basement
1st floor
basement
1st floor
basement
1st floor
basement
1st floor
basement
1st floor
basement
1st floor
Average Humidity Ratio
(lb water/lb air)
System
On
0.294
0.463
0.505
0.580
0.613
0.752
0.538
0.747
0.448
0.482
0.480
0.605
System
Off
0.553
0.642
0.574
0.653
0.687
0.797
0.572
0.626
0.551
0.542
0.587
0.652
Difference
Off-On
0.259
0.179
0.069
0.073
0.074
0.045
0.034
-0.121
0.103
0.060
0.108
0.047
Average Relative Humidity
(%>
System
On
38.5
29.4
24.8
26.5
43.8
41.6
37.9
31.5
30.9
30.5
35.2
31.2
System
Off
43.6
38.7
35.1
38.7
45.3
45.2
44.9
34.0
36.2
33.3
41.0
38.0
Difference
Off-Oa
5.1
9.2
10.2
12.2
L5
3.6
7.0
2.5
5.3
2.8
5.8
6.8

-------
               DESIGN OF RADON REDUCTION TECHNIQUES
                     FOR CRAWL-SPACE HOUSES:
               ASSESSMENT OF THE EXISTING DATA BASE

          by:  D. Bruce Henschel
               Air and Energy Engineering Research Laboratory
               U. S. Environmental Protection Agency
               Research Triangle Park, NC  27711
                             ABSTRACT

     An evaluation has been completed of the alternative technolo-
gies  for  reducing  radon  concentrations in  the living  areas of
houses having crawl-space  foundations.  Sub-membrane depressuriza-
tion  (SMD)  is consistently  the most effective  technique,  often
providing radon reductions  of  80-98% in  the  living area.  Forced
crawl-space depressurization is second, giving 70-96%.  Crawl-space
depressurization  is  less  well demonstrated  than  is   SMD,  and
performance will vary with crawl-space tightness and weather, but
it will be a primary option when high radon reductions are needed
in inaccessible crawl spaces.  Natural crawl-space ventilation and
forced crawl-space pressurization each provides roughly 50% reduc-
tion or lass in the living area. Crawl-space sealing and barriers
(as stand-alone methods) usually give little or no reduction.

     The  paper also  provides  a  detailed review  — based  upon
available data — of the effects of  individual house design, house
operation, system design,  system operation, and geological varia-
bles, on the performance of SMD and of crawl-space depressurization
systems.   From this review,   the major data  gaps are:   1)  the
specific conditions under  which crawl-space depressurization might
be preferred over  SMD,  including the relative operating costs of
the two approaches; 2)  the optimal method for distributing suction
beneath a  SMD  membrane (i.e.,  individual suction  pipes  vs.  sub-
membrane  perforated piping) under  different conditions;  3)  the
optimal degree  of  SMD membrane  sealing  required under  different
conditions; 4)  the conditions under which it may  be possible to
leave  a  portion of  the crawl-space  floor  uncovered by  the SMD
membrane,  if the crawl space is partially inaccessible; and 5) the
durability of SMD, crawl-space depressurization,  and natural crawl-
space ventilation systems.

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

-------
                           INTRODUCTION

     Based upon data presented in the  literature and upon contacts
with commercial radon mitigators, experience with reducing indoor
radon concentrations  in crawl-space houses is much  more limited
than  is experience  with  mitigating  basement and  slab-on-grade
houses.   This relatively limited experience with crawl-space houses
probably  results from  the facts  that:    1)  crawl-space  houses
represent a  relatively limited  percentage of the total housing
stock nationally (14% of the total new housing starts in the U. S.
between  1976 and 1983,  according to  the  National  Association of
Home Builders); and 2)  houses  with vented  crawl spaces may be less
prone to having elevated radon levels in the living area than are
basements and slabs on grade,  because of less direct contact with
the soil.  Another contributing explanation seems to be that many
crawl-space houses with elevated radon have an adjoining basement
(or slab-on-grade)  wing which is commonly the major radon entry
route into the  house.   The mitigation systems in  such combined-
substructure houses often focus on sub-slab depressurization (SSD)
or sump/drain-tile depressurization (sump/DTD) in  the basement; the
crawl space  might need to be treated only by  isolation from the
livable area, and/or by natural or forced ventilation.  Sometimes
the crawl space can be  ignored altogether.

     Extensive  R&D  and  commercial  experience  exists  in  the
treatment of basement  and slab-on-grade  houses,  especially using
active soil depressurization (ASD)  techniques such as SSD and DTD.
The U. S. Environmental Protection Agency  (EPA) is able to provide
reasonably definitive guidance regarding the design and installa-
tion of SSD and DTD techniques for houses with slabs.  But due to
the more limited experience with crawl-space  houses,  EPA is in a
weaker position to provide guidance for this substructure type.

     Although  crawl-space houses  are only a  relatively limited
fraction of the total U. S. housing stock,  they represent a signif-
icant fraction of the housing stock in some parts of the country,
especially the Northwest  and  the Southeast.   And although crawl-
space houses may be assumed to be less radon-prone than the other
substructure types, many crawl-space houses have been encountered
having radon levels  of 15-20 pCi/L (550-750 Bq/m3)  and  higher in the
living area.  Such levels exist in houses  where the crawl space is
a  significant  (or the  sole)  source  of  the radon, and where the
problem  thus cannot be addressed  by  SSD or DTD in  an adjoining
basement.

     The data review and analysis presented in this paper identify
the areas where further information is needed in order for EPA to
provide adequate guidance in the design and installation of radon
mitigation systems for  crawl-space houses.

     The discussion here assumes that the crawl space has a floor
consisting of bare earth, or of gravel or a plastic vapor barrier
on top of bare earth.   Crawl spaces having a concrete slab (or an

-------
unfinished concrete "wash" floor) would commonly be treated using
SSD, similar to a basement or slab on grade, and are not considered
directly in this paper.


      ALTERNATIVE MITIGATION METHODS FOR CRAWL-SPACE HOUSES

SUB-MEMBRANE DEPRESSURIZATION  (SMD)

     Active SMD — the ASD approach applicable to crawl spaces —
involves installation of a membrane (usually polyethylene sheeting)
over  the crawl-space floor,  and drawing  suction beneath  this
membrane using a  fan.  This  is analogous to installing a plastic
"slab" over the floor, and drawing suction  beneath this slab using
either individual suction pipes (analogous  to SSD) or a segment of
perforated piping  beneath the membrane  (analogous to  DTD).   SMD
probably works by the same mechanisms which apply for SSD and DTD.
One major mechanism is depressurization of  the sub-membrane region
relative to the crawl space  in order to  prevent soil  gas flow up
into the  crawl space  (and from there  up  into  the  living area) .
Because  the  membrane will leak if not  sealed  completely,  other
mechanisms can also come into play (ventilation or depressurization
of the crawl space by the SMD fan).

     SMD  has  consistently been found  to  be the most effective
approach  for  reducing radon  in crawl-space  buildings,  commonly
providing radon reductions of 80 to 98% in the living area when the
crawl space is the sole source of the indoor radon (References 1,
2, and 3).  Commercial radon mitigators working in regions having
a significant number  of crawl-space houses have reported similar
success  with  SMD  (References  4,  5, and 6).   SMD  would be  the
primary  mitigation method considered  in  any house where  radon
reductions greater than  about  50% are  needed (thus  ruling  out
natural ventilation of the  crawl space as  an  option),  where  the
crawl space is a major (or the sole) radon source,  and where the
crawl space is accessible. It may still be  the technique of choice
even where reductions below 50% are needed.  Where the crawl space
is adjoined by a basement  wing, the SMD  system will usually need to
be  supplemented  (or  replaced)  by  a  SSD  or DTD  system in  the
basement  (References  2,  4,  5,   6, 7, 8,  9, and 10).   A national
survey of  mitigators  indicated that  SMD  is the technique  most
commonly utilized  in  crawl-space  buildings by about  one-third of
the mitigators surveyed (Reference 11) .

     The installation  cost for SMD is  typically in the  range of
$1,000-$2,500  (Reference  12).   This is  the highest  installation
cost of any of the crawl-space treatment options, with the possible
exception of crawl-space sealing.  However,  relative to crawl-space
depressurization (where the entire crawl  space is depressurized),
the membrane  reduces  the amount  of air drawn  out  of the  crawl
space,  and hence  the  amount  of conditioned air drawn  out of  the
living area overhead.   The annual operating costs of SMD (for  fan
electricity and  for  the  heating/cooling penalty associated  with

-------
treated house air exhausted by the system) will range between $45
and $200, depending upon the particular conditions  (Reference 13) .
At "baseline" conditions — 50 cfm (25 L/sec)  exhaust, 25% treated
air  in exhaust,  6,550 F°-days  (3,640  C°-days) ,  90 W  fan power
consumption — the annual cost would  be $89.  As discussed later,
this is about half the baseline annual operating cost that might be
expected with crawl-space depressurization, the major competitor of
SMD for mitigating high-radon crawl-space houses.

CRAWL-SPACE DEPRESSURIZATION

     With active  crawl-space  depressurization,  an exhaust fan is
mounted to blow crawl-space air outdoors.  Installation  of the fan
may have to be accompanied by some sealing of the crawl  space, to
reduce the crawl-space leakage area and to thus aid in depressuriz-
ing the  crawl  space.    (As a  minimum,  foundation vents should be
closed.)  The objective is to depressurize the entire crawl space
relative to the  living area,  rather  than depressurizing just the
region beneath a membrane over the floor.  Crawl-space depressur-
ization  works  primarily by reversing the natural  flow patterns
between the crawl space and the  living area (i.e., so that radon-
free living-area air  flows down  into  the crawl space, rather than
radon-containing crawl-space air flowing up into the  living area).

     Crawl-space  depressurization  might  also  be  expected  to
function in part  by  increasing  the ventilation rate of the crawl
space, diluting the crawl-space  radon with  air from outdoors and
from the living area.  However,  the dilution mechanism may not be
important  in practice,  if  in  fact the  crawl  space  is  being
effectively depressurized.   Depressurization of  the crawl space
increases the flow of soil  gas up into the crawl space, largely or
completely  offsetting any  dilution   effects  resulting  from  the
increased flow of outdoor and living-area air into the crawl space.
As a result, radon concentrations in the crawl  space have generally
been found to remain about the same,  or to increase, when the crawl
space is depressurized (References 1,  2, and 3), demonstrating that
the dilution mechanism is providing no net benefit.

     Crawl-space depressurization has consistently proven second
only to SMD in effectiveness in reducing living-area  radon levels.
Living-area reductions ranging from  70  to 96% have been reported
from tests in nine crawl-space buildings where crawl-space exhaust
flows ranged between  50 and 130  cfm  (25 and 60 L/sec) (References
1, 2, 3, and 14).  Performance will depend on the tightness of the
crawl space and  the size of the exhaust fan;  i.e.,  on  the  degree of
crawl-space depressurization that can be achieved relative to the
living  area.    Percentage  reductions  may also  depend  upon  the
weather and upon  homeowner activities, since the  low  degrees of
crawl-space depressurization often achieved relative  to the living
area would seem subject to being overwhelmed.

     Where  pressures  have  been reported,  in buildings without
forced-air ducting in the  crawl  space,  crawl-space depressuriza-

-------
tion has depressurized the crawl space relative  to the living area.
by 0 to 0.004 in. WG  (0 to I Pa) in two structures  (References 3,
13, and 14) .  Higher crawl-space depressurizations were observed in
one  house having  an  unusually  tight  floor and foundation wall
(Reference 14).   However,  in  more typical houses with relatively
leaky floors and foundation walls, it would be expected that crawl-
space depressurizations relative to the living area will commonly
be a  couple  thousandths of an  inch  WG (a fraction of  a Pa) ,  at
best, when exhaust flows are in the range of 50 to 130 cfm  (25 to
60 L/sec).

     Because  crawl-space  depressurization  may  increase  radon
concentrations in the crawl space,  it is important that the system
not  be   overwhelmed   often   by  weather-  or  appliance-induced
depressurizations within the living area.  A near-zero crawl-space
depressurization relative to the living area increases the threat
that the system will sometimes be overwhelmed.

     Maintenance of depressurizations within the crawl space would
be expected to be  more difficult when the crawl space  has a high
effective leakage  area.   Forced-air  furnace ducting in the crawl
space would be expected to increase  the  leakage area  between the
crawl space and the living area; in addition, low-pressure return
ducting in the  crawl  space could draw the  potentially increased
crawl-space radon levels into the circulating house air.  Limited
experience in two houses having return ducting in the crawl space
has shown living-area  radon reductions of  70 to 74%  (Reference 1) ,
at the low end of  the range  for crawl-space depressurization;  in
neither house was the living-area radon concentration reduced below
EPA's guideline of 4 pCi/L (150  Bq/m3)  by this  technique.  But in
this same study,  depressurization of  a  crawl space containing only
high-pressure forced-air  supply  ducts  achieved  living-area radon
reductions greater than 90%.  There  is also evidence  that crawl—
space depressurization might contribute to back-drafting of combus-
tion appliances  in the crawl space, where crawl-space depressuriza-
tions above 0.01 to 0.02 in.  WG  (2.5 to 5 Pa) are achieved.

     The national  survey  of mitigators  (Reference  11)  indicated
that  natural  crawl-space  ventilation   or  forced  ventilation
(including crawl-space pressurization and depressurization systems)
is most commonly used by 28% of the mitigators when treating crawl-
space  houses.    Thus,   crawl-space  depressurization systems  are
preferred by fewer than 28%.  It  is suspected that many mitigators
using  an  exhaust fan  in  crawl spaces are  doing so to increase
ventilation/dilution,  rather than to  depressurize the crawl space.

     The installation  cost of a crawl-space depressurization system
will generally be  much lower  than that of  a SMD system,  perhaps
less than $500 if all that is required is the installation of a fan
and perhaps  some  limited  piping.    The  installation cost  would
increase if additional effort  is  required  to seal the crawl space,
to reduce the leakage area.

-------
     Since some fraction of  the air exhausted by the crawl-space
depressurization fan will be conditioned  air drawn down from the
living area — and since the amount of air exhausted will tend to
be greater  than that  from  a  SMD  system —  the heating/cooling
penalty (and hence the operating cost) associated with crawl-space
depressurization will  be greater than that for  SMD.   The annual
operating cost of  a  crawl-space depressurization system (for fan
electricity plus the  heating/cooling penalty) would be in the range
of  $60-$280,  depending upon  conditions  (Reference   13).    At
"baseline" conditions   — 100  cfm  (50 L/sec)  exhaust,  50% of the
exhaust is  treated air, 6,550  F°-days  (3,640  C°-days), 90  W fan
power consumption  — the operating cost  is $167.   This is about
twice the baseline annual cost of the SMD system, cited previously,
as a result of the increased heating/cooling penalty.

     In summary,  crawl-space depressurization will  be most applic-
able where:   1)  the crawl space is  inaccessible, ruling  out SMD as
an option; 2)  the crawl  space is relatively tight  and well isolated
from the living area to begin with, reducing the need to seal and
isolate the crawl  space; 3)  radon  reductions of  greater than 50%
are required in the living area,  ruling  out natural ventilation as
an option;  and  4)  there is no  combustion appliance in the crawl
space that might back-draft.   But even where conditions  2) through
4) above are favorable for crawl-space depressurization,  SMD should
still be  considered, due to  its potential for  greater and more
consistent radon reductions at reduced operating cost.

CRAWL-SPACE PRESSURIZATION

     In crawl-space  pressurization systems,  a fan is  mounted to
blow outdoor  air  (or  air  from  the  living area)  into  the crawl
space.  Some sealing  or isolation of the crawl  space may be needed
in order to achieve effective pressurization.  Crawl-space pressur-
ization works  by causing the  pressure  of the crawl  space  to be
higher than the pressure in the soil, preventing radon-containing
soil gas from entering the crawl space (and thus from entering the
living area).  Unlike crawl-space depressurization, pressurization
can also work  in part  by a  ventilation/dilution  mechanism, since
radon-free air is being blown into the crawl space.  If the crawl
space cannot be effectively pressurized, the ventilation mechanism
may become the more important mechanism.

     In  limited  testing of crawl-space  pressurization  in  five
structures (References  I,  3, and 14), where outdoor air was being
blown into the crawl  space,  crawl-space pressurization was found to
give living-area radon reductions  usually in  the  range of 35 to
80%.   These reductions are generally comparable to those obtained
using natural  crawl-space  ventilation  (i.e.,  simply  opening the
foundation vents, with no fan).

     This result would suggest that the pressurization  system was
probably not effectively pressurizing the  crawl space.   Indeed, in
one of  the  few cases  where pressure measurements were reported

-------
 (Reference  3),  the  crawl space was still under 0.006 in. WG  (1.5
 Pa) negative pressure relative to outdoors with the pressurization
 system  operating; the resulting radon reductions in the overhead
 occupied space  were only about  40%.  The  best  results with crawl-
 space  pressurization systems  were observed with  two atypically
 tight crawl spaces (Reference 14), where the crawl spaces could in
 fact be pressurized to 0.02 in. WG  (5 Pa) and  greater relative to
 outdoors; in  these  two houses,  indoor radon reductions were over
 80%.

     The installation cost of a crawl-space pressurization system
 would be about  the same as that for crawl-space depressurization:
 perhaps on the order of $500 (including insulation of water pipes),
 unless additional effort is required to seal or isolate the crawl
 space,  or  to install  insulation  under the overhead  floor.   The
 operating  cost  for  pressurization  is  difficult  to  estimate.
 Assuming that the pressurizing air is blown into the crawl space
 from outdoors, the annual operating cost should be no greater than
 that for crawl-space  depressurization,  and  greater than that for
 SMD.

     In  summary,  crawl-space  pressurization  would appear  to be
 applicable  where  the  crawl space  is  particularly  tight,  so that
 effective pressurization  might in fact be  achieved.   Otherwise,
 pressurization  will  be applicable  in the  same  situations where
 natural crawl-space ventilation (with no  fan)  is most applicable:
 1) the required living-area radon reductions are on the order of
 50%, and the crawl space may be a major source; 2)  alternatively,
 the  crawl   space  is  not  a  major  source,   and  the  crawl-space
 pressurization/ventilation system is simply  supplementing a SSD or
 DTD  system  in  an  adjoining basement,  so that  high  crawl-space
 reductions are  not needed;  3) the crawl  space is reasonably well
 isolated to begin with;  4) there  are  no water pipes in the crawl
 space which the system might  cause to freeze during cold weather
 (or, alternatively, any  existing  water pipes can conveniently be
 insulated,   or  the  climate is  mild);  and  5)  ideally, there is
 already insulation beneath the overhead flooring (or the climate is
mild),  reducing  occupant discomfort  and  the  heating  penalty
 resulting from  blowing cold air into the  crawl space.

     Crawl-space  pressurization might  be selected over  natural
ventilation only in crawl  spaces  where there is  not a sufficient
 number of properly distributed foundation vents to permit natural
ventilation to  be easily  implemented.   This  statement  is based
upon:   1)   the available  performance  data,  which suggest  that
natural ventilation  is about  as  effective as pressurization in
reducing living-area radon, unless the crawl space is particularly
tight and thus amenable to true pressurization; and 2) the simpli-
city and reduced costs associated  with opening  existing foundation
vents, compared to the installation and operation of a fan.  Even
where sufficient vents do not exist, the installation of additional
vents and the implementation of natural ventilation  can still be
considered as an alternative to pressurization.

-------
NATURAL CRAWL-SPACE VENTILATION

     In  general,  natural  crawl-space  ventilation  consists  of
opening existing vents in the crawl-space foundation wall  (and/or
installing  new vents), to  increase the natural  infiltration of
outdoor air.   Natural ventilation  works through  two mechanisms.
First, it tends to neutralize the pressure between the crawl space
and outdoors.  By thus reducing crawl-space depressurization during
cold weather,  this technique  reduces the driving  force  drawing
radon-containing soil gas into the crawl space (and thence into the
living area).  Second, by providing  an increased infiltration rate
for fresh outdoor air into the crawl space, the technique  dilutes
any radon that does enter the crawl space.

     The  performance  of  natural  crawl-space  ventilation  will
undoubtedly  depend  upon:   the number  and  location  of  vents; the
tightness of the crawl space; weather  conditions, especially wind
and  temperature;  the presence   of  vent  obstructions   (such  as
shrubbery); and activities by the homeowner,  such as the operation
of depressurizing appliances in the  house or  crawl space.   Because
of the role of weather and homeowner activities, performance will
vary with time.

     In  limited  testing  as  part of  EPA's R&D program,   natural
ventilation was found to provide  radon reductions ranging  from 46
to 83% in the living area of  five buildings in which five to eight
existing foundation vents were opened (References  1 and 3).  These
reductions are based  upon relatively short-term testing,  usually
for 2- to 4-day periods both before and after  the crawl-space vents
were opened; this short measurement  period probably contributed to
the scatter in the results, since performance is expected  to vary
with time.   In a  somewhat longer-term study,  testing over 5 to 7
weeks in  two houses having 6 to  10 vents, provided indoor radon
reductions of 40  to 45%  (Reference  15).   Several-week testing in
three  Spokane houses  provided  indoor reductions of  50  to 60%
resulting from natural ventilation in combination with crawl-space
sealing (Reference 16).  Other short-term testing of natural vent-
ilation, sometimes  in combination with sealing,  has demonstrated
living-area reductions of:   21 to 35% in one house  with 14 vents
(Reference 15); and 18 to 27%  in two houses,  and  75% in a third,
where only two vents  were opened (Reference  2) .   In a mitigation
effort where multiple  vents were retrofit into a number of existing
crawl-space houses  (to provide  1 ft2  of  vent area per 150 ft2 of
floor area,  or 0.1 m2 per  15 m2) , living-area reductions consistent-
ly ranged between 0 and 50%  (Reference 17).

     These limited current data suggest that average indoor reduc-
tions will probably be no greater than about 50% in most cases.

     Some measurements suggest that natural crawl-space ventilation
may increase the  ventilation  rate  of the crawl space by a factor of
3 or 4  (Reference 18). This  result  would suggest that crawl-space
radon concentrations should be reduced by 67 to 75% due to dilution

-------
alone, and by even more due to the reduction in driving force; and,
in fact, crawl-space radon levels are reduced  by about that amount
(Reference 1).  The fact that natural ventilation reduces living-
area  concentrations  by less than that amount suggests  that the
reduction  in  crawl-space  concentrations  is accompanied by  an
increase in the flow of crawl-space air up into the house.  Perhaps
some radon is also flowing into the living area via a route other
than the crawl space  (e.g., through a block fireplace structure).

     If natural crawl-space ventilation can be implemented simply
by opening existing foundation vents, the installation cost will be
near zero.  Where the crawl space has no vents (or too few vents)
and where vents thus need to be retrofitted,  the estimated install-
ation cost is about $100 per vent, assuming that multiple vents are
being  installed  (Reference  17) .   The  installation cost  would
increase when:  1) effort  is required to isolate  the crawl space
from the  living area; 2)  water pipes in the  crawl  space must be
insulated; or 3)  insulation must  be  installed  beneath the overhead,
floor, for occupant comfort or  to reduce  the heating penalty.  The
operating  cost associated  with  natural ventilation  will  be  a
heating and cooling penalty in the  house because  the crawl space
will now be colder in  the winter  (and hotter in the summer).  This
heating/cooling penalty cannot be rigorously quantified.

     In summary,  natural crawl-space ventilation will  likely be
useful primarily where:  1) only a limited indoor radon reduction
is needed; or 2) the crawl space is not a major radon source, and
crawl-space ventilation is simply supplementing SSD  or  DTD in an
adjoining basement.   See the  earlier discussion  regarding crawl-
space pressurization.  Based upon EPA's mitigator survey (Reference
11),  natural  and forced  ventilation  of  the  crawl  space  are
preferred by 28% of the mitigators treating crawl-space structures.
While the survey did not inquire specifically,  it is suspected that
most mitigators favoring natural  ventilation  are  utilizing it in
houses meeting criterion  1) or 2) above.  It is also suspected that
many  mitigators  using forced  ventilation are using the  fan to
increase the  ventilation mechanism,  rather  than to  pressurize or
depressurize the crawl space.

SEALING AND BARRIERS

     There are  two primary approaches  for applying sealing and
barriers.  One approach  involves  sealing a  barrier  such as poly-
ethylene sheeting over the  crawl-space floor (and perhaps over the
interior  face of  the  foundation walls) , to  reduce the  flow of
radon-containing soil  gas into the crawl space. The second approach
involves sealing openings between the crawl space and  the living
area, isolating the crawl  space and  reducing the flow  of radon-
containing crawl-space  air into  the living  area.   This second
approach can include:   closure of relatively small openings  (such
as plumbing penetrations); closure of major openings,  such as exist
when the crawl space is completely open to an adjoining basement;
and sealing the seams in any cold-air return ducting in the crawl

-------
space, to reduce the amount of crawl-space air drawn into this low-
pressure  ducting and  distributed  throughout the  house by  the
forced-air heating/air-conditioning system.

     The  limited data  available suggest  that  sealing  openings
between the crawl space and the living area of existing houses, by
itself, may provide little or no reduction in the radon concentra-
tions in the living area.  In tests on four houses  (References 2,
15,  and  16),  attempts to  isolate  the  crawl space  resulted in
changes in indoor concentrations ranging from a 10% decrease to a
15%  increase.   All  four  houses had  forced-air ducts in the crawl
space, complicating the  isolation  of  the crawl  space.   These
observed changes in radon levels are within the normal variability
of indoor concentrations in a given house.

     The effect of a completely sealed barrier of plastic sheeting
over the crawl-space floor and foundation walls has  been tested in
only one house, resulting in indoor reductions of 31%, based upon
2 to 3 weeks of testing both before and after installation of the
membrane (Reference 16).  Tests by other investigators, without the
membrane's being  completely sealed,  suggest  that  essentially no
indoor reductions are achieved when the barrier is not  fully sealed
(Reference 1).

     The openings between the crawl space and the  living area would
appear  to be   so  numerous,  widely  distributed,   and  sometimes
inaccessible — and the  gas tightness and durability  of sealed
barriers are so suspect  —  that sealing/barriers should never be
relied upon  as a stand-alone mitigation method where  the crawl
space is an important radon source.

     According  to the survey of mitigators (Reference 11), 25% of
the mitigators  preferred some type of sealing or barrier approach
in  treating  crawl-space houses.   Discussions  with mitigators
suggest that, in fact,  sealing and barriers in the crawl space are
most  commonly   utilized  as  a  supplement  to  SSD   or  DTD in an
adjoining basement  when the  crawl  space is not a  major source.
Sealing can be most  important when:  1) the crawl  space opens to an
adjoining basement;  and 2)  depressurization,  pressurization, or
natural ventilation of the crawl-space is to be utilized, and some
tightening of the crawl  space is necessary  in order to allow the
crawl  space  to  be  pressurized/depressurized, or  to reduce  the
heating/cooling penalty in the living area.

     Depending  upon the  nature of the sealing effort undertaken,
sealing/barriers can be potentially time-consuming and expensive.

     In summary, a homeowner may wish to try a sealing or barrier
technique as  a  stand-alone method on a  do-it-yourself basis, to see
how  well  this  approach  might  perform  in that  particular  case.
However,  the  likelihood of achieving any meaningful radon reduction
with these techniques  on a stand-alone basis is so small that a
commercial mitigator would  likely never propose sealing/barriers

-------
except  in  combination with SSD in an  adjoining basement,  with a
crawl-space ventilation/pressurization/depressurization approach,
or with SMD in the crawl space.


  DETAILED REVIEW OF DATA BASE ON CRAWL-SPACE TREATMENT METHODS

     The review  of the  existing  literature on radon mitigation
methods for crawl-space houses focussed on the house and mitigation
system variables which could have  an  impact on system performance.
These variables are listed in Table 1.
   TABLE 1.   VARIABLES ADDRESSED DURING REVIEW OF THE DATA BASE
             FOR CRAWL-SPACE RADON MITIGATION SYSTEMS
House Design Variables
     Nature of crawl-space floor (bare earth vs. gravel, evenness)
     Crawl-space floor dimensions
     Crawl-space accessibility
     Nature of the foundation wall (block vs. poured concrete)
     Presence of adjoining wings (basement or slab on grade)
     Tightness of crawl space

House Operating Variables
     Operation of central furnace fan or exhaust fan

Mitigation System Design Variables
     Method for distributing suction beneath membrane
     Number of suction pipes
     Location of suction pipes
     Diameter, type of suction pipes
     Method of installing suction pipes through membrane
     Nature of membrane
     Extent of membrane
     Degree of sealing of membrane

Mitigation System Operating Variables
     Fan capacity
     Fan in suction vs. pressure

Geology/Climate Variables
     Source term (especially radon concentration in soil gas)
     Permeability of underlying soil
     Climatic conditions

Mitigation System Durability
     Radon reduction performance
     System suctions and flows
     Equipment durability

-------
     The discussion  which follows focusses  primarily on SMD and
crawl-space  depressurization,  since  these  are  the techniques
expected to provide the greatest indoor radon reductions  in crawl-
space houses.

HOUSE DESIGN VARIABLES

Nature of the Crawl-Space Floor

     It might be  expected that crawl spaces having gravel floors
would provide good performance with SMD  systems  having only one
individual suction pipe  (even with relatively large floor areas),
without need for  sub-membrane perforated  piping or for strips of
porous matting  beneath the  membrane to aid  in distributing the
suction.  The gravel should enable good suction field extension,
analogous to SSD systems where a good gravel layer exists beneath
the slab.  When the floor consists of bare earth, a greater effort
to distribute suction would  likely be needed  (more suction pipes,
sub-membrane perforated piping, or sub-membrane porous matting).

     The limited data  in crawl spaces  with gravel floors (Refer-
ences 1 and 9)  confirm that very good indoor  radon reductions  (92-
98%) are achieved with SMD,  but,  because other variables were also
varied, do not verify how large a floor area  can be treated by one
individual SMD suction pipe when gravel is present.

     Crawl spaces with bare earth floors have tended to give lower
reductions, often 80-97%  (References 1, 2, 8, 9, 10,  19, and 20).
But other variables besides  the nature of the floor were also being
varied (e.g., floor  area, adjoining wing, method for  distributing
suction).  Thus, a quantitative comparison of gravel vs. bare earth
(in terms of the number of individual SMD  suction pipes per square
meter of  floor  area  required  to  provide  comparable  indoor radon
reductions) is not possible.

     In most reported studies, where the floor is relatively even
and flat  (Reference  1,  2, 9, and 12),  0.15  to  0.25  mm thick (or
equivalent) polyethylene  sheeting has  generally been found to be
sufficient, with strips of heavier material on top of  the membrane
along expected traffic  routes.  Where  the  floor  is  irregular ~
e.g., with rock outcroppings (References 10, 18,  and 19) — heavier
membrane materials were used, or matting  was placed beneath the
membrane.

Crawl-Space Floor Dimensions

     With SMD systems  having  individual  suction  pipes  (i.e., no
sub-membrane perforated piping) where no gravel is present on the
floor, crawl spaces  as large  as  1,500 ft2 (140 m2)  were reduced
below 4 pCi/L (150 Bq/m3)  in  the living area with  a single centrally
located SMD suction pipe, based on results from 12 houses (Refer-
ences 2, 8, 9,  and 20).   When floor areas  are larger than 1,500 to
2,000 ft2 (140  to 190 m2) ,  multiple individual suction pipes have

-------
consistently been  needed  with bare-earth floors to reduce levels
below 4 pCi/L (Reference 2), or else sub-membrane perforated piping
has been needed  (References 1 and 5).

     The number  of individual suction pipes required for a given
floor area  (or the need for sub-membrane perforated piping) might
be  expected to  increase  in  houses where:   1)  it  is  desired to
achieve lower indoor levels (e.g., 2 pCi/L,  or 75  Bq/m3) ; 2) there
is a high radon  source term; or 3) the crawl space has dimensions
(such as an L shape) or an interior foundation wall which divides
the floor area into two or more segments.  The number of pipes for
a given floor  area (or the need  for  perforated piping)  would be
expected to decrease when there is gravel on the floor.

Crawl-Space Accessibility

     Where the crawl space is largely or completely inaccessible,
SMD will not be a practical option, since it  will be  impractical to
install a membrane directly over the  crawl-space  floor.   Crawl-
space depressurization may be the only alternative in such cases
when  indoor radon  reductions much  greater than  about  50%  are
required.  Natural ventilation will be an option  if only limited
reductions are needed.

     In  some  cases where  the  crawl  space  is  only  partially
inaccessible —  e.g.,  due to limited headroom  at  one  end,  or to
large obstructions — it may still be possible to obtain reasonable
reductions by applying SMD to the accessible portion of the floor
(Reference 1).   Since only anecdotal results have been obtained to
date, it is unknown how much of the floor might be left uncovered
by  membrane under different  conditions while still  providing
meaningful radon reductions.

Nature of the Foundation Wall

     Almost all of EPA's data on 24 SMD systems are for crawl-space
houses having hollow-block foundation walls  (References 1,  2, 8, 9,
10,  19, and  20).   All  but two of  these study houses were reduced
below 4 pCi/L  (150 Bq/m3)  in  the  living  area,  and over half were
reduced below 2 pCi/L (75  Bq/m3) .  Thus, it  is impossible from the
limited data to  separate out  the  effects  of  block  vs.  poured
concrete foundation walls.  However, to the  extent that the blocks
may have been  contributing to radon entry,   the SMD system would
seem to have been largely addressing that entry route.

     On the other hand,  in the two crawl-space study houses having
the highest source term (reflected by living-area levels of 88 to
160 pCi/Li  or 3,250 to 5,900 Bq/m3), a block-wall depressurization
(BWD) component  had to be  added  to the  SMD system in  order to
achieve adequate reductions (References 2  and 19) .  Thus, separate
treatment of block walls  may be  important  under  some conditions
(such as a  high source term) when  the walls  become important entry
routes. Some mitigators report treatment of  block foundation walls

-------
using the SMD system by extending the membrane up the interior face
of the wall,  and  attaching the perimeter edge of the membrane to
the sill plate on  top of the foundation wall (References 6 and 20);
in this  manner, the SMD suction pipe might  be  able  to treat the
walls as well as the floor.

Presence of an Adjoining Wing

     When a  basement wing adjoins the  crawl space,  the basement
will essentially  always have  to be treated  (usually using SSD or
sump/DTD) .  Whether or  not the crawl space  will also have to be
treated  (by SMD or  some other method)  will depend primarily upon
the  importance  of  the  crawl  space   as a radon   entry  route
(References 2,  4,  6, 9, and  16).   Where the crawl  space  is not
particularly  important  —  usually as  determined by  grab radon
measurements in the crawl-space soil or air  — the crawl space is
often treated  solely by isolation from  the  basement, as needed,
possibly supplemented with some natural ventilation (References 4
and 6) .   Good communication beneath the basement  slab — suggesting
that SSD or  DTD in  the basement will be very  effective  — also
seems  to reduce  the  need  for  separate crawl-space  treatment
(Reference 9).

     Where the wing adjoining the crawl space is a slab on grade,
available data are less definitive regarding when one or both wings
need direct treatment.

Tightness of Crawl Space

     When the crawl space  is tightly  isolated from the living area
and from outdoors,  the  house may be a  candidate for crawl-space
depressurization.   A low effective leakage area, as determined by
a blower door,  would be a quantitative  indicator  that the crawl
space is tight.  Among the visual (qualitative) indicators that the
crawl space may be  tight would be:   l)  the absence of forced-air
supply or return ducting penetrating the flooring between the crawl
space and the living area; or 2) the absence of foundation vents.

     Another  house  design factor  that  should be  addressed when
crawl-space depressurization is being considered is whether there
is a combustion appliance in the crawl space which might be back-
drafted  if the crawl space is depressurized.

     Another house  design  factor  somewhat related to crawl-space
tightness is  the  degree  of insulation of the crawl  space (e.g.,
insulation under the living-area flooring).   Such insulation might
make natural  crawl-space  ventilation  (or forced  pressurization
using outdoor air) a more viable option where only limited indoor
radon reduction is needed.

HOUSE OPERATING VARIABLES

     A central forced-air  furnace fan operating  in the crawl space
could depressurize  the  crawl  space,  possibly  helping both  to

-------
overwhelm the sub-membrane depressurizations being created by a SMD
system, and to distribute crawl-space radon throughout the house.

     Very  limited  measurements  (Reference  18)  suggest that  a
central fan might depressurize a crawl space by 0.001 in. WG (0.2
Pa) or  less.   Limited sub-membrane depressurization measurements
with SMD  systems indicate that, depending upon SMD design, sub-
membrane suctions can sometimes drop below 0.001 in. WG within 10
to 15 ft (3 to 4.5 m)  of an  individual suction pipe  (Reference 2).
Thus, in theory,  operation  of  a central  furnace fan in the crawl
space could overwhelm the SMD system at locations remote from the
suction pipes.  In practice, there are no clear data demonstrating
increases  in indoor  radon  levels  when the  central  fan begins
operating.

     Tracer gas measurements in two houses (Reference 18) confirm
that operation of a central forced-air furnace fan located in the
crawl space will  in  fact significantly increase the inter-zonal air
flows between the crawl  space  and the  living  area.   Thus, if the
SMD  system  is  indeed  overwhelmed,  the  central  fan will  help
distribute the radon throughout the house,  as expected.

MITIGATION SYSTEM DESIGN VARIABLES

Method of Distributing Suction Beneath SMD Membrane

     Suction can be distributed beneath the SMD membrane either by
installing  individual  suction  pipes vertically  down  through the
membrane (analogous to SSD), or by drawing suction on some pattern
of perforated piping laid beneath the membrane (analogous to DTD).
Installation of perforated piping may increase  installation cost by
perhaps $85  (Reference  12).  However, since  the measurable sub-
membrane suction field may not extend more than about 10 to 15 ft
(3 to 4.5 m) from individual suction pipes when the floor is bare
earth (Reference 2), sub-membrane piping could be a significant aid
in distributing the suction field.  In addition, if leaks develop
in the membrane near the suction pipe penetration, distribution of
the suction through perforated piping would be expected to reduce
the amount of short-circuiting  resulting from such membrane leaks.

     Data from SMD  systems  are  too limited  — and too many other
house and system variables were being varied simultaneously — to
enable definitive guidance regarding the conditions under which one
approach or  the  other would be more  cost-effective or otherwise
preferred.    Intuitively,  sub-membrane  piping  would  be  less
necessary when:  1)  there is gravel on the floor, giving good sub-
membrane communication; and/or 2) the crawl space is smaller than
about 1,OOQ to 1,500 ft2 (90 to 140 m2),  increasing the likelihood
that one individual  suction  pipe would be sufficient; and/or 3) the
source term  is not  high, and reduction of indoor  levels below 2
pCi/L (75 Bq/m3)  is  not  the  objective.

    ' Among  EPA's SMD  study houses,  all seven  which used  sub-
membrane piping were reduced below 2 pCi/L (75 Bq/m3) in the living

-------
area, even though one was as large as 2,700 ft2,  or 250 m2  (Refer-
ences If 7, and 10).  By comparison,  none  of the  study houses with
individual-pipe  SMD systems were  reduced below  2  pCi/L,  except
where  the crawl-space  was  a relatively  small  wing  adjoining a
basement that was being treated by SSD or DTD (References 8  and 9) .

     Some  commercial  mitigators  use  the individual-pipe approach
almost exclusively, unless  the crawl space is unusually large or
complex (References 5 and 6).  Others use sub-membrane perforated
piping routinely  (Reference 4).

Number and Location of  Suction Pipes

     With individual-pipe SMD systems — when there is no gravel on
the  floor to  aid in  distribution  of the  suction field  — one
suction pipe  is  generally sufficient to reduce indoor concentra-
tions to  4 pCi/L (150 Bq/m3)  and less  with crawl-space floors as
large as 1,500 ft2 (140 m2) .  With crawl spaces of  2,000 ft2 (190 m2)
and larger, two or more pipes are needed (Reference 2) .  The number
of pipes needed for a given crawl-space floor area would increase
if the source term were high, or if indoor levels  below 2  pCi/L  (75
Bq/m3) were desired.  The number would likely decrease if there were
a gravel floor.

     One  mitigator  has reported treating bare-earth crawl-space
floors  as large as  2,500 ft2  (230 m2)  with a single individual
suction pipe (Reference 6).  This is accomplished  by elevating much
of the membrane above the floor,  creating an air space underneath
to facilitate suction field extension.  Permeable matting beneath
the membrane might  accomplish a  similar objective, but would add
significantly to the cost.

     With  individual-pipe SMD systems,  the suction  pipe  almost
always penetrates the membrane  at a central  location.   Multiple
pipes are spaced uniformly.

     With  SMD  systems  using sub-membrane  piping, one straight or
snaked length of perforated  piping is commonly laid down the middle
of the  crawl  space,  generally parallel to the long  walls  of the
house (References l, 4, and 8).   More extensive interior patterns
have sometimes been tested  in research studies (References 1 and
10).  In a few cases,  the perforated  piping has been looped around
the crawl-space perimeter (Reference  1).   Piping  looped beside the
foundation walls  might improve  the chances that the system will
treat radon  entry routes  associated with block walls;  however,
since the  suction will  be drawn  immediately beside the perimeter
seam of the membrane, any leaks at that seam could degrade system
performance.   Data are currently  inadequate to enable guidance
regarding   the   preferred   piping   pattern   under   different
circumstances.   In  commercial practice,  good experience has been
obtained with a single  central length of piping  (Reference 4).

-------
Diameter and Type of Suction Piping

     The flows commonly encountered in SMD systems operating with
90-W fans are:  20 to 100  cfm  (10 to 50 L/sec) for individual-pipe
systems (with the higher flows being obtained when there is gravel
on the floor); and 100 to  130 cfm (50 to 60 L/sec) for systems with
sub-membrane piping (and no gravel).  At these flows, the commonly
utilized 4-in.  (10-cm)  diameter piping  should be  sufficient to
avoid undue suction loss due to flow friction.

     Either Schedule  40 or lightweight polyvinyl  chloride (PVC)
piping can be considered for the rigid piping above the membrane.
Flexible piping might be  considered  for  the sub-membrane piping,
although use  of rigid  perforated  pipe below the  membrane could
reduce the risk of the piping's being crushed underfoot.

Method of Installing Suction Pipes Through Membrane

     Individual SMD  suction pipes  must be  installed  through the
membrane in  a manner which prevents  their  open end  beneath the
membrane from resting flush against the dirt floor, thus restrict-
ing air flow into the pipe.  Also,  it is desirable to create some
form of cavity  beneath  the membrane where the pipe penetrates —
analogous to the sub-slab pit often excavated beneath SSD suction
pipes — to reduce suction losses as the sub-membrane gas acceler-
ates to pipe velocity.

     Researchers  and mitigators have  utilized several creative
designs for accomplishing these goals.   One simple approach is to
install an inverted tee  fitting on the bottom of the pipe, with the
cap of the tee resting on  the soil and the two open ends  of the cap
directed  horizontally  beneath  the membrane.   The  sub-membrane
cavity is created by attaching the membrane to the vertical suction
pipe (with a hose clamp and caulk)  perhaps 1 ft (0.3 m)  above the
tee, causing the  membrane to  be elevated above the floor at that
point.  Another approach  (References 2 and  5)  is  to:   excavate a
pit  beneath  the  membrane at  the  point where  the suction pipe
penetrates; place a sheet  of plywood over  the  pit  beneath the
membrane; and install the pipe through a hole in the plywood with
the open end of the  pipe  suspended in  the pit.   Other mitigators
have used supports beneath the membrane (such as an inverted bucket
or a specially  fabricated wooden flange) to support the vertical
suction pipe with its open end above the floor,  while at the same
time elevating  the membrane above the floor to create a cavity
(References 6 and 9).

