EPA/625/R-93/011
                                                 October 1993
 Radon Reduction Techniques for
     Existing Detached Houses


 Technical Guidance (Third Edition)
                    for
Active Soil Depressurization Systems
                D. Bruce Henschel

      Air and Energy Engineering Research Laboratory
Office of Environmental Engineering and Technology Demonstration
          Office of Research and Development
         U.S. Environmental Protection Agency
       Research Triangle Park, North Carolina 27711
                                        Printed on Recycled Paper

-------
                 EPA  Disclaimer Notice
The U.S. Environmental Protection Agency (EPA) strives to provide accurate, complete,
and useful information. However, neither EPA nor any person contributing to the prepara-
tion of this document makes any warranty, expressed or  implied, with respect to the
usefulness  or effectiveness of any  information, method,  or process disclosed in this
material. Nor does EPA assume any liability for the use of, or for damages arising from the
use of, any information, method, or process disclosed in this document.

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

-------
                                 Preface
This technical guidance document has been prepared to serve as a comprehensive aid in the
detailed selection, design, installation, and operation of indoor radon reduction measures
for existing houses based on active soil depressurization techniques. It is intended for use
by radon mitigation contractors, building contractors, concerned homeowners, state and
local officials, and other interested persons.

This document is the third edition of EPA's technical guidance for indoor radon reduction
techniques. This document addresses primarily radon reduction techniques based on active
soil depressurization technology, which is one of the most widely used approaches for
reducing radon in existing houses. The document also addresses active soil pressurization
and passive soil depressurization techniques, because these less widely used techniques
bear a number of similarities to active depressurization systems.

This edition incorporates additional and updated information on active soil depressuriza-
tion techniques, reflecting new results and perspectives that  have been obtained in this
developing field since the second edition of EPA's technical guidance (EPA/625/5-87/019)
was published in January 1988. Thus, this document should be viewed as replacing Section
5 ("Soil Ventilation") of the second edition.

This document does not provide guidance regarding indoor radon reduction techniques
other  than active soil depressurization (and active soil pressurization and passive soil
depressurization). Persons interested in other techniques, including house ventilation, entry
route  sealing, house pressure adjustments, air cleaners, and well water treatment, are
referred to the second edition of the technical guidance document.

Homeowners and occupants who are interested in a brief overview of the alternative radon
reduction techniques available, and of the steps to follow in getting a radon reduction
system installed in their home, are referred to the booklet entitled Consumer's Guide to
Radon Reduction, EPA-402-K92-003. Copies of that booklet,  and of the second and third
editions of the detailed technical guidance document, can be obtained  from the state
agencies and the EPA regional offices listed in Section 15. Copies of the second and third
editions of the technical guidance document can also be obtained from

                             ORD Publications  Office
                   Center for Environmental Research Information
                       U.S. Environmental Protection Agency
                        26 West Martin Luther King Drive
                           Cincinnati, OH 45268-1072

-------

-------
                                      Contents
Preface	iii
Figures	•	•	xi
Tables	xiii
Acknowledgments	xiv
Glossary	xv
Metric Equivalents	xxii
HOW TO USE THIS DOCUMENT	H-l

 1  Introduction..	1
    1.1    Purpose	1
    1.2   Scope	2
    1.3    Content	2
    1.4   Reason for Focus on Active Soil Depressurization	3

 2  Principles, Applicability, and Past Performance of Soil Depressurization Systems	5
    2.1   Principles of Active Soil Depressurization	5
          2.1.1   m Active Sub-Slab Depressurization (SSD)	7
          2.1.2   Active Sump/Drain-Tile Depressurization (Sump/DTD)	7
          2.1.3    Active Drain-Tile Depressurization (Above-Grade/Dry-Well Discharge)	11
          2.1.4   Active Block-Wall Depressurization (BWD)	13
     :     2.1.5    Active Sub-Membrane Depressurization (SMD) in Crawl Spaces	13
    2.2   Applicability of Active Soil Depressurization	18
          2.2.1   Active Sub-Slab Depressurization	19
          2.2.2   Active Sump/Drain-Tile Depressurization	20
          2.2.3   Active Drain-Tile Depressurization (Above-Grade/Dry-Well Discharge)	 21
          2.2.4   Active Block-Wall Depressurization	22
          2.2.5   Active Sub-Membrane Depressurization	23
     2.3   Performance of Active Soil Depressurization Systems	24
          2.3.1   Active Sub-Slab Depressurization	25
          2.3.2   Active Sump/Drain-Tile Depressurization	37
          2.3.3   Active Drain-Tile Depressurization (Above-Grade/Dry-Well Discharge)	41
           2.3.4  thrive Block-Wall Depressurization	44
           2.3.5   TAc&ye Sub-Membrane Depressurization	52
     2.4   Performanceof Active Soil Pressurization Systems	64
           2.4.1    Acti\ Sub-Slab Pressurization	65
           2.4.2   Active Block-Waill Pressurization	67
     2.5   Performance of Passive Soil Depressurization Systems	67

-------
                            Contents (continued)
3  Pre-Mitigation Diagnostic Test Procedures for Soil Depressurization Systems	73
   3.1   General	73
         3.1.1   Purposes of Pre-Mitigation Diagnostics	73
         3.1.2   Diagnostic Tests That Can Be Considered	73
   3.2   Procedures for the Visual Survey	76
   3.3   Procedures for Sub-Slab Suction Field Extension Measurements	81
         3.3.1   Qualitative Assessment of Suction Field Extension	83
         3.3.2   Quantitative Assessment of Suction Field Extension	88
   3.4   Procedures for Radon Grab Sampling and Sniffing	92
   3.5   Procedures for Other Types of Diagnostic Tests	98
         3.5.1   Sub-Slab Flows	98
         3.5.2   Well Water Radon Analysis	100
         3.5.3   Measurements to Determine Significance of Building Materials
                as a Radon Source	101
         3.5.4   Pressure Differential Measurements Across the House Shell	103
         3.5.5   Blower Door Measurements	106
         3.5.6   Tracer Gas Testing	107

4  Design and Installation of Active Sub-Slab Depressurization Systems	109
   4.1   Selection of the Number of Suction Pipes	109
         4.1.1   Houses Having Good Sub-Slab Communication	109
         4.1.2   Houses Having Marginal or Poor Sub-Slab Communication	110
   4.2   Selection of Suction Pipe Location	112
         4.2.1   Houses Having Good Sub-Slab Communication	112
         4.2.2   Houses Having Marginal or Poor Sub-Slab Communication	114
   43   Selection of Suction Pipe Type and Diameter	115
         4.3.1   Type of Suction Pipe	115
         4.3.2   Diameter of Suction Pipe	116
   4.4   Selection of the Suction Fan	118
         4.4.1   Centrifugal In-Line Tubular Fans	120
         4.4.2   Li-Line Radial Blowers	123
         4.4.3   High-Suction/Low-Flow Blowers	123
   4.5   Installation of Suction Pipes Beneath the Slab	125
         4.5.1   Vertical Pipe Installed Down Through Drilled Hole Indoors	125
         4.5.2   Vertical Pipe Installed Down Through Large Hole Indoors (Large Sub-Slab Pit).... 130
         4.5.3   Vertical Pipe Installed Into Unused Sump Pit Indoors	134
         4.5.4   Horizontal Pipe Installed Through Foundation Wall from Outdoors	134
         4.5.5   Horizontal Pipe Installed Through Foundation Wall from Inside Basement	137
         4.5.6   Suction Pipe Installed from Inside Garage	 137
   4.6   Design/Installation of the Piping Network and Fan	140
         4.6.1   Suction Loss in the Piping Network	140
         4.6.2   Considerations in Pipe Routing Between Suction Points and Fan	145
         4.6.3   Considerations in Pipe Installation Between Suction Points and Fan	148
                                             vi

-------
                            Contents (continued)
         4.6.4   Design and Installation of the Fan and Exhaust Piping—Interior Stacks	151
         4.6.5   Design and Installation of the Fan and Exhaust Piping—
                Exterior and Garage Stacks	4	162
   4.7   Slab Sealing in Conjunction with SSD Systems	176
         4.7.1   Perimeter Channel Drains	177
     ;    4.7.2   Other Major Holes Through Slab	 180
         4.7.3   Untrapped Floor Drains Connected to the Soil	181
         4.7.4   Sumps	.......	181
         4.7.5   Intermediate Openings Through the Slab	182
         4.7.6   Openings in Block Foundation Walls	.,	183
   4.8   Gauges/Alarms and Labelling	183
         4.8.1   Gauges and Alarms	184
         4.8.2   System Labelling	185

5  Design and Installation of Active Drain-Tile Depressurization Systems
   (Sump Depressurization)	187
   5.1   Selection of the Number of Suction Pipes:
         Need for a SSD Component to Sump/DTD System	187
   5.2   Selection of DTD Suction Pipe Location:  At Sump or Remote	188
   5.3   Selection of Suction Pipe Type and Diameter	189
   5.4   Selection of Suction Fan	190
   5.5   Installation of a Suction Pipe into the Drain Tile Loop	190
         5.5.1   Suction Drawn on Tiles Remote  from Sump	190
         5.5.2   Suction Drawn on Capped Sump	191
   5.6   Design/Installation of the Piping Network and Fan	195
   5.7   Slab Sealing in Conjunction with Sump/DTD Systems	195
   5.8   Gauges/Alarms and Labelling	196

6  Design and Installation of Active Drain-Tile Depressurization Systems
   (Above-Grade/Dry-Well Discharge)	:	197
   6.1   Selection of the Number of Suction Pipes: Need for a SSD Component to the
         DTD/Remote Discharge System	198
   6.2   Selection of DTD Suction Pipe Location	199
   6.3   Selection of Suction Pipe Type and Diameter.	200
   6.4   Selection of the Suction Fan	200
   6.5   Installation of a Suction Pipe into the Drain Tiles	201
   6.6   Design/Installation of the Piping Network and Fan	;	203
   6.7   Slab Sealing in Conjunction with DTD/Remote Discharge Systems	204
   6.8   Gauges/Alarms and Labelling	204

7  Design and Installation of Active Block-Wall Depressurization Systems	207
   7.1   Selection of the Number of Suction Pipes	210
         7.1.1   Individual-Pipe Variation	210
         7.1.2   Baseboard Duct Variation	210
                                             vii

-------
                             Contents (continued)
   7.2   Selection of Suction Pipe Location	211
         7.2.1    Individual-Pipe Variation	211
         7.2.2    Baseboard Duct Variation	213
   7.3   Selection of Suction Pipe Type and Diameter	214
         7.3.1    Individual-Pipe Variation	214
         7.3.2    Baseboard Duct Variation	214
   7.4   Selection of the Suction Fan	215
   7.5   Installation of SuctionPipes Into the Block Walls	216
         7.5.1    Individual-Pipe Variation	216
         7.5.2    Baseboard Duct Variation	217
   7.6   Design/Installation of the Piping Network and Fan	219
   7.7   Wall and Slab Sealing in Conjunction with BWD Systems	220
         7.7.1    Wall Openings	220
         7.7.2    Slab Openings	223
   7.8   Gauges/Alarms and Labelling	223

8  Design and Installation of Active Sub-Membrane Depressurization Systems	225
   8.1   Selection of the Approach for Distributing Suction, and the Number of Suction Pipes	225
         8.1.1    Approach for Distributing Suction	225
         8.1.2    Number of Suction Pipes	227
   8.2   Selection of Suction Pipe Location	228
         8.2.1    Individual-Pipe/SMD Approach	228
         8.2.2    Sub-Membrane Piping/SMD Approach	228
   8.3   Selection of Suction Pipe Type and Diameter	229
         8.3.1    Individual-Pipe/SMD Approach	229
         8.3.2    Sub-Membrane Piping/SMD Approach	230
   8.4   Selection of the Suction Fan	231
   8.5   Installation of the Membrane and the Suction Pipes	232
         8.5.1    Installation of the Membrane	232
         8.5.2    Installation of the SuctionPipes	235
   8.6   Design/Installation of the Piping Network and Fan	241
   8.7   Sealing in Conjunction with SMD Systems	242
         8.7.1    Sealing the Membrane	242
         8.7.2    Sealing the Block Foundation Wall	242
   8.8   Gauges/Alarms and Labelling	242

9  Considerations for Active Soil Pressurization Systems	245
   9.1   Selection of the Number of Pressurization Pipes	245
         9.1.1    Sub-Slab Systems	245
         9.1.2    Block-Wall Systems	247
   9.2   Selection of Pressurization Pipe Location	247
   9.3   Selection of Pressurization Pipe Type and Diameter	247
   9.4   Selection of Pressurization Fan	,	248
                                             viii

-------
                            Contents (continued)
   9.5   Installation of Pressurization Pipes Beneath the Slab or Into the Block Walls	248
         9.5.1    Sub-Slab Systems	248
         9.5.2    Block-Wall Systems	249
   9.6   Design/Installation of Piping Network and Fan	249
   9.7   Sealing in Conjunction with Soil Pressurization Systems	251
   9.8   Gauges/Alarms and Labelling	251

10 Considerations for Passive Soil Depressurization Systems	253
   10.1  Selection of the Number of Suction Pipes	253
         10.1.1   Passive SSD Systems	253
         10.1.2   Passive DTD Systems	254
         10.1.3   Passive SMD Systems	254
   10.2  Selection of Suction Pipe Location	254
         10.2.1   Passive SSD Systems	;.....	254
         10.2.2   Passive DTD Systems	255
         10.2.3   Passive SMD Systems	255
   10.3  Selection of Suction Pipe Type and Diameter	256
   10.4  Selection of a Supplemental Suction Fan	256
   10.5  Installation of Suction Pipes	256
         10.5.1   Passive SSD Systems	256
         10.5.2   Passive DTD Systems	256
         10.5.3   Passive SMD Systems	257
   10.6  Design/Installation of Piping Network	257
   10.7  Sealing in Conjunction with Passive Soil Depressurization Systems	259
   10.8  Gauges/Alarms and Labelling	259

11 Post-Mitigation Diagnostic Test Procedures for Soil Depressurization Systems	261
   11.1  General	 261
         11.1.1   Purposes of Post-Mitigation Diagnostics	261
         11.1.2   Diagnostic Tests that Can Be Considered	261
   11.2  Procedures for the Visual Inspection	263
   11.3  Procedures for Suction (and Flow) Measurements in the System Piping	263
   11.4  Procedures for Indoor Radon Measurements	265
         11.4.1   Initial Short-Term Radon Measurement	265
         11.4.2   Subsequent Radon Measurements	265
   11.5  Procedures for Checking Combustion Appliance Backdrafting	265
         11.5.1   Backdrafting Test Procedures for Mitigators	266
         11.5.2   Backup Backdrafting Test Methods for House Occupants	268
         11.5.3   Steps Required'When Backdrafting Is Observed	268
   11.6  Procedures for Suction Field Extension Measurements	269
         11.6.1   Qualitative Check with Chemical Smoke	269
         11.6.2   Quantitative Measurement with Micromanometer	270
   11.7  Procedures for Radon Grab Sampling and Sniffing	272
   11.8  Procedures for Chemical Smoke Flow Visualization Tests	273

                                              ix

-------
                            Contents  (continued)
   11.9  Procedures to Test for Exhaust Re-Entrainment	273
   11.10 Procedures for Well Water Radon Analysis	274
   11.11 Procedures to Determine the Significance of Building Materials as a Radon Source	274

12 Operation and Maintenance Requirements for Active Soil Depressuri/ation Systems	275
   12.1  Instructions Following Installation	275
   12.2  Operating and Maintenance Requirements	275
         12.2.1  System Fan	275
         12.2.2  Piping Network, and System and Foundation Seals	277
         12.2.3  Follow-Up Indoor Radon Measurements	,	278

13 Installation and Operating Costs for Active Soil Depressurization Systems	281
   13.1  Sub-Slab Depressurization Costs	282
         13.1.1  SSD Installation Costs	282
         13.1.2  SSD Operating Costs	283
   13.2  Sump/Drain-Tile Depressurization Costs	.,	284
         13.2.1  Sump/DTD Installation Costs	284
         13.2.2  Sump/DTD Operating Costs	284
   13.3  Drain-Tile Depressurization/Remote Discharge Costs	285
         13.3.1  DTD/Remote Discharge Installation Costs	285
         13.3.2  DTD/Remote Discharge Operating Costs	285
   13.4  Block-Wall Depressurization Costs	285
         13.4.1  BWD Installation Costs	285
         13.4.2  BWD Operating Costs	286
   13.5  Sub-Membrane Depressurization Costs	286
         13.5.1  SMD Installation Costs	286
         13.5.2  SMD Operating Costs	287
   13.6  Active Soil Pressurization Costs	288
         13.6.1  Soil Pressurization Installation Costs	288
         13.6.2  Soil Pressurization Operating Costs	288
   13.7  Passive Soil Depressurization Costs	288
         13.7.1  Passive Soil Depressurization Installation Costs	;	288
         13.7.2  Passive Soil Depressurization Operating Costs	289
   13.8  Summary of ASD Installation and Operating Costs	289

14 References	'.	291

15 Sources of Information	299

-------
                                         Figures
H-l   A summary of the steps to be followed in using this document.	H-2
   1   Sub-slab depressurization (SSD) using pipes inserted down through the slab from indoors	8
   2   Sub-slab depressurization (SSD) using pipes inserted horizontally through the foundation
      wall from outdoors	••— 9
   3   Drain-tile depressurization (DTD) where the tiles drain to a sump in the basement	10
   4   Drain-tile depressurization (DTD) where the tiles drain to an above-grade discharge
      remote from the house.	12
   5   Block-wall depressurization (BWD) using the individual-pipe approach	 14
   6   Sub-membrane depressurization (SMD) for the case where individual suction pipes
      penetrate the membrane (SSD analogue)	15
   7   Sub-membrane depressurization (SMD) for the case where suction is drawn on
      perforated piping beneath the membrane (DTD analogue)	16
   8   Example of a visual survey form	77
   9   Experimental configuration for quantitative pre-mitigation sub-slab suction field
      extension.and flow diagnostics using a vacuum cleaner	83
  10  Example of siting vacuum cleaner suction hole and sub-slab suction test holes for
      quantitative suction field extension testing, when slab is fully finished..:	89
  11  Example of graphical interpretation of the results torn quantitative suction field
      extension measurements	»	91
  12  Procedure for calculating sample radon concentrations from measured counts
      when using the alpha scintillation cell technique for grab sampling	,	96
  13   Suction loss per 100 linear ft of straight piping, as a function of gas volumetric flow rate
       and pipe diameter, for circular pipe having smooth walls	117
   14   Alternative approaches for installing the SSDsuction pipe in the slab, in cases where
       the weight of the suction pipe is not supported at the slab penetration	127
   15   Alternative approaches for installing the SSD suction pipe in the slab, in cases where
       the weight of the suction pipe is supported at the slab penetration	128
   16   Some alternative approaches for supporting the weight of a suction pipe when there is
       a horizontal piping run, in cases where the pipe is not supported at the slab penetration	131
   17   Some alternative approaches for supporting tfe weight of a suction pipe when the pipe
       rises directly up through the overhead flooring with no horizontal run	132
   18   A typical configuration for treating an adjoiijing slab on grade using a horizontal suction
       pipe penetrating the stem wall from inside debasement	138
                                               ,^~"~
   19  One possible method for angling a suction pij| beneath the living-area slab of a
        slab-on-grade  house, from inside an adjoining{arage	,	139
   20   One representative SSD piping configuration i|ustrating an interior exhaust stack	141
   21   One representative SSD piping configuration illustrating an exterior exhaust stack	142
   22   Two of the alternative methods for providing proper drainage when a horizontal piping
        run must be routed over or under obstructions, creating low points in the piping	147
                                                XI

-------
                              Figures (continued)
 23   A representative method for mounting a fan in the attic above an interior stack	155
 24   Some options for supporting a fan and exhaust piping in an attic in cases where
      there is no horizontal piping run in the attic.	157
 25   Some alternative approaches that can sometimes be considered for wiring SSD fans
      mounted in attics	159
 26   Two possible configurations for vent caps on SSD stacks	160
 27   A representative method for mounting an exterior fan at the base of an exterior stack	164
 28   Some of the alternative methods for offsetting exterior stacks against the house	167
 29   Some alternative methods for supporting an exterior stack against the side of a house	169
 30   Options when an exterior stack reaches the roof overhang	170
 31   Some alternative methods for supporting exterior stacks where they are routed around
      aroof overhang	171
32A  Cross-section of a sealed perimeter channel drain, illustrating water drainage channels
      both above and below the seal	178
32B   Method for sealing the sump and the slab channel leading to the sump in cases where
      perimeter channel drains empty into a sump in the basement	179
 33   Some alternative sump cover designs (illustrated for the case where suction is being
      drawn at the sump)	193
 34   Some approaches for connecting individual BWD suction pipes into a SSD system	208
 35   Block-wall depressurization (BWD) using the baseboard-duct approach	209
 36   One specific example of a baseboard duct BWD configuration with a major
      SSD component, for the case  where a sump and sump pump are installed
      as part of the radon mitigation system	218
 37   Some options for closing major wall openings at the top of block foundation walls,
      in conjunction with BWD systems	221
 38   Some alternative approaches for increasing block wall treatment by a SMD system	236
 39   Some of the alternative  approaches for installing individual SMD suction pipes through
      the membrane	238
 40   Sub-slab pressurization  using  one typical approach	246
                                              xii

-------
                                     Tables
1,  Examples of In-Line Tubular Fans Which Have Been Marketed for Radon Mitigation	121
2  Performance Characteristics of Some In-Line Radial Blowers Being Marketed for
   Radon Reduction Systems	123
3  Performance Characteristics of Some High-Suction/Low-Flow Blowers Suitable for
   Radon Reduction Systems	124
4  Number of Feet of Straight Pipe Required to Create the Same Suction Loss As Created
   by the Flow Obstruction in One Fitting	143
5  Assumptions Used in Estimating Annual Operating Costs for ASD Systems	282
6  Summary of Installation and Operating Costs for ASD and Related Systems	290
7  Radon Contacts for Individual States	300
8  Radiation Contacts for EPA Regional Offices	303
                                           XIII

-------
                      Acknowledgments
These acknowledgments must begin by recognizing the contributions of the large number
of researchers and mitigators who, through their efforts over the years, have defined the
state of the art of radon mitigation technology. This manual is a documentation of the
combined contributions of these many individuals. Many of these individuals are recog-
nized through citations that appear in the list of references (Section 14), although it was not
possible through this mechanism to acknowledge everyone who has aided in the advance-
ment of this field.

Special recognition is given to several individuals who devoted particularly large amounts
of time in reviewing the draft of this document and in providing consultation to help ensure
that the manual rigorously reflects field experience around the U.S. These individuals
include Terry M.  Brennan of Camroden Associates, Inc., Oriskany, NY; William P.
Brodhead of WPB Enterprises, Inc., Riegelsville, PA; Douglas L. Kladder and Steven R.
Jelinek of Colorado Vintage Companies, Inc., Colorado Springs, CO;  Marc Messing of
Infiltec Radon Control,  Falls Church, VA;  and John  W. Anderson, Jr., and Jack  C.
Bartholomew, Jr., of Quality Conservation, Spokane, WA.

Appreciation is also expressed to the large number of other individuals  who reviewed the
draft of the manual. Their comments have significantly improved the accuracy and
completeness of this document Of the reviewers outside of EPA, the author is particularly
indebted to William J. Angell, Director of the Midwest Universities Radon Consortium,
University of Minnesota; Richard S. Bairibridge of Aarden Testing; Kenneth J. Gadsby of
Princeton University; David H. Saunders of the National Association of Home Builders
Research Center, Arthur G. Scott of Arthur Scott and Associates; and Bradley H. Turk of
Mountain West Technical Associates.

Of the reviewers within the Agency, particular appreciation is expressed to David J. Price
and David M. Murane of the Radon Division,  Office of Radiation and Indoor Air; and
representatives of EPA's regional offices, especially Larainne G. Koehler of Region 2,
Lewis K. Felleisen of Region 3, Paul Wagner of Region 4, James C. Benetti of Region 5,
Stephen G. Chambers of Region 7, Philip C. Nyberg of Region 8, and  David Holguin of
Region 9.

The author also wishes to express particular appreciation to Frank T. Princiotta, Everett L.
Plyler, Michael C. Osborne, and Timothy M. Dyess of EPA's Air and Energy Engineering
Research Laboratory, for  their  technical input, guidance, and support throughout the
preparation of this document
                                      xiv

-------
                                                Glossary
ABS (acrylonitrile butadiene styrene)—A plastic that is resis-
   tant to deterioration (e.g., by soil chemicals), similar to
   PVC. ABS is used to make rigid piping that is commonly
   used; e.g., for residential sewer lines and for perforated
   drain tiles.

Active soil depressurization (ASD)—A class of techniques
   for reducing radon concentrations inside buildings. These
   techniques function by drawing radon-containing soil gas
   away from the  foundation  and exhausting it outdoors
   before it can enter the building.

Aggregate—As used here, aggregate refers to gravel or crushed
   rock that is placed beneath concrete slabs during construc-
   tion to provide an even, well-supported base for the con-
   crete and  to provide a capillary break for moisture pur-
   poses. The term  "gravel" may refer to crushed rock (e.g.,
   pea gravel) or to naturally occurring material (e.g., river-
   run gravel).  The presence of  sub-slab aggregate often
   results in good sub-slab communication. The optimal ag-
   gregate  from the standpoint of radon mitigation is clean,
   coarse aggregate, without substantial fine material to block
   the open spaces between the larger rocks.

   The term "aggregate" is also sometimes used in some
   areas to refer to  sand or sand/pebble mixtures, which can
   also be used to support slabs and provide a capillary break.
   However, in this document, the term is used only to refer
   to gravel or crushed rock.

Air changes per hour (ach)—The number of times within 1
   hour that; the volume of air inside a house would nominally
   be replaced, given the rate at which outdoor air is infiltrat-
   ing the house. If a house has 1 ach, it means that all of the
   air in the house will  be nominally replaced in a  1-hour
   period.

Alarm—As used here, a device that gives a visual or auditory
   signal (such as a light or a buzzer) when the suction in an
   ASD system moves outside the acceptable operating range
   for that system.  An alarm may or may not also include a
   gauge to provide a reading of the actual suction in the
   system.

Alpha particles—A positively charged sub-atomic particle,
   comparable to the nucleus of a helium atom, emitted
   during  decay of certain radioactive elements, such as
   radon and some of its progeny.  The type of radiation
   responsible for the lung cancer risk associated with radon
   decay products. Many of the measurement devices used to
   detect radon are based on the detection of alpha particles.
Backdrafting (of combustion appliances)—A condition where
   the normal movement of combustion products up a flue,
   resulting from the buoyant forces on the hot gases,  is
   reversed, so that the combustion products can enter the
   house. Backdrafting of combustion appliances (such as
   fireplaces and furnaces) can occur when depressurization
   in the house overwhelms  the buoyant force on the hot
   gases. Backdrafting can also be caused by high air pres-
   sures at the chimney or flue termination.

Backer rod—A compressible, closed-cell polyethylene foam
   material, which is formed into ropes or cords of alternative
   diameters. Backer rod can be force-fit into wide cracks and
   similar openings to serve as a support for caulking mate-
   rial.

Band joist—Also called header joist, header plate, or rim
   joist. A board (typically 2 x 10 in.1) that rests (on its 2-in.
   dimension) on top of the sill plate around the perimeter of
   the house. The ends of the floor joists are nailed into the
   header joist that maintains spacing between the floor joists.

Baseboard duct—A continuous system of  sheet metal or
   plastic channel ducting that is sealed over the joint be-
   tween the wall and floor around the entire perimeter of the
   basement Holes drilled into hollow blocks  in the wall
   allow suction to be drawn on the walls and joint to remove
   radon through the ducts to a release point away from the
   inside of the house.

Basement—A type of house construction where the bottom
   livable level has a slab (or occasionally an earthen floor)
   that averages 3 ft  or more below grade level on one or
   more sides of the house.

Block cavities, block voids—The air space(s) within concrete
   block or cinder block often used to construct foundation
   walls. The block cavities form an interconnected matrix
   within a finished wall.

Block-wall depressurization (BWD)—A variation of the ASD
   technology, where the system is attempting to depressur-
   ize the interconnected cavities inside the hollow-block
   foundation wall.
     Readers more familiar with metric units may use the equivalents
     listed at the end of the front matter.
                                                         xv

-------
                                       Glossary  (continued)
Blower door—A device consisting of an instrumented fan that
   can be mounted in an existing doorway of a house. By
   determining  the air flows through this fan required to
   achieve different degrees of house depressurization, the
   blower door permits determination of the tightness of the
   house shell, and an estimation of the natural filtration rate.

Chemical smoke—An inert fine powder, resembling smoke,
   which is released at selected locations during diagnostics
   in order to visualize the direction of air movement at those
   locations. Chemical smoke might be used, for example, to
   determine whether soil  gas  appears to be entering the
   house through selected  openings in the slab. Chemical
   smoke can be dispensed from specially designed guns,
   bottles, or tubes, often by squeezing a rubber bulb on one
   end of the device or by squeezing the sides of the plastic
   bottle.

Cold air return—The registers and ducting that withdraw
   house air from various parts of the house and direct it to a
   central forced-air furnace or heat pump. The return ducting
   is at low pressure relative to the house because the central
   furnace fan draws air out  of the house through this ducting.

Communication (as in "sub-slab communication")—A mea-
   sure of how well openings beneath the slab (e.g., through
   porous gravel or soil under the slab) connect the sub-slab
   region, permitting suctions (or flows) generated  at one
   point to extend to other points beneath the slab. Sub-slab
   communication is classified here in three categories: good,
   marginal, and poor. The concept of communication can
   also be applied to communication between the sub-slab
   region and the cavities in block foundation walls, commu-
   nication beneath crawl-space membranes, etc.

Contractor (as  in  "radon contractor")—A  building trades
   professional who works  for profit to correct radon prob-
   lems;  a radon remediation expert. Also referred to as a
   radon mitigator. Through EPA's Radon Contractor Profi-
   ciency Program (RCPP), contractors can voluntarily dem-
   onstrate their proficiency. Some state radiological health
   offices also maintain lists of qualified professionals.

Conveclive movement—As used here, the bulk flow of radon-
   containing soil gas into the house as the result of pressure
   differences between the house and the soil. Distinguished
   from diffusive movement.

Coring drill—A. large power drill that can cut circular cores
   (e.g., of 4- to 5-in. diameter) out of concrete slabs. Coring
   drills can be operated dry (e.g., with a carbide bit) or wet
   (e.g., with a diamond bit).

Crawl space—An area beneath the  living space in some
   houses, where the floor of the lowest living area is elevated
   above grade level. This  space (which generally provides
   only enough head room  for a person to crawl in) is not
   living  space  but  often contains utilities. Distinguished
   from slab-on-grade or basement construction.

Crawl-space depressurization—A radon reduction approach
   that has sometimes been applied to crawl-space houses,
   where an exhaust fan (blowing crawl-space air outdoors)
   causes the crawl space to become depressurized relative to
   the living area above. This approach prevents radon-con-
   taining crawl-space air from flowing up into the living
   area. Appears to be second only to SMD as an effective
   alternative for treating crawl-space houses.

Cubic feet per minute (cfm)—A measure of the volume of a
   fluid (measured in cubic  feet) flowing within  a fixed
   period of time (expressed in minutes).

Depressurization—In houses, a condition that exists when the
   air pressure inside the house is slightly lower than the air
   pressure outside or the soil gas pressure. The lower levels
   of houses are essentially always depressurized during cold
   weather because of the buoyant force on the warm indoor
   air (creating the natural thermal stack effect). Houses can
   also be depressurized by winds and by appliances that
   exhaust indoor air. ASD systems attempt to depressurize
   the soil (i.e.,  to reduce the soil gas pressure to a value
   lower than the pressure in the house).

Detached houses—Single  family dwellings as opposed to
   apartments, duplexes, townhouses, or condominiums. Those
   dwellings that are typically occupied by one family unit
   and which do not share foundations and/or walls with
   other family dwellings.

Diagnostic testing—Tests conducted  before or after the in-
   stallation of a radon reduction system to aid in deciding
   which radon reduction technology to use, designing  the
   selected system, or evaluating the reasons why an installed
   system is not performing as anticipated.

Diffusive movement—The random movement of individual
   atoms or molecules, such as radon atoms, in the absence of
   (or independent of) bulk (convective) gas flow. Atoms of
   radon can diffuse through tiny openings or even through
   unbroken concrete slabs. Distinguished from convective
   movement.

Drain tile—Perforated piping, usually constructed of flexible
   corrugated black high-density polyethylene or polypropy-
   lene, or of rigid ABS, PVC, or baked clay. Such tiles are
   buried beside the foundation of the house to collect water
   around the foundation and route it away from the founda-
   tion via a sump or a remote discharge.

Drain-tile depressurization (DTD)—A variation pf the ASD
   technology, where the area around the foundation is  de-
   pressurized by drawing suction on  drain tiles.
                                                       XVI

-------
                                      Glossary  (continued)
Effective leakage area—A parameter determined from blower
   door testing, giving a measure of the tightness of the house
   shell. Conceptually, this leakage area reflects the square
   inches of open area through the house shell, through which
   air can infiltrate or exfiltrate.

Entry routes—Pathways by which soil gas can flow into a
   house. Openings through the flooring and walls where the
   house contacts the soil.

EPDM (ethylene propylene diene monomer; a terpolymer of
   that monomer)—A heavy rubberized membrane used for
   waterproofing flat roofs as a substitute for built-up tar-
   and-felt roofs. For SMD systems in  crawl-space houses,
   EPDM is  one logical material to be laid on top of the
   polyethylene sheeting along the routes of expected foot
   traffic within the  crawl space to protect the polyethylene
   from being punctured.

Exfiltration—The movement of indoor air out of the house.

Exhaust fan—A fan oriented so that it blows indoor air out of
   the house. Exhaust fans cause outdoor air (and soil gas) to
   infiltrate at other locations in the house, to compensate for
   the exhausted air.

Expansion joint—A gap  through a  concrete slab, usually
   about 1/2-in. wide and filled with  asphalt-impregnated
   fibrous material. In some regions, such joints are installed
   around the slab perimeter as the wall/floor joint. In other
   cases, they are installed across the middle of the slab
   (perpendicular to  the front and rear walls). They are re-
   ferred to as expansion joints because they would compress
   if the slab ever expanded. They would also reduce crack-
   ing if a segment of the slab shifted vertically relative to the
   foundation walls or relative to another segment of the slab.

Flowable caulk—Refers to caulks (often urethane caulks in
   this document) that are sufficiently  fluid such that they
   tend to flow like a viscous liquid prior to curing. Flowable
   caulks have the  advantage of flowing  into cracks and
   irregularities in the opening being sealed, thus forming an
   effective seal.

Footing(s)—A concrete or stone base, supporting a founda-
   tion wall, that is used to distribute the weight of the house
   over the soil or subgrade underlying the house.

Forced-air furnace  (or heat pump)—A central furnace or
   heat pump that functions by recirculating the house air
   through a  heat exchanger  in the furnace. A forced-air
   furnace is  distinguished from a central hot-water space
   heating system or electric resistance heating.

Furring strip—A small strip of wood (usually 1- by 2-in. or
   1- by 4-in.) that  is commonly attached vertically to the
   interior of block or poured  concrete foundation walls
   inside basements  to support interior panelling being in-
   stalled over the foundation walls; used in lieu of standard
   2- by 4-in. studs. In radon mitigation, one occasional use
   of furring strips can be to attach the crawl-space mem-
   brane for SMD systems to the perimeter foundation wall.

Gamma meter—A portable, hand-held instrument that can be
   used to measure the rate of energy  release by gamma
   radiation in microroentgens per hour.

Gamma radiation—Electromagnetic radiation released from
   the nucleus of some radionuclides during radioactive de-
   cay. Some gamma radiation, caused by radionuclides in
   the surrounding soil and rock and cosmic radiation from
   space, will exist in all houses. In infrequent cases, indoor
   gamma radiation can also result from building materials
   having elevated radionuclide concentrations.

Gauge—As used here, a device that provides a continuous,
   quantitative measurement of the suction within the piping
   of an ASD system.  Gauges may  or may  not  also be
   equipped with an alarm that provides a visual or auditory
   signal if the suction moves outside the acceptable range for
   the system.

Grade (above or below)—The term by which the level of the
   ground surrounding a house is known. In construction,
   grade typically refers to the surface of the ground. Struc-
   tures can be located at grade, below grade, or above grade
   relative to the surface of the ground.

Ground fault interrupter switch-^A switch that  can be in-
   stalled in the power cord leading to masonry drills that are
   being used to drill or core holes through concrete slabs.
   The  switch is intended to reduce the risk of electrical
   shock to the workers by shutting off power to  the drill if
   there were a power surge (e.g., if there were a short circuit
   in the drill) or if the bit hit  an electrical conduit beneath the
   slab.

Heat exchanger—A device  used  to transfer heat from one
   stream to another. In air-to-air heat exchangers for residen-
   tial use, heat from exhausted indoor air is transferred to
   incoming outdoor air, without mixing the two streams.

Heat recovery ventilators—Also known as air-to-ak heater
   exchangers or heat exchangers.

Hollow-block  wall, Block wall—A wall constructed using
   hollow rectangular masonry blocks. The blocks might be
   fabricated using a concrete base (concrete block) or using
   ash remaining after  combustion of  solid  fuels (cinder
   block). Walls constructed using  hollow blocks form an
   interconnected network with their interior hollow cavities.
   Foundation walls are most commonly constructed either of
   hollow block or of poured concrete, although other materi-
   als (such as fieldstone or wood) are sometimes also used.
                                                        XVII

-------
                                      Glossary (continued)
House air—Synonymous with indoor air. The air that occu-
   pies the space within a house.

Indoor air—That air that occupies the space within a house or
   other building.

Infiltration—The movement of outdoor air or soil gas into a
   house. The  infiltration that occurs when all doors and
   windows are closed is referred to in this document as the
   natural closed-house infiltration rate. The reverse of exfil-
   tration.

Installation costs—As used here, the cost to the homeowner
   of having an indoor radon reduction system installed in a
   house. If the system is installed by a professional mitiga-
   tor, installation costs will include labor (including fringe
   benefits), materials, overhead, and profit.

Ionizing radiation—Any type of radiation capable of produc-
   ing ionization in materials it contacts. Ionizing radiation
   includes high energy charged particles, such as alpha and
   beta particles, and non-particulate radiation, such as gamma
   rays and X-rays. By comparison, wave radiation, such as
   visible light and radio  waves,  does not  ionize adjacent
   atoms.

Joist—Any of the parallel horizontal beams (commonly 2- by
   10-in. boards) set from wall to wall to support the flooring
   for a living space or attic overhead. For example, joists in
   the basement ceiling will support the flooring for the first
   floor. If the basement has a plasterboard ceiling, the ceil-
   ing plasterboard will also be attached to these joists from
   underneath.

Load-bearing—A term referring to walls or other structures
   in a house that contribute to supporting the weight of the
   house.

Make-up air, Outdoor source of draft air (to address com-
   bustion appliance backdrafting)—As used here, an out-
   door supply of fresh air into the house to provide  the
   required draft air (and combustion air) needed for proper
   movement of products of combustion  up the flues of
   combustion appliances. Such make-up air may be needed
   in cases where an ASD system is found to be creating
   backdrafting of combustion appliances through depressur-
   ization of the house. The  term "make-up air" can also be
   used to describe the supply of outdoor air into the house in
   general, to prevent house  depressurization by combustion
   appliances and exhaust fans, in cases  where an ASD
   system has not been installed. "Make-up air" can also be
   used to refer to fresh air drawn into the cold air return of
   forced-air furnace systems to ventilate and perhaps even
   pressurize the house.

Manometer—A pressure-sensing device that displays pres-
   sure differences between  two locations by the level of a
   colored liquid. Two types of such manometers (a U-tube
   and a curved inclined manometer) are commonly used as
   pressure gauges permanently mounted on ASD installa-
   tions.

Magnehelic* gauge—A pressure gauge manufactured by the
   Dwyer Instrument Co. that displays pressures on  a cali-
   brated face. Such gauges are sometimes used as perma-
   nently mounted pressure gauges on ASD installations.

Membrane—As used here, sheeting (commonly polyethyl-
   ene) that is laid over the earthen or gravel floor of a crawl
   space as part of a sub-membrane depressurization system.

Micromanometer—A pressure-sensing device capable of de-
   tecting pressure  differences  as low as  0.001 in. WG.
   Commonly used in diagnostic testing; e.g., to assess sub-
   slab  depressurizations created  by a diagnostic vacuum
   cleaner or a SSD system.

Microroentgen—The roentgen (R) is a unit of measure of the
   total ionizing energy being produced by radiation in a unit
   mass of air. A microroentgen (jiR) is 1 millionth (10"s) of a
   roentgen. Gamma radiation  is commonly measured in
   units of |oR/hr; i.e., the rate  at which ionizing energy is
   released by the gamma rays per mass of air.

Mitigator—See Contractor.

Non-flowable caulk, Gun-grade  caulk—Refers to caulks that
   are sufficiently viscous such that the caulk bead will tend
   to retain its shape prior to  curing. Distinguished from
   flowable caulks. Non-flowable caulks are less effective at
   settling into cracks and irregularities in the opening being
   sealed but are required in cases where the opening does not
   provide a channel to contain the fluid movement of the
   flowable caulks  or  where the  opening  is on  a vertical
   surface.

Operating costs—The costs to the homeowner/occupant of
   continued operation of the radon reduction system. Oper-
   ating costs include electricity to operate the ASD fan, the
   house heating/cooling penalty resulting from the exhaust
   of treated house air by the ASD system, system mainte-
   nance (such as occasional fan repair/replacement), and the
   costs for periodic follow-up radon measurements to ensure
   that the system is continuing to be effective.

Passive soil depressurization—Soil depressurization tech-
   niques that are analogous to ASD systems but which rely
   on natural phenomena (thermal and wind effects) rather
   than an active fan to develop the suction in the system.
   Passive suctions  will be much  lower than fan-developed
   suctions, and the performance of passive soil depressuriza-
   tion  systems will always be lower, less reliable, and more
   variable than that for active systems.
                                                       XVIII

-------
                                      Glossary (continued)
PE (polyethylene)—A polymeric, plastic-like material simi-
   lar to PVC. Rigid polyethylene piping is sometimes used
   in ASD system piping. Thin sheets of polyethylene (usu-
   ally 6 to 10 mils thick) are commonly used as the mem-
   brane over the crawl-space floor for SMD systems.

Perimeter channel drain, Canal drain (sometimes referred to
   as a "French" drain)—A water drainage technique in-
   stalled in basements  of some houses during initial con-
   struction. If present, typically consists of a 1- or 2-in. gap
   between the basement block wall and the concrete floor
   slab around the entire perimeter inside the basement. This
   gap allows  water seeping through block foundation walls
   or flowing  from on top of the slab to drain into the fill
   beneath the slab. Often, this approach is utilizing the sub-
   slab  fill as a dry well. Sometimes,  an interior sub-slab
   drain tile loop (or, rarely, the channel drain itself) channels
   this water to a sump  in the basement. The term "French
   drain"  is sometimes also used to refer to a large gravel-
   filled dry well on the exterior of the house (rather than
   directly under the slab), which drains water away from the
   foundation; that is not the definition intended in this docu-
   ment

Picocurie (pCi)—A unit of measurement of radioactivity. A
   curie is the amount of any radionuclide that undergoes
   exactly 3.7 x 1010 radioactive disintegrations per second. A
   picocurie is one trillionth (10J2) of a curie, or 0.037
   disintegrations per second.

Picocurie per liter (pCi/L)—A common unit of measurement
   of the concentration of radioactivity in a gas. A picocurie
   per liter corresponds  to 0.037 radioactive disintegrations
   per second in every liter of air.

Plenum—A chamber into which air is forced, drawn, or
   collected, prior to distribution to  other locations.

Polyethylene—see PE.

Post-mitigation—Refers to any  steps taken following  the
   installation of a radon reduction system in a house.

Poured concrete wall—A foundation wall constructed by
   pouring concrete within  forms  that are removed after
   constructibn. The most common alternative to hollow-
   block walls.

Pre-mitigation—Refers to any steps taken prior to the instal-
   lation of a radon reduction system in a house.

P-trap—In plumbing  applications,  a horizontal section of
   piping containing a U-shaped dip at one end (resembling a
   horizontal letter "P") installed directly below drains. The
   intent is for water to stand in the U, creating a plug that
   prevents odors or vermin from the sewer from entering the
   house through the drain.
PVC (poly vinyl chloride)—A polymeric, plastic material that
   is resistant to deterioration (e.g., by soil chemicals) and is
   used in a wide variety of products. It is used to make rigid
   piping that is commonly used; e.g., in residential sewer
   lines,  and as the piping for ASD systems. Flexible PVC
   couplings can be used to join sections of rigid PVC piping.

Radon—The  only naturally  occurring radioactive element
   that is a gas. Technically, the term "radon" can refer to any
   of a number of radioactive isotopes having atomic number
   86. In this document, the term is used to refer specifically
   to the isotope radon-222, the primary isotope present
   inside houses. Radon-222 is directly created by the decay
   of radium-226 in the uranium decay chain, and has a half-
   life of 3.82 days. Another common isotope of radon (ra-
   don-220, also known as thoron) is a  decay product of
   radium-224, in the  thorium decay chain; thoron has a
   much shorter half-life (56 seconds) than does radon-222
   and, hence, is generally  not a serious problem inside
   houses. However, where high thorium concentrations exist
   in the soil very near the house or where high soil perme-
   ability permits rapid movement of the thoron into the
   house, thoron can sometimes be an important contributor
   to total radon concentrations.

Radon Contractor Proficiency Program (RCPP)—A volun-
   tary program established by the U.S. Environmental Pro-
   tection Agency through which a radon mitigator, by pass-
   ing an examination and by meeting certain other require-
   ments, can demonstrate proficiency in this field.

Radon decay products (or radon progeny)—The four radioac-
   tive elements that immediately follow radon-222 in the
   decay chain. These elements are polonium-218, lead-214,
   bismuth-214, and polonium-214. These elements have
   such short half-lives that they exist only in the presence of
   radon. The progeny are ultrafine solids that tend to adhere
   to other solids, including dust particles in the  air and solid
   surfaces in a room. They adhere to lung  tissue when
   inhaled; the two decay products that are alpha emitters
   (polonium-218 and polonium-214) can then bombard the
   tissue with alpha particles,  thus creating the health risk
   associated with radon. Also referred to as radon daughters.

Re-entrainment—Used in this document to refer to the flow
   of ASD exhaust gases back into the house after they have
   been discharged outdoors. Re-entrained exhaust can pre-
   vent indoor radon concentrations from being reduced to
   the extent that would otherwise be possible  by the ASD
   system. Discharge of the  ASD exhaust above the house
   eave  and away from openings in  the house shell is in-
   tended to reduce (if not totally eliminate) re-entrainment.
                                                       XIX

-------
                                      Glossary (continued)
Rotary hammer drill, Hammer drill—An electric power drill
   that includes a hammering motion in addition to the rota-
   tion of the drill bit, suitable for drilling through concrete.
   The hammering motion is created by metal-to-metal con-
   tact within the drill. These drills are smaller  and  less
   powerful than are electro-pneumatic roto-stop hammers.

Roto-stop hammer, Electro-pneumatic roto-stop hammer—
   A large electrically driven power  drill that provides a
   hammering motion in addition to the rotation of the drill
   bit, and which is larger and more powerful than a rotary
   hammer drill. The hammering motion is created pneumati-
   cally, using compressed air generated by a compressor
   within the device. The hammer usually requires two hands
   to operate. The term "roto-stop" refers to the fact that the
   device can  be adjusted to eliminate the rotary motion so
   that it can be used as a chisel or electric jackhammer.

Sealing—Measures to close openings through slabs, founda-
   tion walls, crawl-space membranes, or other parts of the
   house. Sealing can be intended to prevent soil gas from
   entering the structure through the particular opening or to
   prevent house  air from leaking out through the opening
   (short-circuiting into an operating ASD system). 'True"
   sealing would refer to a 100% airtight seal, preventing all
   convective air movement through the opening (and, usu-
   ally, preventing diffusive radon movement  as  well). In
   practice, many seals will not be 100% effective at prevent-
   ing convective and diffusive movement. To adequately
   reduce short-circuiting of house air into ASD  systems,
   "true" seals are not necessary.

Silt plate—A horizontal band (typically 2x6 in.) that rests on
   top of a  block or poured concrete  foundation wall and
   extends around the entire perimeter of the house. The ends
   of the floor joists that support the floor above the founda-
   tion wall  rest upon the sill plate.

Slab—A layer of concrete, typically about 4-in. thick, that
   commonly  serves as the floor of any part of a house
   whenever the floor is in direct contact with the underlying
   soil.

Slab below grade—A type of house construction where the
   bottom floor is a slab that averages between 1 and 3 ft
   below grade level on one or more sides.

Slab on grade—A type  of house construction where the
   bottom floor of a house is a slab that is no more than 1 ft
   below grade level on any side of the house.

Sniffing (to estimate radon concentrations)—A specific adap-
   tation of grab sampling techniques for radon measure-
   ment, to obtain a rapid estimate of the radon concentration
   at potential entry routes (e.g., under slabs and inside block
   wall cavities). Relative to standard grab sampling, uses a
   much shorter counting time,  and thus provides  a  less
   quantitative radon measurement. Most commonly used in
   pre-mitigation diagnostic testing.

Soil depressurization—Reducing the soil gas pressure (gen-
   erally relative to the pressures inside a house), usually with
   the objective of preventing the convective flow of soil gas
   up into the house.

Soil gas—Gas that is always present underground in the small
   spaces between particles of the soil or in crevices in rock.
   Major constituents of soil gas include nitrogen and oxygen
   (from the outdoor air), water vapor, and carbon dioxide.
   Since radium-226 is  essentially always present in the soil
   or rock, trace levels of radon-222 will exist in the soil gas.

Soil ventilation—Dilution of the soil gas with air drawn from
   elsewhere, usually from the outdoors or from inside the
   house. Such ventilation reduces the radon concentration in
   the soil gas, thus reducing the amount of radon that would
   enter the house when the soil gas enters. Note the signifi-
   cant distinction between "soil depressurization" and "soil
   ventilation." Sub-slab pressurization probably  functions
   largely by a  soil ventilation mechanism. SSD  systems,
   while often functioning largely by a soil depressurization
   mechanism, may also have a tine soil ventilation compo-
   nent.

Source term—Refers to the "strength" of the radon source in
   the soil and rock underlying a house. This strength is
   determined by the radium content of the soil and rock, the
   fraction of the radon that actually enters the soil gas when
   the radium decays, and the ease of soil gas movement
   through the soil toward the house. In this document, the
   term "source term" is often used to refer to the concentra-
   tion of radon in the soil gas.

Stack effect—The upward movement of house air when the
   weather is cold caused by the buoyant force on the warm
   house air. House air leaks out at the upper levels of the
   house, and outdoor air (and soil gas) leaks in at the lower
   levels to compensate. The continuous exfiltration upstairs
   and infiltration downstairs maintain the stack effect air
   movement, so named because it is similar in principle to
   hot combustion gases rising up a fireplace or furnace flue
   stack.

Sub-membrane depressurization (SMD)—A variation of the
   ASD technology commonly applied to crawl-space houses,
   in which suction is drawn beneath a membrane that has
   been placed over the earthen or gravel crawl-space floor.

Sub-membrane piping—Perforated piping, like drain  tile,
   that has been placed beneath the membrane of SMD
   systems to  aid in the  distribution of the  suction  field
   beneath the membrane.
                                                       xx

-------
                                      Glossary (continued)
Sub-slab depressurization (SSD)—A variation of the ASD
   technology, where suction is drawn beneath the concrete
   slab  in a basement, slab-on-grade, or slab-below-grade
   house.

Suction field extension—A measure of how well suction
   applied at one point (e.g., beneath the slab) extends to
   other parts of the sub-slab region.

Suction pipes—Pipes, usually PVC, ABS, or PE, which are
   installed (e.g., through slabs, walls, and membranes) in
   order to draw suction as part of an ASD system.

Sump—A pit  through a basement floor  slab, designed to
   collect water and thus avoid water problems in the base-
   ment. Water is often directed into the sump by drain tiles
   around the inside or outside of the footings.

Sump pump—A pump to move collected water out of the
   sump pit, to an above-grade discharge remote from the
   house. "Submersible" sump pumps are designed for opera-
   tion with the entire unit near or below the water level in the
   sump, and the motors are thus  designed to be corrosion-
   resistant. Submersible pumps are necessary any time the
   sump pit is to be covered as part of the radon mitigation
   system to resist rusting of the pump motor.

Tight house—-A house with a low air exchange rate. If 0.5 to
   0.9 ach is typical of modern housing, a tight house would
   be one with an exchange rate well below 0.5 ach.

Top voids—The air space in  the top course of blocks in
   hollow-block foundation walls;  that is, the course of block
   to which the sill plate is attached and on which the walls of
   the house rest.

Veneer, Brick veneer—A single layer of brick constructed on
   the  exterior face of an  outer wall of a house or other
   building (e.g., in lieu of wooden siding), to provide protec-
   tion, insulation, and ornamentation. The brick veneer is
   securely attached to the load-bearing frame  or masonry
   wall behind it, to prevent the brick from puling away from
   the house. However, the veneer is not designed to support
   a load itself, other than the weight of the bricks.

Ventilation rate (of a house)—The rate at which outdoor air
   enters the house, displacing house air. The ventilation rate
   depends on the tightness of the house shell, weather condi-
   tions, and the operation of appliances (such as fans) influ-
   encing air movement. Commonly expressed in terms of air
   changes per hour, or cubic feet per minute.

Visual survey (of a house)—A mandatory component of pre-
   mitigation diagnostic  testing. Involves inspection of the
   house to aid in the selection  and design of the  radon
   reduction measure.
Wall/floor joint—The junction between the slab (of a base-
   ment or slab-on-grade house) and the foundation walls. In
   many cases, this junction will be a small crack which is
   perhaps only a hairline crack, or it can be perhaps 1/16-in.
   wide. In other cases, where this joint is a perimeter channel
   drain, the gap will be 1 to 2 in. wide. In still other cases, the
   perimeter wall/floor joint may consist of an  expansion
   joint (a l/2-in.-wide gap filled with asphalt-impregnated
   fibrous material), usually to better accommodate any verti-
   cal shifting of. the slab relative to the foundation walls.
   Where the slab and the footings/foundation wall have been
   poured as a monolithic pour, there may be no crack, other
   than potential settling cracks.

Warm air supply—The ducting and registers that direct heated
   house air from the forced-air furnace, to the various parts
   of the house.  The supply ducting is at elevated pressure
   relative to the house because the central furnace fan is
   blowing air through this ducting.

Waterless trap—A trap,  similar in function to the P-trap
   commonly used in plumbing applications but not requiring
   water in order to block sewer (or soil) gas from flowing up
   into the house via floor drains, etc. The trap is designed
   such that a weighted ball or ring seats in a manner to
   prevent gas entry, even if the water in the trap dries out.
   Very useful in sealing floor drains or providing a  water
   path through sump covers in radon mitigation systems, in
   cases where water flow is likely to be so infrequent that a
   standard  water  trap might dry out. Marketed under the
   trade name DranjerR.

WG,  in.  WG—The term "WG" stands for "water gauge."
   Inches of water is  a unit of measure of pressure (or
   suction);  1  in. of water pressure would be that pressure
   able to sustain the weight of a column of water 1 in. high.
   Atmospheric  pressure (i.e., the  pressure created on the
   surface of the earth by the weight of air in the atmosphere)
   is 33.9 ft, or about 407 in., of water at standard conditions.
   One inch of water gauge (1 in. WG) is the reading that
   would be provided by a pressure measurement device (a
   "gauge") if the pressure actually being measured by the
   device were 1 in. of water greater than atmospheric (i.e., if
   the absolute pressure being measured were 408 in. rather
   than 407 in.  of water). Also  expressed as WC ("water
   column").

Working level (WL)—A unit of measure of the exposure rate
   to radon  and radon progeny defined as the  quantity of
   short-lived progeny that will result in 1.3 x 10s MeV of
   potential  alpha energy per liter of air. Exposures are mea-
   sured in working level months (WLM); e.g., an exposure
   to 1 WL for  1  working month  (170 hours) is 1 WLM.
   These units were developed originally to measure cumula-
   tive work place exposure of underground uranium miners
   to radon and continue to be used today as a measurement
   of human exposure to radon and radon decay products.
                                                        xxi

-------
                      Metric  Equivalents
Although it is EPA's policy to use metric units in its documents, non-metric units are used
in this report for the reader's convenience. Readers more accustomed to the metric system
may use the following factors to convert to that system.
  Non-metric
Times
Yields metric
mil (0.001 in.)                  0.0254
inch (in.)                       2.54
foot (ft)                        30.5
square foot (ft2)                  0.093
cubic foot (ft3)                   28.3
gallon (gal)                     3.78
cubic foot per minute             0.47
(cfm, or ftVmin)
inch of water gauge              249
(in. WG)
degree Fahrenheit (°F)         5/9 (°F-32)
British thermal unit (Btu)         1,060
picocurie per liter                 37
(pCi/L)
microroentgen                2.58 x 1040
(MR)
                    millimeter (mm)
                    centimeter (cm)
                    centimeter (cm)
                    square meter (m2)
                    liter (L)
                    liter (L)
                    liter per second (L/sec)

                    pascal (Pa)

                    degree Centigrade (°C)
                    joule (J)
                    Becquerel per cubic meter
                    (Bq/m3)
                    coulomb per kilogram
                    ofair(C/kg)
                                     XXII

-------
                          How to Use This  Document
This document has been designed as a comprehensive
reference document that will most commonly be used by
professional mitigators or by other persons interested in
the detailed design and installation of  indoor radon         «
reduction systems for existing houses. As a result, the
document is lengthy. Because of the tremendous amount
of material to  be covered, the discussion of the major
individual steps in selecting, diagnosing, designing, and
installing mitigation systems has necessarily been sepa-      IB.
rated into different sections. The size of this document
can complicate its effective use, especially by the first-
time user.
                                                        d
This section is a step-by-step summary of how to use
this document. To some extent, it is also an overview on
how to go about selecting, designing, installing, and
operating a radon reduction system for an existing house.
The steps in this process are summarized in Figure H-l.
                                                        (
Many of the steps involved in this process will often be
conducted by a professional mitigator. However, some
steps (such as long-term operation of the system) will
commonly be  the responsibility of the homeowner or
occupant.
Step 1. Determine that the  house  has  a      1C.
         radon problem and initiate action
         to accomplish mitigation.
   This  step will be  initiated  or conducted by the
   homeowner, relocation firm, etc., prior to involve-
   ment by the mitigator.

   1A.  Measure radon levels in the  house to
         determine whether there is a radon prob-
         lem.

       •  Can be done by homeowner using charcoal
          or alpha-track detectors; or owner can have
          measurement done by a professional firm
          (using any one of a variety of measurement
          methods).

          Use EPA-recommended radon measurement
          protocols. See EPA  402-R-92-004 (Refer-
  ence EPA92d)  and EPA 402-R-92-003
  (EPA93).

  A summary of these measurement protocols
  for homeowners is presented in A Citizen's
  Guide  to Radon,  EPA/402-K92-001
  (EPA92a).

Take temporary measures to reduce ra-
don levels, prior to permanent  mitiga-
tion, as warranted and as feasible.

  The homeowner/occupant can increase house
  ventilation and/or seal major soil gas entry
  routes, as temporary measures while await-
  ing the installation of the permanent mitiga-
  tion system.

  Guidance for carrying out these temporary
  measures is provided in Section 2.3 on page
  24 of the second edition of the Technical
  Guidance, EPA/625/5-87/019 (EPA88a).
  Guidance for identifying radon entry routes
  is provided in Section 2.2 (page 15) of the
  second edition.