     There are  currently  insufficient  data  to suggest the condi-
tions under which one of these approaches might be better than the
others.  The sub-membrane pit/plywood approach may be slightly more
complicated,  but  has  the  advantage of keeping the  membrane flat
against the crawl-space floor, reducing the risk that an elevated
section of membrane might get ruptured by foot traffic.   The sub-
membrane pit might occasionally also expose fissures or permeable

-------
strata beneath the soil surface, which might aid in extending the
suction field.   The  approaches that elevate the  membrane  may be
easier to  implement,  and can aid in extending  the suction field
beneath the membrane  by creating an air space  under  part  of the
membrane.

Nature of Membrane

     In almost  all cases, 6-  to  10-mil (0.15 to  0.25  mm)  thick
polyethylene sheeting  or equivalent is used as the SMD membrane
material.   Most  mitigators use at least 8-mil (0.20 mm) equivalent.
Cross-laminated polyethylene equivalent to  8-  to  10-mil (0.20 to
0.25 mm) thickness is  somewhat more expensive,  but is often used
due to improved  puncture resistance.  Material stabilized to resist
ultraviolet  (UV)  radiation  is preferred.   The  cross-laminated
sheeting can be  sealed effectively  at  seams and at the perimeter
wall using urethane caulk; standard polyethylene does not bond well
with urethane caulk,  so that some other adhesive must be used.

     Heavier material, such as  rubberized roofing material, should
be laid over the membrane along expected traffic routes.

Extent of Membrane

     Mitigators  almost universally cover the  entire crawl-space
floor with the membrane when installing SMD  systems.  Tests in two
houses under favorable conditions (each having gravel on the floor
and having sub-membrane perforated piping) showed that good indoor
radon reductions could be achieved even when only a portion of the
floor was  covered (Reference  1) .   Further testing  is  needed to
determine under what  conditions portions of the floor might be left
uncovered.

Degree of Membrane Sealing

     Most  mitigators usually  seal  the SMD  membrane completely,
including the seams  between  sheets,  and the junction between the
sheeting and  the perimeter foundation wall  and interior support
piers  (References  4,  5,  and  6).  An analysis of SMD installation
costs (Reference 12)  suggests that sealing can add  $60-$120 to seal
seams between sheets; $100-$250 to seal those seams and to seal the
perimeter using a simple bead of caulk  or adhesive;  and  up to $600
to seal sheet seams and to seal the  perimeter by wrapping the edge
of the membrane  around a furring strip which is then attached to
the wall.  Many mitigators seal the perimeter using a simple bead
of caulk or  adhesive; such bonds are  reportedly  so good in some
cases  that  the  membrane  can  rupture before the  bond  is broken,
appearing to make the added expense of a furring  strip unnecessary.

     Data are currently  insufficient to enable guidance regarding
the conditions under which complete sealing  of the membrane can be
avoided.

-------
     Limited  testing  in six  houses with  and  without complete
sealing of the membrane (References 1 and 9)  suggests that sealing
can sometimes provide limited additional reductions  in living-area
radon  concentrations  achieved by the SMD system,  and can reduce
system flows.   However,  in  all  six  cases,  very good  indoor
reductions  were achieved even  without  the membrane sealed.   In
other  crawl-space  study  houses,  the degree of sealing was not
varied;  the  SMO  systems were  tested  with  the membrane  either
completely  sealed  (References  2,  8,   and   10)  or  not  sealed
(Reference  2).  The number  of these other houses is too few, and
the number  of other variables being varied simultaneously is too
great, to permit any assessment of the effects of membrane sealing.
However, all  but one  of the houses without membrane sealing were
reduced below 4 pCi/L  (150 Bq/m3) .

MITIGATION  SYSTEM OPERATING VARIABLES

SMD Fan Capacity

     Most research installations  and most mitigators usually use
the 90-W in-line duct fans common in the radon mitigation industry
(270 cfm, or 125 L/sec, at zero static pressure,  and about 1.5 in.
WG, or 370  Pa,  at  zero flow) for  SMD systems.   Tests have shown
that good radon reductions can be achieved with  less fan capacity
when sub-membrane communication is good; i.e., when there is gravel
on the floor (Reference  1)   or  when the  membrane  is  elevated to
provide an  air  gap beneath  the membrane  (Reference 6).   However,
smaller fans will probably often achieve less of  a radon reduction,
especially when the membrane is not fully sealed  (Reference 1) .  At
least  one mitigator  (Reference 5)  occasionally uses  the larger,
100-W in-line fans  (410 cfm, or 190 L/sec,  at  zero static pressure)
to help extend the sub-membrane suction field. No one has reported
testing low-flow/high-suction fans for SMD installations — 25 to
60 cfm (10  to 30 L/sec)  at zero static  pressure,   and  25  in.  WG
(6,200 Pa) or greater at zero flow.   Such  fans might be applicable
in some low-flow SMD cases,  but could suffer a dramatic reduction
in performance if air flow increased over time due  to leaks.

GEOLOGY/CLIMATE VARIABLES

     Data on two houses having high source terms (References 2 and
19) confirm the expectation  that  an increased source term may
necessitate  greater  care in  the design  of  a  SMD  system  (e.g.,
addition of  a BWD component).  Where the  underlying native soil has
relatively  high permeability, sub-membrane  communication may  be
improved, reducing the  number of  individual  suction pipes  needed
for a given floor area.  In two houses with permeable native soil
(Reference 2), "SMD" performed reasonably well without a membrane
(i.e., suction pipes embedded in the bare earthen floor).

MITIGATION SYSTEM DURABILITY

     Only very limited data  are available  on  the durability of SMD
systems.  In six SMD installations by EPA (References 1,  7, and 9),

-------
no degradation in indoor radon reduction performance was measured
over  periods  of  1  to  3  years  following  installation.    For
commercial SMD installations, mitigators report that they have not
generally had  to revisit SMD installations  under their warranty
programs, and  have generally not  received  complaints  from home-
owners regarding reduced performance or other problems with their
SMD systems  (References 4, 5, and  6).

     There have not been significant reports of fan failures with
SMD installations, beyond  those reported for  other  types  of ASD
systems.  A  couple cases have been reported  in Iowa and Florida
where rodents seriously damaged the membrane within perhaps  6 to 12
months  after  the  SMD  system  was  installed;  however,   other
mitigators  (References 4,  5,  and 6)  report that such membrane
damage has not been common in their areas.
                 CONCLUSIONS REGARDING DATA GAPS

     The major data gaps regarding radon mitigation in crawl-space
houses appear to be:

  1. The conditions  (in  addition  to inaccessibility  of the crawl
     space)  under which crawl-space  depressurization might  be
     preferred over SMD (e.g.,  the tightness of the crawl space in
     quantitative  terms, and  the  exhaust  fan  capacity).   The
     operating cost penalty of crawl-space depressurization  vs. SMD
     needs to be defined, as a function  of  crawl-space tightness
     and fan capacity/radon reduction performance.

  2. The  optimal  method for  distributing  suction  beneath  the
     membrane (i.e.,  the individual-pipe approach vs.  sub-membrane
     perforated piping in various  configurations), under different
     conditions (e.g., crawl-space floor area, nature of floor).

  3. The  optimal  degree  of  membrane  sealing  required  under
     different conditions.

  4. The conditions under which it may be possible to leave some
     portion of the crawl-space floor uncovered by membrane, if a
     portion of the crawl space is inaccessible.

  5. The  durability  of  SMD,   crawl-space depressurization,  and
     natural crawl-space ventilation systems, in terms of both
     radon reduction performance and materials lifetime.


                           REFERENCES

 1.  Findlay, W. 0., A.  Robertson, and A. G. Scott,  "Testing  of
     Indoor  Radon  Reduction  Techniques  in  Central Ohio  Houses:
     Phase  2  (Winter   1988-89),"   EPA-600/8-90-050  (NTIS  PB90-
     222704), May 1990.

-------
 2.  Pyle, B. E., and A.  D.  Williamson,  "Radon Mitigation Studies:
     Nashville Demonstration, " EPA-600/8-90-061 (NTIS PB90-257791) ,
     July 1990.

 3.  Pyle, B. E., and K.  W.  Leovic,  "A Comparison of Radon Mitiga-
     tion Options  for Crawl-Space School Buildings," in Proceed-
     ings;   The 1991 International Symposium on Radon and Radon
     Reduction  Technology.Volume  2f pages 10-73 through 10-84,
     EPA-600/9-91-037D (NTIS  PB92-115369), November 1991.

 4.  Anderson,  J. W., Quality Conservation, Spokane, WA, personal
     communication,  February  6, 1992.

 5.  Howell,  T.,  and D.  L.  Jones,  Radon  Reduction and Testing,
     Inc., Atlanta,  GA, personal communication,  February 5, 1992.

 6.  Shearer, D. J.,  Professional House Doctors,  Inc., Des Moines,
     IA, personal communication, February  6, 1992.

 7.  Scott, A.  G., A. Robertson, and W. O. Findlay, "Installation
     and Testing of Indoor Radon Reduction Techniques in  40 Eastern
     Pennsylvania  Houses,"  EPA-600/8-88-002  (NTIS PB88-156617),
     January  1988.

 8.  Gilroy,  D. G.,  and W.  M. Kaschak,  "Testing of Indoor Radon
     Reduction Techniques in 19 Maryland Houses," EPA-600/8-90-056
     (NTIS PB90-244393), June 1990.

 9.  Messing, M.,  "Testing of  Indoor Radon Reduction Techniques in
     Basement  Houses Having  Adjoining Wings,"  EPA-600/8-90-076
     (NTIS PB91-125831), November 1990.

10.  Dudney, C. S., D. L. Wilson, R. J. Saultz, and T. G. Matthews,
     "One-Year  Follow-Up Study of Performance of Radon Mitigation
     Systems Installed in Tennessee Valley Houses," in Proceedings^
     The 1990 International Symposium on Radon and  Radon Reduction
     Technology.  Volume 2f  pages  7-59 through 7-71, EPA-600/9-91-
     026b (NTIS PB91-234450),  July 1991.

11.  Hoornbeek, J., and J. Lago, "Private Sector Radon Mitigation
     Survey," in Proceedings;  The 1990  International Symposium on
     Radon and  Radon Reduction Technology.   Volume 3f  pages 4-17
     through 4-30, EPA-600/9-91-026C (NTIS PB91-234468),  July 1991.

12.  Henschel,  D.  B.,  "Cost  Analysis  of  Soil  Depressurization
     Techniques for Indoor Radon Reduction," Indoor Airf 1(3) : 337-
     351, September 1991.

13.  Henschel,  D.  B.,  "Indoor  Radon  Reduction  in  Crawl-Space
     Houses:   Evaluation  of the State  of  the Art," submitted to
     Indoor Air. April 1992.

14.  Turk, B. H., Santa Fe, NM,  personal  communication, April 3,
     1991.

-------
15.  Nazaroff, W. W.,  and S. M. Doyle,  "Radon Entry into Houses
     Having a Crawl Space," Health Physics. j48(3): 265-281, 1985.

16.  Turk, B. H., R. J.  Prill,  W.  J.  Fisk,  D.  T. Grimsrud, B. A.
     Moed, and R. G.  Sextro,  "Radon and Remedial  Action in Spokane
     River Valley Homes.  Volume 1," LBL-23430,  Lawrence Berkeley
     Laboratory, Berkeley, CA, December 1987.

17.  Fisher, E.  J., Office of Radiation Programs,  U. S. Environmen-
     tal Protection Agency, Washington, D. C. ,  personal communica-
     tion (January 9, 1992).

18.  Matthews, T. G. , D.  L. Wilson, R. J.  Saultz,  and  C. S. Dudney,
     Oak  Ridge  National  Laboratory,  Oak   Ridge,   TN,  personal
     communication,  May 1989.

19.  Nitschke, I., "Radon Reduction and Radon  Resistant Construc-
     tion Demonstrations in New York," EPA-600/8-89-001 (NTIS PB89-
     151476), January 1989.

20.  Osborne, M. C.,  D.  G. Moore, R.  E. SoutherIan, T. M. Brennan,
     and B.  E.  Pyle,  "Radon  Reduction in Crawl  Space House," J.
     Environ. Engineering. 115(3);  574-589,  1989.

-------
MULTI-POLLUTANT MITIGATION BY MANIPULATION OF CRAWLSPACE PRESSURE DIFFERENTIALS
            by:   Bradley H. Turk
                  Mountain West  Technical Associates
                  105 E. Marcy St. , Rm.  109
                  Santa Fe, NM   87501

                  Gregory Powell
                  Gregory Powell and Associates
                  707 Galisteo St.
                  Santa Fe, NM   87501

                  Eugene Fisher, Jed Harrison, and  Bryan Ligman
                  U.S. EPA Radon Division
                  401 M St. S.W.
                  Washington, DC   20460

                  Terry Brennan
                  Camroden Associates
                  RD #1, Box 222
                  Oriskany, NY   13424

                  Richard Shaughnessy
                  University of  Tulsa
                  2643 E. 22 St.
                  Tulsa, OK    74114
                                    ABSTRACT
       Two  demonstration case  studies  of  manipulating the  pressure  gradient in
 buildings  with crawlspaces  to control both indoor radon levels  and outdoor air
 ventilation rates,  while maintaining  energy efficiency are  discussed.   In one
 study,  the inaccessible crawlspace of a  house  with vented combustion applianc-
 es  was pressurized  with outdoor  air to limit soil gas entry into the crawl-
 space.   The technique  reduced indoor  radon concentrations and stopped back-
 drafting of the appliances.   To  avoid cold floors,  frozen pipes, and an energy
 penalty, the pressurizing air stream  was first tempered by  passing through an
 'earth tube' heat exchange  pipe  buried in the  soil surrounding  the house.   In
 the other  study,  mechanical ventilation  systems were installed  in  two natural-
 ly-ventilated school buildings to inhibit radon-laden crawlspace air from
 entering through the structures'  floors.   In one building,  a heat  recovery
 ventilator was used to depressurize the  crawlspace and to pressurize the
 classrooms.  In the other school  building, classrooms were  pressurized with an
 outdoor air supply  fan.  Besides  source  control of radon, the low  existing
 ventilation rates of the occupied zones  were boosted, while in  one school
 building heat recovery from the  depressurization exhaust  air was made possi-
 ble.

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

-------
                                 INTRODUCTION
      The nature of crawlspaces below buildings encourages and sometimes
demands the consideration of innovative alternatives to radon control.  Most
successful crawlspace mitigation in houses has employed sub-membrane de-
pressurization (SMD) or forced depressurization of the crawlspace (1,2,3).
However, significant air leakage through the crawlspace surfaces, the presence
of other hazardous materials in the crawlspace, backdrafting of combustion
appliances, and limited access to the crawlspace may require that different
mitigation options be considered.  These options can provide an opportunity to
address other issues in the buildings -- such as indoor air quality (IAQ) and
energy use.

      In this paper, we discuss two case studies of New Mexico buildings with
crawlspaces and elevated indoor radon levels.  The first study is of a house
with shallow, inaccessible crawlspaces where forced pressurization of the
crawlspaces using passively tempered air reduced radon entry into the crawl-
spaces and inhibited backdrafting of vented combustion appliances by slightly
pressurizing the occupied space.  The pressurizing air stream was first
tempered by passing through underground 'earth tubes', thereby reducing the
possibility of cold house floors, frozen service pipes, and a heating energy
penalty.  The second, ongoing study is of a cluster of elementary school
buildings with natural ventilation and asbestos-containing material in the
crawlspaces.  To investigate radon mitigation options for buildings having
crawlspaces with asbestos and for improving ventilation and indoor air quality
with a minimum energy penalty, different control techniques were installed in
two buildings.  A heat recovery ventilator (HRV) was installed in one building
to both depressurize the crawlspace and ventilate/pressurize the classrooms,
while a system to pressurize the classrooms with respect to the crawlspace was
installed in another building.

                               SITE  DESCRIPTIONS
HOUSE STUDY

      The house (HEP1) involved in the first study was part of the U.S. EPA's
(EPA) Office of Radiation Programs' (ORP) House Evaluation Program.  The
structure is small (57 m2),  with four  rooms,  and exterior walls constructed of
adobe block.  It is located approximately 20 miles (32 km) north of Santa Fe.
Three different additions were made to the house over its approximately 80
year life.  Each addition is built over a separate, shallow (0.2 to 0.3 m)
crawlspace that has no access.  The floor joists, in at least the kitchen
area, appear to be resting directly on the soil.  The crawlspace exterior
walls are relatively air tight with no air vents and stucco plaster extending
below grade.  The building has a flat, built-up roof with no attic space.
Three vented, LPG-fired space heaters and one domestic water heater are
located in the living space.  Indoor radon levels during the winter before
mitigation were approximately 26 pCi/L.

SCHOOL STUDY

      The three Santa Fe school buildings in the second study are being
jointly investigated as part of the EPA ORP's School Evaluation Program and a

-------
University of Tulsa  study of indoor air quality in schools.  The  buildings are
located together  on  the  same campus and were built approximately  50  to  60
years ago.  Each  building has a combination crawlspace and basement.  Only the
basement in building SF1 contains classrooms.   The crawlspaces  in these
buildings also have  rather air-tight exterior walls covered with  stucco.   They
are not vented to the outdoors.  Each crawlspace is accessible  through  a
hatch(es), with space heights ranging from 0.3 to 1.8 m.  Asbestos-containing
material and debris  from an earlier heating system renovation were found  in
all crawlspaces.   All buildings are single story, naturally ventilated  with
operable windows,  and heated by wall convectors from hot water  boilers.
Exterior walls are constructed of brick, clay tile, and concrete,  while the
flat roofs have some small attic spaces. Other basic building information is
summarized in Table  1.
                    TABLE 1. INFORMATION ON 3 NEW MEXICO SCHOOL BUILDINGS
BUILDING
FIRST FLOOR
FLOOR AREA
[# rooms]*
CRAWLSPACE AREA
(X of substructure)
BASEMENT AREA
(X of substructure)
[# rooms]*
RANGE OF
PRELIMINARY
SHORT-TERM'
SCREENING
(pCi/L)
RANGE OF
PRELIMINARY
LONG-TERM*
SCREENING
(pCi/L)
SF1 6940/645 39 61 1.7-8.6 0.6-3.6
(CONTROL) [8] [3]
SF2 6090/576 82 18 0.7-39 0.5-16
[6]
SF3 5600/520 93 7 2.0-22 0.7-9.6
[8]
* Number of rooms includes commonly occupied classrooms, offices, libraries, and lounges.
' Charcoal canister measurements from 3/1/91 - 3/4/91 and 3/22/91 - 3/25/91.
* Alpha track measurements from 3/12/91 - A/19/91.
Buildings SFl and  SF2  are connected by a heated hallway,  although the sub-
structures do not  appear to interconnect.  As seen from  the  table,  only
buildings SF2 and  SF3  had classrooms with significantly  elevated short-term
and long-term radon measurements during screening.  After being alerted to the
presence of elevated radon levels, the school has adopted an interim policy of
opening windows  at 7:00 am, before students arrive, and  keeping windows at
least partially  open throughout the school day.

                                  STUDY DESIGNS
HOUSE STUDY

      An  initial  attempt to provide radon control by forced crawlspace
depressurization  caused flue gases from the combustion  appliances to backdraft
into the  living area.   However, during the several hours  of its  operation,  the
technique caused  radon levels to be reduced below approximately  2 pCi/L.   Sub-
floor pressure  extension between the crawlspaces during operation of the
system and during subsequent diagnostic measurements indicated that forced
pressurization  of the  crawlspaces should inhibit soil gas transport into  the
crawlspaces.  In  addition,  tests suggested that this type of system would push
air up through  existing openings in the house floor, thereby pressurizing the

-------
living space  and forcing the proper draft on the combustion appliances.

      Because operation of this  system blows outdoor  air directly into  the
crawlspace, cold outdoor temperatures could cause  cold house floors, an
additional heat load for warming the air, and the  possibility of freezing
water pipes in the crawlspace  when air temperatures dropped below freezing.
To alleviate  this problem, the air was tempered by passing through O.lm
(4 inch)  PVC  'earth tube' heat exchanger pipes buried approximately 1.2 m deep
in the soil surrounding the house (4,5).  This earth-air heat exchanger system
has been  most often used to cool air coming into a building (6,7).  Figure  1
depicts the layout of this earth tube system.  The large thermal storage of
the soil  warms the air in the  pipe during the heating season and cools  the  air
during the summer.
            Two Porollel 4" Sched. 40 Earth Tubes Approx. 70' Long:
            4 Ft. Below Grade, Rising to Within 12" of Grade
            at Building Perimeter Where it is Protected by R—20
            Rigid Insulation.
      1/4" Condensate
       e«p Holes
:Pressurization Fan
and Filter Box
 Figure 1. Crawlspace pressurization system and earth tube configuration showing the details of pipe layout,
  pipe penetrations (designated 'M'), and test hole openings (designated 'IF') into the crawlspaces in the
                                New Mexico study house.
       An insulated box  containing the in-line  centrifugal fan was  located on
the  roof of the house to  filter the incoming air and to reduce the noise
associated with fan operation.   Once inside the  house, the pressurization pipe
branched to three penetrations  -- one for each crawlspace.  The flow and
pressure in each crawlspace  branch were adjusted by an in-line ball valve.
The  total air flow delivered to all crawlspaces  is approximately 0.042 m3/s
(90  cfm).

       A Femto-Tech model  R210F  radon monitor was installed to continuously
record hourly radon data  during the course of  the study (12/90 through 4/91).
Flows,  differential pressures (APs), and radon in grab samples were periodi-
cally  measured as the system flow rates and distribution of flows  were varied.
An electronic digital micromanometer (Neotronics model MP20SR) was used to
measure APs and flow pressures  (with a pitot tube).  Grab samples  were
collected in evacuated  300 cc alpha scintillation flasks and the sample
activity was counted using a portable counter/sealer (Pylon model  AB-5).

-------
After the  earth tube heat exchanger was installed,  an on-site data logger
continuously  collected average air temperature data over consecutive 30-minute
intervals  for approximately two weeks.  Air temperatures were monitored in the
house and  outdoor  air,  and in the earth tube air  stream at two locations: 1)
at grade level where the pipe entered the soil, and 2)  after leaving the soil,
where the  pipe entered the house.

      The  continuous temperature data were used in  a mathematical model with
empirically-derived parameters to predict the air temperature at the outlet of
the earth  tube system as the outdoor air temperature changed.  A thermal
network was established for each 3.1 m (10.0 ft)  length of earth tube (7,8).
This network  consisted of concentric cylinders of air,  pipe, and soil around
the pipe (see Figure 2).
                                                [8] ANNUAL AVG. SOIL TEMP

                                                [7] 0.43 M (17.0 IN) SOIL


                                                [6] 0.43 M (17.0 IN) SOIL
                                                [5] 0.20 M (8.0 IN) SOIL
                                                [4] 0.10 M (4.0 IN) SOIL
                                                [3] 0.05 M (2.0 IN) SOIL
                                                [2] PIPE WALL
                                                [1] AIR IN  PIPE
           Figure 2. Cross section of the earth tube and soil showing the thermal network
                            temperature nodes and arrangement.
      For a given inlet air temperature and soil  temperature surrounding each
pipe section, a  matrix of standard heat transfer  equations was solved for the
eight temperature nodes.   The new air temperature in the pipe was then used
for calculations in the succeeding pipe section.   Thus,  for each 30-minute
monitoring interval,  outdoor air would progressively gain (or lose) heat until
it reached the end of the earth tube system, where the  calculated outlet
temperature was  compared to the measured temperature.   The length of the pipe
section was determined primarily by the time to perform the extensive computa-
tions. Values for the numerous parameters required for  the calculations are
from handbooks (7,9,10,11,12) and manufacturers and are shown in inch-pound
units in Table 2.   The air flow rate in the pipe  was assumed to be 0.044 m3/s
(94 cfm).

-------
     TABLE 2. FIXED PARAMETER VALUES USED IN THERMAL NETWORK CALCULATIONS
MATERIAL
Air - Single Pipe
Air - Two Pipes
PVC Pipe
Soil
CONDUCTANCE
(Btu/h-ft3-°F)
3.454
1.985
--

CONDUCTIVITY
(Btu/h-ft-°F)
--
--
0.1201
1.50
iterated
DENSITY
(Ib/ff)
0.063*
0.063*
70
0.20
iterated
SPECIFIC HEAT
(Btu/lb-°F)
0.2447
0.2447
0.45
0.20
* Site elevation is approximately 1680 m (5500 ft) above sea level.

Because some of the parameter values are dependent on specific site conditions
(soil density, soil thermal conductivity, and average temperature of the
surrounding soil), iterative calculations were performed using different
values for these three parameters.  Calculated and measured tube outlet
temperatures over the entire two-week monitoring period were compared graphi-
cally for each iteration.  The values identified in Table 2 as 'iterated'
resulted in the best graphical agreement.  The optimum average temperature of
the soil at node eight (8) and beyond was found to be 8.6 °C (47.5 °F).

SCHOOL STUDY

      Blower door tests of SF2 and SF3 indicated that crawlspace depressuriza-
tion and reduction of indoor radon levels could be achieved with less than
approximately 0.94 m3/s  (2000  cfm) of  exhaust flow from each crawlspace.
However, concern over disturbing asbestos particles with air flow through the
crawlspaces prompted the school district to have the asbestos removed from SF2
in February 1992.  Asbestos-containing material remained in SF1 and SF3.  This
situation provided an opportunity to examine different radon reduction
techniques for buildings with and without existing asbestos problems (SF3 and
SF2, respectively).  In addition, techniques were considered that could
improve ventilation and general IAQ, and reduce the energy penalty associated
with the additional ventilation (SF2).  Because building SF1 had satisfactory
initial indoor radon levels, it was selected to serve as a study control --no
radon or IAQ mitigation work will be conducted.  Mitigation work began in
March 1992 and.was substantially completed by mid-April 1992.

      In building SF2, where asbestos had been removed, the crawlspace is
being depressurized to reverse the pressure gradient across the classroom
floors using the exhaust fan of an air-to-air heat exchanger (HRV).  The heat
exchanger core scrubs available heat from the exhaust air stream and returns
it in the supply air stream to the upstairs classrooms and offices.  The
supply air is routed to the upstairs via newly-installed uninsulated ducts in
the crawlspace.  The supply air is then delivered to slightly pressurize the
rooms through hot water coils in the existing wall convectors.  To meet the
ASHRAE Standard 62-1989 recommendation of 15 cfm (0.007 m3/s)  of  outdoor
ventilation air/occupant (13), the supply air delivery was designed to provide
approximately 450 cfm (0.21 m3/s)  of outdoor air  to each classroom,  and
approximately 2800 cfm (1.32 m3/s) of  total  outdoor air to the building.   In
order to maintain satisfactory energy efficient operation of the heat exchang-

-------
er core, the exhaust air flow rate was designed at approximately 3000 cfm
(1.42 m3/s).   Since  balancing and adjustments  have  not been completed,  air
flow rates at this time tend to be lower than designed.  New heating system
valves and thermostats have been installed to assure that the supply air is
delivered at a suitable temperature.

      The asbestos-containing material still in the crawlspace of SF3 elimi-
nated the selection of any mitigation system that involved either the movement
of air or work in the crawlspace.  Therefore,  a supply air system that
pressurizes the occupied portion of the building to prohibit radon-laden air
from entering from the crawlspace has been installed.   The system was
designed to provide a total of approximately 3600 cfm (1.70 m3/s)  of outdoor
air (using ASHRAE 62-1989 recommendations) from a roof-top fan through
insulated ducts to all classrooms.  Because the system does not have a return
side, ventilation is achieved by supply air exiting the classrooms through air
leaks in the ceilings, walls, and floors, and around doors and windows.  The
incoming supply air is warmed by passing through hot water coils in each room.
The coils are fed by a new glycol/water loop that was added to the hot water
heating system, and the temperature of the supply air is controlled by a
sensing bulb and three-way valve at each coil.  Doors and windows must be kept
closed to maintain a pressurized condition in the classrooms.  To promote the
closed condition by providing better temperature regulation, new heating
system thermostats and valves were installed on the existing wall convectors.

      Two classrooms in each of the three buildings were selected to have
sensors installed to continuously monitor temperature, relative humidity,
carbon dioxide (C02),  differential pressures,  and radon.   Carbon dioxide  is
passively monitored by a Gaztech model 1070 Ventostat, while radon in the
classrooms is monitored with the Femto-Tech model R210F.  Differential
pressures are measured across the floor to the crawlspace and across the
exterior wall (both referenced to the classroom) with a Setra model 264
bidirectional pressure transducer with a range of ± 25 Pa.  Radon is monitored
at one crawlspace location in each building with a Pylon model AB-5 monitor
with a passive scintillation flask.  Temperatures are also monitored in each
crawlspace.  Air flows in the ducting of the mitigation systems are determined
from stationary pitot tubes attached to pressure transducers (as above).
Outdoor wind speed and direction, temperature, humidity, and C02 are monitored
at one location on SF1.  A data logger in each building is connected to all
sensors and stores 30-minute average values.  This instrumentation was in-
stalled from February 1992 through May 1992 and will remain in place into the
heating season of 1992-93.  Ventilation rates were measured in the three
buildings under conditions of windows/doors closed and windows/doors open
using an SF6 dilution/decay technique before mitigation systems were in-
stalled.

       The initial diagnostic measurements showed that reversal of the
pressure gradient in the crawlspaces of SF2 and SF3 will, by itself, control
radon entry.  Calculations indicate that the additional ventilation from the
mitigation systems installed in each building provides redundant radon
control.  By switching the supply and exhaust fans on and off, this provides
an opportunity to study four different mitigation conditions in SF2: 1) crawl-
space depressurization alone, 2) classroom pressurization with ventilation, 3)
classroom ventilation alone, and 4) all (1 through 3) combined.  These and
other tests will be conducted during the summer of 1992 and heating season of
1992-93.

-------
                            RESULTS AND DISCUSSION
HOUSE STUDY

      Data from flow, pressure, and radon measurements are summarized in
Figure 3.   Figure 3B exhibits the flow measurements made in each pipe branch
and the main pipe out of the fan.  Figures 3C to 3H present data from radon
grab samples and AP measurements at each of the 6 test holes drilled through
the floor into the crawlspace.  The AP data shown on the figure are for the
change in AP measurements made between the crawlspace and outdoors with the
system on and the system off.

      The figure shows changes in the various measurements as the crawlspace
pressurization system was modified or turned off ('baseline' periods).
Columns 2 to 4 represent the 'initial' pressurization system that used
untempered outdoor air and one penetration into the crawlspace.  Flow rates
designated as '1/2', '2/3', and  '5/6' of maximum flow were from the position
of the fan speed switch.  The system was then turned off (column 5, 'post rait
1 baseline') to prevent freezing of water pipes.  All remaining columns (6 to
14) represent changes to the pressurization system with three penetrations
(Ml, M2, M3) and the earth tube heat exchanger in place.  Adjustments in the
ball valves to change air flows through different pipe penetrations are
denoted in columns 7 to 13 by the term 'modify'.  Changes to flows in individ-
ual pipes are also noted.

      By directing more flow to various pipes, APs and radon concentrations
are affected differently in each of the crawlspaces.  The most efficient
operation of the mitigation system was found to involve closing pipe M2 and
adjusting the flows in pipes Ml and M3 to be equal.  The data show that indoor
radon levels were reduced to approximately 4 pCi/L after optimization of the
crawlspace pressurization system.  The inability to reduce indoor radon levels
below 4 pCi/L, may be due to several factors.  First, it is possible that
diffusion from the soil floor -- not convective flow of soil gas --is the
more important mechanism for radon entry into the crawlspaces of this house.
This could explain why radon concentrations in the crawlspaces remain elevated
even when crawlspace pressures are high enough to stop radon entry via
convective flow.  In this explanation, reduction in indoor radon levels would
be due to dilution of the crawlspace radon from the ventilation air provided
by the pressurization system.  A second explanation involves the construction
details of the wood floors to the crawlspaces.  Since at least the kitchen
crawlspace is divided into separate zones by the joists that are resting
directly on the soil, air from the pressurization pipe that is forced into one
end of the crawlspace must pass through soil below the joists to reach the
zone farthest from the pipe.  Figure 4 illustrates this process.

-------
   00
   c
   H
   ID
   Cf
   rr
a hi
n  a>
M  en
o  v>
l-l  C
Sil
O  N
3 O

Cf
  cn
r^ tt
S §

•*«
INITIAL
CRAWL
                         PASEMN.E
                INITIAL
                CRAWL
                                                     INITIAL                                                         MODIFY             MODIFY
                                                     CRAWL             EARTH                               MODIFY   Ml  UP    MODIFY   M 1  =• M3

                                                     |f    flfl   ALLIES  MOD.FY  M^N  TO  M^WN IfeW  ^2=0^  S^NG
                                                      FLOW   BASELINE   fyjj.   MJ DOWN   M3 UP  M3  DOWN  SEALJKJG   SEALING   SEALING   BIG FAN   BASELINE
       (A)
       (B)
24-
o-
:
I


20

, r"l r^ H1
•H r"-.
4.4 4.6 1 1
1 1 1 1 1 1
                 20 --
                           MITIGATION PIPE FLOW
                           M1

                            +
                                  M2
                                     +
            M3    TOTAL
            	1	
       (O  g^|
                0.1
                       TEST  HOLE  RADON  &  DELTA  P
                                    IF1
 RADON      DELTA P

H	1	h
                100 -C
                i.o

                0.1
                100 ^

                                    IF2
                                    IF3
                                 n
                                                             n
                                                       i
                                                                                                                         •14
                                                                                                                          12
                                                                                                                          10
                                                                                                                          8
                                                                                                                          6
                                                                                                                         •12
                                                                                                                         •10
                                                                                                                         •8
                                                                                                                         6
                                                                                                                         •4
                                                                                                                         2


                                                                                                                         •14
                                                                                                                         12
                                                                                                                         10
                                                                                                                         8
                                                                                                                         6
                                                                                                                         4
                                                                                                                         2
  g
  •
  S
                100
              »
               io.o
       , s       .
       (G)  g'-
                0.1
                                    IF5
                                 n
                                                             n
                                                      ^	&
                                                                                                                                                            H
                                                                                                                                                            12
                                                                                                                                                            10
                                                                                                                         •4
                                                                                                                         2


                                                                                                                         14
                                                                                                                         12
                                                                                                                         10
                                                                                                                         8

                                                                                                                         4
                                                                                                                         2
                                   IF6
                                                                              n
                                                                                                                    9.8
                                                                                                                                                            :!24?
                                                                                                                                                            10 "^
                                                                                                                                                            B »-
                                                                                                                                                            6 iS
                                                                                                                                                            '* ui
                                                                                                                                                            2 Q

-------
                               HOUSE INTERIOR
        Wood Floor: Joists on Soil—v
                  Decking —y     \
                                         Crawlspaee
                                         Pressure.  -
                                         Mitigation
                                         Pipe
                                                                      i i i—
                                     SOIL
Figure 4.
Air flowing through the soil below the crawlspace floor joists may pick up radon despite the high
                        crawlspace pressures.
The farthest zone would be  pressurized,  but would have acquired radon from  the
air passing through the soil.   In both explanations, we found that crawlspace
radon levels are highest  at test  holes farthest from the penetration of the
pressurization pipe, and  that  this radon would be forced into the first floor
living area.

      A number of cracks  and holes were sealed to reduce the air leakage  in
the floor between the house and the crawlspace - those periods are denoted  by
'sealing'.  There was no  observable change in APs,  or in radon levels in  the
crawlspaces or living area.  To increase air flows in the pipes, a larger fan
was installed for one test  period ('big fan').  Although crawlspace APs were
affected, there was no significant change in indoor radon levels.

      Air forced from the crawlspace by the pressurization system caused  the
first floor living area to  be  slightly pressurized (0.5 Pa) with respect  to
outdoors, at floor level.   With the pressurization system off, a typical  AP
across the living area exterior walls of this house was approximately -3  Pa.
This change in living area  pressures caused the flues of the combustion appli-
ances to draft properly --  reducing flue gas entry into the house.

      The pressurizing air  appeared to slightly increase the house ventilation
rate from 0.3 ach (estimated from blower door test data).  On two separate
occasions with the mitigation  system operating, ventilation rates were
estimated by observing the  radon  dilution rate for several hours after the
system was turned on.  The  ventilation rates were calculated to be 0.4 ach  for
a low pressurization flow condition and 0.6 ach for a maximum flow condition.

Earth Tube Temperature and  Energy Use Predictions
      Figure 5 shows  the  actual measured temperatures for the earth  tube

-------
   system  and the calculated outlet temperatures using the optimal parameter
   values  in the thermal network model.  Note  that the calculated and actual
   outlet  temperature curves converge after  approximately eight days, then
   diverge again after a break in the data  (because of instrument failure)  from
   March 24 to March 25.  The convergence is due to the time required for the
   calculational procedure to establish a realistic temperature profile for the
   soil close to the earth tube.  The divergence probably results from the
   calculational procedure not having the history of the missing outdoor air
   temperatures to accurately predict tube and soil temperatures starting on
   March 25,   In general, though, the calculated temperatures very closely
   predict the measured temperatures.  The effectiveness of the earth tube  heat
   exchanger in moderating incoming air temperatures is demonstrated by the
   stability of the outlet temperature despite large changes in the inlet
   (outdoor) air temperatures.
65
C 18 . 3D

60

n C 1 5 . 6}
t? 55

Oi C 1 2 . BD
0)
-o 50

c 10 , 03
LL
45
nj *
« C ^ . 23
i_i 40
w C 4 • 4}
cr
3 35
< C 1 • ^}
cr
HI 30
a
2 C - 1 , 1D
LU
t- 25
C- 3 . 93
20

£
5




i





s jfc-k^ <*.
"^SS^*"^^

I
', .. 2:
} \\ ||:
: ' !.!':!
\ 'f, s
\K ' \
\
\








£
^

















^
\^






i



i
s
::
- *• _/
»". r
• •

i :

? *: '^^vi'^
•i _SrT^ ._jl-* X^. V**^V ^ ~**f
s:^^*1^* ^^ y
/~ '
\ '


f il
i
"*
:
: =
f :
S

*^s



-. :. ";
i- • *d i
::
u.

7 i
/-IvLr^S. fi .j
f ^^ A ^o^yV^^^v"'
^


.

;




\

x :
~ :
"^~




*_
"





V.
Cv/
N/^
**'






i\
•


, \

;*:











^
'


















0:00 0:00 0:00 0:00 0:00 0:00 0:00 0:00 0:00 0.0
14Mgr 16Mar 18Mar 2QMar 22Mar 24Mar 26Mar SBMar 30Msr 1Ap
                                             DATE
                                   MEASURED TUBE OUTLET TEMPERATURE

                                   CALCULATED TUBE OUTLET  TEMPERATURE

                                   MEASURED TUBE INLET TEMPERATURE
Figure 5.  Measured inlet and outlet temperatures for the earth tube are compared with predicted outlet temperatures,
                 Note the large diurnal temperature swings in the tube inlet temperature.
         The  energy saved by tempering the incoming air with the earth tube heat
   exchanger  was  computed using simulated daily  average outdoor temperatures for
   an entire  year (base data from the nearby U.S.  National Weather Service

-------
reporting station at Espanola) .   The thermal network model (with optimized
parameter values) was  then applied to the year-long outdoor temperature
profile.   Figure 6 shows  the simulated outdoor  temperatures and corresponding
earth tube outlet temperatures  that were calculated.   Differences in the daily
average energy necessary  to heat the tempered and untempered air are also
shown for a heating season from September 26 through May 17 assuming that all
pressurizing air would be forced to the upstairs  living area.  The heating
season difference in total energy usage (3160 kwH for outside air vs.  2490  kwH
for earth tube air) is approximately 670 kwH (0.25 x 10" J) or approximately
$35 (using a thermal conversion efficiency of 70% and propane at $0.26/L
      The minimum  tube outlet temperatures  that are likely to be encountered
were calculated  for  unusually cold conditions  that were simulated for  a 10-day
period in January.   With simulated outdoor  temperatures dropping as  low as
-28.9 °C (-20 °F) for this period, the  calculated tube outlet  temperatures
never went below 4.9 °C  (40.8 °F) .  This  suggests that it is unlikely that
tube outlet  temperatures will be cold  enough to freeze pipes  in the  crawl -
space.  All  tube outlet temperatures calculated for the above predictions were
based on the assumption that air would be blown through the earth tube system
for the entire year.  In this way, energy from the warm air during the summer
would be stored  in the soil to be recovered during the colder months.
                                                                          I
                                                                          LU
                                                                          in
                                                                          D
                                                                          a.
                                                                          LU
                                                                          Z
                                                                          LU
                                                                          <
                                                                          O
<
o
       25
          JUL  AUG  SEP  OCT  NOV  DEC  JAN  FEB  MAR  APR MAY  J UN
                                       DATE
                             CALCULATED TUBE OUTLET TEMPERATURE
                             SIMULATED OUTSIDE TEMPERATUPE

                             ENERGY USAGE USING EARTH TUBE
                            — ENERGY USAGE USING OUTSIDE AIR
   Figure 6.  Daily average temperatures predicted using the thermal network model and simulated outdoor
     temperatures.  The daily energy necessary to heat tempered air from an earth tube and untempered
                          outdoor air are shown as separate curves.

-------
      There are  several  drawbacks to the system installed in this house.
First, forced pressurization of the crawlspaces can possibly push other
pollutants that  are  specific to the crawlspaces (such as termiticides and
bioaerosols) up  into the occupied areas.  Although no measurements were made
of these pollutants  in the  occupied zone,  there was no evidence to suggest the
presence of other  important pollutant sources in the crawlspaces.  The second
disadvantage involves the installation of the buried earth tube piping.
Because a large  mass of  soil is necessary for the heat reservoir, a sizable
space near the house is  disturbed by excavation equipment.  Care must be
exercised during the excavation to avoid disturbing underground utilities and
to assure an adequate slope for condensate to drain to weep holes.  If
condensate is permitted  to  collect in the pipe, biological growths (e.g.,
molds and fungi) may flourish and be transported to the indoor air.

SCHOOL STUDY

      Because installation  of mitigation systems was not substantially
completed until  mid-April,  only preliminary data from a 7-day pre-mitigation
period and a 5-day post-mitigation period are presented here.  Figure 7 is an
example of data  from Rm.  15 and the crawlspace in SF2 (with the HRV) that
shows continuous plots for  radon and APs during the pre-mitigation and post-
mitigation periods.
O
a
z
O
a
                                            +
                                                                         •- 8
                                                                         •- 12
                                                                               <
                                                                               a
                                                                               LU
                                                                               Q
0:00    0:00   0:00    0:00
2Apr    5Apr   BApr    11Apr
                                 0:00
                                         0:00
                                         17Apr
0:00
20Apr
0:00
23Apr
0:00
2BApr
                                     DATE
                            RM  15 RADON
                            CRAWLSPACE RADON
                            RM  15 CRAWL-1st FLOOR DELTA P
                      	• RM  15 OUTS IDE-1st FLOOR DELTA P

 Figure 7. Continuous data during pre- and post-mitigation periods for Rm. 15 and the crawlspace in SF2.

-------
      Each of these periods includes a weekend when the school buildings were
closed and unoccupied (April 4 and 5, and April 23 and 24).  The pre-mitiga-
tion data shows that the open-windows policy is effective at reducing indoor
radon levels during the school day.   However, operation of the HRV system that
both depressurizes the crawlspaces and pressurizes/ventilates the classrooms
lowers radon levels even more during the school day and keeps the levels low
during unoccupied periods.   With the HRV operating, the AP data clearly show
that the crawlspace is depressurized approximately 7 Pa with respect to the
first floor, and that the first floor is pressurized approximately 3 Pa with
respect to the outdoors.  When windows and doors are opened during the school
day of the post-mitigation period, the first floor pressurization is reduced.

      These data are summarized on Table 3 along with C02 measurements  and
data from SF1 (control) and SF3 (classroom pressurization).  Results are shown
for the two classrooms in each building with continuous monitoring equipment.

      During the pre-mitigation period in all three buildings, the indoor
radon concentrations over the weekend tend to be slightly higher than for the
week nights (data not shown), and much higher than for the school day.   This
result illustrates the limitation of using radon monitors that continuously
integrate over 24-hour periods (or longer) to accurately measure the occupant
exposure in schools (or other intermittently occupied structures).

      The pressurization system in SF3 is very effective in controlling indoor
(and crawlspace) radon levels.  The pressurization of the building is quite
robust, even though the mitigation system was not delivering designed air
flows during the post-mitigation period.  Pressurization is significantly
diminished during the school day when doors and windows were opened by the
faculty and students.  The data show that indoor radon levels also fell in the
control building (SF1) during the warmer post-mitigation period, although
crawlspace radon levels were not reduced as much as in the two buildings with
mitigation systems operating.  Part of the change is probably due to the lower
driving forces (APs) resulting from milder outdoor temperatures during the
second period.

      Average C02 concentrations  during the  occupied hours were as high as
1900 ppm in Rm. 17 (SF2), with a maximum 30-minute average of 4500 ppm and a
maximum single measurement of 4800 ppm.  Carbon dioxide levels are signifi-
cantly reduced in the four classrooms of SF2 and SF3 by operation of the radon
control systems, while C0a  levels in SF1 change very little during the  second
monitoring period.  These data suggest that the systems, despite not operating
at designed capacity, are improving the general IAQ, as well as reducing
indoor radon levels, in SF2 and SF3.

-------
  Table 3.  Summary of Selected Pre-  and Post-Mitigation Data  from Santa Fe  Schools
                                               .(April  1992).

Building
SF1*
PRE-mhig.

POST-mitig.

SF2
PRE*4Dltlg.

POST-mitig.

SF3
PRE-autig.

POST-mhig.


Period

Statistic

Weekend:*
WeekDajr.t
Weekend:
WeekDay:
Mean
Period (fan)
Mean
Period (bn)
Mean
Period (hrs)
Mean
Period (hrs)

Weekend:
WeekDay:
Weekend:
WeekDay:

Weekend:
WeekDay.
Weekend:
WeekDay:
Mean
Period (hrs)
Mean
Period (his)
Mean
Period (fars)
Mean
Period (hrs)

Mean
Period (hrs)
Mean
Period (hrs)
Mean
Period (hrs)
Mean
Period (hrs)

Radon Concentration
Rm9
5.0
60.5
2.6
40.0
0.9
60.5
0.5
24.0
Rml5
19.0
60.5
8.0
40.0
OS
603
0.6
233
Rml
28.
45.5
9.4
293
0.4
60.0
0.6
23.5
Rml3 I
55
60.5
2.7
40.0
13
60S
0.7
24.0
Km 17 |
63
60S
2.6
40.0
0.4
603
OS
24.0
1 Rm6
12.
603
4.1
40.0
0.6
603
0.6
24.0
(pCi/1)
Crawl
29.
603
17.
40.0
20.
603
7.6
24.0
Crawl
41.
603
29.
40.0
11.0
603
10.
24.0
Crawl
68.
603
49.
40.0
0.4
603
10.
24.0
Differential Pressure (Pa)*
CrawHst Ft
Rm9 .
0.4
573
-0.1
40.0
0.2
573
-02
24.0
RmlS
0.7
573
03
40.0
-6.9
573
-73
233
Rml
03
573
03
40.0
-4.1
573
0.8
24.0
Outdoor-lst Fir
Rm9
23
563
1.1
40.0
13
573
0.6
24.0
RmlS
0.9
573
03
40.0
-32
573
-2.1
24.0
Rml
2.1
57.0
0.6
40.0
-6&
573
-OS
24.0

Carbon Dioxide (ppm)
Rm9 |
450
573
1100
253
400
57.0
1250
24.0
RmlS
750
573
1000
25.0
350
573
750
23.0
1 Rml
600
573
950
253
350
573
600
24.0
Rml3
350
603
750
253
350
603
650
24.0
Rml7
550
573
1900
253
400
573
850
23.0
Rm 6
600
573
700
253
300
573
350
24.0
* Building SF1 is the control building where no mitigation system was installed.
f The 'weekend* is defined as beginning Friday at 17:00 and ending Monday at 05:00.
  The 'weekday' is defined as beginning at 07:00 and ending at 15:00 for a day falling on Monday through Friday.
* All differential pressures are referenced to the first floor classroom.

-------
      Ventilation rate measurements made before mitigation are summarized in
Table 4.   Although ventilation rates have not been measured since installation
of the mitigation systems, estimates of ventilation for individual classrooms
in SF2 and SF3 have been made based on measured air flow from the supply
diffusers.  At the air flow rates for the post-mitigation period, ventilation
rates would be approximately 4.3 and 4.8 ach for Rm. 1 and Rm. 6, respective-
ly, in SF3; and 1.6 and 1.4 ach for Rm. 15 and Rm. 17, respectively, in SF2.
Clearly,  ventilation rates have been greatly increased in SF3 and have
probably been slightly increased in SF2.  Larger increases are expected after
completion of system balancing and adjustments.


   TABLE 4. VENTILATION RATE MEASUREMENTS BEFORE INSTALLATION OF MITIGATION
                                    SYSTEMS
BUILDING
SF1:

SF2:

SF3:

Building Avg.
Classroom Range
Building Avg.
Classroom Range
Building Avg.
Classroom Range
WINDOWS/DOORS OPEN
(ACH)
0.8
1.0 - 1.6
1.1
0.7 - 1.1
1.7
1.6 - 2.1
WINDOWS/DOORS CLOSED
(ACH)
0.3
0.5 to 0.6
0.2
0.1 to 0.3
0.4
0.3 - 0.4
                                    SUMMARY
      The two case studies presented in this paper suggest that alternative
techniques for radon control exist that may also improve the general air
quality in a building.  Furthermore, the additional energy usage to provide
these benefits can be lessened through the use of innovative or existing  'off
the shelf technologies.  The specific control technique used in the house
study may not be applicable to a large number of other houses, since its
crawlspace construction is not often found in other regions of the United
States.  However, the backdrafting of appliances by depressurization of the
crawlspace may occur occasionally - and will require consideration of control
methods such as the one described here.  In addition to providing passive
heating of ventilation air during the winter months, the earth tube heat
exchanger can also furnish some space cooling and dehumidification during the
summer months.

      The systems demonstrated in the school study may have more widespread
application.  For schools with asbestos in the crawlspace or basement, the
classroom pressurization technique can not only significantly reduce indoor
radon levels, but also boost classroom ventilation rates (for buildings with
natural ventilation) and create an improvement in general IAQ, without
disturbing the asbestos.  This technique also may be an effective radon and
IAQ control technique for buildings with slab-on-grade construction.  The HRV

-------
installed in the other school building features a commercially-available unit
that not only provides radon control and improved IAQ, but also will reduce
the amount of energy necessary to heat the additional incoming outdoor air.
Because the mitigation systems in the schools represent newer and more complex
technology, the maintenance staff must be properly trained so that the systems
continue to be effective over the long term.

                               ACKNOWLEDGEMENTS
The authors gratefully appreciate the technical assistance of Ron Simon, John
Anderson, and Francis Offermann; the identification of study sites by Ron
Mitchell of the State of New Mexico Environment Department; the administrative
support of Mike Miller of the EPA's Region VI Radiation Office and John
MacWaters of SC&A; engineering and construction management of M&E Engineering;
the interest and efforts of the construction contractors; and the cooperation
of the homeowner in the house study, and the faculty, staff, school district
officials, and student body at the Santa Fe Public Schools.  These studies
were funded by the U.S EPA Office of Radiation Programs' Radon Division.

                                  REFERENCES
1.    Henschel, D.B.  Indoor radon reduction in crawl-space houses: evaluation
      of the state of the art.  Submitted to Indoor Air.  1992.

2.    Pyle, B.E. , and Leovic, K.W.  A comparison of radon mitigation options
      for crawl space school buildings.  In: Proceedings: The 1991 Interna-
      tional Symposium on Radon and Radon Reduction Technology, Vol. 2, EPA-
      600/9-91-037b (NTIS PB92-115369), 1992, pp. 10-73 to 10-84.

3.    Pyle, B.E., and Williamson, A.D.  Radon mitigation studies: Nashville
      demonstration.  EPA-600/8-90-061 (NTIS PB90-257791), U.S. Environmental
      Protection Agency, Research Triangle Park, NC, 1990.

4.    Nuess, R.M.  Personal communication. 1991.

5.    Nuess, R.M., and Prill, R.J.  Radon control - towards a systems ap-
      proach.   In: Proceedings: The 1991 International Symposium on Radon and
      Radon Reduction Technology, Vol. 2, EPA-600/9-91-037b (NTIS PB92-
      115369),  1992.

6.    Francis,  C.E.  Cooling with earth tubes.  Solar Age. January 1984, pp.
      30 to 33

7.    Zoellick, W.  Predicted and observed performance of a buried earth-air
      heat exchanger cooling system.   Publication reference unknown.

8.    Kohler, J.T., and Sullivan, P.W.  TEANET user's manual.  Total Environ-
      mental Action, Harrisville, NH,  1979.

9.    Avallone, E.A., and Baumeister,  T.  Marks' Standard Handbook for
      Mechanical  Engineers. Ninth Edition.  McGraw-Hill, 1986, pp. 4-56, 8-
      198.

-------
10.    American Society for Heating,  Refrigeration,  and Air Conditioning Engi-
      neers.   1989 ASHRAE Handbook of Fundamentals.  ASHRAE,  Atlanta,  GA,
      1989,  pp. 3.14, 22.21,  22.8.

11.    American Society for Heating,  Refrigeration,  and Air Conditioning Engi-
      neers.   1991 ASHRAE HVAC Applications.   ASHRAE, Atlanta, GA, 1991, pp.
      11.4.

12.    Chemical Rubber Company,  gRC Handbook of Chemistry and Physics 62nd
      Edition.  CRC Press, Boca Raton, FL, 1982, pp. E-2, F-49.

13.    American Society for Heating,  Refrigeration,  and Air Conditioning Engi-
      neers.   ASHRAE Standard, Ventilation for acceptable indoor air quality,
      ASHRAE 62-1989, ASHRAE, Atlanta, GA, 1989, p. 10.

-------
                                                                     VII-6
     TWO EXPERIMENTS ON EFFECTS OF CRAWLSPACE VENTILATING ON
             RADON LEVELS IN ENERGY EFFICIENT HOMES.

           by:  T.D. Sterling1*; E.D. Mclntyre2,  E.M.  Sterling3

                 Professor,  School of Computing Science,
               Faculty of Applied Sciences,
               Simon Fraser University,
               Burnaby, B.C., Canada, V5A IS6.
               Telephone: (604)291-4685/733-1348/681-2701,
               Facsimile: (604)681-2702.

                 Author for correspondence.

               2 Hughes Baldwin Architects,
               300-1508 West 12th Avenue,
               Vancouver, B.C., Canada, V6J 1H2.

                 President,  Theodor D Sterling and Associates
               Ltd.,
               250-1122 Mainland Street,
               Vancouver, B.C., Canada, V6B 5L1.
                            ABSTRACT
     The reduction of radon levels in the living area of a
tightly sealed energy-efficient unoccupied home with a naturally
ventilated crawlspace in comparison with an older, traditionally
constructed home with a tightly closed, un-ventilated basement,
suggests that mechanical ventilation of a crawlspace might be an
effective mean to reducing radon levels in the home.  Radon
concentrations were measured in the main living area of
unoccupied energy-efficient residences.  The houses were of
woodframe construction with crawlspaces, a common construction
technology in many parts of North America.  Radon levels were
measured under three conditions in the test homes:  (a) crawlspace
completely enclosed, thereby preventing ventilation;  (b)
crawlspace naturally ventilated with square vents located on
opposite sides of the structure; and (c) mechanical ventilation
of the crawlspace using a 120 cfm supply fan, providing
approximately two air changes per hour to the crawlspace.
Results showed a 60% reduction in indoor radon levels in the test

-------
home with the mechanically ventilated crawlspace compared to the
test home with the un-ventilated crawlspace, and a 30% reduction
compared to the test home with the naturally ventilated
crawlspace.  The work described in this paper was not funded by
the U.S. Environmental Protection Agency and therefore the
contents do not necessarily reflect the views of the Agency and
no official endorsement should be inferred.

-------
                           INTRODUCTION
     Indoor radon levels are directly related to the ventilation
or fresh air exchange rate measured in air changes per hour
(ach).   However, comparing indoor radon levels for several
ventilated houses rarely results in a significant or meaningful
correlation because of wide variations in the rate of entry of
radon into the indoor environment, even within the same community
(1).  On the other hand, for a given rate of entry, indoor radon
levels are higher in energy conserving houses with low
ventilation rates below 0.5 ach than for conventional houses with
0.5-1.5 ach.  A study of 21 energy-efficient homes and 14
conventional homes in New York State found indoor radon levels in
energy-efficient homes to average 2.7 pCi/L on a year round basis
compared to 0.89 pCi/L for conventional homes.(2)  Nazaroff
estimated on the basis of measurements in energy-efficient houses
and crude estimates of average source levels for radon that, 36%
of English and 75% of American residences would have indoor radon
levels exceeding 0.02 WL if they were retrofitted to reduce
ventilation to 0.1 ach.(3)


     The increased likelihood of radon contamination of energy-
efficient dwellings and the impact on occupant health makes it
important to explore ways and means to decrease exposure to radon
in energy-efficient dwellings through appropriate building
design.


                          EXPERIMENT #1
     The effectiveness of direct outdoor ventilation of a
crawlspace or basement beneath an energy-efficient house in
reducing occupant exposure to radon contamination was evaluated
by comparing indoor radon levels in three immediately adjacent
residences during the time period October to December.


     House 1:  a new, energy-efficient home erected over an
unheated crawlspace with a vapor barrier consisting of 3/4 inch
plywood between the crawlspace and the house.  The crawlspace was
naturally cross-ventilated by vents located in the foundation
wall.


     House 2:  a 67-year old, well built, non-energy-efficient
house with a standard non-mechanically ventilated basement with
operable windows.  Furnace and central air supply ducts were on

-------
the basement level.  Walls and ceilings were insulated.  The
house was empty during the test period, with windows and doors
tightly closed.


     House 3:  a 50 year-old house similar in design to House 2
This house was occupied during the test period.


     All three houses were two-storeyed homes of woodframe
construction.
     Terradex Trac Etch Type SF Radon Detectors were placed in
the crawlspace or basements and living areas of Houses l and 3
for 88 days and of House 2 for 75 days.  In all locations, the
Trac Etch Detectors were pinned to interior vertical partition
walls a minimum of 9 feet away from exterior windows and midway
between the floor and ceiling.  The detectors were then collected
and pCi/L values determined.


RESULTS
     Table 1 summarizes the pCi/L levels for the three houses.
The low radon level in the ventilated crawlspace is as expected,
as constant ventilation prevents buildup of any contaminant.


     Despite being tightly enclosed, House 1 registered 0.82
pCi/L in the living area, approximately half as much as House 3,
which was a standard, non-mechanically ventilated structure, even
though House 3 was ventilated to the comfort of the occupants.
House 2, which was not tightened for purposes of energy
conservation but had been unoccupied with windows and doors
constantly closed, registered 2.04 pCi/L in the living area.


     The effectiveness of ventilated crawlspaces plus a vapor
barrier between the crawlspace and house can be evaluated by
comparing Houses 1 and 2.  While neither structure had been
occupied during the test period, each differs to some extent in
its rate of air infiltration.  There was little, if any,
ventilation in House 1, which was a new, specially tightly
constructed structure.  However, there was some air exchange in
House 2, which was a 67-year old structure that had not been
built to be airtight.  Thus, an estimate of effect of ventilated
crawlspace and bottom vapor barriers errs toward the conservative
side.

-------
     The effect of ventilation on radon levels emerges from a
comparison of Houses  2 and  3.  For similar houses, normal
ventilation associated with occupancy seems to reduce pCi/L by
approximately 25%.  If the  same reduction due to normal
ventilation is assumed to apply to House 1, then radon levels in
that house would be approximately 0.61 pCi/L when the ventilation
system in that house  was turned on.


     Assuming the values measured in Houses l, 2, and 3 are
typical, one can estimate that the use of a ventilated crawlspace
and a bottom-side vapor barrier can reduce indoor radon levels by
approximately 65%.


     Table 1 also shows expected yearly WLMs for occupants under
conditions represented by Houses 1, 2, and 3.  Estimates make the
assumption that an employed occupant spends 13.4 hours indoors
(4,891 hours per year) and  a housekeeper spends 20.5 hours
indoors (7,482.5 hours per  year).(4)  One conclusion flowing from
such a comparison is  that exposure to radon by housekeepers may
approach levels deemed excessive by EPA or Canadian authorities
if they should: (a) live in geographic regions with high
radiation background,  (b) live in tightly closed houses that (c)
lack ventilated crawlspace  and bottom-side vapor barriers.(2)


                          EXPERIMENT #2
     This experiment  investigated the effect of mechanically
ventilating crawlspaces as a means of reducing radon levels in
houses.  A new duplex (i.e. two adjoining but separate buildings)
was constructed adjacent to the houses for which we had already
obtained indoor radon measurements.  Each unit in the new duplex
structure was built on a crawlspace.  The crawlspaces of each
unit were separated by an insulated, woodframe firewall
incorporating an air  vapor barrier.  The two new units were
identical in all relevant parameters and size to the previously
constructed airtight  homes with ventilated crawlspaces.  It was
thus possible to compare indoor radon levels as a function of
different crawlspace  ventilation in an unoccupied, new, sealed
building, wherein:


     Condition 1:  The crawlspace is completely enclosed,
     preventing any ventilation;


     Condition 2:  The crawlspace is naturally ventilated; and

-------
     Condition 3:  The crawlspace is mechanically ventilated.


     Condition 1 was accomplished by blocking the supply and
exhaust vents in the crawlspace exterior walls.  Blocking off the
supply and exhaust vents was accomplished with duct tape and 6
mil polyethylene.  Mechanical ventilation for Condition 3 was
accomplished by installing a fan with a capacity of 120 cfm and
ducted to the 6-inch diameter supply vent.  A 6-inch diameter
exhaust vent was installed at the opposite end of the crawlspace.
Insofar as the crawlspace contained 3,600 cubic feet, the use of
this small fan ensured two air changes per hour.  Under Condition
2, a 2-foot-square supply vent and a 2-foot-square exhaust vent
were located on opposite sides of the crawlspace (see figure 1).


     During each measuring period, one unit had the supply and
exhaust vents in its crawlspace blocked for a period of 30 days,
while the other house had the ventilation ports opened and the
fan operating 24 hours a day.  At the end of each measuring
period, the ventilation conditions were reversed, and new
monitors installed.


     Again, radon levels were estimated during four measuring
periods using etch-type radon detectors.  Detectors were kept in
place for 30 days.  In all units during each measuring period one
recorder was installed on each floor of the unit and one recorder
was installed in the crawlspace.  In all locations, the detectors
were pinned each time in the same position to interior vertical
partition walls a minimum of 9 feet away from exterior windows
and midway between the floor and ceiling.


RESULTS
     Table 2 gives the average values observed for Conditions 1,
2, and 3.


     Under Condition 1 with the crawlspace un-ventilated, the
mean level of radon measured on the main floor was 1.94 pCi/L.
Under Condition 2 with the crawlspace naturally ventilated, the
mean radon level on the main floor was 0.82 pCi/L.  Under
Condition 3 with the crawlspace mechanically ventilated, the mean
radon level on the main floor was 0.57 pCi/L.  Compared to a non-
ventilated crawlspace, mechanically ventilating a crawlspace led
to an approximate reduction of 70% in indoor radon levels.
Mechanically ventilating the crawlspace with a small fan (no more
than 2 ach), yielded a 30% reduction on radon levels over what
would be achieved by having a naturally ventilated crawlspace.

-------
     Table 2 also shows expected yearly WLMs for occupants under
Conditions 1, 2, and 3.  Estimates were based on the assumption
that an unemployed occupant, such as a homemaker or a very small
child, spends 20.5 hours per day indoors (or 7,482.5 hours per
year).(4)


     In conclusion, such a comparison shows that exposure to
radon by homemakers or small children may approach levels
declared excessive by EPA or Canadian authorities, if they should
(a) live in geographic regions with high radiation background;
and  (b) live in sealed houses without ventilated crawlspaces.
Homemakers or small children are least exposed in houses where
ventilation of the crawlspace is enhanced by mechanical means.
(Retired persons and invalids would be exposed at a similar rate
as homemakers, but exposure would have relatively less severe
consequences for older than for younger persons.)


     While our findings are based on a small sample of
observations in Vancouver and on relatively short measuring
periods,  they are in line with expectations based on the effects
of mechanical ventilation and, to a limited extent, on the
findings of others.  Indoor radon levels have been found to be
higher in energy conserving houses with low ventilation rates,
below 0.5 ach, than for conventional houses with 0.5 - 1.5 ach.
A study of 21 energy-efficient homes and 14 conventional homes in
New York State found indoor radon levels in energy-efficient
homes to average 2.7 pCi/L on a year round basis, compared to
0.89 pCi/L for conventional homes.(2)  Insofar as one major entry
of radon into a structure is through gas emanating from the soil
beneath the structure, an air/vapor barrier between crawlspaces
and living areas, combined with a mechanically ventilated
crawlspace, should ensure minimal entry of radon gas into a
residence.  Extra care in placement of pipes and other direct
connections from the soil into the building is of additional
importance.


     Insofar as a link between radon exposure and lung cancer
seems to be well established, inclusion of mechanical ventilation
in the crawlspaces might be highly recommended by construction
standards organizations.
    In light of the very low radon levels measured, longer
measuring periods than 30 days would have been desirable.
However, the arrangements that had been made to keep the houses
unoccupied permitted measurements only over a 120-day period.

-------
                         ACKNOWLEDGMENTS
     This study was made possible, in part, through the Canadian
governmental policy of granting special research tax credits to
research-oriented Canadian corporations.  Buildings were
financed, constructed by, and all research supported by TDS
Limited.  David Mclntyre designed the buildings and supervised
their construction.  Elia Sterling served as research architect.
Part of the study was supported by a special grant from the
Council for Tobacco Research.

-------
                           REFERENCES
1.  Nero A.V.,  Boegel M.L., Hollowell C.D.,  Intersoll J.G.,
    Nazaroff W.W.  Radon concentrations and infiltration rates
    measured in conventional and energy-efficient houses.  Health
    Phy.. 45:401, 1983.

2.  Fleischer R.L., Turner L.G.  Indoor radon measurements in the
    New York capital district.  Health Phv..  46:999, 1984.

3.  Nazaroff W.W., Boegel M.L., Hollowell C.D., Roseme G.D.  The
    use of mechanical ventilation with heat recovery for
    controlling radon and radon-daughter concentrations in
    houses.  Atmos. Envir.. 15:263, 1981.

4.  Szalai A. The use of time.  The Hague: Mouton, 1972.

-------
    TABLE 1.  RESULTS OF RADON MEASUREMENTS IN THREE HOUSES IN VANCOUVER
Yearly WLM*

Unoccupied sealed house
with ventilated crawlspace.
Unoccupied house with
partially occupied basement.
Occupied house with
basement.
pCi/L
Crawlspace 0.07
1st Floor 0.82
Basement 0.38
1st Floor 2.04
Basement 0.82
1st Floor 1.57
WL*'
0.0041
0.0102
0.0079
Employed *
0.029
0.073
0.056
Unemployed*1
0.045
0.112
0.086
'   1 WLM = 680 hours of exposure to 1 WL.
   1 WL = 200 pCi/L, assuming an equilibrium between radon and its daughters of 0.5. U.S. EPA
   remedial action if WL > 0.015.
*  Estimated average of 13.4 hours per day spent indoors at home (4,891 hours per year).
**  Estimated average of 20.5 hours per day spent indoors at home (7,482.5 hours per year).

-------
TABLE 2.  COMPARISON OF EXPOSURE TO RADON IN MAIN
     LIVING AREAS OF HOUSES WITH CRAWLSPACES
 (1) WITHOUT VENTILATION; (2) NATURALLY VENTILATED;
AND (3) ENHANCED VENTILATION VIA MECHANICAL MEANS




Average radon level.
WL
WLM for employed occupant
(based on 4,891 hours per year).
WLM for unemployed occupant
(based on 7,482 hours per year).

Crawlspace
Sealed
1.94
0.0097
0.069

0.112

Crawlspace
Ventilated
Naturally
0.82
0.004
0.029

0.043

Crawlspace
Ventilated
With Fan
0.57
0.0028
0.020

0.031


-------
SUPPLY
 VENT
                                      EXHAUST VENT
                  Unit A
FAN
     SUPPLY
     VENT
 FAN
EXHAUST VENT
                  UnitB
        Figure 1.   Crawlspace ventilation in experimental homes.

-------
               Session VIII
Radon Occurrence in the Natural Environment

-------
                                                                    VIII-1
          INDOOR RADON AND THE RADON POTENTIAL OF SOILS
               by: D.J. Steck and M. J. Bergmann
                   Department of Physics
                   St. John's University
                   Collegeville, MN 56321
                             ABSTRACT
     Subsurface soil samples, collected adjacent to 110 Upper
Midwest houses, were measured to determine radon emanation,
diffusivity, and permeability. The soilborne radon
characteristics of the samples of glacial till that covers most
of the region varied by a factor of ten or more.  A Monte Carlo
model was used to calculate indoor radon concentrations in
idealized basements sited in a homogeneous medium whose radon
properties matched the soil sample properties. Measured basement
radon concentrations correlate with the model predictions. The
correlation is more useful to assess clusters of homes than
individual dwellings.

-------
                           INTRODUCTION

     Direct airborne radon measurements can be a time-consuming
and costly way to survey the potential radon problems in a large
sample of homes. At sites were no home exists, no direct indoor
measurements are possible. Thus, it would be useful to find an
alternative, indirect technique for predicting indoor radon
concentrations in a home, a cluster of homes, or a site for a
future home. Indoor radon concentrations can vary significantly
from place to place. The Upper Midwest region of the United
States contains broad areas of modestly elevated indoor radon
with widely scattered small towns. Within this region, localized
areas of high radon concentrations (hot spots), often the size of
neighborhoods, have been found  (1). Despite widespread publicity,
few homes have been measured for radon. Thus, radon
concentrations are poorly known in most populated areas outside
the major cities. In such an environment, an inexpensive,
effective radon assessment technique would be quite useful to
public health officials.

     Since soils have been identified as a primary source for
indoor radon  (2), it seems reasonable to base a radon potential
assessment protocol on soil characteristics. This approach is not
new. In fact, soil-based assessment protocols have been applied
to a limited number of dwellings with varying degrees of success
(3). Since many of these methods include intensive, in situ,
measurements they are not cost- competitive with direct airborne
surveys (3).  In the hopes of finding an efficient screening
method for homes with elevated radon, we have investigated the
utility of simple, inexpensive soil sample tests that are used
with homeowner-supplied information to predict annual-average
radon concentrations in homes.
                             METHODS

INDOOR RADON MEASUREMENTS

     More than 230 homes, located in Minnesota, Wisconsin, and
Michigan, were monitored for indoor radon concentrations and
sampled for radon source material (soil and water). Airborne
radon concentrations in basements were measured for a period of 1
yr.or longer with alpha-track detectors. Typical uncertainty in
these data is 15%.  Waterborne radon concentrations were measured
with a liquid scintillation technique. Waterborne concentrations
were too low  (median value 9 kBq m"3 ) to be considered a
significant radon source. Other experimental details have been
reported previously  (4).

-------
SOIL SAMPLES

     Participating homeowners were asked to collect soil samples
from a depth greater than 30 cm at four separate locations within
1 m of their basement walls. We selected 110 representative soil
samples for analysis.  In the lab, the samples were mixed,
categorized into types (sand, silt, clay) by microscopic
examination of grain sizes,  and their porosities were measured.
In order to replicate in situ soil characteristics, water was
added until the concentration equaled the average of the wilting
and saturated concentrations ( 2% to 20% by weight). Local
building practices call for soils near homes to be compacted to
90% Proctor density  (5).  We packed the moist samples into PVC
columns (5 cm diameter, x 6 cm) and compacted them with a
instrument patterned after ASTM Standard 78(6).

EMANATION AND 222Rn CONCENTRATION

     The soil columns were sealed and allowed to age. Then the
pore gas was extracted into a scintillation cell. The gas in the
cells was allowed to age for 3 hours, counted, and recounted
after 12 hours to help eliminate any contimantion from thoron
daughters. Results from this system were reproducible to within
10% with a total uncertainty of approximately 15%.

     Gamma spectroscopy was used to measure the total soilborne
222Rn  concentration to an accuracy of 10%.  (4)

DIFFUSION

     The effective, textural diffusivity was measured using a
scintillation chamber patterned after Cohen, et al.  (7). Soil
columns served as leakage pathways for a continuously-monitored
scintillation chamber filled with a high concentration of radon.
The time dependence of the count rate over a 10 to 16 hour period
was fit to extract the diffusion constant. This procedure yielded
effective diffusion lengths to an accuracy of 8%.

PERMEABILITY

     We developed a computer-interfaced permeameter capable of
measuring permeabilities in the range from 10"10 m2  to 10"15m2. In
this device, pressure gradients and flow rates, close to expected
field values, were measured by electronic sensors. Usually, five
separate sets of pressure - flow data were collected and then  fit
to extract the permeability for each soil sample. Although the
accuracy of the fit is generally better than 10%, a comparison
between the experimental and theoretical permeabilities for two
columns constructed with glass beads  ( 1 mm and 0.1 mm), suggest
that the permeabilities are accurate to with a factor of two.

-------
Since we have observed order of magnitude variation in a single
sample's permeability as a function of moisture and compaction, a
factor of two in uncertainty seems acceptable.

TRANSPORT MODELS

     Ninety percent of the homes that we surveyed have basements
that normally extend 2 meters or more into the ground. We have
limited our analysis to basement-type homes with no known nearby
bedrock. In order to find a tractable solution to the radon
transport problem, we have made a number of simplifying
assumptions, namely that:

•  soil is the primary source of indoor radon;
•  the soil is homogeneous;
•  the indoor and outdoor airborne radon concentrations are low
  compared to the pore gas concentrations;
•  the effects of soil moisture are represented by the transport
  characteristics of moist soil samples;
•  the radon entry can be separated into a component through the
  basement walls and a component through the floor;
•  radon enters the basement walls via diffusion and
  pressure-driven flow;
•  radon enters the basement floor via diffusion;
•  the diffusive component is governed by Pick's Law and the
  pressure-driven component is governed by Darcy's Law (1)
•  the basement pressure is uniform.

     After analyzing our data with two previously published
models (8,9), we developed a Monte Carlo model simulation to see
if we could improve the correlation between the theoretical
predictions and the measured indoor radon concentrations. We used
a simple, two-dimensional geometry (see Figure 1), where the
solution to Laplace's equation yields a simple vector field for
the pressure-driven velocity, namely:

                                   a f\              ,..,
                                 ~T°              (D

where k is the permeability, P is the pressure difference between
the outside and the basement, eis the porosity, p. is the
viscosity of air, and r is the distance from the basement wall -
ground intercept. We model diffusion as a macroscopic random
walk. The bulk diffusion constant,  D, generates a step size, d,
as:

                    d=^6DAt                        (2)

     In order to assure that each test atom can take at least 30
steps before reaching the basement, we determine the time

-------
increment,  At, between each random step as the minimum of three
characteristic times of convection, diffusion and decay, e.g.
                    At  =
                         MIN'-f,
where T is the mean life of radon. The model is used to
approximate the probability, P(x,y), that a radon atom generated
in the soil at (x,y) will reach the basement. We define the
effective thickness, A, of soil that contributes to the radon
entry as:
                              P(x,y)dxdy
 4)
where F is the basement depth. Then the radon flux, F, can be
expressed as:
                  F = E ATw
where w is the width of  the basement.
                              p=o
(5)
 Figure 1. Monte Carlo  model geometry,  pressures,  and streamlines
          A  sample  trajectory is  shown.

-------
 Two dimensionless parameters, p  and 5, are used to characterize
the transport distances  (in units of basement depth: T) due to
pressure-driven and diffusive flow respectively:
     We populated the parameter space with 63 data points
distributed log - uniformly in the region bounded by 0.001< p<
30 and 0.001 < 8 < 1. We interpolated values of A between these
solution points. In order to calculate the probability for each
case, we placed a large (>30,000) number of particles in a
truncated region of space next to the basement. After the initial
solution has been calculated,  the region is extended, a new
solution calculated and the process repeated until the
probability changed by less than 3%. Each solution required 100
CPU minutes a Silicon Graphics 4D25 workstation. We compared our
solutions at extremum points (p,8 = 0) with the Tanner model  (9) .

  House characteristics, such as basement pressure differential,
volume, air exchange rate, must be estimated in order to generate
the model predictions. Although we have collected and coded some
information from homeowners that may be useful in estimating
these parameters,  we chose to include only the depth of the
basement in the first analysis. The standard basement depth used
by the model was 2.5 m. For the other input parameters, we
selected a "standard" environment of 8 Pa basement
depressurization,  basement dimensions of 10 x 15 x 2 . 5 m, and an
air exchange rate of 0.3 ach.  We calculated the diffusive flux
through the basement floor using a simple one dimensional
solution (2)  and recently - published slab diffusion coefficients
(10).

                             RESULTS

     Table 1 summarizes the key statistics of the basement radon
and soil source characteristics. Figures 2 through 4 show the
complete distributions.

     The basement radon concentrations, shown in Figure 2,
indicate that elevated indoor radon concentrations are common in
the Upper Midwest . Although the observed soil source parameters
span a broad range, they generally are somewhat below the average
values that have been previously reported (2) . We discovered that
classifying soil types by grain size was not a significant
distinction vis a vis the source or transport characteristics.
For example,  even though the mean diffusion length and
permeability for clays were smaller than sands, the range of
variation within the clays overlapped the range of variation of

-------
      Table 1. Airborne and  soilborne radon  characteristics.
Measure
Basement Rn
Emanation
Diffusion Length
Permeability
Effective Thickness*
Unit
Bq nf 3
kBq m°
cm
nr
cm
Mean
150r
II4
120
10-12F
140
SD
2*
6
60
10r
60
Range'
45
3
45
10
20
to
to 3
to
600
2
155
~15 to 10'"
to
200
* 95% cumulative fraction interval
+ Log normal distributions; geometric mean and standard deviations listed.
± This corresponds to an emanation fraction of 0.4 and an equilibrium radon
  concentration in the available pore spaces of approximately
  40 kBq m° (lOOOpCi/L)
s Monte Carlo model generated parameter

sand parameters. This may reflect the  wide variety of  grain sizes
that occur in a typical  sample of our  glacial till. We note that
while the permeability  spans four orders of magnitude,  the rest
of the variables (including the basement Rn and the effective
source thickness) only  span one order  of magnitude or  less.
     .99
  z
  o
  M
  H
  CJ
  HI
  >
  H
  _J
  D
  X

  O
             -1	1—r
                                             1—i—r
  i

  r
i fe
     .01 b
                   i  -I—L_L-
                        100

                RADON CONCENTRATION  (Bq m~-3)
                                               _
                                               4


                                               5
                                                     •
                                               u

                                               -i
                                            0_0_3
                                              1000
Figure  2  .Basement radon concentration distribution.

-------
   ETT


.99 p-
               r r-rr
•n-i-T-i-rrr;T3
             J
 z
 o
 a
 <
 a:
 u.
  .
 D
 O
    •sp
 -1*
       E
   .01
      i  i  I  i i  i  i i  i  i i  i  I  i i  i i  I  i i

          5        10       15      20

              EMANATION  (kBq m~-3)

                                             «J
                                              25
Figure 3. Soil sample emanation distribution.



2
O
H
0
tr
u.
1 1 i 1
.99 f-
\~
u
P
.9 r-
1
L
L
 LU
 D
 O
                       T—r~i—r~r
                                          I—[-
    ,01
                                                 _
-
r
20
.*
•
l i l 1 l 1 i i
40 60 80 100

12O 140 160
                  DIFFUSION LENGTH (cm)
Figure 4.  Soil sample diffusion length distribution,

-------
   .99
z
o
o
<
GC
u_
HI
>
l-l
U
    . 1
   01
      IT
      l.E-15  l.E-14  l.E-13  l.E-12  l.E-11  l.E-10
                     PERMEABILITY  ( nT 2  )


Figure 5  Soil sample permeability distribution
 Previously published models failed to produce a significant
correlation between the predicted and observed radon
concentrations. Our Monte Carlo model produced a better
correlation including an order of magnitude agreement between the
observed and predicted indoor radon concentrations. While the
probability that this fit could result from a random set of data
is small (  p<0.01), the percent of the variation explained by the
model ( 6%; R2= 0.06)  is  also small.  The dimensionless  parameters
for pressure (p) and  diffusion  (5) calculated  from  the  soil
samples are shown in Figure 6. Figure 7 illustrates the Monte
Carlo model-generated dependence of the effective source
thickness   (A) on p and 5.  The effective thickness distribution
of the soil samples is shown in Figure 7.  A comparison between
'the model  predictions and measured radon in all 110 homes is
shown in Figure 8.

                            DISCUSSION

     The radon environment in the Upper Midwest is rich and
complex. Although median basement concentrations are 150 Bq m~3,

-------
     10
  2    1
 PU
  0)
 OH  .1
    .01
        -*.  .  r  . r
                                                  lit
                          Diffusion Parameter (<5)
Figure 6  Soil sample pressure  (p ) and diffusion  (5) parameters
      0.30
      -0.30
      (P)
      -x.so
      -2.50
                           tog  (A)
                                 -0.2
               -2.50    -Z.UO    -T.5O    -l.DO    -O.SO    0.00

                                (6)
Figure 7 Effective thickness  (A)contour plots.

-------
concentrations in excess of 4000 Bq rrf3 have been observed. Yet,
the glacial till surrounding the houses generally is not high in
radon content or transport properties. However, the climate
encourages construction of basements, heating of homes from
October to May, and cooling of homes from June to September.

     Thus, soilborne radon can enter through a large, relatively
permeable shell (most basement walls are cement block or
fieldstone) which is depressurized relative to the soil for a
substantial fraction of the time. The region receives ample
precipitation  ( approximately 700 mm annually) that sustains
forests in the east and agricultural prairies in the west. Since
soil moisture plays a pivotal role in radon availability  (11),
its seasonal and cyclic variation can dramatically influence
indoor radon. In particular, the role of snow and frost (which
often penetrates several meters into the ground in some years)  is
not well documented. In addition, the static water table  is often
near the surface and varies cyclically.  These factors make
predicting indoor radon concentrations a difficult task in our
region. For example, approximately half of the indoor basement
radon measurements used in this study were made during years of a
severe drought in the north central part of the region. Yet, the
moisture content of the sample material was selected to
approximate historical average soil conditions.

     Although our model results do indicate a good correlation
between the measured and predicted indoor radon concentrations,
numerous factors associated with the particular environment of
each home complicate any attempts to simply model the radon entry
and retention. Below is a partial list of those factors,  not
included in our model, that could be expected to affect the
indoor radon in a particular basement:

•  source variability
  * soil inhomogeneity
  » structural versus textural transport properties
  * vertical and horizontal moisture distributions
  * near surface water tables
  * compactness of soil adjacent to basement walls

•  structural variability
  * transport properties of subsurface structure/holes
  * depressurization: HVAC systems
  » basement - superstructure mixing
  » local meteorological conditions; wind, precipitation
  * superstructure infiltration / ventilation rate

     We are presently trying to incorporate information about the
individual homes into our model in order to improve  its
performance. Unfortunately, a preliminary analysis of the utility

-------
    .99
 cc
 L.


 HI
 >
 _J

 D
 U
     . 1
    .01
      ;!  I  I  1  [  I  I  I  I
                             I  I  I  I  I I  I  I  I  I  I  I  I
      : l  I  i  i  I  i  I  i  i  I
                                 i  i I  i  i  i  i  I  i  i
                50       100       150       200

                  EFFECTIVE THICKNESS  (  cm )
                                                     250
Figure 8 Soil sample effective thickness distribution.
cn

*E



CD
c
cr

a
tu
    800
    600
    400
  cc
  UJ

  In 200
  o
      0 1=
                 100        20O        300

                   PREDICTED  Rn (Bq/m~3)
                                                400
Figure 9 Model predictions compared to  observed  radon

         concentrations. The  solid line is the regression  fit

-------
of the information supplied by homeowners showed no significant
correlations between any categorical responses and indoor radon.
However,  we may be able to improve the model by adjusting a
combination of parameters, e.g. the basement pressure,  entry
area,  ventilation rates, and shell transport characteristics.

     Since a number of these factors might not vary over a small
geographic locale, we tested the model for 20 localized clusters
of homes  by calculating the average predicted indoor radon
concentration and comparing it with the average measured radon
concentration for each cluster. An acceptable cluster was defined
as more than 3 houses located within a 5 km radius.  The
comparison is shown in Figure 10. The model is significantly more
successful for clusters (R2 =  0.34  )  than individual  houses.  In
fact,  if  we define the coefficient of variation (COV) of a
prediction as the average deviation of the data from the fit,
expressed as a percentage, then the COV for our model is 100%  for
individual homes but only 35% for clusters. These COV's compare
favorably with COV's of screening surveys based on short-term
airborne  measurements using a single detector in each house  (1).
     400
   T 300
   E
   \
   IT
   CQ
     200
Q
111
>
QC
HI
S
o
     100
                                     1
                      100
                    PREDICTED Rn
                                  200
                               (Bq/nrT3)
300
 Figure 10 Average radon concentrations for clusters of houses as
          predicted by the model compared to the measured
          concentrations.

-------
                           CONCLUSIONS

     Indoor radon concentrations in basements can be predicted
from soilborne radon parameters using a simple model. At present,
the model is most useful for predicting average concentrations in
localized clusters of houses. Work is underway to investigate
whether the utility of the model can be improved through the
modelling of additional transport, entry, and retention factors.
                        AC KNOWLEDGEMENTS
     We would like to thank Dr. Carol Meger, Norm Tyrell, Fred
Monette, Adam Whitten, Dr. Phil Byrne, and Greg Taft for
assistance with various parts of the instrument development, data
collection, and analysis. Special thanks is given to Dr. Tom
Kirkman who provided valuable suggestions and criticisms
throughout the project.

     This work was supported, in part, by a grant from the
Minnesota Private College Research Foundation with funds provided
by the Blandin Foundation of Grand Rapids, Minnesota.