Arrange for permanent mitigation of the
house.

  The homeowner, relocation firm, etc., will
  talk with one or more professional mitiga-
  tors prior to selecting one to install a perma-
  nent mitigation system in the house.

  Owners should not consider installing a sys-
  tem on a do-it-yourself basis unless they
  feel fully conversant with the principles of
  mitigation and with the information in this
  manual.

  See Section 2.5.1 (page 41) of the second
  edition of Technical Guidance.

  For assistance in locating  and selecting a
  mitigator, see the sources of information
  listed in Section 15 of this document.
                                                H-1

-------
                           Step 1. Determine if house has a radon problem
                                  1 A.  Measure radon levels (References EPA92d and EPA93).
                                  1B.  Take temporary measures to reduce radon, if warranted.
                                  1C.  Arrange for mitigation.
                          Step 2. Select the appropriate radon reduction technique
                                  2A.  Conduct a visual inspection (Section 3.2).
                                       -  Should a technique other than ASD be considered?
                                          (In the large majority of cases, ASD will give the greatest,
                                          most reliable, best demonstrated, and most cost-effective
                                          reductions.)
                                       -  Assuming that ASD is the technique of choice, which variation
                                          of the ASD technology is appropriate? (Sections 2.2 and 3.2)
                                  2B.  Conduct other diagnostic testing if needed in special cases to
                                       assess whether techniques other than ASD should be considered
                                       (Section 3.5). (Such diagnostics will often not be necessary.)
                           Step 3. Design the ASD radon reduction system
                                  3A.  Conduct a visual inspection (Section 3.2).
                                  SB.  Conduct other diagnostic testing if needed (Sections 3.3, 3.4, 3.5)
                                       (Such diagnostics will not always be necessary.)
                                  3C.  Develop the detailed design (Sections 4 through 10).
                                       -  Select number of suction pipes (Sections 4.1,5.1, etc.).
                                       -  Select location of suction pipes (Sections 4.2,5.2, etc.).
                                       -  Select suction pipe type and diameter (Section 4.3, 5.3, etc.).
                                       -  Select suction fan (Section 4.4, 5.4, etc.).
                                       -  Design piping network and exhaust system (Section 4.6, 5.6, etc.).
                                       -  Identify sealing effort required (Section 4.7, 5.7, etc.).
                                  3D.  Estimate the installation  cost (Section 13).
                                                             T
Step 4.
Install the ASD radon reduction system
- Installation of suction pipes (Section 4.5, 5.5, etc.).
- Installation of piping network and fan (Section 4.6, 5.6, etc.).
- Completion of sealing steps (Sections 4.7, 5.7, etc.).
- Installation of gauges/alarms and labelling (Section 4.8, 5.8, etc.).
1
Step 5.

Confirm that installed ASD system is operating properly
5A. Conduct routine post-mitigation diagnostics to confirm proper
performance (Sections 11.2 through 11.5).
SB. Conduct trouble-shooting diagnostics when system is not performing
adequately (Sections 11.6 through 11.11).
1
Step 6.

Ensure proper long-term operation and maintenance
6A. Provide basic operation/maintenance information to owner/occupant
(Section 12.1).
6B. Continue to operate and maintain the system (owner/occupant
responsibility) (Section 12.2).
6C. Make periodic follow-up indoor radon measurements (Section 12.2.3).
6D. Owner/occupant should understand system operating costs
(Section 13).
Rguro H-1.        A summary of the steps to be followed in using this document.

                                                              H-2

-------
Step 2. Select the appropriate radon  re-
         duction technique for that house.

   The selection of a mitigation system depends upon
   house characteristics and radon levels. In the major-
   ity of cases, professional mltigators will determine
   that the most efficient, reliable, and cost-effective
   radon reduction technique will be some suitable
   variation of the active soil depressurization (ASD)
   technology. But in some cases, other approaches can
   or should be considered.

   2A.  Conduct a visual inspection of the house,
         in accordance  with Section 3.2 of  this
         document.

       •   Are a combination of factors present  that
           might complicate the application of ASD to
           that house?

              Suspected  poor sub-slab communica-
              tion.

           -   Fieldstone foundation walls.

           -   Inaccessible crawl space.

              Complex substructure.

              High degree of finish.

           -   Well water or building materials strongly
              suspected of being the radon source.

       •   Are factors present suggesting that specific
           other radon reduction approaches (other than
           ASD) might be candidates?

              A tight basement, suggesting the possi-
              bility of applying basement pressuriza-
              tion; occupants whose life-style would
              be amenable to a basement pressuriza-
              tion approach.

              A tight and/or inaccessible crawl space,
              suggesting the potential  for  applying
              crawl-space depressurization.

              Relatively low pre-mitigation radon con-
              centrations, so that only perhaps  50%
              radon reduction is required, in which
              case a heat recovery ventilator might
              be considered.
    Low pre-mitigation radon concentra-
    tions, combined with major soil gas en-
    try routes, which might suggest that a
    stand-alone sealing approach could be
    considered.

    Well-drained, gravelly, native soil, sug-
    gesting that active  soil pressurization
    may be preferred over ASD.

    Suspected high well water radon con-
    centrations, suggesting that well  water
    treatment may be needed instead of, or
    in addition to, ASD.

    Suspected high-radium building materi-
    als contributing to indoor radon (rare in
    most places), in which case some type
    of  barrier coating  or source removal
    might be considered.

If there are no factors ruling out ASD, which
variation of the ASD technology is appro-
priate? (See Section 2.2 of this document,
in addition to Section 3.2).

    Sub-slab  depressurization (SSD), pre-
    ferred in almost all houses having slabs
    (i.e., basements and slabs on grade)
    where drain tiles are not present.

-   Drain-tile depressurization  in  cases
    where the tiles drain to  a sump in the
    basement (sump/DTD);  an alternative
    to SSD when a sump having drain tiles
    is present.

-   Drain-tile depressurization  in  cases
    where the tiles drain to an above-grade
    discharge or dry well (DTD/remote dis-
    charge); an alternative to SSD that can
    be  considered when such  tiles  are
    present

    Block-wall depressurization (BWD),
    usually used only as a  supplement to
    SSD, DTD, or SMD in cases where
    these other techniques do not adequately
    treat entry routes associated with the
    block walls.

    Sub-membrane   depressurization
    (SMD), the only ASD variation  appli-
    cable  in crawl-space houses having
    earth- or gravel-floored crawl spaces.
                                                  H-3

-------
2B,  Conduct any other pre-mitigation diag-
     nostic  testing required to enable final
     selection of the appropriate radon reduc-
     tion approach forthat house. (See appro-
     priate portions of Section 3 of this docu-
     ment.)
     In many cases, no diagnostic testing beyond
     the visual survey will be needed to make the
     final selection of the radon reduction tech-
     nique.

   •   Blower door testing (Section 3.5.5) to de-
       termine if the basement is sufficiently tight
       to warrant practical consideration of base-
       ment pressurization. Basement pressuriza-
       tion is usually considered only when appli-
       cation of ASD is complicated (e.g., by poor
       sub-slab communication and/or by field-
       stone foundation walls) and alternatives to
       ASD are thus being weighed.

   •   Blower door testing (Section 3.5.5) to de-
       termine if a crawl space is sufficiently tight
       to make crawl-space depressurizatton an
       effective alternative to SMD.

   •   Blower door testing (Section 3.5.5) to de-
       termine if the house is sufficiently tight to
       warrant practical consideration of a heat
       recovery ventilator to achieve the desired
       degree of radon reduction. (Usually con-
       ducted only where alternatives to ASD must
       be weighed; e.g., due to poor communica-
       tion, fieldstone walls, and/or homeowner
       preference for an HRV.)

   •   Well water analysis (Section 3.5-2) to de-
       termine whether well 'water treatment should
       be considered  as a supplement to, or re-
       placement for,  ASD in order to adequately
       reduce  airborne radon concentrations.
       (Would usually be conducted only in cases
       where wells in the area have been observed
       to contain significantly elevated radon lev-
       els.)

   •   Gamma radiation measurements or flux mea-
       surements  (Section 3.5.3) to determine
       whether building materials  are a significant
       source of radon and, hence, whether barri-
       ers or source removal may be needed. (Of-
       ten conducted only where local experience
           suggests that building materials may be a
           source.)
Step 3. Design the radon reduction tech-
         nique for that house.

    Since this document addresses  only ASD, the fol-
    lowing-discussion focuses on the design  of ASD
    systems. For the design of other radon reduction
    techniques, refer to the appropriate section of the
    second edition of Technical Guidance, EPA/625/5-
    87/019 (EPA88a).

    3A.  Conduct a visual inspection of the house,
         in accordance with Section  3.2  of this
         document.

         The visual survey discussed in Step 2A above,
         in connection with selection of the reduction
         technique, will also provide most of the infor-
         mation needed for effective design of the ASD
         system. Often,  this survey will be the  only
         "pre-mitigation diagnostic testing" needed to
         design an ASD system.

       •   Factors that would influence the  number
           and location  of SSD suction pipes (e.g.,
           observed sub-slab aggregate, house floor
           plan and finish, sub-slab utilities).

       •   Factors that would influence tfie design of a
           crawl-space SMD system (e.g., size of crawl
           space, nature of the crawl-space floor, ac-
           cessibility).

       •   Factors that would influence the routing of
           the ASD exhaust piping (e.g., house finish,
           accessibility of an existing  utility chase,
           presence or absence of  an attic, location of
           an adjoining garage).

       •   Factors that would influence the degree of
           slab or wall sealing that would be required
           (e.g., the presence of a perimeter channel
           drain between the slab  and the foundation
           wall).

       •   Driving  forces for radon entry that might
           influence ASD design  (i.e., major exhaust
           fans that could depressurize the house suffi-
           ciently to provide a major challenge to the
           system). See also Section 2.2 on page 15 of
           the second edition of Technical Guidance.
                                              H-4

-------
3B.   Conduct any other pre-mitigation diag-
      nostic testing needed to permit effective
      design of the ASD system. (See appropri-
      ate portions of Section 3 of this docu-
      ment.)

    •   If any pre-mitigation diagnostic test is re-
       quired in addition to the visual survey, this
       test will most often be a qualitative assess-
       ment of sub-slab communication (Section
       3.3.1 of this document).

           Tells  qualitatively whether sub-slab
           communication is good, marginal,  or
           poor.

       -   Is conducted primarily when visual evi-
           dence or other experience in the area
           provides no clue regarding the  general
           nature of the sub-slab communication;
           i.e., whether the system is likely to need
           one SSD suction pipe or several.

    •   Other diagnostic tests to aid in ASD design
       (beyond the qualitative communication test)
       will often not be cost-effective  for profes-
       sional mitigators. These other tests include

           Radon grab sampling or "sniffing" be-
           neath the slab, inside block walls, and at
        '•   suspected soil gas entry routes (Section
           3.4).

           Quantitative measurement  of sub-slab
           communication (Section 3.3.2).

           Quantitative measurement of the flows
           produced by the sub-slab region (Sec-
           tion 3.5.1).

3C.   Develop the detailed design of the ASD
      system,  using one of the following sec-
      tions of this document:

    •   Section 4 (for SSD systems)

    •   Section 5 (for sump/DTD systems)

    •   Section 6 (for DTD/remote discharge sys-
       tems)

       Section 7 (for BWD systems)

    •   Section 8 (for crawl-space SMD systems)

    •   Section 9 (for active soil pressurization
       systems)
       •   Section 10 (for passive soil depressuriza-
           tion systems)

The information base supporting the design guidance for
the ASD systems in Sections 4 through 8 is presented
in Section 2.3 of this document. Data supporting the
guidance for active pressurization systems  are in Sec-
tion 2.4, and for passive depressurization systems in
Section 2.5.

           Guidance for selecting the number of suc-
           tion pipes is presented in the first sub-sec-
           tion within Sections 4 through 10 (e.g., in
           Section 4.1 for SSD systems).

           Guidance for selecting the location of the
           suction pipes is presented in  the second
           sub-section (e.g., Section  4.2).

           Guidance for selecting the type and diam-
           eter of the suction pipes is presented in the
           third sub-section (e.g^, Section 4.3).

       •   Guidance for selecting the  suction fan is
           presented  in the fourth  sub-section (e.g.,
           Section 4.4).

       •   Guidance for the design of the piping net-
           work and exhaust system is presented in
           the sixth sub-section (e.g., Section 4.6).

       •   Guidance for the design of the sealing ef-
           fort required in conjunction with the ASD
           system is presented in the seventh sub-sec-
           tion (e.g., Section 4.7).

    3D.  Estimate the costs to install the designed
         ASD system.

       «   The cost for installation by a professional
           mitigator will depend on the house and miti-
           gation system characteristics, themitigator's
           practices,  and  the mitigator's  labor rates
           (including fringe benefits), materials  costs,
           overhead, and profit margin.

       •   Typical installation costs for each variation
           of the  ASD technology  are presented in
           Section 13 of this document (e.g., hi Sec-
           tion 13.1.1 for SSD systems, Section 13.2.1
           for sump/DTD systems, etc.).

           -  See summary of installation costs for
              the alternative ASD variations in Table
              6, Section 13.8.
                                               H-5

-------
Step 4. Install  the  radon reduction  sys-
         tem in that house.

   Since this document addresses only ASD, the fol-
   lowing discussion focuses on the  installation of
   ASD systems. For installation of other radon reduc-
   tion techniques, refer to the appropriate section of
   the second edition of Technical Guidance (EPA/
   625/5-87/019).

   4A.  Proceed with the installation of the sys-
         tem in accordance with the design, using
         the appropriate section of this document
         (Sections 4 through 10).

         As with the design guidance in Step 3C above,
         the data base supporting this installation guid-
         ance is presented in Sections 2.3,2.4, and 2.5
         of this document.

       •  Guidance for installing the suction pipes in
          the slab, wall, or membrane is presented in
          the  fifth  sub-section within Sections  4
          through 10 (e.g., in  Section 4.5  for SSD
          systems).

       •  Guidance for installing the piping network
          and the exhaust system is presented in the
          sixth sub-section (e.g., Section 4.6).

       •  Guidance for completing the sealing steps
          required in conjunction with the ASD sys-
          tem is presented in the seventh sub-section
          (e.g., Section 4.7).

       •  Guidance for  installing gauges and/or
          alarms on the systems, and for labelling the
          system  components, is  presented in the
          eighth sub-section (e.g., Section 4.8).
Step 5. Confirm that the installed system
         is operating properly.

   5A.  Conduct post-mitigation diagnostic tests
         that are required in all cases, even when
         the ASD system appears to be operating
         weU, as described in Section 11  of this
         document.
       •   Complete a visual  inspection  of the  in-
           stalled system as a routine quality assurance
           step, as described in Section 11.2, to con-
           firm that all  details have been completed
           properly.
    •   Measure suction (and possibly flow) in the
       system piping, a step that can be completed
       while a pressure gauge is being installed in
       the system piping. See Section 11.3.

    •   Measure the indoor radon concentrations
       achieved by the system since this is the real
       measure of system performance, as described
       in Section 11.4.

           Complete a short-term radon measure-
           ment within 30 days after system instal-
           lation or ensure that the homeowner/
           occupant completes such a measurement
           (Section 11.4.1).

       -   Recommend  that the owner/occupant
           also make a long-term radon measure-
           ment (Section 11.4.2).

    •   Test for combustion appliance  backdraft-
       ing, as described in Section 11.5.

SB.  Conduct post-mitigation diagnostic  tests
     on a trouble-shooting basis, to determine
     how a system that is not performing ad-
     equately can  be improved.

    •   Conduct suction field  extension measure-
       ments beneath the slab, inside the block
       wall, or beneath the SMD membrane (Sec-
       tion 11.6).

           Probably the most effective and  most
           commonly used post-mitigation diag-
           nostic when ASD systems do  not per-
           form well.

           Most common reason for  inadequate
           performance is that suction field is not
           extending adequately.

    •   Conduct radon grab sampling or "sniff-
       ing" measurements (Section 11.7).

           Can be used to assess the relative im-
           portance of alternative potential entry
           routes.

           As one specific example, high residual
           radon concentrations inside an individual
           block foundation wall in houses having
           a SSD,  DTD,  or SMD  system could
           suggest that a BWD component should
           be added to the system to treat that wall.
                                                 H-6

-------
           Conduct chemical smoke flow visualiza-
           tion tests, as described in Section 11.8, to
           help identify potential entry routes not be-
           ing adequately depressurized.

           Conduct well  water radon  analyses  or
           gamma surveys to assess building materi-
           als as a radon source, as discussed in Sec-
           tions 11.10 and 11.11, if these tests were
           not conducted prior to mitigation and if all
           other possible reasons for elevated residual
           post-mitigation radon levels have been elimi-
           nated.
Step 6. Ensure  proper long-term opera-
         tion and maintenance of the ASD
         system.
   6A.  Provide the homeowner!occupant  with
         information to help ensure proper opera-
         tion and maintenance of the system (Sec-
         tion 12.1).

         This step is the responsibility of the profes-
         sional mitigator.

       •   According to EPA's interim mitigation stan-
           dards (Reference EPA91b), the owner/oc-
           cupant shall be provided with the following
           information.

              A description of the system, at a central
              label.

              The  system  characteristics representa-
              tive of proper operation (e.g., the proper
              range of readings that can be expected
              on any pressure gauge mounted on the
              system piping).

              The name and telephone number of the
              mitigator, to contact if the system stops
              performing properly.

              A statement indicating any required
              maintenance by the owner/occupant.

           Other information that might be provided,
           depending on the mitigator's practice, in-
           cludes

              Copies of warranties and manufacturer's
              brochures.

           -   The operating principles of the system.
           Corrective action that can be taken if
           the system operation ever moves out-
           side the  acceptable range (including
           steps the occupant can take directly prior
           to calling a professional mitigator for
           repairs).

6B.  Continue to operate and maintain the
     system properly.

     This step is the responsibility of the home-
     owner or occupant.

    •   Ensure continued proper operation of the
       system fan (Section 12.2,1 of this docu-
       ment).

           When the house is occupied, do not turn
           the fan off or down to a lower power
           setting than that at which it was left by
           the mitigator.

           Best radon reduction performance is ob-
           tained when the fans are operated at full
           power.

           Take appropriate action if the suctions
           in the system piping, as indicated on a
           gauge,  fall below  (or rise above)  the
           acceptable range marked on the gauge
           by the mitigator; or, if there is no gauge,
           if a system alarm or some other indica-
           tor suggests that the system may not be
           operating properly. (See specific proce-
           dures in Section 12.2.1.)

       -   A professional mitigator may have to be
           called in to make repairs.

       -   There are some simple diagnostic steps
           the owner or occupant can take directly
           prior to calling a mitigator.

           At some interval (4 to 10 years), the fan
           will have  to be repaired or replaced.

    «   Periodically inspect the system piping in an
       effort to ensure that blockage by ice or
       moisture does not occur (see Section 12.2.2).

    »   Periodically inspect seals in the system pip-
       ing, and seals that were installed in the
       house foundation as part of the mitigation
       system (Section 12.2.2).
                                                 H-7

-------
           Repair seals as necessary.

           With crawl-space SMD systems, "seal
           repair" may involve occasional major
           repairs or replacement of membrane.

6C.  Make periodic follow-up indoor radon
     measurements to confirm that the miti-
     gation system is continuing to perform
     adequately (Sections 12.2.3  and 13.1.2).

   •   EPA recommends re-testing at least once
       every 2 years.

       -   Can be done directly by homeowner/
           occupant using a charcoal or alpha-track
           detector; or, a professional measurement
           firm could be hired.

           Measurement should be made in fre-
           quently occupied part of house.

       -   Long-term measurement (e.g., with al-
           pha-track detector) is probably advis-
           able unless some recent apparent change
           in system performance warrants a more
           rapid measurement (e.g., using a char-
           coal detector).
    •   Measurements more frequent than once ev-
       ery 2 years are advisable if there is any
       apparent change in system performance (e.g.,
       a change in system suctions) or if the house
       has been modified.

6D.  EPA recommends that the  owner/occu-
     pant be advised of the annual operating
     costs associated with operation and main-
     tenance oftheASD system (Section 13).

       The operating costs consist of four elements:

           The electricity to operate the fan.

           The house heating/cooling penalty, re-
           sulting from treated house air being ex-
           hausted by the system.

           The cost of system maintenance (repair/
           replacement of fans on  all  ASD sys-
           tems, repair of seals, and occasional
           major repair/replacement of membrane
           for crawl-space SMD systems).

           The cost of periodic follow-up indoor
           radon measurements.

       The range of values for these  operating
       costs are summarized in Table 6 in Section
       13.8, for the alternative ASD variations.
                                             H-8

-------
                                              Section  1
                                            introduction
1.1    Purpose

This technical guidance document is designed to aid in the
selection, design, and operation of indoor radon reduction
techniques using soil depressurization in existmg houses. In
particular, it draws from the experience of numerous research-
ers and radon mitigators over the past eight years and distills
this information into practical guidance. The guidance should
enable the design, installation, and evaluation of soil depres-
surization systems with increased confidence, with reduced
system costs, and/or improved system performance,  under a
variety of conditions.

As the term is used here, "guidance" represents a recommen-
dation of what EPA considers to be a reasonable course of
action for a given set of conditions, based on experience to
date. Often, alternative reasonable courses of action are pos-
sible. Where this situation occurs, this document attempts to
provide readers with sufficient background to permit an in-
formed  decision for their specific sets of circumstances. A
mitigator will sometimes face special circumstances where
the guidance  in this document might riot represent the pre-
ferred approach in that  specific  case; a  careful, informed
judgement of the necessary approach must be made in such
special cases. Following the EPA guidance will not, in all
cases, guarantee a fully successful mitigation system. How-
ever, by effectively drawing upon the prior experience, use of
the guidance should reduce the risk of installing an unsuccess-
ful system and should facilitate making the subsequent modi-
fications needed to achieve success.

The term "guidance" must be distinguished from the term
"standard." Guidance is a recommendation of what generally
appears to be good practice. In many cases here, the guidance
is a recommendation of multiple alternative courses of action
that appear reasonable, from among which the user can choose.
Users of this document have the option  of following these
recommendations or of  using a different approach where
warranted by their particular circumstances.

By comparison, standards are prescribed procedures or re-
quirements which must be met. As discussed later, EPA has
issued interim standards and is in the process of developing
final standards that radon mitigation contractors must follow
if they wish to be listed under the Agency's Radon Contractor
Proficiency Program (RCPP). Many of these standards are
mentioned at appropriate locations in this document Where
individual standards are mentioned, the text clearly indicates
which features are standards (and are thus considered manda-
tory for RCPP-listed mitigators) and which features are guid-
ance (and are thus subject to the judgement of the user).

This document updates and replaces Section 5 (Soil Ventila-
tion) of the second edition of EPA's Technical Guidance
(EPA88a). Other EPA publications providing information on
indoor radon reduction in houses include

 -  Consumer's Guide to Radon Reduction (EPA92c), a book-
    let which provides a concise overview for homeowners of
    alternative radon reduction techniques.

 -  Application of Radon Reduction Methods (EPA89a), which
    is intended to direct a user through the steps of diagnos-
    ing a radon problem, selecting a reduction method, and
    designing/installing/operating a radon reduction system
    in existing houses, with less detail than is presented in the
    technical guidance documents.

 -  Radon Contractor Proficiency Program Interim Radon
    Mitigation Standards (EPA91b), which list the criteria
    that commercial mitigators are expected to meet if they
    are listed  as  proficient under  the RCPP. Final radon
    mitigation standards, which will replace the interim stan-
    dards, are currently in preparation.

 -  Radon Reduction in New Construction: An Interim Guide
    (EPA87a), a brochure summarizing features that can be
    incorporated into  a house during construction to reduce
    indoor radon levels.

 -  Radon-Resistant Construction Techniques for New Resi-
    dential Construction:  Technical Guidance  (EPA91a),
    EPA's technical guidance document specifically for ra-
    don reduction in houses during construction.

Further information on the indoor radon problem in general
can be obtained from the following EPA publications, among
others:

 -  A Citizen's Guide to Radon (Second Edition) (EPA92a), a
    booklet providing background concerning radon and sum-
    marizing health risks, measurement methods, and reduc-
    tion approaches.

 -  Technical Support Document for the Second Edition of
    the Citizen's  Guide to Radon  (EPA92b); this support
    document provides the information supporting the guid-
    ance in the Citizen's Guide.

-------
 -  Radon Reference Manual (EPA87b), which provides ad-
    ditional background information on indoor radon (e.g.,
    nature and origin, health effects, geographic distribution,
    and measurement).

1.2    Scope

This technical guidance document addresses the design, in-
stallation, and operation of soil depressurization systems for
radon reduction in existing houses. Alternative radon reduc-
tion methods, including structure ventilation, sealing of entry
routes, structure pressure adjustments, air cleaning, and well
water treatment, are not addressed here. Until the third edition
of the Technical Guidance is expanded in the future to update
these other reduction methods, readers interested in these
other methods are referred to the second edition.

The emphasis in this document is on active soil depressuriza-
tion, i.e., on systems which use a fan to depressurize the soil.
However, active soil pressurization, where the system fan is
reversed to  blow  outdoor air into the soil,  is  covered  in
Section 9. In addition, passive soil depressurization  is also
addressed in Section 10; with passive systems, natural phe-
nomena, including the indoor vs. outdoor temperatures and
the flow of winds over the roof-line, are relied upon to provide
the depressurization instead of a fan.

Tliis document focuses on  the retrofit of radon reduction
methods into existing houses as distinguished from the incor-
poration or radon-resistant features into new  houses  during
construction. Separate guidance has been  issued on the sub-
ject of new construction (EPA87a, EPA91a). The soil depres-
surization methods described in this document will be gener-
ally applicable to new construction. However, incorporation
of reduction methods during the construction phase permits
certain house construction features to be modified to improve
system performance. Thus, the approach to system design and
installation and the utilization of other reduction approaches
(such as sealing) can be quite different for new construction
relative to retrofit into existing houses.

As reflected by the title, this document focuses on detached
houses, as distinguished from multi-family dwellings  (apart-
ments,  condominiums)  and from large buildings, such  as
schools and commercial buildings. Separate technical guid-
ance has been issued for school buildings (EPA89b). The
active soil depressurization methods described in this docu-
ment will be  generally applicable to  these other types  of
buildings. However, because of the size and of some  impor-
tant differences in construction ntethods used  in these larger
buildings, the approach used in system design and installation
of the soil depressurization  system can be different.  More-
over, oilier reduction approaches (such as building pressure
adjustment through modifications to the ventilation system)
can more often play a role in these large buildings.

As discussed  in Section  2.1, this document addresses all
variations of the soil depressurization approach, including
sub-slab depressurization,  drain-tile depressurization, block-
wall depressurization, and  sub-membrane  depressurization in
crawl spaces. One or more of these variations  can be applied
to each  of the house substructure types (basements, slabs on
grade, crawl spaces, and combinations of these). These tech-
niques can be applied to address essentially any pre-mitiga-
tion radon concentration, from  extremely elevated to only
slightly elevated.

Soil  depressurization  techniques can treat only that radon
entering the house as a component of soil gas. Thus,  this
document addresses only those cases where the radon prob-
lem is due  to  naturally occurring  uranium/radium in the
underlying soil  and rock or to contaminated fill underneath
the house (e.g., uranium mill tailings). Soil depressurization
techniques cannot treat cases where the airborne radon in the
house enters with the well water or emanates from the build-
ing materials (due to, for example, contaminated aggregate
used in the concrete).

1.3    Content

This document consists of five major elements.

 1. General information on soil depressurization technology,
    including the principles involved with each variation of
    soil depressurization (Section 2.1) and the conditions
    under which each variation is particularly applicable or
    inapplicable (Section 2.2). This information also includes
    a detailed summary of the available data demonstrating
    the performance of each  variation as a function of the
    range of house design, house operation, system  design,
    system operation, and geology/climate variables (Sec-
    tions 2.3,2.4, and 2.5).

 2. A description of the diagnostic test procedures that can be
    used prior to mitigation to enable cost-effective design
    and installation of systems. See Section 3. Procedures are
    described for the full range of possible pre-mitigation
    diagnostic tests. Emphasis is placed on the three  pre-
    mitigation diagnostic tests most commonly found to be
    cost-effective in aiding the design of soil depressurization
    systems: a  visual survey of the house; sub-slab  suction
    field extension measurements; and radon grab sampling
    and sniffing.

 3. Detailed guidance on the design and installation of each
    variation of the soil  depressurization  technique. This
    detailed guidance is included in Sections  4 through 8,
    with one section devoted to each of the soil depressuriza-
    tion variations. Guidance on design and installation of
   , active soil pressurization  and of passive soil depressur-
    ization—to the extent possible,  given the much more
    limited experience with these two  approaches—is  pre-
    sented in Sections 9 and 10, respectively.

 4. A description of the diagnostic test procedures that should
    be used following installation of a system to ensure that
    the system  is operating properly and/or to diagnose the
    cause of the problem if it is not performing adequately.
    See Section 11.

 5. A  description of system operation and maintenance re-
    quirements (Section 12).

-------
In addition, estimated installation and operating cost ranges
for these systems are presented in Section 13. These cost
ranges should be inclusive of most, but not all, of the soil
depressurization installations throughout the continental U.S.,
in 1991 dollars.

1.4    Reason for Focus on Active
Soil Depressurization
This revision to the second edition of EPA's Technical Guid-
ance focuses on active soil depressurization for two reasons.

 1. Active soil depressurization is the most consistently ef-
    fective radon reduction method in existing  houses, and is
    the technique most widely used by commercial mitiga-
    tors.

    Active soil depressurization systems have consistently
    been found to provide high indoor radon reductions with
    good reliability. Most commercial mitigators appear to
    include active soil depressurization as the central compo-
    nent in most of their installations, especially when reduc-
    tions greater than perhaps 50% are required. By compari-
    son, the other mitigation approaches—house or crawl-
    space ventilation (with or without heat recovery), entry
    route sealing, house pressurization (or crawl-space pres-
    surization or depressurization), and indoor air cleaners—
    have each been found to be distinctly less effective and/or
    less reliable than active soil depressurization. In addition,
    some of these techniques offer other drawbacks—e.g., a
    potential impact on occupant comfort and life-style in the
    case of house pressurization and uncertainty regarding
    the actual effect on health risk in the case of some of the
    particle-removal based air  cleaners.  In part for those
    reasons,  these  other techniques  are less well demon-
    strated than is active soil depressurization.  Many of these
    other techniques offer little, if any, reduction in installa-
    tion or operating cost relative to active soil depressuriza-
    tion. These other techniques are most commonly used

    a) by homeowners on a do-it-yourself basis. Some of
       these techniques can sometimes be implemented eas-
       ily without special skills, and some of  them are rela-
       tively inexpensive when a mitigator's labor costs do
       not have to be incurred.

    b) by mitigators, either in conjunction with active soil
       depressurization or in circum stances where active soil
       depressurization is not applicable.

 2. The available data for active soil depressurization has
    increased the most significantly since the second edition
    was published in 1988; hence, it is for this technique that
    the most significant improvements in technical guidance
    can be provided.

This document includes sections on active soil pressurization
and on passive soil depressurization, as well as on active soil
depressurization. The pressurization and passive techniques
are included here because they each have certain similarities
to active soil  depressurization, and because this technical
guidance document thus provides the most appropriate con-
text in which to present these techniques. However, it must be
recognized that

 -  Pressurization and passive techniques  are far less well
    demonstrated than  is active soil depressurization; hence,
    less rigorous technical guidance can be provided for these
    other techniques, and the guidance that is presented here
    is less well supported.

 -  Passive soil depressurization techniques will always be
    less effective than active soil depressurization. The effec-
    tiveness of passive soil depressurization techniques in
    existing houses is unpredictable, highly variable,  and
    often moderate, at  best. Passive systems will likely find
    their greatest application in new construction, where fea-
    tures can be incorporated into the house during construc-
    tion to help improve passive performance.

 -  Although active soil pressurization techniques will occa-
    sionally provide greater radon reductions than does active
    soil depressurization  in a given house (usually where the
    underlying native soil is highly permeable), these occa-
    sions appear to be fairly infrequent

-------

-------
                                            Section 2
           Principles,  Applicability, and  Past Performance
                       of  Soil  Depressurization Systems
2.1    Principles of Active Soil
Depressurization

The general principle of soil depressurization is to draw
radon-containing soil gas away from the house foundation
before it can enter and to exhaust this soil gas  outdoors.
Where a fan is used to create the necessary  suction, the
approach is referred to as active soil depressurization (ASD).
The vast majority of soil depressurization systems that have
been installed are ASD systems.

There are several common variations of the ASD process. For
the purposes of this document, these variations are defined as
follows:

   - Sub-slab depressurization (SSD). One or more suction
     pipes are inserted into the aggregate or soil  beneath a
     concrete slab (either vertically down through the slab
     from the space above, or horizontally through a foun-
     dation wall  below slab level, from outdoors or from
     inside an adjoining basement). Suction is then drawn
     on these pipes using the fan, with the collected soil gas
     then vented  outdoors.

   - Drain-tile depressurization (DTD). Some houses have
     a  loop of perforated drain tile immediately beside the
     footing, either inside or outside the footing, for water
     drainage purposes. The nature of the DTD system
     depends  on how the drain tiles have been designed to
     direct the water away from the house:

     —  where the tiles drain to a sump inside a basement,
         the sump is capped, and the fan draws suction on
         the sump/drain tile network (referred to as "sump/
         DTD");

     —  where the tiles drain to an above-grade discharge at
         a low spot on the lot or to a dry well, the tile loop is
         tapped into at an appropriate point outdoors, and the
         fan connected to depressurize the loop (referred to
         as "DTD/remote discharge").

   - Block-wall depressurization (BWD). One or more indi-
     vidual suction pipes are inserted into the void network
     within a block foundation wall and connected to the
     fan. Alternatively, in what is referred to as the "base-
     board duct" approach,  a series of holes is drilled into
      the void network around the perimeter of a basement,
      just above slab level;  these holes are  then enclosed
      within a plenum ("baseboard") sealed to the wall and
      slab around the perimeter, and the plenum is connected
      to a fan.

   -  Sub-membrane depressurization (SMD), for crawl-space
      houses  having dirt floors. Plastic sheeting is placed
      over some or all of the dirt floor, creating a plastic
      "slab." One or more individual suction pipes penetrate
      this sheeting to draw suction under the  plastic (analo-
      gous to SSD), or suction is drawn on a loop of perfo-
      rated drain tiles placed under the plastic (analogous to
      DTD).

The primary mechanism by which ASD often  functions is to
create a negative pressure in the soil or aggregate immediately
under/beside  the  foundation, i.e., under the  concrete floor
slab, or inside the hollow-block foundation wall, or under-
neath a membrane laid over a dirt crawl-space floor. If the gas
pressure in the soil under/beside the foundation surface is
negative relative to that inside the house, then flows through
any openings through the foundation (e.g., through slab cracks
and block pores) will  consist of clean house air flowing
outward through these openings, rather than soil gas flowing
inward. For the system to be effective, this soil depressuriza-
tion must be maintained at least near the major openings/entry
routes. (Good depressurization can be less crucial; e.g., in
central, uncracked regions of slabs where there are no entry
routes for convective soil gas flow into the house.) Especially
in cases where a SSD or DTD system is found to be maintain-
ing an excellent suction field in the aggregate underneath a
slab, it is clear that flows through the slab openings have thus
been reversed everywhere.

A second mechanism by which ASD appears to function—to
a greater or lesser extent in different circumstances—is dilu-
tion of the radon-containing soil gas beneath the slab and
around the foundation. ASD systems can draw house air and
outdoor air down into the soil around the foundation, diluting
the radon in the soil gas. Thus, even when a negative soil gas
pressure is not being maintained immediately under/beside
the foundation in some places, any soil gas entering the house
at those locations may contain less radon, and the ASD system
can sometimes still be effective if the dilution is sufficient.

-------
Most ASD systems probably function through a. combination
Of the two mechanisms above. Where the slab is fairly tight
(so that little house air is drawn into the soil) and the native
soil is fairly tight (so that little outdoor air is drawn into the
soil), the soil depressurization mechanism is undoubtedly the
predominant component. But where the slab is leaky and the
native soil is highly permeable, so that more air is drawn into
the soil, the soil gas dilution component probably becomes
increasingly important. In cases of extremely permeable na-
tive soils, where flows of outdoor air into the system become
so high that it is difficult to maintain an adequate suction field,
the dilution mechanism may become the predominant compo-
nent. As discussed later, in such cases, it is sometimes prefer-
able to  use soil pressurization techniques  (which function
largely by dilution) rather than to attempt soil depressuriza-
tion.

Perhaps a third mechanism, referred to here as "air-barrier
shielding," might also sometimes play a role. Operation of
ASD systems is postulated to create flows of outdoor air down
through the soil into the system. This subtle flow of ambient
air under essentially unmeasurable pressure gradients may be
creating a shield  around the foundation,  diverting soil gas
which would otherwise flow  toward the foundation. This
mechanism could explain why SSD systems in houses having
marginal sub-slab communication sometimes  obtain good
radon reductions  despite the failure  of the SSD system to
establish measurable depressurizations everywhere beneath
the slab. It could also explain why SSD systems in basement
houses with block foundation walls can achieve high radon
reductions despite failure of the SSD system to create measur-
able depressurizations  within  the  wall cavities; the walls
become a less important source because the soil gas is inter-
cepted before it can enter the void network. Of course, this
"air-barrier shielding" could also be interpreted as nothing
more than a variation of the two mechanisms listed previ-
ously. That  is,  the soil is in fact being depressurized at the
remote location where the conceptual air barrier diverts the
soil gas flow, in  accordance with  the first mechanism; the
depressurization is just too small to measure. Or, the soil gas
is just being diluted by the ambient air flow, in accordance
with the second mechanism.

In practice, ASD systems are commonly designed assuming
that the first mechanism above is the sole mechanism that can
be relied upon. That is, effective depressurization of the soil
and fill material  is assumed to be necessary immediately
under/beside the exterior surface of the foundation, at least in
the vicinity of openings through the foundation. In particular,
ASD variations involving depressurization beneath a concrete
floor slab (SSD, and DTD iix cases where the drain tile loop is
inside the footings) must achieve effective depressurization
immediately under the slab. (There are specific cases,  with
some ASD  variations, where comprehensive depressuriza-
tions are difficult and often  unnecessary to maintain  under/
beside some foundation surfaces;  these cases will be  dis-
cussed later.) To maintain effective soil depressurization  near
all  entry routes through the foundation, the ASD system
requires a suitable combination of the following factors.
   -  Adequate communication within the aggregate and soil
      immediately under the slab (or under the membrane
      covering the dirt floor in the crawl-space). With good
      communication, the suction field generated by a suc-
      tion point can extend beneath the slab to entry routes
      remote from the suction point. Good communication
      thus reduces the need to locate suction points close to
      all of the entry routes. Where SSD  or DTD are being
      relied upon to treat entry routes associated with the
      foundation wall, or where  BWD systems are being
      relied upon to treat slab-related entry routes, communi-
      cation between the sub-slab  and  the void network
      within the block wall, or between the sub-slab and the
      soil on the outside of the footing/foundation wall, can
      sometimes also be of concern  (in addition to the com-
      munication within the sub-slab aggregate/soil).

   -  Fans sufficiently powerful to develop adequate suc-
      tions in the system piping at the flows that are encoun-
      tered. Even where communication is satisfactory, fans
      developing adequate suction  increase the likelihood
      that sufficient depressurization will extend through the
      aggregate, soil, or wall voids, remote from the suction
      points. Where communication is very poor, it becomes
      all the more important that the fans achieve at least
      some minimum suction (although  very  high-suction
      fans will not necessarily provide significant additional
      extension of measurable soil depressurization).

   -  A system design intended to minimize pressure loss in
      the system, so that fan suction is effectively used in
      establishing a suction field in the soil rather than in
      simply moving gas through the system piping. Among
      the steps that can by implemented to  reduce pressure
      loss in the system are a) use of system piping with a
      sufficiently large cross-sectional area; b)  reducing the
      length of the piping runs and the number of elbows and
      other flow obstructions; and c) excavating a hole under
      the slab under SSD pipes, to reduce pressure loss as the
      soil gas accelerates to piping velocity.

   -  Closure of major openings  in the slab or foundation
      wall beneath or beside which the soil is being depres-
      surized (e.g., closure of important slab openings when
      sub-slab depressurization is used). If such  openings are
      not adequately closed, indoor air will flow out through
      these openings and will enter  the suction system. The
      ability of the soil suction field to extend effectively to
      remote entry routes could be seriously  hindered by
      house air flowing into the soil/aggregate through un-
      closed openings.

These factors will be repeatedly addressed in the subsequent
discussions of the individual ASD variations.

The general principles indicated above are discussed further
below, as they pertain to each of the specific variations of the
ASD technique.
   -  Suction points located sufficiently close to the entry
      routes.

-------
2.7.1  Active Sub-Slab Depressurization
(SSD)

In active SSD, a fan is used to draw soil gas away from the
foundation by means of individual suction pipes which are
inserted into the region under a concrete slalx The pipes are
commonly inserted vertically  downward through the slab
from inside the house, as illustrated in Figure 1. When more
convenient, they can also be inserted horizontally through a
foundation wall at a level beneath the slab, as in Figure 2.
Horizontal penetration through a foundation  wall  is most
likely to be convenient when the slab is near grade level, and
the penetration can be made from outdoors (as shown in
Figure 2)  or through the stub wall from inside an adjoining
basement.

The intent of the SSD system is to create a low-pressure
region underneath the entire slab. If a depressurization can be
maintained under the slab which is sufficiently large so as not
to be overwhelmed by depressurizations created inside the
house by weather effects and homeowner activities, this would
prevent soil gas from entering the house through cracks and
other openings in the slab. It could also reduce or prevent soil
gas entry into the void network inside hollow-block founda-
tion walls in the region around  the  footings, thus at least
partially preventing radon entry via the walls. Sometimes, the
depressurization can extend under the footings or through the
block walls to inhibit soil gas entry into the house through
below-grade openings in the exterior face of the foundation
wall.       .

Were the system to function solely by the primary mechanism
discussed earlier, i.e., by maintaining a measurable depressur-
ization in the soil everywhere that it contacts the foundation, a
soil depressurization of about 0.015 in. WG, measured during
mild weather, would  nominally be required to  ensure that
subsequent cold weather and winds would rarely depressurize
the house sufficiently to overwhelm  the system. If exhaust
appliances were off during the measurement, the soil would
nominally have to be depressurized by an additional 0.01 to
0.02 in. WG to ensure that the system would not be over-
whelmed  when these  appliances  were turned on. However,
some experience suggests that the other mechanisms men-
tioned earlier, including soil gas  dilution and perhaps air-
barrier shielding, can come into  play to varying degrees,
depending upon the circumstances. These other mechanisms
could explain why good radon reductions are often achieved
by SSD systems even in cases where portions of the sub-slab
are only marginally depressurized, to an extent far less than
the nominally required 0.025 to 0.035 in. WG. They could
also explain why SSD systems appear to prevent radon entry
through wall-related entry routes in many (but definitely not
all) cases, even when no depressurization (or only minimal
depressurization) can be measured inside the wall voids.

The central issues  with  SSD systems are the number of
suction pipes needed, where  they must be  placed, and the
suction that the fan must maintain in the pipes, in  order to
establish an adequate sub-slab depressurization near all (or at
least the major) soil gas entry routes. The resolution of these
issues is determined largely by the communication beneath
the slab, i.e.,  by the ease with which suction at one point can
extend to other parts of the slab and to the surrounding soil.
Where a good and uniform layer of aggregate (gravel or
crushed rock) was placed under the slab during construction,
communication immediately underneath the slab will gener-
ally be very good. In such cases, one (or perhaps two) suction
pipes will generally be sufficient, and can be located with
some flexibility, even if the communication within the native
soil, underlying the aggregate, is poor. Where there is good
aggregate, the system can be pictured as using the aggregate
layer as a large collector or plenum, into which the soil gas in
the vicinity of the house is drawn and then exhausted out-
doors. Where there is not a good layer of aggregate, or where
the layer is uneven or interrupted to a significant degree, more
suction pipes will commonly  be needed,  and their location
near to the major entry routes (such as the wall/floor joint)
will be increasingly important, depending upon the communi-
cation within the underlying soil.

Although poor communication within the underlying soil will
not impact the ability of SSD to depressurize slab-related
entry routes when sub-slab aggregate is  present,  it would
hinder the suction field from extending through the soil under
the footings. Thus, it could impact the ability of SSD to
depressurize entry routes associated with the outer face of the
foundation walls. It would also reduce the flow of outdoor air
down through the soil, thus reducing the possible roles of soil-
gas dilution and air-barrier shielding as mechanisms in deter-
mining SSD performance.

Where drain tiles are located inside the footings under the
slab, the drain-tile depressurization approaches described in
Sections 2.1.2 and 2.1.3 are essentially a variation of SSD.
However, in this document, the term "sub-slab depressuriza-
tion" is used only to refer to cases where individual pipes are
inserted into the sub-slab region in the manner depicted in
Figures 1 and 2, or in closely-related adaptations of that
approach.

SSD has been one of the most widely applied and effective
approaches used by the radon  mitigation industry, especially
in treating high-radon houses.  SSD should be one of the first
techniques considered in any house, especially where there is
no sump and where sump/ DTD is thus not an option.

2.1.2 Active Sump/Drain-Tile
Depressurization (Sump/DTD)

Drain tiles surround part or all of some houses in the vicinity
of the  footings to collect water and drain  it away from the
foundation. The drain tiles may be perforated rigid plastic
(usually ABS or PVC), perforated flexible plastic (high-den-
sity polyethylene or polypropylene), or porous clay. Drain
tiles will generally be located  right beside, or just above, the
perimeter footings.  They can be either on the side of the
footings away from the house (in which case they are referred
to here as "exterior" drain tiles), or on the side toward the
house  (referred to as "interior" drain tiles). In houses with
slabs, interior drain tiles would be under the slab, embedded
in any sub-slab aggregate. Sometimes (although not com-
monly) interior drain tiles are  not located beside the footings
but extend underneath the slab in some  different pattern.
Exterior drain tiles are  usually buried in a bed of aggregate

-------
                  Exhaust (Released
                  Above Eave)

  Exhaust Option 2:    »-f •>
  Exhaust Stack on       •
  House Exterior1
Exhaust Option 1:
Exhaust Stack through
                                                                       Floor
                                                                Strapping (or
                                                                Other Support)
                                            House Interior
                                               To Exhaust Fan
                                               Mounted in Attic
  Exhaust Fan
  (Rated for
  Exterior Use
  or Enclosed)
              V.	A
    Flexible
    Coupling
                                                            Sealant Around
                                                            Suction Pipe3
                                                               Open Hole
                                                               (6" to 18"
                                                               Radius)
                        Slope Horizontal Legs
                        Down toward Sub-Slab
                        Hole, to Permit Condensate
                        Drainage
                                                                                    Connection to Other
                                                                                    Suction Point(s),
                                                                                    If Any
                                                             Notes:
                                                             1.  Detail for interior and exterior
                                                                stacks shown in later figures.

                                                                Options for effectively
                                                                supporting horizontal and
                                                                vertical piping runs shown in
                                                                later figures.
                                                                Detail shown for piping
                                                                penetrations through slab is one
                                                                option among several. Other
                                                                options shown in later figures.
                                                             4.  Closing of various slab openings
                                                                will sometimes be important for
                                                                good SSD performance.
                                                                                House Air Through
                                                                                Unclosed Openings
Figure 1.  Sub-slab depressurization (SSD) using pipes inserted down through the slab from indoors.


                                                     8

-------
                               90° Elbow
                                                                               Notes:

                                                                               1
                      Suction Pipes
                      Through Wall
 One possible configuration for a multi-pipe system
                                                Exhaust (Released
                                                Above Eave)
                         The exterior downspout
                         exhaust stack illustrated
                         here is one of several
                         possible stack
                         configurations, as
                         discussed later.
                      2. Sealing pipe penetration
                         through wall is
                         important to reduce
                         leakage of outdoor air
                         and air from block
                         cavities into SSD system.
                              3"X4" Rectangular
                              Metal Downspout

                            4" Round-to-3"X4"
                            Rectangular
                            Adapter
                                    Flexible
                                   Coupling

                   Exhaust Fan (Rated for
                   Exterior Use or Enclosed)
                          Grade Level
                               Suction Pipe

               Header Connecting to Other Suction
               Points If Any (See Inset Above)

                        Horizontal Pipe Sloped
                        Down Toward Pit for
                        Condensate Drainage
       45° Fittings (4" Dia.
       Round PVC) to Offset
      ! Stack against House
             Stud
             Wallboard
             Sill Plate
                                                                           House Air Leakage
                                                                           Through Wall/Floor Joint
                         Suction Pit
 Figure 2.   Sub-slab depressurization (SSD) using pipes inserted horizontally through the foundation wall from outdoors.
beside the footings; this exterior aggregate bed is sometimes
(but not always) covered with a material such as geotechnical
cloth or roofing felt, intended to reduce pluggage of the gravel
bed with dirt.

Both interior and exterior drain tiles can discharge collected
water by either one of three methods. If the lot is sufficiently
sloped, the water can be routed to an above-grade discharge at
a low point on the lot, remote from the house. As an alterna-
tive, the water can be directed to a dry well away from the
house. Or, in basement houses when the lot is not sufficiently
sloped, the water is  drained to a sump  inside the basement,
from which the water is pumped to an above-grade discharge.
This section addresses  the case where the tiles drain to a
sump.                                     ,
A typical sump/DTD system is illustrated in Figure 3. An air-
tight cover is sealed over the sump pit. Suction is then drawn
on the drain tile network by connecting a suction pipe to the
sump/drain tile system. While the suction pipe can be installed
through the sump cover, it can be advisable instead to connect
the suction pipe to the tiles at a location remote from the
sump, as illustrated in the figure. Installation of the suction
pipe remote from the sump will facilitate subsequent mainte-
nance of the sump pump, and will reduce the suction loss that
will result if air leaks develop around the sump cover (e.g., as
a result of improper re-installation of the cover after subse-
quent sump pump maintenance).

The drain tiles shown in Figure 3 are interior tiles, although
exterior tiles can also drain to a sump. Where the drain tiles
                                                       :9

-------
      Exhaust Option 2:
      Exterior Stack
                                                             Notes:
                                                             1.
                                Exhaust Option 1:
                                Interior Stack5

                                To Exhaust Fan
                                Mounted in Attic
                                                             3.
                                                             4.
                                                              Figure depicts suction pipe installed
                                                              remote from sump. Suction pipe could
                                                              also be installed through sump cover.

                                                              Detail shown for pipe penetration
                                                              through slab and connection to drain
                                                              tile can vary. Other options are
                                                              shown later.

                                                              Alternative sump cover designs are
                                                              discussed later.

                                                              Options for supporting horizontal and
                                                              vertical piping runs are shown in later
                                                              figures.

                                                              Detail for interior and exterior stacks
                                                              is shown in later figures.

                                                              Closing various slab openings,
                                                              especially the perimeter wall/floor
                                                              joint, will sometimes be important for
                                                              good sump/DTD performance.
                  6&r Flexible
                       Coupling
Strapping
(or Other4
Support)
Exhaust Fan
(Rated for
Exterior Use
or Enclosed)
                                              Straight
                                              Fitting
                                                                   Water
                                                                   Discharge
                                                                   Pipe
                                            Slope Horizontal
                                            Pipe Down Toward
                                            Suction Pipe
                                                                  Caulk or
                                                                  Grommets
                                                                  to Seal
                                                                  Penetrations
                                            Sealant Around
                                            Suction Pipe
                                              Masonry
                                              Bolt
                                                     ,- Sump
                                                    / Cover
                               Hole in
                               Drain Tile
                               Near Suction
                               Pipe
                                      Gasket (or
Grade Level
                                      Silicone Caulk)
                                                                               Existing Interior
                                                                               Drain Tile Loop
                                                                               Circling House
                                                                               Sump Liner

                                                                               Submersible
                                                                               Sump Pump
Figure 3.  Drain-tile depressurization (DTD) where the tiles drain to a sump in the basement.

                                                     10

-------
are exterior tiles, and where the suction pipe laps into these
exterior tiles remote from the sump, the piping will connect to
the tiles outside the house and will look more like that shown
in Figure 4 for DTD/remote discharge systems. (Of course,
when exterior tiles drain to a sump, the remote discharge line
in Figure 4 will be absent, and the suction pipe will thus be
connecting to the  drain tile loop immediately  beside the
footings rather than to the discharge line.)

Where the tiles are interior tiles, the principles involved with
this DTD system are basically the same as those for the SSD
variation described in the previous section. However, DTD on
interior tiles offers two key advantages. First, the tiles provide
a network that helps distribute the suction under the slab and
are located in a zone which will necessarily always have been
excavated and backfilled during construction (and hence will
generally have some communication, even under houses where
little or no aggregate has been placed). Second, the tiles are
located right beside two of the major soil gas entry routes: the
joint between the perimeter foundation wall arid the concrete
slab inside the house; and the perimeter footing region where
soil gas can enter the void network inside block foundation
walls. Thus, interior drain tiles provide a convenient, in-place
network that enables suction to be  easily and  effectively
drawn over a wide area, particularly where it is usually needed
the most. These features are particularly important in cases
where sub-slab communication is marginal  or poor; they are
less important in cases where there is a good layer of sub-slab
aggregate.

Because sump/DTD with interior tiles generally  ensures ef-
fective suction immediately beside the footings, this approach
may be more likely than SSD to treat the block walls at points
around the perimeter where the tiles are present.

Where the tiles draining into the sump are  exterior tiles, the
sump/DTD system will still be expected to divert soil gas
away from the void network in block foundation walls. How-
ever, because  the suction on exterior tiles is being applied
outside the footings, any suction field created beneath the slab
will depend upon extension of the exterior suction through the
bottom course of blocks or beneath the footings, into the sub-
slab region. This extension, in turn,  will depend upon the
permeability of the underlying native soil. Limited suction
field data from sump/DTD systems with exterior tile loops
suggest that, as expected, the sub-slab depressurizations cre-
ated by such systems can be lower than those by systems
having interior loops. Suction on an exterior loop does appear
to treat the wall/floor joint inside the footings, at locations
where tile is present, but the reduced sub-slab depressuriza-
tions might result in lesser treatment of slab-related entry
routes toward the central portion of the slab.

The chances of achieving effective treatment with sump suc-
tion (with either interior or exterior tiles) are greatest when the
tiles form a nearly complete loop around the perimeter (i.e.,
around at least three sides of the house). However, good
performance can sometimes be achieved even when there is
only a partial loop, on one or two sides. This is especially true
when there is an interior loop and aggregate under the slab.
With exterior  tiles, good performance with partial loops ap-
pears to depend on good permeability in the native soil.
Because of the typical effectiveness of sump/DTD, this ap-
proach should be among the first considered in any basement
house having a sump with drain tiles.  Sump/DTD will com-
monly be the preferred approach when the sump connects to
an interior loop of tiles, whether or not there is a layer of sub-
slab aggregate. However, when there is aggregate, SSD is a
competitive option, and SSD might still be selected instead of
sump/DTD under some conditions  (e.g., if there is a high
degree of floor/wall finish at the location where the suction
pipe would have to be inserted into the drain tiles). Where the
drain tiles form an exterior  loop, SSD in the basement may
sometimes be preferred over sump/DTD, especially if there is
sub-slab aggregate, since SSD will apply the suction beneath
the slab rather than outside the footings. However, sump/DTD
can still be a good choice if there is a reasonably complete
exterior loop of tiles (i.e., on at least three sides of the house),
especially if the owner wants to keep all of the system piping
outside the  house  (tapping the suction pipe into  the tiles
outdoors, remote from the sump). If there are only exterior
tiles, and if these tiles do not form a reasonably complete loop,
suction on the tiles will not be delivered directly to  some
portion of the perimeter; this will be  of greatest concern when
the native soil has low permeability. In such cases, SSD might
warrant greater consideration as the  initial approach.

2.1.3  Active Drain-Tile Depressurization
(Above-Grade/Dry-Well Discharge)

Where the tiles drain to an above-grade discharge or to a dry
well, a different basic DTD design is required, as illustrated in
Figure 4. A  check valve is installed in the discharge line, to
prevent outdoor air from being drawn into the system via the
discharge line. Suction is then drawn on the drain tile net-
work, as illustrated. The suction pipe will often be connected
to the drain  tile loop immediately beside the footings; it can
also connect to the discharge line, as shown in Figure 4. The
drain tiles shown in the figure are exterior tiles, which will
commonly be the case when there is remote discharge, al-
though interior tiles can also drain remote from the house.

The principles involved with DTD/remote discharge are the
same as those described in  the preceding  section for sump/
DTD.

Because DTD/remote discharge can be effective, and because
the system  will be entirely outside  the house  (making  it
potentially less obtrusive and less expensive than other ASD
options), this approach should be among those considered in
any case where remotely-discharging  drain tiles exist. As in
the sump/DTD case, DTD/remote discharge will most often
be  the preferred approach when there is an interior loop of
tiles, especially when there is not a good layer of sub-slab
aggregate; a good aggregate layer  would make SSD more
competitive. DTD/remote discharge will  also often be se-
lected when there is a reasonably complete exterior loop (i.e.,
around at least three sides of the house). When the drain tile
loop is exterior to the footings, there is an increased chance
that SSD may turn out to be the technique of choice, rather
than DTD/remote  discharge, because SSD will  apply the
suction directly under the slab.
                                                        11

-------
  Notes:

  1.  For clarity, suction pipe is depicted
     installed in the drain tile discharge
     line. Commonly, the suction pipe might
     instead be installed directly in the
     tile loop beside the footings.

  2.  Figure depicts one option for
     connecting the rigid 4" diam. suction
     pipe to 4" flexible corrugated
     drain tile. Alternative methods of
     connection to 3" and 4" flexible
     tile, or to rigid plastic or clay
     tile, are presented in text.

  3.  The reverse flow valve shown
     represents one design currently
     on the market. Other valve
     configurations are available.

  4.  The exterior downspout exhaust
     stack illustrated here is one of
     several possible stack
     configurations.
  5. Closing various slab openings,
     especially the perimeter joint,
     will sometimes be important for
     good DTD performance.
                                                  Exhaust (Released
                                                  Above Eave)
                                                                3"X4" Rectangular
                                                                Metal Downspout

                                                                4" Round-to-3"X4"
                                                                Rectangular Adapter
  Reverse Flow
  Valve (In
  PVC Pipe), to
  Prevent Air from
  Being Drawn into
  System from
  Discharge3
  Above-Grade
  Discharge
                                               Exhaust Fan
                                               (Rated for
                                               Exterior Use
                                               or Enclosed)
                    Riser
                    Connecting
                    Drain Tile
                    to Fan
                                             Rigid PVC
                                           . :T-Fitting
  Discharge Line
! (Flexible        ..
 . Corrugated
''Piping)2
                                                                              -.,' Footing  .
Rigid PVC
Straight
Fitting3 -
                                •   •."/' ; ;  . •  Existing Drain Tile  ' .
                               "•  •..;."      Loop Circling House1-^'      '.'•''•'.
Figure 4.   Drain-tile depressurization (DTD) where the tiles drain to an above-grade discharge remote from the house.

                                                      12

-------
2.1.4  Active Block-Wall Depressurization
(BWD)

When the foundation wall is constructed of hollow concrete or
cinder blocks, the interconnected network of block cavities
within the wall can serve as a conduit for soil gas. Soil gas
which enters the wall through mortar joint cracks, pores, and
other openings in the exterior face of the blocks below grade
can move either vertically or laterally throughout the  wall
inside this void network. Where the house has a basement,
with the interior face of the blocks inside the basement, the
soil gas can then be drawn into the house through any open-
ings in the interior face, such as holes  around utility penetra-
tions, mortar joint cracks, and  the pores  in the block itself.
Even more importantly, if the  cavities in the top course of
block open to the basement, the walls  will act as a chimney,
with soil gas flowing up through the void network and into the
basement through these uncapped top voids. Even in slab-on-
grade or crawl-space houses with hollow-block foundations, it
is sometimes possible (if the top cavities  are not closed) for
soil gas in the block voids to flow up into the wooden framing
resting on top of the wall, and hence into the house, even
though the foundation wall itself does  not extend up into the
living area.  ;

The principle of BWD is to draw the soil gas out of this  void
network using a fan drawing suction  on  the voids. The fan
would presumably increase the flow of soil gas into the voids,
but would draw the soil gas into the system piping and exhaust
it outdoors rather than permitting it to enter the basement.
Where the block walls are the primary  entry route, and where
the BWD system is able to adequately depressurize the  void
network,  this approach would most directly treat that entry
route. The  depressurization created within  the voids by a
BWD system will sometimes extend under the slab, depend-
ing upon the communication between  the voids and the  sub-
slab, and upon the communication within the sub-slab fill.
Thus, the wall/floor joint, and perhaps some slab-related entry
routes more remote from the walls, will  sometimes also be
treated by a BWD system.

Figure 5 illustrates one method for implementing BWD.  This
approach, referred to as the "individual pipe"  approach, in-
volves insertion of one or two suction pipes  into the  void
network  in each wall  to be treated; these pipes  are  then
connected to one or more fans. A second approach for imple-
menting BWD, referred to as the "baseboard duct" approach,
has been  used occasionally. In this approach, a series of holes
is drilled into the void network around  the perimeter  of a
basement, just above slab level; these holes are then enclosed
within a  plenum ("baseboard") sealed to the wall and slab
around the perimeter, and the plenum is connected to a fan.
For clarity, Figure 5 shows BWD being used as a stand-alone
technique; however, more commonly where BWD is used, the
primary mitigation system will be a SSD system, with depres-
surization of one or  more selected walls implemented  as a
component of that SSD system.

A key problem with BWD is that the numerous and often-
concealed wall openings (especially open top wall voids, and
block pores) are  very difficult to close adequately. Thus,
despite efforts to close these openings,  large amounts of house
air (and possibly outdoor air) will leak into the BWD system
through these openings. Therefore, it has often proven to be
difficult or impractical to maintain sufficient depressurization
throughout the entire wall. Thus, the wall-related entry routes
have sometimes not been adequately treated (along with cen-
trally located slab-related routes), with the result that radon
reductions have not been consistent either within a given
house over time, or between houses, when BWD is used as a
stand-alone method. As an added concern, substantial house
air leakage into a BWD system has sometimes depressurized
the basement sufficiently to cause backdrafting of fireplaces
and other combustion appliances (as well as increasing the
heating/cooling penalty of the system). Where backdrafting
occurs, an outdoor supply of combustion air must be pro-
vided, or else the BWD system might be operated in pressure
instead of suction. Basement depressurization resulting from
BWD systems can also increase soil gas influx through wall-
and slab-related entry routes not adequately depressurized by
the system, thus reducing net radon reduction performance,
and sometimes even increasing basement concentrations (e.g.,
House 19 in Reference Sc89).

In view of these concerns, BWD is looked upon as a technique
which would be used largely as an occasional supplement to
SSD systems, rather than as a method that would commonly
be used by itself. The role of BWD as a supplement to SSD
can be very important  in  some cases. Even where a SSD
system has very effectively depressurized the entire sub-slab
region, occasional cases have been encountered where this
sub-slab depressurization did not adequately prevent soil gas
from entering the void network, requiring simultaneous treat-
ment of  the walls  using  BWD (e.g., Houses .3 and 16  in
Reference Fi91).

2.1.5  Active Sub-Membrane
Depressurization (SMD) in Crawl Spaces

In houses having a crawl space with a dirt floor (including dirt
floors covered with gravel and/or with a plastic vapor barrier),
a variation of SSD or DTD can be implemented if a "slab" (in
the form  of a plastic membrane) is placed over the dirt floor.
Suction can be drawn underneath this plastic "slab," either
using individual suction pipes penetrating the membrane
(analogous to SSD), or through suction on a length or a loop
of perforated drain  tile placed beneath the plastic (analogous
to DTD). Examples of both SMD approaches are illustrated in
Figures 6 and 7.

SMD has been demonstrated to be the most effective radon
reduction method for crawl-space houses. It should be one of
the first methods considered in any crawl-space house where
the crawl space is accessible for system installation and where
the required radon  reductions are sufficiently great (greater
than about 50%) such that natural ventilation of the crawl
space will not be adequate. Use of an exhaust fan to depres-
surize the entire crawl space,  preventing radon-containing
crawl-space air from flowing up into the living area, has also
been found to be effective,  as well as less expensive than
SMD to  install. However, crawl-space depressurization is
usually less effective than SMD, and has a much  higher
heating penalty, since  much of the air exhausted by  the
                                                        13

-------
                                                                  Notes:
          Exhaust Option 2
          Exhaust Released
          Above Have
 Flexible
 Coupling
    Exhaust Fan
    (Rated for
    Exterior Use
    or Enclosed)
1.  Closure of top block voids can be
   very important to avoid degradation
   of BWD performance and increased
   heating/cooling penalty caused by
   excessive leakage of house air into
   the system.

2.  Options for use of individual pipe
   BWD as a supplement to SSD are
   illustrated in a later figure.
r Veneer Gap
          Exhaust Option 1
                                         To Exhaust Fan
                                         Mounted in Attic

                                                           — Strapping (or
                                                              Other Support)
      Close Top Voids1
      Brick Veneer
                                                        Suction Pipe
                                                        Close Major Mortar Cracks and
                                                        Holes in Wall
                                                     Basement Air Through Block Pores,
                                                     Unclosed Cracks, and Holes
      Concrete Block
                                                           6" Dia.
                                                           Collection
                                                           Pipe
                                                                                     From Connections
                                                                                     into Other Walk;
Figure 5.  Block-wall depressurization (BWD) using the individual-pipe approach.

                                                    14

-------
        Exhaust Option 2

       Exhaust Released
       Above Eave
                     Flexible
                     Coupling

                     Exhaust Fan
                     (Rated for
                     Exterior Use
                     or Enclosed)
    Hollow-Block
    Foundation Wall
 Grade Level
  Notes:

  1.  The specific configuration depicted for the pipe penetration
     through the membrane is one of a number of alternatives.
     Other options are shown in a later figure.

  2.  The membrane seams must always be sealed near the suction
     point. Sealing of more remote seams may not always be
     necessary, but is advisable.
                                         3.
     The membrane can often be effectively sealed against the
     foundation wall using a continuous bead of properly selected
     sealant (urethane caulk for cross-laminated polyethylenes,
     other adhesive for regular polyethylenes). Other options
     for sealing the membrane against the wall are discussed in
     text.

               Exhaust Option 1
                                                       To Exhaust Fan
                                                       Mounted in Attic
                                         Crawl
                                         Space
Membrane Sealed
Against Wall
with Bead of
Caulk or  23
Adhesive'

Hose Clamp and
Caulk, Sealing
Membrane
                                      •  • • • •  Dirt Floor in  •
                                        .'  '.'  Crawl Space .

                                       PVCT-Fitting Under
                                       Membrane, to Support •
                                       Pipe and to Help  1
                                       Distribute Suction
                                                                           Strapping (or Other
                                                                           Support) Will Sometimes
                                                                           Be Necessary
                                                                                    Connection to
                                                                                    Other Suction
                                                                                    Point(s), If Any
                             Slope Horizontal Legs
                             Down Toward Membrane
                                                                           PVC Suction Pipe
Semi-Rigid Plastic
Plate Resting on Top
of the T-Fitting, to
Prevent Membrane from
Being Sucked into the
Ends of the T-Fitting1
                                                                            Membrane
                                                                                       Sealant
                         Adjoining Sheets of
                         Membrane Overlapped
                         by about 12 Inches
                         Sealed with Caulk or '
                         Other Adhesive2    —
Figure 6.  Sub-membrane depressurization (SMD) for the case where individual suction pipes penetrate the membrane (SSD analogue).


                                                    15

-------
                                            Notes:
       Exhaust Option 2
       Exhaust Released
       Above Eave
                     Flexible
                     Coupling

                     Exhaust Fan
                     (Rated for
                     Exterior Use
                     or Enclosed)
              1. The perforated piping is depicted here as a straight
                length down the center of the crawl space.
                Alternative configurations are discussed in the text.

              2. The perforated piping depicted in the inset is
                flexible corrugated piping. Rigid perforated
                Schedule 40 pipe could also be used.

              3. The membrane seams should always be sealed near
                the perforated piping. Sealing of remote seams may
                not always be necessary, but is advisable.

              4. The membrane can often be effectively sealed against
                the foundation wall using a continuous bead of an
                appropriate sealant. Other options for sealing the
                membrane against the wall are discussed in the text.

                        Exhaust Option 1
                                                       To Exhaust Fan
                                                       Mounted in Attic
                                       - Floor
                                          Strapping (or Other
                                          Support) Will Sometimes
                                          Be Necessary
                      L
        Sealant
       Hollow-Block
       Foundation
       Wa|l.	
Grade Level
                                       Joist

                                       Slope Horizontal
                                       Leg Down Toward
                                       Membrane
i— Membrane Sealed
   Against Wall with
   Bead of Qaulkor
   Adhesive **
                  Sealant

                  Hose
                  Clamp
                                       Length of Perforated
                                       Pipe at Central Location, •
                                       Connected to Suction Pipe
                                       Using Rigid PVC T-Fittinq
                                       or 90° Elbow (See Inset)
                                     Dirt Floor in
                                     Crawl Space .

                                        Adjoining Sheets of
                                        Membrane Overlapped
                                        by About 12 Inches,
                                        Sealed with Caulk 3
                                        or Other Adhesive
Figure 7.  Sub-membrane depressurization (SMD) for the case where suction is drawn on perforated piping beneath the membrane (DTD
         analogue).
                                                    16

-------
depressurization fan will be treated house air drawn down
from the living area.

The principles involved in SMD are similar to those for SSD
and DTD.

   -  If adequate  suction can be maintained  beneath  the
      plastic, soil gas entry into the crawl space should be
      reduced or eliminated. By this sub-membrane depres-
      surization  mechanism,  gas flows through seams or
      punctures in the membrane should be crawl-space air
      being drawn down into the sub-membrane region, rather
      than soil gas flowing up into the crawl  space, from
      which it might enter the living area.

   -  The other mechanisms discussed earlier for ASD sys-
      tems may also come into play: dilution of the soil gas
      radon levels under the membrane, resulting from the
      increased flows of air into  the sub-membrane region
      (from the crawl space, the block walls, and outdoors);
      and air-barrier shielding.

   -  In crawl  spaces having block foundation walls,  the
      SMD system might help prevent soil gas entry through
      the block cavities into the  crawl space and into  the
      living area overhead, analogous to SSD and  DTD.
      However, there are not sufficient data in houses where
      block crawl-space walls are a major living-area entry
      route to assess how consistently or reliably SMD will
      thus treat the walls.

In addition, there may be an element of crawl-space depres-
surization contributing to the performance of SMD systems.
Significant leakage of crawl-space air into the SMD system
through membrane openings can cause such depressurization
of the entire crawl space relative to the living area, reducing or
completely reversing the normal flow of crawl-space air up
into  the living  area.  This can be an effective mitigation
method in its own right (He92).

Where the crawl space has a floor of bare earth, which is often
the case—i.e., where there is not a layer of gravel on the floor
beneath the membrane—suction beneath a SMD membrane is
analogous to SSD with no aggregate under the slab. Thus, just
as with SSD in houses having poor sub-slab communication,
the sub-membrane depressurizations maintained in SMD sys-
tems are generally fairly low, and measurable suction field
extension beneath  the membrane is limited.  These lower
depressurizations may mean that SMD systems will be less
likely;  e.g., to treat wall-related entry routes, in cases where
that might be necessary. However, SMD systems have gener-
ally proven to be very effective in crawl-space houses, despite
reduced sub-membrane suctions.

Where the crawl space is large, where significant reductions
in indoor radon concentrations are desired (e.g., to 2 pCi/L or
less), and/or where the soil comprising the crawl-space floor
has very low permeability (with no gravel on the floor), it has
sometimes been found necessary to take steps to help the sub-
membrane suction field  extend adequately around the crawl
space.  The most common steps are installing additional suc-
tion  pipes  through  the  membrane (analogous  to  installing
multiple suction pipes in a SSD system), or connecting the
suction system to perforated piping under the membrane (as in
Figure 7).

A key remaining issue in the design of SMD systems is how
much effort should be made to reduce crawl-space air leakage
into the system by sealing the membrane against perimeter
foundation walls and interior piers, and at seams between
sheets of plastic. Some results suggest that such comprehen-
sive sealing of the membrane may not always be necessary in
order to achieve good radon reductions in the living area. This
result is probably a commentary on the limited sub-membrane
depressurizations needed to adequately reduce soil gas flow
into the crawl space in some cases. It may also be a commen-
tary on the relative benefits of the mechanisms other than sub-
membrane depressurization discussed above, especially sub-
membrane soil gas dilution and crawl-space depressurization.

Comprehensive sealing of the membrane is often conducted
routinely by mitigators. Mitigators report that complete seal-
ing is needed in order to ensure the best radon reduction
performance. Complete sealing also will help avoid possible
backdrafting of combustion appliances in the crawl space as a
result of the crawl-space depressurization that can occur if the
SMD fan draws too  much crawl-space air  out through the
unsealed seams. Furthermore, complete sealing will reduce
the amount of treated air drawn  out  of the living area and
exhausted by  the SMD system, thus reducing the heating/
cooling penalty. Until further research confirms the condi-
tions  under which incomplete membrane sealing is accept-
able, it is advisable to seal the membrane everywhere.

This issue extends to the question of whether the entire crawl-
space floor even has to be covered by the membrane. This
question is important because portions of crawl spaces are
often  inaccessible or are occupied by, for example, the fur-
nace and water heater. Good radon reductions have some-
times been achieved even in cases where the dirt floor was not
entirely covered. This result could be suggesting that a) some
sections of the dirt floor are less significant sources of radon
than others; b) suction on a portion  of the soil can extend
sufficiently to intercept the soil gas before it reaches the
uncovered sections of floor; and/or c)  depressurization of the
entire crawl space is contributing  to the observed reductions,
through a  different mechanism than soil depressurization.
Until  this question is better understood, it is advisable to
ensure that the entire crawl-space floor is covered by the
membrane, if at all possible.

In isolated cases, relatively good radon reductions have been
achieved with "sub-membrane depressurization" without the
membrane, i.e., by drawing suction on pipes embedded in the
bare soil in the crawl space. Referred  to here as "site ventila-
tion," this approach  is analogous to the "radon wells" in
Sweden, and would appear to require certain geological char-
acteristics in order to be successful. The soil would have to be
relatively permeable  laterally, and less permeable vertically
toward grade,  so that the suction field would extend through
the soil over the area of the crawl space, but would not be
dissipated by extensive leakage of crawl-space air down from
above. A strata of porous, gravelly soil, capped with a layer of
clay soil, would be an ideal example of a geology conducive
                                                        17

-------
to this approach. Site ventilation cannot be expected to be
sufficient by itself in most locations. However, the fact that
SMD can sometimes be effective even when the membrane
does not cover the entire floor could be indicating that a site
ventilation component may exist in some SMD installations.

2.2    Applicability of Active Soil
Depressurization
Where properly designed and operated, ASD systems (espe-
cially SSD, DTD, and SMD) have consistently demonstrated
high,  reliable radon reductions in a wide range of houses.
Radon reductions greater than 90% are common when pre-
mitigation levels are significantly elevated. Performance has
been demonstrated in thousands of commercial installations
(and numerous experimental installations) over the past 8
years, in houses  having the full range of substructure types.
Installation costs of  ASD systems are typically moderate.
Exact costs can vary from house to house: a range of $800-
$2,500 appears generally representative for most commer-
cially installed ASD systems (He91b, He91c, Ho91, EPA92c),
with an average  cost of $1,135 for SSD and DTD obtained
from  an EPA survey of private-sector mitigation  (Ho91).
Other radon reduction approaches have one or more of the
following disadvantages, relative to ASD:

a)  Some other approaches are more expensive to install than
    ASD. Air-to-air  heat exchangers for house ventilation
    have an average  installation cost of $1,606 according to
    Reference Ho91. Entry route sealing as a stand-alone
    method can  sometimes be more expensive than ASD,
    depending upon the extent of the sealing and the amount
    of floor/wall finish that must be removed and re-installed
    to permit the sealing.

b)  Most other approaches are less effective than ASD. Air-
    to-air heat exchangers provide reductions no greater than
    25-75%; entry route sealing as a  stand-alone  method
    provides 0-50%;  natural crawl-space ventilation provides
    0-50%.

c)  Other approaches are generally less well demonstrated.
    House pressurization has been successfully tested in only
    a limited number of houses; most of the other alternatives
    to ASD listed in a) and b) are also less well demonstrated
    than ASD.

In view of their demonstrated high radon reductions in a wide
variety of houses, and in view of their moderate cost, ASD
systems should be one of the first approaches considered in
essentially any house. These systems are applicable in treating
houses having very high pre-mitigation radon concentrations
(e.g., above 100 pCi/L). ASD will generally be required in
order to reliably achieve 4 pCi/L and less in any house where
the pre-mitigation level is above about 8 to 16 pCi/L. (Only
house pressurization and crawl-space depressurization appear
to offer potential for achieving reductions as great as ASD,
and these other techniques are much less well demonstrated,
ate not as widely applicable, and can offer other complica-
tions.) And ASD can  also be cost-effective in treating houses
having only slightly elevated pre-mitigation levels (e.g., 4 to
10 pCI/L), potentially providing greater and more  reliable
reductions at less long-term capital and operating cost than the
other alternatives for achieving moderate reductions, namely,
house ventilation and entry route sealing.

Techniques other than ASD would be considered most seri-
ously in unusually difficult houses having combinations of the
following complications:

a)  very poor sub-slab communication, requiring a large
    number of suction pipes;

b)  fieldstone foundation walls which are an important entry
    route, since such walls can be difficult to treat and are not
    amenable to BWD;

c)  a very high degree of floor/wall/ceiling finish on the story
    in contact with the soil, thus complicating the placement
    and routing of suction pipes;

d)  homeowner resistance to aspects of the ASD system, e.g.,
    appearance, need to maintain a fan, etc.

Generally, any one of the above complications, by itself,
could be overcome by suitable design. However, especially
when there  are multiple problems,  the effort required to
overcome them could make the ASD system more expensive
(or would create a greater aesthetic  impact) than would an
alternative mitigation approach, such as house pressurization
or house ventilation. Of course, the  applicability of one of
these other approaches would itself depend upon other charac-
teristics of the house, such as the pre-mitigation radon level
and the natural ventilation rate of the house (which would
determine the effectiveness of house ventilation), and  the
tightness of the basement (which would impact basement
pressurization). Mitigation experience to date suggests that
only a few percent of the houses in this country are likely to be
truly inappropriate for ASD (Bro90, Mes90b, Sh90, St90,
We90).

Where ASD  is being considered for a specific  house,  the
applicable variation of the technology is usually  selected as
follows:

a)  In basement houses having sumps with drain tiles, sump/
    DTD will be one of the first ASD options considered.
    Sump/DTD will most likely be successful when the drain
    tiles form a loop inside the  footings, or when there is
    good aggregate beneath the slab regardless of the drain
    tile configuration. Sump/DTD will least likely be suc-
    cessful when the drain tiles are outside the footings and
    form only a partial loop (i.e., a  loop around fewer that
    three sides of the house).

b)  In basement houses having drain tiles  draining to a re-
    mote above-grade discharge or  dry well, DTD/remote
    discharge will be one of the first options considered, with
    the same considerations listed in a) above for the sump/
    DTD case. It should be anticipated that a supplemental
    SSD system could turn out to be required, in addition to
    the  DTD system if installed, if  the drain tiles form an
    exterior loop around fewer than three sides of the house.
                                                        18

-------
c)  In any basement or slab-on-grade house, SSD will gener-
    ally always be one of the first ASD variations considered.
    Where the house has drain tiles, SSD can be considered
    as an alternative or  supplement to sump/DTD or DTD/
    remote discharge. The exact design of the SSD system
    will have to be developed reflecting the degree of sub-
    slab communication, degree of interior finish, etc.

d)  BWD would generally never be installed initially as a
    stand-alone mitigation method unless  prior experience
    with similar houses in the locality indicated that the block
    walls were the predominant source and were sufficiently
    leak-tight  (e.g., top  voids capped), such that BWD of-
    fered clear potential for being  the most cost-effective
    approach in that house. Likewise, unless experience with
    other similar houses in the locality suggested otherwise, it
    would generally not be efficient to install a BWD compo-
    nent on a SSD system, until initial operation of the SSD
    system confirmed that it was not effectively treating the
    wall-related  entry routes and that a BWD  component is
    indeed required.

e)  SMD would be the ASD technique used in any crawl-
    space house, or hi the crawl-space wing of a combined-
    substructure house if that wing is found to need treat-
    ment.

The subsections below list in further detail the specific house
features which determine when each of the individual ASD
variations is likely to be most applicable.

2.2.1  Active Sub-Slab Depressurization

SSD will be most applicable under the following conditions.

 •  In any house having a concrete floor  slab (basements,
    slabs on grade, and paved crawl spaces).

 •  In houses  having good communication immediately be-
    neath the slab, although houses having poor communica-
    tion can also be treated. Good  communication is most
    commonly associated with a good layer of aggregate
    (gravel or crushed rock) under the slab. Poor communica-
    tion can result from  a) lack of, or uneven distribution of,
    aggregate  (combined with a relatively  impermeable na-
    tive soil underlying the slab); or from b) sub-slab obstruc-
    tions (such as forced-air supply ducts and interior foot-
    ings or grade beams) which interrupt the aggregate.

    With good communication, design of  a SSD system is
    simplified, with only one or two suction pipes commonly
    needed, and with flexibility in choosing where the pipes
    are located.  Poor communication does not render SSD
    inapplicable; however, it does require more care in select-
    ing the number and location of suction pipes, and perhaps
    also in other design aspects  (such as a higher-suction
    fan). Mitigation performance  might still be reduced de-
    spite this increased care. Very poor communication (re-
    quiring a large number of carefully located suction pipes)
    combined with extensive interior finish  (complicating the
    siting and installation of pipes indoors) could warrant
    more serious consideration of mitigation techniques other
than SSD, or of installation of the SSD pipes from
outdoors. Some obstructions that cause poor communica-
tion, such as sub-slab forced-air supply ducts, will some-
times not significantly increase the number of suction
pipes required, if there is a layer of aggregate (Fi90,
He91a). In such cases, the system may be functioning
more by soil gas dilution or by air-barrier shielding (as
discussed in Section 2.1),  rather than  by creating an
unambiguous depressurization beneath the slab.

In houses which a) do not have drain tiles (so that DTD is
not an option); or b) which have only a partial loop of
tiles outside the footings (so that DTD is likely to leave
some portion  of the perimeter .foundation  untreated).
However, if sub-slab communication is good, SSD can
sometimes still be the best selection even in a house
having drain tiles, in cases where the system pipe routing
is simplified by avoiding the need to tap into the drain
tiles; the extent or condition of the drain tiles is question-
able; or rain gutter downspouts, window well drains, or
other such drains empty into the drain tiles, making the
tiles difficult to seal and depressurize.

In houses having any pre-mitigation radon concentration
(high, moderate, or only slightly elevated).

SSD is capable of achieving the very high radon reduc-
tions required in houses having extremely high pre-miti-
gation concentrations (e.g., above 100 pCi/L). It (along
with DTD, where drain tiles exist) is the most reliable
approach available for providing the necessary reductions
in any basement or slab-on-grade house requiring more
than about 50-75% radon reduction (i.e., any house hav-
ing pre-mitigation concentrations greater than 8 to 16
pCi/L). The one other approach which appears capable of
achieving reductions greater than 50-75%, in addition to
ASD, is house pressurization. This technique has been far
less well demonstrated than ASD, and is applicable pri-
marily when the basement can be reasonably well iso-
lated from the living area to permit pressurization of the
basement. (The presence of forced-air furnace ducting
between the basement and the living area could seriously
complicate efforts to isolate the two levels.)

Even when pre-mitigation concentrations are less than 8
to 16 pCi/L, and other less effective techniques can be
considered (such as house ventilation and entry route
sealing), SSD is still a viable candidate approach. It can
be less expensive than some of these other approaches
(e.g., air-to-air heat exchangers  and extensive sealing
efforts), and will commonly provide much greater radon
reductions. For example, a house having a pre-mitigation
level of 6 pCi/L might be expected to be reduced to 3 pCi/
L by a house ventilation approach that doubled the venti-
lation  rate,  but  could  realistically be  expected to be
reduced to 1 to 2 pCi/L with SSD. This is important if an
objective is to achieve near-ambient levels indoors.

In houses with hollow-block foundation walls as well as
poured concrete foundation walls. SSD can also be appli-
cable in houses having fieldstone walls, although there is
an increased chance that some other mitigation technique
                                                        19

-------
    will be needed to supplement or replace SSD. While
    hollow-block walls increase the significance of wall-
    related entry routes, and thus increase the chance that
    some additional wall-related treatment might be needed,
    adequate reductions can commonly  be achieved  with
    SSD alone in houses having such walls. If there is reason-
    ably good sub-slab communication, and/or if the  SSD
    suction pipes are located sufficiently close to the walls,
    SSD often (but not always) appears to adequately reduce
    or prevent soil gas entry through wall-related routes. In
    the case of block walls, the system is probably intercept-
    ing the soil gas in the vicinity of the footings, reducing or
    preventing its entry into the void network through mortar
    joint cracks and other openings near the base of the wall.
    Depending upon the permeability of the native soil, the
    suction field may also extend under the footings, possibly
    treating entry routes on the exterior face of block or
    ficldstone walls.

    Depending upon the nature and significance of the wall-
    related entry routes, and the ability of SSD to intercept
    soil gas before it  reaches the exterior face of the wall,
    there are definitely cases SSD does not adequately reduce
    or prevent entry  through wall-related routes. In those
    cases, SSD will have to be supplemented by some addi-
    tional wall treatment (e.g., BWD treating selected block
    walls).

    In houses having  fieldstone foundation walls, SSD can
    still be a reasonable choice (or a reasonable component of
    a combined mitigation system) if the slab is an important
    entry route. However, if sub-slab communication is poor
    and if the fieldstone wall is a major entry route, SSD may
    not be able to develop an adequate suction field or air-
    flow in the soil to intercept the soil gas entering through
    the wall. In such cases, a technique other than SSD may
    be needed—basement pressurization,  house ventilation,
    or sealing or isolation/ventilation of the fieldstone wall.

 •  In houses where  at least a portion of the slab is not
    finished, so that suction pipes can be installed through the
    slab where required in an aesthetically acceptable manner
    without disturbing the existing  wall, floor and ceiling
    finish. However, even where the slab is entirely finished,
    pipes  can usually be  a) installed in  an inconspicuous
    location (e.g.,  in closets);  or b) concealed behind new
    finish  (e.g., boxed in  behind  new wall-board); or c)
    inserted under the slab from outside the living area, as in
    Figure 2, Some of these steps  in finished houses will
    increase costs, but will often not be so severe as to render
    SSD inapplicable  in practice. Almost all houses have at
    least some portion of the slab unfinished (e.g., a utility
    room); where sub-slab communication is good, that lim-
    ited unfinished space can be sufficient, wherever it hap-
    pens to be located on the slab.

In summary, SSD can be one of the first options considered in
any house having a slab. It can be considered regardless of the
sub-slab communication, thepre-mitigation radon concentra-
tion, the nature of the foundation wall, or the degree of interior
finish. In general, only very poor sub-slab communication,
combined with heavy interior finish which limits suction pipe
placement (or combined with fieldstone foundation walls),
will render SSD impractical on the basis of technical and/or
cost considerations. In houses having slabs and no drain tiles
(so that DTD is not an option), and having pre-mitigation
levels above 8-16 pCi/L,  the only other options available
besides SSD for achieving post-mitigation levels of 4 pCi/L
and less are a) BWD; b) basement/house pressurization; or c)
measures involving modifications to the house, e.g., removal
of the existing slab and pouring a new slab over a good bed of
clean, coarse aggregate. These other options will not always
be applicable or practical, either.

2.2.2  Active Sump/Drain-Tile
Depressurization

Sump/DTD will be most applicable under the following con-
ditions.

 •  In any (basement or "slab-below-grade") house having a
    sump pit with drain tiles entering the sump. Sump/DTD
    should always be one  of the first mitigation approaches
    considered when a sump with tiles is present. It should be
    noted that sometimes a sump pit will exist but will not
    have tiles draining into it; suction on such a sump pit
    would not be DTD, but rather, would be SSD, with the
    tile-less sump simply serving as aready-made hole through
    the slab.

 •  In houses where the tiles that drain into the sump form a
    loop beside the footings that is nearly complete (i.e., on at
    least three sides of the house). This is especially true if
    either a) the tiles are outside the footings, and sub-slab
    communication is good to marginal; or b) they are inside
    the footings, and the communication beneath the slab is
    marginal to poor. A nearly complete loop is not important
    if the tiles are under the slab and if there is good commu-
    nication immediately beneath the slab. (An incomplete
    loop would also exist in the case where a portion of a
    complete loop is damaged or blocked with silt.) Sump/
    DTD can also be tried as the initial approach even when
    the  loop is not nearly complete and communication is
    marginal, but there would be an increased chance that this
    initial installation might subsequently have to be supple-
    mented with a SSD system.

    Where the tile loop is outside the footings, the system will
    be depending upon extension of the suction field through
    the  native soil beneath the footings, through the block
    foundation wall, or through below-grade utility channels
    penetrating the foundation, to treat the sub-slab region.
    Likewise, the system will be depending upon extension
    of the field through the native soil to treat sections of the
    footing  where the tile  does not exist. Probably for these
    reasons, the best results with exterior tiles have been
    observed (with good to marginal sub-slab communica-
    tion) when the tiles exist on at least three sides of the
    house (Fi91, K192). Similar concerns exist when the tiles
    are  inside the footings, but sub-slab communication is
    marginal to poor. There have been cases where moderate
    to high radon reductions (40 to 90+%) have been achieved
    with exterior loops on  less than three sides (M87, Sc88,
    He89, K192); the higher reductions with partial exterior
                                                        20

-------
loops likely have been obtained in cases where the native
soil had relatively good permeability.

Where the tiles are inside the footings, and where com-
munication is  good,  the extent  of the tiles  is far  less
significant. Just as a SSD system can perform very well
with only one SSD suction pipe when communication is
very good, likewise a sump/DTD system can perform
wen with a limited segment of drain tile when communi-
cation is good.

In houses without potential major soil gas entry routes
remote from the perimeter walls, in cases with exterior
drain tiles. Such remote  entry routes could include, for
example, interior load-bearing walls (especially hollow-
block walls) or fireplace structures which penetrate the
slab and rest on footings underneath the slab; and interior
expansion joints in the slab. While the suction applied to
an exterior loop of tiles has been shown to extend under
the slab (Fi91, K192), this extension may be weaker than
that with SSD  or with interior tiles. The perimeter foun-
dation walls and wall/floor joint are receiving the stron-
gest  treatment; these perimeter entry routes are com-
monly  the major ones (especially with block walls), so
that effective treatment of these routes is often sufficient
where there are not major interior entry routes. Where
there are interior routes remote from the perimeter, these
routes will provide the system with  an increased chal-
lenge. Available data suggest that, if the exterior loop is
complete and if the fan performance is sufficient, DTD
on exterior loops can produce significant reductions in
indoor radon in houses with such interior entry routes,
especially if the sub-slab communication is good and/or
if the permeability of the native soil is relatively good.
However, the risk of reduced performance is increased.

In houses having any pre-mitigation radon concentration
(high, moderate, or only slightly elevated). In  view of the
high radon removals and the moderate  cost of sump/
DTD systems, they can  be considered for treating any
pre-mitigation level, as discussed for  the case of SSD in
Section 2.2.1.

In houses having any type of foundation wall (block,
poured concrete, or fieldstone), as discussed  for SSD in
Section 2.2.1.

In houses where the area over the slab is largely finished
living space. The installation can generally be  confined to
the area immediately over the sump, if the suction pipe
taps  into the sump. If the suction pipe taps into the riles at
a point remote from the sump,  the installation will be
confined to that one point around the perimeter, which
can be selected to minimize the impact; if the tiles form
an exterior loop, this point will be outdoors. Thus, in
houses having moderate to poor sub-slab communication,
so that a SSD system would require  multiple suction
pipes, it is likely that a sump/DTD system can be installed
with a less significant aesthetic impact or with less sig-
nificant modifications to  the existing finish, compared to
SSD.
 •  In houses where extensive wall finish or other obstruc-
    tions do not hinder installation of a suction pipe into the
    sump or into the perimeter tiles remote from the sump. If
    access to the sump and tiles is constrained, and if there is
    good sub-slab communication, SSD may be preferred
    over sump/DTD even in cases  where a sump with a
    nearly-complete loop of tiles is present.

 •  In houses where the drain tiles do not become flooded,
    i.e., where the sump pump is operating properly. If the
    drain tiles become blocked with water, the suction being
    drawn by the fan will not be distributed around the tile
    loop.

In some cases, there will be some uncertainty whether the tiles
around a given house form a complete loop, or whether they
are partially silted shut, or whether they are inside or outside
the footings. In such cases, judgement must be used. If there is
a reasonable likelihood that the tiles go around three sides of
the house, the advantages of the sump/DTD approach might
make it cost-effective to attempt before proceeding to SSD or
some other approach, especially if only moderate radon re-
ductions (50 to 85%) are required.

As indicated above, there will be cases where SSD may be the
most applicable technique for a given house, rather  than
sump/DTD, even in cases where a sump and a complete loop
of tiles is present. These cases include, for example, houses
where the sump or tiles cannot be conveniently accessed, or
houses where there is a drainage problem.

2.2.3  Active Drain-Tile Depressurization
(Above-Grade/Dry-Well Discharge)

DTD/remote discharge will be most applicable under the
following conditions.

 •  In any (basement  or "slab-below-grade") house having
    drain tiles draining to  an above-grade discharge or dry
    well. DTD/remote discharge should be one of the first
    mitigation approaches considered when a such a drain tile
    system is present.

 •  In houses where the drain tiles form  a loop beside the
    footings that is nearly complete (i.e., on at least three
    sides of the house), if either a) the tiles are outside the
    footings, and sub-slab communication is good to mar-
    ginal; or b) if they are inside the footings, as will com-
    monly  be the case with remote discharge, and the com-
    munication beneath the slab is marginal to poor. If the
    tiles are under the slab and if there is good communica-
    tion immediately beneath the slab, it will be sufficient if
    the tiles form only a partial loop (less than three sides of
    the house)  or form some other pattern. Partial exterior
    loops may be sufficient in cases where the permeability
    of the native soil is relatively good. The rationale for this
    criterion is  as described previously in Section 2.2.2, for
    the sump/DTD case.

 •  In houses without potential major soil gas entry routes
    remote from the perimeter walls, in cases with exterior
    drain tiles, as discussed in Section 2.2.2.
                                                   21

-------
 •  In houses having any pre-mitigation radon concentration
    (high, moderate, or only slightly elevated), as discussed
    for in Sections 2.2.1 and 2.2.2.

 •  In houses having any type of foundation wall (block,
    poured concrete, or fieldstone), as discussed for SSD in
    Section 2.2.1.

 •  In houses where the area over the slab is heavily finished
    living space. Because the DTD/remote discharge system
    is usually installed entirely outside the house, this system
    will tend to be less expensive and less  obtrusive than
    other approaches that could necessitate modifications in
    the finished space.

 •  In houses where the drain tiles do not become flooded, for
    the reason discussed in Section 2.2.2. Hooding would be
    most likely in cases where the tiles drain to a dry well.

2.2.4 Active Block-Wall Depressurization

BWD applies only to houses having hollow-block foundation
walls. Among block-wall houses, BWD will be most appli-
cable under the following conditions.

 •  In houses where a SSD system has already been installed,
    and where post-mitigation diagnostic testing and indoor
    radon measurements indicate that this SSD system is not
    adequately reducing radon entry through the block walls.

    BWD systems, as stand-alone installations (i.e., without a
    SSD component), have  generally  proven less effective
    and less consistent than SSD systems. Therefore, a SSD
    system will often be the first choice, with a BWD compo-
    nent being added only if the SSD system by itself proves
    unable to adequately reduce wall-related entry. Where
    post-mitigation diagnostics indicate that the initial SSD
    system is not adequately reducing radon concentrations
    inside the block walls, a combined SSD+BWD system
    may either be required, or may  be preferable to the
    possible option of adding additional SSD pipes  near the
    walls.

    It can be difficult to predict the need for a BWD supple-
    ment prior to the installation of the initial SSD system; as
    a result, the BWD component may often be added follow-
    ing the initial installation. SSD systems may sometimes
    treat wall-related entry through interception of the soil
    gas before it enters the void network,  rather  than by
    actually creating a measurable depressurization within
    the block cavities. As a result, if pre-mitigation  sub-slab
    suction field extension measurements are conducted us-
    ing a diagnostic vacuum cleaner, failure of the vacuum
    cleaner to create measurable depressurization inside the
    cavities might not necessarily mean that an operating
    SSD system will notadequately reduce radon entry through
    the walls. In addition, high radon concentrations in the
    walls prior to mitigation might well not reflect  the con-
    centrations that  will exist after a SSD  system begins
    operation. Thus, the ability to use pre-mitigation  diagnos-
    tics to foresee when  a SSD system  will need to be
    supplemented by a BWD component will depend upon
 experience in a given locale, where trends may become
 apparent of the conditions under which a BWD compo-
 nent is typically required.

 In houses where one or more of the block walls is a
 particularly important entry route, especially in cases
 where sub-slab communication is sufficiently poor such
 that it will be more difficult for a SSD system to address
 this wall-related entry. Notwithstanding the fact that SSD
 will usually be the first technique of choice, as discussed
 above, many mitigators can relate experiences where a
 stand-alone BWD system treating one or two selected
 walls proved to be extremely effective.

 In houses where there are no major openings in the block
 walls, or where the openings are accessible for reason-
 ably convenient closure. This includes not only the pe-
 rimeter foundation walls, but also any interior block walls
 which are to be treated by the BWD system, and which
 penetrate the slab and rest on footings underneath the
 slab. Block walls  are commonly so leaky  that large
 amounts of air are drawn through the walls and into the
 BWD system from the basement  (assuming a basement
 house, the substructure to which BWD is most commonly
 applied). In addition to creating a significant house heat-
 ing/cooling penalty, this leakage makes it very difficult to
 maintain a suction field inside the wall cavities, poten-
 tially resulting in radon reductions which are insufficient
 and unpredictable, and which are variable over time. It is
 impractical to make a block wall air-tight; even painting
 the wall to close the block pores and hairline mortar joint
 cracks would not make the wall air-tight, and such exten-
 sive effort  is  not  generally required for BWD  to be
 reasonably successful. However, it is crucial that major
 openings not be present  in the wall to be treated, or, if
 present,  that they be closed.

 In particular, BWD treatment of a wall will most likely be
 successful in cases where

-  a course of solid cap block closes the top of the wall.
   Or, if  there is no solid cap block, the open voids in the
   top course are accessible for effective closure.

-  there is no fireplace or chimney structure built into the
   wall, potentially concealing routes for air leakage and
   soil gas entry.

-  there  is no  exterior brick veneer, concealing  a gap
   between the veneer and the interior block or sheathing
   through which air can flow  down into the void net-
   work.

-  the  block does not have particularly  high porosity,
   since high porosity facilitates air flow through the face
   of the block. True cinder block (as distinguished from
   concrete block) is often highly porous. Concrete block,
   which is much more common than cinder block, will
   occasionally have higher-than-normal  porosity when
   there is a reduced amount of cement present in the mix
   from  which the  blocks were fabricated. Particularly
   porous blocks are characterized by more sharply-de-
                                                        22

-------
   fined grains of aggregate on the surface, deeper pits
   between the grains,  and a rougher texture. In less
   porous blocks, these features are more smoothed out by
   the cement.

-  the wall is reasonably integral, and does not contain an
   excessive number of wide mortar joint cracks or miss-
   ing mortar. (All walls will have some hairline mortar
   joint cracks.)

 In houses where the block wall is reasonably accessible,
 i.e., is not covered with sheetrock or panelling, and is not
 otherwise blocked. Effective treatment of a wall will
 require one or more individual pipes penetrating that
 wall, or a baseboard duct running the length of the wall. If
 the wall is difficult to access, this can add significantly to
 the installation cost.

 In houses where there are no obvious major slab-related
 soil gas entry routes remote from the wall, In cases where
 BWD is being considered as a stand-alone method with-
 out SSD. EPA data (He87, Fi91) indicate that BWD by
 itself can create a weak suction field beneath the slab,
 potentially treating some slab-related entry routes remote
 from the wall. However, this suction field will not always
 extend effectively under the slab, even if the wall open-
 ings are  effectively closed. Thus, houses with  badly
 cracked slabs, for example, would not be good candidates
 for BWD, except in conjunction with SSD.

 In conjunction with SSD, individual-pipe BWD is appli-
 cable in houses having any pre-mitigation radon concen-
 tration (high, moderate, or slightly elevated), as discussed
 in Section 2.2.1. When used in conjunction with SSD, the
 BWD treatment can some-times be limited to perhaps
 one or two walls which diagnostics indicate are the major
 entry routes not being treated by the  SSD system. The
 likely moderate cost of such a SSD+BWD system, com-
 bined with its high effectiveness, makes it a good candi-
 date even when only a moderate degree of radon reduc-
 tion is required.

 As a stand-alone technique, however, BWD can some-
 times be more expensive, if all walls must be treated. This
 could limit its applicability to houses having pre-mitiga-
 tion concentrations greater than 8 to 16 pCi/L, where the
 required level of reduction would be sufficient to justify
 the increased system cost

 The increased costs result from the possible need to treat
 all walls in the stand-alone case rather than just one or
 two, which increases the cost of pipe/duct installation and
 of wall sealing. In the case  of the baseboard duct BWD
 variation, installation of the baseboard duct around the
 entire perimeter has consistently made this  the most
 expensive ASD approach. The average installation cost
 of the  baseboard variation  according to EPA's survey
 was $1,588  (Ho91),  compared to $1,135 for SSD, al-
 though costs above $2,000 could be anticipated in some
 cases. In the case of the individual-pipe BWD variation,
 the average cost reported in the survey was $1,045;  this
 cost is somewhat less than that of SSD, suggesting  that
    not all of the walls may have had to be treated by some of
    the respondents to the survey. EPA's  experience has
    indicated that where any significant closure of wall open-
    ings is required (e.g., closure of open voids in the top
    course of block), and where all walls must be treated, the
    labor involved will almost always result in an installation
    cost greater than the average reported in the survey.

 •   In houses of any substructure where the block foundation
    walls  can provide an entry route into the living area.
    BWD has most commonly been used in houses having
    basements.  However, on occasion,  BWD components
    have also been found to be an important addition to SSD
    installations in slab-on-grade houses, in some cases where
    open voids on the top of the stem foundation wall (at slab
    level)  provided access to the living area. No  data have
    been found defining  experience  with BWD  in  crawl-
    space houses having block foundations.

 •   The baseboard duct BWD approach would best be con-
    sidered in houses having a perimeter channel drain (some-
    times referred to as a "canal" drain  or "French" drain)
    around the perimeter wall/floor joint. SSD would still be
    one of the first  mitigation approaches considered, even
    where a perimeter channel drain is present. However, the
    drain  should  be closed  in an appropriate manner, as
    discussed in Section 4.7.1, in order to ensure effective,
    reliable performance of the SSD system and to reduce the
    heating/cooling  penalty; this closure will  increase the
    cost of the SSD system. Application of the baseboard
    duct BWD  approach where a perimeter channel drain is
    present covers the drain, a step  which is likely  to be
    required regardless of the mitigation approach used, while
    a) taking advantage of this ready-made access under the
    slab to provide sub-slab treatment around the entire slab
    perimeter; and b) uniformly treating the wall voids close
    to the footing region. When used as  a stand-alone tech-
    nique, baseboard  duct BWD over a perimeter channel
    drain is in fact a combination of SSD and BWD.

2.2.5 Active Sub-Membrane
Depressurization

SMD will be most applicable under the following conditions:

 •   In houses having dirt-floored crawl spaces (including dirt
    floors covered with gravel or a vapor barrier),  or  in
    basement houses where a portion of  the basement has a
    dirt floor. SMD has consistently been found to be the
    most effective approach for reducing radon concentra-
    tions  in houses with  dirt-floored crawl spaces (Fi90,
    Py90), although not necessarily the least expensive;  it
    should always be  one of the first approaches considered
    for such houses. Crawl spaces having  concrete slab floors
    would generally be treated using SSD. Crawl  spaces
    having concrete wash floors (i.e., floors consisting of an
    unfinished layer of concrete about 2 in. thick or less) may
    be treated using SMD if the concrete wash is too badly
    cracked to enable  SSD.

 •   In crawl-space houses where the  crawl space is reason-
    ably accessible, i.e., has adequate headroom, a reason-
                                                     23

-------
ably level floor, and limited obstructions (such as furnace
and water heater, interior load-bearing walls and support
piers, storage, etc.).  It  has been found that complete
coverage of the crawl-space floor with the plastic mem-
brane, and sealing the membrane at all junctions, are
often  not necessary. Thus, inability to access some lim-
ited portion of a crawl space due to one or more of the
above problems does not necessarily render SMD inap-
plicable. However, SMD  will provide its best perfor-
mance when the crawl space is entirely or largely acces-
sible.  SMD will be inapplicable in crawl spaces which are
largely or  completely inaccessible (e.g., "suspended
floors," which sometimes provide no more than 12 in. of
headroom anywhere  in  the "crawl  space," and which
sometimes have no access to the "crawl space").

In crawl-space houses where natural crawl-space ventila-
tion (opening foundation vents) is not an option for radon
reduction, i.e., in houses  where radon reductions of more
than about 50% are required in the living area, or where
cold winters and the presence of water pipes in the crawl
space complicate or discourage the use of natural ventila-
tion.

Available data (Na85, Tu87, Fi89, Py90, Py91, Fis92)
suggest that the living-area reductions achievable by open-
ing foundation vents are variable, and typically no greater
than about  50%. Thus, where greater reductions are re-
quired, natural ventilation is not a reliable option. Even
where reductions  no greater  than 50%  are  needed to
achieve, for example, 4  pCi/L in the living area, SMD
could still be a desirable option, since it could provide
even greater reductions and thus reduce living area con-
centrations to even lower values.

Also,  in cold climates, open vents could result in water
pipe freezing in the crawl  space, and cold floors in the
living area above. While the pipes can be insulated, these
difficulties  might discourage some homeowners from
using  natural  ventilation  on a year-around basis.

In crawl-space houses where crawl-space depressuriza-
tion is less  likely to be a viable radon mitigation option,
i.e., in houses where the crawl space is not well-isolated
from  the outdoors and/or from the living area, or where
there  are combustion appliances (such  as a gas-fired
furnace) in the crawl space.

Crawl-space depressurization has consistently been found
to be  second  only to SMD in effectiveness for reducing
radon levels  in crawl-space houses (Fi89, Fi90, Py90,
Py91, Tu91c). It is less  expensive than SMD to install,
but it  is commonly more expensive to operate because it
generally draws a greater amount of treated house air out
of the living area.

Crawl-space depressurization  will be most effective in
cases  where the crawl space is well isolated, i.e., where
the perimeter walls are  relatively tight, and where the
flooring between the crawl space and the overhead living
area (or the wall between the crawl space and any adjoin-
ing basement) are relatively tight. Where the crawl space
    is not well isolated, there is an increased chance that a)
    crawl-space depressurization will not be able to effec-
    tively depressurize the crawl space, and will thus not be
    able to provide effective radon reductions at reasonable
    exhaust fan capacity; b) the amount of treated house air
    drawn down  from the living  area will  be high,  thus
    increasing the heating and cooling penalty  associated
    with crawl-space depressurization; and/or c) the installa-
    tion cost of the crawl-space depressurization system will
    be increased, due to  the effort required to provide the
    needed isolation. Poor crawl-space isolation thus would
    tend to make crawl-space depressurization less competi-
    tive with SMD.  Features that  tend to make the crawl
    space less tight include forced-air heating ducts penetrat-
    ing the flooring between the crawl space and the living
    area, other unclosed openings through the flooring, ab-
    sence of a wall separating the crawl space from an
    adjoining basement, and the presence of foundation vents.

    However, even where the crawl space is relatively tight,
    SMD  will commonly provide greater radon reductions at
    lower exhaust flows (i.e., lower heating/cooling penalty).
    Thus, SMD should always be considered, even where the
    house may be amenable to crawl-space depressurization.

    The presence  of combustion appliances  in  the crawl
    space could make crawl-space depressurization inappli-
    cable, due to concerns that that approach could cause
    back-drafting of the appliances.

 •  In crawl-space houses having either  hollow-block or
    poured concrete foundation walls. In the relatively  lim-
    ited testing to date in crawl-space houses,  there has been
    no  clear  indication that  possible radon entry into the
    crawl space or living area via the void network inside
    block walls presents a problem that cannot be treated by
    SMD. To be  safe, when  treating crawl-space houses
    having block foundations, some mitigators design the
    SMD  system  in  a manner to increase its likelihood of
    treating the wall cavities.

2.3    Performance of Active  Soil
Depressurization Systems
This section reviews the experience to date with each of the
ASD variations, to suggest the performance  that might be
expected with these techniques under various conditions.
Available  results are summarized from research, develop-
ment, and demonstration projects, and from the experience of
commercial radon mitigators.

The subsequent discussions for each ASD variation will ad-
dress the current understanding of the effects of the following
classes of variables on system performance:

a)  House design variables, including, for example, sub-slab
    communication, forced-air ducts and other obstructions
    under the slab, the nature of any drain tile loop, founda-
    tion wall material of construction, house size, presence of
    an  adjoining  wing on the house, presence  of  major/
    inaccessible radon entry routes, and crawl-space features,
    such as accessibility and nature of the floor.
                                                    24

-------
b)  House operating variables, including, for example, the
    characteristics of the heating/ventilating/air conditioning
    (HVAC) system, and the operation of any exhaust fans.

c)  Mitigation system design variables,  including, for ex-
    ample, the number, nature, and location of SSD suction
    pipes, the nature of any suction pit excavated under the
    slab where the SSD pipes penetrate,  the degree of slab
    sealing carried out in conjunction with ASD installation,
    configuration of the SMD system in the crawl space, etc.

d)  Mitigation system operating variables, including, for ex-
    ample, the capacity of the suction fan.

e)  Geology and climate variables that could influence sys-
    tem operation and performance, e.g., the source strength
    (or the concentration of radon in the soil gas), the perme-
    ability of the underlying native soil, and weather condi-
    tions that could influence radon entry (temperature, winds,
    precipitation).

Much of the information summarized here has become avail-
able since  the second edition of this document was published
in January 1988 (see Section 5 in Reference EPA88a). While
EPA88a accurately indicates the general performance of (and
confidence in) these systems, there is now a broader and more
extensive  understanding of the conditions which influence
system performance.

2.3.1  Active Sub-Slab Depressurization

Active SSD  has  been one of the most widely used radon
reduction techniques. While there is not an accurate count of
the number of SSD installations that have been made over the
past six years, the number appears to be greater than ten
thousand (Ho91). Such systems have been installed through-
out the United States, and in some other countries.

SSD systems have consistently provided radon reductions of
80 to  99%, except in cases where sub-slab communication
was extremely poor, where well water  was a significant
contributor to airborne radon concentrations, or where certain
design problems arose. In some cases,  a BWD or SMD
component had to be added to the SSD  system to achieve
these high reductions. The highest percentage reductions are
obtained when the pre-mitigation level is highly elevated.
Because the system  can never reduce indoor concentrations
below outdoor levels (which average roughly 0.4 pCi/L around
the country),  a SSD system could not achieve percentage
reductions as great as 90% when pre-mitigation levels are in
the range of 4 pCi/L.

The EPA research, development and demonstration program
has tested SSD (sometimes in combination with other ASD
variations) in a total of 85 basement houses, representing a
range of house design and geology/climate variables (He87,
Mi87, Sc88, Fi89, He89, Ni89, Tu89, Du90, Gi90, Mes90a,
Py90, Du91).  These 85 houses  had  pre-mitigation values
ranging from a low of about 10 pCi/L, to a high of over 1,000
pCi/L. Over 80% of these houses were reduced to  post-
mitigation values below 4  pCi/L  in the basement (and typi-
cally even lower values upstairs); over 50% were reduced to 2
pCi/L and less, and about 25% were reduced to 1 pCi/L and
less. Those houses still above 4 pCi/L remain elevated due to
one or more of the following reasons: extremely poor sub-slab
communication; re-entrainment of SSD exhaust back into the
house, in some cases where the radon concentrations in the
exhausts were dramatically elevated (Mi87, Fi91, He91d);
and contributions to the airborne levels from radon in the well
water (Ni89, Fi91, He91d). The re-entrainment problem can
be addressed through additional  care in the design of the
system exhaust; the problem with the well water would re-
quire that a water treatment step be installed. The problem of
very poor communication may be more difficult to address,
with possible solutions including the addition of more suction
pipes, increasing the fan capacity or making other changes in
system design, adding, for example, a BWD component, or
attempting another mitigation approach, such as  basement
pressurization.

The percentage of these basement houses that were reduced to
2 pCi/L and less could presumably also be increased by
further improvements on the current SSD installations. At the
time that many of these installations were made early in the
program, efforts commonly stopped once the ability to achieve
EPA's initial guideline of 4 pCi/L was demonstrated.

The post-mitigation  measurements reported for the 85 base-
ment houses represent a range of measurement methods and
durations, ranging from several-day measurements with con-
tinuous monitors and charcoal detectors, to 3- to  12-month
alpha-track detector measurements. Thus, some of these mea-
surements better represent the long-term performance of these
SSD systems than do others. However, it is clear that, overall,
the SSD systems are being very effective.

The EPA program has tested SSD in 40 slab-on-grade houses,
representing a range of house and geology/climate variables
(Fi89,  Fo89, Fi90,  Gi90, Mes90a,  Roe91,  Tu91b,  Tu91c,
Fo92). These 40 houses fall into two categories: 11 represent
the mid-Atlantic coast and the Midwest (Fi89, Fi90, Gi90,
Mes90a), and are characterized by a good layer of aggregate
under the slab;  the remaining 29 represent Florida and the
Southwest (Fo89, Roe91, Tu91b, Tu91c, Fo92), and are char-
acterized by low-permeability packed sand (and/or clay) un-
der the slab, generally with no aggregate. The pre-mitigation
indoor levels ranged between 7 and 30 pCi/L at all locations,
except for nine of the Florida houses (Fo89, Roe91, Fo92),
where some individual pre-mitigation readings were as high
as 40 to 100 pCi/L.

Of the EPA slab-on-grade study houses having good aggre-
gate, one- or two-pipe SSD systems were sufficient to reduce
all 11 of them below 4 pCi/L, 10 of them to 2 pCi/L and less,
and 7 of them (almost two-thirds) to 1 pCi/L and less. These
reductions were achieved despite the presence of sub-slab
forced-air heating supply ducts,  disrupting the communica-
tion beneath the slabs in a number of the houses. However, of
the 29 houses in Florida and the Southwest with poor sub-slab
communication, only 19 of these houses (about two-thirds)
have to date been reduced to 4 pCi/L and less, despite the use
of multiple suction pipes (as many as nine pipes in one Florida
house). These 19 houses reduced to 4 pCi/L and less represent
5 of the 6 of New Mexico houses, and 14 of the 23 Florida
                                                       25

-------
houses. Seven of the 29 houses (about one-quarter) have been
reduced to 2 pCi/L and less, and  only two (both  in New
Mexico) have been reduced to 1 pCi/L and less, based upon
short-term measurements. The apparent conclusion is that,
where a good aggregate layer exists, relatively simple SSD
systems can provide excellent reductions  in slab-on-grade
houses, despite sub-slab obstructions. But where the material
directly under the slab is a low-permeability packed sand/
clay, more extensive SSD systems will be needed, and re-
sidual  (post-mitigation) radon levels will tend to be higher.
The difficulty experienced in achieving residual levels below
2 to 4 pCi/L in the Florida houses could also be in part due to
the higher pre-mitigation values in some of these houses.

This experience with SSD  in basement and slab-on-grade
houses under the EPA R.D&D program is  also generally
reflected in the experience  of other researchers (Er84) and
commercial mitigators. Respondents to EPA's mitigator sur-
vey (Ho91) indicated that SSD (or, where appropriate, DTD),
sometimes in conjunction with sealing, is the technique used
in about 90% of basement and slab-on-grade houses. In aNew
Jersey survey (DeP91), 75 to 85% out of almost 1,2QO com-
mercial installations  surveyed in that state were found to
involve SSD. In addition, discussions with a number of miti-
gators  who have  installed 250 mitigation systems or more
apiece, confirm that SSD is very effective over a broad range
of applications (Bar90, Bro90, Mes90b, Sh90, St90, We90).
According to these mitigators, one- or two-pipe SSD systems
are sufficient in a  large majority of the houses mitigated. For
some more difficult basement and slab-on-grade houses (as
few as 10 to 20% of the total in some geographical locations),
three, four, or more suction pipes are  required in order for
SSD to be effective, or a BWD component is required. In only
about  1% of the  houses treated by these mitigators were
conditions so unfavorable that SSD was practically not appli-
cable; these cases usually involved extremely poor sub-slab
communication  (such as wet clay under the slab), high  de-
grees of interior finish, irregular house configurations, and
additions to the house.

The following discussion summarizes the results to  date, as
they address each  of the house, mitigation system, and geol-
ogy/climate variables that can influence system design, opera-
tion, and performance.

a)  House design variables

 •  Sub-slab communication. Numerous studies in both
    basement and slab-on-grade houses having good commu-
    nication (Tu89, Du90, Fi89, Gi90, Mes90a, Fi91), as well
    as commercial experience, have consistently demonstrated
    that such houses can be reduced below 4 pCi/L, and
    generally below 2 pCi/L, with  only one or two suction
    pipes. "Good  communication" generally coincides with a
    layer of aggregate beneath the slab. (See definition of
    "aggregate" in the Glossary.) Where communication is
    poor (Fo89, Tu89, Py90, Du91, Fi91, Roe91, Tu91b,
    Fo92), reductions below 4 pCi/L are still commonly
    achieved, but generally require more suction pipes and
    more careful  pipe placement to ensure that an adequate
    suction field is established everywhere  beneath the slab.
Ability to reliably achieve 2 pCi/L and less is not well
demonstrated in poor-communication houses.

Nature of foundation wall. In a majority of basement
houses where sub-slab communication is good, SSD per-
formance appears generally similar regardless of whether
the foundation wall is constructed of poured concrete or
of hollow block (He87, Gi90). Often, any reduced perfor-
mance observed in  block-foundation basement houses
relative to poured-foundation houses appears small, per-
haps even within the uncertainty  of the radon measure-
ments (He87). However, there will be some percentage of
cases where SSD clearly is not adequately treating  the
entry routes associated with block walls, and a BWD
component will be required, or will be more  practical
than attempting to address the wall entry routes through
more SSD pipes (He87, Sc88, M89, Tu89, Py90, Sh90).
The potential need to add a BWD component appears to
be greatest when sub-slab communication is poor, hin-
dering interception of soil gas by the SSD system before
it enters the void network. In infrequent cases, presum-
ably where  the walls are particularly important radon
sources,  wall treatment  has sometimes been  required
even when communication has been reasonably good,
and when the SSD system has appeared to be effectively
depressurizing the entire sub-slab  (Section 7.3 in Fi91).