     The work described in this paper was not funded by the U.S.
Environmental Protection Agency and therefore the contents do not
necessarily reflect the views of the Agency and no official
endorsement should be inferred.
                            REFERENCES
1. Steck, D.J.  Spatial and temporal indoor radon variations.
   Health Physics . 62: 351, 1992

2. Nazzaroff, W.W., Moed, B.,  Sextro, R.G. Soil as a source of
   indoor radon: generation, migration, and entry, in: W.W.
    Nazaroff and A.V. Nero  (eds.) Radon and its Decay Products in
   Indoor Air. Wiley, Interscience New York, N.Y.,  1988

3. Scott,A.G. Site characterization  for radon supply potential: a
   progress review. Health Physics 62: 422, 1992

-------
4.  Steck,  D.J. Variations of radon sources and indoor radon along
   the southwestern edge of the Canadian Shield. Environment
   International 15: 271, 1989

5.  Gary Traut, private communication, Braun Engineering, St.
    Cloud,  MN

6.  American Standard Testing Methods for moisture - density
   relation of soil and soil-bearing aggregate D 698, 1978

7.  Cohen,  B.L., Rakowski, J., and Nason, R. A simple compact
   apparatus for measuring diffusion properties of radon through
   soils and other materials. Health Physics 50: 133, 1986

8.  Nazaroff, W.W. and Sextro, R.G. Technique for measuring indoor
  222Rn  source potential  of  soils.  Environmental  Science and
   Technology 23: 157, 1989

9.  Tanner,  A.B. A tentative protocol for measurement of radon
   availability from the ground. Radiation Protection Dosimetry
   24: 79,  1988.

10.Nielsen, K. K. and Rogers, V.C. Radon entry into dwellings
   through concrete floors. In: Proceedings of The 1991
   International symposium on Radon and Radon Reduction
   Technology. U.S.E.P.A. Philadelphia, Pa. 1991

11.Rogers,  V.C. and Nielsen K. K. Correlations for predicting air
   permeabilities and 222Rn diffusion coefficients of soils.
   Health Physics 61: 225, 1991

-------
NATURE AND EXTENT OF A 226-RA ANOMALY IN THE WESTERN SWISS JURA MOUNTAINS
                    by :   Heinz Surbeck
                          Federal Office of Public Health
                          Rad.Surv.Sec. (SUeR)
                          c/o Physics Institute,  University
                          Perolles
                          CH-1700 Fribourg, Switzerland
                                 ABSTRACT
     High  indoor  radon concentrations (up to 5 kBq/m3 in  living  rooms)
are  abundant in the Western Swiss Jura Mountains,   a Karst  terrain.  The
search  for the radon sources has led to the discovery of very high  226Ra
activities (up to 850 Bq/kg) in soils covering the low activity  limestone
bedrock.

     Nature  and extent of this radium anomaly are presented and  possible
origins are discussed.

     A striking feature of this anomaly is a strong disequilibrium in  the
23aU  series.  The 236Ra is in equilibrium with its parent  nuclide  230Th
but the precursors of the 230Th are present at far lower levels only.

     Many  soils  of the Western Jura Mountains show  a  loess-like  grain
size distribution,  pointing to an aeolian origin.   The dust has  probably
been  deposited at the end of the Hurra glacial period.  Leaching  of  this
very  fine  material may have led to the observed  disequilibrium  in  the
2aaU series.

-------
                               INTRODUCTION
    Those of you mainly interested in radon remedial actions may  consider
the  reserach presented in this paper as sheer luxury.  But  sometimes  it
pays  off to invest in a better understanding of the nature and extent  of
a radon source prior to mitigation.  If,  as an example,  it can be  shown
that there is a widespread radium anomaly of natural origin,  you will not
waste your time and money to clean up tiny artificial sources.

    High  indoor radon concentrations are frequently found in the  Western
Swiss Jura Mountains (1,2,3).  These mountains consist of at least several
hundred  meters  of Jurassic and partly Cretaceous  limestone.  They  have
been folded up during the late phase of the alpine orogeny.  The peaks are
at  about 1600 m above sea level.  Karst features like sinkholes or  caves
are  abundant.  The thin soil cover on the mountain heights  (above  about
900  m) is younger than the last glacial period (Wurm) and contains no  or
only  very few alpine material.  Windblown (aeolian) material  of  unknown
origin is present in many of the soils (4).

    Despite  the low 226Ra concentration in the limestone bedrock and  the
lack  of  alpine  material with possibly  higher  activities,  high  22SRa
activities  have  been found in many soil samples from  the  Western  Jura
Mountains.

    It  was tempting to blame the watch industry for the  enhanced  radium
concentrations;  an important part of the Swiss watch industry is situated
in this region.  Until about 1965 this industry processed large quantities
of  226Ra.  Radium activated luminous paint was applied onto  watch  dials
and hands.

    Radium contaminations  are clearly present by they do not explain  the
high  activities  found in soils sampled at the top of  the  mountains.  A
strong  evidence for the natural origin is also the presence  of  enhanced
230Th activities in the high-radium soil samples.  230Th is the  precursor
of  the 226Ra in the Z3BU series.  230Th has not been found in samples  of
luminous paint used by the watch industry and therefore can be used as  an
indicator for the natural origin of the 226Ra (5,6).

    Whereas  it seems to be clear that the watch industry can not be  made
responsible  for the widespread radium anomaly in the Western  Swiss  Jura
Mountains, the origin of this anomaly reamains unclear.

-------
    In  this  paper I present what is already known about the  extent  and
the  nature  of this anomaly,  how this information has  been  gained   by
gamma  spectrometry  and  how  it may help to reveal  the  origin  of  the
enhanced 226Ra activities.

    The list below summarizes important characteristics of the anomaly.

    (1) The highest 32SRa activities found is nearly 900 Bq/kg dry weight.

    (2) The 22SRa is in equilibrium (within experimental error) with it's
        natural precursor 230Th.

    (3) There is a strong disequilibrium in the "au series. The first
        member with enhanced activity is the 230Th.

    (4) There is an extremely large radon emanation from the soils with
        the enhanced aasRa activity. At least half of the "3Rn produced
        can escape to the pore space and is thus available for transport.

    (5) The specific 2a*Ra activity as a function of the grain size
        peaks at intermediate grain sizes (near 200 um). This is in clear
        contrast to the activity vs. grain size pattern found in "normal"
        soils from the Swiss Plateau. In "normal" soils the specific
        activity decreases monotonically  with increasing grain size.

    (6) Clearly enhanced activities (>150 Bq/kg dry weight) have so far
        only been found at altitudes higher than 900 m above sea level.

    (7) The anomaly extends over a strip of at least 100 km lenght and
        some km width. There is a tendency to higher values towards the
        south-west.

    Figure 1 helps to locate the sites mentioned in this paper.

-------
                Western Swiss
                Jura Mountains
            Chasseral
       Tete de Ran
    Chasseron
   Ml. Tendre
Marchairuz^
          *V
La
     Figure 1 . Swiss map with (he locations mentioned in the text and in the other
              Figures. The dashed lines roughly represent the axis oF the
              respective geographic unit.

-------
                          MATERIALS AND METHODS
SAMPLING
    In  general soil samples are taken to a depth of 10 cm and  are  split
into  two  subsamples of 0 to 5 cm and 5 to 10  cm.  The  standard  sample
surface is 10 cm x 10 cm.  Some of the samples have been taken with a 7 cm
diameter soil auger down to a depth of 50 cm.   At most places in the  Jura
Mountains the soil (A and B horizon) has a thickness of hardly 30 cm,  but
in small depressions the soil can be as deep as 2 m.
SAMPLE PREPARATION
    The  soils  are dried under a heating lamp.   The temperature  is  kept
below  60 °C to avoid destruction of the clay  mineral  structure.  During
drying  the soils are carefully stirred by hand to avoid the formation  of
large hard blocks.  After drying,  a part of the grain size fraction <  32
urn is separated by dry sieving.

    The  separation  of  the fractions larger than 32 urn is  done  by  wet
sieving  in two step.  About 500 g can be treated in one  batch.  For  the
first  step,  the coarse separation,  a vibrating sieve machine  is  used.
About  50  liters of water flow through the soil  during  this  step.  The
fractions are then carefully washed by hand.  A sharp water jet is used to
dissociate  the  agglomerates that formed during  drying.  Gentle  rubbing
with  the fingertips accelerates this process.  The clay fraction  staying
suspended and the floating organic material are discarded.  The  separated
fractions are then dried under a heating lamp.  During the second  sieving
step  there  are again about 50 liters of water flowing through  the  soil
sample  of  about  500  g.  The quantitiy of water  used  for  both  steps
corresponds  to  about the quantity of rainwater that has  passed  through
the  soil  during the last 10 years.  The water used for  wet  sieving  is
ordinary tap water,  a non-chlorinated groundwater.  It shows,  apart from
about 3 Bq/liter of 222Rn, no measurable radioactivity.

    Repeated  washing with large quantities of water showed that at  least
for these soils from the Jura Mountains there is no measurable leaching by
wet sieving.

-------
    Just  before measurement,  each grain size fraction is saturated  with
water and dried under a heating lamp within about 1 hour.  This  procedure
removes any radon that has escaped to the pore space.  The radon remaining
in  the sample after this treatment is the part that could not  escape  to
the pore space.
MEASUREMENT
    If  there  is  sufficient material,  40 cm3 of the  dried  sample  are
packed  into a flat polystyrene box.  In this case the gamma  spectrum  is
taken  with  a 25% relative  efficiency  HP-Ge-detector.  Smaller  samples
(grain  size  fractions)  are measured with  a  well-type  HP-Ge-detector.
Construction  and shielding materials of this detector have been  selected
for  low activity.  This makes it possible to reliably  determine  natural
and   artificial  radionuclides  in  samples  of  only  some   grams   and
measurement times of 2 days.

    The  calibration  of both detectors is based on measurements  of  IAEA
reference  samples (SOIL-6,  SDN-2,  SDA-1,  RG-series).  The  calibration
points  for  the  25%  detector  are  fitted  with  a  single   5th-degree
polynomial (log(effficiency) vs.  log(energy)) over the energy range  from
20 keV to 2 MeV.  The well-type detector is calibrated for each gamma line
individually.

    The  gamma spectra are evaluated with a interactive program  developed
by Intertechnique (INTERGAMMA).  Mainly below 100 keV the operator has  to
improve  the marking of the peaks.  This is very important for a  reliable
determination of the 230Th from it's weak 67.6 keV peak.

    22
-------
    Therefore  the  fraction of the radon produced that can  escape  under
the most favorable conditions to the pore space and is thus available  for
transport  can  be  determined  from  the  difference  between  the  22SRa
activity  and  the activities of the radon daughters in  a  freshly  dried
sample.
                          RESULTS AND DISCUSSION
    Figure 2 shows the disequilibrium in the 23aU series for some  samples
from  the  Harchairuz-Nt.Tendre region.  Whereas 230Th and  22SRa  are  in
equilibrium  (within measuring errors) the 23*Th activities are at only  a
fraction  of  the 32*Ra levels.  The equilibrium between  226Ra  and  it's
precursor 230Th is a strong evidence for the natural origin of the  226Ra.
In  samples of luminous paint used by the watch industry I have  found  no
230Th  (less  than  1/10  of  the  226Ra  activity) (5,6).

    According  to alpha spectrometry measurements carried out on  some  of
my  sample by a different laboratory,  23
-------
1200
s 1000
^m
w
"3
* 800
•o
I 600
o-
ffi
jj 400 •
3
=
« 200

a)
a
0
2
•D
$ 5
I ™
O 5
9
0 g
ffi
o
I I I I I
1200 •
1000 -
-
800 -
-
600
•
400 -


200 -
0.

<


b)
>
T



o

o


o
r n
I I I
              .*  o to  t  o
              z  s s  ™  z
              i  A 


4



{

.1 *


0

0 I I I I I
i 0 to t o
M N «^ t}I CSI
f £ a) A A
£ £ GC Q. 0.

a
d
o
•_
T3
m
**
3
u
 Ml. Tendre #0 , 0-5 cm depth
         d> Mt.Tendre *0 , 5-10 cm depth

-------
this fallout.

    The  236Ra  activity  concentration as a function of  the  grain  size
(Fig.3)  peaks  at very small and at intermediate  grain  sizes.  This  is
different  from  the behaviour in "normal" soils from the  Swiss  Plateau,
where  the  activity decreases monotonically with  increasing  grain  size
(3,9).

   .A hint to understand this strange pattern found in the soils from  the
Jura  Mountains may come from the activity vs.  grain size for  137Cs  and
13*Cs.  Both  cesium  were  brought  in by the  rainwater  and  have  been
adsorbed  by the soil from solution.  In soils from the Swiss Plateau  the
13TCs  activity  vs.  grain  size  shows that  the  cation  adsorption  or
exchange  capacity (c.e.c.) is highest for the smallest grains  (<32  ym).
This can be attributed to the presence of clay minerals (< 2 jim) in the   <
32  urn fraction.  The 137Cs activity pattern in the Jura  Mountains  soils
shows  the presence of a high-c.e.c.  substance not only in  the  smallest
fraction but also in intermediate grain sizes.  The similarity between the
13'rCs and the 5*26Ra acitivity vs.  grain size patterns points to a  common
origin.  aasRa  may therefore have been adsorbed from solution  too.  This
would also explain why such a large part of the radon produced can  escape
to the pore space.  The nature of the high-c.e.c.  substance in the  large
grains remains unclear.  That this exchange capacity is still present  can
be seen .from the -l3*Cs activities vs.  grain size. 13'*Cs has its origin in
the fallout after the 1986 Chernobyl accident.

    Mainly  the  soils  from  Marchairuz  contain  many  dark  grains   of
different  diameters.  With the skill and  the patience of a  gold  panner
grain  size  fractions larger than 64 urn of these soils can  be  separated
into  a lightweight dark and a heavier light part.  The  lightweight  part
consist of dark,  very hard grains. They are not an artefact of the drying
process,  they  are  already present in the fresh  soil.  The  light  part
consist  mainly  of  small  pieces of limestone  and  some  clear  calicte
crystals.   The   lower  part  of  figure  3  shows  that   the   activity
concentrations  are  far  higher  in the dark grains  than  in  the  light
grains.

    Dark,  nearly  indestructible grains in soils from the  (French)  Jura
Mountains  have  also been described by Gaiffe (10).  She has  found  that
they  are composed of organic material and clay minerals held together  by
a  cement  containing iron hydroxide.  Either these clay minerals  or  the
iron  hydroxide  may be the reason for the strong adsorption  capacity  of
the grains in my soil samples.

    Enhanced  a26Ra  concentrations  have so far only been  found  in  the
Western  Swiss  Jura  Mountains and only at altitudes higher  than  900   m
above  sea level  (11).  This altitude corresponds to about the  upper  ice
margin at the end of the last glacial period.  It is therefore tempting to
attribute  the radium anomaly to an aeolian deposition during the  retreat
of the glaciers,  with the Jura Mountains free of ice. According to Pochon

-------
1UUU •
O)
f 800 -
3
•a 600 •
ci
o 400
.&
I 200
o
n J 	 	 	 	 	 	

o .
MfMf
* t 0 0 o
• 0 • " * • 3
B •
0 0 0 0 0 0 0
"ill III
CM m in o o o o
n to CM in o o o
V i *- CM in o o
CM ' ' ' T ^
rn ro in C9
to CM in o o
«• CM o a
. in o
1
( 1000 -
§ 800 -
"3
3
* 600 •
^g
5
9 400 -
00
| 200
a
0.
f ^




: \
B
^ 0
1







|_

b^
N
M

c
2
d








250
f 200-
3
>
5 150
*
a 100
f 50
•3
a
0
t *
r f i
. A T j,
^ T * *

I I I I
CM n in o
n e CM in
v t *• N
CM ' '
ro w in
(O N
^™

I
/"
1 250
•£ 200
3
•o 150-
d
m 100
1 50 •
a
n .
A

I
I

*
f*
i*
*

i i
e o
o a
in o
i
0 '
in o
CM 0
in






J

t
I
*

I -
e E
e i
o *.
*? o
I l^
O 7
0 _
o .E
fc
^





                        M
                      ^ c
                                                            I      I
Figure 3 . Activities as a function of the grain size for the soil sample
         Harchairuz *91.2 .0-10 cm depth. Error bars are +/-2 sigma.
         O : Th-234 . • : Ra-226 . • : Pb-214 . D  : Pb-210
         X : ftc-228 . A : Cs-137 , A : Cs-134  x 10 . + : K-40 .
         The Cs-134 activity in the light grain part is below the
         detection limit of 4 Bq/kg dry weight.

-------
           Ra-226 activity
         I Bqflcg dry weight 1
      < Error approx. +/- 15 %
                                        anticlinal
                                        axis
                                        1450m
                                        above
                                        sea level
               10m
E
e
o
in
                a>
                                                         500 m
                                                           b>
                                                Tele de Ran
Marchairuz Chasseron Chassera
(390)
1680m


@@ # @ @ ©
1450m 1610m 1425m 1550m
1683 m Altitude aboue sea leuel
120km
c>
                                                                     fl
Figure 4 . Spatial variation oF the Ra-226 activity in soil samples from the
         Western Swiss Jura Mountains. Soils have been dried bul
         otherwise untreated,  a) Region oF Ht. Tendre, 1680 m above
         sea level. b> Region oF Marchairuz. all sampling points within
         +/-20 m at 1360 m above sea level,  c) Samples taken near the
         top oFthe mountains, in small depressions. The values shown are
         the highest Found in several samples oFthe respective location.

-------
many  soils  in the Western Jura Mountains contain aeolian  material  from
this  period,  but  the  origin  and  the  original  composition  of  this
windblown material are unknown (4).

    The  activities in the soil samples from even nearby points  can  vary
considerably. Even on the 10 m scale spatial variations can be a factor of
2 as can be seen in figure 4. This figure also shows the variations on the
medium (km) and the large (100 km) scale.

    The  anomaly extends over a strip of at least 100 km lenght  and  some
km  width.  There is a tendency to higher values towards  the  south-west.
Soil  samples from farther west would be very  interesting.  Unfortunately
"farther West" lies in France and until know it has been impossible to get
soil samples from this part of the world.

    It  would  be  a really big surprise if the  radium  anomaly  and  the
extreme  disequilibrium  in  the 23au series in  the  Western  Swiss  Jura
Mountains  would  be  unique.  There are  similar  geologic  and  climatic
conditions all over the world. Have a closer look at your soils !
                             ACKNOWLEDGEMENTS
    I would like to thank Giovanni Ferreri,  Christophe Murith and Georges
Filler for their help to sample and prepare the soils and H.R.  von Gunten
for the alpha spectrometry data.
                The work described in this paper was not funded
                by the U.S. Environmental Protection Agency and
                therefore the contents do not necessarily reflect
                the views of the Agency and no official endorsement
                should be inferred.
                                REFERENCES
 1.  Lauffenburger, T. and Auf der Maur, A., The Concentration
     of Radon in a Town where Radium-Activated  Paints  were
     used. In; Proceedings of the 6 th Int.  Congress of  IRPA,
     Berlin (West), May 1984

 2.  Surbeck,  H., The Search for Radon Sources, a Multi-
     disciplinary Task.
     Rad.Prot.Dosim.,24,1/4 (1988) 431-434.

-------
3.  Surbeck, H.  and  Filler, G.  A  Closer Look  at  the  Natural
    Radioactivity  in  Soils. In : Proceedings of The   1988
    Symposium  on Radon and  Radon  Reduction  Technology,
    Denver CO, October 17-21, 1988, Environmental Protection
    Agency Report EPA-600/9-89-006, Research Triangle Park, 1989

4.  Pochon, M.  Origine  et evolution des sols du  Haut-Jura
    suisse.  Memoires de la Societe Helvetique des Sciences
    Naturelles. Vol.XC. 1976.

5.  Surbeck, H., Radium und Radon im Boden, messtechnische und
    geologische Aspekte, In: Tagungsbericht Kolloquium Messung
    von Radon und Radon-Folgeprodukten, May 6-7, Berlin, FS-
    91-56T, Fachverband fuer Strahlenschutz, 1991, ISBN 3-8249-0028-9.

6.  Surbeck, H., The Search for Radon Sources, a Multi-
    disciplinary Task, In : Proc. of the 2nd Workshop on Radon
    Monitoring in Radioprotection, Environmental and/or Earth
    Sciences, November 25 - December 6, Trieste, Italy, in press.

7.  Tanner A.B., Radon Migration in the Ground:a Supplementary
    Review. In : Gesell, T.F. and Lowder, W.M. (eds.),
    Proc.Nat.Rad.Env.Ill, 5, Conf-780422, US Dept. of Commerce,
    Nat.Techn.Inf.Serv., Springfield,VA, 1980

8.  Fleischer, R.L., Moisture and Rn-222 Emanation,
    Health Phys. 52/6, 797-799, 1987

9.  Surbeck, H. and volkle, H., Radionuclide Content vs. Grain Size
    in Soil Samples. The Science of the Total Environment 69 (1988)
    379-389.

10. Gaiffe, M., Processus pedogenetiques dans le karst jurassien,
    analyse de la complexation organo-minerale en ambiance calcique,
    Ph.D. thesis, Univ. de Franche-Comte, Besancon, France, 1987,
    ISSN 0150-679X.

11. Surbeck, H., Volkle, H., Zeller, W., Radon in Switzerland, ^n  :
    Proc. of The 1991 Internat. Symposium on Radon and Radon Reduction
    Technology, Philadelphia PA, April 2-5, 1991, Environmental
    Protection Agency Report EPA-600/9-91-37a-d, National Technical
    Information Services (NTIS), Springfield VA, 1992.

-------
                                                                            VIII-3
        RADON POTENTIAL OF THE GLACIATED UPPER MTDWF.ST-
     GEOLOGIC AND CLIMATIC CONTROLS ON SPATIAL VARIATION

                           R. Randall Schumann
                           U.S. Geological Survey
                MS 939 Federal Center, Denver, CO 80225-0046

      Many areas of the United States underlain by soils derived from continental
glacial deposits generate elevated indoor radon levels (£ 4 pCi/L). For example,
Iowa (71 percent), North Dakota (63 percent), and Minnesota (46 percent) have
some of the highest percentages of homes with elevated indoor radon levels in the
State/EPA Indoor Radon Survey. Determining the radon potential of glaciated
areas is complicated by several problems: 1) surface radioactivity is generally
uncharacteristically low in glaciated areas and does not  appear to  correlate well
with indoor radon values; 2) because glaciers redistribute the bedrock they
override and entrain, the composition and physical properties of till soils do not
necessarily reflect those of the underlying bedrock; and 3) where glacial cover is
thin (less than about 10 m), the radon potential may be a complex product of the
glacial cover and the underlying bedrock. Data from field studies in North Dakota
and Minnesota suggest that radium is leached from the near-surface soil
horizons and accumulates deeper in the soil profile, providing a radon source at
depth which is not readily detectable by surface radioactivity.  Crushing and
grinding of rocks by glaciers increases the mobility of uranium and radium in the
resulting tills, allowing them to move readily downward through the soil profile
with other mobile ions as the soils are leached.  Soil-gas  and indoor radon values
in North Dakota are generally higher than those in northern Minnesota. The
higher radon levels in North Dakota are likely due to enhanced emanation and
relatively higher radium concentrations in the clay-rich soils, whereas the
generally higher permeability and deeper weathering profiles of the sandy till
soils in Minnesota allow soil gas to be drawn into structures from  a larger source
volume, enhancing indoor radon levels in these areas. Some of the highest indoor
radon levels in North Dakota are associated with deposits of glacial Lake Agassiz
and other glaciolacustrine  deposits.  In contrast, glacial  lake deposits in
Wisconsin and Michigan are typically associated with low radon levels.
Differences in source-rock  composition and soil moisture conditions are likely
responsible for this  dissimilarity.

-------
                                                                           VIII-4
Title:  EPA's National Radon Potential Map

Author: Sharon Wirth, U. S. EPA, Office of Radiation Programs

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

-------
  Session IX
Radon Surveys

-------
                            ABSTRACT

              Comparing the National and State/EPA
                    Residential Radon Survevs
by:       Jeffrey L. Phillips
          Frank Marcinowski
          U.S. Environmental Protection Agency
          Office of Radiation Programs
          Washington, D.C.  20460
     The purpose of this paper is to examine the scope and
findings of two extensive EPA radon surveys, the National
Residential Radon Survey (NRRS), and the State/EPA Residential
Radon Survey (SRRS).  Both surveys provide estimates of radon
distribution across the U.S. while utilizing different survey
designs and concentrating on different units of analysis.  This
paper will attempt to provide a comparative account which
outlines both complimentary and deviant characteristics of each
data set.

     Statistical parameters, survey goals and design, survey
implementation, and survey analysis will provide points of
discussion and comparison.

     The NRRS was initiated in 1989 and completed in 1991.  The
SRRS was initiated in 1986 and is still ongoing, with a total of
40 States participating to date.

-------
             FINAL  REPORT

     RADON TESTING IN NORTH DAKOTA
          DAY CARE FACILITIES
               JUNE 1992
              Prepared  By:

           Arlen L. Jacobson
       Radiation Control Program
 Division of Environmental Engineering
North Dakota State Department of Health
     and Consolidated Laboratories

-------
                            ABSTRACT

     The North Dakota  State Department of Health and Consolidated
Laboratories'  survey of day care facilities  in  North Dakota has
gathered data that will be useful in the  mitigation of radon in
those facilities.  While many of the county,  city,  and facility
arithmetic means were less than 4 picocuries  per  liter,  a large
number were above this level.  The data will enable the Department
to map the concentrations in day care facilities and prove useful
in predicting radon "hot spots."

                          INTRODUCTION

Background

     During the second and third years of a three-year grant from
the U.S. Environmental Protection  Agency (EPA),  the North Dakota
State Department of Health and Consolidated Laboratories (NDSDHCL)
began testing for radon in 2,000  licensed day care facilities.

     The day  care  facilities  are licensed  by the  North Dakota
Department of Human Services.

     During January 1991, the Department mailed 2,019 application
packets to  day care  operators.   A second  mailing was  sent to
1,596 day care facilities in April  1991.  Of  the  2,000 day care
facilities solicited,  411 facilities responded.

Sampling

     The Department followed EPA guidance for sampling protocol.
Every  third facility was  sent a  duplicate  and  every  seventh
facility received a field blank.

Determination of Sampling Locations

1.   At least one detector was placed on each level of the house.

2.   After locating one  detector on each level, the locations are
     prioritized (based upon room  occupancy).

     Highest Priority  Rooms:  living  room, recreation room, dining
     room, and family room.

     Lower Priority Rooms:   bedroom, sleeping room,  TV room, toy
     room, kitchen, entry,  etc.

     Not tested:  cloak room, bathroom, pantry, furnace room, etc.

-------
      The Department followed EPA guidance to formulate its QA Plan
 and used the  National Database  format  for recording  collected
 data.

 Precision

      Precision was  determined  by utilizing  a  coefficient  of
 variation computer program.

 Accuracy

      Accuracy  was determined from a series of samples  that  were
 exposed in  radon chambers  in  laboratories  at  the  Office  of
 Radiation  Programs -  National Air and  Radiation Environmental
 Laboratory/  Las  Vegas,  Nevada facility  or  Montgomery,  Alabama
 facility.   In  all, 24 sampling vials  were sent to Las  Vegas  and
 181 sampling vials to  Montgomery.

      Data was  verified by use  of  a QA computer program.

Type of Sample

     A  "Type of Sample" field was  added to record the  sample type.
 Selections  for  this field were from the  following:

 1.   NR  A normal or  regular  sample.

 2.   FB  A field blank that is to remain closed in a  room that is
          not being tested.

 3.   OD  Original duplicate.

 4.   CD  Complimentary duplicate.

 5.   AC   Accuracy check,  a sample  that has  been placed  in a
          chamber  of known radon concentration.

                           PARAMETERS

Geological Regions

     Table 1 contains  the  arithmetic  and  geometric means  for
selected day care  parameters.  The geological region with the most
samples had the lowest average radon concentration.

Basement Sump

     Many buildings are built on flood plains  and a large number
have sumps  in   basements.   Sumps  were  especially prevalent in
cities built near  rivers.  It  has been noted  previously,  by the

-------
Department, that basements with sumps generally have higher radon
concentrations.  The radon concentration can usually be lowered by
covering the sump.   A surprising 440 samples, about  26%, were from
homes containing sumps.

Mobile Homes

     Of 545  day care  facilities, 29 (5.32%)  were  mobile homes.
Mobile home day care facilities were given one sample to be placed
in  the living  room unless  the  facility was  scheduled  for  a
duplicate sample or field blank sample.  Only 2.79% of the total
samples were from mobile homes.  The average radon concentration
for mobile homes was near 4 pCi/1.

Heat Distribution

     An "other"  category  was  added to the  Federal Database for
heat distribution systems  to include combination systems and wood
burners that are only  used during  the winter heating season.
Forced air was  the  most popular type of heat distribution system
(72.12%).      Individual   heating  units   should  be   further
investigated.

Test Location Floor

     The  most  popular  sampling  floors  were the first  floor
(61.63%) and the basement (31.49%).

Type of Facility

     The facility breakdown is as follows.   The family type of
facility (64.29%), Group  (28.29%), Center (5.28%),  In Home (0%),
Multiple (1.01%), Half-Day (0.65%), Pre-School (0.12%), and Public
Approval (0.36%). Note, the family and group facilities are more
than 90% of the samples.

Basement,  Crawl Space,  or Slab-on-Grade

     The basement, crawl space, and slab-on-grade building styles
can be found  in combinations  in some older homes.  Older homes
generally have  crawl spaces  for small  houses and  basements for
larger houses.   Some houses are a combination of all three types.
Slab-on-grade is not a  popular building style in  North Dakota,
comprising only 3.68%  of the  samples.

Type of Building

     Single  family  buildings  comprise 85.05%  of  the  samples.
Following  Federal   guidelines,  the Department  categorized  the
buildings  such  that most  buildings were contained in the first

-------
four types.  The most popular other type of building selected was
churches.

Room Type

     The relatively large average radon concentration of bedrooms
prompted the following investigation:

             Arith. Mean
             Radon Cone.   % of      % of      % of       % of
   Room Type    in pCi/1    Samples   Basement  1st Floor  2nd Floor

   Bedroom       5.10       29.89     26.19      58.33      4.63
   Family Room   5.61       12.63     66.20      31.46      2.35
   Living Room   4.57       20.88     6.25      91.48      2.27
   Other Room    5.31       11.63     22.96      72.96      4.08

     The remaining room  types all contained less than  5%  of the
samples.  The  large  percentages of bedrooms and family rooms in
the  basement  is a   contributing  factor  to  their  high  radon
concentration.

Maximum Number of Children/Type  of Facility

     The most  prevalent  maximum number of children  found  in day
care  licenses  are 7  (59.31%),  followed  by  12  (13.05%),  and  6
(5.22%).  The maximum number of children licensed for a center is
140 children.   Arithmetic and geometric  means  are  contained in
Table 2.

Testing Company

     Two different radon laboratories were  used to  analyze day
care samples.  The first  laboratory analyzed 1054 samples and the
second  company  analyzed  1568  samples.     Differences between
companies are explained  in the Variance section.

Two Testing Seasons

     Two   different   radon   testing  seasons   were   involved,
October 1990   through   April 1991  and   October 1991  through
April 1992.  A case  can be  made concerning the  two  different
seasons, yet a  case  can  also be made about  differences within a
season.   In North Dakota  it makes a difference whether one samples
in October, April, or January.

     Figure 1,  "North Dakota  County Data"  contains  the essential
arithmetic  means,   standard  deviation,   number   of   samples,
geological  region,  and  weather station.    For  county data  see
Table 3 and for city  data see Table 4.

-------
                       QUALITY ASSURANCE

Field Blanks

     Nine field blanks with the accuracy checks were sent to the
second testing company.  The results were 5 samples with 0.0 pCi/1
and 4 samples with 0.1 pCi/1 radon concentration.  Of the 93 field
blanks sent to day care facilities most were 0.1 pCi/1, however,
some were  greater than  1 pCi/1.    It is assumed  the day  care
operator opened the field blank for those greater than 1  pCi/1.

Accuracy

     The first company's  mean accuracy  for  a 5.9 pCi/1 standard
was 72.7% with a standard deviation of  4.73.   Twelve samples of
11.9 pCi/1 concentration averaged  93.5% with a standard deviation
of 13.44.

     The second  company averaged  110.7%  (Std.  Dev.  12.37)  for
85 samples for a standard of 5.4+/-0.4  pCi/1.  The other standard,
12.8+/-0.8 pCi/1 averaged 110.5% with  standard deviation of 4.86.

Coefficient of Variation

     Eighty-one duplicate  pairs were  analyzed during  the first
radon season.  The coefficient of variation  arithmetic mean was
16.74% with a standard deviation of 29.61.

     During the second radon season,  the average coefficient of
variation was 9.49%,  standard deviation 11.42.

                            VARIANCE

     Utilizing S.A.S. PROC NESTED software,  the  variance of the
parameters were calculated.  The results follow:

-------
PARAMETER
County
City
Building
Geological Regions
Basement Sump
Mobile Homes
Heat Distribution
% OF TOTAL
VARIANCE
21.5
23.3
90.6
6.2
9.2
1.2
3.1
PARAMETER
Test LOG. Floor
Type of Facility
Bsmt Crawl Space
Slab-on-Grade
Type of Building
Type of Room
Maximum Children
Maximum Children
Type of Facility
% OF TOTAL
VARIANCE
5.2
1.90
1.5
5.8
0.8
4.7
5.6
     Testing  companies
program.
were  separated  by  a  PROC  UNIVARIATE

Initial Testing
Company
Final Testing
Company
N
593
1093
ARITH.
MEAN
4.57
5.63
ARITH.
STD.DEV.
6.05
7.29
LOG 10
MEAN
2.83
3.63
LOG 10
STD.DEV.
2.73
2.50
              CROSS VARIANCE IN % OF TOTAL VARIANCE
Maximum Children        0
Type of Facility        5.62

City                   23.28
Geological Region       0

County                 16.06
Building               74.69
Basement Sump           0

County                 16.06
Geological Region       0
Building               74.69

City                   10.20
Building               80.46
Basement Sump           0
          Building
          Geological  Region

          City
          Basement  Sump

          County
          City
          Basement  Sump

          City
          Geological  Region
          Building
80.46
 0

 0
50.44

12.29
 0
44.24

10.20
 0
80.46

-------
     There    are    47 counties,    104 cities,    545 buildings,
3 geological regions,  and Yes or No  indicating whether or not the
facility contains  a basement sump.  Major contributors to variance
are the county, the city, the building,  and the sump.

                 WEATHER ANOMALIES AND RAINFALL

     The parameters the  Department  uses  from the three national
weather stations at Bismarck, Fargo,  and Williston are a departure
from the 30-year temperature average, inches of rainfall, inches
of snow, average  wind speed, percent of  possible  sunshine,  and
weather conditions,  e.g., hail, fog, rain, snow, etc.

     Figure 2 indicates  the data  is  log normal.   The results as
computed by the S.A.S. PROC UNIVARIATE program are:

               N                   1686
               Mean                3.32
               Std.Dev.            2.60
               W:  Normal          0.98441
               Median              3.30

                    SUMMARY  AND CONCLUSIONS

     Conducting a  survey using the  National DataBase  is  labor
intensive.    Often the  instructions  to  testing applicants  are
either not understood  or not followed.    In  future surveys,  the
Department should conduct  organizational  meetings  in  which  the
applications can be completed and  questions answered.

     More effort  and expertise is   needed  in  the area of  the
mapping of radon concentrations.

     This survey shows day care facilities have less  of a  radon
concentration than do homes.   The problem is that both average
concentrations in  North Dakota are greater than 4 pCi/1, day care
facilities (5.18)  and homes (7.00).

     The  Department   should   continue to  work  with  day  care
operators to help them mitigate their radon problems.

-------
                       Table 1
                       DAT CARE
Arithnetic and Geometric Means for Selected Par*
iters
Arithmetic and Geometric Me
Parameters
% Of
Samples
ans for TO and OP Samples in pCi/1
Arithmetic
Mean
Mean
Std.Dev.

Geometric
Mean
Mean
Std.Dev.

Geological Region
1
2
3
All
11.33
35.94
52.73
100.00
5.24
6.99
4.08
5.26
8.34
8.87
4.26
6.89
3.14
4.26
2.84
3.32
Basement Sump
Yes | 26.10
No | 73.90
7.52
4.46
9.36
5.56
4.55
2.97
2.51
2.77
2.41
2.60

2.74
2.48
Mobile Home
Yes
NO
Heat Distribution
Hot Air
Hot Hater
Radiant
Individual Units
Other
2.79
97.21

72.12
14.18
4.21
6.88
2.49
3.81
5.30
4.96
6.94

5.43
3.64
4.18
7.59
4.76
6.82
4.08
3.27
11.77
5.70
2.02
3.37

3.50
2.32
2.81
4.31
3.41
2.96
2.58

2.54
2.61
2.71
2.79
2.08
Test Location Floor
Basement
First
Second and Above
31.49
61.63
6.88
6.91
4.62
3.42
8.07
6.36
3.32
4.65
2.89
2.46
2.46
2.59
2.26
Type of Facility
Family License
In Home License
Multiple
Half-Day
Preschool
64.29

1.01
0.65
0.12
4.96

3.40
3.72
5.15
5.45

2.80
2.27
1.95
3.34

2.01
2.85
4.77
2.46

3.22
3.23
1.49

-------
ArithMtic and G«<
    Table  1
    OAT GARB
»tric Means  for Selected Pan
  (Continued)
                                                  iters
Arithmetic and Q«<
^— ^^— — ~— — —
Parameters
aaetric M<
^^SSS^S^^^^^^^S
% of
samoles

tans for
.
MR and "" fl*nples in pci/l
Arithmetic
Mean
Mean
Type of Facility (Cont'd)
center
Group
Public Approval
Type of Basement
Basement
Crawl Space
Slab on Grade
other
5.28
28.29
0.36

17.16
11.51
3.68
7.06
3.79
6.37
1.05

5.54
4.65
4.69
3.44
Std.Dev.

Geometric
Mean
1 	
1 Std.Dev.
I

5.30
9.62
0.84
2.15
3.69
0.75

6.66
9.27
7.35
3.39
3.63
2.50
2.54
2.32
2.67
2.76
2.30

2.52
2.75
2.82
2.56
Type of Building
single Family
Multi-Family
Business
School
Other
85.05
4.86
3.32
2.43
4.33
5.64
3.03
3.84
3.01
2.52
7.26
1.75
5.70
4.39
1.87
2.54
2.53
1.76
1.78
1.87
3.63
1.89
3.27
2.71
2.32
Room Type
Bedroom
Family Room
Living Room
unfinished
Basement
Office
Classroom
other
Finished Basement
Kitchen
Recreation Room
Dining Room
29.89
12.63
20.88
3.44
0.24
4.98
11.63
1.48
3.02
2.49
2.49
5.10
5.61
4.57
6.85
5.68
3.29
5.31
7.13
5.37
7.00
5.31
6.60
6.20
7.22
6.17
4.46
4.12
6.99
4.73
8.82
6.36
7.65
3.33
3.92
2.74
5.11
3.50
2.03
3.37
4.92
3.21
4.90
3.20
2.50
2.34
2.71
2.09
3.03
2.67
2.49
2.99
2.46
2.39
2.60
                       10

-------
Arithaetic and Ge<
   Table 1
   DAT CARE
trie Mean* for Selected Par.
  (Continued)
Arithmetic and Geometric Means for MR and OD 8
Parameters
% Of
Samples

Arithmetic
Mean
Mean
Std.Dev.

aaples in pCi/1
~^ —
Geometric
Mean
Mean
Std . Dev .

Room Type (Cont'd)
Playroom
TV Room
Sleeping Room
Nursery






















3.91
0.36
1.72
0.77






















7.01
4.80
6.45
8.02






















7.57
3.29
7.34
15.16






















4.46
3.67
3.91
2.84






















2.62
2.14
2.65
4.09






















                        11

-------
                   Table 2
                   DAT CARE
Maximm Number of Children and Type of Facility
Arithmetic and Geometric Means for HR and OD S»«ple« in DCi/1
Maximum Number of
children/Type of Facility

6 Family
7 Family
8 Group
9 Group
10 Group
11 Family
1 1 Group
12 Group
13 Group
14 Family
14 Group
14 Half-Day
15 Group
16 Pre-School
1 6 Group
1 7 Group
18 Group
18 Half -Day
24 Pre-School
30 Public Approval
32 Center
35 Center
45 center
5 0 center
55 Center
56 Center
58 center
% Of
samples

5.22
59.31
1.01
1.42
0.71
0.30
1.60
13.05
0.65
0.12
2.08
0.12
1.84
0.06
0.53
0.18
4.21
4.74
0.06
0.12
0.06
0.47
0.24
0.95
0.30
0.12
0.30
Arithmetic
Mean
Mean

3.75
5.05
4.71
5.33
2.78
8.68
3.81
7.04
6.14
4.10
4.77
0.85
6.06
3.20
17.63
6.97
4.87
4.36
7.10
2.20
3.20
7.11
1.95
2.48
1.04
12.90
1.56
Std . Dev .

2.95
5.63
2.47
3.84
1.88
1.27
5.47
9.55
4.36
1.10
3.66
0.25
4.72
0
30.48
12.46
0.80
2.01
0
0.20
0
5.16
0.15
0.77
0.37
12.40
0.37
Geometric
Mean
Mean
2.62
3.39
4.09
3.94
2.17
8.59
2.21
4.26
4.61
3.94
3.56
0.81
3.57
3.20
3.34
3.31
4.84
3.77
7.10
2.19
3.20
3.53
1.94
2.38
0.96
3.56
1.51
Std . Dev .

2.53
2.45
1.71
2.23
2.03
1.15
2.60
2.74
2.20
1.32
2.29
1.35
3.57
0
5.27
2.96
1.12
1.78
0
1.10
1
4.35
1.08
1.32
1.51
7.11
1.32
                    12

-------
     Table 2
     DAT CJUtB
of Children and Type of Facility
   (Continued)
AriJ-haetie and Geometric MMP« for HR and OD Savple* in DCi/1
Maximum Number of
children/Tvce of Facility

60 Multiple License
61 Multiple License
64 Multiple License
65 Center
75 center
80 Center
83 Multiple License
85 Center
95 Center
108 Center
116 center
120 Center
123 Center
135 center
140 Center












% Of
Samples

0.59
0.12
0.30
0.30
0.12
0.71
0.12
0.24
0.30
0.30
0.06
0.24
0.18
0.24
0.30












Arithmetic
Mean
Mean

1.35
2.75
0.40
1.68
3.80
15.42
6.50
1.05
2.64
3.08
0.70
3.28
7.73
1.20
1.12












Std.Dev.

0.05
0.35
0.06
0.19
0
8.42
0.30
0.18
0.77
0.55
0
0.50
6.58
0.19
0.24












Geometric
Mean
Mean
1.35
2.73
0.39
1.67
3.80
11.69
6.49
1.03
2.53
3.03
0.70
3.24
4.90
1.18
1.09












Std.Dev.

1.04
1.14
1.18
1.12
1
2.35
1.05
1.19
1.34
1.20
1
1.15
2.76
1.18
1.27












     13

-------
                           Ttbl* 3
COUNTY
                   DAYCARE AND 1988 STATE STUDY

         RNCOWCinBD  PERCE8TGT4   PCTGT20  RNCONC1998  PCTCT41988  PCTGT2088
ADAMS
BARNES
BENSON
BILLINGS
BOTTINEAU
BOWMAN
BURKE
BURLEIGH
CASS
CAVALIER
DICKEY
DIVIDE
DUNN
EDDY
ENMONS
FOSTER
GOLDEN VALLEY
GRAND FORKS
GRANT
GRIGGS
HETTINGER
KIDDER
LAMOURE
LOGAN
MCHENRY
MCINTOSH
MCKENZIE
MCLEAN
MERCER
MORTON
MOUMTRAIL
NELSON
OLIVER
PEMBINA
PIERCE
RAMSEY
RANSOM
RENVILLE
HIGHLAND
ROLETTE
SARGENT
SHERIDAN
SIOUX
SLOPE
STARK
STEELE
STUTSMAM
TOWKER
TRAILL
WALSH
WARD
WELLS
WILLIAMS
ALLCOUNTIES
13.30
8.29
0.00
0.00
3.58
20.62
2.27
3.76
7.37
3.95
4.32
3.89
2.23
3.63
3.38
0.00
1.81
10.98
9.05
3.09
2.44
3.02
3.97
2.60
4.94
6.26
2.70
3.84
3.44
3.70
3. 10
3.20
0.00
4.23
2.43
3.06
10.38
3.03
1.82
2.40
3.00
0.00
0.00
0.00
3.41
6.87
4.12
6.33
6.30
13.33
3.23
2.79
2.44
3.18
83.33
73.76
0.00
0.00
41.67
76.92
0.00
29.15
62.91
45.83
60.00
25.00
0.00
30.00
31.58
0.00
12.50
83.61
100.00
37.03
22.22
20.00
44.44
7.14
33.71
80.00
0.00
16.67
.23.00
28.73
29.17
0.00
0.00
25.00
10.00
37.50
78.57
33.33
94,34
3.00
40.00
0.00
0.00
0.00
27.69
100.00
49.26
100.00
51.72
68.42
22.05
33.71
13.04
41.40
16.67
13.13
0.00
0.00
0.00
38.46
0.00
0.95
4.69
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
10.07
0.00
0.00
0.00
0.00
0.00
0.00
7.14
0.00
0.00
0.00
0.00
1.25
4.17
0.00
0.00
0.00
0.00
0.00
7.14
0.00
0.00
0.00
20.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
31.38
0.00
0.00
0.00
2.97
9.30
8.00
7.20
8.70
6.00
0.70
2.<0
4.90
7.90
3.70
3.50
8.80
8.70
4.60
6.60
3.70
4.00
1.70
8.10
3.30
7.20
4.00
5.20
3.70
3.60
7.10
3.50
.00
.10
.60
.30
.30
.40
.40
.80
.60
.40
.10
.00
.50
.50
.00
.90
.20
.00
.20
.80
.10
.80
0.50
4.20
3.60
4.60
7.00
78.30
78.90
62.30
77.80
60.60
71.00
0.00
48.50
69.00
42.90
81.80
100.00
75.00
50.00
CO. 00
57.10
42.90
90.10
60.90
20.00
61.30
50.00
60.00
76.90
33.30
44.40
50.00
70.60
73.90
53.50
60.00
69.30
84.20
72.90
29.40
77.80
57.10
66.70
60.90
50.00
70.00
60.00
50.00
42.90
60.10
42.90
65.00
50.00
65.40
75.50
45.50
27.30
47.10
63.00
13.00
7.90
12.50
11.10
3.00
6.50
0.00
1.00
1.00
0.00
0.00
0.00
10.70
0.00
6.70
0.00
0.00
11.00
4.30
0.00
6.50
0.00
0.00
0.00
0.00
11.10
0.00
0.00
6.50
3.00
10.00
0.00
0.00
13.60
0.00
0.00
0.00
0.00
4.30
0.00
0.00
0.00
0.00
0.00
4.90
0.00
2.30
10.00
3.80
10.20
1.50
9.10
0.00
4.00
                              Legend

 RNCONCNROD - Arithmetic mean of day  care  radon
                concentrations
 RNCONC1988 - Arithmetic mean of the  1988  State Study
                radon  concentrations
              PERCENTGT20 - Percent of day  care samples
                greater  than 4 pCi/1 and  20 pCi/1,
                respectively.
 PCTGT41988, PCTGT2088  - Percent of 1988 State Study
                samples  greater than 4 pCi/1 and
                20 pCi/1,  respectively.
PERCENTGT4,
                           14

-------
                                  Table  4
                    Day Care Arithmetic and Geometric Means
                          and Standard Deviations
COUNTY

ADAMS

BARNES
BENSON
BILLINGS
BOTTIMEAU

BOWMAN
BURKE
BURLEIGH
CASS
CAVALIER

DICKEY

DIVIDE

DUNN

EDDY

EMMONS
FOSTER
GOLDEN VALLEY

GRAND FORKS
CITY


HETTINGER

LITCHVILLE
VALLEY CITY



BOTTINEAU

BOWMAN
SCRANTON

BOWBELLS
LIGNITE

BALDWIN
BISMARCK
LINCOLN
MENCKEN
WING

CASSELTON
FARGO
HARWOOD
KINDRED
MAPLETON
WEST FARGO
WHEATLAND

LANGDON

ELLENDALE

CROSBY

KILLDEER

NEW ROCKFORD

HAZ ELTON
LINTON


BEACH

EMERADO
GRAND FORKS
LARIMORE
THOMPSON
N

6
6
33
4
29


12
12
13
10
3
7
5
2
422
3
383
26
3
4
213
5
134
11
5
3
53
2
24
24
5
5
8
8
8
8
4
4
19
4
15

8
8
139
3
132
1
3
ARITH
MEAN
13.30
13.30
8.80
6.33
9.14
0.00
0.00
3.35
3.35
22.74
27.15
8.03
2.27
2.28
2.25
3.72
2.37
3.77
2.13
6.43
8.40
7.58
9.70
7.72
5.78
5.22
1.53
8.04
5.60
3.80
3.80
4.32
4.32
3.84
3.84
2.13
2.13
3.03
3.03
3.55
1.85
4.01
0.00
1.75
1.75
11.38
7.73
11.63
0.90
7.43
AMSTD
DEV
7.42
7.42
7.90
0.65
8.37
0.00
0.00
1.81
1.81
22.37
23.80
0.26
0.24
0.28
0.05
3.64
0.19
3.68
1.31
0.56
5.97
9.89
3.01
8.65
5.80
0.27
0.54
13.84
1.40
2.37
2.37
2.37
2.37
0.64
0.64
0.73
0.73
1.78
1.78
1.95
1.48
1.80
0.00
1.89
1.89
10.97
6.58
11.15
0.00
0.25
GEO
MEAN
9.91
9.91
6.12
6.29
6.10
0.00
0.00
2.86
2.86
12.28
13.95
8.03
2.26
2.26
2.25
2.71
2.36
2.76
1.78
6.41
6.01
4.61
9.26
4.83
2.73
5.21
1.45
4.49
5.42
2.73
2.73
3.50
3.50
3.79
3.79
2.01
2.01
2.45
2.45
3.01
1.44
3.66
0.00
1.28
1.28
8.41
4.90
8.69
0.90
7.43
GEOSTD
DEV
2.53
2.53
2.37
1.10
2.51
0.00
0.00
1.77
1.77
3.14
3.59
1.03
1.11
1.13
1.02
2.23
1.09
2.23
1.86
1.09
2.37
2.80
1.39
2.77
4.55
1.05
1.39
2.66
1.29
2.66
2.66
2.01
2.01
1.17
1.17
1.38
1.38
1.96
1.96
1.84
1.93
1.50
0.00
1.96
1.96
2.22
2.76
2.17
1.00
1.03
                               15

-------
                                 Table  4

                    Day Care Arithmetic and Geometric Means
                          and Standard Deviations
COUNTY

GRANT

GRIGGS


HETTINGER



KIDDER

LAMOURE


LOGAN


MCHENRY



MCINTOSH


MCKENZXE

MCLEAN




MERCER


MORTON



MOUNTRAIL




 NELSON

 OLIVER
 PEMBINA


 PIERCE
CITY
N
ARITH AMSTD
MEAN DEV

ELGIN

COOPERSTOWN
HANNAFORO

NEW ENGLAND
MOTT
REGENT

STEELE

EDGELEY
LAMOURE

CACKLE
NAPOLEON

TOWNER
UP HAM
VELVA

ASHLEY
WISHER

WATFORD CITY

GARRISON
MAX
TURTLE LAKE
WASHBURN

BEULAH
STANTON
HEBRON
MANDAN
NEW SALEM
PARSHALL
PLAZA
NEW TOWN
STANLEY
LAKOTA

CAVALIER
ST. THOMAS
2
2
27
23
4
9
3
3
3
5
5
9
5
4
14
5
9
14
6
4
4
20
4
16
10
10
24
7
3
8
6
8
5
3
80
2
74
4
24
4
3
5
12
1
1
16
12
4
10
9.05
9.05
3.31
2.70
6.78
2.54
1.70
4.83
1.10
3.02
3.02
4.07
1.08
7.80
2.64
2.52
2.70
4.94
7.83
2.50
3.03
6.16
6.80
5.99
2.79
2.79
3.60
1.60
0.90
7.15
2.55
3.45
4.26
2.10
3.70
4.95
3.67
3.58
4.90
1.95
19.73
3.44
2.79
3.20
3.20
0.00
4.23
4.76
2.65
2.45
0.75
0.75
2.02
1.38
1.55
1.92
1.24
1.09
0.57
0.92
0.92
3.35
0.19
0.47
0.93
0.37
1.12
5.77
7.82
0.07
1.56
2.72
0.57
3.01
0.72
0.72
3.75
1.06
0.00
4.60
0.73
1.38
1.14
0.14
3.17
1.05
3.29
0.13
5.95
1.57
2.95
0.46
2.12
0.00
0.00
0.00
4.11
4.63
0.32
2.57
GEO GEOSTD
MEAN DEV
9,02
9.02
2.74
2.35
6.60
1.80
1.27
4.72
0.98
2.87
2.87
2.58
1.06
7.79
2.50
2.49
2.51
3.38
4.91
2.50
2.60
4.69
6.78
4.28
2.70
2.70
2.40
1.29
0.90
5.89
2.43
3.20
4.12
2.10
2.87
4.84
2.79
3.57
2.69
1.17
19.53
3.41
1.96
3.20
3.20
0.00
2.77
2.82
2.63
1.68
1.09
1.09
1.89
1.74
1.26
2.40
2.19
1.25
1.60
1.39
1.39
2.72
1.20
1.06
1.36
1.15
1.45
2.17
2.64
1.03
1.75
2.72
1.09
3.00
1.30
1.30
2.41
1.92
1.00
1.83
1.37
1.47
1.30
1.07
2.07
1.24
2.11
1.04
3.16
3.20
1.15
1.14
2.51
1.00
1.00
0.00
2.44
2.79
1.13
2.38
                                16

-------
                                  Table  4

                    Day Care Arithmetic and Geometric Means
                          and Standard Deviations
COUNTY



RAMSEY


RANSOM


RENVILLE

HIGHLAND
ROLETTE
SARGENT
SHERIDAN
SIOUX
SLOPE
STARK
STUTSMAN



STEELE

TOWNER

TRAILL
WALSH

WARD
CITY

RUGBY
WOLFORD

DEVILS LAKE
STARKWEATHER

ENDERLIN
LISBON

GLENBURN

BARNEY
COLFAX
HANKINSON
WAHPETON
WYNDMERE

DUNSEITH
ROLETTE
ROLLA
ST. JOHN

FORMAN
KILNOR




BELFIELD
DICKINSON
SOUTH HEART

JAMESTOWN
MONTPELIER
STREETER

HOPE

CANDO

BUXTON
HATTON
HILLS BORO
HAYVILLE
PORTLAND

GRAFTON

BERTHOLD
MI NOT
N

7
3
8
5
3
14
3
11
9
9
53
1
6
4
38
4
20
5
9
2
4
5
1
4



65
3
57
5
136
129
4
3
3
3
3
3
29
3
5
6
6
9
19
19
68
1
67
ARITH
MEAN
1.97
3.57
3.80
4.26
3.03
10.99
1.80
13.49
3.14
3.14
1.83
0.10
0.98
0.80
1.99
3.05
2.40
3.50
2.32
3.80
0.50
7.50
25.50
3.00
0.00
0.00
0.00
3.41
4.17
3.32
3.98
4.18
3.97
8.43
7.70
6.87
6.87
6.13
6.13
6.46
2.70
8.58
4.02
8.57
6.77
14.66
14.66
3.24
2.90
3.25
AMSTD
DEV
0.22
4.48
2.17
2.54
0.93
6.37
0.80
4.71
1.41
1.41
1.40
0.00
0.38
0.07
1.46
0.82
1.44
1.70
0.53
0.00
0.07
9.22
0.00
2.23
0.00
0.00
0.00
2.61
0.26
2.65
2.77
2.48
2.33
2.11
0.94
3.43
3.43
1.25
1.25
3.47
0.22
3.91
0.91
3.84
2.37
13.15
13.15
2.88
0.00
2.90
GEO
MEAN
1.96
1.17
3.36
3.66
2.90
8.20
1.63
12.74
2.69
2.69
1.36
0.10
0.93
0.80
1.52
2.95
1.91
3.21
2.25
3.80
0.49
3.82
25.50
2.38
0.00
0.00
0.00
2.60
4.16
2.52
2.72
3.27
3.11
8.18
7.64
6.13
6.13
6.02
6.02
5.58
2.69
7.58
3.93
7.45
6.25
9.29
9.29
2.28
2.90
2.28
GEOSTD
DEV
1.13
4.54
1.60
1.69
1.35
2.50
1.55
1.40
1.87
1.87
2.24
1.00
1.38
1.09
2.14
1.28
2.11
1.47
1.30
1,00
1.15
3.04
1.00
1.91
0.00
0.00
0.00
2.15
1.07
2.14
2.63
2.28
2.27
1.27
1.13
1.59
1.59
1.21
1.21
1.72
1.09
1.68
1.23
1.77
1.52
2.67
2.67
2.43
1.00
2.44
                                17

-------
                                    Table 4

                     Day Care  Arithmetic and Geometric 'leans
                            and  Standard Deviations
COUNTY

WELLS

WILLIAMS
CITY


HARVEY

ALAMO
RAY
WILDROSE
WILLISTON
N

14
14
46
4
5
3
34
ARITH
MEAN
2.79
2.79
2.42
4.23
1.62
7.67
1.86
AMSTD
DEV
1.73
1.73
2.32
1.15
1.70
5.35
0.97
GEO
MEAN
2.31
2.31
1.66
4.08
0.78
2.36
1.62
GEOSTD
DEV
1.83
1.83
2.52
1.29
3.50
9.34
1.70
                                    18

-------
                                 riouu 1


                          North Dtkoti County Data
 Top Line: Geolo|K»l Re|ion. Weitber Statioa


 Middle Un< Numb«r of Stmptei


 Bottom Une Arithmetic Mcu,  Standard Devittioa
                   Daycare  Radon Study 1992
  100.01
-
-
G
o
5

£
—
G
-
CJ
c
O   1.0
o

c

1
£
    :  .
                                                     —i—

                                                      K
—i—

 .:
                          20   30   40   SO  60   70




                         Percentage Less Than
M
                                   19

-------
     Ventilation, Climatology &£ Radon Activity in Four Minnesota Schools

                 Tim Burkhardt,  Earnest Tare, &  Laura Oatman

           Indoor Air Quality Unit, Minnesota Department of Health

                              Minneapolis, MN 55414
                                     ABSTRACT

Radon concentrations were measured in two Minnesota schools and analyzed by examining the
relation between HVAC system operation and radon levels; differences between short-term, long-
term, and continuous testing devices; and associations between atmospheric pressure and radon
levels.  Multiple monitoring was performed in the schools using charcoal canisters, alpha-track
detectors, and Honeywell continuous monitors.  Climatologic data was examined for correlations
to radon concentrations. HVAC operation (on versus off) was found to significantly alter radon
concentration; long-term alpha-track tests were found to give lower results than short-term charcoal
canisters or continuous monitors; atmospheric pressure appeared to be inversely related to radon
concentration.
                                  INTRODUCTION

In early 1989, the U.S. Environmental Protection Agency (EPA) conducted a screening program
to assess radon levels in homes and schools across the country. Sixteen schools were screened in
Minnesota.  EPA then assisted the Minnesota Department of Health (MDH) in conducting a
comprehensive study of those schools with high screening levels.  The four schools with the highest
screening levels (range: 5.5-10.7 pCi/L) were chosen for an extensive one year study.  Selected
results of the national study including the Minnesota schools were reported at  the 1991
International Radon Symposium in Philadelphia, PA (1,2,3). This paper presents a further
analysis of selected findings in two of the four Minnesota schools.
* Present address: Health & Environment Digest, Freshwater Foundation, Wayzata, MN.

-------
The two schools are Gideon Pond Elementary in Burnsville, MN (EPA number MNO) and
Sweeney Elementary in Shakopee, MN (EPA number MN4).  Gideon has two occupied floors, 12
occupied classrooms in contact with die ground, and is 27 years old; Sweeney has one occupied
floor, 24 occupied classrooms in contact with the ground, and is 23 years old.

Bodi buildings are of slab-on-grade construction with the lowest level partially below grade.  First
floor construction is of concrete in both buildings. Utility pipes are located mostly or
completely overhead.

The schools were both equipped with air handlers, univents and radiant heating but differ in the
configurations of their heating, ventilation and air conditioning (HVAC) systems. For example,
some rooms were not ventilated by central air handlers; sizes of fresh air intakes were varied.
Where relevant to radon levels, these differences are discussed below.
                                       METHODS

Four schools were selected for extensive study based on their high screening levels and their
proximity to the Minnesota Department of Health. Detailed methods for the study were
developed by EPA and are given elsewhere (4).

Each classroom in contact with die ground was tested using charcoal canisters, alpha-track
detectors, and Honeywell continuous monitors.  The year-long testing period was divided into
four seasons. In each season a nine-day period was selected and divided into diree "rounds" of
about three days each, with different ventilation conditions in effect during each round (TABLE
1). Round 1 was over a weekend with the HVAC system off; Round 2 was during the week under
normal HVAC conditions (on during die day and set back during the night);  Round 3 was during
the second weekend with die HVAC system on continuously.

During each round of each season, charcoal canisters were placed in all occupied classrooms that
were in contact widi the ground. In other words, in each season, one charcoal canister was used for
each of three three-day periods (rounds). Alpha-track detectors were placed and collected every
diree months. Continuous monitors were set for 24-hour intervals except during die three rounds
when diey were set for four-hour intervals. Three to four continuous monitors were used in each
school and were placed in different rooms. The continuous monitors did not remain in any one
room for the entire twelve months due to needs of the school and a desire by the researchers to
collect data from as many rooms as possible.

-------
The detectors were provided by the EPA and placed in the schools by MDH with the assistance of
officials from each school. Continuous monitors were provided by Honeywell. Climatologic
data was collected at the Minneapolis/St. Paul International Airport by the National Oceanic and
Atmospheric Administration (NOAA).
                                        RESULTS

Data collection was interrupted during each season and in each school for various reasons such as
temporary power shut down, monitor moved from original location, monitor tampered with,
results illegible.  In addition, power shortages during Season 1 in August 1989, prevented
fulfillment of the Round 3 requirement of continuous HYAC operation in two of the four schools
studied. Adherence to the schedule during the other three seasons is not clearly documented and
test results from these periods are not analyzed here.

Results presented here are for Gideon and Sweeney schools, Season 1, Rounds 1-3.  Gideon and
Sweeney were selected from the four schools studied for the quality of data available, especially
adherence to the nine-day HVAC schedule discussed above.  Season 1 was chosen because good
data were available and because the ground remained unfrozen during the entire testing period
(August through October), allowing for better analysis of precipitation data.
VENTILATION EFFECTS
Effects of changing ventilation pattens on measured radon concentrations can be seen in Figures 1
and 2. For the charcoal canister and continuous monitor data, each column represents a three-day
average  for all the rooms tested in that school.

In general, radon levels in both schools were greatest when the HVAC system was off, least when
the system was continuously on, and intermediate when the system was in normal operation (i.e, on
during the school day and set back during the night).  Although the absolute radon levels differ, the
same trend is seen whether measured either by charcoal canisters or continuous monitors.  On
average,  a four- to six-fold decrease in radon concentration was seen with HVAC continuously on
compared  to completely off.  Long-term alpha-track  concentrations are provided for reference
and are approximately equal to the concentrations measured by the short-term testing devices when
the HVAC system was on continuously.

The trend  with respect to HVAC status was also seen  in a third school (Cleveland Elementary-
Secondary School; MN2) that did not follow the HVAC operating schedule.  Here, the HVAC

-------
system was not turned on continuously during Round 3 and tadon concentrations were lowest
during normal operation (Round 2) and elevated in both Round  1 and Round 3.  In a fourth
school (Lincoln Elementary School; MN6), however, where the HVAC schedule also was not
followed, the trend was not observed.  Specifically, Round 1 concentrations were greater than
Round 2, as expected, but Round 3 concentrations were the lowest of the three when they were
expected to be approximately equal to  Round 1 and greater than Round 2.
COMPARISON OF TESTING DEVICES
A comparison of selected results from charcoal canisters, continuous monitors and alpha-track
detectors is given in Figure 3.  The long-term alpha-track reading is considerably lower than the
readings from the two short-term measurement devices.  In the data shown, charcoal canister
readings were from two to four times greater than the long-term alpha-track reading. Particularly
for Rooms 9 and 11, the continuous monitor data highlights the potential daily changes in radon
concentration in a given classroom.

Results from Round 2 (HVAC normal) were selected for Figure 3 because they represent the
building under normal operating conditions (with students present and warm weather ventilation).
Results from Round 1 (HVAC off) were similar.  Results for Round 3 (HVAC on continuously)
were not included in the comparison because the very low radon levels measured are too near
detection limits to  assure significance.

All of the readings are specific for each room.  Continuous monitor data represents an average of
the six four-hour readings given during a 24-hour period. The charcoal canister reading is a three-
day average and the alpha-track reading is a 70-day average that includes the 9-day test period.

In addition to the differences among the measurement devices, it is notable that radon levels in
different rooms varied from each other considerably, thus emphasizing the importance of testing
every room. It is notable also that rooms 9 and 11 are both basement rooms (a full level below
grade) and room 11 has no windows or univents.
CLIMATOLOGIC EFFECTS
A plot of radon concentration (Gideon Pond Elementary School, Room 1) versus atmospheric
pressure is given in Figure 4. Rainfall is also noted. The trend suggested here supports the
observation that radon levels increase with decreasing pressure and vice versa.  The two highest
radon readings in this time period occurred simultaneously with the second- and fourth-lowest
pressure readings (days 16 and 21).  Low radon readings occurred on a number of days that also

-------
had some of the highest pressure readings (days 1-3, 8, 11-15, 18-19, 22). While these results do
not explain or make certain the relationship between radon concentration and pressure, they add
positively to the other data collected on this subject.

From the year's worth of survey data collected, this time period (May 27, 1990- June 19,1990)
was selected for the large number of precipitation events and pressure changes that occurred, and
because it is a period when the ground is completely unfrozen.  This time period also does not
overlap with the HVAC schedules described in TABLE  1, thus avoiding confounding from the
intentional ventilation changes in the school.

Climatologic data was collected at the Minneapolis/ St.Paul International Airport,
approximately 20 miles from Gideon Pond Elementary. Because atmospheric pressure varies with
altitude, the absolute pressure reading at the school is slightly different from that reported here.
However, because the difference is both slight and  systematic, it does not change die effect shown
here (5).
                                      DISCUSSION

The results discussed here add useful information to die knowledge base of radon behavior in
schools. In particular, this study found that HVAC operation can directly and significantly affect
the radon concentration in a school classroom.  Improving ventilation to reduce high radon levels
is therefore recommended as another method of improving indoor air quality in schools.

Comparison of testing devices suggests that a long-term (in this case, 70-day) alpha-track test
gives a lower reading than short-term charcoal canister tests.  Continuous monitor readings vary as
expected and on any single day are likely to be higher than the long-term alpha-track data.
Assuming alpha-track results represent die true radon concentration most accurately, these results
support the use of a long-term radon test over a short-term test.  This is consistent with Minnesota
Department of Health radon testing recommendations.  The observed variation in radon
concentration among different rooms in a given school also demonstrates the importance of testing
each occupied classroom in contact with the ground.

Finally, radon concentration appears to be inversely associated with atmospheric pressure. While
the relationship is not fully understood, the data collected in this study support the previously
suggested association. A better understanding of this effect might be useful for improving radon
testing and mitigation protocols.

-------
                              ACKNOWLEDGEMENTS

The Minnesota Department of Health appreciates the cooperation and assistance received from the
staffs of Gideon Pond, Sweeney, and Lincoln Elementary Schools and from Cleveland
Elementary-Secondary School.  We also thank Gregory Spoden of the Minnesota Department of
Natural Resources State Climatology Office for providing climatologic data and Honeywell for
the loan of their radon continuous monitors.

The work described in this paper was partially funded by the United States Environmental
Protection Agency but has not been reviewed in accordance with dieir peer and administrative
review process and therefore the contents do not necessarily reflect the views of die Agency and no
official endorsement should be inferred.
                                    REFERENCES

 1. Schmidt, A.L.  The results of EPA's school protocol development study.  In: Proceedings of
 die 1991 International Symposium on Radon and Radon Reduction Technology.  United States
 Environmental Protection Agency, Philadelphia, PA,  April 1991, Volume 5, Session 10.

 2. Fisher, G. Diagnostic evaluation of 26 U.S. schools— EPA's school evaluation program.  In:
 Proceedings of the 1991 International Symposium on  Radon and Radon Reduction Technology.
 United States Environmental Protection Agency, Philadelphia, PA,  April 1991, Volume 5,
 Session 10.

 3. Schmidt, A. L., MacWaters, John T., and Chmelynski, Harry.  Seasonal variation in short-term
 and long-term radon measurements in schools. In:  Proceedings of the 1991 International
 Symposium on Radon and Radon  Reduction Technology. United States Environmental
 Protection Agency, Philadelphia,  PA,  April 1991.

 4. School Protocol Development  Study. United States Environmental Protection Agency,
 Office of Radiation Programs, June 1989.

 5. Spoden, Gregory. State Climatology Office, Minnesota Department of Natural Resources.
 Personal  communication, May 1992.

-------
TABLE 1.  MINNESOTA SCHOOLS RADON TESTING SCHEDULE.
SEASON

l
(SUMMER)


2
(FALL)


3
(WINTER)


4
(SPRING)

ROUND

1
2
3

1
2
3

1
2
3

1
2
3
HVAC STATUS

OFF
NORMAL
ON

OFF
NORMAL
ON

OFF
NORMAL
ON

OFF
NORMAL
ON
DATE

7/28/89-7/31/89
7/31/89-8/2/89
8/4/89-8/7/89

l_l 1/3/89-11/6/89
11/6/89-11/8/89
11/10/89-11/13/89

2/2/90-2/5/90
2/5/90-2/7/90
2/9/90-2/12/90

5/4/90-5/7/90
5/7/90-5/9/90
5/11/90-5/14/90

-------
                                          CHARCOAL CANISTER
                    2.5 -

                      2 :
                                                                      0.41
                                OFF               NORMAL               ON
                                 HVAC STATUS (THREE 3-OAY PERIODS: 7/28/89-8/7/89)



                                          CONT1NOUS MONITOR
                                OFF               NORMAL               ON

                                HVAC STATUS (THREE 3-OAY PERIODS: 7/28/89-8/7/89)


                                        ALPHA-TRACK DETECTOR
                                       70-DAY AVERAGE (6/23/89-8/30/89)
Figure 1. Radon concentration and HVAC status: Gideon Pond Elementary School. Concentrations
         shown represent an average of all rooms tested with each testing device during the time period
         specified.

-------
                                        CHARCOAL CANISTER
                  3.5
                   3
                  Z.5
             [Rn].   2
             pCi/L  1-5
                   1
                  0.5
                   0
                             3.0S
                             OFF               NORMAL               ON
                              HVAC STATUS (THREE 3-DAY PERIODS: 7/28/89-8/7/89)

                                        CONTINUOUS MONITOR
                              OFF               NORMAL               ON
                              HVAC STATUS (THREE 3-DAY PERIODS: 7/28/89-8/7/89)

                                      ALPHA-TRACK DETECTOR
0.56
                                                  7
                                     70-OAY AVERAGE (6/23/89-8/30/89)
Figure 2. Radon concentration and HVAC status: Sweeney Elementary School. Concentrations shown
         represent an average of all rooms tested with each testing device during the time period
         specified.

-------
                  2.5 -

                   2 -
                            7/31/89
                                             ROOM 1
     8/1/89
                                                                  8/2/89
                            7/31/89
                                             ROOM 9
    8/1/89
8/2/89
             [Rn].
             pCi/L
                            7/31/89
                                             ROOM 11
     8/1/89
8/2/89
            Alpha-track
Charcoal canister
       Continuous monitor
Figure 3. Comparison of radon measurement devices in Gideon Pond Elementary School. Charcoal
         canister and alpha-track results are three-day and 70-day averages, respectively.

-------
        3


      2.5


        2
pCI/L
        1


      0.5

^B
q



-

,1 — 1,1 |,






j 	 4 j 	 i+j 	 L(_ 	 -^
1234





\
\



5




/
/
\/


V
R


/

R



-H 	 -t • " — i '
6 7










8















\ ,
•
, — |
R
9




B.--*-,
/



_...




,n,
10 11




/
V



12




/







p
/





13










DAY (May 2 7- June 19. 1990)


1







Radon concentration







\







14










R=







\
P



•!,
__—

V

R
/
/
/


i —

R



15 16 17
'RAINFALL


p
/





18









I





.



I — 1

19





\













| 	
R






V


R



/




20 21

Atmospheric pressure




•
/


—

22






















R
r 29.4
29.2


29

28.8
28.4
28.2
to
23


                                                                                                                WINCHES
          Figure 4.  Radon concentration and atmospheric pressure in Gideon Pond Elementary School, Room 1.
                    Pressure data was collected at the Minneapolis/St. Paul International Airport.

-------
     ESTIMATES  FROM THE  IT.  S.  ENVIRONMENTAL  PROTECTION AGENCY'S
                NATIONAL SCHOOL RADON SURVEY (NSRS)
             By: Lisa A.  Ratcliff
                 U.  S. Environmental Protection Agency
                 Office of Radiation Programs
                 401 M Street,  S. W.
                 Washington,  D.  C.   20460

                 Jane Williams  Bergsten
                 Research Triangle Institute
                 P.  O. Box 12194
                 Research Triangle Park,  NC  27709
                              ABSTRACT


     During the winter of 1990-91, a nationwide probability sample
of approximately 1,000 public schools participated in the NSRS.
In approximately 100 of the schools, both long-term and short-term
measurements were made in all schoolrooms, yielding alpha track
and diffusion barrier charcoal canister measurements for about
4,000 schoolrooms.  In abut 900 schools, short-term radon
measurements were made in all schoolrooms having ground contact,
yielding diffusion barrier charcoal canister measurements for
about 28,000 schoolrooms.  National and regional estimates of
radon levels in schools are presented.

     Approximately 78 percent of the nation's public schools have
no rooms with screening radon levels over 4 pCi/L, approximately
11 percent have either 1 or 2 such rooms, and an additional 11
percent have 3 or more rooms over 4 pCi/L.   Approximately 3
percent of the nation's public schools have one or more rooms with
screening levels over 10 pCi/L.

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

-------
                                                                         X-1
Title:  EPA's Revised School Radon Measurement Guidance

Author:  Chris Bayham, U. S. EPA, Office of Radiation Programs

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

-------
                                                                      X-2
                   RADON IN COMMERCIAL BUILDINGS
             By:  Harry Grafton and Ayotunde Oyelakin
                 Columbus Health Department
                 Environmental Health Division
                 181 S.  Washington Blvd.
                 Columbus,  OH  43215
                             ABSTRACT
     Radon screening and followup data reported for 250 Municipal
Buildings are presented.  Data are organized to group buildings by
structural characteristics and by occupant activity.
Prioritization guidelines for large building testing projects are
suggested.  Commercial building radon levels are contrasted to
residential radon levels to demonstrate the significance of radon
testing in the workplace.

-------
                                                                                 X-3
                  IOWA MULTIRESIDENTIAL  BUILDING RADON STUDY

           by:    James w.  Cain
                  Iowa State  University  Energy Extension
                  Ames,  Iowa   50011
                                   ABSTRACT

      This  study tested radon levels in duplex and other multiresidential
ouildings.   Fifty-nine units in twenty-six buildings were tested.   Long-term
(9 month) and short—term (3 day)  tests were used to evaluate radon levels.
Twenty units were tested using both test devices.  The averages for the duplex
units are 1.8 pCi/L for the long-term tests and 2.9 pCi/L for the  short-term
tests.  The averages for the multiresidential buildings are 3.8 pCi/L for the
long-term tests and 4.3 for the short-term tests.  Radon levels above 4 pCi/L
are found on every level of multiresidential buildings and duplexes.  Fifteen
percent of  the second floor units tested,  and twenty percent of the third
floor units, were above 4 pCi/L.   This study finds elevated radon  levels in
all types of buildings and on all the floors, and recommends that  testing be
encouraged  in all dwelling units regardless of the location in a building,  or
the type of building.

-------
                  IOWA MULTIRESIDENTIAL BUILDING RADON STUDY


      A substantial amount of testing has been conducted on single family
homes in Iowa.  A study conducted by the Iowa Radon Project including 7100
dwellings found 70% of the dwellings have tests of 4 pCi/L or greater.1   The
U.S. Environmental Protection Agency-Iowa Department of Public Health survey
of 1381 homes found 71% of dwellings have test results of 4 pCi/L or greater
(1989).  The findings of these two studies are essentially the same.  These
screening surveys indicate a large percentage of Iowa's single family homes
should have additional testing and are potential candidates for mitigation.
Indeed, to date, no Environmental Protection Agency supported state survey has
found as high a percentage of dwellings at or above the action level.

      The Iowa Department of Public Health, the Iowa Radon Project, and other
Iowa organizations have expended considerable effort to convince home owners
to test.  An assessment of public awareness conducted by the Iowa Radon
Project in 1990 indicates some knowledge about radon is possessed by a very
large part of the population.2  Although the portion of the  population which
has tested is small, the message appears to be getting to people.  The
message, that every home should be tested, clearly means single family homes.
This message is based on solid research.  Little research has been conducted
in Iowa on multi-family structures.  Therefore, this study focuses on multi-
residential buildings.

      A target population of fixed income families was chosen as a mechanism
to gain entree into multiresidental buildings.  First, fixed income families
living in duplexes and multi-family buildings were identified.  Selection for
participation was structured to produce a sample that included a variety of
multiresidential building types—2 family, 4 family, 6 family, 12 family, and
larger buildings.

      The study was conducted in two phases.  The first phase is long-term
radon testing conducted with alpha-track detectors.  The testing period was
nine months.  The number of units in the first phase is smaller than the
second phase; not all units were included in phase one.

      The second phase is short-term radon testing.  The short-term tests were
conducted with charcoal canisters with a typical exposure of three days.  The
intent was to include all of the participants from the first phase plus
additional units.

      As the study proceeded, the opportunity arose to include a 48-unit
building.  This building, the Project, was not available for inclusion in the
first phase.  Therefore, no long-term tests were conducted in the building.
During phase two, we were able to conduct short-term tests in 31  (34?) units
of the Project, in the office, and in the common room.
^reiner,  T.H.; Hodges,  L.;  and Cain,  J.  Radon in Iowa.   Paper presented at
 1989 International Winter Meeting, American Society of Agricultural Engi
 neers, New Orleans, Louisiana.  December 12-15, 1989. p. 2.
2Cain,  J.W.,  Radon Public Awareness: A 1990  Survey of Iowa Residents.   Iowa
 Radon Project, Iowa State University Energy Extension, Ames, Iowa, 1991. p.6.

-------
      The
and multi
agencies
Agencies,
groups.
a number
asked to
                 SAMPLE  SELECTION AND  LOSSES

 first task was to identify fixed income families living in duplexes
-family buildings.  This was accomplished by contacting a number of
which provide services to fixed income families: Community Action
 elderly meal sites, housing authorities, and similar community
These service providers were given an explanation of the program and
of return post cards for potential participants.  Each agency was
inform eligible families and give them a post card to send in.
      The  post  card asked the potential participant to provide an address,
indicate the type of building they lived in (mobile home,  duplex, apartment
building,  or single family),  and affirm they were a fixed income household.
Not counting the responses from people living in single family homes and
mobile homes,  52 post cards were returned in time for selection.  From these
responses  a  sample of participants was selected.

      The  original pool included 11 families in duplexes and 31 families in
other multi-family buildings.  Two samples were drawn from the pool.  The
first sample was for the long-term tests (alpha tracks), and the second sample
was for the  short-term tests (charcoal canisters).   The short-term pool
includes all the units in the long-term pool.  Two households were eliminated
because they were not in central Iowa.

      For  phase one a sample of 24 units was desired, so an initial sample of
30 households was selected.  The extra households were included to cover
anticipated  losses.  Some families withdrew during the placement of the detec-
tors, and  additional selections were made to replace those households.  A
variety of building sizes  (sized by number of units) was desired in the
sample. Therefore, it was determined that no more than 25% of the units would
be duplexes.  Households were selected to produce a sample with a structural
profile to match the goal.

      The  short-term sample included all the qualified families, including
everyone in  the long-term sample.  The profile of buildings in the pool
assured that a variety of building sizes are represented in the sample.

      Among  the families that withdrew during the first phase, long-term
testing, were three families that resided in the same building.  These
individuals  indicated management would not permit them to participate.  After
some communication with the management, an agreement was reached to include
the three  units in the short term sample and to test the remainder of the
building.  This building is the Project.


                     TABLE I.   SAMPLE SELECTION AND LOSSES


Duplexes
Multi-Family
Project
TOTALS
LONG TERM SAMPLE
Drawn
9
23

32
Tests
6
15

21
Losses
3
8

11
SHORT TERM SAMPLE
Drawn
10
22

32
Tests
8
19
31*
58
Losses
2
3

5
FINAL

9
19
31*
59
    * Does not include tests conducted in non—living areas

-------
      Sample losses in the long-term phase include families that withdrew and
those households where the testing could not be completed.  Two families
moved, two families threw the alpha track away, and one family lost the detec-
tor.  Sample losses in the short-term phase include the families that could
not be reached to arrange testing.

      Table I summarizes the sample selection and loss information.  Combined,
the final sample includes 9 families in duplexes and 19 families in multi-
family buildings plus the additional units in the Project.


                              TESTING PROCEDURES

ALPHA TRACK DETECTORS

      The placement of the long-term, alpha track, detectors started on May
29, 1990 and continued until June 28, 1990.  The devices were placed by staff
from the Iowa Department of Public Health  (IDPH).  The IDPH staff contacted
each household by phone for an appointment to place the detectors.  The IDPH
staff tried to contact each of the families at least twice while the detectors
were in the homes to remind them about the study and answer potential ques-
tions.  Because of these efforts the number of detectors/families lost was
minimal.

      Iowa Department of Public Health staff began to collect the alpha track
detectors on March 22, 1991 and had completed the collection by March 29,
1991.  The long term tests were exposed for an average 291 days (9.7 months)
over three seasons—summer, fall, and winter.

CHARCOAL CANISTERS

      The placement of the short-term, charcoal detectors, started on November
26, 1990 and was completed by January 14, 1991.  The canisters were placed by
staff from the Iowa Department of Public Health (IDPH).  The IDPH staff
contacted each household by phone to set an appointment.  Those households in
the long-term phase of the study were also contacted for an appointment.

      IDPH staff collected the charcoal canisters—starting on November 30,
1990 and completed the pick-ups by January 18, 1991.  The typical test was
three days in duration.  Except for one over-exposed canister (24 days), the
longest exposure was five days and the shortest was three days.

      The Project residents were notified by management of the day and time
staff would be in the building to place detectors.  Cooperation among the
tenants was very good.  All but thirteen apartments were tested.  Those not
included were not home the morning the canisters were placed.  In addition to
the apartments, tests were conducted in the boiler room, the office,  beauty
shop, and the community room.  There were no long-term tests in the Project.

OTHER TESTING ISSUES

      The schedule for this research was set to produce the best test results
within the grant period.  The long-term tests were placed as soon as possible
in an effort to get as long a test period as possible.   By placing the
canisters in June, the long-term devices were exposed for an average of 9.7
months.  The plan for the short-term detectors was to wait until the heating
season was well underway before placement began.  So short-term testing was
started after Thanksgiving.  The expectation was that all the short-term
testing would be completed before Christmas.  However,  a few short-term tests
were performed in January.

-------
      No attempt was made to select which unit in a building was to be used in
this study.  The tested unit(s)  could be located anywhere in the building,
from basement to top floor.  Placement of detectors was in living areas, not
necessarily the lowest livable area of the structure.  In the case of most
multiresidental buildings, this distinction is not particularly important.
However, four of the duplexes have basements.  The only basement test conduct-
ed in a duplex was where the basement was a living area,  within dwelling
units the test devices were placed in bedrooms or living rooms.

      The research plan included duplicate tests on ten percent of the
dwellings.  Six of the thirty alpha tracks sent to the laboratory for process-
ing are duplicates, 20% of the total.  Nine of the seventy-seven charcoal
canisters sent for processing are duplicates, 11.7% of the total.  The
duplicate detector results were essentially the same as those for the primary
detectors, most variation in results is within three or four tenths of a pico
curie per liter.

      The research plan also called for 5 percent of the tests to be blanks.
Two blank alpha tracks were sent to the laboratory, 6.6% of the total detec-
tors analyzed.  Six blank charcoal canisters were sent for analysis, 7.8% of
the sample.  The results  for the unexposed canisters are as expected: one,
two, and three tenths of  a pico curie per liter.


                                    RESULTS

      Summary results for all testing is presented in Table II.  Twenty-one
units were tested with alpha track detectors.  The highest alpha track result
is 14.4 pCi/L and the lowest is 0.7 pCi/L.  The average is 3.3 pCi/L.1   Nearly
24% of all the units tested above 4 pCi/L.
                      TABLE II.  SUMMARIZED TEST RESULTS


Number of
Units
Average
Result
Highest
Result
Lowest Re-
sult
% Below 4
pCi/L
% Above 4
pCi/L
ALPHA TRACK TESTS
ALL TESTS
21
3.3
14.4
.7
76.2
23.8
CHARCOAL TESTS
ALL TESTS
62
4.1
31.1
.5
62.9
37.1
WITHOUT PROJECT
UNITS
27
4.7
31.1
.5
59.3
40.7
'There  is  a  95%  confidence that  the  mean  for the  alpha  tracks  for these  types
 of dwellings is between 1.4 and 3.8 pCi/L.

-------
      Comparing the results from the short-term charcoal tests with those for
the long-term tests, presented in Table II, shows the short-term tests are
similar although a bit higher.  Short-term tests were conducted on 62 units.
The aggregate average short-term result is 4.1 pCi/L1;  with the  lowest being
.5 pCi/L and the highest 31.1.  About 37% of these units tested above 4 pCi/L.
While these are not screening tests, the difference between these results and
screening test studies of single family homes is substantial.  When all the
extra units in the project are removed from the sample, the percent of
charcoal tests above 4 pCi/L increases.


DUPLEXES

      The results for the tests conducted in the duplexes are presented in
Table III.  Six duplex units completed the alpha track test phase of the
study.  The highest score is a 5 pCi/L and the lowest is .7 pCi/L.  The
average for the six duplexes is 1.8 pCi/L.  All but one of the tests was
conducted on the first floor.  The one test conducted in a basement is the
high reading, the only unit to test above 4 pCi/L.

      Eight duplex units participated in the short-term phase of the study.
The highest is 7 pCi/L and the lowest is .5 pCi/L.  The average is 2.9 pCi/L.
Two units have tests above 4 pCi/L, and one of those is a basement test.
Except for the basement unit, testing was conducted on the first floor of the
duplexes.  Only one alpha track test and two charcoal tests did produce
results above the action level.
                          TABLE III.  DUPLEX RESULTS

Number of Units
Average
High
Low
Above 4 pCi/L
ALPHA TRACK
6
1.8
5.0
0.7
16.7 %
CHARCOAL
8
2.9
7.1
0.5
25.0 %
MULTI-FAMILY BUILDINGS

      Tests were conducted in 17 different multi-family buildings.  Sixteen of
these buildings are discussed here.  The Project had a multiple number of
tests and is discussed separately later.  These buildings had only one unit
tested.  A summary of information for these multi-family buildings is present-
ed in Table IV.  Except for the project, this figure shows the data broken
down according to buildings and the size of the building, in units.