An attempt was made to study the effect of the foundation
wall on  SSD performance  in  slab-on-grade houses in
Ohio having good sub-slab aggregate (Fi90, He91a). As
it turned out, any effect of wall  material  could not be
separated out from the effect of house size, because all of
the large houses (generally greater than 1,700 ft2) had
block foundations, and all of the small ones (less than
1,400 ft2) had poured concrete. All of the smaller houses
with poured foundations  were reduced from pre-mitiga-
tion levels of 13 to 30 pCi/L down to 1 pCi/L and less
with one suction pipe. However, the larger, block-foun-
dation houses were  usually reduced  only to within  the
range of 1  to 3 pCi/L,  and sometimes required two
suction pipes to reach that range. The one small (1,100
ft2) house with a block foundation  did perform  more
poorly than the small poured-foundation houses, achiev-
ing only 1.5 to 2.5 pCi/L with one suction pipe. One slab-
on-grade house with good aggregate tested in Maryland
(Mes90a) had a block foundation and was larger than any
of the Ohio houses, yet was reduced from pre-mitigation
levels of 7 to 16 pCi/L down to below 1 pCi/L with only
one SSD pipe, comparable to the Ohio poured-foundation
houses. This Maryland house had styrofoam insulation
board on the interior face of the foundation wall below
the slab, possibly reducing the short-circuiting of outdoor
air into the system through the wall and thus contributing
to the good performance. (In all of the houses, the slab
was elevated above grade by 6 to 12 inches in at least
some locations around the perimeter, creating the possi-
bility that outdoor air could leak into the system through
the walls.)

From the above data, it appears that air leakage into  the
system via the block walls might  be contributing (along
with house size) to reduced SSD performance in slab-on-
                                                       26

-------
grade houses having good aggregate. This effect could be
reduced when there is insulation board inside the block
foundation. Any detrimental effect of the block walls
does not appear adequate to prevent achieving  levels
below 4 pCi/L with one to two suction pipes in slab-on-
grade houses with good aggregate, even  when sub-slab
forced-air supply ducts are present, and even when the
suction pipes are placed near the perimeter. However,
wall effects could increase the number of pipes required
to achieve near-ambient indoor levels in block-founda-
tion houses.  The relatively  small effect of the  block
foundation in the good-aggregate case might be attributed
to adequate extension of the suction field beneath the
slab, through the aggregate, despite any leakage of out-
door air that might be occurring through  the foundation
walls.

But in slab-on-grade houses in  Florida, where sub-slab
communication  is often very poor, block walls appear to
have a more significant impact on SSD performance. In
one house (House C4) where SSD suction pipes were
tested both toward the slab interior  and near the slab
perimeter, pre-mitigation indoor levels of 38 to 67 pCi/L
were reduced to 5.5 pCi/L by the operation of two suction
pipes toward the interior of the slab, but only to 9.4 pCi/L
by two pipes near the perimeter (according to 3- to 4-day
continuous radon monitor measurements) (Fo92). The
poorer performance with the perimeter pipe locations is
presumably due to outdoor air  short-circuiting into the
system via the walls, which commonly extend perhaps 12
in. above grade. In a second house (House Cl), originally
at 40-70 pCi/L, four interior pipes  generally  reduced
indoor levels to 2.2-3.4 pCi/L, whereas five perimeter
pipes achieved a level of only 4.5 pCi/L. In a third house
(C5),  originally at about 20 pCi/L, two  interior pipes
reduced levels to 8.0 pCi/L, whereas two perimeter pipes
achieved only 11.0 pCi/L.  This  wall effect is confirmed
by measurements in test slabs and by computer modelling
(Fu91). Where  communication beneath the slab is very
poor, the impact of air leakage  into the system through
block walls might be expected to have a more visible
effect.

Not all slab-on-grade houses having very poor communi-
cation and block  foundations have exhibited the same
difficulties encountered in the Florida houses. Three such
houses were tested in New Mexico  (Houses AL02, AL03,
and AL04 in Tu91b). Two of these  houses have been
reduced below 4 pCi/L with one to three suction  pipes,
and one has been reduced below 2 pCi/L, despite  place-
ment of the pipes near the perimeters. This difference in
achievable radon  levels between  the Florida and New
Mexico  houses  may be due to better suction field  exten-
sion around the  foundation or through the underlying soil
in New Mexico  resulting from the geology or dry climate
in the Southwest,  or resulting  from  local construction
practices (e.g.,  the presence of insulation board around
the foundation in some cases). It could also result, in part,
from the fact that the pre-mitigation levels in New Mexico
were lower, 7 to  18 pCi/L in these three houses, com-
pared to levels  up to 103 pCi/L in a few of the Florida
houses.
With  either basement or slab-on-grade houses  having
block foundation walls,  another  wall-related variable
which could influence the performance of SSD alone is
whether the voids in the top course of block were  capped
during construction. If the top voids are capped, there is
an increased likelihood that better  reductions will be
achieved with fewer SSD pipes, and that a BWD compo-
nent will not be needed in the ASD system. In basement
houses, the closure of the top voids during construction
would be apparent from the presence of a solid cap block
as the top course. In slab-on-grade houses,  the  closure
would take the form of either a) a solid top block (which
could be an L-block), with the slab poured inside the stem
wall (or to the notch in the L); or b)  the slab poured on top
of the stem wall, closing the top cavity. The nature of
block stem wall in a slab-on-grade house can be difficult
to determine.

House size. Where there is a  good layer  of sub-slab
aggregate  and where there are no  complicating factors,
one or two SSD pipes have sometimes treated residential
basement slabs and slabs on grade as large  as 1,850 to
2,700 ft2, reducing indoor levels  to 1  to 2 pCi/L and less
(Fi89, Fi90,  Mes90a). Such reductions  in  such large
nouses have sometimes been observed even where some
complicating factor is present (e.g.,  sub-slab forced-air
supply ducts, very high pre-mitigation concentrations, or
block foundation walls, as discussed above). Where such
complicating factors are present, one or two pipes will
commonly be sufficient to reduce large houses to 3 to 4
pCi/L and less, if not below 1 to 2 pCi/L (Fi90, Mes90a),
when there is good aggregate. In such cases, house size is
not a significant variable.

Underscoring this  conclusion from  residential  testing,
tests on large slab-on-grade schools and commercial build-
ings have demonstrated  that where communication is
good  and  there are not sub-slab obstructions, a single
suction pipe (and a properly sized fan) can be sufficient to
treat slab areas as great as 50,000 ft2 (Cr92b). No house
would be anywhere near this size.

But where sub-slab communication is not good, large
house slabs will require a greater number of suction pipes
located properly in order to provide sufficient  suction
field extension.

Adjoining wings. In basement houses having an adjoin-
ing slab-on-grade or crawl-space  wing, a SSD  system
treating the basement slab alone, without direct treatment
of the adjoining wing, is sometimes sufficient to provide
the needed radon reductions throughout the entire house
(Sc88, Fi89, Mes90a). The adjoining wing is most likely
to also require direct treatment when a) the adjoining
wing provides important entry routes, and has distinctly
elevated soil gas radon concentrations under its founda-
tion; and b) communication beneath the basement slab is
poor,  hindering the extension of  the basement  suction
field under the adjoining  wing,  and  hindering intercep-
tion of soil gas before it reaches the foundation of the
adjoining wing (Mes90a).
                                                    27

-------
Sub-slab obstructions. Obstructions can be present under
the slabs in basement and slab-on-grade houses which
could potentially interrupt suction field extension even in
cases where aggregate is present. Such obstructions  in-
clude footings  supporting interior load-bearing  walls,
grade beams, sub-slab forced-air supply or return ducts,
and sunken living rooms in slab-on-grade houses.

Testing in seven slab-on-grade houses  in  Ohio having
sub-slab forced-air supply ducts has indicated that, where
a good layer of aggregate exists, SSD performance is not
dramatically reduced by such ducts, even though sub-slab
depressurization measurements suggest that the ducts are
interrupting  suction field extension (Fi90,  He91a). The
ducts did appear to increase the likelihood that two suc-
tion pipes would be required to reduce the house below 4
pCi/L, and that the fan would have to be operated at full
capacity. However,  even the largest house in the Ohio
project (2,600 ft2 slab) was reduced below 2 pCi/L with
two suction pipes (89% reduction), despite sub-slab ducts.
In one large (2,700 ft2) slab-on-grade house in Maryland
having sub-slab supply ducts and good aggregate, levels
were reduced below 1 pCi/L (over 90% reduction) with a
single suction pipe (Mes90a).

Even in one 1,900 ft2 slab-on-grade house (House AL04)
in New Mexico having no aggregate (slab resting directly
on dry sand/clay soil), the house has been reduced to 1 to
2 pCi/L (a reduction of about 85%) with only three SSD
pipes despite the presence  of sub-slab supply ducts
(Tu91b). In one other house with sub-slab supply ducts
and no aggregate (House AL02, having a slab of 2,000
ft1),  levels were reduced below 4 pCi/L with only one
pipe; however, these post-mitigation levels tend to spike
with decreases in barometric pressure, for reasons that
may or may not be associated with the sub-slab ducts.

More data  would be required in poor-communication
houses before a conclusion could be drawn regarding
how consistently the degree of success seen with sub-slab
ducts in the New Mexico project might be expected in
other such  houses.  In addition, no data exist  on SSD
performance  in  houses  having forced-air  return ducts
under the slab (with or without aggregate). Since return
ducts operate at low pressure, they would be competing
with the SSD system for the soil gas, and would provide
SSD systems with a greater challenge than do supply
ducts.

There is not a definitive data base available on the effects
of the other types of obstructions under house slabs. In
the testing of SSD systems in large slab-on-grade school
buildings, which are more  likely than  houses to have
interior footings and grade beams, experience has been
that, even when aggregate is present, slab treatment by a
given SSD  pipe will not reliably extend past an interior
footing supporting an interior block load-bearing wall
which completely bisects the slab (Cr91).  However, in
testing in houses where such  an interior footing com-
monly  appears in the  form of a stem wall separating a
basement from an adjoining slab-on-grade wing, suction
on the  basement slab alone has been found adequate to
    treat the adjoining slab on grade when there is good
    aggregate, as discussed previously (Mes90a). This effect
    might sometimes be due to interception of soil gas before
    it reaches the slab-on-grade foundation, and not always to
    a measurable extension of the basement suction field into
    the aggregate beneath the slab on grade. The permeability
    of the underlying native soil could also play a role. Where
    aggregate is present,  grade beams (thickened slabs) are
    likely to be poured on top of the aggregate, so that suction
    fields have been observed to extend past grade beams
    (Cr91). The effect of interior footings and grade beams
    on SSD systems will likely depend upon the permeability
    of the underlying soil, to permit the suction to extend
    under the interior obstruction.

    Where sub-slab  communication is  poor, it would be
    anticipated that interior obstructions would generally cre-
    ate a more serious problem  than they do in the good-
    communication case.

b)  House operating variables

 •  Operation  of central furnace fan, and  exhaust
    fans. Operation of a central furnace fan in a basement
    house, when the cold air return ductwork is in  the base-
    ment, has been found to cause incremental increases in
    basement depressurization typically ranging between 0.002
    and  0.02 in. WG (Mi87, Ha89, Hu89, Ma89b, Tu89,
    Tu90). Similar increases in depressurization have been
    observed in rooms within slab-on-grade houses, when
    cold air returns are in those rooms (Cu92). This depres-
    surization results because the low-pressure cold air return
    ducting is typically very leaky, so that the central fan is
    withdrawing significant amounts of air from the base-
    ment and distributing most of it to other locations in the
    house.  Various exhaust fans in basement houses have
    been found to be creating increases in basement depres-
    surization ranging from 0.001 to 0.02 in. WG for clothes
    driers (Mi87, Ma89b, Tu90), 0.004 to 0.008 in. WG for a
    whole-house fan (Tu90), and 0.008 to 0.02 in. WG for an
    attic fan (Tu90). Again, similar increases in depressuriza-
    tion  have been observed  in the living area of slab-on-
    grade houses from operation of such exhaust fans (Cu92).

    SSD systems can be designed to develop sufficient sub-
    slab  depressurizations during cold weather with the ex-
    haust appliances off,  so that the system will nominally
    not be overwhelmed by these additional basement de-
    pressurizations when the  appliances  come on. Where a
    good layer of aggregate exists, sub-slab depressurizations
    sometimes well in excess of 0.01 to 0.02 in. WG can be
    maintained under most or all of the  slab by one or two
    suction pipes, during cold weather with the appliances off
    (e.g., Gi90, Mes90a, Fi91). Where there is not an aggre-
    gate layer, more suction pipes  would be required to
    maintain sub-slab depressurizations  that great,  and it is
    more important to locate the pipes near the entry routes
    (especially  the wall/floor joint) to help ensure that the
    depressurization extends to those entry routes.

    Depending upon site-specific factors, there may not nec-
    essarily be a significant  impact on  long-term average
                                                    28

-------
    indoor concentrations if the pressure differential across
    some  portion of the slab  is occasionally reversed by
    operation of these exhaust fans. Moreover,  since SSD
    seems to work by mechanisms in addition to soil depres-
    surization (in particular, by soil gas dilution), it may in
    fact not be necessary to guarantee that the sub-slab de-
    .pressurizations being established by the system are greater
    at every sub-slab location than every potential basement
    depressurization that the system may ever encounter.
    However, where the SSD system can reasonably be de-
    signed to provide sub-slab depressurizations of about
    0.01 to 0.02 in. WG everywhere during cold weather with
    the appliances off, in order to ensure that the system will
    essentially never be overwhelmed, it is advisable to do so.

    Where slab pressure measurements are made during mild
    weather, the  (conservative) target sub-slab depressuriza-
    tion with exhaust appliances off would be increased to
    0.025  to 0.035 in. WG, to include the further basement
    depressurization caused by the thermal stack effect dur-
    ing cold weather. (This conservative maximum basement
    depressurization is thought to be high; combustion appli-
    ances  would back-draft if such basement depressuriza-
    tions were maintained for an extended period.) The stack
    effect is discussed further in 2.3. le (Climatic conditions).

    Where forced-air supply ducts are located under the slab,
    then the SSD system faces an even greater challenge. Not
    only must the system overcome any depressurization of
    the  house caused by leaky cold air returns, but it must
    also overcome pressurization of the sub-slab region caused
    by the supply ducts. Testing in a total of eight slab-on-
    grade houses having sub-slab supply ducts (Fi89, Fi90,
    Mes90a) has demonstrated that, where good aggregate is
    present, the  SSD system can generally maintain good
    radon reductions despite operation of a central furnace
    system having such ducts. There are no data available on
    the effects of f orced-air return ducts under the slab, which
    would present the SSD system with an entirely different
    challenge than do supply ducts.

c)  Mitigation system design variables

 •  Number of suction pipes. Where there has been a good
    aggregate layer beneath the slab, radon reductions to 2
    pCi/L and less have commonly been achieved in both
    basement and slab-on-grade houses using only one or two
    suction pipes (Tu89, Bro90, Du90, Fi89, Gi90, Mes90a,
    Mes90b, We90, Fi91). Residential slab areas as great as
    1,850 ft2 (Fi89) to 2,700 ft2 (Mes90a) have been treated
    by a single suction pipe under favorable conditions, as
    have school slabs as large as 50,000 ft2 (Cr92b).

    But when communication is marginal or poor, more pipes
    will be required to maintain adequate sub-slab depressur-
    ization at the major entry routes. Mitigators report two to
    four suction pipes commonly being required in "typical"
    marginal-communication houses (Bro90, Mes90b), about
    one pipe per 350 to 750 ft2.  However, depending upon
    how poor the communication is, how widely the entry
    routes are distributed, and how large the house is, even
    more  pipes can be required. Three to five pipes (corre-
sponding to one pipe per 350 to 750 ft2) have proven
inadequate to reliably achieve less than 4 pCi/L in some
Florida slab-on-grade houses having poor sub-slab com-
munication (Fo92). Some commercial installations have
been reported having as many as eight (St90) to eleven
(Bro90) suction pipes in worst-case basement houses
where communication was truly poor. In one large base-
ment house (2,260 ft2 slab)  with no measurable suction
field extension beneath the slab, 20 suction pipes (about
one pipe per 100 ft2) reduced 3-month winter-quarter
concentrations in the basement to 2 pCi/L (Sc88), raising
doubts regarding the viability of achieving near-ambient
radon concentrations with SSD alone in that house, at
least in the basement during cold weather.

Where communication is moderate to poor, various in-
vestigators have used pre-mitigation diagnostic tests to
aid in deciding how many suction pipes will be needed at
what locations  in order to adequately depressurize the
sub-slab (EPA88b, Fo90, Fi91, Tu91a, andTu91b, among
others). These tests commonly involve use of a vacuum
cleaner to determine  the ease of suction field extension
under the slab. A large number of mitigators report that
the vacuum cleaner testing, although useful as a qualita-
tive indicator of good vs. poor communication, generally
appears to over-predict the number of SSD pipes actually
required to achieve the required radon reductions when
used for quantitative design (Fo90, Fi91, Si91, Sau92).
That is, the sub-slab suction field generated by the vacuum
cleaner is more limited than it should be, suggesting that
more SSD pipes will be needed to provide the desired
depressurization than are in  fact required to adequately
reduce indoor radon  levels. Perhaps in some cases, the
disagreement between the vacuum cleaner and the SSD
results occurs because the vacuum cleaner diagnostic test
has not been conducted properly—e.g., insufficient time
has been allowed for the vacuum cleaner to establish its
suction field at remote test  holes (Hi92). (See Section
3.3.2.) This result may also  be indicating that the SSD
system is working by mechanisms in addition to depres-
surization—e.g., by soil gas dilution, which the vacuum
cleaner flows are sometimes too low to reproduce.

Location of suction pipes. Where there is a good layer
of aggregate, the one or two suction pipes can usually be
located just about anywhere, as necessary to avoid fin-
ished areas and to accommodate the homeowners' living
patterns. If there is no constraint regarding pipe location,
placement of the pipes near what would appear to be the
more important potential entry routes would intuitively
seem to be advisable (e.g., toward the fully below-grade
block foundation wall in a walk-out basement); but the
pipes should never be placed too close to a major slab
opening,  such as an  unsealed perimeter channel drain,
which would permit excessive air short-circuiting into the
system. If there are constraints on pipe location, this is
generally not a problem when sub-slab communication is
good; there are numerous instances  where  large slabs
have been very effectively treated (reducing indoor levels
to 1 to 2 pCi/L) with a single pipe at one end. Examples
include House 21 in Reference Sc88, House 43 in Refer-
ence Fi89, and House 488 in Reference Mes90a.
                                                       29

-------
Where the sub-slab region can conveniently be accessed
from outdoors, as is often the case in slab-on-grade
houses, pipes penetrating through the foundation wall
from outdoors appear to  give reductions comparable to
those achieved when the pipes penetrate down through
theslab from indoors when communication is good (Fi90).
Such exterior penetrations have even been successfully
used in basement houses, in cases where the basement is
so heavily finished that interior pipes are not practical
(K189).

When communication is moderate to poor, more care is
required in locating the multiple suction pipes near the
major entry routes, since the suction field will not extend
so far from the pipes. General experience, together with
measurements of the "radon entry potential" around slabs
(Tu91a, Tu91b), have indicated that the wall/floor joint,
and block foundation walls where present, are  consis-
tently among the major entry routes. Interior slab loca-
tions are usually major routes only where there is some
major slab opening at an interior site (e.g., a cold joint, or
an interior block foundation wall or fireplace structure
which penetrates the slab). Consequently, suction pipes
are commonly placed near the perimeter walls when
communication is poor (e.g., Sc88, Tu91b),  with pipes
placed at interior locations primarily when there is an
interior entry route. Again, a suction pipe should not be
placed too close to a major unsealed slab opening, to
avoid excessive short-circuiting of house air into the
system.

In addition to having  the pipes near the major entry
routes, location of the pipes around the perimeter in poor-
communication cases also takes advantage of the fact that
sub-slab communication  will generally  be best in the
vicinity of the footings. The region around the footings
had to be excavated and backfilled during construction;
thus, the soil will tend to be looser and more permeable at
that location, and some soil subsidence may even have
taken place there  after the slab was poured, possibly
creating an air gap under  the slab.

The one case where perimeter placement of suction pipes
appears to be undesirable in poor-communication houses
has been in slab-on-grade houses with block foundations
and shallow footings in  Florida (Fo90,  Fo92),  as dis-
cussed above under Nature of foundation wall. A similar
result has been reported in Arizona and southern Califor-
nia (K192). Because of the extremely poor sub-slab com-
munication in the  Florida houses, location of the SSD
pipes near the perimeter probably is resulting in  enough
outdoor air short-circuiting into the SSD system through
the above-grade block foundation wall and perhaps  through
the soil under the  footings, to unacceptably  reduce the
already-weak suction field extension through the sub-
slab  fill material. Location of suction pipes near perim-
eter  block walls in slab-on-grade houses has not been
found to be a significant problem at other sites, including
Ohio (Fi90) and Maryland (Mes90a), where aggregate is
present under the slabs. Nor did it appear to be a major
problem in New Mexico slab-on-grade houses,  despite
the fact that the communication under the New Mexico
slabs appears as poor as that under the Florida slabs.

Size of suction pipes.  The suction loss in the system
piping is determined by the gas velocity in the pipe (and
hence the pipe diameter), as  well as by the length of
piping and the number of elbows and other obstructions
in the piping. For a given piping configuration, the pipe
diameter necessary to adequately reduce suction loss can
be calculated based upon the capabilities of the SSD fan
being used and upon the suction that must be maintained
beneath the  slab. Four-inch diameter pipe has been the
size  most commonly  used, since 4-in. PVC piping is
readily available, and provides reasonably low  suction
loss over the range of flows typical of most SSD installa-
tions (commonly 25 to 100 cfm in houses having good
communication). Three-in. diameter piping can  usually
be considered where desirable for aesthetic reasons (e.g.,
to permit concealment within stud walls), where flows
are sufficiently low, and/or the piping run is sufficiently
short, and/or fan capacity is adequate.  When very  low
flows (e.g.,  10 cfm) are  expected in a particular  suction
pipe in poor-communication houses, 2-in. piping can be
considered for that particular pipe.

Another consideration in selecting pipe  size is to reduce
flow noise in the piping. As  flo
-------
The benefits of increasing the pit size in poor-communi-
cation houses will depend upon the nature of the sub-slab
fill material and of the underlying soil (i.e., the presence
of fractures and strata in the underlying material). In one
study (Ma89a), increasing pit depth from zero to 4 ft
increased the effective distance over which the measur-
able suction field extended, by factors ranging up to 10 in
four poor-communication basement houses in Tennessee.
These increases resulted from reduced acceleration pres-
sure losses, as predicted by  a mathematical model  of
suction field extension,  and also from  increases some-
times observed in the effective permeability of the sub-
slab soil/fill, presumably achieved through intersection of
fractures and strata.

In another study (Fo89, Fo92), in four poor-communica-
tion slab-on-grade houses in Florida, tests were  made
with three different pit configurations (no pit, pit 10 in.
square by 12 in. deep, pit 15 in. square by 19 in. deep).
The results showed that the pits could increase measur-
able sub-slab depressurizations by up to 0.005 to 0.02 in.
WG at test holes 15 to 20 ft from the SSD suction pipe.
However, there was not a clear relationship between pit
size and the resulting increase in depressurization.

In a third  study (Py90), the effects of a range of pit
dimensions were tested in four basement houses in Ten-
nessee having poor communication. In this study, wide,
shallow pits (28 in. square by 10 in. deep) provided better
sub-slab depressurizations in two houses (by 0.002 to
0.05 in. WG) at distances of 8 to 25 ft from the suction
pipe,  than did narrow, deep pits (4  in. square by 30 in.
deep). In two other houses, pit sizes were increased from
zero (no pit) to, first, 10 in. square by 12 in. deep (10 x 12
in.); then to 20 x 16 in.; and then to 24 x 18 inches. Sub-
slab depressurizations generally increased with increas-
ing pit size, with some important differences between the
two houses. In the first of these latter two houses, which
had better communication than did the other, the depres-
surizations were increased by 0.001 to 0.064 in. WG at
distances of 9 to 38 ft from the suction pipe, and increas-
ing the pit size above 20 x 16 in. did not provide signifi-
cant improvements. But in the second house, no increase
in pit size was able to make the measurable suction field
extend beyond 8 ft from the suction pipe; within that 8-ft
distance, the depressurization continued to increase dis-
tinctly with  increasing  pit size, even  above 20 x  16
inches.

From the above results,  it is apparent that pits can defi-
nitely be helpful in extending the suction field in houses
having  poor sub-slab communication, consistent with
theory. The most cost-effective size for a pit will be site-
specific. Depending upon how low the sub-slab perme-
ability is,  and depending  upon whether fractures  are
present in the sub-slab material, pits will not always serve
to extend the measurable suction field  to distances sig-
nificantly greater  than  can be  achieved without pits.
However, pits will usually increase the measurable de-
pressurization that is achieved within this distance.
    All three of the studies discussed above focused on the
    effect of pit size on measured sub-slab depressurizations,
    not the effect of the pits on the actual radon reduction
    performances of the SSD systems. Given the previous
    observations that measurable depressurizations every-
    where under the slab are not always required for effective
    radon reductions, the effect of the pits on radon reduc-
    tions might have been better than their  effect on the
    depressurizations in some of the cases where depressur-
    izations were not significantly increased by the pits.

    In practice, most mitigators excavate a sub-slab pit of 6 to
    18 in. radius, the size that can be conveniently prepared
    by reaching through the hole that has been drilled through
    the slab  for the suction  pipe. From the  standpoint of
    reducing suction losses due to soil gas acceleration, cal-
    culations indicate that there would be no significant ben-
    efit in making the pit any larger  than 12 to 18 in. (Br92).
    The only potential benefit of a larger pit would be in-
    creased possibility of intersecting permeable fissures or
    strata.

 •  Degree of slab sealing. Sealing of the wall/floor joint
    and of other slab openings would generally be expected
    to improve SSD performance, by improving the distribu-
    tion of the suction field  beneath the slab. Slab sealing
    would also be expected to reduce the amount of house air
    withdrawn by the system, thus reducing the house heating
    and cooling penalty. A variety of generally anecdotal
    measurements  have been made  by individual investiga-
    tors regarding the effect of slab sealing on system perfor-
    mance, usually in cases where some major opening (such
    as an initially unclosed perimeter channel drain) appeared
    to be degrading the performance of an  existing SSD
    system. The effects of such sealing would be expected to
    be site-specific, depending upon such factors as the size
    of the slab opening, the sub-slab communication, and the
    proximity of the suction pipe to  the opening.

d)  Mitigation system operating variables

 •  Fan capacity. The centrifugal in-line tubular fans cur-
    rently being installed in many residential mitigation sys-
    tems typically range in size from approximately 50 watts
    with 4-in. diameter couplings (capable of moving  about
    125 cfm at zero static pressure) to approximately 90 watts
    with 6-in. couplings (capable of moving about 270 cfm at
    zero static pressure). Fans as small as 29 watts have been
    used by some mitigators (Str91). In general, the 90-watt
    fans will provide more effective treatment of entry routes
    in a given house, lower indoor concentrations, and better
    insurance against the system being  temporarily  over-
    whelmed by weather effects or homeowner activities,
    compared to the 50-watt fans. However, these larger fans
    create a higher system operating cost, due to increased
    power  consumption and an increased heating/cooling
    penalty resulting from withdrawal of a greater amount of
    treated air out of the house.

    Where a 90-watt in-line fan at full power is not necessary
    to achieve the desired reductions, an operating cost sav-
    ings of roughly $70 per year (He91b, He91c) might be
                                                     31

-------
achieved by using a 50-watt fan, or by operating a 90-
watt fan at reduced power. But this reduction in operating
cost will essentially always be accompanied by an in-
crease in the residual radon levels in the house, even if
levels remain  below 4 pCi/L. Therefore, it is recom-
mended that a safety factor be used in selecting the fan
size and operating  conditions, rather than  risking in-
creases in homeowner exposure by under-sizing the fan
in an effort to achieve relatively modest reductions  in
operating cost.

The annual cost savings of S70  from switching to the
smaller  fan translates into only $5  to $6  per  month
savings in electricity and heating/cooling costs, a differ-
ential which many homeowners may not be able to distin-
guish within the normal monthly variations in their utility
bills. The reductions in the capital cost of the fan would
also be only modest, with the 90-watt  fans currently
being no more than $10 more expensive than the 50-watt
fans (He91b, He91c). On a national scale, however, con-
sistent use of the larger fan will result in increased
pollutant release from electrical power plants and in-
creased consumption of natural resources (coal, oil and
gas). The trade off between reduced indoor radon levels
with the larger fan, on the one hand, vs. increased energy
costs, increased power generation requirements, and in-
creased  resource consumption on the other hand, is a
decision that each individual mitigator and homeowner
will have to make for their individual situations.

Testing in both basement and slab-on-grade houses has
shown that the 90-watt centrifugal in-line fans can some-
times  be operated at reduced power and still achieve
indoor levels of 4 pCi/L and less, if communication is
good, if the source strength is not too high, and if the
house is not too large. However, even where there is good
aggregate,  a reduction in fan power results  in some
increase in radon levels, even though levels remain below
4 pCi/L. In one basement house in New Jersey (House 3)
having good communication and a 2-pipe SSD system
(Du90), a 90-watt fan reduced basement concentrations
from 180 pCi/L to 0.8-1.0 pCi/L in two separate 6-day
measurements with the fan at full power; levels increased
to 1.8 to 2.3 pCi/L in two 6-day measurements with the
fan reduced to 75% of full power, and to 3.7 to 4.6 pCi/L
in two several-day measurements with the fan reduced to
50%. In  another good-communication basement house
having an  adjoining paved  crawl space (House 7  in
Du90), with a SSD + BWD variation involving suction
on a baseboard duct over a perimeter channel drain, post-
mitigation levels in the basement increased from 0.3 pCi/
L to 2.7 pCi/L when the 90-watt fan was reduced from
full power  to 50% of full power. (Pre-mitigation levels
were 34 pCi/L.)

In another study (Ma88) in three basement and basement-
plus-crawl-space houses having good communication and
two-pipe SSD systems (OR-15, -17, and -18), reduction
of a 90-watt fan to 25 to 50% of full flow resulted in only
limited increases, if any, in indoor radon levels.  In one
bascment-plus-slab-on-grade  house in Washington state
having good communication and a one-pipe SSD system
(House ESP113  in Pr87), living-area radon levels  in-
creased  from 4.3 to 5.0 pCi/L when fan power was
reduced sufficiently to decrease the suction in the system
piping from 1.7 to 0.8 in. WG, and the flows from 36 to
23 cfm.

The effect of reduced fan capacity was also tested in nine
slab-on-grade houses in Ohio having good sub-slab ag-
gregate (Fi90, He91a). In most houses having two SSD
pipes, the effect of fan  capacity was tested with both
pipes operating, and also with each pipe operating indi-
vidually. Where the system was  reliably treating the
house with  the fan at full power, decreasing the 90-watt
fan from full power to one-eighth of full power increased
indoor radon levels from an average of 1.1 pCi/L to an
average of 2.2 pCi/L (averaged over all houses). In three
of these cases, decreasing fan power did not increase
indoor levels. Only in cases where the SSD system was
marginal with the fan at full power, were indoor levels
increased above 4 pCi/L when fan power was reduced.

Occasionally, the 90-watt centrifugal in-line tubular fans
will not be adequate for a given SSD installation. Under
these circumstances, the question shifts from one of how
small the fan can be, to one of whether an even larger fan
(or multiple fans) might be warranted.

If the problem in a given installation is that system flows
are too high (so that the 90-watt fan cannot move enough
air to develop adequate sub-slab depressurization), then a
higher-volume fan is needed. Higher-volume centrifugal
in-line fans that have most commonly been used include
100-watt units  with either 6- or 8-in. couplings, capable
of moving  up  to 410 cfm at zero static pressure. (In-
creased volume can also be achieved by operating two
smaller fans in parallel; however, for a given total volu-
metric flow rate, a single larger fan will generally have
lower capital and operating costs than two smaller ones
having a combined volumetric flow capacity equal to that
of the one larger fan.) From a practical standpoint, SSD
systems in  residential applications  will not commonly
have this problem of the flows being too high. If flows are
so high that a 410-cfm fan appears to be needed, it is
likely that air is  short-circuiting into the system some-
where, and  the need is for some additional sealing rather
than for a larger fan.

More commonly, the problem with SSD systems where
the 90-watt fan is inadequate will be that the 90-watt fan
is drawing  very little flow. This will be the case when
sub-slab communication is poor. In poor-communication
cases, the typical 50- and 90-watt centrifugal tubular fans
operate at maximum suction and  at low  flows, at the
extreme end of their normal performance curve.  Under
these conditions, some investigators have considered us-
ing high-suction, low-flow fans  and blowers better de-
signed to operate at those conditions  (EPA88b, Fo90,
Cra91). High-suction fans and blowers can generate from
perhaps 4 to more than 40 in. WG suction, compared to 1
to 2 in. WG maximum for  the centrifugal in-line fans.
These high-suction units have the disadvantages of sig-
nificantly greater capital and operating costs; in addition,
                                                   32

-------
units not manufactured specifically for radon mitigation
are commonly much more noisy. Several high-suctions
blower currently being marketed for radon reduction are
quiet, but have a higher capital cost (more than $600,
compared to less than $100 for a 90-watt in-line fan) and
a higher power consumption (generally greater than 150
watts) (Cra91, Ra92). Another disadvantage of some of
these blowers is that since their performance curve is
designed for low-flow  operation, a relatively small in-
crease in flows (due, for example, to leaks that develop in
the system over time)  could  cause the fan to become
overwhelmed, losing its effectiveness.

High-suction regenerative blowers were tested  in two
basement houses in New York, each having poor commu-
nication and pre-mitigation concentration basement con-
centrations of 21  pCi/L (Houses OP-01 and OP-09 in
Ni89). In each house, the blower drew on four SSD pipes
around each basement  perimeter. In House OP-09, this
high-suction system reduced basement concentrations to
3.4 pCi/L. But in OP-01, the regenerative blower reduced
basement levels to  only 11 pCi/L. By  comparison,  a
typical tubular fan drawing on a single, central SSD pipe
in House OP-01 provided only slightly poorer reductions,
to 14 pCi/L. It was subsequently found that the block
foundation wall was the primary  entry route requiring
direct treatment in this  particular house, and OP-01 was
ultimately  reduced  below  4  pCi/L  with a combined
SSD+BWD system. The regenerative blower was clearly
not much more able than the in-line fan to extend  suction
to the points (perhaps outside the foundation) where soil
gas was entering the  wall void  network,  despite the
location of suction pipes immediately beside the wall.

In testing on one of the poor-communication slab-on-
grade houses in Florida (Fo89), measurements were made
comparing the sub-slab depressurizations created by the
standard 90-watt centrifugal tubular fan (<2 in. WG), and
by two different blowers capable of developing 2.3 and 6
in. WG static pressure. The results showed that the high-
suction blowers increased the sub-slab depressurization
at test holes (within about 15 ft of the suction hole) where
the in-line fan had generated measurable depressurization
to begin with. Sub-slab depressurization at these holes
was increased by an amount proportional to the increased
suction in the SSD pipe. However, the blowers  did not
extend the measurable  suction field to greater distances
than the in-line fan had provided. These tests did not
include measurements of the effects of the different fans/
blowers on the radon levels in the house.

One firm reports having tested a high-suction  blower
(typically operating in the range of 4 to 14 in. WG) in 42
basement houses in New  England having very poor
communi-cation, with packed sand and/or clay beneath
the slab, and with system flows of about 20 cfm (Cra91).
Indoor levels were reportedly reduced below 2 pCi/L in
all of these houses, usually with only two suction pipes,
based  upon  short-term post-mitigation  measurements.
Three-quarters  of the houses were  reportedly reduced
below  1 pCi/L. Pre-mitigation indoor concentrations av-
eraged about 20 pCi/L. The two suction pipes corre-
    sponded to about one pipe per 500 ft2 of slab area, on the
    average.

    In summary, available data on the use of high-suction
    fans/blowers are limited. In very poor-communication
    houses with low flows, such blowers would be operating
    more comfortably on their performance curve than would
    the standard in-line fans, and they thus might have a
    longer lifetime. However, at the present time, the limited
    results are mixed regarding whether radon reduction per-
    formance would be significantly improved by such blow-
    ers. It has not been demonstrated that such high-suction
    fans/blowers will in fact consistently result in an exten-
    sion of a sub-slab suction field to portions of the slab that
    would not be reached by a standard 90-watt in-line cen-
    trifugal fan, although some additional extension would
    intuitively be expected. Accordingly, definitive guidance
    cannot currently be given regarding under what condi-
    tions such blowers  would be worth the increased cost
    (and sometimes increased noise) that they would entail.

 •  Fan in  suction vs. pressure. The case where the SSD
    fan is reversed, so that it pressurizes the sub-slab region
    with outdoor air rather than depressurizing it, will be
    addressed in Section 2.4.

e)  Geology/climate variables

 •  Source strength. "Source strength" refers to the amount
    of radon which the underlying soil can supply to a house,
    and depends upon both the radon concentration in the soil
    gas and the rate at which the soil gas can move through
    the soil. Most  commonly, a high source strength results
    from a high radon concentration in the soil gas; in this
    document, soil gas concentrations greater than 2,000 pCi/
    L will be referred to as "high."

    High source strengths necessitate more care in the design
    of SSD systems, even where sub-slab communication is
    good, because:

    a) any soil gas entry route left untreated by the system
       will be potentially significant, in view of the concen-
       tration of radon that can enter the house through that
       opening; and

    b) re-entrainment of the SSD system exhaust back into
       the house can be more significant, in view of the
       radon concentrations that will exist in the  exhaust
       (with re-entrainment of only 0.1% of the  exhaust
       being sufficient in worst-case houses to create indoor
       levels of 4 pCi/L and greater) (Fi91).

    Basement and slab-on-grade houses having high source
    strengths have commonly demonstrated the greatest diffi-
    culty in achieving post-mitigation indoor levels signifi-
    cantly below 4 pCi/L (Sc88, Fo89, Fi91, Fo92). Such
    houses may have difficulty achieving near-ambient in-
    door levels.

 •  Permeability of underlying soil. In most cases, the
    performance of a SSD system is determined by the com-
                                                    33

-------
munication within (the penneability of) the material im-
mediately below the slab, rather than that of the native
soil. The material below the slab will commonly be
gravel or crushed rock, or perhaps native soil that has
been disturbed during construction of the house;  such
materials will generally have better permeabih'ty than the
undisturbed native soil. If the native  soil is  relatively
impermeable but  the material directly under the slab is
fairly permeable,  good SSD performance is commonly
achieved (e.g., Fi90).

However, the permeability of the native soil can be
important in some cases. Soils having moderate perme-
ability might aid the SSD system in treating entry routes
associated with block foundation walls, either through
extension of the sub-slab suction field under footings to
the exterior face of the wall, or through dilution of the soil
gas with ambient  air drawn through down the  soil. This
statement is based on the observation that aBWD compo-
nent has more commonly seemed to be needed, in addi-
tion to or instead of SSD, when the sub-slab communica-
tion has been very poor (Ni89, Py90, Fi91).

Where the permeability of the native soil is  relatively
high, SSD performance can suffer, apparently because
outdoor air flowing down through the soil and into the
system can interfere with the extension of the suction
field. This problem was encountered hi basement houses
built on glacially deposited, excessively drained sandy
and gravelly soils  in the Spokane River Valley (Tu87). In
5 of these houses  where both pressurization and depres-
surization of the sub-slab were tested, reversing the fan to
pressurize the sub-slab (an approach which depends upon
the ability to maintain high flows through the system)
was consistently found to provide better radon reductions
than did SSD (which is better able to maintain adequate
suction fields in cases where flows are moderate to low).
Similar results were observed in 1 house built on a very
well-drained gravel soil in New York (Kn90).

Another way of stating the above observation is that at
least two  mechanisms are contributing to SSD perfor-
mance, as discussed earlier: soil depressurization and soil
gas dilution. With relatively tight native soils, soil de-
pressurization  is likely the more important of the  two
mechanisms, and  SSD is the technique  of choice.  But
with highly permeable native soils, high air flows are
established through the soil; soil gas dilution becomes the
predominant mechanism, and maintenance of a sub-slab
suction field becomes more difficult, with the result that
active soil pressurization may become the technique of
choice.

Climatic conditions. Weather conditions can influence
the performance of SSD systems in several ways.

a) Cold temperatures will increase the depressurization
   created by the  thermal stack effect on the lower level
   of the house, depending upon the height of the house
   and the temperature difference between indoors  and
   outdoors. This  depressurization will increase the driv-
   ing force for  radon  entry, and will  give  the SSD
system a greater indoor depressurization to overcome.
Basement depressurization in a two-story house cre-
ated by the stack effect during cold weather would be
approximately 0.015 in. WG (Sau89).

If sub-slab depressurizations being created by a SSD
system were being measured during  mild  weather
with exhaust appliances off, the conservative rule of
thumb would thus be that the system should be de-
signed to maintain a depressurization of at least 0.015
in. WG everywhere to avoid being overwhelmed by
the stack effect when cold weather arrives.  In addi-
tion, to avoid being overwhelmed by the incremental
basement depressurization created when exhaust ap-
pliances are turned on during cold weather, the SSD
system should nominally maintain an additional sub-
slab depressurization of up to 0.01 to 0.02 in. WG, as
discussed previously in Section 2.3.Ib. Thus, ideally,
sub-slab depressurizations  measured during mild
weather with appliances off should total about 0.025
to 0.035 in. WG everywhere in order to ensure that the
system will never be overwhelmed during cold weather
with the appliances on.

But as re-iterated several places in this document, this
target depressurization is usually a very conservative
design goal. Commonly,  sub-slab depressurizations
much less than these  ideal targets will still provide
satisfactory SSD performance.  Thus,  an expensive
upgrade of a SSD system in an attempt to achieve
these high depressurizations is often unnecessary.
However, where the SSD system can reasonably be
designed to achieve such depressurizations, it is prob-
ably advisable to do so.

Furthermore, this conservative maximum basement
depressurization  of 0.025  to 0.035  in. WG due to
thermal and appliance effects is thought to be high for
many cases. Houses which  are in milder climates,
which are leaky, or which do not have some of the
major depressurizing appliances (e.g., clothes driers,
whole house fans, central furnace fans) will encounter
lower worst-case basement depressurizations. In addi-
tion, the upper end of the range assumes that the major
depressurizing appliances  are operating  during the
coldest weather; among these appliances, whole-house
and attic fans will in fact not be  operated in cold
weather,  and clothes  driers will be operated only
intermittently. Combustion appliances in the base-
ment would backdraft if depressurizations as great as
0.035 in. WG were actually maintained for any ex-
tended period. But although this basement depressur-
ization range may be conservatively high for many
houses, it is used throughout this  document as  a
reasonable,  conservative design tool which can be
useful as long as it is properly understood.

Where an aggregate layer exists beneath the slab, SSD
systems can commonly achieve sufficient sub-slab
depressurizations to compensate for the worst-case
stack effect and exhaust appliance effects with one or
two suction pipes (e.g., Gi90, Mes90a, Fi91). Where
                                                    34

-------
   sub-slab communication is not good, more suction
   pipes, and more careful placement of the pipes, would
   be necessary to achieve 0.025 to 0.035 in. WG.

   Where slab pressure measurements are made during
   cold weather with exhaust appliances on—i.e., with
   the system experiencing its worst-case  challenge—
   any measurable sub-slab depressurization should be
   sufficient (0.001 to 0.002 in. WG).

b) Winds blowing against a house will create negative
   pressures across the house shell on the downwind side
   at the roof, and will create positive pressures on the
   upwind side. The net effect of these pressure changes
   on SSD system performance will vary from house to
   house, depending upon where the major openings
   through the house shell are, and where the major soil
   gas entry routes are. Data on wind speeds and direc-
   tions around mitigated houses have been collected in
   only a few studies (Tu87, Ma88, Tu88a, Du90, Tu91b).
   The data from these studies has not been fully  ana-
   lyzed. From the analysis to date, no generally appli-
   cable, unambiguous effect of wind velocity has been
   demonstrated.

c) Decreases in barometric pressure sometimes appear to   f)
   create  significant  short-term spikes in indoor radon
   levels and brief degradations of SSD performance.  In    •
   two slab-on-grade houses in New Mexico (Tu91b),
   decreases in barometric pressure were found to  con-
   sistently cause spikes to 10 to 20 pCi/L, where the
   SSD system was otherwise maintaining levels below
   2 to 4 pCi/L. This barometric pressure effect appeared
   to be independent of other weather-related variables
   often associated  with  barometric pressure changes,
   namely, winds and precipitation. Similarly, others
   have reported that indoor radon concentrations com-
   monly spike when barometric pressure falls (K192).
   The probable mechanism responsible for this baro-    •
   metric effect is that the  indoor pressure decreases
   rapidly in conjunction  with the drop in barometric
   pressure, while the soil gas pressure decreases more
   slowly. This lag in equilibration of the soil gas pres-
   sure results in a temporary increase in the house
   depressurization  relative to the soil, increasing the
   driving force dramatically (by an amount potentially
   on the order of inches of water) and overwhelming the
   SSD system.

d) Precipitation, in  the form of rain or snow, might
   influence SSD performance via two different mecha-
   nisms.

   In the first mechanism, the resulting cap of snow  or
   water-saturated soil at grade level can divert soil gas
   toward the foundation which would otherwise escape
   to the outdoor air remote from the house. As one
   example  of how  this  might create  a problem, this
   diverted soil gas might result in increased entry via
   the exterior face of a block foundation wall which is
   not being treated by the SSD system. In this hypo-
   thetical case, failure of the system to treat the exterior
   face might not be a problem when the soil gas is not
   thus diverted by the cap. Another possible way of
   viewing this could be that the moisture cap on the
   surface could be blocking the flow  of ambient air
   down  through the soil into  the  SSD  system,  thus
   reducing the benefits of the soil gas dilution and air-
   barrier shielding mechanisms postulated at the begin-
   ning of Section 2.1

   In the second mechanism by which precipitation might
   affect SSD performance, moisture in the fill under the
   slab could reduce the communication within the fill,
   possibly reducing performance (unless the  moisture
   also blocks radon entry). Such changes in sub-slab
   communication with precipitation have been reported
   in Florida (Fo90), although the actual effects on SSD
   performance have not been closely studied.

   Of the investigators who have measured precipitation
   as part of mitigation projects (Tu87, Ma88, Tu89,
   Du90, Tu91b), those reporting results to date gener-
   ally indicate no discernible effect of precipitation on
   SSD performance (Tu89, Du90). However, more de-
   finitive data are needed.

Mitigation system durability

A number  of investigators have made measurements
around SSD systems, installed either as parts of a R,D&D
project or as a commercial installation, in order to assess
how well the systems were performing  1  to 4 years
following installation (Ni89, Pr89,  Du91, Fi90, Fi91,
Gad91, Ha91).  These studies have addressed both the
radon concentrations being maintained in the houses, and
the reasons for any observed degradation in performance
(decreases in system suctions and flows, hardware fail-
ure, or homeowner intervention).

Radon reduction  performance.  In general, except
where a SSD fan failed or where the system was turned
off by the homeowner, systems that were installed as part
of a R,D&D project have maintained fairly consistent
indoor radon concentrations over the 1- to 4-year periods
covered by the various studies.

One of the largest and longest-term durability studies
(Fi91) has addressed 38 basement houses in eastern Penn-
sylvania having very high pre-mitigation concentrations,
where systems were installed under an EPA project dur-
ing the period 1985-87 (Sc88). Twenty  of these houses
had SSD systems (one with aBWD component). Winter-
quarter alpha-track detector measurements in the base-
ments and living areas were compared for the winters of
1986-87,1987-88, and 1988-89 (Sc89, Sc90b). In 7 of the
20 SSD houses, the 1988-89 winter-quarter reading was
greater than the average over all three winters by 1.0 pCi/
L or more, suggesting a possible degradation in system
performance over the two to four years since installation.
The value of  1.0 pCi/L represents the estimated  95%
confidence interval in the measurements, considering
both the  alpha-track measurement uncertainty and the
natural radon variations in the house. Testing indicated
                                                    35

-------
that, in four of the seven houses varying by more than 1.0
pCi/L, the increase was likely due to variations in re-
entrainment of the  SSD exhaust (containing up to 8,000
pCi/L), or in  radon released from well water. In a fifth
house, the increase was  demonstrated  to be due to  a
dramatic increase  in radon entry through an opening
through the poured concrete foundation wall, well above
the slab; basement concentrations were reduced to 2 pCi/
L by closing that wall opening. In only two of the seven
houses could  the increase not be explained by one of the
factors above. In these two, 1988-89 readings varied from
the three-winter average by 1.0 and 1.3 pCi/L, respec-
tively. In view of the conservative nature of the ± 1.0 pCi/
L confidence interval, and in view of the small absolute
variation in concentrations over the three winters, it is not
clear that the observed increases in fact represent any real
system degradation. Measurements of system piping flows/
suctions, and  of sub-slab depressurizations, confirm that
in none of these cases  were the radon increases due to
reduced  fan performance,  or to failure to adequately
depressurize the sub-slab.

In summary,  it was  concluded from the Pennsylvania
testing that, overall, none of the SSD systems have expe-
rienced any significant degradation over the years in their
ability to treat  the radon entry routes,  except in cases
where the fan failed (discussed later) or where the home-
owner turned  the system off.

In another study (Ha91, Gad91), three- to  six-month
alpha-track detector  measurements were made in ten
basement houses in New Jersey over a 2-year period,
after SSD systems had been installed as part of an earlier
RtD&D project (Tu89, Du90). In none of the houses did
there appear to be any significant deterioration in radon
levels in the house, except in two cases where the home-
owner turned off the fan. (Two of the houses did show
seasonal variations in radon levels.)

In a third study (Du91), continuous radon measurements
were conducted for a  year or more  in four  basement
houses in the Tennessee Valley having SSD systems (two
having SMD components) installed under  a R,D&D
project (Ma88,  Ma89a, Ma89b).  While radon measure-
ments did increase slightly in all four houses over the
year, in no case was the increase greater than 0.9 pCi/L
(with the average increase being 0.5 pCi/L), and in no
case did the  indoor levels  exceed 2.7 pCi/L. No clear
degradation of system performance is apparent.

In 10 houses  in Ohio having SSD systems (Fi90), quar-
terly alpha-track measurements conducted over a 1-year
period generally remained in the range  of 1 to 2 pCi/L
and less, with no quarter-to-quarter deterioration appar-
ent over the year. No seasonal variation was apparent

In an earlier study (Ni89),  14 New York houses which
had received radon mitigation systems as part of a 1984
project (Ni8S) were re-visited in 1986-87 to assess the
long-term effectiveness of  these early installations. Of
the six  basement houses  having SSD  systems,  radon
levels appeared to have  increased by 4 to 20 pCi/L in two
of the six, and marginally (by 1 to 3 pCi/L) in three
others. The increases resulted from condensate plugging
the exhaust piping in one house, partial blockage of the
grade-level exhaust pipe in two others, and radon leakage
out of the pressure side of the fan (which was inside the
basement) in one or two others. These problems reflect
the fact that these SSD installations were some of the first
in the U.S., and were among those contributing to the
understanding that now exists regarding sloping exhaust
piping to avoid condensate accumulation, and avoiding
fans inside the house.

While some preliminary measurements have been made
to assess  the durability of SSD  systems  installed by
private mitigators (Ha91, Gad91), definitive results are
not yet available.

System suctions and flows. System suctions and flows
have remained relatively steady over time in those dura-
bility test projects where those data have been reported
(Pr89, Fi91). In the SSD  installations in Pennsylvania
(Fi91), system conditions  remained remarkably steady
over the two to four years that the systems had been
operating, except when the fan failed. It was found that,
when the capacitor in the fan circuitry fails, the fan can
continue to operate for an extended period at dramatically
reduced suction and flow, as discussed later.

Sub-slab depressurizations. In four basement houses
in the Tennessee Valley (Du91), sub-slab depressuriza-
tions showed no discernible change over  the course of a
year,  when the  mitigation systems were  operating nor-
mally. In 20 Pennsylvania houses having SSD systems
(K91), sub-slab depressurizations remained generally high
after 2 to 4 years of operation.

Equipment durability. Of the 20 fans operating since
1985-87 on  SSD systems in the Pennsylvania project,
four have failed to date, all due to failure of the capacitor
in the fan circuitry (F191). All 20 of these fans are 90-watt
centrifugal in-line tubular fans from one  vendor. Of the
14 SSD installations in New Jersey (Tu89, Du90), three
90-watt tubular fans have failed since 1986-87, two due
to capacitor failure and one due to bearing failure (Ha91,
Gad91). One private mitigator who has installed approxi-
mately 100 fans of the same type that is in the Pennsylva-
nia study houses, estimates that roughly 10 failures have
occurred over the past 5 years, about  one-third due to
capacitors and two-thirds due to bearings  (Mes90b). The
vendor of these particular fans has indicated that, at least
in the earlier models, the capacitors installed in the fans
had a rated operating lifetime of 40,000 hours (about 4.5
years of continuous operation), in which  case additional
failures can be  expected  in  the  near future as these
installations age. Depending upon the brand of fan, ca-
pacitors can often be replaced relatively easily.

When the capacitor fails,  the fan will commonly stop
operating after a relatively short time because the coils in
the motor will burn out. However, in one case (Fi91), the
fan continued to operate at dramatically  reduced power
for a year or more after the capacitor failed. Due to the
                                                    36

-------
    drop in fan performance, radon levels in the house in-
    creased significantly (from 1.9 to 8.8 pCi/L in the base-
    ment, compared to pre-mitigation values of 148 pCi/L).
    Yet, because the fan was running, it would have sounded
    to the homeowner as if it were operating normally. This
    result underscores the need  for an flow- or suction-
    actuated alarm or indicator on SSD systems.

    If power to the fan is interrupted after the capacitor fails,
    the fan will not re-start when power is restored.

 •   Homeowner intervention.  Homeowners have occa-
    sionally turned SSD systems off, commonly due to fan
    noise, or in an effort to reduce the cost of electricity when
    windows are open during mild weather (Fi91,  Ha91).
    Fans have been turned off when the owners were away. In
    two instances, the fans were turned off because the speed
    controller on the fan  created radio  interference (Fi90,
    Ha91). Fans were recorded as having been turned off in
    eight of the 38 study houses in Pennsylvania (Fi91), and
    in two of the eight houses tested during the durability
    testing in New Jersey (Ha91).

    With the centrifugal tubular fans commonly used in SSD
    systems, fan noise can be largely eliminated if the fan is
    mounted properly with flexible couplings, to avoid vibra-
    tion in the exhaust piping and the house structural mem-
    bers.

    Turning the fan off when windows are open during mild
    weather should be discouraged, since it is generally im-
    practical to ensure that the fan is consistently turned back
    on when windows are closed. Annual alpha-track mea-
    surements  in the Pennsylvania study houses (Sc90b),
    where pre-mitigation levels were very high, showed that
    the open windows did not generally compensate for the
    system  being off. The result was that annual average
    levels in houses with the  fans off during the summer,
    were usually greater than the winter-quarter values in
    these same houses, since the systems were usually left on
    all winter.

    There is usually no problem with  fans being turned off
    while the house is unoccupied,  as long as the fans are
    consistently turned back on when the owners return.

    In those instances where radio interference occurs due to
    the  speed controller, options include a) re-locating the
    controller in a manner to shorten and re-locate the wire
    between the controller and the fan,  thus modifying the
    source of the interference; b) using a  more expensive
    controller;  or c) eliminating the controller, operating the
    fan at full power.

2.3.2  Active Sump/Drain-Tile
Depressurization

Active sump/DTD is probably the second most used variation
of the ASD technology, after SSD. Sump/DTD is likely the
most common  system installed in  houses  having  a sump
connected to drain tiles. (As discussed in Sections 2.1.2 and
2.2.2, SSD can sometimes be selected rather than sump/DTD
even when a sump with tiles is present.) While the exact count
is uncertain, sump/DTD systems have likely been installed in
thousands of houses.

As with SSD systems, sump/DTD systems have commonly
provided radon reductions of 80 to 99%. Exceptions  have
generally occurred primarily in cases where poor sub-slab
communication is combined with other problems, such as an
incomplete drain tile loop and a high source strength. The
highest percentage reductions are obtained  when the pre-
mitigation level is highly elevated. In some cases where the
basement house had an adjoining slab-on-grade  or crawl-
space wing, the sump/DTD system had to be combined with a
SSD or SMD component treating the adjoining wing.

The EPA research program has tested sump/DTD in a total of
16 basement houses, sometimes in conjunction with SSD or
SMD treating an adjoining slab-on-grade or crawl-space wing
(He87, Mi87, Sc88, Fi89, Ni89, Du90, Gi90, Mes90a). These
houses had pre-mitigation concentrations ranging from about
10 pCi/L to greater than 2,000 pCi/L. In all 16 houses, an
interior loop  of tiles (inside the footings) drained into the
sump, although some houses (Mi87) reportedly had exterior
loops also; none of the sumps had only exterior tiles. All but 1
of these 16 houses were reduced  below 4 pCi/L; 11  were
reduced below 2 pCi/L, and 7 (more than 40%) were reduced
below 1 pCi/L.

All seven of the houses reduced below 1 pCi/L were in the
Washington, D. C., area (Gi90, Mes90a), where  achieving
such low concentrations has generally proven to be easier than
in many other parts of the country. The readings below 1 pCi/
L have been confirmed by annual alpha-track measurements
in five of the seven houses. The sub-slab communication was
good in all seven houses, and pre-mitigation concentrations
were generally below 10-20 pCi/L.

The  one house which was not reduced below 4 pCi/L was
House C32D in Mi87, which remained elevated (with post-
mitigation readings of 5-13 pCi/L) even though the sump/
DTD system  was creating a very effective depressurization
(over 0.1 in. WG) everywhere beneath the basement slab
(Br92). This house probably remained elevated for two rea-
sons.

   -  The house had an extremely high source strength, with
      one pre-mitigation measurement as high as 1,357 pCi/
      L. As discussed previously,

      —  a high source strength increases the significance of
         any entry route left untreated by the system. Al-
         though the slab was effectively depressurized, it is
         believed that the sump/DTD system was not ad-
         equately preventing the  flow of very high-radon
         soil gas via the block wall cavities.

      —  a high source  strength means that any re-entrain-
         ment of system exhaust will present an increased
         problem, since radon levels in the exhaust will also
         likely be very high. Re-entrainment had been an
         early problem in this house, and some reduced re-
         entrainment may have been continuing even after
                                                       37

-------
         initial modifications had been made to the system to
         take care of this problem.
complete interior loop (Br92). The nature and extent of the
tiles in the other house were not reported.
   -  The house had an adjoining crawl-space wing which
      was not treated.

All four of the houses which were reduced below 4 but not
below 2 pCi/L were among the earlier houses tested in the
EPA program, and most of them  offered various complica-
tions. Two of them (Houses C30A and C39A in Mi87) had
extremely high source strengths, and had adjoining slabs on
grade which might not have been being adequately depressur-
ized by the SSD component treating the slab-on-grade wing.
Another house (House AR-20 in Ni89) had an unclosed
perimeter channel drain at the time that the post-mitigation
reading of 2.3 pCi/L was obtained, and showed a notable
degradation in system performance a year later when the
measurement was repeated with the perimeter channel drain
closed, casting some doubt on the final results.

In summary, sump/DTD with interior drain tile loops appears
consistently able to reduce even high-level basement houses
to 4 pCi/L and less, except in instances where there is  a
combination of difficulties such as high source strength and
untreated adjoining wings. Incomplete drain tile loops com-
bined with poor sub-slab communication might also be ex-
pected to present a problem. Where communication is reason-
ably good, sump/DTD with interior tiles will likely achieve 2
pCi/L and less in the large majority of cases.

The EPA data base does  not include testing in houses where
an exterior loop (outside the footings) drains into the sump.
However, as discussed later, mitigators in some regions of the
country have reported extensive experience with such systems
(K192).

The post-mitigation measurements reported for these 16 houses
represent a range of measurement methods  and durations.
Some of these measurements are from annual  alpha-track
measurements, while others are from 4-day to 2-week con-
tinuous radon monitor readings. However, it is  clear  that,
overall, the sump/DTD systems are being very effective.

The positive experience under the EPA R.D&D program with
sump/DTD systems having interior tiles, is generally reflected
in the experience of other investigators and  of commercial
mitigators.

Radon reductions of 70 to 95% were reported with four early
sump/DTD installations  in basement  houses in New York
(Ni85). Three of these houses were reduced below 4 pCi/L in
the basement, one of them below 2 pCi/L. These reductions
were achieved despite a variety of complications, including a)
two of the basements had adjoining crawl spaces (which were
cither actively or passively ventilated as part of the mitigation
effort); b) one of the houses (with a crawl space) had only a
partial exterior drain tile loop, which would generally be
expected to give poorer performance than interior loops; and
c) the 24-watt fans used were much smaller than the 90-watt
units commonly  used today. All  four houses had pebble
aggregate under the slab (Br92). In addition to the one house
having an partial exterior loop, two of the other houses had a
Active sump/DTD systems were installed in a number of
block basement houses as part of remedial work in several
mining communities in Canada (Ar82). Radon reductions of
60 to 80% were reported for these early installations. Some of
the key details regarding these installations were not reported,
such as the nature and extent of the drain tile loops, and the
nature of the sub-slab communication. In addition, other steps
(such as source removal) were commonly implemented in
conjunction with the DTD, so that the effects of the DTD
system alone cannot always be separated out.

Numerous commercial mitigators  have reported installing
large numbers of sump/DTD systems in cases where the drain
tiles form interior loops, inside the  footings  (e.g.,  Mes91,
Bro92). The performance of these systems is reported to be
generally high, although comprehensive performance data for
all of these installations are not available.

Mitigators working in some regions of the country, such as the
eastern slopes of the Rocky Mountains, have reported exten-
sive experience with sump/DTD systems in cases where the
tiles which drain into the sump are outside the footings
(K192). These systems have been reported to give consistently
good performance when  the exterior tiles form  a complete
loop around full basements.  The  underlying soil in these
houses is commonly decomposed granite, a gravel/clay mix
with a permeability that is variable, but sometimes relatively
high. There is often no imported crushed rock placed under
these  basement slabs, with the native decomposed granite
having been used in lieu of crushed rock. The good perfor-
mance of these exterior-loop systems, despite the absence of
crushed rock, may be due in some cases to the relatively good
permeability which sometimes exists in the native soil.

The following discussion summarizes the results to date with
sump/DTD systems, addressing each of the house, mitigation
system, and geology/climate variables that can influence sys-
tem design, operation, and performance. In many cases, the
effects of the variables on sump/DTD systems are similar to
the effects discussed for SSD systems in Section 2.3.1; in
those cases, the discussion here is abbreviated.

a)  House design variables

 •  Sub-slab  communication.  Among the  EPA  study
    houses where communication was known to be good (13
    of the 16 houses), sump/DTD reduced levels below 4
    pCi/L in all but one case. That one case was House C32D
    in Mi87, where a very  high source strength  and an
    adjoining wing were complications. In  10 of these 13
    good-communication houses, levels fell below 2 pCi/L.
    Even in the one EPA study house where the communica-
    tion was reported to be marginal (House AR-16 in Ni89),
    levels were reduced to below 2 pCi/L. In the two EPA
    study houses where the communication was not reported
    (Mi87), levels were reduced below 4 pCi/L, despite high
    source  strengths.  In the four sump/DTD installations
    reported in Ni85, all of which involved good communica-
    tion, levels were reduced below 4 pCi/L by three of them,
                                                       38

-------
despite the use of low-power (24-watt) fans and, in some
cases, incomplete loops and untreated adjoining wings. In
all of these houses (except for one or two in Reference
Ni85), the drain tiles were inside the footings.

Where sub-slab communication is good and where the
tiles are inside the footings, mitigators report routinely
achieving concentrations well below 4 pCJ/L with sump/
DTD (Mes91, Bro92).

Thus, when the tiles are inside the footings, sump/DTD
consistently gives good performance when communica-
tion is good (Sc88, Fi89, Ni89, Du90, Gi90, Mes90a,
Mes91, Bro92). Even where communication is marginal,
sump/DTD might be expected to give reasonable perfor-
mance with interior tiles (Ni89). The drain tiles would be
distributing the suction near the major entry routes (the
perimeter wall/floor joint and the base of the block foun-
dation walls). Sump/DTD systems with interior tiles are
most likely to be inadequate as stand-alone methods
when there is a combination of complications, including
poor sub-slab communication, an incomplete tile loop, a
high source strength, and/or an untreated adjoining wing
which is an important radon source. But even in those
cases, sump/DTD could still be an important component
of any system that is installed.

As  indicated earlier,  good radon reductions have been
reported in basements having no sub-slab crushed rock
and with exterior drain tiles, when the tiles form a nearly
complete loop around the house (K192). At least in some
cases, relatively good permeability in the underlying
decomposed granite may be contributing to this effective-
ness.

Extent of drain tile loop. Since most of EPA's study
houses and many of the commercially-mitigated houses
appear to have nearly complete interior drain tile loops,
there are not sufficient data to quantify the effects of this
variable. Where there is good communication and where
the drain tiles are inside the footings, it will probably not
be so important that a complete loop be present. Where
the communication is poor,  or when the tiles are outside
the footings as in Colorado (K192), the need for a com-
plete loop will likely be more important, to ensure that
suction reaches the entire perimeter. The effect of incom-
plete exterior loops is discussed in Section 2.3.3a, in
connection with drain tiles that discharge remote from the
house.

Nature a/foundation wall.  Where sub-slab communi-
cation is reasonably good, the nature of the foundation
wall (block vs. poured concrete) does not appear to make
any difference on sump/DTD performance. Of the 11
EPA study houses reduced below 2 pCi/L by sump/DTD,
and of the 7 reduced below 1  pCi/L, more than half had
block foundations. However, the data are too limited to
determine whether block foundations become more of a
problem when communication is poor.

Of the five EPA study houses not reduced to 2 pCi/L or
less, four  had block foundations; however, as indicated
earlier, some of these had other complications which may
have been playing more of a role than was the block wall,
such as very high source  strengths and inadequately-
treated adjoining wings. In  the one house which was not
reduced below 4 pCi/L  (House C32D in NC87), post-
mitigation diagnostics suggested that the block founda-
tion walls were apparently still a radon source, despite
excellent depressurization of the sub-slab by the sump/
DTD system (Br92); but again, the continued importance
of the block walls was likely due to the unusually high
source strength under this house. The one  house with
marginal communication that was reduced below 2 pCi/L
(Ni89) did happen to have a poured concrete wall.

In summary, sump/DTD with complete interior drain tile
loops would generally be expected to do a fairly good job
at treating block foundation walls. The suction is being
distributed around the entire perimeter of the wall at its
base, thus potentially preventing soil gas from entering
the void network inside the wall. However, where poten-
tially complicating factors  exist, such as a high  source
strength (or perhaps such as poor sub-slab communica-
tion and an incomplete tile loop), the limited data suggest
that the presence of a block wall rather than  a poured
concrete wall may increase the chances that system per-
formance will be reduced.

House size.  One of the EPA study houses had a base-
ment floor slab larger than 2,000 ft2, and was reduced
below 2 pCi/L with the sump/DTD system (Fi89). This
house had a complete interior drain tile loop and aggre-
gate beneath the slab. With the  drain tiles aiding in
distribution of the suction, it would be expected that
sump/DTD should be able to treat fairly large slabs. This
would especially be true when the tile loop is  complete
and/or where there is good  sub-slab communication.

Adjoining wings. The one study directly addressing the
need to treat adjoining wings in basement houses (Mes90a)
found that, with either a SSD or a sump/DTD system in
the basement, the need to treat the adjoining slab on grade
or crawl space was greatest when a) the adjoining wing
provides important entry routes; and b) communication
beneath the basement slab is poor. Five of EPA's 16
sump/DTD study houses were not reduced to 2 pCi/L or
less; of these, 4 had adjoining wings which were poten-
tially important sources, in view of the source strengths in
those studies (Mi87, Sc88). In one of those cases (House
C32D in Mi87), the adjoining wing was not being  treated
at all.

In some cases where the drain tiles are outside the foot-
ings, these tiles have sometimes been observed  to  extend
around the slab-on-grade wing as well as the basement in
split-level houses (K192). In such cases, sump/DTD would
have an increased likelihood of treating both wings effec-
tively. (It should be noted that, when there are interior
tiles leading to the basement sump, it is much less likely
that these tiles Will extend around the adjoining slab on
grade.)
                                                    39

-------
 •  Sub-slab obstructions. No data are available on the
    ability of sump/DTD systems to deal with sub-slab ob-
    structions, such as interior footings or forced-air ducts. If
    the tiles draining into the sump are interior tiles, then,
    unless the obstructions interrupt the tiles, it would be
    anticipated that the sump/DTD system would likely do a
    better job than SSD  in distributing  the suction field
    around these obstructions. If the tiles are exterior tiles,
    and especially if the obstruction is an interior footing
    supporting a load-bearing wall penetrating the slab, or is
    otherwise an interior entry route, it would be anticipated
    that the obstruction would have a greater likelihood of
    degrading sump/ DTD performance.

b)  House operating variables

 •  Operation  of central furnace fan, and exhaust
    fans. See comments  in Section 2.3.1b, regarding the
    effect of these variables on SSD systems.

c)  Mitigation system design variables

 •  Size of suction pipe. Active sump/DTD systems gener-
    ally have system flows toward the upper end of the range
    observed, for SSD systems. Because the drain tiles facili-
    tate air collection from around the entire perimeter, flows
    are generally greater than 50 cfm, and are often greater
    than 100 cfm, with the 90-watt fans. At these  flows,
    mitigators will probably often find that pipe diameters of
    at least 4 in. will be desirable to  limit the suction loss and
    pipe flow noise.

 »  Degree of slab sealing. No  data have been reported
    quantifying the effect of slab sealing on sump/DTD per-
    formance. Since the drain tile loop enables the system to
    effectively draw air from around the entire perimeter, it
    would be expected that caulking the wall/floor joint,
    wherever accessible, might significantly reduce the amount
    of air entrained by the system, if the joint is more than a
    hairline crack. Such reduced entrainment would reduce
    the heating/cooling penalty, if not improve system per-
    formance.

d)  Mitigation system operating variables

 •  Fan capacity. Except possibly for a few of the earliest
    installations, all 16 of the sump/DTD systems installed
    under the EPA program have used 90-watt centrifugal in-
    line tubular fans with 6-in.  couplings, operating  at full
    capacity. Some mitigators report that, where communica-
    tion is good, good radon reductions can be achieved with
    smaller  fans (e.g., 50-watt  fans with 5-in. couplings),
    despite the lower flows and suctions with those fans
    (Mes91).

    Data have not been reported quantifying the effect of fan
    capacity on system performance. Because of the higher
    flows sometimes observed from sump/DTD systems, use
    of smaller, lower-flow fans might sometimes reduce the
    radon reduction performance of the system.
    It would not generally be expected that higher-flow fans,
    larger than the 90-watt fans (capable of moving up to 270
    cfm at zero static pressure), would be needed on residen-
    tial sump/DTD systems. Flows too great to be handled by
    the 90-watt fans would usually be suggesting that air was
    short-circuiting into the system.

 •  Fan in suction vs. pressure. See Section 2.4. No data
    have been found where a system has been installed to
    pressurize a sump/drain tile network. Soils so poorly
    drained that a sump is necessary in the house, will prob-
    ably not be sufficiently permeable to warrant sump/drain
    tile pressurization rather than depressurization.

    However, cases may occasionally be encountered where
    a sump has been installed in a house built  on a highly
    permeable native soil, because of a particular builder's
    standard procedures or because of code requirements in
    the region. In such cases,  sump pressurization might be
    considered.

e)  Geology/climate variables

 •  Source strength. As with SSD systems (Section 2.3.le),
    sump/DTD systems installed in houses having high source
    strengths have demonstrated the greatest difficulty in
    achieving post-mitigation  indoor levels of 2 pCi/L and
    less (Mi87).

 •  Permeability of the underlying soil. See the discus-
    sion concerning SSD systems (Section 2.3.le). Although
    the data base with sump/DTD systems is limited, it might
    be expected that, with the drain tiles located immediately
    beside the foundation walls, these systems might some-
    times be better able  than SSD  systems  to treat  entry
    routes associated with block foundation walls, when the
    permeability of the underlying soil is poor. When the
    drain tiles are outside the footings, as commonly reported
    in Colorado (K192), relatively good permeability in the
    underlying soil may be among the factors that can con-
    tribute to extension of the  suction field beneath the foot-
    ings and beneath  the slab (which, in that part of the
    country, is commonly poured on the native soil with no
    crushed rock).

 •  Climatic conditions. Although less data are available
    for sump/DTD systems than for SSD systems regarding
    the effects of  climatic  variables, the SSD  results dis-
    cussed in Section 2.3. le would be expected to be gener-
    ally applicable to sump/DTD systems as well.

f)  Mitigation system durability

 •  Radon reduction performance.  Of the 38 EPA miti-
    gation installations in eastern Pennsylvania which  have
    been monitored since 1985-87 (Fi91), one (House 29)
    involved  sump/DTD with interior  tiles. Winter-quarter
    alpha-track detector measurements in that house over the
    3 years following installation remained very steady, with
    no one winter's reading varying from the three-winter
    average (1.9 pCi/L) by more than + 0.5 pCi/L.
                                                       40

-------
    In  another study (Gad91), 4- to  6-month alpha-track
    measurements were repeated over a 2-year period in one
    house in New Jersey having a sump/DTD system (House
    4 from Du90). Concentrations in the basement remained
    steady between 2 and 3 pCi/L during all five measure-
    ment periods.

    In two houses in Ohio having sump/DTD systems (Fi90),
    quarterly alpha-track measurements over a 1-year period
    generally remained steady between 1 and 2 pCi/L in the
    basement, except during one quarter in one of the houses
    when the fan failed due to improper mounting.

    During the year following installation of five sump/DTD
    systems in Maryland (Mes90a), annual alpha-track mea-
    surements were conducted (Mes90c). All five houses
    remained at 1 pCi/L and less over the year.

 •  System suctions and flows. System suctions and flows
    remained steady over the 3 years since installation, in the
    one sump/DTD installation among the 38 systems  that
    have been monitored in Pennsylvania (Fi91).

 •  Sub-slab depressurizations. In the one Pennsylvania
    house (Fi91), sub-slab depressurizations remained very
    high (0.64-0.69 in. WG) in the 3 years since installation.

 •  Equipment durability. Nine of the 16 EPA sump/DTD
    installations were monitored for a year or more following
    installation. Only one of the nine fans failed,  a wall-
    mounted unit in Ohio (Fi90). That failure was the result
    of improper mounting, rather than to an inherent problem
    with the fan.

 •  Homeowner intervention.  No  cases of homeowner
    intervention were reported for EPA's sump/DTD instal-
    lations.

    In general, the frequency of such intervention for sump/
    DTD systems would be expected to be somewhat greater
    than that discussed in Section 2.3.le in connection with
    SSD systems. The owner or service personnel will occa-
    sionally have to remove and re-install the sump cover for
    sump pump maintenance, and there would be some risk
    that the cover may not be re-installed properly.

2.3.3  Active Drain-Tile Depressurization
(Above-Grade/Dry-Well Discharge)

Active DTD, in cases where the tiles discharge remote from
the house (to an above-grade outfall or to an underground dry
well), is less common than is sump/DTD in many parts of the
country, but is fairly common in some locations.

Drain-tile systems with remote discharges commonly (but do
not always) involve exterior drain tile  loops, outside the
footings. DTD systems on exterior tiles are expected to be
more dependent upon the presence of nearly complete loops
to ensure adequate treatment. That is, the loop should ideally
extend around all four sides of the house in full basements, or
around the three buried sides in walk-out basements. The need
for a nearly complete loop is expected because the suction
field from incomplete exterior loops will not always effec-
tively extend through relatively impermeable native soil around
the outside of the footings to reach sections of the foundation
where tiles are not present, to treat the foundation wall and the
wall/floor joint at those locations. Also, the exterior suction
field can be hindered in extending underneath the footings or
through the foundation wall to reach any entry routes that may
exist in the central portion of the slab. However, cases have
been reported where good reductions were achieved with
DTD/remote discharge when the exterior tiles extended be-
side only one or two sides of the house; this good performance
may be the result of relatively high permeability in the native
soil.

By comparison,  interior drain tiles (more common when the
tiles  drain to a sump) can better extend suction around the
perimeter and to interior  entry routes even when the loop is
not complete, at least when aggregate is present beneath the
slab. Because results to  date from active DTD/remote dis-
charge systems are all from systems treating  exterior tiles, the
observed performance with DTD/remote discharge are vari-
able, depending upon the extent of the drain tile loop and the
permeability of the native soil, probably among other factors.

Most of EPA's direct data on the performance of DTD/remote
discharge systems are from tests in eight houses in Pennsylva-
nia (Sc88, He89, Fi91). In all of these houses, the tiles were
outside the footings, and drained to an above-grade discharge.
The houses were all basements with block foundation walls.

Five of these houses had essentially complete exterior drain
tile loops (Houses 10,12,15,26 and 27). From winter-quarter
alpha-track measurements in these houses over the 3 years
following installation, three of these houses were reduced
below 4 pCi/L in the basement, and two (Houses 15 and 26)
were reduced below 2 pCi/L. Of the two houses not reduced
below  4 pCi/L according to  the alpha-track measurements
(Houses 10 and 27), it has been demonstrated (Fi91) that the
failure of the system to achieve levels below 4 pCi/L was not
due to inadequate treatment of the foundation by  the DTD/
remote discharge system, but rather, was due to a) re-emtain-
ment of the system exhaust (radon levels in the exhausts were
650 to 2,300 pCi/L); and b) in the case of House 10, radon
released from well water (which contained 26,000 pCi/L). Re-
direction of the system exhausts, and treatment of the well
water in House 10, reduced basement concentrations in both
houses to below 2.5 pCi/L.

In summary, in  all five EPA study houses  having complete
exterior loops, DTD/remote  discharge  reduced all five to
below 4 pCi/L in the basement (actually, below 3 pCi/L), and
reduced two below 2 pCi/L, when supplemented in one case
by well water treatment. These concentrations generally cor-
respond to radon reductions of 90 to 99+%, except in the case
of House 12, where the relatively low pre-mitigation concen-
tration (11 pCi/L) caused the percentage reduction to be lower
(77%).

The remaining three houses with DTD/remote discharge sys-
tems in the Pennsylvania study had incomplete exterior loops
(Houses 13,14, and 16). In none of these houses was the DTD
system by itself able to reduce the house below about 10 pCi/
                                                      41

-------
L (Sc88, He89). In House 13, where the tiles may have been
beside portions of three sides of the house, the DTD system by
itself reduced basement concentrations to 11 pCi/L (84%
reduction), based upon short-term measurements. Adding a
separate one-pipe SSD system to supplement the DTD system
in House 13 decreased levels to 2.5 pCi/L, based on 3- to 4-
month alpha-track measurements during each of three win-
ters. The other two houses had tiles beside only two sides of
the house, or had  an entire wing without tiles. The DTD
system reduced basement levels  to 15 pCi/L  in House 14
(74% reduction) and 150 pCi/L in House 16 (37% reduction)
(Sc88, He89). In these cases,  the DTD systems were aban-
doned in favor of BWD systems.

In summary, in the three EPA study houses where the exterior
loop was incomplete, the DTD system sometimes provided
moderate to high reductions (74 to 84%) even when the tiles
were beside only two or three sides of the house. In one case,
only low reductions (37%) were achieved when the tiles did
not extend around a wing of the house. Thus, even when the
tiles are incomplete, depressurization of an exterior drain tile
loop can sometimes be considered where only moderate re-
ductions are needed. However, with the high pre-mitigation
levels in the Pennsylvania houses, where high reductions were
required, in no case was DTD adequate by itself to reduce
levels below 4 pCi/L when the loop was not complete.

Commercial mitigators have reported good success with DTD/
remote discharge in hundreds of houses in and near Colorado
where exterior drain tile loops extended around all four sides
of the house (full basements) or around the three buried sides
(walk-out basements) (K192). In a few cases, good reductions
were reported even in cases where the tiles extended beside
only one  or  two sides of the house. These  results  were
obtained despite the fact that there was no crushed rock
beneath the slab; the slab was resting directly on graded native
soil, which is commonly decomposed granite (a  gravel/clay
mix of variable permeability). The effective performance is
attributed to possibly good permeability in the native soil in
some cases, permitting the suction field to extend through the
soil beneath the footings and beneath the slab; and the ten-
dency of the exterior drain tiles to intersect channels through
which utility lines penetrate  through the foundation, thus
providing a relatively high-permeability route for the suction
field to extend through the foundation wall/footings and be-
neath the slab (K192).

In one other early study (Sa84), 80% radon reduction (to a
post-mitigation level of 8 pCi/L) was achieved in one house
by suction on an exterior drain tile loop  with above-grade
discharge. The tiles apparently extended around the  three
sides of the house that were below grade, forming a U with
each leg of the U discharging to grade. Suction was drawn by
two 50- to 60-watt squirrel-cage blowers,  one on each leg of
the U; squirrel-cage blowers tend to produce higher flows and
lower suctions than do the 90-watt centrifugal fans commonly
used today. This was a "berm" house with poured concrete
foundation walls, good sub-slab aggregate, and forced-air
supply ducts beneath the slab.
The following discussion summarizes the results to date with
DTD/remote discharge systems, addressing the  individual
variables.

a)  House design variables

 •  Sub-slab communication. The effect of sub-slab com-
    munication on the performance of the exterior DTD
    systems discussed above is unclear.

    Many of the numerous Colorado houses discussed above
    (K192) did not have crushed rock beneath the slab, sug-
    gesting limited communication; yet, good radon reduc-
    tions were reported. This result could be suggesting that
    sub-slab communication may not be critical if the exte-
    rior tiles are beside all three or four buried sides of the
    house. The permeability of the native soil in this region is
    variable; in some cases,  this permeability may have been
    high enough to help compensate for the lack of crushed
    rock. Also,  it is not recorded to what extent these houses
    may have had soil gas entry routes at the interior of the
    slab, which might have challenged the DTD system in
    these potentially poor-communication cases if the suction
    field were extending only weakly inward from the perim-
    eter. Sub-slab communication would be expected to be
    less important for exterior DTD performance  if the main
    soil gas entry routes are the  perimeter walls  and the
    perimeter wall/floor joint.

    Among the EPA study houses, the two Pennsylvania
    houses giving the best results above (Houses  15 and 26,
    both having complete loops and achieving less than 2
    pCi/L in the basement), were the two houses having the
    poorest communication (Fi91). This result would suggest
    that, with exterior tiles (which are probably functioning
    by treating the base of the block foundation wall and the
    wall/floor joint), the completeness of the tile loop is more
    important than good communication beneath the slab.
    The other three houses  having complete loops and ulti-
    mately being reduced to 2 to 4 pCi/L (Houses  10,12, and
    27), had much better communication than did Houses 15
    and 26.

    Among these five Pennsylvania houses having complete
    exterior loops, the two houses having poor sub-slab com-
    munication had  the lowest sub-slab  depressurizations
    created by the exterior DTD system (with no depressur-
    ization measured at some of the  test holes in House 26,
    and respectable readings of generally 0.014-0.048 in.
    WG in House 15). By comparison,  two  of the three
    houses having good communication achieved better sub-
    slab depressurizations,  0.056-0.085 in. WG; the third
    house with good communication (House 12) had readings
    more  comparable to the poor-communication houses,
    0.014-0.018 in. WG. (For reference, the one sump/DTD
    installation in Pennsylvania with interior tiles and good
    communication achieved  sub-slab  depressurizations of
    0.625-0.685 in. WG.) From the above measurements, it
    would appear that good sub-slab communication  im-
    proves the chances that adequate sub-slab depressuriza-
    tions will be achieved  with exterior DTD systems, in-
    creasing the likelihood that any entry routes in the central
                                                       42

-------
portion of the slab would be treated. However, from the
House 12 results, good sub-slab communication does not
ensure that these systems will necessarily achieve high
depressurizations. And,  especially  from the House 26
results, where both communication and sub-slab depres-
surizations were poor, but where levels below 2 pCi/L
were consistently achieved nevertheless, it is apparent
that good communication and good depressurization are
not necessarily required for good radon reductions. This
confirms that treatment of the perimeter block walls and
the wall/floor joint (generally aided by having a complete
drain  tile  loop), rather  than  achieving good sub-slab
depressurization, is the key requirement for exterior DTD
systems.

The three Pennsylvania houses having only partial exte-
rior loops (Houses  13,  14, and 16) each had variable
communication, ranging from excellent under some por-
tions of each slab to very poor under other parts (Fi91).
House 16 had generally good communication beneath the
basement slab, but very poor communication beneath the
adjoining  paved crawl-space  addition where  the drain
tiles  did not extend. Thus, in these houses, it is not
possible to assess to what extent the inadequate perfor-
mance of the DTD/remote discharge system was due to
the incomplete tile loops,  and to  what extent it was
impacted by poor sub-slab communication. But, as indi-
cated above, it is suspected that, in the absence of major
entry routes in the central portion of the slab, the com-
pleteness of the loop is probably  the more important
variable.

The one house addressed in Sa84, which had essentially a
complete loop (on the three  sides  of the house below
grade), had good aggregate. However, due to the sub-slab
forced-air ducts, communication was probably not good.
The lower percentage reduction (80%) and higher re-
sidual radon levels (8 pCi/L) in this house were more
likely due to the unusual features of this house (a berm
house) and to the design of this early DTD system, than to
the nature of the sub-slab communication.

Extent of drain tile loop. As discussed above, the
exterior tiles common with DTD/ remote discharge sys-
tems may function to a large extent by treating the base of
the foundation wall and the wall/floor joint. The extent to
which the suction field from  these systems will extend
beneath the slab is quite variable, and is not always
clearly related to how good the sub-slab communication
is. In addition,  with the exterior tiles, the suction field
probably extends only weakly through the native soil
outside the footings to regions where the tiles are not
present, if the native soil has relatively low permeability.
Under these conditions,  the extent of the drain tile loop
could be a very important variable in determining the
effectiveness of these systems with exterior tiles. It may
often be important that  the tiles be located on three or
four sides of the house.

However, experience with DTD/remote discharge in Colo-
rado indicates that suction on exterior tiles can sometimes
be effective even in cases where the tiles are on only one
or two sides of the house (K192). Success with  such
partial loops appears to  depend upon relatively good
permeability in the native soil underlying the foundation.
Under this condition, the suction field from the partial
loop might be expected to better extend around the por-
tion of the perimeter without tiles, and beneath the slab.

Entry routes in center of slab. The available data are
not adequate to quantify the effects  of interior entry
routes on the performance of exterior DTD systems. Only
one of the Pennsylvania houses had a major interior route
(House 27, which had a hollow-block structure penetrat-
ing the middle of the slab to support a fireplace on the
floor above) (He87). The system in House 27 achieved
levels below 4 pCi/L (93% reduction) despite this interior
entry route, perhaps because the complete loop and the
good sub-slab communication resulted in sub-slab de-
pressurizations of 0.056-0.081 in. WG.

From the preceding discussion, it would be expected that
interior entry routes will likely be most important in cases
where poor sub-slab communication, poor permeability
in the native soil, an incomplete drain tile loop, or other
factors prevent the  suction  field from  the exterior tiles
from extending beneath the slab.

Nature of the foundation wall. The  available data are
not sufficient to determine the impact  if the foundation
wall were poured concrete rather than block.

All of the houses in Pennsylvania had block foundations.
Complete drain tile loops are ideally located to treat entry
through block walls, so that the nature  of the foundation
wall may not be particularly important when the loop is
complete. Since poured concrete walls substantially re-
duce wall-related entry routes,  exterior DTD systems
with partial loops may perform better  when the wall is
poured, since leaving a  portion  of the wall  untreated
would then be less of a problem. Poured concrete walls
might further hinder the extension of the exterior DTD
suction field into the region beneath the slab, compared to
relatively porous block walls.

House size. The data are insufficient to enable a defini-
tive  statement regarding  the effect of  house size. Only
one  of the five Pennsylvania houses  having  complete
loops had a slab larger than 1,000 ft2 (House 10, where
the slab was 1,300 ft2). Presumably, if  the exterior drain
tile loop were complete and were thus treating the entire
perimeter, the size of the slab would not be particularly
important, unless there was a significant entry route in the
central portion of the slab.

Adjoining wings. Experience in Colorado suggests that,
in many cases, exterior tiles around a basement will also
extend around an adjoining slab-on-grade wing (K192).
In such cases, it would be anticipated  that DTD/remote
discharge may effectively treat both wings.

The reader should be aware that, in other parts of the
country, it is possible that exterior drain tiles will not
extend around the adjoining wing, possibly leaving un-
                                                    43

-------
    treated both the adjoining wing and the side of the base-
    ment abutting the wing. None of the EPA study houses in
    Pennsylvania where DTD/remote discharge has been ef-
    fectively tested had adjoining wings.

 •  Sub-slab obstructions. The house in Sa84, with sub-
    slab forced-air ducts, is the only house reported with an
    exterior DTD system having sub-slab obstructions. The
    other unique features of this berm house prevent a clear
    interpretation of the effects of  these ducts on system
    performance.

    Where an obstruction is not associated with an interior
    entry route, it might be expected that the obstruction will
    not significantly interfere with exterior DTD performance,
    since the system appears to perform  largely by treating
    the foundation wall and wall/floor joint immediately ad-
    jacent to the tiles. But where the obstructions are associ-
    ated  with interior entry routes—e.g., an interior load-
    bearing wall or sub-slab ducts—it could be expected that
    the obstruction will sometimes degrade system perfor-
    mance, by providing an entry route potentially not effec-
    tively treatable by the exterior system.

b)  House operating variables

 •  Operation of  central furnace fan, and exhaust
    fans. See comments in Section 2.3.1b. To the extent that
    exterior DTD/remote discharge systems will sometimes
    provide lesser sub-slab depressurizations that will SSD or
    sump/interior DTD systems, the exterior DTD system
    may sometimes be more prone to being overwhelmed by
    basement depressurizations created by these fans.

c)  Mitigation system design variables

 •  Size of suction pipe. Four-in. diameter pipe has been
    used on all of the EPA-sponsored DTD/remote discharge
    installations (Sc88). The flows from all five of the sys-
    tems still operating in Pennsylvania have been above 100
    cfm, confirming that pipe of at least that size will gener-
    ally be needed. Flows in  the Colorado systems have
    generally been somewhat lower, 50 to 80 cfm (K192); 4-
    in. piping has routinely been used in the Colorado instal-
    lations as well. The 4-in. piping is also convenient for
    connecting to the exterior drain tiles, which are usually
    cither 3 or 4 in. in diameter.

 •  Degree of slab sealing. No data have been reported
    quantifying the effect of slab sealing on DTD/remote
    discharge performance. As with the sump/DTD systems
    (Section 2.3.2c), because the tiles are immediately beside
    the footings, caulking the wall/floor joint would usually
    be a desirable step wherever die joint is accessible and
    whenever it is more than a hairline crack.

d)  Mitigation system operating variables

 •  Fan capacity. All five of the exterior DTD systems in
    Pennsylvania have  the 90-watt centrifugal tubular fans.
    Given the relatively high flows in these systems (>100
    cfm) and the need for high suctions in an attempt to
    extend the suction field under the slab, the 90-watt fans
    will usually be the logical choice, rather than a smaller
    unit. The 90-watt fans have also routinely been used in
    the Colorado installations.

 •  Fan in suction vs. pressure. See Section 2.4. Soils so
    poorly drained that  drain tiles are necessary will not
    usually be sufficiently permeable to warrant DTD pres-
    surization rather than depressurization.

e)  Geology/climate variables

    See the discussion for these variables in Sections 2.3.le.

f)  Mitigation system  durability

 •  Radon reduction performance. The only durability
    data on DTD/remote discharge systems are from the five
    houses in Pennsylvania (Fi91). In all cases except House
    10, where exhaust re-entrainment and well water contri-
    butions caused larger fluctuations, none of the winter-
    quarter alpha-track measurements over a three-year pe-
    riod varied from the three-winter average by more than ±
    1 pCi/L.  In no case was a progressive degradation in
    performance apparent.

 •  System  suctions  and flows. In all five Pennsylvania
    houses, the suctions and flows in the DTD suction pipe
    remained steady, within the accuracy of the measure-
    ment, over the 3 to 4 years following installation.

 •  Sub-slab depressurizations. No degradation in sub-
    slab depressurizations were apparent over the 3 to 4 years
    after installation.

 •  Equipment durability. One of the five 90-watt DTD/
    remote discharge fans (in House 15) failed over the 4
    years since installation. This fan failed due to bearing
    failure, the only fan in Pennsylvania to fail for a reason
    other than capacitor failure. This failure rate (one in five)
    is  roughly the same as the rate  for fans on the  SSD
    systems.

 •  Homeowner intervention. In one case (House 15), the
    homeowner turned the fan off due to the noise created by
    the worn bearings.

2.3.4 Active Block-Wall Depressurization

BWD has occasionally been used as a supplement to SSD in
basement houses, in both R&D study houses and in commer-
cial installations, when one or more of the block foundation
walls had appeared to be an important radon entry route which
has not been adequately treated by SSD alone. The BWD
component in these combined SSD+BWD systems is usually
accomplished using the "individual pipe" approach, i.e., by
inserting a PVC pipe into one or more of the block cavities in
the wall(s) to be treated, and tying this pipe  into the  SSD
system piping.

In the large majority of cases, SSD alone will be sufficient,
and the BWD component will be unnecessary. However, in
                                                       44

-------
some parts of the country—in particular where basement sub-
slab communication is poor or where the source strength is
high—mitigators are commonly prepared to supplement SSD
systems with BWD components in block-foundation houses,
where high  radon  concentrations inside the walls  or other
local conditions suggest that a BWD leg into one or more of
the walls might be necessary (Mes90b, Sh9Q, Jo91).

BWD as a stand-alone method, without SSD, has been tested
in a few R&D study houses having basements. It does not
appear to have been  widely  used commercially, although
mitigators have occasionally encountered isolated cases where
one or two  block  walls were particularly important radon
entry routes and where a stand-alone BWD system treating
those walls proved to be extremely effective. One firm mar-
kets a "baseboard-duct" BWD system nationally through lic-
ensees (E188),  sometimes used in conjunction with basement
water control.  But even this BWD system often has a SSD
component;  the baseboard ducts are often connected to  a
retrofitted, perforated sump crock installed through the base-
ment slab to handle water collected in the baseboard ducting,
and the suction is drawn on this sump (providing SSD) as well
as on the baseboard channel drain.

From the limited available data, BWD as a stand-alone method
can be effective in basement houses suited to this approach.
Well-suited houses include houses which permit good closure
of all major wall openings and which do not ha.ve major slab-
related entry routes remote from the walls. Usually, stand-
alone BWD  will be considered primarily in houses where the
walls are the predominant entry route, and where marginal or
poor sub-slab communication will prevent a SSD system from
adequately treating the walls.

Stand-alone BWD has also been made to perform reasonably
well in  less suitable houses, though this has often required
some effort  to adequately  close wall openings and to boost
suction in the walls (e.g., with multiple fans or more suction
pipes). Moreover, the ability to reliably achieve indoor levels
below 4 pCi/L with stand-alone BWD systems in such less
amenable houses has not been consistently demonstrated.
EPA's experience has suggested than one cannot always
reliably predict which houses will be truly suitable, nor how
much effort will be required to make the stand-alone BWD
system give the desired reductions. In addition, due to the
amount of house air that can be drawn into the system through
the walls, the house heating/cooling penalty and the threat of
combustion appliance backdrafting are likely to be increased,
especially in houses less amenable to the BWD approach.

Results from stand-alone BWD systems  have been reported
only for seven basement houses in Pennsylvania (He87, Sc88),
one basement house having an adjoining slab on grade in New
Jersey (Tu89),  and three basement houses in New York (Ni89).
(The Pennsylvania study also included five other basement
nouses having block-wall pressurization systems,  discussed
in Section 2.4.)

Of the seven houses in Pennsylvania, six (Houses 3,7,8,14,
16, and 19) involved the individual-pipe variation of the BWD
process. Li these five, there were one or more pipes installed
in each of the perimeter walls, with House 16 having nine
pipes. Among these six houses having stand-alone individual-
pipe BWD systems,

   -  All six of the houses except House 19 achieved radon
      reductions of 92 to 99% in the basement. The percent-
      age reductions are high because the pre-mitigation
      levels were extremely elevated, as great as 400 pCi/L.
      Thus, despite the high  percentages, only three of the
      houses were reduced below 4 pCi/L (with two of these
      three being reduced to 2 pCi/L and less, and one to 1
      pCi/L).

   -  Four of these houses (Houses 3, 8, 14, and 19) were
      particularly suitable for BWD: the walls were unfin-
      ished and accessible, the top block voids were gener-
      ally accessible for closure, and there were no complica-
      tions such as a block fireplace structure in the walls,
      exterior veneer, or high block porosity. Houses 3, 8,
      and 14 were reduced below 4 pCi/L with a single fan
      depressurizing the system; Houses 3 and  14 were re-
      duced below 2 pCi/L. These low post-mitigation levels
      have been confirmed by winter-quarter alpha-track mea-
      surements over the 3 years following installation. These
      three houses, which were apparent beforehand as being
      distinctly amenable to the BWD approach, are the only
      Pennsylvania houses where BWD has been unambigu-
      ously successful.

   -  The fourth "amenable" house above (House 19) has
      consistently averaged 31 pCi/L in the basement since
      installation (no reduction from the pre-mitigation value
      of 32 pCi/L) (Fi91). ALL major openings in the walls
      had been effectively closed, and smoke tracer testing
      confirmed that the single fan was depressurizing the
      wall voids every-where. The slab in this house is badly
      cracked, and diagnostic testing confirmed that soil gas
      is entering the house through these cracks.  Thus, the
      BWD system would appear to be effectively treating
      the walls, but the suction field is not extending beneath
      the slab to treat the slab-related routes remote from the
      walls. As a result, the ability to identify a priori houses
      which will be truly amenable to stand-alone BWD is in
      question. As a minimum, the absence of potentially
      important slab-related  entry routes remote froni the
      walls would appear to be one of the criteria for deter-
      mining amenability. The walls (and the wall/floor joint)
      must be the primary (or sole) entry routes.

      As a point of interest, the BWD system in House 19—
      so unsuccessful in reducing basement concentrations—
      is extremely effective in reducing levels upstairs, to
      below  1 pCi/L (Sc89, Fi91). The  BWD  system is
      thought to be depressurizing the basement, preventing
      the flow of the 31 pCi/L basement air upstairs.

   -  Two of the six Pennsylvania houses having stand-alone
      individual-pipe BWD systems (Houses 7  and 16) of-
      fered difficulties in wall closure. These difficulties
      included inaccessible top voids, and exterior brick ve-
      neer.  While radon reductions of 92 and 98% were
      achieved in these houses, respectively, basement levels
      remained elevated (32 pCi/L in House 7 according to
                                                        .45

-------
      short-term measurements, 5.3 pCi/L in House 16 based
      upon three winters of alpha-track measurements). To
      achieve these reductions, various  special steps were
      taken in system design (using two  suction fans in
      House 7, using 6-in. diameter piping for the major
      piping run in House 16 to reduce pressure loss). Had
      the pre-mitigation concentrations in these two houses
      been lower, the percentage reductions would have been
      lower, but it is possible that post-mitigation levels
      might have gotten below 4 pCi/L. These results indi-
      cate that in houses not amenable to stand-alone BWD,
      good reductions can sometimes still be achieved, but
      special steps may be required, and the ability to reduce
      levels below 4 pCi/L is uncertain. In these cases, stand-
      alone BWD is not the best choice.

      In House 16, the nine-pipe BWD  system was subse-
      quently temporarily replaced with a three-pipe SSD
      system, to determine which would perform better (Fi91).
      The two SSD pipes in the basement very effectively
      depressurized the entire basement slab (to 0.323-0.363
      in. WG), although the one SSD pipe in the adjoining
      paved  crawl-space addition achieved more marginal
      sub-slab depressurizations in  that wing (0.001-0.020
      in. WG). Even with apparently excellent treatment of
      the basement slab, the SSD system  alone resulted in
      distinctly higher indoor radon levels than did the BWD
      system alone (13 pCi/L in the basement, compared to 4
      pCi/L with the BWD system, and 5 pCi/L upstairs,
      compared to 2 pCi/L with the BWD system). While the
      marginal treatment of the paved crawl space by the
      SSD system might have been contributing to the re-
      duced performance of that system, these results suggest
      there  can  be cases where SSD systems cannot ad-
      equately treat all of the wall-related entry routes, even
      when the system is depressurizing the sub-slab very
      well. Thus, there will be cases where the SSD system
      must be supplemented with a BWD system. Possibly,
      placement of additional  SSD pipes near the  walls in
      House 16 could have improved the wall treatment by
      the SSD system. However, House 16 would appear to
      be a case where a combination of SSD + BWD could
      be the optimum approach for reliably reducing the
      house below 4 pCi/L.

The one house tested with a stand-alone BWD system in New
Jersey (Tu89) was an  individual-pipe system in a basement
house having an adjoining slab on grade (House LBL-10).
Two BWD suction pipes were installed in the wall voids of
the stem wall separating the two wings. This relatively simple
BWD system reduced basement concentrations from 146 pCi/
L to 3 pCi/L. By comparison, a stand-alone SSD system tested
in this house (with two SSD pipes extending all the way
through the stem wall to beneath the adjoining slab on grade)
gave about the same  results. The success of the two-pipe
BWD system in New Jersey is very different from the experi-
ence with the more extensive systems in Pennsylvania. The
difference might result from the stem wall being a particularly
important source in the New Jersey house, and/or  to some
depressurization beneath the adjoining slab being obtained by
the BWD system.
The  three New York houses (AR-01, OP-01, and OP-16)
having stand-alone BWD systems were all individual-pipe
installations in houses having marginal sub-slab communica-
tion  (Ni89). In the two houses which either had solid cap
blocks (AR-01) or where a substantial effort was  undertaken
to close the voids in the top course of block (OP-16), suction
on one to three walls reduced radon levels to about 2 pCi/L
(from pre-mitigation levels of 17 to 55 pCi/L). BWD had been
selected for these houses because higher radon levels had
been measured inside the block cavities than beneath the slab,
suggesting that wall entry was important.

Even in the third New York house (OP-01), where no wall
sealing was performed and where suction was drawn only on
one wall, levels were reportedly reduced from 21 to about 3
pCi/L. In House OP-01, this one-wall BWD system gave
lower post-mitigation readings than did a one-pipe SSD sys-
tem which  was tested back-to-back  with the BWD system.
The SSD system alone gave readings of 5 to 7 pCi/L, com-
pared to the 3 pCi/L for the BWD system alone. The apparent
success of the BWD system in House OP-01 was achieved
despite several factors which would suggest that it would not
be amenable to BWD as a stand-alone method: unscalable top
block voids; the ability to treat only one of the walls; high sub-
slab  radon  concentrations; and an extensively cracked slab,
offering entry routes remote from the foundation walls. This
success in OP-01 may have been due, at least in part, to the
following factors (Br92):

   -  the block foundation walls extended all the way up to
      the attic. Thus, even though the top voids (in the attic)
      could not be sealed, there would be an increased resis-
      tance to air flow from the top voids down into the
      BWD  system, increasing the suctions that  could be
      maintained inside the cavities at basement level.

   -  three suction pipes were installed in the one wall, with
      the result that suction is  thought to have  extended
      around the  corners into the adjoining walls. Thus, at
      least parts of two other walls were also being treated.
      Perhaps these three walls were the major entry routes.

   -  the post-mitigation level was probably depressed by
      the fact that this measurement was made during mild
      weather. Cold-weather post-mitigation results were not
      reported.

One of the seven Pennsylvania houses having a stand-alone
BWD system (House 11) received the baseboard-duct varia-
tion  of the technology (He87, Sc88). This particular house
was in fact one (end) unit in a multi-unit townhouse structure,
with a perimeter channel drain around the entire  structure.
Only moderate radon reductions (about 65%, to 21 pCi/L)
were achieved. The testing in this one unit could not fairly
represent the results  that might be  achieved in  a detached
house  where the full perimeter channel drain could be ac-
cessed; in addition, the owners of this house discontinued
participation in the project before testing could be completed.
Thus, the results from this townhouse are not felt to be a fair
representation of the performance that might be expected with
a stand-alone baseboard-duct system in a detached house.
                                                        46

-------
 BWD (generally the individual-pipe variation) has been tested
 in a number of cases as a supplement to a SSD system. The
 benefits of adding a BWD component to a SSD system will
 depend upon the significance of the walls as an entry routes,
 and the ability of the SSD system by itself to address wall-
 related entry. A BWD component is most commonly needed
 when sub-slab communication is poor,  hindering the exten-
 sion of the SSD suction field around the base of the walls.

 In three houses in Pennsylvania, SSD  systems were tested
 with and without a BWD component. In each case, the BWD
 component consisted of one or more individual pipes into
 each of the perimeter walls, connected  into the SSD piping
 and treated by a single fan. In one house with moderate to
 good communication (House 3), a one-pipe SSD system de-
 pressurized the sub-slab to 0.024-0.093 in. WG, and was
 adequate to reduce basement levels to about 3 pCi/L; a
 supplemental five-pipe BWD component further reduced lev-
 els below 2 pCi/L (Fi91). In the second house,  House 20,
 which had poor communication (He87, Sc88), a BWD com-
 ponent was necessary in order to reduce basement concentra-
 tions to near 4 pCi/L; the SSD system by itself was marginally
 inadequate, achieving 5 to 8 pCi/L in the basement. In this
 house, the BWD component consisted of one 2-in. diameter
 (equivalent) wall suction pipe tapped into each of the five 4-
 in. SSD pipes, which were immediately beside the walls. The
 BWD pipes caused suction in the SSD pipes to fall from about
 0.9 in. WG to about 0.2 in. WG. (Well water treatment was
 also necessary in House 20, as discussed in Fi91.)

 But in the third house in Pennsylvania (House 7), with vari-
 able communication (ranging from low to good), addition of a
 BWD component caused performance to  degrade signifi-
 cantly (increasing basement concentrations from about 4 to 26
 pCi/L). Analogous to House 20, the BWD component in
 House 7 was achieved by tapping one 2-in. (equivalent) wall
 suction pipe into each of the seven 4-in. SSD pipes.  As in
 House 20, the air flow out of the BWD component of the
 system in House 7 caused the suction  in the  SSD pipes to
 decrease significantly, from about 0.9 in. WG to about 0.1 in.
 WG. But in House 7, the reduced suction in the SSD piping
 resulted in a significant reduction in the performance of the
 SSD component of the system which more than offset any
 benefits from the BWD treatment of the walls.

 In one basement house in New Jersey  having an adjoining
 paved crawl space and having poor communication (LBL-12
 in Tu89), SSD in the basement reduced levels only to 5 pCi/L.
 Addition of a suction pipe into the cavities of the block wall
 separating  the two  wings,  and  connecting it to the SSD
 system, reduced levels to 2.3 pCi/L.

 In testing in New York (Ni89), BWD was tested in conjunc-
 tion with SSD in four houses. One of these houses (OP-01), a
 full basement house, has been discussed earlier in connection
 with testing of BWD alone and SSD alone. In House OP-01,
 combined operation of the one-pipe SSD system  and the
 BWD system on one wall reduced basement levels from 21
^-pCi/L to about 3 pCi/L during cold weather. By comparison,
 SSD alone had achieved only 5-7 pCi/L;  BWD alone had
 achieved 3 pCi/L, but that had been during  mild weather.
The other three New York houses had been basement houses
having adjoining slabs on grade (AR-04, -05, and -09), report-
edly with relatively good communication beneath the base-
ment slab. Grab radon measurements inside the block cavities
were in all cases comparable to, or higher than, the radon
measured beneath the basement slab, indicating the impor-
tance of wall-related entry. In each case, a SSD system in the
basement was supplemented by a BWD leg treating the stem
wall between the two wings. In all three houses, adding the
BWD component made a significant improvement compared
to the SSD component alone. In two of the cases, the BWD
component was required to reduce basement concentrations
below 4 pCi/L (achieving levels of about 2 pCi/L).

In these three New York houses (Ni89), and in New Jersey
House LBL-12 discussed previously (Tu89), the depressur-
ization of the stem  wall may have also been providing some
depressurization beneath the adjoining slab on grade. To that
extent, these results would be generally consistent with results
observed by others (Mes90a), indicating that, in basement
houses having  adjoining slabs, SSD beneath the adjoining
slab is sometimes required in addition to SSD in the basement.
Where a BWD component is being added to a basement SSD
system,  the stem  wall separating  the basement from any
adjoining wing is commonly the first wall to which suction is
applied; it will often contain the highest radon levels, and, as
stated above, suction on this stem wall may also be treating
the adjoining wing.

In three basement houses in Tennessee having adjoining crawl
spaces (Py90), with poor communication beneath the base-
ment slab, a BWD component (along with sealing of the top
block voids and, in one case, the surface of the highly-porous
block wall) was required, in addition to  two-pipe, two-fan
SSD systems, in order to reduce levels to 4 pCi/L. In the two
houses where crawl-space soil was exposed, a SMD compo-
nent was required also.

The  preceding discussion has focused on BWD (and SSD +
BWD) systems  in basement houses, since almost all of the
BWD data have been obtained on that  substructure type.
BWD components to SSD systems have been tested in three
cases in slab-on-grade houses.

In one slab-on-grade house in Ohio having a block foundation
wall beneath the slab perimeter (House 1 in He91a), suction
was  drawn on a hole cored horizontally all the way through
the foundation wall from outdoors, below slab level. The 90-
watt fan was mounted directly over this hole, with no pipe
extending through the wall, and with no attempt to close the
block cores surrounding the hole. Under these circumstances,
there should be a significant  BWD component  to this
SSD+BWD system. This  combined system reduced indoor
levels from about 20 to about 3 pCi/L. By comparison, when
the fan was remounted on a pipe inserted through the wall,
and  the surrounding voids foamed in an effort to reduce the
BWD component, levels rose to about 8 pCi/L. Levels fell
again when the pipe and foam were removed, re-establishing
the BWD component. Thus, the BWD component can clearly
be important in some cases. Based upon results from some of
the other slab-on-grade houses tested in the Ohio project, the
BWD component appeared  to be most important where the
                                                        47

-------
SSD system was marginal (i.e., in large slabs having only one
suction point).

In two other slab-on-grade houses in New Jersey (Os89a), a
combination of SSD + BWD was tested, again using suction
pipes penetrating horizontally through the block foundation
wall from outdoors. In each of these houses, there were three
suction pipes manifolded to a single fan, with one pipe pen-
etrating all the way through the wall into the sub-slab region,
and the other two pipes terminating inside the block cavities.
Radon reductions of 99% were obtained in both of these
houses, reducing indoor levels from pre-mitigation values of
700-1,000 pCi/L, to post-mitigation  concentrations of 6-8
pCi/L. (The effects of deleting the BWD component were not
tested.)

The  importance of the BWD component in these slab-on-
grade houses probably resulted  from  the treatment of wall-
rclatod soil gas entry, and from extension of a sub-slab suction
field through the relatively permeable backfill material around
the perimeter immediately inside the footings. As discussed in
Section 2.3.1c, in  connection with perimeter  placement  of
SSD pipes in Florida slab-on-grade houses (Fo90, Fo92), air
Short-circuiting into the SSD+BWD system through the po-
rous blocks might sometimes be expected to degrade system
performance. In  the Ohio house discussed above, the
SSD+BWD system was subject to being overwhelmed when
the forced-air furnace fan (with sub-slab supply ducts) was
operated continuously. Thus, reliable guidance cannot cur-
rently  be offered regarding  the conditions under which a
BWD component can best be added to a SSD system in slab-
on-grade houses.

The following discussion summarizes the results to date with
BWD, addressing the individual variables.

a)  House design variables

 •  Sub-slab communication. A stand-alone BWD sys-
    tem  (or a BWD component on a SSD system) appears
    most applicable when sub-slab communication is poor
    (Ni89, Tu89, Py90, Sh90, Fi91, Jo91). This is undoubt-
    edly because the poor sub-slab communication prevents
    the SSD system from effectively intercepting the soil gas
    before it enters the wall void network.

    However, where the walls are major entry routes, BWD
    components can occasionally still be important even when
    sub-slab communication is good. The results from Houses
    3 and 16 in Pennsylvania (Fi91) demonstrate that, even
    when communication is  good and a SSD system can
    effectively depressurize the entire sub-slab, a BWD com-
    ponent can still provide improved performance. In the
    case of House 16, a stand-alone BWD system provided
    better reductions than a stand-alone SSD system. The
    unusually high source strengths under these houses may
    be the explanation why effective sub-slab depressuriza-
    tion by the SSD system was unable to adequately reduce
    radon entry through the walls.

    On the other hand, where  sub-slab communication  is
    marginal, adding a multi-pipe BWD component to a SSD
system can  sometimes actually reduce overall system
performance. In House 7 in Pennsylvania, high suctions
were needed in the SSD system piping in order to achieve
adequate suction field extension beneath the slab; addi-
tion of the BWD pipes resulted in high air flows from the
walls which reduced the system suction and degraded the
overall performance.

Access to close major wall openings. Best results
with BWD systems have consistently been achieved in
cases where major wall openings either have not existed
or have been accessible for closure (He87, Sc88, Ni89,
Py90). It is particularly important that the block voids at
the top of the wall be closed, either with a solid cap block
installed during construction, or through a sealing effort
during mitigation. When major wall openings exist and
cannot be effectively closed, there is an increased likeli-
hood had poorer BWD performance will be achieved
(e.g., Houses 7 and 16 in Sc88). One house where good
BWD performance was achieved despite the inability to
close the top block voids (House OP-01  in Ni89) was
atypical; the block walls extended up into the attic, creat-
ing increased flow resistance for air flowing down through
the top voids, thus increasing  the suction that the BWD
system was able to maintain in the cavities at basement
level.

Entry routes in center of slab. The data from House
19 in Pennsylvania (He87, Sc88) suggest that slab-related
entry routes remote from the walls (such as extensive slab
cracking) can sometimes make it impossible for a stand-
alone BWD system to adequately treat the house, necessi-
tating a SSD component. However, occasional cases have
been reported where a stand-alone system gave adequate
performance despite extensive slab cracking (House OP-
01  in Ni89). The ability of  stand-alone  BWD to ad-
equately treat such houses will depend upon a) whether
the perimeter walls are  the  predominant entry route,
despite the slab  cracks; and/or b) the ability of BWD
suction to extend under  the  slab to treat the interior
cracks.

Stand-alone BWD systems have been found to create
measurable (but often marginal) depressurizations be-
neath the slab; depressurizations ranging from 0.001 to
0.012 in. WG have been measured under the central slab
in Houses 8, 14, and 16 in Pennsylvania (Fi91). Under
these conditions, some ability to treat interior slab entry
routes would be anticipated. Nevertheless, it is recom-
mended that SSD components always be  considered in
conjunction with BWD systems, especially when there
are interior slab-related entry routes.

Nature of the foundation wall. BWD is applicable, of
course, only with block foundation  walls. Best perfor-
mance has generally been observed where the walls: a)
have solid cap blocks as the top course, or have top void
accessible for closure; b) walls not containing complicat-
ing entry routes such as block fireplace structures; and c)
are not unusually porous.
                                                       48

-------
House size. The data are insufficient to enable a defini-
tive statement regarding the effect of house size. Among
the houses where stand-alone BWD systems have given
the best results, many  have had relatively small slabs
(570 to 860 ft2), but one (House 14 in Pennsylvania) had
a slab of 1,300 ft2. And among the houses where stand-
alone BWD systems have given poorer results, one had a
small slab (570 ft2), one a larger slab (1,100 ft2). Intu-
itively, houses having  longer  perimeters are likely to
require more BWD suction pipes with stand-alone sys-
tems; for example, House 16, with 1,040 ft2 of slab area,
received nine BWD pipes, and is still not reliably below 4
pCi/L in the basement.

There is no evidence that slab size plays a role in deter-
mining whether a  BWD  leg must be added to a SSD
system, unless perhaps the increased slab size is com-
bined with poor communication.

Adjoining wings. In a total of four basement houses
having adjoining slab-on-grade  or paved crawl-space
wings (Ni89, Tu89), depressurization of the block stem
wall separating  the wings  proved to be necessary in
conjunction with SSD  in the basement. This could be
suggesting that, with adjoining wings, the stem wall can
be a particularly important entry route, perhaps because
the soil adjoining the basement beside that wall is "capped"
with a slab that helps direct the soil gas toward that wall.

SSD beneath an adjoining slab on grade or paved crawl
space, as a supplement to SSD beneath the basement slab,
has sometimes been found to be necessary to effectively
treat such combined-substructure houses (Sc88, Mes90a).
The BWD treatment of the stem wall in the four houses in
Ni89 and Tu89 may well have also been providing a SSD
component under the adjoining slab. And likewise, the
SSD treatment of the adjoining wings in Sc88 and Mes90a
was likely also providing a BWD  component treating the
stem wall. In all cases in Sc88  and Mes90a, the SSD
pipes treating the adjoining slab  were inserted horizon-
tally through the stem wall from inside the basement;
hence, the suction  beneath the adjoining slab was being
generated immediately beside the stem wall.

In summary, it is  likely that, in  combined-substructure
houses with block foundations, the radon entry associated
with the adjoining wing is a combination of a) entry into
the basement through the block stem wall; and b) entry
directly into the adjoining wing  through  routes in that
wing. The relative importance of these two pathways will
likely  vary depending upon site-specific factors. The
stem-wall BWD components added in Ni89 and Tu89,
and  the adjoining-slab SSD legs  added in Sc88  and
Mes90a, were both likely providing a combination of
SSD + BWD treating both pathways. The adjoining-slab
SSD approach, where  the pipes penetrate all the way
through the wall into the region beneath the adjoining
slab, would likely be better at treating the entry directly
into the adjoining wing.
    The data base for stand-alone BWD systems in com-
    bined-substructure houses is inadequate to permit an as-
    sessment regarding the effect of an adjoining wing on the
    design and performance of such systems. BWD  as a
    stand-alone technique has been reported in only one
    house having an adjoining wing (House  16 in Sc88,
    where three individual BWD pipes in an adjoining paved
    crawl space supplemented six BWD pipes in the base-
    ment). In this case, the treatment of the walls in  both
    wings was  inadequate to reliably reduce the basement
    from 395 pCi/L to below 4 pCi/L (achieving 5.3 pCi/L
    according to  winter-quarter alpha-track averages). The
    living area was reduced below 2 pCi/L, according to the
    winter-quarter alpha-tracks. It is suspected that residual
    radon levels would have been higher if BWD pipes had
    not been extended into the walls of the adjoining crawl
    space. Poor communication beneath the crawl-space slab
    was probably preventing the BWD pipe in the basement
    stem wall from establishing much of a SSD component
    beneath that adjoining  slab. However, testing was not
    conducted with and without the BWD pipes in the walls
    of the adjoining wing; thus, it is not possible to quantify
    how important it was to treat the adjoining wing in that
    way, nor to understand the complications being  created
    by the adjoining wing in this one house.

b)  House operating variables

 •  Operation of central furnace fan,  and exhaust
    fans. As discussed in Section 2.3. Ib, central furnace fans
    (with cold air returns in the basement), and various other
    exhaust fans (including whole-house exhaust fans, attic
    fans, and clothes driers), have been observed to cause
    basement depressurizations of 0.001 to 0.02 in. WG. No
    quantitative measurements have been reported  for the
    depressurizations created by BWD systems inside block
    walls. However, pressure measurements in  the BWD
    piping are generally low, 0.01 to 0.15 in. WG  (Sc88),
    one-tenth or less of what is commonly measured  in SSD
    pipes. These low suctions in the piping suggest that the
    depressurizations inside the wall are probably quite low,
    and perhaps not measurable in some locations. Stand-
    alone BWD systems have been reported to create sub-
    slab  depressurizations ranging from 0.001 to 0.012  in.
    WG in  three Pennsylvania houses having moderate to
    good sub-slab communication (Fi91); these depressuriza-
    tions are much lower than are commonly measured with
    SSD systems when communication is good.