      The long-term and short-term tests were conducted in the same units.  A
wide range of building sizes are represented in the sample.  Apartments to be
tested were not selected by location in the building.  Only in the case of the
four unit and twelve unit buildings are there more than one building of any
size in the sample.
JThere is a 95% confidence that the mean for the charcoal tests for these
 types of dwellings is between 2.6 and  4.4 pCi/L.

-------
      Four different buildings have units with long-term tests above the
Action level.   Two of the buildings are four unit structures.   One is a twelve
unit building and one is a twenty-two unit building.

      Eight buildings have short-term tests above the action level.  Four of
these buildings are four unit buildings.  One is an eight unit building.  One
is a twelve unit building.  One is a twenty-two unit building, one is twenty-
four units.  All of the buildings which had long-term tests above the action
level have short-term tests above the action level.
                  TABLE IV.  MULTI-FAMILY BUILDINGS BY SIZE
SIZE IN UNITS

48
24
22
20
12

8
6
5
4







Totals
# OF
BUILDINGS
1
1
1
1
2

1
1
1
8







17
# OF
UNITS
34
1
1
1
2

1
1
1
8







50
ALPHA
TESTS
*
2.7
14.4
1.5
4.3
0.7
2.1
1.1

2.0
5.5
1.7
2.5
1.1
0.7
3.4
14.0

CHARCOAL
TESTS
*
4.9
31.1
0.8
6.8
1.6
4.5
1.2
1.8
1.2
5.5
4.4
3.7
0.8
1.6
6.1
15.2

              Summarized separately.
      Table V summarizes the results for the Project tests and the rest of the
buildings in the study, the regular sample.  The Project results will be
discussed later.  Table V shows that for the regular sample, both long-term
and short-term tests were taken in 15 of the buildings.  One building had only
a short-term test.  The average for the long-term tests is 3.8 pCi/L, with a
high of 14.4 pCi/L and a low of .7 pCi/L.  The average for the short-term
tests is 5.4 pCi/L, with a high of 31.1 pCi/L and a low of .8 pCi/L.  Most of
the tests, 13 of 19, were conducted in first floor apartments.  Three were in
basement units, one in a second floor unit and two in third floor apartments.
Four of the buildings have basements, the remaining twelve are built slab-on-
grade.

-------
      The charcoal tests results are slightly higher than those for the alpha
track tests, see Table V.  The average of all the alpha track tests is 3.8
pCi/L and the charcoal test average is 5.4 pCi/L.  Ten units had higher
charcoal tests results than alpha track results,  three were lower,  and two
were the same.  The percentage of units above 4 pCi/L is much higher for the
charcoal tests, 42%, than the percentage of alpha track units,  27%.
                       TABLE V.  MULTI-FAMILY BUILDINGS

ALPHA TRACK TESTS
Highest
Lowest
Average
Above 4 pCi/L
CHARCOAL TESTS
Highest
Lowest
Average
Above 4 pCi/L
REGULAR
SAMPLE
15
14.4
0.7
3.8
27 %
19
31.1
0.8
5.4
42 %
PROJECT





35
9.5
0.9
3.7
34 %
TOTAL





54
31.1
0.8
4.3
39 %
THE PROJECT

      The Project is a two and three story building, built in the general
shape of a U.  The two legs of the U are primarily two stories and the
connecting base of the U is three stories.  The building is constructed on a
slab.  The building contains 48 dwelling units, an office, a community room,
the boiler room, and a beauty shop area.  Thirty-six of these spaces were
included in the testing program.  On both the first and second floors 13 areas
were tested.  On the third floor 9 apartments were tested.

      All of the tests on The Project building are short-term tests.  The
tests were all conducted between Monday, December 10 and Thursday December 13,
1990.  Within the project building a variety of results were obtained.  The
range extends from a low of 1.1 pCi/L to a high of 9.5.  The average reading
for the building is 3.7 pCi/L.

      The boiler room test is 3.3 pCi/L.  Nine of the 15 areas tested on the
first floor have readings above 4 pCi/L  (60 % of the areas tested).   First
floor tests range from a low of 1.8 pCi/L to a high of 9.5 pCi/L.

      Thirteen second floor areas in The Project were tested.  The average of
the second floor readings is 2.3 pCi/L.  The lowest second floor reading is .9
pCi/L and the highest is 6.2 pCi/L.  Eight of the test results are below 2

-------
pCi/L (53%).  Two second floor (15 %)  readings are above the action level (4.6
and 6.2 pCi/L).

      The Project's third floor tests range from a low of 2.2 pCi/L to a high
of 6.2 pCi/L.  Six of the nine third floor tests (66%) are between 2 and 3
pCi/L.  Two third floor units (22 %)  have readings above the action level (4.1
and 6.2 pCi/L).   The average of the third floor readings is 3.2 pCi/L.

      The third floor readings are slightly higher than those on the second
floor.  The second floor average is 2.3 pCi/L while the third floor average is
3.2 pCi/L.  The lowest second floor reading is .9 pCi/L while the lowest third
floor reading is 2.2 pCi/L.  Eight second floor readings are below 2 pCi/1,
but no third floor readings are below 2 pCi/L.  On the other hand, the highest
reading on either the second or third floor is the same, 6.2 pCi/L.


RESULTS BY BUILDING LEVEL LOCATION OF UNIT

      One hypothesis expects the radon levels to decline as one goes up from
one level to another.  In this theory radon levels in multi-family buildings
are expected to be lower on upper floors because forced air heating is not
common in such buildings and because apartments are separate units.  However,
this pattern does can occur in single family homes even when heated with a
forced air system.  There are many routes in multi-family buildings through
which radon can migrate from the lowest level to the top floors; elevator
shafts, plumbing and chimney chases,  and electrial wiring chases or cavities.
The test results by floor are summarized in Table VI.  In considering the
radon levels on various floors of buildings all tests are included, including
the Project.

Basements

      Five tested areas are located in basements.  Four of these areas have
both long-term and short-term tests.  On the short-term tests four of the five
areas are above the action level.  One of these apartments tested at 4.9 pCi/L
with the short-term test, but is 2.7 pCi/L on the long-term test, considerably
below the action level.
                          TABLE VI.   RESULTS BY FLOOR
FLOOR
Basement
First
Second
Third
TOTAL
UNITS
4
34
12
10
ALPHA TRACK TESTS
Units
4
16
0
1
% > 4 pCi/L
75.0
12.5

0.0
CHARCOAL TESTS
Units
4
32
12
10
% > 4 pCi/L
100.0
43.8
16.7
20.0
# WITH
BOTH TESTS
3
15
0
1
 First Floors

      The first floor test results can be divided  into those on units in
 buildings constructed directly on the ground, slab-on-grade, and those with

-------
units in buildings with basements and/or crawlspaces.   A total of 35 first
floor areas were tested.  Eight of the 35 first floor test areas are in
duplexes and 27 are in multi-family buildings.

      Long-term tests were conducted in 16 first floor areas (duplexes and
multi-family buildings). The highest of these long-term test results is 14.4
pCi/L and the lowest is .7 pCi/L.  The average long-term reading is 2.6 pCi/L.
Two of the units are above the action level (5.5 and 14.4 pCi/L).  Of the 14
tests below 4 pCi/L ten of them are below 2 pCi/L,  62.5% of all the long-term
first floor tests.

      The short-term tests display a greater variability with a high of 31.1
pCi/L and a low of .5 pCi/L. The average short-term reading for the first
floor units 4.6 pCi/L. There are 15 units  (44%) with short-term test results
above 4 pCi/L and the remaining 19 (56%) are below the action level.

      Six of the first floor tests were taken in buildings with basements and
two with crawlspaces.  Four of the units over basements have long-term tests.
The average is 1.3 pCi/L and none of the long-term results is over the action
level.  The average for the short-term tests on all six of the basement units
is 2.4 pCi/L, one unit exceeded the action level.  The two units built over
crawlspace are similar in their results; neither unit is over the action
level, the long-term test average is 1.4 pCi/L and the short-term test average
is 1.5 pCi/L.

      More than half  (27) of the first floor test areas are in slab-on-grade
buildings.  In 11 cases long-term tests were conducted.  The average long-term
reading is 3.2 pCi/L.  Two of the 11 are above 4 pCi/L.  In 26 cases short-
term tests were conducted with an average of 5.3 pCi/L.  A total of 14 of the
26 (54 %) are above the action level.

Second Floors

      There are no long-term tests on any second floor apartments, but there
are 13 short-term tests.  All of the second floor tests results are from the
Project.  The average short-term test for second floor apartments is 2.3
pCi/L.  Eight  (61.5%) of the test results under 2 pCi/L.  Only two units
(15.4%) have tests above the action level  (4.6 and 6.2 pCi/L).  The highest
test is 6.2 pCi/L and the lowest is  .9 pCi/L.

Third Floors

      There is one long-term test for a third floor unit and ten short-term
tests.  All but one of the third floor test results are from the Project.  The
long-term test  (0.7 pCi/L) is well below the action level.  The average of the
short-term tests is 3.1 pCi/L.  The short-term test results range from a low
of 1.6 pCi/L to a high of 6.2 pCi/L.  Six of the third floor units  (60%) have
readings between 2 and 3 pCi/1; seven  (70%) have readings less than 3 pCi/L.
Two units  (20%) have test results above the action level.
                                   FINDINGS

DUPLEXES

      Data was collected on nine duplex units occupied by fixed income
families.  The duplex  averages are below the action level;  long-term average
1.8 pCi/L, short-term  average 2.9 pCi/L.  One long-term test and two short-
term  tests are above the action level.  Although the sample is small, elevated
radon levels can occur and the percentage of scores above the action level

-------
(16.7% of long-term and 25% of short-term)  suggests the number of duplexes in
the housing stock might be substantial.1  A  recommendation  to  test  dwelling
units in duplexes is prudent.

MULTI-FAMILY STRUCTURES

      The results of testing in multi-family buildings indicates that testing
all units is an_appropriate activity.  Tests were conducted in 17 different
buildings and nine buildings have at least one test above the EPA action level
of 4 pCi/L.  The project building has several apartments with short-term
readings above the action level.  Four of the 15 long-term tests (27%)  are
above the action level, and 19 of the 50 short-term tests  (38%)  are above the
action level.

      The average long-term readings are lower than expected,  but four of the
readings are above the action level.  While 17 buildings is not a large
sample,  many different sizes of buildings are represented.   Readings above the
action level are found in different sizes of buildings.

BY FLOOR

      Long-term tests in basement apartments were above the action level in
three of four long-term tests and four of five short term cases.  Two of 16
(12.5 %) first floor apartments have long-term test results above the action
level and 44 % of the short-term first floor tests are above the action level.
In two second and two third floor apartments the short-term tests are above
the action level.  Although most of the tests that are above the action level
were conducted in basement or first floor apartments, these results clearly
indicate elevated levels can occur on the upper floors of multi-family build-
ings.  Testing of units in multi family buildings, including the second and
third floor apartments, is recommended.

      The data connecting radon levels with basement and crawlspace construc-
tion practices are not conclusive.  Although none of the first floor units
over basements or crawlspaces have readings above the action level, there are
only 8 units in this category.  The results for slab-on-grade buildings are
somewhat skewed by the fact that the Project accounts for more than half of
the readings above the action level.  However, it is clear that with slab-on-
grade construction elevated radon levels are possible.


                                  CONCLUSION


      In summary, this study concludes that the testing of all dwelling units
in duplexes and multi-family structures is advised.  Despite the fact that
multi-family buildings and duplexes seem to have lower frequency of units
above the action level than found in Iowa single-family homes, apartments in
these type of structures can have radon above the 4 pCi/L level.   Testing
should be encouraged in all structures and on all floors.

      The work described in this paper was not funded by the U.S. Environmen-
tal Protection Agency and therefore the contents do not necessarily reflect
the views of the Agency and no official endorsement should be inferred.
 One  ineligible  unit  adjacent  to a  participant  was  tested in  the  short-term
 phase.   In this unit the short-term readings are 15.6 pCi/L in the basement
 and 14.9 pCi/L on the first floor.  The other unit in this building tested at
 3.3 pCi/L on the first floor.

-------
                                  REFERENCES
1.   Greiner,  T.H.,  Hodges,  L., and Cain, J.  Radon in Iowa.  Paper presented
     at 1989 International Winter Meeting, American Society of Agricultural
     Engineers, New Orleans, Louisiana.  December 12-15, 1989.

2.   Cain, J.W. Radon Public Awareness: A 1990 Survey of Iowa Residents.  Iowa
     Radon Project,  Iowa State University Energy Extension, Ames, Iowa, 1991.

-------
                                                                            X-4
Title:  Airflow in Large Buildings

Author: Andrew Persily, U. S. Department of Commerce, Building and
            Fire Research Laboratory

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

-------
                                                                      X-5
         MEETING VENTILATION GUIDELINES WHILE CONTROLLING
                         RADON  IN SCHOOLS
             By:  Gene Fisher and Bryan Ligman
                 U.  S.  Environmental Protection Agency
                 Office of Radiation Programs
                 401 M Street S.  W.
                 Washington,  D.  C.   20460

                 Terry Brennan
                 Camroden Associates,  Inc.
                 RD  #1,  Box 222
                 Oriskany,  NY  13424

                 William Turner
                 H.  L.  Turner Group
                 Concord,  NH  03301

                 Richard Shaughnessy
                 University of Tulsa
                 Tulsa,  OK
                             ABSTRACT
     As part of a continuing radon in schools technology
development effort, EPA's School Evaluation Team has performed
radon mitigation in schools by the method of
ventilation/pressurization control technology.  Ventilation rates
were increased, at a minimum, to meet the American Society of
Heating, Refrigeration and Air Conditioning Engineers  (ASHRAE)
standard, "Ventilation for Acceptable Indoor Air Quality"  (ASHRAE
62-1989).

     This paper presents the results and the preliminary
evaluations which lead to the Team's decision to implement this
technology.  Factors considered include energy penalties, comfort,
indoor air quality (IAQ), building shell tightness, and equipment
costs.  Cost benefits of heat recovery ventilation were also
considered.  Earlier results of the SEP Team's efforts have
indicated a severe ventilation problem within the schools of this
nation.  A holistic approach to radon mitigation in schools and
other large buildings, which controls radon as well as improving
overall IAQ, should be the goal of radon remediation where
practical.

-------
                                                                               X-6
          REDUCTION IN A BELGIAN SCHOOL :  FROM RESEARCH TO APPLICATION
           by  :   P. Cohilis,  P. Wouters and P.  Voordecker
                  Belgian Building Research Institute (CSTC-WTCB)
                  Rue d'Arlon 53/10
                  1040 Brussels, Belgium
                                  ABSTRACT
     Several  kind of measurements were  performed  in  a  school,  situated in
the province of Luxembourg (south of Belgium),  in which radon
concentrations higher than 1000 Bq/m3 were measured. These experiments
included tracer gas techniques, pressurization techniques and CO2 and radon
concentration measurements, in order to define appropriate remedial
actions. As a consequence of the results of these studies, two mechanical
ventilation systems producing an over-pressure in the rooms and improving
the global ventilation were installed and tested in two rooms of the
school. Their use was completely successful in achieving radon reduction.
The possibility of using a sub-slab depressurization system was also
evaluated, and it was shown that this method could also be applied to
reduce radon levels in the school. However, it was decided to consider the
solving of the radon problem in a larger context and to suggest, to the
school authorities, the use of the method based on the pressurization of
the rooms. Indeed, as many Belgian schools have general indoor air quality
problems  (too low air change rates), solutions aiming to solve several
problems at the same time should be considered in first instance.

-------
                                INTRODUCTION


     The average indoor radon concentration in Belgium is almost 50 Bq/in3,
with a clear distinction between, roughly, the north  (mean value :
39 Bq/m3)  and the south (mean value :  11 Bq/m3)  parts of the country,  due
to geological factors  (1).  The highest indoor radon concentrations may be
found in the south :  the available data indicates that a very high
proportion of the buildings expected to have radon concentrations of more
than 400 Bq/in3  (the European action level for existing buildings) are
situated in the province of Luxembourg.

     A number of radon concentration measurements were made in schools of
this province, during the last two years, on the initiative of the
educational responsibles. These measurements allowed the identification of
a few school buildings with an important radon problem  (2).

     The present paper describes the investigations which were performed in
one of these schools in order to define the most appropriate remedial
actions. In this school, the radon concentrations measured in some
classrooms were higher than 1000
             EVALUATION OF THE CHARACTERISTICS OF THE BUILDING
PRESENTATION OF THE BUILDING


     Figure 1 presents a schematic view of the studied school. It is a
modern building of the middle of the eighties. The  rooms are  distributed
among the basement, the ground level and the  first  floor. All the
investigations were performed on the right part of  the building  (see arrows
in figure 1), which is the  zone where high radon concentrations  were
measured. In fact, the occupied rooms presenting a  radon problem were  rooms
CL.l to CL.9, situated at the ground level. All those rooms have a  floor
heating system and small opening windows at the top of fixed  windows.  They
also have large sash-windows usable for intensive ventilation. But  because
of comfort problems these windows  are not frequently used during most  part
of the year.
It is important to note that only  some of the rooms of the ground level are
situated above the basement (rooms CL.l to CL.4 and part of the  central
hall). The other  rooms of this level are constructed directly on the soil
 (ground-supported floor), with a gravel layer (30 cm thick) immediately
below the slabs.
 RADON ENTRY  PATHS  AND TRANSPORT


      The use of tracer gases  (N2O, SFg,  ...) is an attractive way to study
 the  radon  transport  within a  building.  For this purpose,  a  tracer gas  (N2O)
 was  injected in  the  basement  hall and N2O concentrations  were measured,  in
 function of  time,  in all the  rooms of the measure  zone defined  by the
 arrows in  figure  1.  The  MATE-System (Multi-purpose Automated Tracer  gas
 Equipment)  (3)  was used  for the  monitoring.

-------
                                                        J:


                                                                         FIRST FLOOR
                 L r
                          CL. D
       BASEMENT

Figure 1.   Schematic view of the school.  The  arrows
           investigations were performed.
                                                   indicate the zone of the building where the

-------
     These tracer gas measurements allow,  among other things,  to check the
hypothesis following which the basement could be the principal source of
the high radon  (222Rn) concentrations measured in the ground level rooms.
In order to make this check, the evolution with time of the N2O
concentrations measured in the rooms must be compared to radon
concentrations measured during the same periods, in the same rooms.

     Four measurement periods were defined,  corresponding to different
experimental conditions and dates  (see table 1).


TABLE 1.   MEASUREMENT PERIODS AND RELATED CONDITIONS FOR THE TRACER GAS
           EXPERIMENTS
       Period
  Date
(Jul.  Day)
Conditions
                       226-229
                       229-233
                       233-236
                       236-240
                  basement entry
                  room doors
                  room windows

                  basement entry
                  room doors
                  room windows

                  basement entry
                  room doors
                  room windows

                  basement entry
                  room doors
                  room windows
         closed
         open
         closed

         open
         open
         closed

         open
         closed
         closed

         open
         closed
         closed except in
         CL.2 and CL.8
     The results of the tracer gas measurements are given in table 2, where
we present a  "pollution indicator" obtained by dividing, for each room,  the
measured concentration by  the  concentration measured in the injection  room
 (basement hall), and  averaging over  each period. More detailed results may
be found elsewhere  (4).
 TABLE 2.   N2O  POLLUTION  INDICATORS  (X  100) FOR SEVERAL ROOMS, AVERAGED
           OVER EACH MEASUREMENT PERIOD
Room

Basement hall
Ground floor hall
CL.l to CL.4 (mean value)
Period
1
100
2
1.5
2
100
20
19
3
100
42
3
4
100
30
6 (CL.l, 3 and 4) ;
 CL.5  to  CL.9  (mean value)       1      16      5

 First floor (mean value)        2      19     29
                                    0.5  (CL.2)
                                (CL.5, 6, 7 and 9);
                                     0 (CL.8)
                                        18

-------
     Radon concentrations were measured with charcoal canisters exposed
 during 24 hours  in different  rooms (one measurement  per  period).  Table  3
 presents the  results  of these measurements.


 TABLE 3.   RADON CONCENTRATIONS  IN SEVERAL ROOMS  (Bq/m^) DURING THE TRACER
           GAS EXPERIMENTS	

           Room                                Period
	1	2	3	4	

       Basement  hall          1150     1080       990       1160
       Ground floor hall      650       530        870        550
       CL.8                    515       460        440        40
       CL.2                    560       550        160   < background
     The results  of  the  N2O  and radon  concentration measurements  (tables 2
 and 3)  indicate :

    that the closing of the access to the basement has no influence on the
    radon concentrations measured in the rooms but has a net effect on the
    N20 concentrations measured in the same rooms.

 -  that the closing of the doors of the rooms allows  the reduction of the
    radon concentration in room CL.2 (located above the basement)  but not  in
    room CL.8 (located directly on the soil). The conclusion is different
    for N2O : when the doors of the rooms are closed,  the N2O concentrations
    in  the rooms become negligible.

     that the opening of the small windows of rooms CL.2 and CL.8  highly
    reduces the radon concentrations measured in these rooms. The  same
    conclusion comes from the N2O concentration measurements :  the opening
    of  the windows of the rooms  (doors closed !)  is sufficient  to  ensure
    enough ventilation to reduce dramatically the gas  concentrations.

     The above  remarks indicate that the basement does not  act as  a  major
 "source room" for the rooms located at the ground level. The high radon
 concentrations measured in this school  (essentially in the  occupied rooms
 of the ground floor) are related to openings (cracks, joints,  ...) between
 the rooms and the underlying soil.


 AIRTIGHTNESS CHARACTERISTICS OF THE BUILDING


     The degree of permeability to air of the studied zone  of  the  building
 was measured with the pressurization method. For an indoor/outdoor pressure
 difference of 50 Pa, an air change rate of 4.0 h~l (^50 value)  was
 measured.  This value corresponds to a "rather well airtight" building in
 the belgian context.

     The permeability to air of each one of  the  various rooms  is also  an
 important parameter which may have an influence on the choice  of  an
 appropriate remedial action. For example, it will be  vain to try  to apply
 the pressurization method on a very leaky room!

-------
     In order to calculate the  total air-flow  Qtot  across all the walls of
a specific room, for a given value of the inside/outside pressure
difference AP for that room, it is necessary to make a quantitative
evaluation of its airtightness  characteristics. Pressurization measurements
allow us to make such evaluations. In fact,  it is interesting to separate,
for each room, the air-flow QOut across the walls separating the room from
outside the building ("outer walls") from the air-flow Q^n across the walls
separating the room from other rooms in the building ("inner walls") :


       Qtot  =   Qsut   +  Qin
                                                                       U)
             -   cc«   .  |APCUJN""   +   c
                                        lfl
where COut anc* Nout characterize the outer walls,  Cin and Nin characterize
the inner walls, APout is the pressure difference between the room and the
world outside the building and APin is the pressure difference between the
considered room and the remainder part of the building.

     The knowledge of Qtot (and of Qout  and Qin)  for the rooms of the
building is rather important.  Let's take again the example of the
pressurization method : a low value for Qin (comparatively to QOut)» for a
particular room, means that the production of an over-pressure in that room
will have a weak influence on the pressure in an adjoining room.  In other
words, the over-pressure produced in a given room by a ventilation system
will not be significantly affected by what happens in the neighbouring
rooms. By another way, two similar rooms with rather different Qtot values
for the same values for AP may need the use of ventilation
(pressurization) systems with rather different characteristics in order to
produce similar values of AP in those two rooms.

     The experimental method employed to determine Cout,  Nout,  Cin and Nj_n
is based on the use of two variable-flow fans. For each room, one of the
fans controls the pressure difference between the considered room and the
remainder part of the building, and the other one allows the control of the
pressure difference between this remainder part of the building and the
exterior. More details may be found elsewhere (5).
The values for Cout, Nout, Cin and Nin were determined for some rooms of
the school. It was concluded that the values for Cj.n are low comparatively
to the Cout values, for each one of these rooms : the principal
airtightness defaults in rooms CL.l to CL.9 are situated on their outer
walls. These rooms are well separated from the rest of the building when
the doors are closed.
                         POSSIBLE REMEDIAL ACTIONS


     Several radon reduction techniques may be applied to schools (6) .  It
appears, from the above described characteristics of the building, that two
types of remedial actions should be tested in this school-building,
together with the sealing of radon entry routes : the pressurization of the
classrooms and the subslab depressurization. These two techniques have the
same goal  : to reverse the driving pressure difference between the soil and
the house, in order to obtain a higher pressure in the lower part of the

-------
building than underneath the entire slab (to avoid the entering of soil gas
\_. . . ^^-vrs^f^^t* T rt T"l I
by convection).
EVALUATION Or THE CLASSROOM PRESSURIZATION METHOD
Two different ventilation systems were installed in the school, one of them
in room CL.7 (ventilation system called "RENS") and the other one in room
CL.8 (system called "VENT"). These systems are air supply systems, then
producing an over-pressure in the rooms.
The air-flow versus pressure difference curves corresponding to these two
systems were measured in our laboratories, in an experimentation chamber.
They are presented in figures 2 and 3. For each system the measurements
were mace for three ventilator velocities (minimum, medium and maximum) and
the entrance grids were always kept completely open. In figures 2 and 3
those measurements are represented by different symbols on solid lines.

     As the over-pressures produced in the rooms by these systems also
depend on the airtightness characteristics of the  rooms, we also present in
figures 2 and 3 the characteristic curves of the rooms  (dashed lines)
calculated using an expression similar to expression  (!) with the
parameters Cout' ^out' *-in and Nin defined in the  previous paragraph  (and
experimentally determined as explained above). The intersection points
between the dashed lines and the solid lines, in figures 2 and 3, indicate
the air flows and the over-pressures expected for  rooms CL.7 and CL.8 with
the used ventilation systems.
      400
Minimum
Medium
                                        A  Maximum
Room CL.7
     Figure 2.    Characteristic curves for ventilation system RENS  (three
                  ventilator velocities : minimum, medium and maximum) and
                  for  room CL.7  (dashed line).

-------
     1000
Minimum
                                 40           60
                                   DP [Pa]

                           Medium      A Maximum
                                              80
Room CL.8
               100
     Figure  3.     Characteristic  curves  for  ventilation  system VENT  (three
                  ventilator  velocities  : minimum, medium,  maximum)  and for
                  room CL.8  (dashed  line).
     Radon  concentrations  were measured before  and after  the  installation
of the ventilation systems,  in order to evaluate the influence of their
operation en the concentrations  measured in the rooms where they were
installed and also in the adjoining rooms (rooms CL.6 and CL.9,  see
figure 1).
The radon concentrations were measured with charcoal canisters exposed in
the rooms during two to three days.  Four measurement periods were defined,
corresponding to different experimental conditions and dates (see table 4).
Table 5 presents the results of  these measurements. It can be seen that the
two ventilation systems are very efficient in reducing the radon
concentrations in rooms CL.7 and CL.8. This reduction is a result of two
effects  : the increased ventilation of the rooms and the over-pressure
created in these rooms. It is interesting to point out that the
consequences of using these ventilation systems in rooms CL.7 and CL.8 are
different for rooms CL.6 and CL.9.  The radon concentration in room CL.6
seems to be independent of the use of the ventilation system (RENS) in the
adjoining room CL.7.  The situation is very different for room CL.9 which
shows a large increase of its radon concentration when the ventilation
system (VENT) is ON in room CL.8  (note that the values measured in room
CL.5, which is far away from the other rooms, can be, to some extend, used
as normalisation values).  This  is probably due to the effect of the
"pressure barrier" in room CL.8, which is much greater than in room CL.7 as
shown in figures 2 and 3.

-------
TABLE 4.    MEASUREMENT PERIODS AND RELATED CONDITIONS DURING THE EVALUATION
           OF THE PRESSURIZATION METHOD.      	
   Period
Conditions
                    doors  and windows  closed  in  all the  rooms
                    cracks and joints  between the  floor  and the walls
                    sealed in rooms CL.7  and  CL.8
                    doors and windows as  in  period 1
                    cracks and joints in  rooms  CL.7 and CL.8 as in period 1
                    rooms CL.7 and CL.8 :  ventilation systems installed
                                          (but OFF)
                    doors and windows as in period 1
                    cracks and joints in rooms CL.7 and CL.8 as in period 1
                    room CL.7  :  ventilation system  (RENS) ON during all
                                 the period, at maximum velocity
                    room CL.8  :  ventilation system  (VENT) ON during all
                                 the period, at maximum velocity
                    doors and windows as in period 1
                    cracks and joints in rooms CL.7 and CL.8 as in period 1
                    room CL.7  :  ventilation  system (RENS)  ON  during all
                                  the  period,  at medium velocity
                    room CL.8  :  ventilation  system (VENT)  ON  during all
                                  the  period,  at minimum velocity
 TABLE  5.    RADON CONCENTRATIONS  IN Bq/m3  ("-" MEANS  "NO MEASURE")  FOR
            PERIODS DEFINED IN TABLE 4           	
        Room
                                            Period
CL.5
CL.6
CL.7
CL.8
CL.9

-
1 335
450
~
190
2 020
1 390
247
610

-
42
24
^
180
1 830
120
19
1 680
     It can be  said,  in  conclusion, that the use of these ventilation
 systems  was  completely successful,  for both systems.   This  is in part  due
 to the fact  that the air-tightness  of the  rooms  is rather good (as it  was
 shown  by pressurization measurements)  so  it was  rather easy to produce an
 over-pressure of a few Pascals (especially with  system VENT). Ventilation
 systems  similar to one of these two systems,  installed in each room,  could
 be an  efficient radon reduction method in  this  school. Moreover,  their use
 allows the reduction of the high C02  levels which  can  be measured in the

-------
rooms during the occupation periods.  This  is illustrated in figure  4  which
shows,  for one of the rooms,  the evolution with time of the C02
concentration for two experimental conditions  : fan ON and a small  window
open (29 April), and fan OFF and all  windows closed  (30 April).
  CO ^
  §43

  c
  a>
  o
  c
  o
  o
  o
  u
             4 -
                                                       F/sn
                                                       Door
                                Small window
                                                      Window
                                                                C02
                  i
                  0
I
4
I
8
12
16
20
i
0
i
4
i
8
12
16
20
                           10  14  18  22  2   6   10  14  18  22
                    Time  (hour)    29 and 30  april
     Figure 4.    Evolution,  with time,  of  the C02 concentration in a
                 classroom for two experimental conditions.
EVALUATION OF THE SU3SLAB DEPRESSURIZATION METHOD
     The subslab depressurization (SSD)  technique  is an alternative to  the
classroom pressurization method. It is well known  that the probability of
success of the SSD method depends on several factors,  in  particular  on the
permeability below the slab  (7). As some of the rooms  of  the studied
building are constructed directly on the soil,  with a  gravel layer
underneath tr.e slabs, it appeared that it could be interesting to install
and test an SSO system in one of. these rooms.

-------
     The  gravel  layer  below  the  slab of room CL.7 was depressurized with
the help of a variable-flow extraction fan.  A  rectangular opening was
performed through one of the exterior  walls,  in  a  parallel direction  to and
below the slab,  up to the gravel layer. A pipe (100 mm diameter)  was
introduced inside this opening,  in order  to  allow  the  connection of  the
fan. Afterwards, the rectangular aperture was  sealed with a PUR foam.

     To  evaluate the effectiveness of  the  depressurization of  the gravel
layer it is necessary to measure the pressure  at several points of this
layer. For this purpose several regularly-spaced small apertures were
drilled through the exterior walls of  room CL.7  (and also of the adjoining
rooms),  below the slab, up to the gravel. This allows  the measurement,  with
a micro-manometer, of the pressure difference  between several points  of the
gravel layer and the exterior of the building. All these "measurement
points" are connected to the micro-manometer with the help of a selection
system.  Most of the measurement points are located on the perimeter of room
CL.7, so they are very close to the joints between the slab and the walls
of that room. Figure 5 shows the position of all the measurement points
 (numbered from  1 to 12) .
                        -U,
                                                  CL.5
                                   Hall
i— 1
                                             CL.6
 — 2
   3 4
                      CL. 9
                           12
                                 CL. 8
                                                  CL.7
                                   11
                                                          — 5
                                                                FAN
                                                  10 9  8
     Figure 5.    Position  of  the  various  "measurement points"  (numbered
                  from 1  to 12)  for  the evaluation of the SSD technique.

-------
     We have measured, with the apparatus described above, the pressure
differences AP between each measurement  point  (in the gravel layer) and
the outside, for different values of  the  flow Q measured through the fan.
Figure 6 presents the results of these measurements.  It  shows,  for
different Q values, the evolution of  the  pressure difference AP in
function of the number of the measurement point.
        0


       -10

       -20

       -30

       -40

       -50


       -60

       -70
                                                                 11
       _^- 29.6 m3/h
5     4    5     6     7    I     9     10
           Measurement point
. 45 m3/h   _^_ 79 m3/h   _^ 101 m3/h  _^ 124 m3/h
         12
                                 5    6     7     8
                                  Measurement point
                                    10
   11
           148 m3/h
 164m3/h
274 m3/h
     Figure 6.    Pressure  difference between the gravel layer and the
                  outside in  function of the number of the "measurement
                  point"  (see figure 5), for different values of the flow
                  measured  through the fan.

-------
     It  appears that it is rather easy to depressurize the gravel layer
situated below the slab of room CL."7  : AP is of the order of several tens
of Pa,  even for the measurement points which are the most distant from the
extraction fan. The obtained depressions are believed to be sufficient to
prevent the entry of radon from the soil into the room (this is probably
true for most of the values of the flow Q measured across the extraction
fan, but it could be interesting to confirm this by performing some radon
concentration measurements in the room).
The AP values measured at the points  1, 2, 11 and 12 indicate that it  is
very difficult to create large depressions below the adjoining rooms CL.6
and CL.8 with an extraction fan connected to the gravel layer situated
below room CL.7. That means that an extraction point should be necessary
for each room of this school. This is probably due to the fact that the
walls of the rooms penetrate deeply in the soil, thus preventing the
connection between the gravel layers  situated below the slabs of the rooms.

     In  conclusion,  the subslab depressurization technique seems rather
easy to apply with success when, as it is the case for some of the rooms of
this school, a gravel layer is present below the slab. However, one has to
keep in mind that, for the particular case of the studied building, a
suction point will be necessary for each one of the rooms. Moreover, the
method may not be applied to all the  rooms of the ground level.


                         DISCUSSION AND CONCLUSION
     As a consequence of the results of the investigations presented in the
previous paragraphs, the use of the pressurization method was suggested to
the school authorities. It was decided to seal the cracks and joints
between the floors and the walls,  and to install mechanical ventilation
systems  (air supply  !) in rooms CL.l to CL.9. The characteristics of these
systems were defined according to  the indications obtained during the
evaluation of the effectiveness of systems RENS and VENT presented above. A
special attention was given to the comfort aspects : noise, draught
problems, ... .

     All the systems will be connected to an automatic time-switch and will
be used in the daytime, i.e. during a period of time slightly longer than
the occupation period of the rooms. Indeed, radon concentrations
measurements performed with an active detector indicate that it is more
than enough to start the systems a few hours before the arrival of the
occupants.
It is clear that an  over-pressure  will be difficult to obtain in a room in
which the occupants  let the door and the windows open! This is a
disadvantage of the  method  : the need of an active co-operation of the
occupants.

     In addition to their effectiveness in reducing the radon
concentrations in this school, these air-supply systems will reduce the CO2
levels in the rooms. This is an advantage. In fact, the solving of the
radon problem should be seen in a  large context, the final objective being,
in our opinion, to  improve the indoor air quality and climate. From this
point of view, several aspects should be taken into account, together with
the radon levels  :  combustion products levels, odour levels, thermal
comfort,  	 When possible, solutions aiming to solve several problems
at the same time, without creating other problems, should be considered in
first instance.

-------
                             ACKNOWLEDGEMENTS
     This  work  was  partly  financed  by  the  Radiation  Protection Programme of
the Commission of the European Communities.

     The authors  are  indebted  to Messrs. Bossicard,  R.,  Hotton,  M,  and
L'Heureux, D. for assistance during some of the experiments, and to Dr. A.
Poffijn and Dr. F.  Tondeur for fruitful discussions.

     The work described in this paper  was  not  funded by the U.S.
Environmental Protection Agency and therefore the contents do not
necessarily reflect the views of the Agency and no official endorsement
should be inferred.
                                 REFERENCES
     Poffijn,  A.,  Charlet,  J.M.,  Cottens,  E.,  Hallez, S., Vanmarcke, H. and
     Wouters,  P.  Radon in  Belgium :  the current situation and plans for
     the future. In :  Proceedings of  the 1991 International Symposium on
     Radon and Radon Reduction Technologies.  Philadelphia, USA.  1991.
     Vol. 5,  VI, p. 7.


     Poffijn,  A.,  Uyttenhove,  J., Drouget,  B. and Tondeur, F.  The radon
     problem in schools and public buildings in Belgium, submitted to
     Radiation Protection Dosimetry as part of the Fifth International
     Symposium on the Natural Radiation Environment.  Salzburg, Austria.
     September 22-28,  1991.
     Roulet, C.-A. and Vandaele, L.  Airflow patterns within buildings :
     measurement techniques.  IEA, Technical Note AIVC 34.  December 1991.
     p. 111-72.
     Cohilis, P., Kouters, P., Verheyden, J., Vandaele, L., Bossicard, R.,
     L'Heureux, D. and Voordecker, P.  A case-study concerning radon
     problems in schools  : Libramont - Belgium.  CSTC-WTCB internal report,
     November 1990.
5.   Cohilis, P., Wouters, P. and Voordecker, P.  Intervention du CSTC a
     Libramont  : mesures de pressurisation et evaluation de systemes de
     ventilation dans le cadre de la problematique du radon.  CSCT-WTCB
     internal report.  February 1991.
     Saum, D., Craig, A.B. and Leovic, K.  Radon mitigation in schools.
     ASHRAE journal.  January-February 1990.
 7.   Cohilis, P., Wouters, P. and L'Heureux, D.  Prediction of the
     performance of various strategies of subfloor ventilation as remedial
     action  for  radon problems.  In  : Proceedings of the llth AIVC
     Conference.  Belgirate, Italy.  September 1990.  Vol. 2, p. 17.

-------
                                                                   X-7
   MULTIPLE MITIGATION APPROACHES APPLIED TO A SCHOOL WITH LOW
                        PERMEABILITY SOIL

                               by

                         D. Bruce Harris
                           USEPA, AEERL
                             RTP, NC

                               and

                Eugene Moreau and Robert Stilwell
               Maine  Department  of  Human Services
                  Division  of  Health Engineering
                           Augusta, ME

     A series  of mitigation  approaches have  been applied  to a
amiti-wing elementary school  with  low permeability soil.  Each of
the three wings have been mitigated using a different technique. An
active soil depressurization (ASD)  system deployed  in one wing was
used to compare the performance of  inline axial flow fans and high
vacuum  blowers.   A  high  vacuum blower ASD  system  and  a  room
pressurzation system using unit  ventilators were compared in the
second wing. The  third wing used a heat recovery ventilator. The
results of long term monitoring of each system will be discussed.

-------
                                                                      X-8
                GENERAL  INDOOR  AIR  INVESTIGATIONS  IN
                 SCHOOLS WITH ELEVATED  RADON  LEVELS
             By: Terry Brennan
                 Camroden Associates,  Inc.
                 RD #1,  Box 222
                 Oriskany, NY  13424

                 Gene Fisher and Bryan Ligman
                 U. S. Environmental Protection Agency
                 Office of Radiation Programs
                 401 M Street S. W.
                 Washington,  D. C.   20460

                 Richard Shaughnessy
                 University of Tulsa
                 Tulsa,  OK

                 William Turner and Fred McKnight
                 H. L. Turner Group
                 Concord,  NH  03301
                              ABSTRACT
     As part of the EPA School Evaluation Program, several public
schools were visited.  The schools were being investigated because
of elevated indoor radon levels.  A variety of additional indoor
air measurements and observations were made in the schools.  The
goals were to characterize important factors related to general
indoor air quality and to assess the impact of radon control
methods on air quality problems other than radon.  Schools were
inspected and measurements were made.  Potential sources of
airborne contaminants and conditions that increase the risk of
indoor air problems were cataloged.  Measurements of peak room air
carbon dioxide, carbon monoxide, bioaerosols, organic compounds,
and respirable particulates were made.  Methodologies for carpet
sampling for fungi and bacteria were developed and tested.

-------
                                                                    X-9
  COMPARISON OF ASP AND HVAC SYSTEM CONTROL IN SCHOOL BUILDINGS

          by:  Bobby E. Pyle
               Southern Research Institute
               Birmingham,  Alabama  35255

               Kelly W. Leovic,  Timothy M. Dyess,
               and D. Bruce Harris
               Air and Energy Engineering Research Laboratory
               U.S. Environmental Protection Agency
               Research Triangle Park, North Carolina  27711

                            ABSTRACT

     Active soil depressurization (ASD) and heating, ventilating,
and air-conditioning (HVAC)  system control are the two most widely
applicable  radon  mitigation techniques  for school  buildings.
School officials  often need to select which  technique is better
suited for a particular building;  however,  the two techniques have
not been compared previously in the same building.

     This paper presents direct  comparisons of the radon reduction
capabilities of ASD and HVAC control in  two schools.  The primary
HVAC systems for  both  schools are unit ventilators, although one
school also has  individual  room exhausts  and  the  other  has an
office area with a central  air handling unit.   The schools are
located  in  South Dakota and  Minnesota,  and  data  were collected
using continuous dataloggers.  In the  South Dakota school the unit
ventilators  provided only  marginal radon reduction while the ASD
system consistently maintained radon levels below 4  pCi/L.   In the
Minnesota school,  although both  systems maintained radon  levels
below  2  pCi/L, the  HVAC systems provided slightly better radon
control than ASD when  sufficient  outdoor air  was supplied.

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

-------
                          INTRODUCTION

     Since 1988 the U.S. Environmental Protection Agency's (EPA's)
Air and Energy Engineering Research Laboratory (AEERL) in Research
Triangle Park, NC,  has conducted various levels of radon mitigation
research in 50  schools located  in 13  states.  Mitigation systems
have been installed in 24 of the schools: 14 installations are ASD
systems; 4 are HVAC system control; and 6 are a combination of both
ASD and HVAC control (1).

     Initially,  AEERL's  radon  mitigation  research  in  schools
focussed on applying the most successful technique developed for
residential  houses  — ASD.    Because  of  complicated  subslab
structures  and subslab  fill material  that  sometimes make  ASD
systems expensive  to  install,  and because of indoor  air quality
concerns  (such as  carbon  dioxide),   radon  reduction  using  HVAC
systems  has  also  been researched.    In more  recent  projects,
continuous datalogging equipment has been used to monitor schools
with  HVAC  systems  —  either  central  HVAC  systems  or  unit
ventilators (UVs) 	 to identify the conditions under which HVAC
systems  can be used  to  control  the radon  levels.    This  paper
presents results from  two  schools in which ASD  systems  and HVAC
system control were compared in  the same building.   The schools are
located in  South Dakota and Minnesota.   Those more familiar with
metric units should refer to the conversion factors listed in Table
1.
SOUTH DAKOTA SCHOOL

     Initial radon measurements in the Rapid City area schools were
made  from  November  1990  to  February  1991 using  alpha  track
detectors  (ATDs).    Based  on  these  results, eight  schools  were
identified  by  AEERL as  potential  research schools.    The school
ultimately  were selected for this  study,  Lincoln Elementary,  had
the  highest average  radon  level,  18.8  pCi/L.   This  school  is
located  in  western Rapid  City  on a  hillside  of  exposed  and
crumbling shale.