    Based upon the above  discussion, it is expected that
    stand-alone BWD systems will be much more  subject
    than SSD systems to being overwhelmed by the basement
    depressurizations created by appliances.

c)  Mitigation system design variables

 •  Method of distributing suction to walls (individual-
    pipe vs. baseboard-duct approach). Insufficient data are
    available to permit a meaningful comparison of the rela-
    tive performance of these two  approaches.  Most  pub-
    lished data address the individual-pipe approach. In no
                                                    49

-------
cases have both approaches been tested back-to-back in a
single house, to compare performance.

Number of suction pipes. In all six of the Pennsylvania
houses where the individual-pipe BWD approach has
been tested as a stand-alone measure, one or more BWD
pipes were installed in each perimeter foundation wall
and in each interior load-bearing  block wall that pen-
etrated the slab. Where there was a  discontinuity in  a
wall, dividing the wall into two  sections, a pipe was
installed in each section. As indicated previously, this
complete coverage of the walls was adequate to reduce
the three most amenable houses below 2 to 4 pCi/L in the
basement during the winter, but was inadequate to reduce
the less amenable houses below 4 pCi/L in the basement.
No testing was conducted to reduce the number of pipes
in the amenable houses, to determine whether fewer pipes
would have been sufficient in those  cases. Qualitative
(smoke tracer) measurements suggested that suction on
one perimeter wall would sometimes  extend around the
comer into the adjoining  perimeter wall, but would not
consistently do so.

The one house in New Jersey having a stand-alone indi-
vidual-pipe BWD  system (Tu89) achieved substantial
reduction (from 146 to 3 pCi/L in the basement) with two
BWD pipes in only one of the basement walls (the stem
wall beside the adjoining slab on grade).

Of the two amenable houses in New York having stand-
alone BWD systems (Ni89), reductions to about 2 pCi/L
in the basement were achieved with BWD pipes in only
one to three of the perimeter walls. The one less amenable
house (OP-01) appears to have been significantly reduced
with three suction pipes in one wall, although there are
not cold-weather data with only the BWD system operat-
ing to confirm this favorable result.

In summary, with stand-alone individual-pipe BWD sys-
tems, there are some limited data (from New Jersey and
New York) suggesting that installation of a suction pipe
into each wall is not always necessary, especially when
the house is amenable to BWD. This result is consistent
with isolated experiences of commercial mitigators. This
situation will occasionally occur when one or two walls
are the predominant entry routes, perhaps due to a high
source strength near those walls. However, there are
other limited data (from Pennsylvania) suggesting that,
when a house is not amenable to BWD, one pipe in each
wall will not be sufficient Given  the limitations of the
data, if a stand-alone individual-pipe BWD  system  is
being considered for a given house, it would  appear
logical during the planning process to anticipate that one
pipe will be required on each perimeter and interior load-
bearing wall, unless or until evidence becomes available
indicating that fewer pipes will be sufficient

No data  are available from stand-alone baseboard-duct
BWD systems, indicating whether it will sometimes be
sufficient to install baseboard ducts on only some of the
walls, or along only a portion of a wall. Presumably, the
baseboard-duct approach would be at least comparable to
the individual-pipe approach, in terms of its ability to be
successful when fewer than all four walls are directly
treated.

Location of suction pipes. No definitive data  are
available regarding the optimum point at which to insert
an individual BWD suction pipe into a block wall. Com-
monly, pipes have been installed roughly  midway  be-
tween the two ends of the wall, although  obstructions
have often required that the pipes be installed off this
mid-point Where multiple pipes have been installed in a
single wall, the pipes have been installed at roughly equal
distances from the two ends of the wall and from each
other. In the Pennsylvania houses (Sc88),  BWD pipes
were generally installed near the bottom of the wall, in an
effort to improve the treatment of the wall/floor joint and
the footing region, and to improve extension of the suc-
tion field under the slab.

Size of suction pipes. Where BWD has been used as a
stand-alone technique, flows have been high, due to  the
leakiness of the walls. Under these conditions, piping of
at least 4 in. diameter has essentially always been used, to
reduce pressure losses in the piping. In  some cases in
Pennsylvania, to further reduce pressure losses, the 4-in.
diameter legs into the walls were connected  to a 6-in.
diameter trunk line leading  to the exhaust fan. Since
exhaust flows with the 90-watt centrifugal fans were
sometimes as great as 150 to 200 cfm with these systems,
piping pressure losses were sometimes as great as 0.25 in.
WG. But in none of the Pennsylvania cases did pressure
losses between the fan and the wall entry points appear to
be a primary cause of poor system performance.

Where a BWD component has been tapped into a SSD
system, and the combined piping directed to a single fan,
it has commonly been necessary to restrict flow into  the
BWD piping. Otherwise, system suction can be reduced
to such an extent that performance of the SSD component
deteriorates, offsetting any benefits adding of the BWD
component. This restriction of BWD flow can be accom-
plished by using 4-in. pipe for both the SSD and  the
BWD piping, and by installing dampers or valves in  the
BWD legs to allow the flow from those  pipes to be
reduced. Perhaps more commonly, this  restriction  has
instead been achieved by reducing the BWD legs to 1 or 2
in. in diameter.

As  discussed previously, tapping 2-in.  BWD suction
pipes into 4-in. diameter SSD pipes in  two  houses in
Pennsylvania (Houses 7  and 20 in Sc88) resulted in
comparable drops in the suctions in the SSD pipes in both
houses (from about 0.9 to about 0.1-0.2 in.  WG). How-
ever, the impacts on basement radon concentrations were
different in  the two houses. In House  20, the BWD
component improved performance, whereas in House 7,
performance degraded, despite effective depressurization
of the block walls. Apparently because of the specific
characteristics of House 7, a pipe diameter even smaller
than 2 in. would be warranted, if a BWD component is to
be helpful at all.
                                                   50

-------
    Commercial midgators also report using 1- to 2-in. diam-
    eter BWD suction pipes, when adding BWD components
    to SSD systems (Mes90b, K192). Some mitigators also
    use dampered or valved 4-in. pipe for the BWD legs,
    although other mitigators express concern regarding pos-
    sibly inadequate reliability of the dampers and high costs
    of the valves. The successful SSD + BWD installations
    that have been reported using 4-in. pipe for the BWD legs
    without dampers or valves, have been cases where the
    BWD component consisted of one or two pipes treating
    only the stem wall between a basement and an adjoining
    slab on grade (Ni89, Tu89).

 •  Degree of wall sealing. Best performance with stand-
    alone BWD systems has consistently been achieved in
    houses where the block wall had a top course of solid cap
    blocks, or where the open voids in the top course were
    accessible for effective closure. However, the actual ef-
    fect of closing the top voids has not been well quantified;
    in almost all cases, the  closure was completed before
    system performance measurements were made. In only
    one case have results been reported with and without the
    top voids closed (House OP-16 in Ni89). In that house,
    pre-mitigation concentrations in the basement (about 55
    pCi/L) fell  only to 23 pCi/L when a stand-alone indi-
    vidual-pipe BWD system was activated without closing
    the top voids (according to a 1-week continuous measure-
    ment made  in March). Closure of the top voids reduced
    basement levels to 2.3 pCi/L, based upon a 5-day con-
    tinuous measurement the following My. The difference
    in weather between the two measurements, as well as the
    closure of the top voids, could have contributed to the
    much lower reading obtained after closure.

    The  effects of  other types of wall closure  on BWD
    performance have not been well defined. In one Tennes-
    see house where a basement SSD  system was being
    supplemented by BWD on a block wall on the interior of
    a stone foundation wall (House DW43 in Py90), coating
    the porous face of the block wall with a surface bonding
    cement reduced basement radon levels  from about 18
    pCi/L to 4-5 pCi/L with  the system operating.  Such
    coating of the block surface is usually necessary only
    when the block is unusually porous.

d)  Mitigation system operating variables

 •  Fan capacity. Most of the testing of BWD systems, and
    of SSD+BWD systems, appears to have been conducted
    using the 90-watt, 270-cfm  centrifugal in-line tubular
    fans  at full power. In view of the generally high flows
    expected from BWD systems, use of smaller fans would
    not appear to be advisable. No data have been reported
    quantifying the effect of fan capacity on BWD perfor-
    mance.

 •  Fan in suction vs. pressure. Block-wall systems with
    the fans  operating to  pressurize the  wall  cavities, are
    discussed in Section 2.4.
e)  Geology/climate variables

 •  Source strength. High source strength could be contrib-
    uting to some of the apparent effects of other variables
    discussed earlier for stand-alone BWD systems. Both of
    the Pennsylvania houses which offered difficulties in
    wall closure and which were not consistently reduced
    below 4 pCi/L in the basement  (Houses 7 and 16) also
    likely had the highest source strengths of any  of the
    houses tested with  stand-alone  BWD systems.  As  an
    indicator that the source strengths were likely high, the
    pre-mitigation radon concentrations in these basements
    were 402 and 395 pCi/L, respectively, the highest among
    the tested houses. Thus, source strength, as well as diffi-
    culties in wall closure, could have been contributing to
    the lesser success in these two houses.

    On the other hand, the one house in New York (OP-01)
    where BWD successfully reduced levels below 4 pCi/L
    despite the fact that suction was drawn  on only one
    uncapped wall, also had one of the lowest source strengths.
    Cold-weather pre-mitigation radon levels in the basement
    were 21 pCi/L. The relatively lower source strength could
    have facilitated the apparent success of BWD  in this
    house.

    Intuitively consistent with the above results is the fact
    that both of the houses which were reduced below 2 pCi/
    L with stand-alone BWD systems (House 14 in Pennsyl-
    vania, House AR-01 in New York)  appeared to have
    relatively low source strengths (with pre-mitigation base-
    ment concentrations of 36 and 17 pCi/L), combined with
    the ability to effectively close the top voids. The remain-
    ing houses which were reduced  below 4 pCi/L, but not
    below 2 pCi/L, all had higher pre-mitigation levels (55 to
    350 pCi/L).

    However, it is noted that even houses which apparently
    have relatively high source strengths (with pre-mitigation
    basement concentrations of 150 to 350 pCi/L) have been
    reduced below 4 pCi/L with stand-alone BWD systems,
    where the house was amenable.  And conversely, in one
    house where the source strength appeared relatively lower
    (House 19 in Pennsylvania, with a pre-mitigation base-
    ment level of 32 pCi/L), BWD was unsuccessful. Thus,
    the source strength (or the pre-mitigation indoor concen-
    tration, as a surrogate for source strength) does not, by
    itself, suggest whether a stand-alone BWD system might
    be successful.

    A high source strength could be an important factor in
    determining whether a BWD component needs to  be
    added to a SSD system. Failure of a stand-alone SSD
    system to fully treat a block wall will be more  serious
    when the remaining soil gas which continues to enter the
    wall has extremely elevated radon concentrations.

  •  Permeability of underlying sott. There are no data
    relating the performance of BWD systems (or the need to
    add a BWD component to a SSD system) to soil  perme-
    ability. However, it might be anticipated that poor soil
    permeability could increase the  likelihood that a BWD
                                                        51

-------
    component might be needed on a SSD system, since this
    could reduce the ability of the SSD system to treat the
    exterior face of the block wall, or to intercept the soil gas
    before it reaches the foundation.

 •  Climatic  conditions.  Because of the relatively  low
    depressurizations created  inside the wall cavities  and
    beneath the slab by stand-alone BWD systems, it would
    be expected that BWD will be much more subject than
    SSD to being overwhelmed by weather-induced depres-
    surizations of the basement. The thermally induced de-
    pressurization of the basement during cold weather would
    be about 0.015 in. WG in a two-story house (Sau89). By
    comparison, the sub-slab  depressurizations maintained
    by stand-alone BWD systems have been measured rang-
    ing from 0.001 to 0.012 in. WG in houses having moder-
    ate to good sub-slab communication (Fi91). Depressur-
    izations inside the walls have not been reported, but they
    arc likely well below the suctions of 0.01 to 0.15 in. WG
    measured in BWD pipes near the point where they pen-
    etrate into  the wall.

f)   Mitigation system durability

 •  Radon reduction performance. Of the 38 EPA mitiga-
    tion  installations in eastern Pennsylvania which have
    been monitored since 1985-87 (Fi91), five have been
    stand-alone BWD systems. Winter-quarter alpha-track
    measurements in those houses over the 3 years following
    installation have remained very steady in all five houses.
    Only  in one house (House 19) did any one winter's
    reading vary from the three-winter average by more than
    ± 0.5 pCi/L. And even in House 19, where depressuriza-
    tion of the basement by basement air leakage into the
    BWD system apparently was increasing radon entry
    through slab cracks, basement concentrations remained
    remarkably steady over the years (31.3 ± 2.5 pCi/L).

 •  System suctions and flows. System suctions and flows
    remained steady over the 3 years since installation, in the
    five stand-alone BWD installations that have been moni-
    tored in Pennsylvania (Fi91).

 •  Equipment durability. There have been no fan failures
    among the five stand-alone BWD installations under the
    Pennsylvania project. Two of these five fans are the
    standard 90-watt centrifugal tubular units; the remaining
    three are 90-watt wall-mounted centrifugal units having
    comparable fan curves.

 •  Homeowner intervention. Two of the five BWD fans
    in Pennsylvania are known to have been turned off at one
    time or another over the 3 years following installation. In
    one case, the fan became unplugged accidentally; in the
    second, the owner turned the fan off during the summer
    when windows were commonly opened. This experience
    is consistent with that for other ASD techniques.
2.3.5  Active Sub-Membrane
Depressurization

Active SMD has consistently proven to be the most effective
approach for treating crawl-space houses. Radon reductions in
the living area of the house have commonly been reduced by
80 to 98% by SMD in EPA study houses. Lesser reductions
have been  observed in  some  cases, where pre-mitigation
levels were low, where there was an untreated adjoining wing,
or where a combination of factors prevented adequate distri-
bution of suction beneath the membrane (large crawl space,
very poor soil permeability, inadequate number  of  suction
pipes through the membrane, or  inadequate sealing of the
membrane). In houses where a basement or a slab-on-grade
living wing adjoins the crawl space, a SSD or DTD system
treating the adjoining wing is often advisable, in addition to
(or perhaps  instead of) the SMD component in the crawl
space.

SMD has occasionally been tested back-to-back against other
crawl-space treatment techniques: natural crawl-space venti-
lation (i.e., opening foundation vents); forced (fan-assisted)
crawl-space ventilation with the fan mounted to blow crawl-
space air outdoors (in an effort to achieve crawl-space depres-
surizati9n); and forced ventilation with the fan blowing out-
door air into the crawl space (in an effort to achieve crawl-
space pressurization). SMD has always provided distinctly
better reductions than have the other approaches (Fi90, Py90,
Py91, He92). The technique  which has generally  proven
second in effectiveness,  after SMD, has been crawl-space
depressurization. There is undoubtedly a crawl-space ventila-
tion/ depressurization component  to SMD  systems, as well,
since crawl-space air leaks into the SMD system through
openings in the membrane and is  exhausted. However, con-
centrating the ventilation/depressurization in the  region be-
neath the membrane appears to more effectively intercept the
radon and prevent its entry into the living area.

Based on EPA's mitigator survey (Ho91), commercial mitiga-
tors report using SMD over one-third of the time in crawl-
space houses. According to this survey, about one-third of the
crawl-space houses are treated using natural crawl-space ven-
tilation or forced ventilation (crawl-space pressurization or
depressurization), and the remaining  third are treated using
barriers  or sealing. From discussions with mitigators in re-
gions where crawl-space houses are relatively common (Sh91,
An92, How92,  K192), SMD is the technique of choice in
houses having an accessible crawl space with no adjoining
wing. Where there is an adjoining basement or slab-on-grade
wing which is being treated using SSD or DTD, the crawl
space wing may be treated using one of the ventilation ap-
proaches or sealing rather than SMD, although sometimes a
SMD component is added to supplement the SSD system.
Where the crawl space is inaccessible for installation of a
SMD system, a crawl-space ventilation approach is  usually
employed (Tu91c, He92, K192).

EPA has obtained results from testing SMD systems in 10
crawl-space houses not having adjoining wings (Ni89, Os89b,
H90, Py90, Du91). Pre-mitigation concentrations in the living
area typically ranged from 3 to 33 pCi/L, although one house
had a level of 160 pCi/L. Nine of these houses were reduced
                                                       52

-------
below 4 pCi/L, and six were reduced below 2 pCi/L. The
house having the pre-mitigation level of 160 pCi/L was among
those reduced below 2 pCi/L, although the SMD system had
to be supplemented with a BWD component and a well water
treatment system to achieve this reduction (Ni89). The appar-
ent performance in this latter house may be exaggerated, since
post-mitigation measurements made in April are being com-
pared against pre-mitigation measurements made the preced-
ing October through December. The one house not reduced
below 4 pCi/L had an unusually shaped crawl space (100 ft
long and 23 ft wide), and was reduced from 33 to 5 pCi/L in
the living area (Py90). In an effort to treat this long house with
highly impermeable soil, porous matting was placed beneath
portions of the membrane to aid in extension of the suction
field, two suction pipes penetrated the membrane, and the
membrane was sealed everywhere (Py90).

All of the houses except one had radon reductions of 80% or
greater in the living area. The house having a lesser percent-
age reduction (63%) experienced this relatively low percent-
age because of low pre-mitigation concentrations, 7 pCi/L
(Py90). A second house with low pre-mitigation levels (3 pCi/
L) might also have had living-area reductions below 80%, but
this is uncertain because post-mitigation levels in  the living
area were not reported (Os89b).

The installations in the 10 houses above vary in their designs.
In some cases, more than one suction pipe penetrated the
membrane. In some cases, the SMD suction pipe was drawing
suction on a length or a loop of perforated piping placed under
the membrane.  In some cases, the  membrane was sealed
everywhere (at seams between sheets, around the crawl-space
perimeter, around interior piers); in other cases, it was sealed
only in some locations, or nowhere.  In a  few cases, the
membrane did not cover the entire crawl-space floor. In two
cases, porous matting was placed beneath at least a portion of
the membrane to improve suction field distribution.  In a
number of  these cases, the system performance might have
been improved,by taking additional steps, such as increasing
the number of suction pipes or  increasing the membrane
sealing effort. Such additional steps might have increased the
number of the houses reduced below 2 pCi/L.

The pre- and post-mitigation measurements reported for the
10 houses represent a range of measurement methods and
durations, ranging from several-day measurements  with con-
tinuous  monitors or charcoal detectors, to 3- to  12-month
alpha-track detector measurements. Thus, some of these mea-
surements better represent the long-term performance of these
SMD systems than do others. However, it is clear that, over-
all, the SMD systems are being very effective.

In addition to the results from the  10 "pure"  crawl-space
houses (having no adjoining wings), EPA has also obtained
results from SMD systems in 14 houses where a basement or
slab-on-grade living wing adjoined the crawl space (Sc88,
Gi90, Mes90a, Py90, Du91). In 11 of the 14 houses, a base-
ment adjoined the crawl space; in all of these houses, the
SMD system was supplemented with a SSD or DTD system in
the basement. In the three houses where a slab  on grade
adjoined the crawl space, the slab-on-grade  wing was not
directly treated.
In all 11 houses where an adjoining basement was also being
treated, the combined SSD/DTD + SMD system achieved
radon reductions greater than 90%. All of die houses  were
reduced below 4 pCi/L, and 9 of the 11 were reduced below 2
pCi/L in the basement.

Among the three houses with an untreated adjoining slab-on-
grade living wing (Py90), two were reduced below 4 pCi/L,
and none were reduced below 2 pCi/L in the living area above
the crawl space. Both houses reduced below 4 pCi/L experi-
enced radon reductions of 80% or greater. The one having the
best results (92% reduction, to a post-mitigation level of 2.2
pCi/L) had no membrane at all, but was being treated by four
suction pipes buried in bare soil (i.e., what is referred to in this
document  as site  depressurization). The one  house not re-
duced  below 4 pCi/L achieved only moderate reductions
(from 28 to 15 pCi/L, a reduction of 46%). It is not clear why
the  SMD  system in this last house performed so poorly.
Perhaps the slab on grade was an important radon source,
although it was relatively small (a 300 ft2 converted garage)
compared  to the crawl space (900 ft2).

The experience of commercial mitigators with SMD systems
has been more limited than with SSD and DTD systems.
Houses having a crawl-space substructure represent a rela-
tively limited percentage of the housing stock (about 14% of
the  total new housing starts  in the U.S. between  1976 and
1983, according to the National Association of Home Build-
ers). Moreover, crawl-space houses may be  less prone to
having elevated radon levels in the living area. Mitigators
working in regions  having a relatively high  percentage of
crawl-space houses report that properly designed SMD sys-
tems are consistently very effective, commonly reducing in-
door radon levels below 2 pCi/L (Sh91, An92, How92, K192).

The following discussion summarizes  the results to date for
each of the variables that can influence system design, opera-
tion, and performance.

a)  House design variables

 •  Nature of crawl-space floor. Many crawl spaces have
    floors comprised of bare native soil. In some cases, this
    soil floor is covered with a layer of gravel, and/or with a
    plastic vapor barrier. Crawl spaces having a concrete slab
    as a floor, or an integral  unfinished concrete "wash"
    floor,  are usually considered for treatment using SSD,
    and are thus not considered in this discussion of SMD.

    Where the crawl-space floor is bare soil, the permeability
    in the native soil is often poor. Despite this, good perfor-
    mance of SMD systems has generally  been achieved
    without special provisions  to extend the suction field,
    unless the crawl space has been larger than about 1,500
    ft2. In those cases where permeability was poor and the
    crawl space was large, performance has been improved
    through the use of perforate4 piping under the membrane
    (Fi90, Du91, An92, K192), multiple  individual suction
    pipes  (Py90), or sub-membrane matting (Py90). As dis-
    cussed later, the effects of one vs. two suction pipes in
    such cases have been defined (Py90). Unfortunately, the
    systems have not been tested with and without perforated
                                                       53

-------
piping, or with and without matting, so that the exact
effects of those design selections are not clear.

In one house in Tennessee where the permeability of the
floor soU seemed relatively good from visual appearance
(House DW31 in Py90), a reduction of 92% was achieved
in the living area (to 2.2 pCi/L) with no membrane at all
(i.e., site depressurization). In that case, four suction
pipes were installed, drawing suction on covered pits in
the four quadrants of the crawl space. Testing of a simpler
site depressurization system in four crawl-space houses
in Ohio (Fi90), where the clay soil was impermeable, and
where the system consisted of a  single pipe embedded
about 10 ft deep into the soil outside the crawl space,
gave no measurable indoor reductions. Site depressuriza-
tion would be expected to work  best in cases where a
suction could be drawn  on a relatively permeable soil
layer capped by a relatively impermeable layer. House
DW31 in Tennessee is the only house where the site
depressurization approach has been successfully demon-
strated to date in the U.S.

Where the crawl-space floor is covered with gravel, it
would be expected that the gravel  would facilitate the
distribution of suction beneath the membrane with a
single suction pipe, analogous to the effect of gravel
beneath a basement slab. Gravel could also increase
flows beneath the membrane, facilitating the leakage of
crawl-space air into the system via any unsealed seams in
the membrane. This increased air flow may or may not
make complete sealing of the membrane more important
when gravel is present. Gravel was present on the floors
of 5 of the 24 EPA study houses discussed previously
(Fi90, Mes90a). The crawl spaces with gravel in Fi90
were relatively large, one as big as 2,700 ft?; however, the
SMD systems all used sub-membrane perforated piping,
so that any role of the gravel in enabling a single pipe to
treat such large houses was not determined. The crawl
spaces with gravel in Mes90a were small (300 ft2) wings
adjoining basements; such small wings would be ex-
pected to be effectively treated whether there were gravel
or not, and,  in addition, the effects of the basement
treatment masked  any role  of the  gravel. The crawl
spaces in a given project having gravel floors did not
produce  higher flows or lower  suctions in  the  SMD
piping than did those in the same project having floors of
bare soil.

In summary, it would be expected  that gravel could
improve  SMD performance, by improving suction field
extension. However, the available data are not sufficient
to permit determination of the effect of gravel floors.

In many  cases, a pre-existing vapor barrier will already
be down  on some portion of the crawl-space floor. Such
pre-existing barriers may not be sufficiently complete, or
sufficiently neatly deployed, to serve as the membrane
for a SMD system. However, such existing sheeting has
commonly been incorporated into the membrane for the
system, with  the pre-existing plastic being straightened
out and overlapped, and with new plastic being put down
over areas not already covered. Where the pre-existing
plastic is in good condition, it does not generally appear
that SMD performance is degraded by making use of the
existing plastic (Fi90).

Most of the crawl-space houses tested in the EPA projects
have had relatively flat, even floors, which facilitated
placement of the membrane sheets over the floor. More
irregular floors, such as those with rock outoroppings,
could impact the ability to cover the entire floor, or could
require sub-membrane matting (or a thicker membrane)
to reduce the risk of membrane punctures due to traffic in
the crawl space. Typically, with flat floors, 6- to 10-mil
thick polyethylene membranes (or thinner cross-lami-
nated membranes of equivalent thickness), without mat-
ting, have been used both in the EPA projects and com-
mercially.  More expensive 15- to 20-mil polyethylene
has been used in two of the EPA projects where floors
were flat but heavy traffic was expected (Os89b, Gi90);
reinforced resin industrial pit liner material has been used
in one study where extremely rugged floors were encoun-
tered in northern Alabama (Ma88). In one house in New
York having rock outcroppings in the crawl space (House
OP-05 in Ni89), matting was placed beneath the mem-
brane where it covered the rocks, apparently in part to
improve sub-membrane communication as well as to
provide puncture resistance.

Crawl space accessibility. In some cases, inaccessible
portions of the crawl space can apparently be left uncov-
ered by the membrane without  serious degradation in
SMD performance.

In one house in Ohio where the central portion of the
crawl space  was inaccessible due a forced-air air condi-
tioning unit, no attempt was made to cover the central
crawl space (House 24 in F190). (Pre-existing  vapor
barrier material was present in the central portion of this
slab, but it  did not cover the floor completely.) New
membrane for the SMD system was  extended out from
the perimeter wall for the width of one sheet (about 8 ft),
around the entire perimeter. Suction was drawn on a loop
of perforated piping placed around the perimeter, beneath
the membrane. Living area concentrations were reduced
from 16 to below 2 pCi/L with this system. Because the
perforated piping was immediately beside the perimeter
wall, it was felt necessary to carefully seal the perimeter
of the membrane to the wall, to reduce crawl-space air
leakage into the system. The results from this house show
that complete coverage of the floor is not always neces-
sary. However, in this case, the cost saving achieved by
avoiding coverage of the central area was partially offset
by the cost involved in sealing the membrane to the
perimeter wall, a step which might have been less neces-
sary had the suction been drawn toward the central  por-
tion of the slab.

In some cases, the crawl space is largely or completely
inaccessible. For example, some crawl spaces have  only
about a foot of headroom throughout ("suspended floors").
In such cases, SMD is not an option. Forced exhaust
ventilation of the crawl space (crawl-space depressuriza-
tion) will commonly be preferred in such cases; however,
                                                   54

-------
if backdrafting of combustion appliances is a concern,
crawl-space pressurization may be preferred (Tu91c).

Nature offoundation wall There are insufficient data
on SMD systems to enable an assessment of the effect of
block vs. poured concrete foundation walls. All but three
of EPA's 24 crawl-space study houses had block founda-
tions; the three having poured foundations did not achieve
better reductions than did those having block founda-
tions. Since all but 2 of the 24 houses were reduced below
4 pCi/L, and since 15 were reduced below 2 pCi/L, it
would appear that to the extent that the blocks may have
been contributing to radon entry, the SMD system was
largely addressing that entry route.

Some mitigators take steps to help ensure that a SMD
system will treat the block wall (Sh91, An92, K192).
Steps include, for example, extending the membrane up
the entire interior face of the wall inside the crawl space
(Sh91), or drilling holes through the interior block face
beneath the level at which the membrane is attached
(KHZ). There are no data  defining  the effectiveness of
these steps.

House size. In summary, the data are too limited, given
the number of variables being varied, to make a definitive
statement regarding the effect of house size on SMD
performance. However, it would appear that where perfo-
rated piping is placed beneath the membrane, or perhaps
if gravel is present on the crawl-space floor, the size of
the crawl space is not  critical.  With perforated  piping
and/or gravel, crawl spaces as large as 2,700 ft2 have
consistently been reduced below 2 pCi/L.

But where  perforated piping is  not placed beneath the
membrane, and where there is  no gravel on the floor,
SMD performance  seems  consistently to be reduced.
Levels still often appear to be reduced below 4 pCi/L
under these conditions, even with crawl-space floor areas
as large as 1,500 to 2,000 ft2.  However, without  sub-
membrane  perforated piping or gravel, there is  an in-
creased likelihood that multiple suction pipes through the
membrane will be required; and/or living-area concentra-
tions will be less effectively reduced (or may even remain
elevated).

Accordingly, with crawl spaces of 1,500 to 2,000 ft2
where  no gravel is present, and, in  some cases, in even
smaller crawl spaces, serious consideration  should be
given to installing perforated piping beneath the mem-
brane,  especially if it is desired to reduce living-area
concentrations to 2 pCi/L and less. Some mitigators have
reported often achieving levels below 2 pC3/L  with SMD
systems having one individual suction pipe and no sub-
membrane  perforated piping (Sh91, How92). However,
the data from EPA's relatively limited number of study
houses suggests that living-area levels below 2 pCi/L
may not be reliably achieved, even in relatively small
crawl spaces, unless perforated piping or gravel are present.

One EPA study house having a crawl-space floor area of
2,700 ft2 was reduced from 17 pCi/L to below 1-2 pCi/L
in the living area with a SMD system and no adjoining
basement wing (House 22 in Fi90). The good perfor-
mance in that particular house may be due in part to the
fact that there was gravel over the floor, and a perimeter
loop of perforated piping was installed, facilitating distri-
bution of the suction field beneath the membrane. In a
second house, having a crawl space of 2,050 ft2 (House
28 in Fi90), living area levels were also brought below 1-
2 pCi/L even though there was no gravel to aid in suction
field distribution. But again, two parallel lengths of
perforated piping were installed beneath the mem-
brane; in addition, the membrane was sealed every-
where,  and the pre-mitigation levels  were  only
slightly elevated (5 pCi/L).

All of the other 13 crawl-space study houses that were
reduced below 2 pCi/L in the living area were smaller
(with crawl-space floor areas ranging between 300 and
1,300 ft2). Each of these other houses also had one or
more other factors working in their favor: low pre-mitiga-
tion  concentrations (as low as 3 to 5 pCi/L in several
cases); an adjoining basement wing that was also being
treated with SSD or DTD (responsible for most of the
measured radon reductions); sub-membrane perforated
piping; gravel on the floor; or, in one case, simultaneous
BWD and well water treatment supplementing the SMD
system.

Among the 15 EPA study houses that were reduced to 2
pCi/L and less (out of the 24 houses total), only three
houses were reduced below 2 pCi/L in the living area by
SMD alone (without  SSD  in  an  adjoining basement)
which did not have sub-membrane perforated piping to
help distribute the suction field in the crawl space. One of
the three houses had a 300 ft2 crawl space and an adjoin-
ing basement, but obtained  levels below 2 pCi/L when
only the SMD leg of the mitigation system was operating
(House 582 in Mes90a). The second house had a 930 ft2
crawl space (Os89b). The third house (House OP-05 in
Ni89) had a 1,500 ft2 crawl space (Br92). Each of these
houses had other circumstances that may have aided in
achieving the low levels. The house with the 300 ft2 crawl
space, in addition to having only a small crawl  space
wing, had a gravel floor and a low pre-mitigation radon
concentration (4.3 pCi/L). The second house had low pre-
mitigation levels in the living area to begin with (3 pCi/L,
based on grab  samples), and did not  have good post-
mitigation measurements in the living area.  The third
house, House OP-05, included fibrous  matting  under
portions of the membrane over bedrock outcroppings, to
aid in suction field extension. Moreover, this house re-
quired BWD and water treatment components to be re-
duced below 2 pCi/L, and the low post-mitigation mea-
surement was made during mild weather (April).

The nine EPA study houses not reduced to 2 pCi/L or less
had  neither perforated piping nor gravel beneath the
membrane. Seven of these nine houses were reduced to
living-area concentrations below 4 pCi/L, though not
below 2 pCi/L; these seven houses had crawl-space floor
areas ranging from 300 to 2,000 ft2. Only one of these
seven houses had a crawl-space floor area greater than
                                                    55

-------
1,500 ft2 (House DW29 in Py90, with a floor area of
2,000 ft2 and a pre-mitigation level of 16 pCi/L). In this
house, a one-pipe system achieved only 7 pCi/L in the
living area, and a second pipe had to be installed to
reduce living-area levels to 3 pCi/L.

Of the two EPA study houses not reduced below 4 pCi/L
by SMD, one (House DW27 in Py90) had an unusually
elongated floor plan, with a crawl-space floor area of
2300 ft2, which likely contributed  to its poor perfor-
mance. Even a second individual suction pipe through the
membrane could not reduce levels below 5 pCi/L. The
other house (DW60 in Py90) had a crawl-space floor area
of only  900 ft2, but had an untreated adjoining slab-on-
grade wing. Again, neither had sub-membrane perforated
piping or gravel on the floor.

Adjoining  wings. Where the EPA study houses have
had basement wings adjoining the crawl spaces, treat-
ment of the basement (using SSD or DTD) has commonly
supplemented the crawl-space SMD system. In most
cases, the relative contributions of the SSD/DTD compo-
nent and the SMD component have not been isolated, so
that  it cannot be determined how necessary the SMD
component was (or how necessary the SSD/DTD compo-
nent was).

In three basement-plus-crawl-space houses in Maryland
where the individual contributions of the two components
were separated out (Mes90a), the two components to-
gether always achieved lower indoor radon levels than
did either component alone. But in two of the three
houses,  basement treatment alone was sufficient to re-
duce levels below 1-2 pCi/L, and the SMD component
had  only a minor incremental effect The one house
where the SMD component was important (House 1357)
was  the only house where basement sub-slab communi-
cation was poor, and where the crawl-space appeared to
be an important radon source based upon soil radon
concentrations. In that house, basement concentrations
were reduced from 11 to 3 pCi/L with SMD treatment
only, whereas basement SSD by itself (with a one-pipe
SSD system) reduced  levels only to 6 pCi/L. The two
components together reduced the basement to 1.2 pCi/L.
The success of the basement-only treatment in the other
two houses could be due in part to the low pre-mitigation
concentrations in these other houses (4.2-4.8 pCi/L).

The three  EPA  study  houses which had  slab-on-
grade living areas adjoining the crawl spaces were
all treated with SMD only, with no attempt to apply
SSD to the adjoining slab  (Py90). In two of these
houses, treatment of the crawl  space alone  was
sufficient to reduce living-area concentrations from
16-26 pCi/L, down to 2-3  pCi/L. These two houses
include  one (DW31) where the slab-on-grade whig was
almost as large as the crawl space, and where the SMD
system  included four  suction pipes and no membrane
(i.e., site depressurization).  However, the third house
with an adjoining slab on grade was reduced only from 28
to 15 pCi/L, despite the fact that the crawl space was
relatively small  (900  rP and the adjoining  slab was
    significantly smaller (a converted garage having only 300
    ft2).

    Of the 15 EPA crawl-space study houses reduced below 2
    pCi/L, 9 had adjoining basement wings which were being
    treated with SSD or DTD, supplementing the crawl-space
    SMD system. Stated another way, out of the 11 study
    houses having an adjoining basement wing being treated
    by SSD or DTD, only 2 were not reduced below 2 pCi/L.

    While not definitive, the above data, taken overall, seem
    to be clearly suggesting that any wing adjoining the crawl
    space should normally be treated in addition to (or instead
    of) SMD in the crawl space. The relative importance of
    treating the adjoining wing vs. the crawl space may vary
    from house to  house. However, in most cases, SSD or
    DTD in  the adjoining wing will probably be important;
    and, in many cases, it will likely be more important than
    SMD in  the crawl space.

b)  House operating variables

 •  Operation of central furnace fan. Investigators
    who  have measured sub-membrane depressurizations
    in SMD  systems not having perforated piping or gravel
    beneath the membrane, consistently report that  de-
    pressurizations drop below 0.01 in. WG within 6 to 10
    ft of the suction pipe (Os89b, Py90), and below 0.001
    in. WG within 10 to 15 ft (Py90).

    Where a central forced-air furnace is present in the crawl
    space, operation of the furnace fan can be expected to
    have several complex effects. One effect will be that the
    leaky cold-air return ducts will create some depressuriza-
    tion of the crawl space, mitigated to some extent by the
    general leakiness of crawl spaces. This depressurization
    would work against the sub-membrane depressurization
    created by the SMD system, tending to draw sub-mem-
    brane gases up into the crawl space through leaks in the
    membrane. A second effect will be that the radon-con-
    taining crawl-space air that is sucked into the return
    ducting will be distributed throughout the house, dramati-
    cally increasing the interzonal transfer of air between the
    crawl space and the living area

    The question is whether the furnace fan will depressurize
    the crawl space sufficiently to overwhelm the sub-mem-
    brane depressurization, drawing more  soil gas into the
    crawl space and then helping to distribute it throughout
    the house. Based upon the data from  Os89b and Py90
    above, a crawl-space depressurization as low as 0.001 in.
    WG  could  nominally be sufficient to overwhelm  the
    system.

    Limited data have been reported for two Alabama houses
    regarding the effects of a central furnace fan in the crawl
    space on crawl-space pressures (Houses HU11 and HU12
    in Ma89b). In HU12, central  fan operating decreased the
    pressure in the crawl space (relative to  outdoors) by less
    than  0.001 in.  WG. Nominally, this increase in crawl-
    space depressurization could help to  overwhelm sub-
    membrane depressurization  at points  remote from the
                                                   56

-------
    SMD suction pipes. But in House HUH, operation of the
    central fan appeared to increase the crawl space pressure
    (by somewhat more than 0.001 in. WG) relative to out-
    doors. If this measurement is correct, the central fan
    would be aiding the SMD system, increasing the effective
    sub-membrane depressurization relative to the crawl space.
    These reported central fan effects are somewhat in ques-
    tion. The pressure effects  are small (often  below the
    sensitivity of the measurement device); and the numbers
    are the average of several measurements made over the
    course of a day, so that temporal variations could have
    influenced the results.

    The measurements of crawl-space pressures in Ma89b
    were not accompanied by measurements of crawl-space
    radon concentrations  or  of SMD system performance.
    Thus, it  is unknown whether the measured pressure ef-
    fects would have in fact overwhelmed a SMD system, or
    have increased indoor radon levels.

    The effects of  the central furnace fan on interzonal air
    transfer between the crawl space and the living area have
    been reported on these same two houses (Ma89b). These
    data, based upon freon tracer testing, confirm that inter-
    zonal flows can increase significantly (e.g., from about
    50 to about 200 cfm in HU11). Thus, to the extent that a
    central fan in the crawl space did increase crawl-space
    radon levels, the forced-air system would serve to distrib-
    ute the radon through  the house.

c)  Mitigation system design variables

 •  Method of distributing suction beneath membrane.
    Two primary approaches are considered for distributing
    suction:  1) drawing suction  on perforated piping beneath
    the membrane (analogous to DTD); and 2) installing one
    or more individual  suction pipes through the membrane
    at various points, with no sub-membrane perforated pip-
    ing (analogous  to SSD). No data exist comparing a SMD
    system in a single house with and without perforated
    piping beneath the membrane.

    In summary, as discussed previously under the section on
    House size, the data are not fully definitive, given the
    number of variables that were being varied. However, the
    trend clearly suggests that perforated piping may be
    important in a crawl space of any size if the living area is
    to be reduced to 2  pCi/L or less. And, especially if the
    crawl-space floor area is larger than 1,500 to 2,000 ft2,
    perforated piping may be important to ensure levels be-
    low 4 pCi/L, at least without multiple suction pipes.

    All nine of EPA's SMD installations using sub-mem-
    brane perforated piping  were reduced below 2 pCi/L,
    regardless of floor area (up to 2,700 ft2). Only 3 of the 12
    .stand-alone SMD systems without perforated piping re-
    duced the living area below 2 pCi/L, and all of these had
    special circumstances which aided the achievement of
    this low level. Of the two houses larger than 1,500 ft2 not
    having sub-membrane perforated piping,  one required
    two suction pipes to be reduced below 4 pCi/L, and the
other remained above 4 pCi/L despite the use of two
suction pipes.

In addition to the results on  these EPA-sponsored
installations, commercial mitigators report routinely
using sub-membrane perforated piping whenever lev-
els below 2 pCi/L are desired and/or when the crawl
space is sufficiently large (An92, How92, K192).

All nine of the EPA study houses which have used sub-
membrane piping have been reduced to 2 pCi/L and less
in the living area (Sc88, Fi90, Gi90, Du91). The pre-
mitigation concentrations in these houses were typically
15-30 pCi/L, but were 40-60 pCi/L in two cases. These
reductions were achieved despite relatively large crawl
spaces (most ranging from 1,000 to 2,700  ft2, except in
three houses where the crawl-space was a 200 to 600 ft2
wing adjoining a basement). In three of the four "pure"
crawl-space houses in Ohio (Fi90), this  good perfor-
mance might have been aided by the presence of gravel
on the  crawl-space floor. In the two Alabama houses
(Du91), the two Maryland houses (Gi90), and the one
Pennsylvania house (Sc88), which had basements adjoin-
ing the crawl spaces, simultaneous SSD or DTD in the
basement wing was undoubtedly  contributing  signifi-
cantly to the high reductions achieved by the SSD + SMD
(or DTD + SMD) systems.

Among the nine essentially "pure" crawl-space houses
where no  perforated piping was installed beneath the
membrane (Ni89, Os89b, Py90), in only two cases were
concentrations in the living area reduced below 2 pCi/L
(although levels were reduced below 4 pCi/L in all but
two of the houses, as discussed in the section on House
size). Among the two houses where concentrations were
reduced below 2 pCi/L, one had a very low pre-mitiga-
tion  level of 3 pCi/L (Os89b); the second required a
BWD and a water treatment component, in addition to a
three-pipe SMD system, to achieve this level in mild
weather (Ni89).

Six of EPA's crawl-space study houses had adjoining
basements being treated by SSD or DTD, with no perfo-
rated piping beneath the  membrane in the SMD compo-
nent (Mes90a, Py90). Of these  six, four were reduced
below 2 pCi/L without the perforated piping (Mes90a).
However, the crawl-space wings in these four combined-
substructure houses were relatively small (300 to 600 ft2),
and the basement SSD or DTD component of the mitiga-
tion system was playing  the major role in achieving the
high reductions. In only one of these houses (House 582)
was  the SMD  component adequate by itself to reduce
basement concentrations below 2 pCi/L; and this house
had the benefit of a low pre-mitigation level (less than 5
pCi/L in the basement)  and gravel on the crawl-space
floor to aid in extending the suction field.

The  other  two houses with adjoining  basements and
without sub-membrane perforated piping on the SMD leg
of the system (Houses DW14 and DW78 in Py90) were
reduced below 4 pCi/L, but not below 2 pCi/L. These two
                                                       57

-------
houses had crawl-space wings of about 300 and 1,100 ft2,
respectively.

Another approach that has been considered to aid  in
suction field extension beneath the membrane is to place
porous matting beneath the membrane. This approach has
the additional benefit of providing support to reduce the
risk of membrane punctures, but the matting would add
significantly to the installation cost. Sub-membrane mat-
ting has been tested in only two houses (Ni89, Py90). The
results have not been definitive, with one of the houses
still above 4 pCi/L (Py90), and the other requiring a
BWD and water treatment component (and three SMD
pipes) to be reduced below 4 pCi/L (Ni89).

Number of suction pipes (where perforated piping is
not placed beneath membrane). No systematic study has
been conducted to assess the effect of the number  of
individual SMD suction pipes penetrating the membrane
in cases where no sub-membrane perforated  piping is
used. In general, only one suction pipe has been installed.
The only houses in which the effect of one vs. two suction
pipes has been tested have  been those cases where the
first pipe proved to be insufficient to reduce living-area
concentrations below 4 pCi/L.

In summary, a single pipe has commonly been sufficient
to reduce crawl-space houses as large as 1,500 ft2 below 4
pGi/L in the living area, even when there is no gravel on
the floor. This is the case despite the fact that the limited
sub-membrane suction field extension data suggest that,
theoretically, a single pipe should not be able to effec-
tively treat an area that large, and despite the fact that the
membrane has not always been fully sealed. However,
one pipe has generally been insufficient to reduce these
houses below 2 pCi/L, except where the crawl space is
small (600 ft2 or less, from  the available data) and is a
wing adjoining a basement which is also being treated.

Data are limited from houses having crawl spaces larger
than 1,500 ft2. However, these limited data suggest that,
with such larger crawl spaces, more than one suction pipe
will likely be needed, at least when there is no gravel on
the floor. If gravel were present to help extend the suction
field, it might be expected that a one-pipe system would
do a better job. Intuitively, complete sealing of the mem-
brane should help reduce the number of suction pipes
needed in these larger houses, although the available data
are too limited to permit a definitive statement regarding
the benefits of membrane sealing.

Where sub-membrane depressurizations have been mea-
sured (Os89b, Py90), measurable depressurizations (i.e.,
down to 0.001 in. WG) have  usually been found to extend
no more than about 10 to 15 ft from the suction pipe in
cases where the crawl-space floor is bare soil (no gravel,
no  matting). The rule of thumb discussed in Section
2.3. le for basement houses  is that, ideally, sub-slab de-
pressurizations of approximately 0.015 in. WG (mea-
sured during mild weather) would be desirable to ensure
that a SSD system is  not overwhelmed by the thermal
stack effect during cold weather. Sub-slab depressuriza-
tions below 0.002 to 0.005 in. WG (measured during cold
weather) may be sufficient; however, such limited de-
pressurizations may occasionally be overwhelmed  by
exhaust appliance operation. Sub-membrane depressur-
izations have been found to drop below that value within
6 to 10 ft of the SMD suction pipe. Thus, on the basis of
measured sub-membrane depressurizations, one would
estimate that, where no gravel is present to aid suction
field extension, one SMD suction pipe would be required
roughly every 100 to 400 ft2 if a depressurization of 0.01
in. WG were desired, or every 400 to 900 ft2 if a depres-
surization of 0.001 in. WG were sufficient.

But in apparent contradiction to this calculation, houses
with crawl spaces as large as 1,500 ft2 have consistently
been reduced below 4 pCi/L in the living area (from pre-
mitigation levels as high as  20  pCi/L) with a single
suction pipe penetrating the membrane at a central loca-
tion. This has been the case even when there is no gravel
on the floor to help distribute the suction field, and when
the membrane is  not fully sealed.

The EPA program has included  13 crawl-space  study
houses (including 6 with adjoining wings) having  crawl
spaces of 1,500 ft2 or smaller, where SMD systems have
been installed with no sub-membrane perforated piping
(Ni89, Os89b, Gi90, Mes90a, Py90). Of these 13, in only
one case was one pipe insufficient to reduce living-area
concentrations below 4 pCi/L (House DW60  in Py90,
with a 900 ft2 crawl space, which had a small adjoining
slab-on-grade living area). On the other hand, the only
houses among these 13 which were reduced below 2 pCi/
L with a single pipe were four of the houses where  the
crawl space was  a relatively small (300 to 600 ft2) wing
adjoining a basement which was also being treated  using
SSD or DTD (Gi90, Mes90a). It is known that the  base-
ment SSD or DTD component, not the SMD component,
was responsible for the low levels achieved in those four
houses.

Data are available from  only  two houses having crawl
spaces larger than 1,500 ft2, where no perforated piping
was placed beneath the membrane (Py90). Neither house
had gravel  on the crawl-space floor. The first house,
House DW27, had an unusually long, narrow crawl  space
(approximately 100 ft by 23 ft, with a total floor area of
2,000 ft2). One SMD suction pipe, toward one end of this
crawl space, could reduce the living-area concentrations
in this house only from 33 to 8-12 pCi/L.  Adding a
second suction pipe, located toward the other end of the
crawl space, could reduce levels only to 5 pCi/L, despite
the fact that the membrane was sealed everywhere (around
the perimeter and around piers, and at seams between
sheets), and despite the  fact  that porous matting was
strategically placed beneath portions of the membrane to
improve suction  field distribution. A small section (250
ft2) at one end of the crawl space had been paved, and
served as a basement on the  same level as the crawl-
space; this small basement was not treated with SSD, and
thus could have been partly responsible for the failure of
this house to be reduced below 4 pCi/L.
                                                    58

-------
 The second house having a large crawl space, House
 DW29, had a crawl-space floor area of 2,000 ft2 and an
 adjoining slab-on-grade living area (a converted garage)
 of 600 ft2. With the membrane sealed around piers and
 seams near the pipe penetrations but nowhere else, and
 with no treatment of the slab on grade, one pipe toward
 one end of the crawl space reduced living-area concentra-
 tions from 16 to  6-7 pCi/L. A second pipe, toward the
 other end, reduced levels to 3 pCi/L.

 There may be several reasons why a single SMD suction
 pipe is able to treat a floor  area greater than would be
 estimated based upon the 0.015 in. WG rule of thumb.

-  Sub-membrane  depressurizations much smaller than
   0.015  in. WG—perhaps  too small to be measured
   quantitatively—may be sufficient to make the SMD
   system effective. Chemical smoke visualization testing
   around the  unsealed perimeter of many of the mem-
   branes in Reference Py90 confirmed that, although the
   measurable sub-membrane depressurization extended
   only 10 to 15 ft from the  suction point, some suction
   was in fact extending all the way to the perimeter, since
   smoke was being clearly drawn beneath the membrane
   at that location (Br92).

-  The rule of thumb is based on the assumption that the
   SMD  systems function only by the mechanism of
   reversing the direction of  air flow between across the
   membrane.  That mechanism  would suggest that the
   sub-membrane region must in  fact be depressurized
   essentially everywhere almost all of the time if the
   SMD system is to maintain good performance. In fact,
   another mechanism can also contribute—ventilation of
   the  sub-membrane region, and  dilution of sub-mem-
   brane radon concentrations.

-  Or, depressurization of the entire crawl space, created
   by leakage of crawl-space air into the SMD system,
   could be reducing air flow up into the living area from
   the crawl space and could be increasing crawl-space
   ventilation, thus introducing a third mechanism.

 Results have  been reported  from  one house where site
 depressurization was tested in the crawl space, i.e., "SMD"
 with no membrane (House DW31 in Py90). The house
 had a moderately sized  crawl  space  (800 ft2), and an
 adjoining slab on grade living area of comparable size. In
 this system, the suction pipes terminated in pits excavated
 in the soil floor, covered with treated ply wood. The effect
 of two vs. four suction pipes was tested. With two pipes
 operating, one in each of two diagonally opposing quad-
 rants of the crawl space, living-area concentrations were
 reduced from 26 to about  10  pCi/L. With four pipes
 operating, one in each quadrant, levels fell to 2-4 pCi/L.
 Suction measurements in test holes drilled about 1 ft deep
 into the soil  floor confirmed that, with  all four pipes
 operating, the soil was depressurized by 0.005 in. WG or
 greater at distances of 6 ft from the suction pipes.
Location of suction pipes (or of perforated piping).
The available data do not permit an assessment of the
effect of pipe location.

Where perforated piping  has been placed beneath the
membrane, the data are insufficient to identify any differ-
ence between placement of the piping as a loop around
the perimeter vs. alternative configurations  in a more
central location. Some mitigators who use perforated
piping prefer a  strip of piping down the middle of the
crawl space, for simplicity and to minimize air leakage
into the system through the perimeter seam and through
the block walls. Other researchers suggest a perimeter
loop of piping,  because  wind effects may cause the
greatest radon flux out of the soil to  occur around the
perimeter (Sc90a); also, a perimeter loop would  be ex-
pected to provide the best treatment of block foundation
walls, if the walls are thought to be a source. All perfo-
rated piping configurations have appeared to work well.

Where no perforated piping has been placed beneath the
membrane, single-suction-pipe systems have generally
had the one pipe penetrating the membrane at a central
location. Two-pipe systems have had one pipe in each
half of  the crawl space,  sometimes toward  (but never
immediately beside) the perimeter wall. Since the suction
field appears to extend in roughly the same manner in all
directions from the suction pipes (Os89b,  Py90), i.e.,
since  sub-membrane communication appears generally
uniform, such central location of the pipes would seem to
make sense. Intuitively, it would be desirable to locate the
pipe penetrations away from unsealed seams in the mem-
brane, to reduce the short-circuiting of crawl-space  air
into the system. For example, if the perimeter  of the
membrane is not sealed against the foundation  wall, it
would seem desirable to locate the pipes at least 6 to 10 ft
from the walls, since that appears to be about how far the
measurable suction field may extend.

Size of suction pipes. The lowest SMD system flows
would be expected in cases  where the membrane  is
completely sealed, there is no gravel on the floor, and
there  is not perforated piping beneath the membrane.
Where system flow measurements have been reported
under these conditions, flows have been between about
20 and 40 cfm with the standard 90-watt fans or equiva-
lent operating at full power (Os89b, Mes90a, How92).
Where the membrane is not completely sealed but condi-
tions  are otherwise similar (no  gravel, no perforated
piping), limited data suggest a somewhat higher range, 30
to 100 cfm (Py90, Br92). Where gravel is present but
conditions are otherwise similar (membrane completely
sealed, no perforated piping), again, limited data suggest
a somewhat higher range,  30 to 100 cfm (Mes90a).

This range of flows  from SMD systems with no sub-
membrane perforated piping (20 to 100 cfm) is  almost
identical to that reported in Section 2.3.Ic for SSD sys-
tems. As discussed in that earlier section,  the  proper
piping size for a given installation will depend upon the
specific piping configuration and fan performance curve
for that particular installation. However, in general, the
                                                    59

-------
commonly used 4-in. diameter piping will probably be a
reasonable choice. Three-in. piping can usually be con-
sidered  in cases'where desired  for aesthetic reasons,
where the flows  are sufficiently low, the  piping run
sufficiently short,  and/or the fan performance curve suit-
able. In selecting  the pipe size, it should be recognized
that flows from SMD systems may increase over time as
some of the membrane seals break, or as the membrane
becomes punctured.

Where the membrane is completely sealed and there is no
gravel,  as in the lowest-flow case above,  but where
suction is being drawn on sub-membrane perforated pip-
ing, the observed range of SMD flows  is 20  to 110 cfm
(Sc88, Fi90, Bo91, K192). This range is  similar to that
observed above where there is no sub-membrane piping
(including the no-perforated-piping cases without com-
plete sealing  and with gravel). Thus, where the mem-
brane is completely sealed and where there is no gravel
with systems having sub-membrane piping, the consider-
ations in selecting the proper pipe diameter are the same
as those discussed above for systems without such piping.

The highest SMD flows might be expected in systems
where the membrane is not completely sealed, there is
gravel on the floor, and suction is being drawn on perfo-
rated piping under the membrane. In two such installa-
tions where flows were reported (Fi90), flows were in-
deed high, 190 to 200 cfm with a 90-watt fan operating at
full capacity. In each of these cases, the perforated piping
formed  a  loop  around the crawl-space perimeter; the
membrane was sealed against the foundation wall, but did
not always completely cover the interior. In two  other
installations where the membranes were completely sealed
but where conditions were otherwise similar (gravel,
perforated piping), flows  were 100 cfm in one 300 ft2
crawl space and 220 cfm in a second, 1,300 ft2 crawl
space. The particularly high flows in  the larger house
may have resulted in part because the perforated piping
matrix included three lengths  of parallel piping,  con-
nected to a single header.

In systems having flows in the range of 200 cfm, the use
of the common 4-in. piping could result in a significant
suction  loss through the piping, depending upon the
configuration of the piping system. Thus, if the system in
fact has to be operated at such high flows,  a mitigator
may sometimes have to consider the use of larger piping.
However, in the three houses having the flows between
190 and 220 cfm with the fan at full power, the systems
were so effective with the perforated piping and the
gravel, that the fans  could  be  operated at significantly
reduced power and still maintain indoor levels below 2
pCi/L. The final installations in these  houses, with the
fans left operating at reduced power, typically had flows
in the range of 50 to 100  cfm, comparable to that  in
typical SSD systems. Thus, again, 4-in. piping may com-
monly be sufficient, but may not be crucial in some cases
if there is a pressing need to use smaller-diameter piping
for part of the piping route.
Pits beneath membrane where pipes penetrate. In
houses having slabs with poor sub-slab communication,
an attempt is sometimes made to improve suction field
extension by excavating a pit beneath the slab at the point
where the SSD pipe penetrates. This pit could function
both by reducing the pressure losses as the sub-slab gases
accelerate to pipe velocity, and by intersecting fissures or
good-communication strata beneath the soil surface. A
pit beneath the  SMD membrane at the  point where an
individual suction pipe penetrates (in cases where sub-
membrane perforated piping is not being installed) might
be envisioned as working by similar mechanisms.

The  available data are not sufficient  to  quantify  the
performance benefits of excavating pits of various sizes,
compared against other methods for installing individual
suction pipes through the membrane.

Pits  were excavated in the eight SMD installations in
Reference Py90, including one in which there  was no
membrane (site depressurization). No results are reported
assessing the effect of pit size.

More commonly in commercial installations, no pit is
excavated. Instead, the suction pipe is supported in some
manner on the floor of the crawl space, and the membrane
is sealed around the pipe at a height of perhaps a foot
above floor level (see Figure 6). Various approaches for
accomplishing this are described later, in Section 8.5.2.
This method effectively provides a "pit" beneath  the
plastic "slab" which would serve the function of reducing
the pressure losses from gas acceleration. The only differ-
ence is that in this case, the "pit" is created by raising the
"slab" rather than by excavating the soil.

Although raising the membrane will reduce pressure losses
due to acceleration as effectively as would excavation of
a pit, it will not accomplish the second objective of a pit,
namely, to intersect fissures and strata in the soil beneath
the SMD membrane. Since SMD  probably  functions
largely by sweeping away the radon that enters the region
between the membrane and the soil, the failure to help
extend a suction field through sub-surface fissures may
not be important. In any event, with the limited data
available, no improvement in radon reductions has been
demonstrated for  the excavated pit versus the raised
membrane approach.

Nature of membrane. Almost all of the EPA projects
and the reported commercial installations have used poly-
ethylene sheeting as the membrane material. Where the
crawl-space  floor is relatively smooth  and not much
traffic is expected, standard polyethylene sheeting as thin
as 6 mil has been used (Py90). Such standard sheeting is
subject to puncture, and is protected by strips of heavier
material, such as ethylene propylene  diene  monomer
(EPDM), a rubberized roofing material, along expected
traffic routes in the crawl space. More commonly, thicker,
8- to 10-mil standard polyethylene, or  cross-laminated
sheeting which is 8- to 10-mil equivalent, and is report-
edly much more puncture-resistant than the standard ma-
terial, has been used (Bro90, Fi90, Mes90a, Jo91, Sh91,
                                                   60

-------
An92, K192). In a few cases, especially in walk-in crawl
spaces where heavy  traffic is expected, 15- to  20-mil
material (or equivalent) has been used, even where the
floor has been relatively smooth (Os89b, Gi90).