Building Description

     The  original  building  was  constructed  in 1951  with  10
classrooms, a cafeteria/gymnasium,  and several offices and special
purpose rooms.  The original  building had approximately 14,280 ft2.
Three classrooms and a library were added to the west side of the
original building in 1957,  increasing the total area to 22,132 ft2.
The plan of the building is  shown in Figure  1.

     The design drawings indicate  that the original building has
be low-grade, poured concrete, foundation walls around the perimeter
of the building and under  the corridor walls but not between the

-------
classrooms.  The slabs are poured on compressed soil covered with
a minimum of 6 inches of gravel.  The addition, Rooms 17-20, has a
subslab utility tunnel along the outside walls.  However, because
of the lower radon levels in the addition, it was not included in
this investigation.

     Floor/wall cracks were observed in many of the classrooms and
are  thought   to  be   a  major  contributor  to   radon  entry.
Unfortunately cabinets and closets  typically surround three of the
four walls  in  each classroom (the  fourth wall is the chalkboard)
making access  to and sealing of the floor/wall cracks difficult.
One  crack  observed  in  Room 16  — a  corner  room with  both the
highest ATD (36.7 pCi/L) and followup measurements (35.9 and 14.1
pCi/L)  — was about  0.5  inch wide.

HVAC System Description  and Initial Measurements

     Each  classroom  has a UV  for  heating and ventilating.   The
units are located along  the outside wall.  Most of the classrooms
also have  a fan-powered exhaust located  in the  closet along the
wall parallel to the corridor.

     On May 1,  1991,  carbon dioxide  (CO2) measurements were made in
most of  the rooms.   The measurements, in most  cases,  were made
while the rooms were  occupied and the hall doors were open.  The CO2
concentrations ranged from 700  TO 4000 parts per million (ppm) with
an average  value of  1687  ppm.  Because  the  American  Society of
Heating, Refrigerating,  and Air-Conditioning  Engineers (ASHRAE)
recommends  CO2 levels below  1000 ppm the researchers recommended
that school personnel make the necessary repairs to bring the UVs
up  to   design  specifications.     This  was  before  continuous
datalogging began.

     Measurements of UV airflows and differential pressures across
the building shell were  made in July 1991.  The measurements were
taken under various building operating conditions in all classrooms
in the original building:

     • UV speed  (high, medium, and  low)
     • UV outdoor air damper (completely open and completely
       closed)
     • closet exhaust systems  (on and  off, each room)
     • room-to-corridor  door position  (open and closed)

     For these spot  measurements the  greatest negative pressure
differentials across the building shell occurred when the hallway
door was closed  and  the exhaust fan on.  With the UV outdoor air
damper  open  and the exhaust fan  operating,  the  differential
pressure for the classrooms averaged -0.022 inch water column  (WC) .
The  differential pressure averaged -0.032 inch  WC with  the UV
outdoor air damper closed.

-------
     Operation of  the exhaust  fans  (one  for each  room)  had  a
dramatic effect on the pressures in the room and tended to override
any pressurizing  abilities of  the UVs.  The total exhaust measured
for the 10 classroom fans was 5,554 cfm while the maximum outdoor
air capability of  the UVs  totaled 1,590  cfm — 3.5 times  more
exhaust than  supply.   The exhaust  fans  are  new and draw  630  to
1,021 cfm per room,  compared to the old fans that exhausted about
100 cfm per room.

     Opening the  hallway doors tended to have a neutralizing effect
on the classroom differential pressures.   The negative  pressures
caused by exhaust fan  operation were near zero  when  the hallway
doors were open.

     The UVs were able to pressurize  the rooms  on average, 0.003
inch WC with the exhaust fan was off and the hallway door closed.
A positive pressure was  also observed when the  UVs were on,  the
exhaust fan was off, and the hallway door was open.

Pre—Mitioation Radon Levels

     Two sets  of radon measurements were made in  May 1991 using
charcoal canisters provided by AEERL.  These results are shown in
Figure 1, together with the initial ATD measurements.   The values
circled are the  Winter 1990-91  ATD results;  those in parentheses
were made with the building closed and all UVs off over the weekend
of May 3-5, 1991; the final set of measurements was made with the
building closed  and the UVs operating over the weekend of May 17-
19, 1991.  The average school radon level with the UVs  off was 10.9
pCi/L.   The  highest level, 35.9 pCi/L,  was  measured  in Room 16.
The lowest level, 1.3  pCi/L,  was measured in  Room 9.  With the UVs
on, the school average radon level dropped to 5.7  pCi/L with a high
of 14.1 pCi/L measured in Room  16  and a  low of 0.6 pCi/L in Room
12.

     The relative changes in radon levels with the UVs on compared
to UVs off are seen  in Figure  2.  In half of the rooms  (Rooms 1, 3,
11, 13, 16, Office,  Gym,  Principal's Office, Teacher's Lounge, and
Health Room) the radon levels dropped with the UVs on compared to
the  levels  measured with the UVs  off.   This would  be expected
because of the outdoor air provided by the UVs.  In  the other rooms
the levels increased with the UVs operating  (Rooms 20, 9,  18, and
the Boiler Room  showed the largest increases).   The increases in
these classroom  radon  levels are thought to  be  due to closed or
inoperable outdoor air intake dampers in  those rooms.(As  mentioned
previously,  the UVs  were   repaired  prior  to  the  continuous
datalogging.)  The higher radon level  in the  Boiler Room might be
due to increased depressurization with the UVs operating.

     The outdoor air temperatures were significantly different for
the  two  measurement periods  so these  data should be interpreted
cautiously.   Over the  period May 3-5, 1991 (UVs off)  the  average

-------
high and low temperatures were 46.7 and 31.7°F, respectively.  Over
the period May 17-19, 1991 (UVs on)  these average temperatures were
67.3 and 47.3°F.  These large temperature differentials during the
two charcoal canister measurements  make comparison of the relative
influence of the UVs difficult.   The  influence of the UVs on radon
levels will be discussed later in this paper based on the results
from the continuous datalogger.

Subslab Pressure Field Extension (PFE) Measurements

     PFE  measurements  were made by  AEERL and  Southern Research
Institute.   The subslab PFE was found  to be much  better in the
north wing than in the east wing.   ASD systems were installed as
shown in  Figure  1.   Because of the different PFE measured in the
two wings  of  the building,  the systems  for the two wings must be
different.

     The  PFE  in the north wing was  better than that in the east
wing.  As indicated in Figure 1, one suction point was placed on
each side of  the corridor,  because subslab footings separate the
two areas.  One point was placed  in Room 14 (to treat Rooms 12, 16,
the Teacher's Lounge,  and the Health Clinic),  and one point was
placed  in Room 11  (to treat Rooms 9,  13,  and the Office  areas).
Both systems  use  8  inch diameter,  Schedule 40 polyvinyl  chloride
 (PVC) piping for both  vertical  and horizontal  pipe runs.  Each
suction point uses  a fan rated at  approximately  300  cfm at 1 inch
WC.  The  fans are located on the roof, away  from any air  intakes.

     Because  of the  poorer PFE measured during  the  diagnostic
testing  in the east wing, one  suction point  was  installed in each
of the  four rooms.  The overhead piping for  Rooms  1  and 3 and for
Rooms 2  and 4 were  manifolded  together so  that  only  two fans were
needed.   In Rooms 1 and 3, the piping was run from the rooms across
the hall  above the ceiling tile and up through the chase above the
closets  in Room 4.  The pipes  were joined  and  exited the  building
through an old window  opening (now closed  with plywood sheeting).
The piping from Room 2 was  run across the  bathrooms  and joined to
the piping from the suction point in Room 4.  A single pipe was run
 around  the room to  exit through the  old window openings above the
 closet.

     The fan flow rate was  expected  to be  much lower than that  in
 the north wing,  because of  the lower PFE.  A fan capable of a high
 pressure head was also anticipated.   Due  to  the low flow  rate and
 high pressure expected in these east wing  systems,  the piping was
 much smaller in diameter  than the  systems  in the north wing.  All
 piping  in this wing was 2 inch diameter Schedule 40 PVC.  The fans
 used in this  wing were capable of moving roughly 25 cfm at a static
 pressure of  around 30-35 inch WC  and 50  cfm at 0  inch WC static
 pressure.

-------
Datalogger Installation

     A computer based datalogger monitoring system was installed in
the school  in July 1991.   Remote monitoring  of the  school  was
carried out  via a  telephone modem.  Data were  downloaded to  a
personal computer  in  Birmingham, Alabama  for  data analysis  and
storage. This system monitored the following parameters  every 30
minutes:

     •    Ten continuous radon monitors measured the radon levels
          in Rooms 1,  2, 3,  4, 9, 11, 12,  13,  14, and 16.

     •    Nine  pressure  transducers   monitored    differential
          pressures between both the rooms and the subs lab areas in
          Rooms 1,  4,  11,  and 14.   These  pressure differentials
          were referenced to the pressure in the hallway  which in
          turn was monitored relative to  outdoor pressure.

     •    The temperatures in Rooms 1,  4, 11, 16, the hallway,  and
          outdoors were monitored.

      •   The operation of the UVs in  Rooms 1,  4,  11,  and 14  was
          monitored.   This  included  the  on/off  times  and  the
          positions of the outdoor air dampers.

     •    The on/off  times  of  the  closet exhaust fans   and  the
          position of the doors  (open  or closed)  for Rooms 1, 4,
          11, and 16 were also monitored.

     •    A weather station was used to continuously monitor wind
          speed and direction and precipitation.

Results

     The test matrix  used to evaluate both  the UVs and  the  ASD
system's effects on radon is  shown in  Table 2.   Testing  occurred
November 18, 1991, through January 31,  1992.   The results of this
series of tests are summarized in Figures 3, 4, 5, and 6 for Rooms
1, 4,  13,  and 16,  respectively.   In  each of these figures  the
results are shown first (left) with the ASD off to evaluate the UV
alone and second  (right) with the ASD  operating to evaluate both
ASD alone and ASD in conjunction  with  the  UVs.   The results with
the ASD system off are discussed first,  followed by  a discussion of
the results with the ASD system on.

     Room 1  was  the only classroom  with  a door directly  to  the
outdoors.   In Figure 3 it is quite evident that opening the outside
door was very effective  in lowering  room radon levels.  However,
this approach is not recommended as a  practical year-round approach
to radon control. It is interesting to note that opening  the hall
door also lowered the radon  levels for all modes of operation of
the UV and the exhaust fan.   In  fact, the radon levels in all four

-------
rooms decreased when the hall door was  open.  This was most likely
due  to  increased  ventilation  due to  dilution  of  the  radon
concentrations together with pressure neutralization, reducing the
driving force.

     With the ASD systems off  (left-hand graphs in Figures  3, 4, 5,
and  6)  operation  of  the  closet  exhaust  fans  without  the  UVs
resulted generally  in  either  lowering  the  radon levels or having
little effect.  One exception is seen in Figure 4 where operation
of the  exhaust fan  in Room 4 with the  hall door  open actually
increased  the  radon  levels.    Other research has  shown  that
operation of  exhaust fans  can either  decrease  or  increase radon
levels, depending on  the  relative leakage  between above grade
(decreases radon  levels through infiltration of outdoor  air)  or
below grade (increases radon levels through infiltration of radon-
containing soil gas).

     Operation of the  UVs  (with  ASD and  exhaust fan off)  reduced
the radon levels in Rooms 1, 4, and 16 by about 10%.  However, UV
operation increased  the levels in  Room  13 by approximately 25%.
These results are not consistent with the  charcoal canister results
obtained in  May 1991  (shown  in  Figure 2)  .   However,  during the
period  that  these  continuous  radon  measurements  were  taken
(November 1991 to January  1992)  the average  outdoor high and low
temperatures were 44.2 and 19.9°F,  considerably  lower than in May,
1991 resulting in less  outdoor air supplied through the UV dampers.
It  is possible  that the  UV in  Room  13  pulled radon from the
floor/wall crack, distributing it to  the room and increasing the
radon levels.

     With the ASD  systems operating  (the right-hand  graphs  in
Figures 3,  4, 5,  and 6), the average radon levels in all classrooms
dropped to  less than  3  pCi/L and  in  some cases to less than 1
pCi/L.  The percent reductions ranged  from  about 80 to 99% with an
average reduction of about 90%.   These reductions appear to be
fairly independent of the mode of operation of the UVs and exhaust
fans, indicating that ASD and HVAC together did  not outperform ASD
alone.  Thus, while the UV  and exhaust  fan  operation may lower the
radon  levels  in  many  of  the  classrooms, the levels were not
consistently  reduced to below 4  pCi/L.   Only the ASD  systems
effectively and reliably maintained radon  levels below 4 pCi/L.


MINNESOTA SCHOOL

     Nokomis  Elementary School  was initially  researched  by the
Midwestern Universities Radon Consortium (MURC) and used in their
May  1991  training course.    Because the school  presented a good
opportunity  to  compare ASD and HVAC control  in the same school,
AEERL installed  a continuous  datalogger  in the school  in January
1992 to monitor the  systems installed by MURC.

-------
Building Description and Pre-Mitiaation Radon Levels

     Nokomis Elementary is a single story concrete masonry building
with  brick  veneer  and slab-on-grade  foundation  located  in  the
eastern part of St.  Paul.  The total floor area,  as shown in Figure
7, is about 15,000 ft2.  The average radon level in the school was
4.2 pCi/L measured  with charcoal  canisters  in October 1990.   The
highest  and  lowest  concentrations  were  6.4  and  2.8  pCi/L,
respectively.     Subslab   material  is  packed  sand,   and   the
configuration of the subslab footings is unknown.  PFE measurements
indicated that  a pressure  field could be extended  about  15  feet
from the suction point.

     The  heating and  ventilating systems  in the  school are  a
combination of systems. Hot water for these  systems  is provided by
a  gas/oil  fired boiler located in the  northwest  corner  of  the
building.  Each of the nine classrooms (Rooms 101,  102, 103,  104,
105, 106, 107,  108,  and 109) are heated with  finned-tubes along the
outside walls.  In addition, the two classrooms on the south end of
the building (Rooms 107 and 108) have hydronic slab heating (these
classrooms are used  as  kindergarten rooms).  Heating  and cooling in
all of the classrooms is provided by UVs located along the exterior
walls.   The gym/lunchroom, kitchen,  and the  office  complex  are
serviced by a central air handling unit (AHU) with ducted returns.
The unit has provisions for  outdoor  air  intake  at  roof level.
There are fan powered exhausts in the restrooms and in the kitchen.


     A building investigation  conducted by MURC indicated that the
outdoor air  intake  capabilities of the UVs  and AHU unit exceeded
the exhaust fan capacities so that building pressurization appeared
to be a viable option for radon control.  A fan door test conducted
on the building indicated an effective leakage area  of around 500
in.2.   This  indicates that  about 3000 cfm of outdoor air is needed
to pressurize the building adequately to prevent radon entry when
the exhaust fans are on. PFE measurements indicated that a pressure
field could be extended about 15 feet from the suction point.

Implementation of Mitigation Systems

     In May 1991 a private  contractor  of  MURC mitigated the school
using three mitigation systems.   One system  used the UV's in Rooms
101 to 106 to bring in  the  minimum amount of outdoor air necessary
to  pressurize the  rooms and  to  reduce  the radon  entry.  During
normal  operation the  outdoor  air dampers  modulate  above  this
minimum setting  as  determined by indoor  and outdoor temperatures.
This was also  attempted in Rooms  107  and 108; however, the radon
levels  were  not   reduced  below  4  pCi/L  during  the  initial
evaluation.  An ASD system was  then installed in  these two rooms
with  a  single suction  pit  installed in  each  of the two closets.
The pipes from each point were connected and run to a single, roof-


                                8

-------
mounted, exhaust  fan  capable  of moving 420 cfm at  0  inch  we and
about 180 cfm at 1.6 inch WC.  The  locations of the ASD points are
shown in Figure 7.  Although the Gym/Office area in the north end
of the  building has  a HVAC system  with  an outdoor  air intake,
mitigation was accomplished through a combination of HVAC control
and another single-point ASD system.  The suction point was located
in the Boiler Room very close to the Utility Tunnel.  This location
resulted in bypassing the subslab area by pulling air through the
tunnel block  walls  instead  of from the subslab.   Post-mitigation
radon levels with all systems operating averaged below 0.5 pCi/L.

Datalogger Installation

     In January 1992  a datalogger  was installed  in the school in
order to determine  how the various  mitigation  systems performed
relative to each  other and  the  interactions  between each pair of
systems.   Parameters  monitored  were: radon levels  in Rooms 106,
107,  108,   and the Principal's  office;    subslab  differential
pressure measurements  in Rooms  106,  107,  and the  Principal's
office; outdoor air damper  positions on the  UVs  in Rooms 106 and
107 and  on the office  AHU;  temperatures  in Rooms  106,  107, the
Principal's office,  and outdoors; and wind speed and direction.

Results

     The test matrix  shown in Table  3 was developed to evaluate
each of the mitigation systems.  School Officials were  unwilling to
turn the ASD  systems off while the building was occupied (to test
the HVAC systems alone) .   As a result, testing with  the ASD systems
off was done  over the spring  break,  April 10 to 20, 1992.   These
tests should  be representative  of  winter  conditions:  the outdoor
temperatures during  this testing period ranged from  18 to 43°F, with
an average of  30°F.   The testing schedule  was:

     Test if UVs and AHU at normal  damper  operation; ASD  on:  This
     test  was used to  measure the  radon levels  in the post-
     mitigation configuration with the ASD and HVAC systems all
     operating.   Under  normal  conditions  the  outdoor  air  (OA)
     damper modulated based on OA temperature.   This  is typically
     20 to 25% damper open.

     Test  2,  UVs,   AHU,  ASD  off:    To collect  baseline   (no
     mitigation) radon data, all UVs and  the AHU were turned off.
     The ASD  fan outlets on  the roof were  covered with plastic
     garbage  bags and sealed  with  duct tape on Friday afternoon,
     April 10, 1992. The  ASD systems remained off through Test 6.

     Test  3,  UVs  and AHU on, damper  closed:   On  Monday morning,
     April 13, 1992, the UVs in Rooms 101  to  108 and the  AHU were
     turned on but  the outdoor  air dampers were kept closed  (UVs
     and AHU  running  24  hours).   The school was operated in this

-------
     configuration until Wednesday morning,  April 15,  1992.

     Test 3A, UVs  and AHU on,  dampers  closed,  filters out:   On
     Tuesday afternoon,  April  14,  1992,  it  was  found that  the
return    air filters in the  UVs and the AHU were extremely dirty.
The  filters were pulled out of the units and the systems continued
     to operate as described in Test 3,  above.

     Test 4,  UVs  and  AHU  at  10% damper  open:   On  Wednesday
     morning, April 15,  1992,  the outdoor air vents were opened to
     10%  outdoor air on both  the  UVs  and the  AHU  (running  24
     hours).   The  ASD  systems remained off.    The  school  was
     operated in this configuration until Friday morning, April 17,
     1992.

     Test 5, UVs and AHU at  50% damper  open:   On Friday morning,
     April  17,   1992,   the   outdoor  air  vents  were  opened  to
     approximately 50% outdoor  air on both the UVs and the office
     HVAC system  (UVs  and HVAC running 24 hrs) .   The school was
     operated in this configuration until Saturday afternoon, April
     18, 1992.

     Test 6,  UVs  and  AHU at normal damper  open:    On Saturday
     afternoon, April 18, 199.2,  the UVs the office HVAC system were
     changed to normal day/night operation (night setback at 68°F) .
     The ASD systems remained off with exhaust pipes covered.

     Test 7, UVs and AHU at  normal damper operation;  ASD on:  On
     Monday afternoon,  April  20,  1992,  the  ASD system pipes were
     uncovered and  fans turned  on.   The UVs and AHUs returned to
     the post-mitigation configuration.

     Test 8, same as Test 7  (2  weeks  later).

       The  amount  of outdoor air introduced into each of the UVs
and the AHU as a function of  damper position  is shown in Table 4.
Note that in Table  4,  the percent damper opening corresponds to
different percentages of outdoor air for each of the four locations
monitored.   For  example, a  10% damper opening corresponds to 27,
38,  34,  and  19% outdoor air for Rooms  106, 107, 108,  and the
Office,  respectively.    For  simplicity,  the remainder  of this
discussion will refer to the  percent damper opening rather than the
percent outdoor air.  Refer to Table 4 for the specific quantity of
outdoor air for  a  given location and  damper position.

     The  results of the nine tests are shown in  Figure 8.  The
average radon levels for the  Principal's Office, Room 106,  and Room
107  are plotted for each of the  test  conditions.   Radon  levels
increase  from Test  1 to 2 because of stack effect induced  radon
entry.  With the ASD systems operating  (Test  1), the  stack  effect
is reversed.  The operation of the UVs with outdoor  air  also helps


                                10

-------
to reduce the stack effect.

     The results shown under Test 3 and 3A are conditions in which
the UVs and AHU are depressurizing the building because the outdoor
air  dampers   are  completely  closed.    In  addition to  building
depressurization,  it  is  also  possible  that   the  radon  levels
increased from Test  2  to Tests  3 and 3A because background radon
levels continued to build up in  the building after turning off the
ASD systems  (Test 2) .

     With minimum outdoor air (approximately 10%, damper opening in
Table 4) the  radon levels were reduced below 4 pCi/L  during Test 4.
With  about  50%  damper  opening  (Test 5) ,  the  radon  levels  are
reduced to background  levels.  In fact,  the levels  during Test 5
were lower than normal UV and AHU operation together with the ASD
system  operation.    In Test  6  the  UVs  and  the AHU  outdoor  air
dampers   were  returned   to  the  post-mitigation  setting   of
approximately 20 to 25% outdoor  air.  Turning on the  ASD systems in
Test 7  reduced  the levels in the Office and in Room 107 but had
little effect in Room 106 since  there  is no ASD  system in  Room 106.

     Test 8 presents continuous data collected  about 14 days after
Test  7.   Radon  levels  in  the  two  areas  with ASD  systems  are
relatively consistent  with those observed during Tests  1 and 7.
The radon levels in Room 106  (without the ASD system) are slightly
lower than in Tests  1  and 7.

     The  lowest radon levels in these  three  rooms were observed
during  Test  5 when  the  UVs  and AHU  were  operated  at 50% damper
position  (refer to Table  4 for actual percent outdoor air)  and the
ASD  systems were  off.    However,   it  is often   difficult  to
consistently  depend  on  50%   damper  position   when   outdoor
temperatures drop below freezing. The UVs do maintain radon levels
below  4 pCi/L in all three rooms as long as the damper position is
set  to 10% outdoor air (Test 4).

     Results show that radon  levels are consistently reduced below
1 pCi/L in Rooms  107 and  the  Office when the ASD systems  for these
areas  are operating.   During these tests,  the  radon levels in the
rooms  with the ASD systems were  always lower than in the  room with
the  only UV control  (Room  106) .   The  effectiveness  of the ASD
system alone is currently being evaluated.
                                 11

-------
                           CONCLUSIONS

     In  Lincoln,  the  UVs alone  were not  able to  consistently
maintain radon  levels  below 4 pCi/L.   The exhaust  fans  and the
opening of the  classroom-to-hallway  doors  had varying effects on
radon levels.  The ASD systems, however, provided excellent radon
control, even  better than  that  predicted during  the diagnostic
tests.  Operation of the UVs together with the ASD system did not
provide much additional radon reduction over the ASD system alone.

     The UVs were effective in reducing the radon levels to below
4 pCi/L in  Nokomis if a minimum amount of outdoor air was supplied.
If the damper position  was  set at 50%, then average radon levels
were reduced below l pCi/L.  Radon levels were  consistently reduced
below 1  pCi/L  in Rooms  107  and  the Office with the ASD systems
operating and the HVAC  system set to normal outdoor air.  During
these tests, the  radon levels in the  rooms with the ASD systems
were always lower than  in the room with the only UV control.  The
effectiveness of the ASD system alone is currently being evaluated.

     If the PFE measurements indicate that  a properly designed ASD
system will be  effective,  this would be the preferred system for
consistent radon  control.   If in addition, improvement in indoor
air quality or  further  radon reduction is desired, the amount of
outdoor air supplied through the HVAC  system  should be increased.
This will help  to approach the long-term national goal of ambient
radon  levels  in buildings  established in the  1988  Indoor Radon
Abatement Act while also improving indoor  air quality.


                            REFERENCE

1.   Leovic, K.W.,  Craig,  A.B.,  and Harris, D.B.   Update on
     Radon Mitigation Research in Schools.  Paper presented at
     The 1991 Annual American Association  of  Radon Scientist
     and Technologists   National  Fall  Conference, Rockville,
     MD. October  9-12,  1991.

                         ACKNOWLEDGEMENTS

     The authors  would like to express their appreciation to the
Rapid  City and St. Paul school  officials  who assisted in making
these  projects  possible.  Thanks also to Bill Angell of  MURC who
assisted with the project in St.  Paul.
                                12

-------
TABLE 1.  METRIC CONVERSION FACTORS
Non-Metric
cubic foot per minute
(cfm)
degrees Fahrenheit (°F)
foot
inch
inch water column (WC)
picocurie per liter
(pCi/L)
square foot (ft2)
square inch (in.2)
Times
0.47
5/9 (°F-32)
0.305
2.54
249
37
0.093
6.45
Yields Metric
liter per second
degrees Centigrade
meter
centimeter
pascals
becquerels per cubic
meter
square meter
square centimeter
                  13

-------
 TABLE  2.  TEST MATRIX FOR LINCOLN ELEMENTARY SCHOOL
Test
No.
1
2
3
4
5
6
7
8
9
10
11
Test
Dates
11/18
to
24/91
11/25
to
29/91
12/2
to
6/91
12/9
to
16/91
12/16
to
20/91
12/30
to
1/3/92
1/6
to
10/92
1/13
to
17/92
1/20
to
24/92
1/27
to
31/92
12/23
to
27/91
Class
Room
Doors
Normal1
Normal
Closed4
Closed
Closed
Closed
Normal
Normal
Closed
Closed
Closed
Unit
Ventilator
operation
Normal2
Normal
Normal
Normal
Normal
Normal
Normal
Normal
Normal
Normal
Off
Exhaust
Fan
Operation
Off
On
Off
On
Off
On
Off
On
Off
On
On
Outdoor
Air-Damper
Position
Normal3
Normal
Normal
Normal
Open5
Closed6
Normal
Normal
Normal
Normal
Normal
ASD
System
Operation
Off
Off
Off
Off
Off
Off
On
On
On
On
Off
Notes: 1  Determined by teacher preference
       2  On 4:00 am - 8:00 pm weekdays, off over weekend
       3  Controlled by outdoor/inside temperature
       4  Teachers asked to keep closed as much as possible
       5  OA damper mechanically blocked open
       6  OA damper mechanically blocked closed.
                           14

-------
TABLE 3.  TEST MATRIX  FOR  NOKOMIS  ELEMENTARY
Test
Number
1
2
3
3A
4
5
6
7
8
uv
(On/Off)
On
Off
On
On
On
On
On
On
On
UV OA
(%>
Normal
Normal
0
0
10
50
Normal
Normal
Normal
AHU
(On/off)
On
Off
On
On
On
On
On
On
On
AHU OA
(%)
Normal
Normal
0
0
10
50
Normal
Normal
Normal
ASD
(On/Off)
On
Off
Off
Off
Off
Off
Off
On
On
Comments



Filters
Out




Two
Weeks
Later
                       15

-------
TABLE 4.  OUTDOOR AIR FLOW VERSUS DAMPER POSITION FOR UVS
                   AND AHU UNIT IN NOKOMIS ELEMENTARY
Room *
106
106
106
106
107
107
107
107
108
108
108
108
Office
Office
Office
Office
Outdoor Air
Damper Percent
Open
0
10
50
100
0
10
50
100
0
10
50
100
0
10
50
100
Total Unit
Flow (cfm)
900
990
1020
950
926
1109
1050
1009
1050
1190
1200
1205
2025
2160
2280
1960
Outdoor Air
Flow (cfm)
0
265
320
358
0
420
650
660
0
400
513
642
0
420
500
1100
Percent
Outdoor Air
0
27
31
38
0
38
62
65
0
37
43
53
0
19
22
56
    Note that Rooms 106,
    an AHU.
107, and 108 nave UVs and the Office has
                                16

-------
    To High Pressure

    Fans on Roof
Suction
Points
                                                         N
                                                Fan  (T3B)
                                                On Roof
Figure 1.  Floor  plan of Lincoln  Elementary  School, Rapid City,  SD,
          showing  radon  levels  and mitigation  systems.
                                 17

-------
 CO
 o
 o
 o
 03
 EC
 LLI
 CD
 O
 o
 CC
         20
         19
         18
       Library
         16
   Health Room
         14
         13
         12
         11
   Boiler Room
Teacher's Lounge
 Principal's Office
        Gym
       Office
          4
          3
          2
          1
               -100      -50       0       50      100      150      200
                      CHANGE WITH UV ON COMPARED TO OFF (%)
                                                                   250
Figure 2. Changes in radon levels with UVs on compared to off
          using CCs at Lincoln Elementary on 5/91 (data missing for Rms 12 & 14).
                                      18

-------

       fcFFECIS OF ROOM OPERATING CONDITIONS ON
     RADON LEVELS IN ROOM 1, ASD OFF. WEEKDAYS ONLY
        EFFECTS OF ROOM OPERATING CONDITIONS ON
      RADON LEVELS IN ROOM 1. ASD ON. WEEKDAYS ONLY
UV-CkCXO UV-O/EX-1  IN-1/EX-0 UV.1/EX-1

  Unrt V«ntlfator/Exh Fan Oporatlon
                                   Outoki* Door CUMOHU Door Cto*«l

                                 O*»U» Door CtoMiittti Door Open


                             OvMtto Door optnftWI door OP«I
                                   Ou»W« Ooor Cloud/Hall Door Cloud

                                 Ouaidt Door OOM4/HUI Ooor Op«n

                               Outrtrt Deer Op*n/H«l Door Oo^d
uv.o/Ex-o  uv-o/W-i  uv-iyix.o tv-i/tx-t
  Unit Ventilator/Exh Fan Operation
   Figure 3.  Comparison of HVAC  and ASD in Room 1, Lincoln Elementary.

-------
              bF-HtCIS OF ROOM OPERATING CONDITIONS ON
            RADON LEVELS IN ROOM 4, ASD OFF, WEEKDAYS ONLY
' I
 '
         W-CVEX-O     UV-WEX-1    W-1/EX-0    LV-1/EX-I

                Unit V»ntII«tor/Exh Fan Operation
                                               MMDoorOpwi
         EFFECTS OF ROOM OPERATING CONDITIONS ON
       RADON LEVELS IN ROOM 4. ASD ON. WEEKDAYS ONLY
                                                              ,
'
                                                               '
                                                            u
                                                          s
                                                            •
                                                             I
                                                             •



   UV-O/tX-O    UV-O/EX-1     UV.l/EX-0     UV>1/EX-1
          Unit Ventllator/Exh Fan Operation
                                            KWOoorCloMd

                                          HID Doof Op«n
        Figure 4.  Comparison of HVAC and ASD in Room 4, Lincoln Elementary.

-------
:
              EFFECTS OF ROOM OPERATING CONDITIONS ON
           RADON LEVELS IN ROOM 13. ASD OFF, WEEKDAYS ONLY
         UV-WEX-O    UV-WEX-1     W-1/tX-O    UV-1/EX-1
               Unit Ventilator/Exh Fan Operation
                                              Hd Door Opm
     EFFECTS OF ROOM OPERATING CONDITIONS ON
   RADON LEVELS IN ROOM 13. ASD ON, WEEKDAYS ONLY

!»
II
M
'
•
,
.

t









^^ ^m jam
~~L_ _T X /j&BK&y X-fJHBKraJr XH
fPfW / / SJy X vSy XH-,C
UV-0/EX.O     UV-O/EX-I    OV-1/EX-O     UV-1/EX-1
       Unit Ventilator/Exh Fan Operation
         Figure 5.  Comparison of HVAC and ASD  in Room 13, Lincoln Elementary.

-------
              EFFECTS OF ROOM OPERATING CONDmONS ON
            RADON LEVELS IN ROOM 18. ASD OFF, WEEKDAYS ONLY
     JO

          '
i >
• •
                  UV-CVEX-1    UV-1/EX-0    UV-l/tX-1
               Unrt V«n«later/Exh Fan Operation
                                                HMOoordOMd
                                              HaOewOpm
           EFFECTS OF ROOM OPERATING CONDITIONS ON
         RADON LEVELS IN ROOM 16, ASD ON. WEEKDAYS ONLY
i"
                                                           u
                                                         :
                                                         9
                                                              .

                                             Hen Door dowa
     UV-O/EX-O     UV-O/tX.1    UV.T/EX-0     UV.1/EX-1
            Unit Ventilator/Exh Fan Operation
        Figure 6.  Comparison of HVAC and ASD in Room  16, Lincoln Elementary.

-------
                                    UNIT VENTILATORS
CO


r
-i ^

107





108

c


/
\
)









X

AbU












"---AS

PIT




RR RR





109
i i
DP1T


105












106

103












104

101




11
L





1fl9






GYM

	


UTILITY
TUNNEL





STOR


KIT
J
I "• I


STOR


OFF




BOILER
ROOM
Asp pn
•2f^^^
^

Ml IR9F

RR

OFF LNG

      Figure 7.   Floor  plan of Nokomis  Elementary School,  St.  Paul,  MN,
                  showing  radon  mitigation  system locations.

-------
                                          Tast Number.

                                              l-UVi and AHU «l NormaJ Damper Operation; ASO On


                                              2-UV». AHU, ASD Off


                                              3-UV* and AHU On. Damper Closed


                                              3A-UV» and AHU On. Dampen Closed. Piters Out


                                              4-UVs and AHU at 10% D»mper Open

                                              S-UV* and AHU at 50% Damper Open


                                              6-UVs and AHU at Normal Damper Open


                                              7-UV. and AHU at Normal Damper Operation: ASD On


                                              8- Same as Te*1 1 (Two Weeks Later)
                  C
                      Office
3A       4       S

     Test N.aibCT


^^ ROOM 106      2>£
                                                    ROOK 107
Figure 8.  Radon levels during the testing period at Nokomis
             Elementary School.
                                   24

-------
                                                                   X-10
   EFFECTIVENESS OF HVAC SYSTEMS FOR RADON CONTROL IN SCHOOLS


          by:   Kelly W. Leovic, A.  B.  Craig,  and Timothy M.  Dyess
               Air and Energy Engineering Research Laboratory
               U.S. Environmental Protection Agency
               Research Triangle Park,  North Carolina  27711

               Bobby E. Pyle
               Southern Research Institute
               Birmingham,  Alabama  35255


                             ABSTRACT

     Research has demonstrated that heating, ventilating, and air-
conditioning (HVAC) systems in schools can be used to reduce indoor
radon  levels  if  adequate   outdoor air  is  supplied.    The  U.S.
Environmental Protection Agency's  (EPA's) Radon Mitigation Branch
(RMB) has been researching the ability of HVAC systems to reduce
radon  levels  since  1988.    Continuous dataloggers  have  been
installed in 10 schools over the past 2  years to better understand
the influences of various HVAC system operating conditions on radon
levels.  This  paper presents the  results from a Columbus,  Ohio,
school where  a central HVAC system  was used  to reduce elevated
radon levels.  The results from this project, together with other
HVAC  research projects,  are  applied   to  address three general
questions regarding radon mitigation with HVAC systems:

     1) Can HVAC systems be used to reduce radon  levels from above
     20 to below 4 pCi/L?


     2) Can HVAC systems be  used to maintain radon levels below 4
     pCi/L year round  in cold  climates?

     3) How do other  building components (such  as  operation of
     exhaust fans and internal doors) affect HVAC system control of
     radon?

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

-------
                           INTRODUCTION

     EPA's Air and Energy Engineering Research Laboratory (AEERL)
has conducted  radon mitigation  research  in 50  school  buildings
since 1988  (1).  Research  in many of these schools has addressed
radon reduction using active soil depressurization (ASD) systems.
Results indicate that  ASD can be  successfully  applied  for radon
reduction in about 75% of U.S. school buildings.

     AEERL  has also conducted  research on  the use  of heating,
ventilating,   and   air-conditioning   (HVAC)   systems  for  radon
reduction.    Using  the  HVAC  system  to  control radon  can  be
beneficial for the following reasons:

     • radon reduction in the estimated 25% of U.S. schools where
     ASD is not applicable,
     • supplemental  radon  reduction  in  schools where ASD systems
     are  installed in order to  further reduce  radon levels  (to
     reach  the long-term national goal of  ambient  radon levels
     established in the 1988 Indoor Radon Abatement Act), and
     • improved indoor air quality in addition to radon reduction
     through the introduction of additional outdoor air.

     Use of the HVAC system as a radon control technique  depends on
the specific building, but in general, it may be considered in  any
school  that   has  a  HVAC  system  that  supplies  outdoor  air.
Specifically,  use  of the  existing HVAC system  to control radon
levels is not  reasonably applicable  in  schools where:

     •  the  existing  HVAC  system  is   not  designed  to  supply
     conditioned outdoor  air (e.g.,  exhaust-only system, radiant
     heat,  or  fan  coil units),
     •  the  existing HVAC  system does not  consistently supply
     outdoor air during all seasons,  and
     •  radon  control/indoor air quality concerns  in the school
     system are overridden by energy cost concerns.

     Since  1990, AEERL has researched HVAC  control of radon  in 10
schools.  Central HVAC systems and unit ventilators have each been
extensively researched  in four  schools  (for  a total  of  eight
schools), and  specialized  units  (a heat recovery ventilator  and  a
solar  ventilator)  have  been researched in two other  buildings.
Dataloggers were installed in these schools to continuously monitor
parameters  such as radon concentration,  differential pressure,
differential   temperature, percent open  of  outdoor  air damper,
operation of exhaust fans, and opening and closing of doors in  the
building.   AEERL's datalogging system is described in Reference 2.

     This paper  presents recent  data collected in one  of  the
research schools with a central HVAC  system.  The school is located
 in Columbus, Ohio,  and research was conducted from spring 1991  to

-------
spring  1992.  These  recent  data,  together  with data  presented
previously (3) and at this Symposium  — Comparison of ASD and HVAC
System  Control  in School  Buildings  (4)  and  Multiple  Mitigation
Approaches Applied to a  School  With  Low  Permeability Soil (5)  —
are summarized  in the conclusion  of the paper  to  address three
questions on radon reduction with HVAC systems:

     1) Can HVAC systems be used to reduce radon  levels from above
     20 to below 4 pCi/L?

     2) Can HVAC systems be used to maintain radon levels below 4
     pCi/L year round in cold climates?

     3) How  do other building  components (such as  operation  of
     exhaust fans and internal doors) affect HVAC system control of
     radon?
                 RESULTS: COLUMBUS, OHIO, SCHOOL

     The following  sections cover a description of the building,
diagnostic  measurements,  and  analyses  of  the  continuous  data
collected in the school.  Readers more familiar with metric units
should refer to the conversions in Table 1.

BUILDING DESCRIPTION

Structure

     The  original  school  building was  constructed in  1966 and
includes 13 classrooms, offices, and a multipurpose room.   The area
of  this  slab-on-grade  building is 24,000  ft2.   A slab-on-grade
addition was constructed in 1973,  bringing the total area  to 31,000
ft2. A floor plan of the school is  displayed  in Figure 1.  A subslab
utility  tunnel runs  under  three corridors in the  school.   The
tunnel  walls  are  constructed  of  unpainted  concrete   blocks,
facilitating airflow  from the soil  to the tunnel.

HVAC Systems

     The original building has 23 fan-coil units (FCUs) located  in
the subslab utility  tunnel.   The utility  tunnel runs under the
corridors  to the  east, west,  and  south,  as shown  in Figure  2.
Supply air for the  FCUs is  distributed to the tunnel by  a central
fan located in a fan room adjacent to the tunnel.   The  outdoor air
supplied  from  the central fan ranges from 0 to 100%.  Air in the
tunnel  (supplied by  the central fan)  is  then  supplied to each
classroom  by the FCUs.  The air  either passes through hot  water
coils for heating or  bypasses the coils.  The air from  each FCU  is
then  supplied to the  classroom  via a  subslab  duct  which then
distributes  the air  through registers  located along the  outside

-------
walls of  the  classroom.  Return  air from the  classrooms  passes
through openings  in  the classroom-to-corridor wall and  into  the
corridor.  The return air is then pulled into a centrally located
return air grille in  the corridor  near the central fan (Figure 2).
During normal  building operation (7 am to 6 pm) the central fan and
the FCUs run continuously.

     This type of HVAC system is sometimes referred to as a "face
and bypass" system.  Although a number of schools in the Columbus
area have this type of system, it has not been observed by EPA in
schools in other parts of the country.   Central HVAC systems with
subslab  supply  and/or  return air ductwork  have,  however,  been
observed in many of EPA's research schools (1).

     The 1973 addition  has a  central HVAC system for heating and
cooling.  All of the ducting for this system is located overhead.
Since premitigation  radon measurements indicated that  the radon
problem  was  more  urgent in  the  original building  than  in  the
addition, the addition was not part of this research project.

PreMitiqation Radon Levels

     initial radon measurements in the school were made in December
1990 by the Columbus Department of Health.   The measurements were
made with E-Perms and averaged 11.1 pCi/L,  with a maximum reading
of 20.5 pCi/L.  Follow-up charcoal canister measurements were made
the weekend of  March 22-25,  1991, when the  HVAC system was off.
The average radon level was  4.5  pCi/L with a high  of  8.4 pCi/L
measured  in  the  east  tunnel.  The  weather was  relatively mild
during this measurement period, with  a  high of 72°F and an average
of 56°F.  One  would expect these mild weather conditions to reduce
the stack effect  in  the building.   A  reduced  stack effect would
result in reduced radon entry compared to colder weather.

DIAGNOSTIC MEASUREMENTS

     Initial  diagnostic  measurements   conducted in  March  1991
indicated that the utility tunnel and associated air distribution
system were the  major contributing factors to elevated radon levels
in the school.    As  a result, the focus of  this research project
was on the effect of HVAC system operational parameters on radon
levels.   Pressure field extension   (PFE) measurements  were also
conducted as  part of the diagnostics and will  be  presented in a
forthcoming EPA report.

     Inspection  of the  building  HVAC  system  indicated  that the
system  might  not  be  operating per   design.    Reduced  system
maintenance due to budget limitations probably contributed to this
situation.  To obtain  information on the current operation of the
system, a local testing and balancing company was hired to measure
system airflows  in May 1991.   A 60  point traverse on the suction
side of the central fan was made with the fan set at  its design of

-------
550  revolutions per  minute.    Results  showed that  the fan  was
running at 27, 264 cfm: 264 cfm above the design specification of
27,000  cfm.    The results  of  the  airflow measurements  (actual
measurements) for all  23  FCUs,  together with  the design airflow,
are shown in Table 2.  These measurements show that,  although the
overall airflow for the 23 FCUs is only 233 cfm above design, there
is wide variation from design for most of the individual units.