In several houses in Alabama where the crawl-space
floors were  rocky and irregular, heavy-duty reinforced
industrial pit liner material was used as the membrane
(Ma88).

Extent of membrane. In almost all of the EPA study
houses and reported commercial installations, the SMD
membrane has covered the crawl-space floor completely.
The limited exceptions include

a) two houses in Ohio (Houses 22 and 24 In Fi90), where
   an 8-ft wide strip of membrane was plaiced around the
   crawl-space perimeter (covering a perimeter perfo-
   rated piping loop), and was connected to sheets of pre-
   existing vapor barrier which covered the central por-
   tion of the floor to a large extent (but not completely).

b) one house in  Tennessee (House DW31 in  Py90),
   where no membrane was installed  (i.e., where the
   technique used was site depressurization).

In no case  has back-to-back testing been reported of
different degrees of floor coverage.

As discussed previously, excellent results were obtained
in the Ohio  houses, reducing living-area concentrations
from pre-mitigation values of 16-17 pCi/L down to below
2 pCi/L with the SMD fan at reduced power. But the tile
loop, and the gravel on the floors  of both crawl spaces,
were likely contributing to the success of these installa-
tions. Moreover, the  central portions of the floors were
not completely uncovered, since the pre-existing vapor
barrier was providing at least a partial membrane. Thus, it
is not clear how universally, and under what other condi-
tions, partial floor coverage will be satisfactory.  Smoke
tracer testing in House 24 with the SMD system operating
did show  that, at an uncovered portion of the  central
crawl space, flow did appear to  be upward from the
gravel into the crawl  space.

These results in the  Ohio houses suggest that, in  fact,
flow reversal everywhere may not always be necessary
for good performance, and that, under the right condi-
tions, complete floor coverage might not  be needed.
However, the data are so limited that no definitive con-
clusions can be drawn regarding when only partial cover-
age may be acceptable. Thus, at the present time, it is
recommended that the crawl-space floor be completely
covered if this is at all possible.

Because the perforated piping loops were immediately
beside the foundation wall in the two Ohio houses, it was
felt to be necessary to effectively  seal the membrane
against the  wall,  to reduce air short-circuiting into the
system. To ensure a permanent seal, this perimeter seal-
ing was accomplished by adhering the plastic sheeting to
the walls with a bead of sealant,  then mechanically at-
taching the plastic by sandwiching it between the wall
and a furring strip secured with explosively driven nails.
Testing was not conducted to confirm that this rigorous
perimeter sealing  approach was indeed required. But if
such a time-consuming perimeter sealing effort did in
fact prove to consistently be necessary when the tiles are
beside the  foundation wall,  the increased cost of this
sealing effort could largely offset any savings achieved
by not having to  cover the central portion of the floor
(He91b, He91c).

As also discussed previously, the four-pipe site depres-
surization system in Tennessee House DW31 achieved
good results, reducing levels in the living area from 26 to
just above 2 pCi/L. This result demonstrates that, in some
cases, no membrane is required at all. Since these results
were obtained on only one house, it is not at all clear how
widely this approach  will be  applicable. It would be
expected that site depressurization will be applicable only
in cases where the native soil in the floor is reasonably
permeable,  as it visually appeared to be in DW31. Ide-
ally, the permeable soil stratum on which suction is being
drawn  should lie covered by a less permeable layer, to
reduce the short-circuiting of crawl-space air down through
the soil into the system. Terminating the suction pipes in
covered pits, as in DW31, is also likely an important
consideration in the design  of a site depressurization
system, in order to reduce pressure losses resulting from
soil gas acceleration and possibly to intersect more per-
meable fissures/strata.

A variation of site depressurization was attempted in four
crawl-space houses in Ohio  (Fi90). In these houses, a
single 4-in. diameter pipe was embedded about 10 ft deep
into  the soil outside the house, immediately beside the
foundation, and suction was drawn with a 90-watt fan. No
radon reduction was obtained in the living areas of these
houses. The lack of success in these  installations was
likely due primarily to the impermeability of the native
clay soil. Other contributing factors could have been the
inability to excavate a suction pit beneath the pipe, and
the fact that the suction was  not being drawn inside the
foundation.

From these results in Tennessee and Ohio, it would
appear that the no-membrane site depressurization ap-
proach will be applicable only in potentially isolated
cases where special geological conditions are present.

Degree of membrane sealing. Where alternative de-
grees of membrane sealing have been  tested  in a given
house, the limited results  suggest that good performance
can often be obtained even when the membrane  is not
sealed anywhere except in the vicinity of the suction pipe
penetration through the membrane. But to achieve such
good performance, the SMD fan must  have adequate
capacity to handle any increase in flows resulting from
leakage through unsealed seams. The  standard 90-watt
fans appear generally sufficient to handle unsealed mem-
branes, but significantly reduced performance has some-
times been observed when the  fan has been  operated at
reduced power under such conditions.
                                                    61

-------
Although reasonable SMD performance has often been
observed in the relatively limited number of cases where
the membrane has not been completely sealed, the best
radon reductions at a given system flow have generally
been observed when the membrane has been completely
sealed, as would be expected. Where perimeter block
foundation walls are thought to be a potential source,
careful sealing of the membrane will be most likely to
extend the SMD suction field adequately to treat wall-
related entry routes. Moreover, by reducing the amount
of crawl-space air drawn into the  system, sealing will
reduce the risk, of backdrafting combustion appliances in
the crawl space, and should reduce the amount of treated
house air drawn into the system (hence reducing the
heating/cooling penalty). For these reasons, most mitiga-
tors routinely seal SMD membranes everywhere in com-
mercial installations. At the present time, complete seal-
ing of the membrane is recommended, if this is at all
possible. Complete sealing includes sealing at seams
between  sheets; around the crawl-space perimeter; and
around interior piers or any other obstructions. (Sealing
around the suction pipe penetration through the mem-
brane is required in all cases.)

In House 28 in Ohio (Fi90), the initial SMD configura-
tion tested consisted of inserting two individual suction
pipes through the pre-existing vapor barrier, which  cov-
ered most of the floor. The seams between the sheets of
plastic were caulked where possible, but such caulking
was not possible everywhere. No effort was made to seal
the plastic to the perimeter foundation wall. There was no
gravel on the floor. With the 90-watt fan  at full power,
this system reduced living-area concentrations from 5-7
pCi/L, down to 1.0 pCi/L. This initial system was  then
replaced  with  two parallel lengths  of perforated piping
beneath an entirely new membrane. The sheets of the new
membrane were overlapped by a foot, and attached to
each other with polyethylene adhesive. The membrane
was attached to the perimeter walls with sealant and a
furring strip; it was not sealed around interior piers or
around an interior fireplace support structure. With the
fan again at full power, this upgraded system resulted in
living-area concentrations of 0.5 pCi/L. The difference in
radon levels with the two systems is so small that it is not
clear that the improved system had any real effect; and to
the extent it did have an effect, that effect could be due in
part to the addition of perforated piping beneath the
membrane, not just to the membrane sealing effort. The
flows from the two systems were  the same, 130  cfm,
suggesting that the sealing did not reduce air leakage into
the system.

In a second Ohio house (House 33 in Fi90), the SMD
system was tested with and without the membrane being
sealed around the perimeter. (The membrane was sealed
at seams between sheets in all cases.) The system in this
house consisted  of three parallel lengths of sub-mem-
brane perforated piping, on  a  gravel floor. With the
membrane unsealed  around the perimeter,  living-area
concentrations were reduced from 17 pCi/L to 13 pCi/L,
with the fan at very low speed (exhausting 18 cfm); and
0.4 pCi/L, with the fan at maximum speed (exhausting
224 cfm). After the perimeter was sealed, indoor levels
were reduced to  1.0 pCi/L, with the fan at a low speed
which was inadvertently higher than the very low speed
tested prior to sealing (exhausting 54 cfm); and 0.4 pCi/L
with the fan at maximum speed (exhausting 118 cfm).
Thus, at full fan power, similar indoor radon reductions
were achieved with and without perimeter sealing, but
sealing reduced the exhaust rate about in half. This sig-
nificant impact of sealing on flows is probably due in part
to the presence of gravel. The effect of sealing at lower
fan speed cannot be determined, since the reduced speeds
tested before and after sealing were different.

While the comparison and interpretation of the figures
from House 33 is complicated by the differences in fan
speeds  tested before and after perimeter sealing, two
possible effects of the sealing effort are apparent. First, if
the fan capacity is great enough, very good SMD perfor-
mance can be achieved even  with the membrane perim-
eter unsealed.  Second, when the  membrane is sealed,
comparable  performance may sometimes be achieved
with lower system flow rates than is possible without
sealing, thus reducing the risk of backdrafting and, likely,
the heating/cooling penalty.  The  good reductions ob-
served in these houses without complete membrane seal-
ing may have been due in part to  the facts that a)  sub-
membrane perforated piping was being used in both
cases, helping to ensure good distribution of the suction
field even with membrane leakage; and b) the gravel in
House  33, again helping to  ensure good suction field
distribution.

In four Maryland houses where a basement adjoined the
crawl space, the crawl-space SMD system was tested
with and without the membrane sealed to the foundation
wall around the  perimeter (Mes90a). In all cases, the
membrane was sealed at seams between sheets.  The
results  with and without perimeter sealing were  con-
ducted with only the SMD system operating; the SSD or
DTD system in the adjoining basement was disconnected
during this testing. The SMD  systems consisted of single
suction pipes penetrating the membrane at a central loca-
tion. Two of the houses had gravel on the floor, two did
not. In  all four cases, sealing the membrane perimeter
reduced radon concentrations in both the basement and
the upstairs living area, usually by 0.5 to 1.6 pCi/L. (Pre-
mitigation levels in these houses were generally low, 3 to
11 pCi/L; thus,  these relatively small changes in the
absolute radon reduction represented a fairly significant
increase in the percentage reduction.) While the absolute
reductions in radon concentrations were relatively small,
and thus subject to some uncertainty, the fact that levels
consistently  went down after perimeter sealing suggests
that this could be a real effect.

In most other reported studies, the SMD systems were
either tested only with the membrane sealed everywhere
(Ni89, Gi90, Py90, Du91), or with the membrane sealed
nowhere except near the suction pipe penetration (Py90).
The  number of  houses involved  is  too few, and the
number of other variables being varied is too great, to
                                                    62

-------
    permit any assessment of the effects of membrane sealing
    from these other studies.

d)  Mitigation system operating variables

 •  Fan capacity. Most of the reported SMD installations
    have involved the 90-watt fans operating at full power.
    Since fan capacity has generally not been varied, not
    much can be said at this time regarding the ability to
    achieve adequate radon reductions with smaller fans or
    with the 90-watt fans at reduced power. Given the moder-
    ate flows usually measured in SMD systems (commonly
    20 to  100 cfm, as discussed under Size of suction pipes
    above), and given the suction that appears to be necessary
    to extend a  suction field beneath the membrane (see
    Number of suction pipes above), use of 90-watt fans at
    full power in SMD applications appears reasonable.

    The one study reporting the use of 90-watt  fans at re-
    duced power has been  the testing on four crawl-space
    houses in Ohio (Fi90). These houses all had perforated
    piping beneath the membrane, and three had gravel on the
    floor, improving the potential for success with the fans at
    reduced power. In three of these houses, operation of the
    fan at low power, to  exhaust about 50 cfm, reduced
    living-area concentrations below 2 pCi/L, and often to 1
    pCi/L and less (from pre-mitigation levels of 5-17 pCi/L).
    In the fourth house (House 22), the fan had to be operated
    at medium power, exhausting  about 100 cfm, to be re-
    duced below 2 pCi/L.  But even  in  this fourth house,
    operation at low  power (exhausting 50 cfm) was suffi-
    cient to reduce levels from 17 to below 4 pCi/L. To
    achieve these flows and reductions at reduced fan power,
    it was generally necessary to have sealed the membrane
    at seams and around the perimeter.

    Operation of the fan at full power in the Ohio houses
    (exhausting 130 to 224 cfm) gave further marginal reduc-
    tions, reducing living-area concentrations below 1 pCi/L
    in all cases.  Moreover, as the experience in House 33
    showed (see Degree of membrane sealing above), full-
    power operation made sealing of the perimeter and seams
    less necessary; at maximum fan power, indoor  levels
    below 1 pCi/L were achieved even without  sealing the
    membrane perimeter.

    It is re-emphasized  that the  success with reduced fan
    capacity in the Ohio  houses is likely due to the facts that
    there is sub-membrane perforated piping and, in three of
    the houses, gravel on  the floor to help distribute the
    suction. The flows observed in those houses with the fan
    at low to medium power (50 to 100 cfm, often with the
    membrane sealed) are the same as (or higher than) those
    observed with a 90-watt fan at  full power in houses
    without perforated piping or gravel.

    In summary, a 90-watt fan at full power is advisable when
    the SMD system has no  perforated piping  under the
    membrane, when the membrane is not completely sealed,
    and when there is not gravel on the floor. When there is
    perforated piping, the membrane is sealed, and there is
    gravel—i.e., when there is good sub-membrane commu-
    nication, and the necessary flows can be obtained with
    less fan capacity—operation at reduced fan power (or
    with a smaller, 50-watt fan) may be satisfactory. But even
    in this latter case, there will likely always be some
    improvement in performance with full-power operation.
    Where there is doubt regarding whether full-power op-
    eration is required, the operating cost savings resulting
    from operation at reduced power is probably often not
    sufficient to justify the risk of operation at less than full
    power (He91b,He91c).

 •  Fan in suction vs. pressure. Operation of SMD sys-
    tems with the fan reversed to blow outdoor air beneath
    the membrane,  has not been tested. Based  upon the
    experience with fan reversal in other ASD techniques and
    recognizing the general leaHness of SMD membranes
    relative to slabs, as well as the poorer performance gener-
    ally observed when the crawl-space is ventilated with
    forced-air supply from outdoors (crawl-space pressuriza-
    tion) vs. forced-air exhaust from the crawl space (depres-
    surization), it would be  expected that sub-membrane
    pressurizatibn would likely be less effective than SMD.
    See Section 2.4.

e)  Geology/climate variables

 •  Source term. With the possible exception of House OP-
    05 in New York (Ni89), none of the tested crawl-space
    houses has been reported to have high soil gas radon
    levels (above 2,000 pCi/L). Thus, the effects of source
    term on SMD systems cannot be assessed.

 •  Permeability of underlying soil.  Except for House
    DW31 in Tennessee (Py90), which visually appeared to
    have relatively loose soil as the crawl-space floor, all of
    the tested crawl-space houses have apparently been built
    on fairly impermeable clay. Thus, an assessment of the
    effects of soil permeability is not possible based upon the
    data.

    A four-pipe site depressurization system gave good re-
    ductions in House DW31, reducing indoor levels from 26
    to 2.2 pCi/L, suggesting that "SMD without a membrane"
    may become possible when soil permeability is good. In
    general, site depressurization would be expected to work
    best when the permeable soil was in a stratum beneath a
    less permeable  layer which would  serve as a cap to
    reduce the short-circuiting of crawl-space  air into the
    system.

    Intuitively, permeable soil on the crawl-space floor would
    be expected to play a role analogous to that of gravel
    beneath the membrane of a SMD system, namely, help-
    ing to  distribute the suction field under the membrane.
    This could potentially  improve  performance of SMD
    systems not having perforated piping  under the mem-
    brane.

 •  CKmatic conditions. There are no  data indicating the
    effects of temperature, winds, barometric pressure, or
    precipitation on SMD performance. Since sub-membrane
    depressurizations  generally  drop below 0.001  in. WG
                                                       63

-------
    within 10 to 15 ft of a suction pipe, crawl-space depres-
    surizations created by thermal and wind effects might be
    expected to more readily overwhelm  a SMD system
    compared to a SSD system, since sub-slab depressuriza-
    tions are often greater. But on the other hand, a crawl
    space is leakier than a basement,  at least when the crawl
    space is vented and unheated; the crawl space might thus
    be expected to become less depressurized than a base-
    ment, and thus provide less of a challenge to the low sub-
    membrane depressurizations.

f)   Mitigation system durability

 •   Radon reduction performance. Little data are avail-
    able on long-term radon reductions achieved by SMD
    systems. The one DTD + SMD installation in a basement-
    plus-crawl-space house in Pennsylvania has consistently
    maintained  basement concentrations below 2 pCi/L ac-
    cording  to winter-quarter alpha-track detector measure-
    ments over the three years since installation (Sc89, Fi91).
    In three of  the four "pure" crawl-space houses in  Ohio
    where an annual alpha-track detector measurement was
    successfully completed following installation, living-area
    concentrations remained  below 2 pCi/L  (and below 1
    pCi/L in two of them) (Ro90). In two of the four crawl-
    space houses in Maryland having  adjoining basements
    (Mes90a), where annual alpha-track measurements were
    successfully completed following installation, the com-
    bined SSD/DTD + SMD systems in these houses main-
    tained concentrations well below  1 pCi/L in both the
    basement and the upstairs living area (Mes90c).

 •   System suctions and flows. System suctions and flows
    remained steady in the one Pennsylvania house having a
    combined basement DTD system plus crawl-space SMD
    system, over the three years following installation (Fi91).

 •   Equipment durability.  No  fan failures have been re-
    ported for SMD systems in any of the EPA study houses.

    In addition  to fan failures,  another concern in SMD
    installations is the integrity of the membrane over  time,
    including ruptures in the plastic itself, and failures in
    membrane seals. Failures could result from foot traffic in
    the crawl space, or to the drying or embrittlement of the
    sealants and plastic over time due to, for example, UV
    effects. Such failures in  membrane integrity could de-
    grade system performance, especially if the breach oc-
    curred near the penetration of a suction pipe through the
    membrane or for some length beside sub-membrane per-
    forated piping.

    Essentially no data exist on the various potential types of
    membrane failures, or on the effects of any such failures
    on SMD performance. Two mitigators report having ob-
    served cases where rodents had caused extensive damage
    to membranes within less than a year of installation, in
    Iowa and Florida (Wi90, Ba92);  however, other mitiga-
    tors report  that such rodent  damage is not a common
    problem (An92, How92, Sh91).
2.4    Performance  of Active Soil
Pressurization Systems
All of the preceding discussion has addressed soil ventilation
systems where the fan is oriented to draw suction on the soil.
As discussed in Section 2.1,  soil depressurization systems
function by a) maintaining a  negative pressure in the  soil
relative to the pressure inside the house, thus preventing soil
gas from entering the house; and b) probably by one or more
other mechanisms, including true soil ventilation, which in-
volves dilution of the soil gas with air from the house and
outdoors. The first mechanism  probably tends to be the more
important in most cases. Hence, soil depressurization systems
tend to function best in cases where system flows are not
particularly high, because such conditions make it easier for
the fan to maintain reasonable  suctions in the soil.

In some cases where flow is unusually high, it can be desir-
able to reverse the fan, so that the fan blows outdoor air under
the foundation. This approach is referred to as active  soil
pressurization. In high-flow cases, it can sometimes be diffi-
cult for a .soil depressurization system to maintain adequate
suctions in the soil, and the second mechanism listed above
(true  soil  ventilation) becomes increasingly important. In
these high-flow cases, soil pressurization is an alternative to
placing a larger fan on a soil depressurization system.

Specific cases where active soil pressurization has been found
to be a potentially viable alternative include

    1) Sub-slab pressurization systems, as an alternative to
       SSD systems, where the underlying soil is a highly-
       permeable,  well-drained gravel (Tu87, Kn90), or a
       highly permeable, highly fissured shale or limestone
       (Br89). Modeling studies also suggest that sub-slab
       pressurization may perform better than SSD when the
       native soil is highly permeable (Ga92). Outdoor air
       leaking into the system through the permeable soil or
       rock interferes with suction field extension by the
       SSD system. In such  cases, switching to sub-slab
       pressurization may be  preferred over the options of
       adding more  SSD  suction pipes or using a higher-
       capacity SSD fan.

       If the house had drain tiles, drain-tile pressurization
       could be considered instead of DTD when the native
       soil is highly permeable. Houses on such well-drained
       soils are less likely to require  drain tiles, although
       drain tiles may still sometimes be present.

    2) Block-wall pressurization systems, as an alternative to
       BWD,  in cases  where  depressurization of the leaky
       block walls draws enough air out of the basement to
       cause back-drafting of combustion appliances in the
       basement (Sc88).

Where the fan is operated to pressurize the soil, the system is
probably operating by two mechanisms. First, the outdoor air
being blown underneath the slab creates a region around the
foundation that is pressurized relative to the surrounding soil,
thus preventing soil gas from  flowing toward the house by
convection. Second, if the underlying soil is sufficiently po-
                                                       64

-------
rous, the air being forced under the slab (or into the walls)
may flow into the soil with a sufficient velocity to overwhelm
diffusive movement of radon toward the foundation. To the
extent that these two mechanisms are not fully effective, any
radon that continues to enter the sub-slab region (or the walls)
by either convection or diffusion will be forced into the house
by the pressurization system, and system effectiveness will be
reduced.

The failure of pressurization systems to match the perfor-
mance of depressurization systems in many cases, as dis-
cussed later, probably results because the pressurization sys-
tems are not able to establish sufficient pressures or air flows
in the  soil to completely prevent convective and  diffusive
radon movement to the foundation. The reason why pressur-
ization systems often do not perform as well under tight-soil,
low-flow conditions is probably that flows of air into the soil
are not sufficiently high to overwhelm radon diffusion under
those conditions.

Under high-flow conditions, both pressurization and depres-
surization systems may have problems maintaining pressures
(or suctions) everywhere under the slab  (or in the wall). The
reason why pressurization systems sometimes perform better
under these conditions may be that pressurization is at  least
partially forcing the radon away from the foundation, whereas
depressurization draws it toward the foundation.

Another criterion for effective pressurization performance, in
addition to highly permeable soil, may be that the soil gas
radon concentration should not be particularly high (e.g.,
perhaps on the order of 1,000 pCi/L or less) (Br89).

Where outdoor air is being blown into a block wall—where
the high flows are being created by the leakiness of the block
walls rather than by high permeability in the native soil—the
block-wall pressurization system may also be working in part
by basement pressurization or basement ventilation with un-
conditioned outdoor air.

Various potential problems have been suggested which might
result from operation of systems in pressure. These include 1)
potential freezing around the foundation in  cold  climates,
when cold outdoor air is blown beneath the slab, potentially
leading to structural problems; 2) potential freezing inside
framed walls in block-wall pressurization systems, especially
in humidified basements, due to the exfiltration of relatively
moist indoor air created by the basement pressurization com-
ponent  of the system; 3)  possible condensation of indoor
moisture on slab or wall surfaces being cooled by the outdoor
air, during cold weather; and 4) potentially increased levels of
termiticides, spores, or soil moisture inside the house, due to
the increased flow of sub-slab gas up into the house created by
the pressurization system. For the most part, insufficient data
are available regarding soil pressurization systems to enable
an assessment of the seriousness of these potential problems.

2.4.1   Active Sub-Slab Pressurization

Active sub-slab pressurization systems have been reported in
five basement houses near Spokane (Tu87), and in four base-
ment houses in New York (Br89, Kn90), where highly perme-
 able native soils resulted in sub-slab pressurization providing
 greater indoor radon reductions than did SSD. One mitigator
 in the Spokane area (Bar90) has also reported that sub-slab
 pressurization systems sometimes provide better reductions
 than do SSD systems.

 In the five Spokane houses, sub-slab pressurization systems
 with one to four pressurization pipes were initially able to
 reduce all of the houses to 3 pCi/L and less in the living area,
 based on short-term measurements. Pre-mitigation levels in
 these house were typically 15 to 50 pCi/L, although one house
 had a pre-mitigation level of 106-141 pCi/L. To achieve 3
 pCi/L in several of the houses, the sub-slab pressurization fans
 being  used had to be operated at full power. For these fans in
 these houses, full power produced a pressure of  1.25 to 2 in.
 WG in the pipes near the slab, and a flow of about 50-200 cfm
 in the total  system. By comparison, when the fans  were
 reversed to operate  in suction, two of these five houses were
 reduced to 3 to 5 pCi/L in the living area,  and the other three
 were reduced only to 7 to 19 pCi/L.

 In one of the houses in New York (Kn90), a two-pipe SSD
 system with a standard 90-watt in-line duct fan  reduced the
 basement concentrations from  a pre-mitigation level of 169
 pCi/L to a 12-day post-mitigation average of 10.5 pCi/L. The
 sub-slab appeared to be being depressurized everywhere, al-
 though in some locations the depressurization was less than
 0.001  in. WG. When the fan was reversed to pressurize the
 sub-slab, basement levels fell to a 6-day average of 3.3 pCi/L.

, In the other three New  York houses (Br89), sub-slab pressur-
 ization reduced indoor radon to about 2 pCi/L, from levels
 which sometimes rose to several hundred pCi/L. SSD or DTD
 had initially been tested in two of these houses. In one of these
 two, the depressurization system had given no radon reduc-
 tions;  in the other, depressurization could reduce levels only
 to 18 pCi/L.

 The effect of sub-slab  pressurization vs. SSD has also been
 tested in 16 basement houses and 7 slab-on-grade houses
 where the underlying soil had low permeability (Ma88, Sc88,
 Tu89, Du90, Fi90, Py90). In  none of these cases was the
 performance of the pressurization system better than that of
 the SSD system. Occasionally, the performance  was compa-
 rable, but most commonly, the pressurization system gave
 much poorer reductions.

 The 16 basement houses  were in the states of  New Jersey
 (Tu89, Du90), Pennsylvania (Sc88), and Tennessee (Ma88,
 Py90). All but three of these houses had one or two sub-slab
 ventilation pipes; these other three had four to five pipes. The
 communication beneath the basement slab ranged from good
 to poor among these houses. The native soil was commonly a
 low-permeability clay. Half of the basements had adjoining
 slab-on-grade or crawl-space wings; where there was an ad-
 joining slab, there  was often (but not always)  at least one
 ventilation pipe inserted beneath the adjoining slab.

 With the systems in these 16 basement houses  operating in
 suction 14 of the houses (almost 90%) were reduced to 4 pCi/
 L and less; and 10 (about 60%) were reduced to 2 pCi/L and
 less in the basement (from pre-mitigation levels ranging from
                                                        65

-------
12 to 156 pCi/L). By comparison, when the fans were re-
versed to pressurize the sub-slab, the results were dramati-
cally poorer only two of the houses (less than 15%) were
reduced below 4 pCVL; and only two (these same two) were
reduced to 2 pCi/L and less.

Operation in pressure caused post-mitigation levels to in-
crease by 4 to 40 pCi/L in all but two of the 16  houses,
compared to the post-mitigation levels when the system oper-
ated in suction. In those two, the increase was 0 to 1 pCi/L. In
none of these 16 houses did operation in pressure result in
reduced post-mitigation concentrations compared  to opera-
tion  in suction. But in all  cases, operation in pressure did
provide some reduction in indoor levels compared to pre-
mitigation concentrations; i.e., in no case were post-mitiga-
tion concentrations greater than the pre-mitigation levels dur-
ing operation in pressure.

Two of these studies (Sc88, Py90) reported increases in soil
gas odors within the house when the SSD fan was reversed to
operate in  pressure. One of the studies (Py90) measured
roughly five- to ten-fold increases in the termiticide aldrin in
the basement air of one house when the fan was reversed,
although aldrin levels  appeared to have returned to their
original values after the system had operated in pressure for
10 weeks.

The  seven slab-on-grade houses with  low soil permeability
where the SSD fans were reversed were all in Ohio (Fi90). All
of these systems had one or two ventilation pipes. There was a
good layer of aggregate beneath the slab in all of these houses,
although communication in some of them was interrupted by
sub-slab  forced-air supply ducts. The underlying native soil
was  a low-permeability clay.

With the systems in these seven slab-on-grade houses operat-
ing in suction, all of the houses were reduced below 4 pCi/L,
and  six (about 85%) were reduced below 2 pCi/L. With the
fans reversed to  pressurize the sub-slab, reductions were
dramatically poorer: only three of the houses were still below
4 pCi/L, and  only one was  below 2 pCi/L. Operation in
pressure caused post-mitigation levels to increase by 1 to 13
pCi/L, compared to the post-mitigation levels during opera-
tion  in suction. But again, the sub-slab pressurization system
did provide some reduction in radon, compared to pre-mitiga-
tion  concentrations.

In summary, sub-slab pressurization has consistently given
better reductions than has SSD in houses built on well-drained
gravel soils and on highly fractured rock. Sub-slab pressuriza-
tion has consistently given poorer reductions than has SSD in
houses built on low-permeability soils, regardless of whether
or not there is a good layer of aggregate beneath the slab.

There is insufficient experience with active sub-slab pressur-
ixation systems to permit a definitive review of the effects of
the various house design and operating  variables, and the
various system design and operating variables, analogous to
that presented in Section 2.3.1 for SSD. Some of the informa-
tion  presented in Section 2.3.1 would be applicable to, or
could be adapted to, pressurization systems.
Some results have been presented regarding the durability of
the five successful pressurization systems in Spokane (Pr89).
These follow-up measurements included quarterly alpha-track
monitoring during a 2-year period after installation, and sys-
tem inspection and flow measurements after 2 years of opera-
tion.
All five of the Spokane houses had at  least one quarterly
alpha-track above 4 pCi/L in the living area (compared to
short-term post-mitigation measurements indicating about 2
pCi/L and less immediately after installation). In all cases,
homeowners reported that the system had been turned off at
least part of the time during mild weather, undoubtedly con-
tributing to the elevated levels. In one house, a system failure
resulted in significant vibration noise which resulted in the fan
being turned off for most of the 2-year period, so that living-
area concentrations remained at pre-mitigation values (around
30 pCi/L). In a second house, a similar failure resulted in
significant air leaks between the fan and the ventilation pipes,
although the system was left operating. In this second house,
living-area concentrations rose to about 30  pCi/L  toward the
end of the 2-year period (compared to a pre-mitigation value
of 140 pCi/L), after having remained between 2 and 4 pCi/L
for the year and a half prior to the failure. The problems with
the installations in these latter two houses  do not reflect an
inherent problem with the durability of sub-slab pressuriza-
tion systems, but rather, reflect the formative stage of radon
mitigation system design at the time that these systems were
installed (Winter 1985-86).

Pressure and flow measurements in the five Spokane systems
after 2 years of operation showed that system pressures had
increased since installation, and flows had decreased, in all
cases. Inspection in two of the houses showed that dust and
debris had accumulated on the surface of the soil under the
slab, at the point where the ventilation pipe terminated just
below the slab. Presumably, this debris had  been entrained in
the fan inlet, which had no screen, and blown into  the system
piping. Removal of this debris in the two  houses resulted in a
substantial recovery toward original pressures and  flows.  The
apparent increases in  indoor radon levels  in these houses
could be due in part to the reduced flows caused by  this debris,
which may have plugged the interstices between soil particles.
It was recommended that future sub-slab pressurization  sys-
tems be installed with cleanable filters on the fan inlet, and
with gauges to alert the homeowner if system pressure is
changing (Pr89).

In summary, the fact that some radon measurements exceeded
4 pCi/L in all five of the Spokane houses  over the 2 years
following installation does not necessarily reflect an inherent
problem with the durability of sub-slab pressurization  sys-
tems. The elevated levels may result from a combination of a)
correctable problems in the  design of these particular early
installations; and b) homeowner intervention, in turning the
systems off. Accordingly, a definitive statement  cannot be
made from these limited results regarding  the durability of
sub-slab pressurization systems being designed today. If fil-
ters are required in order to prevent debris buildup in the
system, the homeowner is going to have to be alert  to need for
continuing maintenance of this filter.
                                                        66

-------
Because of the limited applicability of, and experience with,
sub-slab pressurization  systems, the emphasis in this docu-
ment is on SSD (Section 4). Sub-slab pressurization is ad-
dressed more briefly, in Section 9.

2.4.2 Active Block-Wall Pressurization

Active block-wall pressurization systems have been reported
in three basement houses in Pennsylvania (Houses 2, 5, and 9
in Sc88). In Houses 2 and 9, the block-wall system was of the
"baseboard duct" design, and each house had two fans blow-
ing into the duct; House 5 had an "individual-pipe" configura-
tion, with  one fan. All  three  houses had a very high source
term, and pre-mitigation concentrations in the basement rang-
ing from 110 to 533 pCi/L.  Li  each case, BWD had been
attempted  first,  but had been converted to a pressurization
system because of back-drafting of wood stoves and fire-
places.

Unfortunately, in none of the Pennsylvania houses was the
same block-wall ventilation  system tested back-to-back  in
suction and pressure, to enable a direct comparison of the two
approaches. In all cases, some other system modification (use
of different fans, additional sealing) accompanied the reversal
of the fan. However, the radon reductions achieved in suction
and  in pressure were generally comparable, with pressure
seeming to give slightly greater reductions  in two  of the
houses, and suction slightly greater in the third house.

In only one of the three houses, House 2, was the block-wall
pressurization system able to reduce basement concentrations
below 4 pCi/L (when supplemented by a well water treatment
system), based  upon short-term measurements immediately
after installation. Houses 5 and 9 achieved basement concen-
trations of 5 and 7 pCi/L, respectively, based upon the short-
term measurements, although about 3 pCi/L of the residual
level in House 9 might have been due to well water.

Winter-quarter;alpha-track detector measurements in these
houses during the three winters following  installation have
shown basement and living-area concentrations holding rela-
tively steady at  values above  4 pCi/L (Fi91). In House 2, the
average basement reading over the three winters has been 4.3
pCi/L; in House 5,4.8 pCi/L; and in House 9,11.5 pCi/L.

These concentrations represent substantial percentage reduc-
tions from the high pre-mitigation levels, but reveal that some
entry routes are not being adequately treated by these systems.
Sub-slab pressure field extension measurements in House 2
indicated that the sub-slab in that house was apparently being
effectively pressurized by the baseboard duct system. (Sub-
slab measurements were not completed  in  the other two
houses.) Presumably, the flows that could be maintained by
the systems were not adequate to sufficiently dilute the high
radon  concentrations around the foundation  of these high-
source-term houses, and were not adequate to prevent some
radon from entering the wall void network. If the radon could
reach the sub-slab  and  wall void regions, the pressurization
system would force it up into the house.

None  of these  three houses were  optimal for block-wall
treatment, due to inaccessible top voids in the block wall, and
due  to fireplace structures.  As discussed regarding BWD
systems in Section 2.3.4, neither of the two stand-alone BWD
systems in this Pennsylvania project that were installed in
non-optimal houses achieved basement concentrations below
4 pCi/L, either.

In summary, from the very limited results available with
block-wall pressurization systems, it is not apparent that block-
wall pressurization is either less or more effective than BWD.
Neither approach appears to be a reliable method for reducing
levels below 4 pCi/L when applied as a stand-alone method,
especially not in houses which have high pre-mitigation levels
and  which  are not optimal for block-wall treatment. BWD
appears to  have an advantage in that it can be used as a
supplement to SSD systems.

2.5   Performance of Passive Soil
Depressurization Systems
All of the preceding discussion of soil depressurization sys-
tems has addressed the case where an electrically powered fan
is being used to draw suction on the soil. It is the use of the fan
that results in these systems  being referred to as active soil
depressurization systems. The fans commonly used are ca-
pable of developing suctions of about 1 in. WG and higher in
the system piping. With that amount of suction in the system
piping, the chances are improved that the suction field will
extend beneath the entire slab and around the foundation, and
that the conservative goal will be met of maintaining about
0.025 to 0.035 in. WG depressurization everywhere beneath
the slab during mild weather with exhaust appliances off. (See
Section 2.3. Ib, Operation of central furnace fan, and exhaust
fans, and Section 2.3.1e, Climatic conditions).

Nominally, soil depressurization systems can also be operated
in the passive mode, i.e., without a fan. Passive systems rely
on natural  phenomena to  develop the suction in the stack.
These natural phenomena include a) thermal effects, created
when the stack temperature is warmer than the outdoor air,
causing the soil gas inside the stack to rise; and b) wind
movement  over the roofline, which creates a low-pressure
region over the roof.

Passive soil depressurization systems may operate by two
mechanisms. One mechanism is soil depressurization, as with
active systems, using the fairly low naturally-induced suctions
listed in the preceding paragraph. The second possible mecha-
nism is development of a  "pressure break" in the aggregate
bed beneath a slab or in the region under a SMD membrane,
i.e.,  an equalization of sub-slab or  sub-membrane pressures
with outdoor pressures, providing a buffer isolating the soil
from the depressurized house. Such a pressure break could
reduce the  house-induced suction  on the soil, and hence
amount of radon drawn out of the soil.

In passive  systems, the  exhaust stack generally extends up
through the house indoors, providing a direct route to the roof
from the suction pipe penetrations  through the slab, sump
cover, or crawl-space membrane. To the extent that passive
systems rely upon the first mechanism above (soil depressur-
ization), the important thermal contribution to the passive
suction would be largely lost during cold weather if the stack
                                                        67

-------
extended up outside the house or through an unheated garage,
as is sometimes done with active systems. The thermal effects
in passive soil depressurization systems are exactly analogous
(and similar in magnitude) to the thermal stack effect which is
drawing soil gas into the house. The difference is that the
stack is providing the soil gas with a direct "thermal bypass"
up to the roof.

The primary problem with passive operation is that the natural
suctions produced in the stacks are quite low. The thermally-
induced depressurization that would be created in a two-story
stack during cold weather would be on the order of 0.015 in.
WG, the same as the  thermally-induced depressurization in
the basement. Depressurizations that have been measured in
passive stacks have consistently been lower than 0.1 in. WG,
and often less than 0.05 in. WG (Gi90). These passive depres-
surizations  are  10  to 100 times lower than those usually
maintained in active system piping. With such low suctions,
passive systems can  have  difficulty maintaining  adequate
depressurizations everywhere beneath the slab or membrane,
to the extent that passive systems rely on the soil depressur-
ization mechanism.

To the extent that passive systems rely on the pressure break
mechanism, an analogous problem may exist, i.e., difficulty in
buffering the soil from the house at all locations using only a
single pressure-relief pipe. From the standpoints of both the
soil depressurization mechanism and the pressure break mecha-
nism, perhaps connection of the passive stack to a network of
perforated piping under the slab or the SMD membrane would
aid in the extension of the suction field or of the pressure-
break buffer. However, there are no definitive data quantify-
ing the benefit of such sub-slab or sub-membrane piping.

An added concern is that the performance of passive systems
may be variable, changing  as the temperatures and winds
change, varying the natural suction in the stack.

One key advantage of passive systems, if they perform well, is
that they avoid the need for homeowner maintenance of a fan.
The risk is thus eliminated that house occupants  might be
subjected to high radon exposures over a long  period if the
homeowner fails to notice or repair a malfunctioning fan, a
potentially  important  consideration. However,  passive sys-
tems will not necessarily be  maintenance-free;  they may
require homeowner attention in maintaining  seals of slab
cracks/openings (since system flows/suctions will be too low
to tolerate much short-circuiting), in ensuring that the stack
outlet does not get restricted by debris such as leaves (since
flows may be too low to blow debris away), and in verifying
that system suctions/flows  are being maintained. Another
advantage of passive  systems  is that they avoid any noise
associated with fan operation  or with the rapid release of
exhaust gas from the stack outlet. (Generally, noise associated
with fan operation in active systems can  be  significantly
reduced by proper mounting of the fan and piping.)

Passive systems will  also avoid the cost of electricity for
operation of a  fan, and, because of their low flows, will
significantly reduce the heating/cooling penalty associated
with active systems. The maximum savings in electricity and
heating/cooling bills resulting from these factors will be on
the order of $7.50 per month (He91b, He91c), an amount
which  many homeowners will not notice,  but an amount
which  would add up  over time and which could have an
impact on national  energy consumption. Fan  maintenance
costs would be eliminated. Eliminating the fan would also
reduce system installation costs by an amount  equal to the
cost of the fan and the associated wiring, although there could
be offsetting installation cost increases due to increased foun-
dation sealing requirements, efforts to improve the extension
of the weak suction field, and the need for an interior stack.

In summary, there are some advantages associated with pas-
sive systems. However, these advantages will be achieved at
the expense of reduced radon reduction performance com-
pared to active systems, based upon data available to date. The
reduced performance of passive systems relative to active
systems likely results because any benefits resulting from the
passive pressure break mechanism are not sufficient to com-
pensate for the greatly reduced role of the soil depressuriza-
tion mechanism caused by the generally weak passive suction
field.

From a practical standpoint, passive soil depressurization will
likely prove to be applicable only in cases where sub-slab
communication is very good.  Good communication will be
necessary to permit reasonable extension of the weak passive
suction field, or for the development of a reasonable pressure
break with a single passive stack.  A tight slab (or SMD
membrane) will also likely be required. Significant air leak-
age into  the sub-slab region from  any source will likely
overwhelm the weak passive suction field; significant open-
ings between the sub-slab region and the house would tend to
defeat the pressure break.

On this basis, it would be expected that passive soil depressur-
ization systems will likely work best in houses where there is
good aggregate under the slab, and where the slab is tight or is
accessible for sealing. The presence of an interior drain tile
loop might be expected to aid in the distribution of the weak
suction fieldj or in development of the pressure break; how-
ever, there are no definitive data demonstrating that such a
drain tile loop will in fact be beneficial. Passive soil depres-
surization will not be applicable in cases where high system
flows would be expected,  such as DTD or SSD systems in
houses having badly cracked slabs than cannot be effectively
sealed, or such as BWD systems. Passive operation of a DTD
or SSD system might be expected to perform best when the
foundation wall is constructed of poured concrete, so that the
system would not have to address radon entry  into (and air
flows from) hollow block foundation walls.

Under  the favorable conditions defined above,  passive sys-
tems can sometimes be sufficient to reduce slightly elevated
houses below 4 pCi/L. However, the performance in a given
house  cannot be predicted prior to installation. Also, the
performance may be expected to vary over time, as outdoor
temperatures and winds change (varying  the suction in the
passive stack) and as exhaust appliances are operated (poten-
tially overwhelming both  the low passive suction  and the
pressure break). Thus, whenever a passive system is installed,
the homeowner should be prepared to make frequent measure-
ments in the house for a period of time after installation, in
                                                        68

-------
order to understand the conditions (such as warm tempera-
tures, low winds, and appliance operation)  that result in the
passive system being overwhelmed. The mitigator or owner
must also be prepared to install or activate a fan on the system
if these measurements show that a fan is necessary.

The available data base on the performance of passive soil
depressurization systems is very limited.

In one project where passive systems were retrofit into exist-
ing houses in Maryland (Gi90), testing was conducted on ten
passive SSD installations (with no perforated piping  beneath
the slab),  two DTD systems, and two  SMD systems.  The
nature and extent of the drain tiles in  the two sump/DTD
systems are unknown. Both of the SMD systems included a
network of perforated tiles beneath a completely sealed mem-
brane. The aggregate beneath the slabs was  good in some
cases, and poor or uneven in other cases.

The passive systems in these 14 existing Maryland houses
typically provided  moderate radon reductions (30 to 70%),
although reductions ranged from  a low of zero to a high  of
90%. In only two of the houses (047 and 079) was the passive
system adequate to reduce indoor levels below 4 pCi/L.

The two Maryland houses giving the best reductions (Houses
004 and 079, achieving up  to 80 to 90% reduction) had SSD
systems with no sub-slab piping, but had good aggregate and
poured concrete foundation walls. Two houses having poor/
uneven aggregate (054 and 074) gave poor performances with
passive SSD systems (0 to 40%). However, good communica-
tion was not sufficient, by itself, to ensure high radon reduc-
tions. Houses with poured foundations consistently gave bet-
ter  reductions  than those with block foundations, with the
highest reductions in block-foundation houses being 50%.

Of the two houses having sumps with at least partial drain tile
loops to help distribute the suction, one  (061) achieved only
low to moderate reductions (15 to 50%), while the other (096)
did somewhat better (35 to 75%). Of the two houses having
passive SMD systems, only one (047) gave reasonably good
reductions (20 to 70%), undoubtedly due in part to  the fact
that the "passive stack" in this house consisted, of the furnace
flue, which increased the passive suction when the  furnace
was operating.

The passive suctions measured  in  these Maryland stacks
ranged from zero to 0.1 in. WG. The passive SSD  systems
were found to create measurable sub-slab depressurizations as
far  as 40 ft away from the suction pipe in one case, but more
commonly, the measurable sub-slab suction did not appear to
extend more than about 20 ft from the pipe.

When these passive soil depressurization systems were acti-
vated using a standard 90-watt in-line duct fan, radon reduc-
tions increased to 90-99% in all cases, reducing indoor levels
to 1 to 2 pCi/L or less (Gi90). In no case did activation of the
system fail to provide significant additional reductions be-
yond those achieved with the passive system.

Another study addressed two  newly constructed basement
houses in  the Washington, D.C.,  area, which  had been built
with a good layer of aggregate, with perforated piping beneath
the slab, and with slab sealing at the wall/floor joint and at
slab penetrations, in order to facilitate radon mitigation. Both
houses had poured concrete foundation walls, aiding the per-
formance of the passive system. A passive stack drawing
suction on the sub-slab piping in these houses provided aver-
age radon reductions of 75  to 90% (Sau91a).

These reductions in the two Washington, D.C. houses were
obtained in both summer and winter, indicating  reasonable
performance even in  warm  weather when  thermal effects
would not be contributing to the natural suction. In one house,
the passive stack was sufficient by itself to reduce the house
below 4 pCi/L, from pre-mitigation levels as high as 20 pCi/L
during the winter. In the second  house, the passive system
reduced winter levels from 29 to 7.5 pCi/L; activation of the
system reduced levels below 1 pCi/L. In the summer,  the
passive stack by itself reduced levels from 2-4 pCi/L to below
1 pCi/L in both houses. The low radon levels achieved during
mild weather (when the passive depressurization mechanism
might be expected to be playing only a minimal role) could be
a commentary on the importance of the pressure break mecha-
nism, as well as the naturally reduced driving forces existing
during mild weather.

While moderate to high radon reductions were achieved with
the passive stack in both houses, indoor levels were subject to
occasional spikes, presumably due to basement depressuriza-
tion caused by weather effects and forced-air furnace opera-
tion.

As an  extension of the previous study, measurements were
completed in 15 newly-constructed houses in the Washington,
D. C., area having passive  stacks  (Sau91b). As with the two
houses discussed in the preceding paragraph, these houses all
had a  good aggregate layer, perforated piping beneath  the
slab, and slab sealing. Based upon 1- to 2-week  continuous
radon measurements with and without the passive stack capped,
the passive systems were providing reductions ranging from 9
to over 90%, with average reductions of 64 to 70%. Radon
concentrations were reduced from an average of 8-18 pCi/L,
to an average of 2.5-6 pCi/L. Eight of these 15 houses were
reduced below 4 pCi/L by the passive system, although a
couple of these houses had pre-mitigation levels below 4
pCi/L  to begin with. Operation of a mitigation fan on six of
these houses provided substantial additional reductions  be-
yond those achieved with the passive system, reducing levels
below  1 pCi/L in all six  cases.  Again, it is believed that
basement depressurization  by forced-air return ducts in the
basement contributed to overwhelming of the passive sys-
tems.

Of the 15 houses in the preceding study (Sau91b), the passive
system provided reductions of at least 50% in all but 5 of
them. All five of these less successful houses exhibited some
problem potentially explaining the reduced passive perfor-
mance. In three of these houses, sub-slab communication was
found  to be lower than would have been expected had the
specified sub-slab aggregate been properly installed. In the
other two, the passive stack was in an unheated garage, thus
reducing the contribution  of thermal effects to the passive
suction.
                                                       69

-------
In another project, passive SSD or DTD was tested in two new
basement houses in Pennsylvania which had been built with a
good aggregate layer, capped sumps with an interior drain tile
loop beneath the slab, and with sealing of the wall/floor joint
(Br91a). Both houses had poured concrete foundation walls.
In the first house (House PA1), a single SSD pipe penetrated
the slab and was not directly connected to the sump or drain
tiles. The passive stack was routed through the roof of a one-
story wing of the house, adjoining the two-story wing over the
basement (Br92). Passive operation of this system in House
PA1 provided essentially no reduction in the basement radon
concentrations (8-10 pCi/L), based upon back-to-back 1-week
measurements with and without the stack capped. Activation
of the system reduced levels to 1-2 pCi/L.

The  system in House PA1 was  later modified to direct the
stack straight up through the two-story wing of the house
above  the basement (Br92).  With this new configuration,
passive operation of the system reportedly reduced basement
levels to about 2 pCi/L, with suctions of about 0.005 to 0.01
in. WG under the slab.

In the second Pennsylvania house (House PA2), where the
system drew suction directly on  the sump, passive operation
reduced basement levels from 5 pCi/L to 1.5 pCi/L, again
based upon 1-week measurements (Br91a).

In addition to these recent research results, one mitigator has
reported occasionally obtaining moderate reductions applying
passive SMD commercially in about 20 crawl-space houses,
reportedly reducing indoor levels from 8 pCi/L down to below
3 pCi/L in "pure" crawl spaces with poured concrete founda-
tion walls (K192). In these particular installations, the mem-
brane was  sealed well, with  a length of perforated piping
beneath. The passive "stack" from the sub-membrane piping
penetrated the crawl-space band joist and exhausted at grade
level. This type of stack forfeits the passive suction that would
be developed by the thermal stack effect, since the stack does
not rise through the heated space. Rather, it relies upon wind-
induced natural suction and on the pressure break mechanism,
along with any reduction resulting purely from the sealing of a
membrane over the crawl-space floor.

Passive soil depressurization systems have also been tested in
a number of earlier remedial efforts in the U.S. and Canada
(Vi79. Ar82, Ta85).

Reference Ar82 summarizes results from passive systems that
were retrofit into a number of existing Canadian and U.S.
houses contaminated with uranium mill tailings. Radon reduc-
tions of 70 to  90% were reported in many of these houses.
However, the interpretation of these reductions, in terms of
the actual performance of the passive system, is complicated
by the fact that the reported reductions often  also include the
effects of other mitigation measures that were implemented
simultaneously. These other measures included removal of
mill tailing source material from under the slab, and various
sealing efforts such as trapping of floor drains. An additional
limitation is that the pre- and post-mitigation radon and work-
ing level measurements were likely obtained by multiple grab
samples in many of the cases, thus providing a less rigorous
measure of system performance.
In another project involving new residential construction in
uranium mining and processing communities in Canada (Vi79),
very extensive sub-slab perforated piping networks were in-
stalled beneath the slabs of a number of houses during con-
struction, apparently in a good bed of aggregate. In 18 of these
houses, a vertical stack extended from this network up through
the interior of the house to the roof. Passive depressurization
of the networks in these 18 houses reportedly gave satisfac-
tory reductions during the winter, maintaining indoor levels at
0.02 WL and less. However, during mild weather, when the
thermal contribution to the natural suction in the stack would
be reduced or eliminated, performance of the passive systems
reportedly degraded. As a result, based upon multiple grab
sample measurements, ten of the houses averaged above 0.02
WL over the entire measurement period, despite the extensive
piping network. Under these conditions, the systems had to be
activated; each of the 18 houses averaged below 0.015  WL
after activation.

A major effort was made to retrofit a passive system into one
existing house having a very high source term (resulting from
naturally occurring radium in the underlying soil and rock)
and having block foundation walls (Ta85). The existing slabs
for both the basement and the adjoining slab below grade in
this house were torn out Some of the underlying soil and rock
was removed, and replaced with aggregate. Loops of perfo-
rated drain tile were placed around the perimeter of each slab,
and beside the footings for an interior load-bearing wall in the
basement. Each loop connected to a passive stack which rose
through the house to the roof. The slabs were re-poured, with
efforts to make the slabs as tight as possible.  Based upon
periodic grab samples over a several-month period after miti-
gation, the radon levels were reportedly reduced by greater
than 99%, from about  13.5 WL before mitigation to below
0.02 WL. One significant spike  (to 1 WL) was measured
during one of the grab sampling campaigns, and a small fan
was  installed in each of the vent stacks  to boost suction,
bringing concentrations back down to acceptable levels. The
fan on one of the two vent stacks is still operated frequently by
the homeowner. Grab samples do not reveal the variations in
radon levels, or the average levels, mat exist between sam-
pling campaigns.

In summary, the available results with passive soil depressur-
ization systems suggest that

   -  Passive  SSD or  DTD systems have commonly been
      shown to give moderate indoor radon reductions, some-
      times as high as 70 to 90% if sub-slab communication
      is good and the slab is tight. However, passive perfor-
      mance is not predictable, and may vary with time.

   -  A given soil depressurization system in a given house
      always gives much higher reductions when operated in
      the active mode, with a fan, than when operated pas-
      sively.

   -  Good sub-slab communication is mandatory for good
      passive  SSD or  DTD performance. However, good
      communication  is not sufficient by itself to ensure
      good performance.
                                                       70

-------
Passive SSD and DTD have often provided their best   Because of the apparent limitations and uncertainties in the
reductions in new construction where steps have been   performance of passive systems, the discussion of system
taken to tighten the slab during construction  (e.g.,   design and installation in the subsequent sections will focus
caulking the wall/floor joint).  Such slab tightening   on active  systems. However, much of the information on
should be conducted in new houses if it is desired to   design and installation of active systems can be readily adapted
improve the chances that a passive system will perform   for use with passive systems, with the guidance given above.
well. However, there are no definitive results quantify-
ing the benefits of such slab sealing.

Best results with passive SSD or DTD systems have
consistently been obtained where the foundation wall is
poured concrete, rather than hollow block.
                                                  71

-------

-------
                                            Section 3
               Pre-Mitigation Diagnostic Test Procedures
                      for Soil  Depressurization Systems
3.1    General

3.1.1  Purposes of Pre-Mitigation
Diagnostics

The primary purposes of diagnostic testing prior to the design
and installation of a radon reduction system are to:

    a)  improve the radon reduction performance of the sys-
       tem ultimately installed; and

    b)  reduce the cost to the homeowner of achieving a given
       level of radon reduction.

If the diagnostic testing does not have a reasonable likelihood
of improving performance and/or reducing costs, there is no
purpose in conducting it.

Pre-mitigation diagnostics can be used to help select which
mitigation technology can best be applied in a specific situa-
tion (e.g., whether basement pressurization or a house ventila-
tion approach might be more suitable than ASD). Once the
decision is made to install an ASD system, the diagnostics can
aid in the design of the ASD system.

3.1.2  Diagnostic Tests That Can Be
Considered

Once a mitigator is familiar with the housing characteristics in
his/her area, and the way in which these characteristics influ-
ence radon entry  and the performance of ASD (and other)
Systems, it will often be possible to select and design an ASD
system with a minimum of pre-mitigation diagnostics.

A pie-installation visual inspection will be the one diagnostic
that will generally be required prior to  final selection and
design of the system in all cases. In many cases, this will be
the only pre-mitigation diagnostic test necessary.  In cases
where the sub-slab communication is poor, uneven, or uncer-
tain, sub-slab suction field extension testing may sometimes
be cost-effective, using a vacuum cleaner or a portable ASD
fan test stand to generate the suction field, in order to facilitate
efficient  determination of where suction  pipes should be
located. Especially where suction field diagnostics are needed,
grab sampling (or  "sniffing") to determine radon concentra-
tions at locations under the slab or around potential soil gas
entry routes may sometimes be helpful in  selecting suction
locations.
Most mitigators will rarely have  to utilize pre-mitigation
diagnostic tests other than those listed in the preceding para-
graph, except in special cases.

A more complete summary of the diagnostics that can be
considered, and the cases under which they might be useful, is
given below. They  are listed in the approximate order of
frequency with which they might be used. While the discus-
sion here focuses on their use as /we-mitigation diagnostics,
some of them can sometimes be used during  mitigation or
after mitigation (see Section 11).

 •  Visual inspection. This "diagnostic test", in one form
    or another, will be required prior to mitigation in every
    house. Most major questions, such as whether  a house is
    a suitable candidate for ASD (or for other techniques),
    whether the house will be relatively simple or relatively
    difficult to mitigate, where suction pipes might be located
    and how the exhaust piping might be routed, etc., can be
    at least partially addressed by the visual inspection, with-
    out any measurements.

 •  Suction field extension. This is the second most com-
    mon  diagnostic  technique for ASD  systems, after visual
    inspection. It is necessary only if: a) the visual inspection
    and the mitigator's prior experience do not provide suffi-
    cient basis for  a reasonable judgement regarding the
    likely sub-slab communication; or  b) inspection and/or
    prior experience suggest that communication will be poor
    or  uneven, but  do not suggest the logical number and
    location of SSD suction pipes. A relatively simple mea-
    surement of sub-slab  depressurizations created at a few
    remote  test holes  by  an industrial  vacuum cleaner can
    help  confirm whether a house has relatively good or
    relatively poor  communication. A more  complex test
    procedure, involving a greater number of test holes, would
    better quantify how far the suction field from a single
    SSD pipe  might extend, and where discontinuities in
    communication exist, to aid in a more rigorous selection
    of pipe  number and location.

    Rather  than a vacuum cleaner, some mitigators might
    elect to use a portable ASD fan test stand to generate the
    suction field. The portable ASD fan would give a more
    rigorous and easily-interpreted indication of how far the
    eventual system suction field will extend, and of the
    flows that can be expected, compared to a vacuum cleaner.
    Where  the flexibility exists, some professionals some-
    times use an ASD fan mounted on the initial suction pipe
                                                     73

-------
installed through the slab during installation, proceeding
on a "design-as-you-instaU" basis. Suction field exten-
sion testing is also common as a post-mitigation diagnos-
tic tool (with the ASD system operating), to assess why
an installed system is not providing the desired perfor-
mance.

Sub-slab flows. As an extension of the suction field
extension test, one could measure the flows developed in
the vacuum cleaner nozzle (or in the piping to the por-
table fan test stand) at various suctions induced by  the
vacuum under the slab. The results could help quantify
the preferred performance curve for the fan to be installed
on the ASD system.

Such sub-slab flow measurements will probably not be
necessary for many mitigators in most houses. An assess-
ment of whether a standard moderate-suction/high-flow
fan will be appropriate, or whether a high-suction/low-flow
fan would be preferred in a given house, can generally be
made based upon the  mitigator's experience and  the
suction field extension results, without the more rigorous
sub-slab flow measurement at multiple suctions. At most,
if suction field extension tests are being performed, a
mhigator might measure the flow in the vacuum cleaner
at the one suction used for that testing, to determine
whether the sub-slab  flows are high or low. If the diag-
nostic vacuum cleaner is not set up to enable flow mea-
surements, some researchers suggest simply listening to
the sound of the vacuum motor as a qualitative indicator
of whether it is moving a lot of air, or only a little (Br9 Ib,
Bro92).

Radon  grab sampling and "sntffing." "Measure-
ment of radon concentrations near potential entry routes—
e.g., under the  slab near slab cracks and openings, or
inside hollow-block foundation walls—can suggest  the
relative  importance of these potential entry routes, and
thus can help guide ASD design. For example, especially
where sub-slab  communication is not good, SSD pipes
might be located toward those cracks/ openings  having
the highest sub-slab concentrations. Block walls having
high radon levels in  the cavities would be those most
likely to warrant a B\V"D component complementing  the
SSD system. Such grab sampling  will likely be of value
primarily when  sub-slab communication is  not good, or
when the radon levels in the soil gas are very high, and
when the location of the ASD suction pipes thus becomes
of increased importance. Poor-communication houses are
the houses  where sub-slab suction field extension mea-
surements are most likely to be helpful, in which case test
holes would be being drilled through the slab for suction
field testing; it would make sense to expand the suction
field measurements to include grab sampling through the
slab holes, and through utility openings in the block
walls.