     During the following winter (1991-92), Columbus Public Schools
maintenance personnel  adjusted  the  FCUs to their original  design
airflow (third column  of  Table  2).   The central  air  handling fan
was set to  maximum opening position with  no  change  in operating
speed and  locked in place.   These  adjustments of the  HVAC fans
helped  to  reduce  the depressurization  in the  tunnel.   School
maintenance personnel also applied two coats of paint to the tunnel
walls and caulked all  floor  and walls  cracks  in  the  tunnel.  The
authors were not informed  of these changes (adjustments to the HVAC
system and sealing of the tunnel) beforehand and,  as a result, were
not able to quantify the  effect  of the individual changes on radon
levels.

ANALYSIS OF CONTINUOUS DATA

     During the spring of 1991,  the  school was instrumented with a
continuous datalogger to record:

     • radon concentrations  (eight locations)
     • differential pressure  (seven locations)
     • temperature (seven locations)
     • percent open of return and outdoor air dampers  (central fan)
     • weather  (humidity, wind speed and direction, rainfall)

Initial plans were to collect data during the spring of 1991 with
the central HVAC fan  supplying specified quantities of outdoor air.
The  test  matrix  for  these  measurements  is  shown  in Table  3.
Unfortunately, school officials were unable to adjust the outdoor
air  damper  as  required   for  the  testing,  so   testing  of  the
parameters in Table  3 was delayed until the winter of 1991-92.

     Note  that the  percent  open of  return air  and  outdoor air
dampers  in  this  paper  refers to  the  actual  damper  position.
Airflow versus damper position would be expected  to be non-linear,
with the greatest increase in airflow in the initial increments of
damper opening.  For example, if the outdoor air  damper  is open to
50%, the quantity of outdoor air likely exceeds 50%.  Calculations
to determine the quantities of outdoor air associated with the 50%
damper  position have  been made and are to be  presented  at the
Symposium.

     Data were collected for the test matrix both before and after
school personnel  adjusted the HVAC  system and sealed the tunnel.
The results are discussed below.

-------
Effect  of  Outdoor  Air  Damper  Position  on  Radon  —  Before
Adjustments

     Figure 3 shows the effect of the outdoor air damper position
on radon levels in the school.  These data were collected prior to
adjustments to the HVAC system and sealing of the tunnel; however,
it is possible that some of the work was initiated during the final
test  (with  100% outdoor  air).   Since school officials  did  not
inform the researchers of the work until it was underway, the exact
timing is not known.

     The top graph in Figure 3 shows radon levels in the west and
east  tunnels,  and the  bottom  graph shows  radon levels  in  five
classrooms and the teacher's lounge.   The average  radon levels
during each of the three  test conditions  in Table 3 (100% return
air,  50%  return  air/50% outdoor air,  and 100%  outdoor air)  were
calculated from  the  continuous  data  collected during  the  test
condition.  Three conclusions are apparent from these data:

      (1) With 100% return air, average radon levels are much higher
     than the previous premitigation E-Perm and charcoal canister
     measurements, averaging over 20 pCi/L.

      (2)  Average radon levels  in  the six rooms  (bottom graph)
      closely track radon  levels  in  the tunnels  (top graph).   The
      levels  in the rooms tend to  be slightly  lower than in the
      tunnels,  supporting  our assertion that the  tunnels are the
      primary source of radon in the school.

      (3) Radon levels are  reduced by about 75% when the outdoor air
      damper position  is increased from 0 to  50%.   No additional
      reduction is observed when the outdoor  air  damper position is
      increased from 50 to  100%.  However, the data collected during
      the  50% return  air/50% outdoor  air damper conditions  are
      limited  to  a 24 hour test and, thus,  should be interpreted
      cautiously.

      (4)  Although  radon levels  are  reduced by about 75% when the
      outdoor air is increased to 50 or  100%, average radon levels
      in the classrooms  still exceed 5  pCi/L.

Effect of Outdoor Air Damper Position on Radon — After Adjustments

      The  test matrix was then  repeated after school maintenance
personnel adjusted the HVAC system to design specifications and
painted the tunnel walls.  The results from these tests  are shown
in  Figure 4.   These data show  similar trends in radon  levels as
Figure 3  (before adjustments).  However, average radon levels are
about 6  pCi/L lower in the west tunnel and 1 pCi/L lower in the
east  tunnel after sealing. These reductions in tunnel radon levels
are also  reflected in  the room radon levels.  During the test run
with  50% return air/50%  outdoor air,   radon  levels  in the rooms

-------
average about 4 pCi/L.  When the outdoor  air is increased to 100%
average radon levels  in the classrooms are about 2.5 pCi/L.

     Figure  5 summarizes  the  average  radon levels  in the  two
tunnels and  six rooms  after adjustment  of  the HVAC  system  and
sealing of the tunnel.  The percent reduction attributed to these
changes is shown above each bar.  All reductions exceeded 10%  with
a high of 57% reduction when 100% outdoor air was supplied.

     Prior to these adjustments, the differential pressure between
the tunnel and the outdoors was negative to neutral (depending on
the percent outdoor air).  After the HVAC system was adjusted and
the tunnel  sealed,  the differential pressures  under all  three
outdoor air conditions were relatively neutral.  This reduction of
negative  pressure  reduced  radon  entry  into the  tunnels  and
consequently, radon levels in the school.


                           CONCLUSIONS

     The data  from the Columbus school  presented  in  this  paper,
together with  data from other  schools  researched by  AEERL,  are
applied to address three questions on radon  reduction with HVAC
systems:

     1) Can HVAC systems be  used to  reduce radon levels from above
     20 to below 4 pCi/L? Data  from this school showed that radon
     levels could  only  be reduced from above  20  pCi/L to below 4
     pCi/L  when  the HVAC  system  was  adjusted to   its  design
     specifications,  100% outdoor air was supplied to the central
     fan,  and tunnel walls were  sealed.   Although radon reductions
     were about  75% prior to adjustment of  the HVAC  system  and
     sealing of  the tunnel, levels  were not reduced  to below 4
     pCi/L even  with 100%  outdoor air.   Data from  other schools
     (3,4,5)  support the conclusion  that  is it difficult to reduce
     radon levels from above 20 pCi/L to below 4 pCi/L using only
     the HVAC system.

     2) Can HVAC systems be used for radon control  year  round in
     cold climates?  The research project in this school indicated
     that,  when  100% outdoor air  is introduced and  major  radon
     entry routes are sealed,  radon  reduction with the HVAC system
     is consistent.  However,  school personnel indicated that this
     condition  is  difficult  to  achieve during  extremely  cold
     weather  due   to  concern   with  energy  costs.     Similar
     observations have been made in other  research schools (3,4,5) ,
     particularly in schools with unit ventilators.

     3) How  do other building  components  (such as operation of
     exhaust fans and internal doors) affect HVAC system control of
     radon?  Other building components were  not  monitored in this
     study.   However, research in other schools (3,4,5) have shown

-------
     that  (a)   opening the  classroom-to-corridor  door  affects
     classroom radon levels (may  increase or decrease depending on
     the school),  and (b) operation of exhaust fans may increase or
     decrease classroom radon levels depending on the building.


                           REFERENCES

1.   Leovic, K.W., A.B. Craig, and  D.B. Harris.   Update on Radon
     Mitigation Research in Schools.  EPA-600/D-91-229 (NTIS PB91-
     242958). Presented  at the  1991  Annual AARST National  Fall
     Conference, Rockville, MD, October 1991.

2.   Harris,  D.B.  and B.E.  Pyle.    Data   Logging  Systems  for
     Monitoring Long-Term Radon Mitigation Experimental Programs in
     Schools and  Other Large Buildings.   Presented at  the  85th
     Annual AWMA Meeting,  Kansas City, MO,  June 21-26,  1992.

3.   Leovic, K.w., D. B. Harris, T. M. Dyess,  B. E. Pyle, T. Borak,
     and D. W.  Saum. HVAC System Complications  and  Controls for
     Radon Reduction in  School Buildings,    Presented at the 3rd
     International   Symposium   on  Radon  and   Radon   Reduction
     Technology, Philadelphia, PA, April 2-5, 1991.

4.   Pyle, B.E.,  K.W.  Leovic,  D.B. Harris,   T.M. Dyess,  and  A.B.
     Craig.  Comparison of ASD and  HVAC System Control  in School
     Buildings. To be presented at the 4th International Symposium
     on  Radon  and  Radon  Reduction Technology,  Minneapolis,  MN,
     September 1992.

5.   Harris, D.B.  and  E. Moreau.   Multiple  Mitigation Approaches
     Applied  to  a School With  Low  Permeability  Soil.   To  be
     presented  at the 4th International  symposium on  Radon and
     Radon Reduction Technology, Minneapolis, MN, September 1992.


                        ACKNOWLEDGEMENTS

     The authors  would like  to thank David  Sands of the Columbus
Public Schools  (Architecture and Engineering Department)  for his
assistance  with  this  project;    Harry Grafton  of the Columbus
Department of Health for conducting radon and  differential pressure
measurements; and Bill Angell of the  Midwest Universities Radon
Consortium for his assistance during the diagnostic measurements in
Columbus.
                                8

-------
TABLE 1.  METRIC CONVERSION FACTORS
Non-Metric
cubic feet per minute
(cfm)
degrees Fahrenheit (°F)
picocuries per liter
(pCi/L)
square feet (ft2)
Times
0.47
5/9 (°F-32)
37
0.093
Yields Metric
liters per second
degrees Centigrade
becquerels per cubic
meter
square meters

-------
TABLE 2. AIRFLOWS FOR FCUS
Room Served by FCUs
1
2
3
4
5
6
7
8
9
10
11
12
Multipurpose Room
Stage
Teacher's Lounge
Music Room
Principal's Office
Supply Room
Conference Room
Office
Nurse's Office
TOTALS
Measured
Airflow (cfm)
868 & 601
705 & 816
1305
1260
1170
1435
1407
1335
1192
1210
1269
1318
3645
839
541
568
345
361
309
476
308
23,283
Design
Airflow (cfm)
1700
1700
1200
1200
1200
1200
1200
1200
1200
1200
1200
1200
4000
1000
450
700
350
300
300
350
200
23,050
             10

-------
            TABLE 3. TEST MATRIX FOR SCHOOL
Test
No.
I
II
III
Duration
1 week
1 week
1 week
HVAC System
Operation (hours)
6:30 am - 4:30 pm
6:30 am - 4:30 pm
6:30 am - 4:30 pm
Return
Air (%)
100
50
0
Outdoor
Air (%)
0
50
100
Notes:    FCUs operate when HVAC system operates.
          HVAC system hours of operation are approximate.
                            11

-------
  5.3
(tunnel)
              N
            5.3
             4.9
            <0.5
                                 <0.5

                             C0.5 \
                   5.5
                              <0.5
                                            
-------
N
u
Return Air
Damper
^^^_
Main
Air Handling x
Unit
Lt/
•*•**>.
1
•••
Outdoor Air
Damper
r
*• 	
^sia^^
--* Filters
c=»ol
i ••
,
Fan Coil
J (Typical)
•*«*» Supply Air
(/"
^


J
r^
-*^ Supply Air
(To Fan Coils)
^ Return Air
>s (From Upstairs)
     Figure 2. Layout of subslab utility tunnel,
                            13

-------
    20
  1
    10
                                                100% O A
                          HVAC Operating Condition
    * - Limited data for chit condition
    20
  -4
  c 10
  cc
100%RA
                                                100%OA
                          HVAC Operating Coadinoa
     - Limited dan (or this condition
Figure 3. Effect of  outdoor  air damper on  radon levels
           (before  sealing). Top,  tunnels;  bottom,  classrooms
           (OA =  outdoor  air;  RA  = return  air.)
                             14

-------
  23
  .
1
i
  1
                        HVAC Operating Condition


                        W-Tunn«l
                                              100%OA
  2C
§
a is

1
•

  10
            100%RA
   50%RA/50%OA

HVAC Operating Condition
                                              lOCK^OA
  Figure 4.  Effect  of outdoor  air damper on radon levels
              (after sealing).   (OA  = outdoor air;  RA =  return

             air.)
                                15

-------
J 20
0
-
-
t
i
:
:
  15
10
*  J
                  17
              10O&RA          50%RA/50%OA          100%OA
                            HVAC Operating Condition

                           Qassrooms V//J Tunnel
 Note: Numbers above bars are percent reduction due to sealing tunnel
          Figure 5.
                   Average  radon  levels  in  tunnels  and   in
                   classrooms  before sealing  and percent  radon
                   reduction due to sealing.  (OA — outdoor air;
                   RA = return air.)
                                16

-------
                                                                   X-11
         RADON PREVENTION IN CONSTRUCTION OF SCHOOLS AND
         OTHER LARGE BUILDINGS—STATUS OF EPA'S PROGRAM


          by:  Alfred B. Craig, D. Bruce Harris,
               and Kelly W. Leovic
               U. S. Environmental Protection Agency
               Air and Energy Engineering Research Laboratory
               Research Triangle Park, North Carolina 27711


                             ABSTRACT


     The Air and Energy Engineering Research Laboratory  (AEERL) of
the U.S.  Environmental Protection Agency (U.S. EPA) started a radon
mitigation  research, development, and demonstration  program in
1985.  Initial studies were on existing  and  new houses,  and the
program  was expanded  in  1987 to  include  mitigation  studies in
existing schools.    AEERL  has  conducted various  levels of radon
mitigation research  in 49 existing schools in  13 states  since that
date.

     In  1990,  a decision was made to  issue a technical guidance
document that  describes how to design  and build schools and other
large buildings which will  be radon resistant and easy to mitigate.
Tentative  design premises  were  developed,  based on  studies in
existing schools, and presented at The 1991 International Symposium
on Radon and Radon  Reduction Technology, Philadelphia, PA, April
2-5, 1991  (1).

     During  1991,   an  opportunity  arose  to  demonstrate  this
recommended  technology in  a 60,000  sq ft hospital  building in
Johnson  City,  TM.   Results of this  project are reported herein.
The  mitigation  system was  extremely effective,  lowering radon
levels to  less than 0.5 pCi/L (detection limit for test used) in
the entire building.  Levels greater than  50 pCi/L were measured
in the building  when both  the air handling  and  mitigation systems
were shut  off.

     The technical guidance document  is in the final stages  of peer
review and will  be  issued  before the end  of this  year.

     This paper also gives  examples of oversized or poorly designed
radon mitigation systems which are  currently being installed in
some new U.  S.  schools.

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

-------
                           INTRODUCTION


     The Air and Energy Engineering Research Laboratory (AEERL) of
the Environmental Protection Agency  (EPA)  initiated an extensive
research, development, and demonstration program on the mitigation
of radon in structures in the U.S.  in 1985.  initial work was done
on  houses,  both existing and new.    In  1987,  this  program was
expanded to include schools and other  large buildings.  Extensive
diagnostics have been carried out in  many  existing schools, and
mitigation systems  have been installed and evaluated in typical
ones.   This work was carried out in 49 existing  schools  in 13
states.

     In 1990, a decision was made to develop technical  guidance for
the  design  and construction  of  new  schools and  other  large
buildings which were  radon resistant and easy to  mitigate.   As a
prelude to developing this technical guidance,  the architectural
plans  and  specifications  of  all  schools  mitigated  by  EPA were
carefully studied to identify those  features which affect radon
entry and ease of  mitigation.   The  results of  these studies are
being used to  develop a technical  guidance manual for the design
of  new schools which  are  radon  resistant and  which can be easily
and inexpensively mitigated if a radon problem is found after the
building is completed.  This document  will be released by the end
of  this year.

           DESIGN FEATURES AFFECTING EASE OF MITIGATION
            WITH ACTIVE SUBSLAB  DEPRESSURIZATION (ASD)


     Review of all  school pressure field extension (PFE) studies
and  examination  of   architectural  plans  have   led   to  the
identification of the following features which affect PFE  and hence
the effectiveness of  ASD:

          Subslab barriers (type,  size, and location)
          Aggregate
                Bulk density (or void volume)
                Particle size (both average size and particle size
                 distribution)
                Particle shape   (naturally  occurring  stone  from
                 moraine  deposits  with rounded corners vs crushed
                 bedrock)
          Subslab suction pit size
        •  Amount of suction applied

SUBSLAB BARRIERS

      Schools have been found to  fall  into one of the following four
categories  listed in  order of ease of mitigation:

-------
     Type 1.   No interior walls extend through slab with the roof
               load  being  carried by posts  (steel  or reinforced
               concrete) extending through the slab to footings.

     Type 2.   Walls  between classrooms  extend through  slab  to
               subslab footings.  Hall walls do not.

     Type 3.   Hall walls extend through slab to subslab footings.
               Walls between classrooms do not.

     Type 4.   All walls extend through slab to subslab footings.

     A recent survey of  the characteristics of about 100 schools
chosen at random on  a national  basis has shown that about 60% of
existing  schools  in  the U.S.  are  Type  1,  the easiest  type  to
mitigate.   Type  1  construction is  also increasing  in popularity
particularly since air conditioning  in schools negates the need for
windows in  all rooms.   In commercial and industrial construction
where dimensions of buildings are large in both directions (length
and  width)  ,  Type   1  (known  architecturally  as post  and  beam
construction) is almost always used.  It is also believed to be the
most economical of the four types.

AGGREGATE

     Aggregate  characteristics  are also very important to  the
mitigation of large slabs.  The following conclusions on the effect
of aggregate properties on  PFE are believed to be accurate although
the effects of these variables have not been quantified:

     1.   PFE is proportional to average  aggregate particle size
          —the smaller the average  particle size, the less the PFE
          (assuming the same relative particle size distribution).

     2.   The narrower the aggregate  particle size distribution
          range, the greater the void volume and the PFE (assuming
          the same average particle size).

     3.   The smoother the shape of the stone, the  lower the void
          volume; hence  moraine stone (with its rounded corners)
          has lower void volume  and  will give less PFE for the same
          average particle size  and particle size distribution than
          crushed aggregate.

     Based  on these  parameters  the  optimum subslab aggregate for
ASD is believed to be crushed bed rock meeting specifications for
#5 stone as defined  in ASTM-33-86 (2).

-------
SUCTION PIT SIZE

     Suction pit si2e also increases in importance as the size of
the building  increases.   Since  cost of  the  suction pit  in  new
construction is relatively independent of size,  the pit as shown
in Figure 1 is recommended.   This type pit with a 6 in.  riser was
used in the Johnson City  hospital  described in the next section.
It is inexpensive and easy to install.  The interface between the
pit void and the aggregate face in this pit (a key variable in PFE)
is 7 sq ft.

AMOUNT OF SUCTION APPLIED

     The larger the suction  fan  the greater is the PFE attained.
A  T3B  Kanalflakt  fan  was   used  in  the  Johnson  City  hospital
described in the next section.   This fan moves 510 cfm of air at
no head.  Initial  and operating  costs of this  size fan (rated at
0.1 HP) are inconsequential in a large building.

       APPLICATION OF OPTIMUM MITIGATION  DESIGN FEATURES TO
        JOHNSON CITY REHABILITATION HOSPITAL CONSTRUCTION


     Late in 1990, an opportunity presented itself to demonstrate
ASD in  a large building  under optimum conditions  in  a hospital
being built in Johnson City,  TN.   The hospital  building is one
story with a floor area of about  60,000 sq ft and is slab-on-grade
construction with  no  foundation  walls penetrating the slab (Type
1).  Mechanical piping,  electrical  conduit, and structural columns
penetrate the slab, and the columns sit on footings below the slab.
These columns support steel  beams overhead which in turn carry the
bar joists for the roof.   This type of construction is referred to
architecturally as post and beam construction.   It is used in most
commercial  and  industrial buildings currently  being built in the
U.S.  All internal walls  are gypsum board on metal  studs, and the
exterior walls are metal stud supporting gypsum board on  the inside
surface and an  exterior  insulation finish system on the outside.
The 4 in. slab was  poured  over a 6 mil vapor barrier underlain with
a  4 in. layer of crushed  aggregate which  was continuous under the
entire  slab.   The slab,  exterior  foundation walls, and footings
were poured monolithically.   The slab was divided into about 15 ft
squares  by a  combination of pour  joints  (1,000  lineal  ft)  and
control  saw joints (5,000 lineal  ft).   No expansion joints were
used.

     EPA was requested to review the  plans and specifications and
to recommend a  radon mitigation  system since the region was known
to have high radon potential.   After this review, the following
recommendations were made to the architect designing the building:

-------
  1. Minimum of  4 in. of crushed  aggregate  meeting the specifi-
     cations  for  #5  stone  as defined  in  ASTM-33-86  "Standard
     Specifications for  Concrete  Aggregate"  (2) carefully placed
     so as not to  include any soil.

  2. Sealing  of  all  pour and  control saw  joints and  any  slab
     penetrations  with a polyurethane caulking.

  3. Installation of one  centrally  located subslab suction pit (SP)
     of the design shown in Figure l in the approximate center of
     the slab with a  6 in.  stack  leading to the roof capped with
     a Kanalflakt  T3B turbo fan capable of  moving 500  cfm at no
     head.

  4. Continuous  operation of  the heating,  ventilating,  and  air
     conditioning  (HVAC) fans in order to pressurize the building
     in all areas except those where negative pressure is necessary
     to control  odors, noxious chemicals,  or infectious diseases
     (toilets, kitchen,  pharmacy, soiled  linens  area,  isolation
     wards, etc).

     All of the recommendations were accepted and incorporated into
the building design.   Upon completion of  the  building  shell  and
sealing of the slab, diagnostic measurements were made to determine
the potential for a radon problem  and to test the effectiveness of
the ASD  system.    Test  holes  were drilled  through the  slab  at
varying distances  from the suction pit, including  a series around
the entire perimeter about 5 ft from the slab edge.  Radon levels
below the  slab  were  measured by  "sniffing" using a Pylon  AB-5
continuous monitor.   Levels from  200  to  1,700 pci/L were found
under the slab.  This is a significant level of radon which could
result  in  indoor  measurements  of  over  50   pCi/L  under  some
conditions of building operation.

     The  depressurization  fan  was  then   turned  on and  subslab
pressure was measured made using a Neotronics roicromanometer.  The
fan removed about 200 cfm of soil  gas at a  vacuum of about 1.5 in.
W. C.  Negative  pressure was 0.47 in. W.  C.  in the suction pit,
0.22 in. 50 ft from the SP, and 0.18 in. at the farthest point on
the perimeter (a  distance of 185 ft) . This  is considered extremely
good PFE.  The PFE data  are shown in Figure 2 and are plotted in
Figure 3,  giving essentially a straight line  on  semi-log paper.
Extrapolation  of  these   data  in  Figure  3  indicates  that  the
mitigation system  would  mitigate  a much larger slab.   At 500 ft
from the suction pit, subslab pressure would have  been about 0.15
in. W.  C., almost  two orders of magnitude greater  than needed for
mitigation.  This indicates that the system as installed could have
mitigated a slab of over  1,000,000  sq ft had it been that large and
built according  to the specifications used  for the Johnson  City
hospital.

-------
     Upon completion of the building,  radon levels were measured
in half of the  building with the HVAC and  ASD  systems off using
open faced charcoal canisters.  Radon levels ranged from less than
0.5 pCi/L (lowest detectable level  with  the open faced canisters
used)  to  52.7  pCi/L.    Highest  levels  were  in the  bathrooms
associated with the patient rooms.   The  patient room  with the
highest bathroom radon  level had  a  reading of 10.5 pCi/L,  higher
than that found in any non-bathroom area in the building.

     The entire building was then  measured with the HVAC system on
and  the  ASD  system off.   Again some  of  the  bathrooms  (under
negative pressure)  had  elevated  radon levels as  did  some  of the
patient rooms.  The bathroom with the  highest radon reading with
the HVAC off was again  the  highest  in  the  building with the HVAC
operating, testing  16.1 pCi/L.

     The final series of tests  was made with both the HVAC and ASD
systems operating.  The 20 bathrooms with the  highest radon levels
in the second series of tests were remeasured.  No measurable radon
levels were  found  in  any  of the  rooms  tested.   This  is not
surprising in view  of the large  negative pressure under the slab
with the installed ASD system in  operation.

                     INSTALLATION OF ASD AT
                  DRANESVILLE ELEMENTARY SCHOOL


     AEERL was involved  in the  development of  the  mitigation
strategy  for Dranesville Elementary School,  Fairfax  County, VA,
when it was  being designed  in 1987.   The  school was designed to
control radon using the HVAC system by  running  the  circulation
fans of the  HVAC system continuously  to keep the building under
positive  pressure  at  all times.   As a backup  precaution, the
subslab was  prepared  for  the possible  use  of ASD by installing a
4 in. layer of #57 stone and carefully sealing all of the expansion
joints in the  slab.    When  the   school  was tested after  it was
completed in the  summer of  1988, radon  in  all  rooms  was below 4
pCi/L  with  the air handling fans  operating.   The  building was
tested in the middle of the  winter 3 years  later: all rooms in one
wing tested  between 4  and  10 pCi/L and EPA was called back to
investigate  the reason for  the   failure of  the  radon mitigation
system.  After  considerable investigation,  it was found the  HVAC
system had been converted to standard  night setback shortly  after
Energy Management System  (EMS) personnel took over control of the
HVAC system  in  the  building.

     Maintenance  department management  decided to  install ASD
rather than  try to convince the EMS  people  to  operate the HVAC
system as designed (continuous fan operation).   At the time the
building was designed,  the  importance  of subslab barriers had not
been recognized and no attempt was made to influence the foundation

-------
design.  Although the design does not fit the definitions for any
of the four  types  of subslab construction,  it is most similar to
Type 3 (see Figure  4) but only one room (the  kindergarten room) was
found  to have  subslab  barriers on all four  sides.   Since one
suction  point  was  expected  to reach  all  of  the  wing  (except
possibly the kindergarten room), the suction  point location was
chosen centrally as shown in Figure 5.   PFE  was measured using the
vacuum cleaner test method and reached all areas  of  the wing under
investigation including the kindergarten room.  The  ASD system was
then installed by maintenance personnel.  Retesting with the system
operating showed that all  of the subslab area of the  wing was under
negative pressure  as shown in  Figure  5.    with the ASD system
operating,  all rooms tested at  less than 1  pCi/L.

     The system as installed did not meet the specifications used
in Johnson  City with regard to pit  size,  stack pipe  size, and
suction fan  size.   Dirt removed from the suction pit was only 5
gallons, meaning  that the  pit was probably  only about  1  ft in
diameter.  The  surface of the interface between  the pit void and
the crushed  aggregate calculates to be  about  1 sq ft compared to
about 7 sq ft at Johnson  City.   The stack  pipe was  only 4 in. in
diameter (compared to 6 in. at Johnson City) and about 75 ft long
resulting  in considerable  pressure loss.   The  fan used  was  a
Kanalflakt T2 capable of  pulling 260 cfm at  no  head rather than
500 cfm as specified for Johnson City.  To  test the effect of fan
size, a Kanalflakt T3B fan, capable of pulling  500 cfm at no head,
was  added  to  the  line as it  exited  the  building.   With this
arrangement, subslab negative pressure  was about doubled at all
test points, indicating that the PFE was about doubled.

                     INSTALLATION  OF ASD AT
                     DUNN ELEMENTARY  SCHOOL


     Installation  of  ASD  at  an  existing  school  of  Type  1
construction in 1991, Dunn Elementary School  in Louisville,KY, gave
EPA an opportunity to test the effect of the  suction pit size on
PFE.   A local mitigator had made diagnostic tests at this school
and had recommended an ASD system consisting of 14 suction points
and three  fan  systems.    EPA was asked to  review the mitigation
strategy.  Based on the contractor's diagnostic data and a review
of the foundation drawings (showing post and beam construction),
EPA believed that the 53,000 sq ft building could  be mitigated with
one suction point.   The  suction point was evaluated using four pit
sizes:  none (just a 6 in. hole through  the  slab), 1 by 1 ft, 2 by
2 ft, and 2 by 3 ft.  Although it was necessary to dig the last two
pits  2 ft deep  to  get the desired  diameter,  it  is  believed that
only that portion of the  pit contacting the aggregate layer (top
4 in.) contributed to PFE.

-------
     PFE was measured at the points noted on Figure 6 for each pit
size using a fan capable of  pulling 510 cfra  at no head.   Results
are plotted in Figure 7.   Increasing the  diameter of the suction
pit definitely increases the PFE.   If negative  pressure for all of
the  points  is  averaged for  each  suction  pit  size,  consistent
results  are obtained.   Increasing  pit  size  increases  subslab
negative pressure as shown in Table 1.

         TABLE  1.   EFFECT OF SUBSLAB SUCTION  PIT SIZE ON
                  SUBSLAB NEGATIVE PRESSURE
Pit diam
ft
0
1
2
3
Aggr./void
surf area
sq ff
0.2
0.8
1.6
2.4
Stepwise
increase
aggr/void
ratio t %
400
200
150
Average
SS neg
press
in. WC
0.014
0.024
0.030
0.034
Increase
in SS
nea P.%
70
25
13
neg pressure
incr/interface
ratio incr"
%
17
12
9
*  Assumes 4  in.  aggregate layer.
b  Increase in subslab negative pressure divided by the stepwise
   increase in  the aggregate to void ratio.

     These  data  show  that  average subslab  negative  pressure
continues to increase as the suction pit size is increased but at
a  decreasing rate  even  when corrected for the actual increase in
size  of  the  void to  aggregate  interface ratio  (last  column).
However,  since (in new construction) cost of the suction pit does
not increase with increasing pit size (within the range used), the
larger  size  is  recommended.  Consequently, the choice of  a 4 ft
pit (Figure 1) at the Johnson City hospital appears to be a valid
decision.

                      OTHER INSTALLATIONS IN
                         NEW CONSTRUCTION
     Several  other  installations  have  been  brought  to  EPA's
attention where architects and/or mitigators have recommended much
more   extensive  mitigation   systems  than  necessary   in  new
construction.    This  has  resulted  from   the  lack  of  field-
demonstrated  technology  (by EPA or others).   Three examples are
described  in this  paper in  order to  illustrate the  amount of
redundancy which is currently  being used.

-------
SCHOOL 1

     This school has an area of about 60,000 sq ft and it is Type
1 construction (post and beam)  with the  exception of the gymnasium
which  is Type  4  (all four  walls  extend through  the  slab  to
footings).  The building is in a radon-prone area, and a decision
was made to rough in an ASD system.  The system was designed with
one suction  point for about every 5,000 sq ft  (ten  points were
installed in locations shown on Figure 8).  An average  of 45 lineal
ft of  4  in.  perforated  pipe was  used  beneath the slab  at each
suction point to improve PFE (represented by dotted lines in Figure
8).  This installation was brought to EPA's attention before -the
building was  completed,  and permission was  obtained  to make PFE
measurements.

     In  order to measure  the  maximum  PFE range  of  one  suction
point, 11 were capped and a fan was temporarily installed on stack
5.   (See Figure 9.)   With  the fan operating,  negative pressures
under the slab were measured through small  holes in the slab using
a micromanometer.   PFE data obtained are shown in Figure 9 as is
the distance  from the suction  point.   Good  PFE was  found in all
cases except  at point O.   (The reason that  this point  was not
reached is not known since  it is closer to stack 5  than some of the
points that were reached.)  Note that excellent negative pressure
was found at points F  and G even though this area  (the gymnasium)
is isolated  by  subslab  walls  from the  location of the active
suction point.  Further investigation showed that  the footings in
the gymnasium area were poured under very wet  conditions and that
coarse crushed  aggregate was  placed in the  trenches before the
footings were poured.  This apparently allowed the PFE  to reach the
area under the gymnasium slab.  Adequate negative pressure was also
found at point H although this  point had two  subslab walls between
it and the activated suction point.

     This evaluation showed that  one suction  point  (rather than
12) would have been sufficient  in this school.  The additional cost
of the extra points was significant as was  the  cost of the subslab
perforated pipe which is not needed if a large suction  pit  is used.

SCHOOL 2

     A local mitigator designed the subslab system for  this school,
essentially treating it as  multiple houses.  Twelve suction points,
with 10  in. diameter suction pits  in the subslab  aggregate, were
installed with 3  in. risers to the attic where a small house type
suction  fan was to be  added to each stack  if a radon problem were
found.   After the  slab  was poured,  a  professional  engineer who
designed the plumbing, mechanical,  and  electrical  systems for the
architect contacted one of  the authors  to  ask  if the  3 in. stacks
could be manifolded and larger fans used.  The plans were revised
to manifold six stacks to one suction fan and three to each of two

-------
other  fans.   The use  of three  fans  rather than one  was chosen
because of the distance between the stacks.  The individual stacks
will have valves for balancing purposes.  This will also allow us
to determine the effectiveness of a single suction point in the six
point  system  with the  other stacks shut off.   EPA  plans to test
this building as soon  as the fans are installed.  Results should
be available by the time of the  symposium.

     Examination of the  foundation plans of this school disclosed
that  it is Type  1  (post and beam).    It is believed that this
building could have been mitigated with one properly designed and
located suction point.

LARGE  BUILDING 1

     In  October 1990,  one  of  the authors  was contacted  by an
architect who had designed a war gaming facility to be built by the
U. S  Army  Corps of Engineers  for  the Department of Defense in a
radon-prone area of Pennsylvania.   The building was four stories
with  a footprint of about  50,000 sq ft.   The mitigation system
contained about 2,500  lineal  ft of  4  in. diameter perforated pipe
consisting of a loop around  the building near its outside walls
with  connecting  cross  lines every 20  ft.   Four stacks were to be
installed  on the four corners  of  the  pipe system  leading  to a
penthouse.   If the building were found to  have a radon problem
after   completion,  the  system  was   to  be  activated  by  the
installation  of four suction fans on these four stacks.  This plan
was  an adaptation of  a plan originally  developed  by  the Atomic
Energy Control Board of  Canada in  the 1970s.

      Examination of the  foundation plans  showed  that the building
design was Type 1 (post and beam), and it was recommended that the
proposed mitigation system be  replaced  with a centrally located
single suction pit and stack  as  in the  Johnson City  Hospital.
Unfortunately,  the  building was on a fast track schedule and the
Corps was unwilling to slow down the construction schedule as  long
as the planned  system  would solve the problem.   They were assured
that  it would, but that it was more expensive than it needed  to be.
A decision was  made  to proceed  with the  mitigation system as
designed to  eliminate  any delay.


      Mitigation of  these three  buildings demonstrates  the  amount
of over-design being  used  on  mitigation  systems in  schools and
other large  buildings.  The reason for this is  the  uncertainty of
what  is  needed and  the  philosophy of  erring  on the side of
redundancy if an error is made.  This is  really unnecessary since
additional suction points can readily be added on a retrofit basis
if tests show that a given system  is not  sufficient to eliminate
a radon problem in  an  entire building.

-------
     It is also  recognized that additional field verification of
the system  as installed  in Johnson City  is needed, and  EPA is
making an effort to locate new  school  buildings and other large
buildings being designed where this can be done.

     It is hoped that EPA's new construction  guide for schools and
other large buildings,  to  be  issued before the end of 1992, will
eliminate the  practice of over-design.  A system similar to the
Johnson City  system is being recommended  in this guidance.  The
document will also recommend that the HVAC system be designed and
operated to meet ASHRAE 62-1989  ventilation  standards (3).  Plans
are  underway  to   see  that   this  guidance  manual  gets  wide
distribution  to  architectural firms designing school buildings.
Meanwhile a  two page  article  has  been placed  in Architecture/
Research, a publication of the American Institute of Architects/
Association of Collegiate  Schools of Architecture (AIA/ACSA)  (4).

                         INSTALLATION COSTS


     Incremental costs of the mitigation  system at Johnson City
were easily ascertained since the  contract  for  the building had
been let  before the mitigation system was  added to the design.
The additional cost of  the radon mitigation  system was covered by
four change orders  for which  the construction contractor charged
$5,367.   Thus,  the system cost  $0.096  per sq ft of floor  space.
Specifications had already called for 4 in. of aggregate under the
slab,  and  there  was   no   charge  for  the  change  in  aggregate
specification.  The other three change  orders covered installation
of the suction pit and  stack to  the roof,  sealing of all pour and
control saw joints with a polyurethane  caulking,  and  installation
of the fan and warning  system.

     A recent study of costs of mitigation  in eight new schools
recently  built gave  costs from $0.35 to $1.10 per sq ft, as much
as  $46,000  for  one  school building  (5).    Hence,  the installed
mitigation system at the Johnson City Hospital cost only a fraction
of the cost of systems currently being installed  in new schools in
the U.S.

                            CONCLUSIONS


     A  low  cost,  single  point  ASD  system,  installed   during
construction,  has  lowered radon  levels  in a  new  single story
hospital  building  to near ambient levels of less than 0.5  pCi/L,
the detection limit of the radon test used.    Levels greater  than
50 pCi/L  were measured in  the building  with  both the HVAC and ASD
systems off and as high as 16 pCi/L with the  HVAC system operating
and  the ASD  system off.   ingredients of this  radon mitigation
system are:

-------
  1. Slab-on-grade post and beam construction with no barriers
     to soil gas flow below the slab.

  2. Continuous layer of coarse, narrow particle size range crushed
     aggregate a minimum of 4 in. thick under the slab.  (ASTM #5
     is the preferred stone.)

  3. Careful sealing  of all  slab cracks and penetrations and the
     use  of a 6  mil  plastic  film between  the slab and  the
     aggregate.

  4. A specially  designed subs lab  suction  pit having  a  void to
     aggregate  interface  area of at least 5  sq ft  and a  6 in.
     diameter stack to the roof.

  6. An exhaust fan (on the stack)  capable of exhausting a
     minimum of 500 cfm at no head.

     Incremental cost for  the mitigation system was less than $0.10
per sq ft compared to a cost range of $0.35  to  $1.05 per sq ft for
eight  schools  recently built  in the U.S.  with more complicated
radon mitigation systems.

     Mitigation systems being  designed for new schools and other
large buildings by architects and/or mitigators are more extensive
and costly than they  need to be.  This has resulted from the lack
of field-demonstrated technology  (by EPA or others).  Application
of  the technology  to be recommended  in EPA's  new  construction
technical  guidance manual  (under  development and  named  "Radon
Prevention in the  Design of Schools and Other Large Buildings") has
the potential to reduce both cost and complexity of these systems
yet achieve very low  radon levels.

                            REFERENCES


1.   Craig, A. B., K. W.  Leovic, and D.  B. Harris,  Design of Radon
     Resistant  and  Easy-to-Mitigate New  School Buildings.   In
     Proceedings:   The 1991  International Symposium on Radon and
     Radon Reduction Technology, Volume l, EPA-600/9-9l-037b  (NTIS
     PB92-115369),  November 1991.

2.   ASTH-33-86 "Standard Specifications for Concrete Aggregate,"
     American  Society for Testing Materials.  May  1986.

3.   ASHRAE-62-1989.    Ventilation For  Acceptable   Indoor  Air
     Quality.   American Society  of  Heating, Refrigerating and Air
     Conditioning  Engineers, Inc.   1989.

-------
4.   Architecture/ Research, Volume 1, Number 1, October 1991.  A
     publication  of   the  AIA/ACSA   Council   on  Architectural
     Research.

5.   Craig,  A.  B,,  K.  W.  Leovic,  and D.  W.  saum,  Cost  and
     Effectiveness  of  Radon Resistant  Features  in New  School
     Buildings,  Healthy  Buildings—IAQ'91, Washington,  D.  C.,
     September 4-8, 1991.

                        CONVERSION FACTORS


Readers more familiar with the metric system may use the following
factors to convert to that  system.

     Non-metric          Multiplied bv            Yields Metric
        cfm                  0.00047                   mx/s
        ft                   0.30                      m
        ft2                  0.093                     m2
        gal.                 0.0038                   mx
        HP                   7.46                      W
        in.                  0.025                     m
        in. we              249                         Pa
        mil                  0.000025                 m
        pCi/L                37                         Bq/ra3

-------
SECTION A
                              •   6" PLASTIC PIPE
                    4'X4'X3/4- TREATED
                      PLYWOOD SHEET
TO STACK
TO ROOF
                                                   8"X8*X8"
                                                   CONCRETE BLOCK

                         SECTION A
                 Figure 1. Subslab suction pit

-------
      I I

    0.180"
       -0.180"
       185'
 -0.193"
J
     n
     A
   102'
                 H
121'
                           U-0.186'
                   •0.188"
                                         173'
                          -0.204"
              -0.225
                                "
                          -0.256"
                          -0.468
                                "
                                     106'
                               -0.2061
                       108'
                                 113'
            i
              -0.215
            Figure 2,  Pressure field extension,

           Johnson City Rehabilitation Hospital

-------
        SUBSLAB PRESSURE, Neg .  inch W.  C.

Figure 3.  Plotted  pressure field extension,
     Oohnson City Rehabilitation Hospital

-------
                                    Ki nder-
                                    garten
                                    Room
        Figure 4.  Subslab walls ,
                 Dranesville Elementary
ConfRm  4    ASD  Slab  Penetration
\
' »«\ u<
• D -0.021 "\
tu \

\
r ".Q 148" i
V« I *TW |
1L5 •" — ""
ttt
111"
\ / -0.006"
Stack -0.99"
117 US
-0.007" E •
f
• -0.66 " IM I
>•

110 1 m
-0.011" C. 3 H*Xo.005»
           Figure 5.  Pressure field extension,
                    Dranesville  Elementary

-------
X PFE measurement locations
                                                                          Suction pit
                                                                           Exhaust fan
                        Figure 6.   Pressure field extension  measurements
                                    Dunn Elementary School

-------
 I

II
M
I)
u
II
L
•I
II
CJ!
H
L
c>
>
<
    0.69
    o.oe-
                       '.••"•'.•• •
                                             Pit Size (depth  x  diameter,


                                                            2x3



                                                            2x2


                                                            1  X  1



                                                           0x0  (no pit)
<100                  100-  149                150-200

                 Distance From Suction  Pit  (ft)

         Figure 7.  Effect  of subslab suction pit size,

                   Dunn  Elementary School
                                                                                      >200

-------





1 	 1
\ X J
\s
/ \
Stack 13
X
\
*» .*
» * *
1 1 V
! \ /
1 *•- 1*

i



\

I— 1 V'*'* v*x /
u "Stack 18 Stack 19
1 /\ j» (Post & beam)
• * ^** 3
Stack 11* p--j



	 	 	 . . 	 	

E
1
'*" »«
10' *
i" ^tack 15 / 5
*• « —
IO «, ^ n
la V-' »
•
" .' V C.
.-' ••-. 2.
5
\
*% S UB^BBS^J "V •* *
\/
Stack 12
^>( St.ck\l7
St.'ck 16 P"~J| \
footing

s ?
Z GYM f
i *
(no post *°
& beam)

1







• 4
st,
'
\
\
*
ick It-^
Figure 8.  ASD stack and perforated pipe location, School 1

-------
                            * I -0.006"
                                210'
                            . K  -0.032"
                                 120'
              of  1/4" drilled 'subtUb"tirho>ts  -0.029" !
                                                 70'
         • M fno • *•«.)  """*"
                    1
        190'
       N -0.003"
I fi52iU?5A'^S*^rT;
                     S,t»ek IS   .•*
                       \    / - o
                  • J     V ./  -0.171"
                  -0.164"  X    26'
                   40'    •'   *«       • c
                              \    -0.120"
                                      42'
                                     -0.144"
                                     '190'.
"i   rr  . 8znrt>111
 UM^J I  *ft -0.094
      LI,!?-
                                                           •r
                                                                         * 0
                                                                       0.000
                                                                        175'
                                                           -0,018"-0.OlinJ -0,003"
                                                     .•*-<096
                                                            ..100
                                                          omi-o.oc
                                                          140',  150
                          Figure 9. Pressure field extension,
                                    from stack 5, School  1
GOVERNMENT PRINTING OFFICE: 1992—648-003/6O.020

-------