Well water  radon  analysis. If it is suspected from
experience in the mitigator's service area that  radon in
well water might be a significant contributor to the indoor
radon levels, and if the homeowner has not had a water
measurement conducted, it can be desirable to conduct a
water measurement. This measurement would permit a
realistic assessment of how much airborne radon is likely
being contributed by the well water, recognizing that an
ASD system could not address the water-related contri-
bution to the indoor air levels. Mitigators working prima-
rily in houses served by municipal water supplies, or in
areas where well water radon concentrations are typically
low, would rarely (if ever) find it necessary to make water
measurements. Mitigators working in areas where very
high concentrations of radon in water sometimes occur
might want to have a water measurement made, if the
homeowner has not already done so, before guaranteeing
that a post-mitigation level of 4 pCi/L will be achieved
with an ASD system alone.

A gamma measurement against the well water pressure
tank, against the water piping, or at a toilet bowl, is a
simple screening measurement that can be made to sug-
gest whether a more rigorous water analysis is warranted.

Flux measurements or gamma measurements. Mea-
surements of radon flux from interior building surfaces
(or, more conveniently, measurements of gamma radia-
tion near the surfaces, if a gamma meter is available) are
needed only in cases where there is reason to suspect that
building materials may be an important contributor to
indoor radon levels. These measurements would qualita-
tively indicate whether the building materials might be an
important radon source, thus limiting the effectiveness of
ASD (which can only address the soil gas source). Most
mitigators will rarely, if ever, have the need to conduct
these diagnostics. Building materials will be a significant
contributor only in unusual cases where, e.g., uranium
mill tailings, or natural  materials with very high radium
contents, have been used as fill around the house or as
aggregate in the concrete. Unless there is some basis for
suspecting building materials to be a source, flux or
gamma measurements would more likely be conducted as
post-mitigation trouble-shooting diagnostics, to determine
why radon levels are still elevated, rather than  as
pre-mitigation diagnostics.

Pressure differential measurements  across the
house shell (above or below grade). Sometimes, usually
for research purposes, pressure measurements are made
between indoors and outdoors above grade. Such mea-
surements will rarely be useful in designing commercial
installations, except in assessing the threat of combustion
appliance back-drafting. In addition, pressure measure-
ments can be made across the slab (with no suction being
drawn under the slab). These below-grade measurements
are conducted in conjunction with the suction field exten-
sion testing discussed earlier.

Pressure differential measurements might be used to aid
in interpreting sub-slab  suction field extension test data.
They could indicate whether the existing driving  force
during the suction field diagnostics were much less than
the estimated  maximum driving force  that  would  be
expected during cold weather with exhaust appliances
operating.  If the existing driving force during the diag-
nostics were much less than the estimated expected maxi-
                                                   74

-------
mum, the additional sub-slab depressurizations needed to
compensate for the higher driving forces could then be
factored into the ASD design. Pressure differential mea-
surements across the shell could also be used to  assess
house depressurization with various house appliances in
operation (e.g., a central furnace fan or a whole-house
exhaust fan), or under different weather conditions, in an
effort to quantify the actual maximum driving forces that
might exist in a particular house (rather than relying on
the conservative rule-of-thumb values, 0.025 to 0.035 in.
WG for combined thermal and exhaust  appliance  ef-
fects).

For the above purposes, the preferred pressure differen-
tial to measure is that across the slab, between the house
and the sub- slab region (with the sub-slab vacuum off).
That is the differential against which the ASD system will
have to compete. The pressure differential between  in-
doors and outdoors above grade will  normally tend to be
slightly greater than that across the slab. Where a mitiga-
tor plans  to  conduct sub-slab  suction field extension
diagnostics anyway, it makes sense  to include pressure
measurements across the slab with the vacuum off, to aid
in data interpretation.  Most mitigators never  make
pre-mitigation indoor-outdoor pressure differential mea-
surements above grade to aid  in interpreting sub-slab
suction field extension diagnostics.

The pressure differential across the shell above grade is
sometimes measured as an indicator of whether the house
might be prone to back-drafting of combustion  appli-
ances. If the house is depressurized relative to outdoors
by a certain amount during cold weather prior to mitiga-
tion—e.g., by 0.02 in. WG or more, according to some
references (CMHC88, TEC92)—then there is a threat of
back-drafting that could be exacerbated by an ASD sys-
tem. Many mitigators use diagnostic approaches other
than pressure measurements across the shell above grade
to assess the risk of back-drafting. Since many mitigators
conduct back-drafting tests as a posMnitigation diagnos-
tic, tests aimed at identifying back-drafting are discussed
in Section 11.

Blower door testing. A blower door is a calibrated fan
which can temporarily be mounted in the doorway of a
house, either blowing house air out or blowing outdoor
air in. From measurements of the fan flows required to
maintain various pressure differentials across the house
shell, one can  calculate the effective leakage area be-
tween the house and outdoors. This type of information is
generally not necessary for the design of an ASD system.
Thus, in the large majority of houses where a mitigator
can tell immediately that ASD is the mitigation technique
of choice, blower door testing will not be conducted.

Blower door testing will be  considered  only in those
houses where house ventilation techniques or basement
pressurization appear to be candidates, because the house
in not amenable to ASD for some reasons (poor commu-
nication, high  finish, complex substructure, fieldstone
foundation walls, etc.). The house leakage area can be
used to estimate natural ventilation rates, which can be
used, say, to select the capacity (or to estimate the effec-
tiveness) of an air-to-air heat exchanger. The blower door
flows required to maintain a given basement pressuriza-
tion would suggest the practicality of (and the required
fan size for) a basement pressurization system. This
information could aid in the decision of whether to force-fit
an ASD system into a non-amenable house, or to try a
house ventilation or basement pressurization approach.

Even where a house in not amenable to ASD, blower door
testing will not always be necessary to aid in evaluation
these other approaches. For example, a house with
pre-mitigation levels above 10 to 15 pCi/L will generally
not be reduced below 4 pCi/L with a typically sized
air-to-air heat exchanger,  since 200 cfm units would
generally be expected  to provided no more that perhaps
50 to 75% radon reduction (EPA88a). Thus, blower door
diagnostics would not be needed to assess the potential
for using an air-to-air  heat exchanger in a house having
very high pre-mitigation levels. As another example,
houses with open stairwells or other significant openings
between the basement and upstairs would not be able to
achieve adequate basement pressurization without high
flows, which would likely create an unacceptable heat-
ing/cooling penalty and unacceptable drafts. No blower
door testing would  be necessary to rule out basement
pressurization in those types of houses.

Tracer gas testing.  Various tracer gases have some-
times been used for various  purposes associated with
radon mitigation work, usually in research applications.
Tracer gas measurements would almost never be used by
a commercial mitigator, especially not as a pre-mitigation
diagnostic tool.

Sulfur hexafluoride (SFS) andperfluorocarbon (PFT) trac-
ers are commonly used to measure house ventilation rate.
House ventilation information is not necessary for the
design of an ASD system, as discussed previously; thus,
these tracer measurements would  not be needed when-
ever it is immediately apparent that ASD is the system of
choice. Where ventilation information would be useful to
assess house ventilation and basement pressurization op-
tions, blower door testing would generally be the ap-
proach of choice to obtain that information. The equip-
ment and level of effort needed to make SF6 measure-
ments are unrealistic  for use in anything other than a
research setting. 'PFTs are relatively simple to  use, but
even in this -case, the costs of the PFT analysis, and the
time that would be required on the part of the mitigator to
deploy and retrieve the PFT emitters and detectors, would
not be practical in most commercial mitigation settings.

In previous research projects, halogenated hydrocarbons
have sometimes been used as tracer gases for qualitative
diagnostics. Halogenated hydrocarbons marketed under
trade names such as Freon* and Genetron* are widely
used as refrigerants as well as blowing agents,  cleaning
agents,  and fire extinguishing materials. They  were at-
tractive for use as tracer gases because they  are widely
available  (e.g.,  through  the refrigeration and
                                                    75

-------
    air-conditioning industry), and portable detectors for the
    gases are relatively inexpensive and simple to use.
will often be all the diagnostics that are required for effective
design of an ASD system.
    Among the pre-mitigation tests that individual mitigators
    have conducted with halogenated hydrocarbons have been:
    a) injection of the tracer beneath an adjoining slab on
    grade, and detection of the tracer in the exhaust from a
    vacuum cleaner drawing suction beneath the basement
    slab, in an effort to better determine  whether suction
    drawn beneath the basement slab would extend  to the
    adjoining slab; and b) injection of the tracer into a drain
    tile entering a sump, and detection of  the tracer in the
    exhaust from a vacuum cleaner drawing suction on a
    second drain tile entering the sump, in an effort to deter-
    mine if the tiles form a complete loop.

    However, EPA no longer recommends the use of most
    halogenated hydrocarbons for radon mitigation diagnos-
    tics, due to global concerns about stratospheric  ozone
    depletion. The original chlorofluorocarbon (CFC) refrig-
    erants, such as R-12, are particularly damaging  to the
    ozone layer; more recent, so-called "transition" refriger-
    ants—liydrochlorofluorocarbons (HCFCs), such as R-22—
    arc less damaging. The venting of either of these classes
    of compounds by the refrigeration industry is banned by
    the Clean Air Act Amendments  of 1990. Although this
    venting ban does not explicitly preclude the release of
    CFCs and HCFCs in small quantities in connection with"
    radon mitigation diagnostics, the use of these gases for
    diagnostics would seem inappropriate.

    Advanced refrigerants—hydrofluorocarbons (HFCs), such
    as R-134a—do not  contribute to ozone depletion, and
    their venting by the refrigeration industry is not currently
    banned. Thus, they are the one refrigerant gas that might
    still be considered for diagnostic use.

    Concerns about the environmental impacts  of haloge-
    nated hydrocarbons aside, most mitigators will probably
    find it technically unnecessary  to conduct diagnostics
    using these gases in commercial application. Sufficient
    information for ASD design will probably be obtained
    from prior experience and from the visual inspection (as
    well as from sub-slab suction  field measurements,  if
    needed), without the use of these gases.

The  following subsections address specific procedures for
conducting those diagnostics which are most commonly use-
ful to mitigators, along with discussion of how the results of
the diagnostic  testing can be used in ASD design. Further
discussion of how the results can be used will be presented in
Sections 4 through 8.

3.2    Procedures for the Visual
Survey
Some type of visual survey will be required in every house
prior to installation, to confirm that ASD is the appropriate
mitigation technology for the house, and  to determine the
bajsic design of the ASD system. The visual survey, together
with the mitigator's prior experience with houses in the area,
There will be two general components to a visual survey. One
component will be an interview with the homeowner, to
determine homeowner expectations and house usage patterns.
If the homeowner saw the house under construction or during
remodelling, the homeowner interview may also provide im-
portant building construction information (such as the pres-
ence of sub-slab aggregate) that might not be apparent during
the mitigator's visit. The second component of the  visual
survey will be an inspection to identify house characteristics
which could influence system design.

Some portion of the "visual survey"  might in fact be con-
ducted over the telephone. However, in no case could all of
the necessary detail be obtained without an actual visit to the
house.

A number of house survey forms have been proposed to aid in
the conduct of the visual inspection (EPA88a, EPA88b, Tu88b,
Fo90, among others). One simple form is presented in
Figure 8.

Every form that has been proposed includes some items that
some mitigators will not really need at least some of the time.
The purpose of the forms is to provide a systematic method to
help an investigator ensure that key items will not be over-
looked during the survey. Because information that may be
important to one mitigator in one house may not be important
to another mitigator in another house, even the most practi-
cally oriented general form will always have some entries that
some mitigators will find unnecessary for system design.

The discussion below reviews the type of information  that
mitigators should always collect, and why it is needed, irre-
spective of the specific survey form that is used.

 •  Pre-mitigation radon measurement results. In most
    cases, the mitigator will not be responsible for conducting
    the pre-mitigation measurements of indoor radon concen-
    trations. However, in order to properly select and design
    the mitigation system, the mitigator must find out from
    the homeowner what the pre-mitigation  levels are, and
    what the measurement method and measurement  condi-
    tions were. For example, a high pre-mitigation concentra-
    tion  would rule out house ventilation as a candidate
    mitigation technique; the large increase that would be
    required in ventilation rate would likely result in unac-
    ceptable energy costs and discomfort,  and would be
    beyond the increase in ventilation  that could realistically
    be achieved by  air-to-air heat exchangers.  Very high
    levels could be suggesting a very high  source term,
    Suggesting the possibility of high radon concentrations in
    the ASD exhaust and hence the need for extra care in
    designing the exhaust configuration.

    To properly  interpret these results, the  mitigator must
    know enough about how the measurements were con-
    ducted to be able to make a reasonable judgement regard-
    ing how representative the results are of annual averages
    in the  house.  The type of information needed  would
                                                       76

-------
 PRE-MITIGATION RADON MEASUREMENTS
          Type of Monitor	           ' Location of Monitor_
          Starting Date of Test	             Test Duration	
          Test Results	•	
                                        RADON MITIGATION PROJECT RECORD
                                             Pre-Mitigation Visual Survey
 CONTRACTOR DATA
          Contractor Name	 RCPP I.D._
          Company Affiliation	 Phone No._
          P. O. Box or Street Address	_	:	
          City	                   State	     ZIP Code_
 PROJECT DATA
          Client or Agent Name_
          Site Address	
          City_	                   State	     ZIP Code_
          Test Conducted:
          a by a RMP contractor (list company name)_
          O by the client in accordance with EPA Measurement Protocols_
          O under unknown conditions
          Comments	
 HOUSE DESCRIPTION
          Foundation Type(s)
            Q Basement             O Slab on Grade               O Crawl Space
            O Combination (Describe)  	
          Foundation Walls
            O Concrete             O Block                  O Stone                  O Wood
            O Other (Describe)	

          Houses built on slabs
          Basement or Slab-on-Grade Floor
            O Concrete             Q Exposed earth           O Wood                  Other	
          Sub-Slab Material (for houses with slabs)
            a Aggregate            Q Soil (type)	;	       a Unknown
          Apparent Sub-Slab Obstructions	
          Slab/Wall Openings Which Could Affect ASD Design	
          Drainage System (Sump, Drain Tiles)    	
                                                                                                    (continued)
Figure 8.  Example of a visual survey form.

                                                          77

-------
          Houses built over crawl spaces
          Crawl-Spacs Floor
             O Exposed Earth         Q Gravel Over Earth        Q Vapor Barrier Over Earth
             Q Concrete or concrete wash                    Q Other	
          Accessibility of Crawl Space:
          Access Door	            Headroom	
          Obstructions in Crawl Space	
          Stories Above Grade or Basement
             O one             O two            O three           Comment_
          Attic Space:          Q yes            Q no
          Other comments about house substructure/design     	
          Heating/cooling system (Describe fuel type, equipment, distribution system)_
          Posstbla sources of significant house depressurization (e.g., attic fan)_
          Features Suggesting Mitigation Approaches Other Than ASD
          Water Supply:     O Private Well              O Public Well              a Public Surface Water System
          Any reason to expect building materials may be a source)	
          Any features suggesting that basement pressurization should be considered)_
          Any features suggesting that house ventilation should be considered)_
          Any features suggesting that crawl-space depressurization (or other crawl-space treatment,
          other than SMD) should be considered)	
          Sketch the envisioned mitigation system, showing:
             - House features which will affect ASD suction pipa location, such as slab or crawl-space floor dimensions, finish or other
               above-grade obstructions, sub-slab obstructions, or accessibility.
             - House features which will affect ASD exhaust pipe routing, including interior (and/or exterior) finish and obstructions.
             - Any seating steps that wilt be necessary.
             - Other steps, required to make mitigation systems other than ASD work.
             -All Important components of the mitigation system.
                                                                                   Form completed by:__
                                                                                            RCPP I.D._
                                                                                                Date:
Figure 8. (continued)

                                                              78

-------
 include: the type of measurement device that was used;
 the duration of the measurement; the general climatic
 conditions during the test; where the measurement was
 made (e.g., confirming that a detector was not placed in a
 crawl space or sump); and the degree to which the house
 remained closed during the test.

 Factors determining which ASD variation is used.
 These factors include:

-  House substructure (basement, slab on grade, crawl
   space, or combination of these). The substructure will
   determine  whether SSD, SMD, or a combination of
   techniques is used.

-  Presence or absence of a sump with visible drain tiles,
   or of exterior drain tiles draining to an above-grade
   discharge (along with  the apparent or reported com-
   pleteness of the drain tile loop). The presence of these
   features will determine whether sump/DTD  or DTD/
   remote discharge are options.

-  Nature of the foundation wall (poured concrete vs.
   hollow-block vs. fieldstone;  if block,  are top voids
   open, and is the wall otherwise very leaky?). This
   information would help  determine whether a BWD
   component might be needed as a supplement to a SSD
   or DTD system, or (along with other information)
   whether BWD might be considered as a stand-alone
   technique.

-  Presence  of a  perimeter channel  drain. If the
   baseboard-duct configuration of BWD is going to be
   considered, houses having perimeter  channel drains
   would be the best candidates for this approach (espe-
   cially when the basement is not finished).

 Factors suggesting house is not amenable to ASD,
 or  is amenable to alternative approaches. These
 factors include:

-  Known poor sub-slab communication  (from observ-
   able or homeowner-reported lack of  aggregate, and
   from experience in other houses in the area); complex
   substructure (e.g., multiple wings); high degree of base-
   ment finish; and/or fieldstone foundation walls. These
   factors, especially in combination, could seriously com-
   plicate application of SSD, BWD, or SSD in combina-
   tion with BWD or SMD.

-  A crawl space which is inaccessible or unusually clut-
   tered, preventing application of SMD.

-  A basement which is relatively tight  (no forced-air
   ducts, stairwell between basement and upstairs has
   door  which can  be closed, no other major  openings
   between the basement and  upstairs such as laundry
   chute), which would suggest the possible applicability
   of basement pressurization.

-  Homeowners amenable to the lifestyle required for
   basement pressurization to be successful (e.g., willing
   to keep the door closed between the basement and
   upstairs, acceptance of drafts and possible heating/
   cooling penalty that might result).

-  A crawl space which is relatively tight (e.g., no forced-air
   ducts, no foundation vents), suggesting the crawl-space
   depressurization might be an option.

-  House  is served by a private well, and a high radon
   concentration has been measured in the well water (or
   there is some  other basis for expecting high radon
   concentrations  in  the water). In such cases, a water
   treatment system  might  need to  be  considered as a
   supplement to ASD.

-  Underlying soil is well-drained gravelly soil, in which
   case soil pressurization might be considered instead of
   ASD.

 Factors  which  would influence  the number and
 positioning of SSD suction pipes.  Such factors in-
 clude:

-  Aggregate observed beneath the slab through slab open-
   ings (e.g., at the  bottom of sump pits, registers for
   sub-slab forced-air supply ducts, utility openings under
   bathtubs, etc.); or, aggregate reported to be present by
   homeowner, from observations of house during con-
   struction; or, aggregate believed to be present based
   upon code requirements or common construction prac-
   tice in the area. Good aggregate would, of course,
   suggest the need for only one or two SSD pipes.

-  Slab size. Especially where sub-slab communication is
   suspected to be poor  or uneven,  a  larger slab may
   suggest the need for additional suction pipes.

-  House  floor plan, degree of floor/wall/ceiling finish,
   and living patterns. Suction pipes will preferentially be
   installed in unfinished areas (such as unfinished por-
   tions of basements, or utility rooms) or in concealed
   areas (such as closets). In unfinished areas, pipes would
   be placed out of the normal traffic patterns.

-  Homeowner plans and preferences. The owners' plans
   to finish a currently unfinished area, or personal prefer-
   ences for aesthetic or other reasons,  would influence
   pipe location.

-  Observed or reported sub-slab utilities (such as sewer
   lines and forced-air ducting) and in-slab utilities (such
   as heating coils built into the slab), which could limit
   locations where pipe penetrations can be made.

-  Exterior driveways, patios, walkways, etc., which would
   affect where below-grade penetrations might be made
   from outdoors.

-  Apparent entry routes through the slab. Such entry
   routes might suggest that suction pipe placement might
   be biased toward those entry routes  in order to help
   ensure effective treatment. (Where the slab openings
                                                    79

-------
   are significant, location of a suction pipe too near such
   entry routes without at least partial  sealing of the
   openings could result in significant air short-circuiting
   into  the system.) Entry routes of concern might in-
   clude: a region toward the interior of  the slab where
   there is extensive cracking; a block structure toward
   the interior of the slab, which penetrates the slab and
   rests on footings underneath; and a perimeter block
   foundation wall that extends the deepest below grade in
   a walk-out basement.

-  Observed or reported sub-slab obstructions which di-
   vide slab into segments, such as interior footings and
   forced-air supply or return ducts. While it will not
   always be necessary to ensure that a SSD suction pipe
   is placed in each segment of the slab if there is good
   aggregate,  it may sometimes be necessary,  and the
   mitigator should be aware that this problem might
   arise.

-  Evidence of water entry into a basement through slab
   or wall cracks, suggesting that drainage is poor in that
   part  of the basement and that suction pipes at that
   location may become blocked by water during wet
   weather.

 Factors which would  influence the design of a
 crawl-space SMD  system. These factors  would in-
 clude:

-  The size of the crawl space. Larger crawl spaces will
   increase membrane materials and installation cost, and
   would increase the likelihood that sub-membrane per-
   forated piping or multiple suction pipes would be needed
   to adequately distribute suction beneath the membrane.

-  The nature of the crawl-space floor. Gravel on the floor
   would facilitate suction field extension under the mem-
   brane.  Irregular  floors, e.g., with protruding rocks,
   could require heavier membranes, and more effort in
   membrane installation.

-  The  accessibility of the crawl space, including head-
   room and major obstructions such as forced-air heat-
   ing/cooling systems. Poor access to some portions of
   the crawl space  could increase installation costs and
   perhaps require  design  modifications  which include
   leaving portions of the floor uncovered. (Poor access to
   the entire crawl space could require the use of a mitiga-
   tion approach other than SMD.)

-  Expected traffic patterns in the crawl space. If particu-
   larly heavy traffic is anticipated over some significant
   portion of the crawl-space floor (e.g., due to the loca-
   tion of appliances in the crawl space), it may be neces-
   sary to place heavier membrane—in extreme cases, 45-
   or 60-mil sheets of EPDM™ (a rubber-like roofing
   membrane), or some other appropriate material—on
   top of (or in place of) the polyethylene membrane in
   the heavy-traffic areas to protect the membrane from
   being punctured.
 Factors which would influence the routing of the
 exhaust piping. These factors include:

-  The availability of a convenient route for an interior
   stack up through the house, through either: closets or
   other concealed areas on any floors above the slab or
   crawl space; or utility chases. The location of such a
   convenient exhaust route  could also influence  SSD
   pipe location.

-  Convenient access to an adjoining  slab-on-grade ga-
   rage, so that the exhaust stack can be routed up through
   the adjoining garage instead of through the house.

-  Locations where  the exhaust piping can reasonably
   penetrate through the basement band joist, and where
   an exterior stack/exhaust can be installed.

-  The presence and accessibility of an attic where the fan
   for an interior stack can be mounted, if an interior stack
   is planned.

-  The nature and degree of finish in the areas through
   which  the exhaust piping may have  to pass. If  an
   exhaust piping route cannot be identified which largely
   avoids finished areas, this could increase installation
   costs and have an aesthetic impact. The nature of the
   basement ceiling is one particularly important element
   of the finish; an unfinished or a suspended ceiling will
   greatly simplify horizontal piping runs, relative to a
   sheetrock ceiling.

-  Any obstructions (such as ducts and utility pipes) which
   may complicate routing, e.g., by requiring that a  hori-
   zontal piping run make a  vertical bend which could
   hinder maintaining proper pipe  slope for condensate
   drainage purposes.

-  Any exterior finish which would influence the ability
   to  penetrate the house shell near grade and extend a
   stack up the outside of the house, where an exterior
   stack is planned. Such finish could include, e.g., exte-
   rior brick or stone cladding at the point where the
   exhaust piping would penetrate the shell, or the nature
   of the exterior siding and roof overhang, if it is aes-
   thetically important to box in or otherwise finish the
   exterior fan and stack.

 Factors affecting  the degree of sealing required.
 Significant slab or wall openings that could require some
 form of closure during installation should be noted. Open-
 ings  of particular importance include:

-  Wide gaps at the perimeter wall/floor joint (including,
   in  the extreme case, a perimeter channel drain). The
   accessibility of this joint should also be noted, since
   that will significantly influence the practicality of (or
   the installation effort required for) closure of this gap.

-  Other major slab  openings, such as sump holes, un-
   trapped floor drains, cold joints, and utility penetra-
   tions.
                                                     80

-------
-  Major wall openings (in particular, open top voids in
   hollowrblock foundation walls). This is of particular
   importance when a BWD component to the ASD sys-
   tem is anticipated.

 Driving forces for radon entry which could influ-
 ence  ASD  design. The  house design and operating
 features which influence radon entry driving forces will
 not often be that important to a mitigator when there is
 good sub-slab communication. Under those  conditions,
 the sub-slab is likely to be depressurized sufficiently to
 withstand the depressurizations  created by  appliances,
 and the flows induced by thermal bypasses. However,
 where communication is marginal, general knowledge of
 the challenges that the system might be facing could aid
 in judging how conservatively  the system should be
 designed. Features that can be noted include:

-  Appliances that can contribute to depressurization of
   the lower level of the house. These appliances include
   combustion appliances (furnaces and boilers, fireplaces,
   wood stoves, and water and space heaters),  if these
   appliances burn fuel (i.e., are not electric), and if they
   are operated with any frequency. Depressurizing appli-
   ances also include exhaust fans (clothes driers, room
   exhaust fans, whole-house exhaust fans, and exhaust
   fans on kitchen ranges).

   Where cold-air return  ducting is located inside the
   livable space, as is commonly the case, central forced-air
   furnace fans can often serve essentially as exhaust fans
   for that portion of the house in which the return ducting
   is present. This situation will exist if there are not
   sufficient  supply registers in that portion of the house
   to compensate for the air withdrawn by return registers
   and by the leaky, low-pressure return ducting. Or, if the
   forced-air supply ducting is located outside the livable
   space (such as in the attic or garage), a similar situation
   could occur.

   If, in aggregate, the appliances present appear as though
   they might be sufficient to create significant depressur-
   ization when operating, the ASD system may warrant
   additional suction pipes or other measures in an effort
   to increase the sub-slab or  sub-membrane depressur-
   izations created by the system.

   The presence of combustion appliances, especially in
   basements, would also alert the mitigator to check for
   back-drafting if the ASD system has unusually high
   exhaust flows. This can be a particular problem if a
   BWD system is being considered.

T  Thermal bypasses, which could increase the flow of
   soil gas into the house through entry routes if the entry
   routes are not adequately depressurized by the ASD
   system. Major thermal bypasses include, e.g., unclosed
   stairwells between house stories, laundry chutes, utility
   chases, chases around  flues, openings associated with
   forced-air ducts, etc. Thermal bypasses will often not
   be a significant concern in ASD design in most cases,
   but if communication is marginal and if the thermal
      bypasses are extensive, the mitigator may want to be
      somewhat more conservative in system design.

In the conduct of a visual survey, the tools most likely to be
required are a flashlight, a screwdriver, and a stiff brush. The
screwdriver might be used, e.g., to pry grilles off of a floor
drain in order to see if it is trapped; the brush might be used to
scrape away dirt and concrete wash at the wall/floor joint in
order to better see the nature of the crack there. Some mitiga-
tors also  utilize heatless chemical smoke devices, to enable
visualization of air flows. Chemical smoke could  indicate
whether air flow is into the house at various openings, sug-
gesting the possible importance of those openings;  it could
also be used around flues of combustion appliances, to assess
whether the appliances are drafting properly or whether the
house is near back-drafting conditions prior to mitigation.

One output from the visual survey will be a recommendation
regarding any further diagnostic testing that is needed before
the mitigation system can be designed. If no further diagnos-
tics are found to be necessary, the final  output of the visual
survey will be the design of the system, as suggested by the
last page of Figure 8 (the sketch of the mitigation system).
The level of written detail desirable for the design depends on
whether the person(s) who conducts the visual survey and the
design will be on-site at the house during the actual installa-
tion, with less detail being needed if the designer is going to
be present to give directions to the workmen. A floor plan of
the lowest level of me house roughly to scale, showing, for
example., the pipe penetrations and routing for the envisioned
system, the fan location, and key house  features influencing
design, will generally be  advisable, to  help the mitigator
visualize the system and determine  materials requirements,
and  to aid in communication with the workmen  and the
homeowner.

3.3    Procedures for Sub-Slab
Suction Field Extension
Measurements

If there is not evidence of a reasonably consistent aggregate
layer beneath the slab, from either the mitigator's observation
or from the homeowner's experience and if the mitigator's
prior experience in the area does not suggest  whether the
house is likely to have aggregate, then sub-slab  suction field
extension measurements may be advisable.

Suction field extension measurements involve generation of a
suction field under the slab prior to installation of the mitiga-
tion  system, and then measuring the suction field  induced
under the slab at test holes remote  from the suction point.
Commonly, this pre-mitigation suction field is generated us-
ing an industrial vacuum  cleaner. Another option that has
been suggested but does not appear to have been used to date
is to utilize an ASD fan, mounted on a 4-in. pipe in a portable
test stand,  to generate the suction.  Use of a fan and pipe
similar to that which is to be used in the ultimate system
would provide more meaningful and easy-to-interpret suction
field and flow data than would a vacuum cleaner, since the
vacuum  cleaner motor has a much different  performance
curve than do ASD fans.
                                                     81

-------
Rather than using a portable ASD fan test stand, some mitiga-
tors use the mitigation system itself to generate the suction
field, installing the first suction pipe, mounting the fan, and
turning the system on. If the results show inadequate exten-
sion of the suction field, additional suction pipes are installed
where needed.  This approach might be viewed as
dMr/ng-mitigation, rather than pre-mitigation, diagnostics.

In the discussion here, it will generally be assumed that a
vacuum cleaner is being used, because vacuum cleaners are
commonly employed, and because more care is usually needed
in the conduct of the test and the interpretation of the results
when a vacuum cleaner is used. Where necessary, it will be
indicated how the procedure would differ if an ASD fan rather
than a vacuum cleaner were being used to generate the suc-
tion.

Suction field extension measurements can be conducted with
two different approaches.

The first, qualitative approach is simply to determine whether
sub-slab communication is relatively good or poor. This ap-
proach involves fewer test holes (e.g., one in each corner of
the basement), and indicates whether suction is generally
extending well in  various  directions. (It can also indicate
whether system flows will be relatively low or high.) This is
the approach to start  with any time that communication is
unknown and might be good. If this test shows that communi-
cation is good (by virtue of good sub-slab depressurizations
induced by the vacuum cleaner in the slab corners), then no
further diagnostics may be needed. But if the results indicate
that suction field extension is marginal, poor, or uneven, more
sub-slab data may be desirable to design the system. In that
case, the testing can be extended to provide more quantitative
results, per the second approach below, by drilling more test
holes and more carefully controlling the speed of the vacuum
cleaner.

The second, more quantitative approach can be useful when
sub-slab communication is marginal or poor. By adjusting the
vacuum to better reproduce SSD flows, and  by  measuring
vacuum-induced sub-slab depressurizations through more test
holes, a more definitive estimate can be developed regarding
how far the suction field from a SSD suction pipe might
extend, and thus how many  SSD pipes might be needed (and
where they might best be located). As discussed later, these
more extensive tests with a vacuum cleaner may not predict
exactly how far the suction  field from a 4-in. SSD pipe will
extend. Most often, the vacuum cleaner  diagnostics tend to
over-predict the number of SSD pipes required. Any diagnos-
tic testing conducted using a SSD fan test stand or using the
initial SSD system to generate the suction field will almost
automatically reproduce SSD flows and thus almost automati-
cally be quantitative, if a sub-slab suction pit is excavated and
if the suction losses in the system piping are simulated.

The experimental setup for quantitative suction field exten-
sion testing using a vacuum cleaner is illustrated in Figure 9.
The details of this configuration will  be discussed later, in
Section 3.3.2. This figure will also be helpful in understanding
the qualitative testing approach, as well as some  of the key
concepts regarding suction field extension testing.
In conducting sub-slab diagnostics using a vacuum cleaner, a
mitigator must be aware that the suction-vs.-flow perfor-
mance curve of the vacuum cleaner will be very different from
that of the ASD fan that will ultimately be installed. Thus,
care must be taken in cases (such as with the quantitative
approach) where it is desired to attempt to reproduce actual
SSD flows and suctions using the vacuum. The approach for
simulating  the SSD system is to adjust the vacuum cleaner
speed so that  the sub-slab  depressurization measured at a
baseline test hole 8 to  12  inches away from  the vacuum
nozzle—i.e., at a point that would be in the sub-slab suction
pit under a SSD pipe, were the SSD pipe to be installed where
the vacuum nozzle is—is about equal to that which the SSD
pipe is expected to produce in the sub-slab pit. See Figure 9.
Simple theory regarding sub-slab flow dynamics predicts that
if the these sub-slab depressurizations  are equal, then  the
flows in the vacuum cleaner should nominally be the same as
the flows in the  SSD system,  despite their very different
performance curves. And under these conditions, the sub-slab
depressurizations  generated by the vacuum at more remote
test holes will nominally be the same as those generated by a
SSD system. If the sub-slab depressurization with the vacuum
at this baseline test hole is not set equal to that expected with
the SSD pipe, then the flows and depressurizations generated
by the vacuum will have no quantitative relationship to those
that will be generated by the SSD system.

In slabs having good aggregate, flows may be sufficiently
high such that the relatively low-flow vacuum cleaner will be
unable to achieve the desired suction at the baseline test hole
as specified above. In these cases, the vacuum will be unable
to simulate a SSD system. But in cases where aggregate is
good and  flows high, quantitative suction  field diagnostics
will be unnecessary anyway.

From a practical  standpoint, some mitigators might find it
more cost-effective to perform the sub-slab diagnostics dur-
ing the installation, as mentioned earlier, when sub-slab com-
munication is uncertain. In this case, additional suction pipes
would  be added  as necessary if  the initial pipe(s) proved
unable to extend a sufficient suction field. An added advan-
tage of proceeding in this manner is that,  since the suction
field is being created by the system fan, there is no concern
about whether a diagnostic vacuum cleaner is properly repre-
senting the system fan. A disadvantage of this approach is that
the mitigator will notknow what complications may be present
until installation is well underway.

If a mitigator decides  to skip the pre-mitigation sub-slab
diagnostics, and if the initial hole(s) drilled through the slab to
install suction pipes show no aggregate or other evidence of
likely poor communication, then it is advisable to proceed
with during-mitigation (or post-mitigation) suction field test-
ing, as discussed above. The cost of conducting these diagnos-
tics while the installation crew is on site is estimated to be
roughly $45 (with a standard deviation of $47) in unfinished
basements (He91b, He91c). By comparison, if the crew leaves
the site and depends upon a post-mitigation radon measure-
ment to alert  them  that the system is  not functioning as
desired, this cost  may increase by roughly $150, due to the
time required for the crew to return to the site. Or, as another
option, the mitigator might elect to simply install a  second
                                                        82

-------
 Pitot tube or other device to measure the flow
 of sub-slab gas into the vacuum cleaner;
 needed if it is desired to measure sub-slab
 flow characteristics to aid in selecting the SSD
 fan having the optimum performance curve.  —
 PVC ball valve, to allow room air to bleed
 into vacuum cleaner intake as necessary to
 achieve the desired sub-slab depressurization
 at the baseline test hole. (This is an alternative
 to the use of a speed controller on the vacuum
 cleaner motor as a method for adjusting the
 induced sub-slab depressurization at the
 baseline hole.)	
 1.25" PVC pipe: rigid pipe facilities mounting
 in slab and enables measurement of flow into
 vacuum, if desired.	
 Exhaust
 Discharged
 Outdoors
                   Flexible
                   Vacuum Hose
   Industrial
   Vacuum
   Cleaner
     Magnehelic® gauge or micromanometer. capable
     of measuring sub-slab depressurizations in the
     range that will be developed by the SSD suction
     pipe (commonly 0.5-1.5 in. WG, sometimes higher).
     Room air bleed into the vacuum cleaner intake
     (or vacuum motor speed) must be adjusted until
     this gauge shows that the vacuum cleaner is
     maintaining the sub-slab depressurization that
     the SSD fan is expected to produce at this
     location (i.e. in the suction pit).
       r Micromanometer. to measure the sub-slab
        depressurization at remote test holes. If the
        vacuum cleaner can be adjusted so that the
        sub-slab depressurization at the baseline test
        hole with the vacuum is identical to that which
        the SSD fan will produce at that location (i.e. in
        the pit beneath the SSD suction pipe), then the
        measured sub-slab depressurizations at remote
        test holes with the vacuum should be identical
       ' to that which the SSD fan will produce at the
        remote holes.
                            Hole
                                      U—S"-12"—«J
  Baseline
  Test Hole
•	Remote
    Test Hole
Figure 9.   Experimental configuration for quantitative pre-mitigation sub-slab suction field extension and flow diagnostics using a vacuum
          cleaner.
suction pipe at a convenient location without spending the
estimated $45 to do the diagnostics. Installation of a second
pipe adds  roughly $135 (standard deviation $44) to $225
(standard deviation $90) to the cost, depending upon degree of
finish. It is a judgement call regarding whether it is a reason-
able gamble to spend  $135-S225 to install a suction pipe
without first spending roughly $45 for diagnostics to see if the
pipe is necessary and where it should optimally be located.

The following discussion describes the equipment and materi-
als needed, the test procedure, and the means for interpreting/
utilizing the test results, for each of the two measurement
approaches.
3.3.1  Qualitative Assessment of Suction
Field Extension
This suction field extension measurement approach provides
a qualitative indication of whether communication is rela-
tively good or poor, and of how uneven it may be.
                                                     83

-------
 Equipment and materials required

-  Industrial vacuum cleaner (assuming, as discussed
   above, that a vacuum cleaner is being used rather than
   an ASD fan). While units having various performance
   characteristics  are available, typical vacuum cleaners
   used in diagnostics can develop suctions up to 80 in.
   WG at zero flow, and flows of perhaps 100 cfm at zero
   static pressure. (If an ASD fan were used instead of a
   vacuum, so that flows during the diagnostics approxi-
   mated ASD system flows, this test could almost auto-
   matically become quantitative,  as  mentioned previ-
   ously.)

-  Sufficient vacuum cleaner  hose to permit the vacuum
   to be located outdoors during the tests, or, if the vacuum
   were indoors,  to  permit the vacuum  exhaust to be
   routed outdoors. Directing the exhaust outdoors will
   avoid the discharge into the house of high-radon soil
   gas and of dust that is not captured by the vacuum. If
   the vacuum itself were outdoors, fugitive soil gas and
   dust escaping from  the vacuum would be prevented
   from entering the house. The vacuum  is depicted in-
   doors in Figure 9 for convenience.

-  A suitable masonry drill for drilling holes  through the
   concrete slab. An electric rotary hammer drill is com-
   monly used for this purpose.

-  Drill bits for  the masonry drill, including: one bit
   slightly larger  than the nozzle of the vacuum cleaner
   (typically about  1.25  in. diameter), to drill the  hole
   through  which the vacuum nozzle  will penetrate the
   slab; and one  bit for the  test holes (1/4- to  1/2-in.
   diameter).

-  Pressure sensing device, for determining sub-slab de-
   pressurizations at the remote test holes. This device can
   be one of the following:

   -- Preferably,  a digital micromanometer as shown in
      Figure 9, capable of detecting pressures as low as
      0.001 in. WG. This device provides a quantitative
      measurement Flexible tubing of a diameter similar
      to that of the test hole is required to connect the
      micromanometer to the sub-slab through the test
      hole.

   — A heatless chemical smoke device. Chemical smoke
      may sometimes be more sensitive than the micro-
      manometer in detecting depressurizations,  since
      smoke might be drawn down into the test hole at
      suctions below 0.001 in. WG. For qualitative as-
      sessments of suction field extension, the qualitative
      indication of depressurization provided by smoke
      sticks may  sometimes be sufficient. However, the
      quantitative result from a micromanometer is ex-
      tremely useful in telling the mitigator whether the
      suction field extension  is marginal or  strong. An-
      other concern about the use of smoke is that the test
      hole will not be sealed by tubing, as with a micro-
      manometer; when the sub-slab suction is marginal,
      the open test hole may be sufficient to neutralize the
      sub-slab depressurization at that location (Gad92).

      Thermally generated smoke (from a punk stick or
      cigarette) should never be used for this purpose.
      The natural thermally-induced rise  of this smoke
      could mask or prevent any flow down into the hole
      in response to sub-slab depressurization. Also, there
      is a fire hazard.

-   Rope caulk or putty, to provide an air-tight seal where
   the vacuum cleaner nozzle and the micromanometer
   tubing penetrate through the slab. In addition, the rope
   caulk can be used to temporarily close  individual test
   holes while measurements are being made at another
   hole, to  prevent the measurements from being influ-
   enced by air leakage through the other holes.

-   Hydraulic  cement or  other non-shrinking cement, to
   permanently close all of the holes following  testing.

-   (Optional) A pilot tube (or some other suitable device,
   such as an anemometer or a calibrated orifice) to deter-
   mine flows in the vacuum cleaner nozzle, as shown in
   Figure 9. Relatively high flows in the nozzle would
   confirm  observed good  communication, and would
   suggest that a relatively high-flow fan is needed on the
   ASD system. (It could also be suggesting air leakage
   into the sub-slab near the vacuum  cleaner, possibly
   explaining poor communication). Relatively low flows
   would tend to confirm poor communication, and possi-
   bly suggest the need for a low-flow, high-suction ASD
   fan. For this  qualitative testing,  some investigators
   suggest  dispensing  with the pitot tube, and simply
   judging whether the vacuum flow is high or low from
   the sound of the vacuum motor (Br91b).

-   Note that the Magnehelic® gauge shown in Figure 9,
   mounted in a baseline test hole near the vacuum nozzle,
   in not included on this list for the qualitative test. The
   additional effort involved in installing this gauge in a
   baseline hole, and in adjusting the vacuum to achieve
   the desired depressurization at that hole, is a key factor
   distinguishing between the qualitative and quantitative
   procedures.

 Test procedure

-   From the house floor plan, select the location for the
   1.25-in. vacuum cleaner suction hole through the slab.

   -  If a SSD system is being considered, the suction
      hole ideally should be located at a site where a SSD
      suction pipe will potentially be installed. This selec-
      tion not only reduces the number of places where
      the slab is penetrated,  but also reduces the potential
      for differences between the diagnostic results and
      the actual system performance,  in  the event that
      communication varies at different locations around
      the slab. (Note that if an ASD fan on a 4-in. pipe is
      being used to generate the suction field rather than a
      vacuum cleaner, the  4-in. hole through the slab
                                                     84

-------
   would almost certainly be drilled at the site of a
   permanent suction pipe.)

~  The goal of locating the vacuum cleaner hole at a
   potential SSD pipe site means that the vacuum hole
   may be  near the slab perimeter, especially if poor
   communication is suspected. Most radon entry will
   often be through the perimeter wall/floor joint and
   block foundation walls, so that it is most important
   to determine that the  suction field is extending
   effectively at the perimeter. The suction hole should
   be at least 6 to 12 in. from the wall, to avoid the
   footing (EPA88b, Gad89, Br92). The vacuum hole
   must not be near any major slab opening. If there is
   a large wall/floor joint—in particular, if there is a
   perimeter channel  drain—substantial  air leakage
   through the perimeter joint can overwhelm the
   vacuum cleaner, which cannot  move a lot of air
   (perhaps 100 cfm). In such cases, the vacuum hole
   should be toward the center of the slab, even if the
   ultimate installation may involve closure of the
   perimeter gap and location of SSD pipes near the
   perimeter.

—  In cases where SSD pipes should not be placed near
   the perimeter walls, such as in slab-on-grade houses
   in Florida, it has been  suggested that the vacuum
   hole be  located between 6 and 15 ft from exterior
   walls (Fo90).

-  Consistent with the selection of sites for SSD pipes,
   the vacuum cleaner suction hole should be in an
   unfinished area (such as an unfinished basement or
   a utility room) or in an inconspicuous area (such as
   a closet), acceptable to the homeowner.

—  As with S SD pipe site selection, the vacuum suction
   hole site should be selected in  an effort to avoid
   sub-slab or in-slab utilities.

~  If there  is a sump with a drain tile loop, and sump/
   DTD is the  intended mitigation technique, the
   vacuum cleaner can draw suction at the sump, avoid-
   ing the need to drill a 1.25-in. hole through the slab.

Select the  location for the suction measurement test
holes.

-  For this qualitative measurement, one test hole in
   each quadrant  of the slab,  generally toward the
   corners, would  be appropriate. Because of the im-
   portance of suction  field  extension around the pe-
   rimeter, location of the test holes near the walls is
   generally preferred, but not so close to a wide wall/
   floor crack that the suction field will have dropped
   near zero. Some investigators suggest locating the
   slab test holes  at about the middle of each wall,
   about 6 in. out from  the wall  (EPA88b). If it is
   suspected that communication is good, some miti-
   gators may wish to start with only one or two test
   holes, at locations the most remote from the suction
   hole (Gad89); by this approach, additional suction
      holes would be drilled only if the pressure measure-
      ments at the first hole(s) indicate that communica-
      tion is not good. The holes should not be immedi-
      ately beside major openings, such as perimeter chan-
      nel drains, since, even with good communication,
      sub-slab depressurizations will have declined to-
      wards zero near such major air leakage points.

   —  As with other slab holes, the test hole locations
      should be selected to  be inconspicuous,  and to
      avoid utility lines.

-  Drill 1/4- to 1/2-in. suction test holes.

   —  If initial testing is being conducted on only one or
      two test holes, because good communication is
      suspected, drill these first.

   —  If  the   hole  is  being   drilled  through
      asbestos-containing floor tile  over the slab, some
      mitigators place a wet sponge (or some other mate-
      rial, such as shaving cream) on top of the tile at the
      drilling site, to prevent asbestos fibers from becom-
      ing airborne during drilling.

   —  Make sure the drill bit penetrates through the slab
      and any vapor barrier, into the sub-slab fill.

   —  Try to assess the nature of the sub-slab fill. One
      approach for doing this (Bro92) is to stop the drill
      just after it penetrates the slab. Then, push the drill
      down into the sub-slab material to determine how
      compact it is. Turn the drill back on for a second,
      then pull it out and inspect the bit. If it is clean, this
      suggests that aggregate may be present, or that the
      soil under the slab may have subsided. If there is
      dirt on the bit, that will indicate that aggregate is not
      present, and will suggest the type and wetness of the
      soil immediately under the slab.

      This information will aid in the interpretation of the
      suction field extension  test results. If aggregate is
      present at all of the initial test holes, a decision may
      be  made not to continue with  the suction field
      extension  test (depending upon prior experience
      with the evenness of aggregate layers in that area),
      avoiding the need to drill the remaining test holes or
      the vacuum cleaner suction hole.

   —  Vacuum up the dust created by  drilling, to avoid
      plugging of the micromanometer sampling tube
      (and of the grab sampler filter, if grab sampling is
      performed),  and to permit effective sealing of the
      gap between the sample tube and the slab. If a
      sub-slab grab sample is  to be drawn  for a radon
      measurement, the  vacuum should  be  operated as
      briefly as possible (for  only a  few seconds), to
      minimize  any artificial  reduction in the sub-slab
      radon level.

   —  Temporarily close test holes with rope caulk.
                                                  85

-------
-- (Optional) If a grab sample is to be taken to deter-
   mine sub-slab radon concentrations at the test holes,
   that sample would be taken at this time. Following
   the vacuuming step above, the hole should remained
   closed with the rope caulk for awhile before the
   sample is taken, to let the sub-slab concentrations
   re-equilibrate after possible dilution due to the drill-
   ing and vacuuming.

Drill  1.25-in. suction hole through  slab for vacuum
cleaner nozzle, and insert nozzle.

- See comments  above for drilling  test holes, for
   cases where asbestos-containing floor tiles are cov-
   ering the slab.

-- Make certain drill bit penetrates through the slab
   and any vapor barrier, into the sub-slab fill.

— Inspect the nature of the material under the slab, as
   discussed above for the test holes. Again, if aggre-
   gate appears to exist everywhere, continuation of
   the suction field extension diagnostics is probably
   unnecessary.

— Vacuum up the dust created  by the drilling, to
   permit effective sealing of  the  gap between the
   nozzle and the slab. Minimize operation of vacuum
   if it is planned to draw sub-slab  grab samples for
   radon measurements.

« (Optional) Any grab sample for radon measure-
   ments would be taken at this time, if planned. See
   comments above in connection with the test holes.

— Insert the nozzle into the hole, to a depth about one
   half the thickness of the slab. Insertion of the nozzle
   all the way to the underside of the slab might result
   in partial blockage of the nozzle by the sub-slab fill.
   Tightly press rope caulk or putty around the circum-
   ference of the nozzle, at the joint between the side
   of the nozzle and the top of the slab. Failure to seal
   this gap will result in significant air leakage down
   through the gap, reducing suction field extension.

   The  vacuum cleaner hose attached to this nozzle
   should be long enough to extend  to the vacuum
   cleaner located outdoors.  Alternatively,  if the
   vacuum is in the house near the  suction point, the
   hose on the vacuum cleaner exhaust should exhaust
   outdoors.

— If the  vacuum is to draw suction  on an existing
   sump, rather than through a hole drilled in the slab,
   seal a sheet of plastic over the sump, and seal the
   vacuum nozzle through the  plastic. Alternatively,
   seal  the vacuum nozzle into an exposed drain tile
   where it enters the sump, and temporarily seal the
   ends of any other drain tiles entering the sump.

Place the house in the condition that will be used
throughout this testing. In general, it is recommended
that this condition be with exhaust appliances operat-
ing, to simulate the most challenging conditions in that
particular house. If no sub-slab depressurizations are
measured under those conditions, the appliances might
be turned off (and perhaps windows opened), to see if
some sub-slab depressurization is  being achieved un-
der non-challenging conditions.

With  the vacuum cleaner off, measure  the  sub-slab
pressurization (or depressurization) at each test hole
with the micromanometer. This will indicate the "base-
line"  pressure difference  across the  slab under the
weather and  appliance operation  conditions existing
during the tests. As discussed in Section 3.5.4, this
measurement will indicate how the existing conditions
compare against the expected maximum slab pressure
differential (i.e., the expected maximum driving forces),
and may thus aid in interpreting the  results  with the
vacuum operating.

~ Remove temporary rope caulk seal over each hole;
   replace after measurement is completed.

— Insert sample tube from micromanometer, to a depth
   about half the thickness of the slab. Press a rope
   caulk seal around the tubing circumference, at joint
   between tubing and top of slab. Failure to effec-
   tively seal this gap will result in serious measure-
   ment error.

— Observe reading on micromanometer. Reading will
   likely  fluctuate somewhat, as a  result of minor
   pressure changes inside the house due to winds, etc.
   Record the observed range, and/or the average.

~ Check the zero on the micromanometer often, pref-
   erably before each reading in cases where  the mea-
   sured pressures are very low.

— If the mitigator has only a chemical smoke device,
   and not a  micromanometer, a quantitative measure
   of existing pressures would not be possible, and this
   baseline measurement would thus  be of less value.
   However, when the smoke testing is done during
   operation  of the vacuum (see below), it would still
   be worthwhile to turn  the vacuum off during the
   testing at each test hole, to see if these is a distinct
   difference in  smoke flow with and  without the
   vacuum operating.

Measure the depressurization created  beneath the slab
at each test hole when the vacuum cleaner is operating.
On vacuum cleaners having speed controllers, operate
the vacuum at full power for this relatively qualitative
test.

-- With tape covering other test holes, insert sample
   tube from micro-manometer into each test hole in
   turn. Seal  gap between the tubing and the top of the
   slab with rope caulk or putty.
                                                  86

-------
   --  Where communication is marginal to poor, it may
      take a minute, or perhaps several minutes, for the
      suction from the vacuum to become established,
      especially at remote test holes.

   -  Observe the reading on the micromanometer at
      each hole. Record the range and/or the average of
      the readings. Compare these values with the values
      measured before the vacuum cleaner was turned on.

   -  Check the zero on the micromanometer often, pref-
      erably before each reading in cases where the mea-
      sured depressurizations are very low.

   -  Be careful if any test holes are near the suction
      point.  The high suctions that can sometimes exist
      near the vacuum cleaner when flows  are low (tens
      of inches of water) can be high enough to damage
      some micromanometers.

   —  If a chemical smoke device is used, release only a
      small quantity of smoke very near to the test hole.
      The smoke patterns can sometimes be obscured by
      air currents in the room when the pressure differ-
      ence across the slab is quite small. Back-lighting
      the smoke with a flashlight can be helpful in seeing
      the smoke patterns (EPA88b).

   —  If the  vacuum  cleaner nozzle is not fitted with a
      pilot tube (or other device for measuring flow), a
      qualitative assessment of whether the flows seem to
      be high or low should be noted from either the
      sound of the vacuum motor (Br91b, Bro92) of from
      the apparent velocity of the vacuum exhaust (Bro92).
      The sound of the motor (or the feel of the exhaust
      jet) at high and low flows can be determined by
      operating the vacuum, first with the nozzle in free
      air (high flow), then with the nozzle  blocked (low
      flow).

   --  (Optional) If the vacuum cleaner nozzle is fitted
      with a flow measurement device, the flows in the
      nozzle should be recorded.

-  All holes that have been drilled through  the slab must
   be permanently cemented closed after testing is com-
   plete.

 Interpretation of results

-  Communication is probably reasonably good and rea-
   sonably uniform if induced sub-slab depressurizations
   could be distinctly measured with the micro-manometer
   at each test hole  (or the smoke flow was distinctly
   down into  each test hole) with the vacuum cleaner
   operating.

   —  The stronger the measured depressurizations, the
      more confident one can be in this result.

   --  Also, if the house was probably near the maximum
      driving  force during the testing (cold weather, ex-
   haust appliances operating), one can be more confi-
   dent in this result.

—  As an option, a mitigator might qualitatively com-
   pare the baseline (vacuum off) pressure differen-
   tials  measured during  the testing  with the
   roughly-projected worst-case maximums for cold
   climates (about 0.025 to 0.035 in. WG). As a quali-
   tative approximation, was the vacuum  developing
   sub-slab depressurizations of the right order to po-
   tentially compensate for worst-case conditions? (See
   Note below.)

—  Relatively high flows in the vacuum cleaner would
   generally tend to support these suction field obser-
   vations.

-  The confidence that communication is  good  is in-
  • creased if aggregate was observed through the test
   holes.

-  If sub-slab depressurizations are marginal but flows
   are high, this  could be indicating that communica-
   tion is good but that the relatively low-flow vacuum
   cleaner is being overwhelmed by the available flow.
   In such cases, the ASD system (with perhaps double
   or treble the flow capacity of the vacuum) might
   perform well. (Marginal depressurizations and high
   flow could also be suggesting a leak through the
   slab near the vacuum.)

—  Note: It is re-emphasized that, in this qualitative
   testing, no effort has been made to try to make the
   vacuum cleaner simulate an actual ASD fan.  Thus,
   the actual sub-slab depressurizations measured with
   the vacuum cleaner will be different from those that
   will be established by the ASD system. Thus, any
   use of the actual sub-slab depressurizations mea-
   sured with the vacuum—e.g., to  assess whether
   they might be sufficient to compensate for worst-case
   house depressurization, as suggested above—can
   be done in only a qualitative manner.

When  communication  thus appears to be reasonably
good and uniform, only one or two SSD pipes will
likely be sufficient, and there will be flexibility  in
selecting where the pipes are to be located. (Or, if a
sump/DTD system is being considered, confidence is
increased that the system will perform well, even if the
drain tile loop is not complete.)

If the results are inconclusive, with distinct depressur-
izations in some test holes, but with no (or marginal)
depressurization in other holes, this may be suggesting
that communication is not uniform or that the suction
field is having difficulty extending to the more remote
portion of the slab, hi this case, the logical next step is
determined based upon the mitigator's experience with
other houses in the area.

~ The mitigator might elect to proceed with the instal-
   lation of a one- to two-pipe SSD system, if one or
                                                    87

-------
         more of the following conditions exist a) flows in
         the vacuum were high, and there are no nearby slab
         openings  through which air could have been
         short-circuiting  into the vacuum, suggesting that
         the vacuum cleaner was simply being overwhelmed
         by high flows from a sub-slab region having good
         communication; b) aggregate appeared to be present;
         c) the poor depressurizations were observed only at
         a few of the most remote test holes; d) the house
         was  tested under challenging conditions  (cold
         weather and exhaust appliances operating), so that
         the risk is reduced that the observed marginal de-
         pressurizations might be overwhelmed by more chal-
         lenging conditions later; and e) local experience
         suggests that houses such as the one tested can be
         adequately  treated  by  one or  two  pipes.
         Post-mitigation sub-slab suction field measurements
         should be performed to confirm that, in fact, the
         slab is being adequately treated.

      —  If conditions a) through e) above are not present, the
         mkigator might elect to proceed with the installa-
         tion, installing  a  conservatively-designed system
         (with   additional   suction  points).  Again,
         post-mitigation suction field measurements should
         be performed.

      —  The mitigator might elect to proceed with the instal-
         lation, conducting suction field measurements dur-
         ing installation.  The first one or two suction pipes
         would be installed, the system fan turned on, and
         those results used to determine where any addi-
         tional suction pipes should be located.

      —  The mitigator might elect to conduct more quantita-
         tive communication diagnostic testing (Section 3.3.2)
         in order to better quantify SSD pipe requirements.

   -  If the results suggest poor suction field extension, with
      no (or only marginal) depressurizations  observed in
      most or all of the test holes, then poor communication
      should be assumed. Under these conditions, the mitiga-
      tor would  be well advised either to conduct more
      quantitative suction field extension diagnostics (Sec-
      tion 3.3.2) prior to  installation, or to proceed with the
      installation doing the diagnostics during installation.

      —  Relatively low vacuum nozzle flows would be ex-
         pected in these cases.

3.3.2  Quantitative Assessment of
Suction Field Extension

As indicated above, one option when sub-slab communication
appears to be marginal or poor is to proceed with quantitative
suction field extension diagnostic testing. The primary dis-
tinction between quantitative and qualitative testing is that,
with quantitative testing, the flows in the diagnostic system
are adjusted to simulate the actual expected ASD flows. This
simulation  can be achieved either by adjusting a vacuum
cleaner to achieve the proper depressurization at a baseline
test hole 8 to 12 in. away from the vacuum nozzle; or, by using
an ASD fan to draw the suction through a 4-in. pipe, including
a suction pit under the pipe and simulating the expected
suction losses in the ASD system piping. Another feature of
quantitative testing is that, in general, more test holes are
drilled at different distances from the suction pipe (rather than
two to four holes at the corners of the slab).

The baseline test hole is located at a distance from the vacuum
nozzle such that if it were a SSD suction pipe rather than a
vacuum nozzle at that location, the baseline hole would be in
the sub-slab suction pit under the pipe. If the sub-slab depres-
surization at the baseline test hole during vacuum  cleaner
operation is adjusted to equal that expected in the SSD suction
pit, then, by simple fluid dynamic theory, the flow in the
vacuum will nominally be the same as that in the SSD pipe,
despite the fact that the vacuum motor has a very different
performance curve from that of the ASD fan. And because the
flows are the same, the sub-slab depressurizations at remote
test holes should nominally be the same.

In many (but  not  all) cases  where the more quantitative
suction field diagnostics have been conducted,  it has  been
found that the vacuum cleaner diagnostics over-predict the
number of SSD pipes that will be required to treat the house
(Ma89a,  Fo90,  FI91, Si91, Sau92). A  variety  of possible
explanations have been offered for why this occurs, including:
a) failure of the diagnostician to allow enough time (some-
times several minutes  or more in tight  soils) for the
vacuum-induced suction field to become established under
the slab  (Hi92); b) failure  to properly adjust the vacuum
cleaner flows; c) inability of a measurement at the baseline
test hole to accurately reflect the suction that a SSD pipe
would produce in an open pit at that location (Hi92, Sau92);
d) small leaks around the sample tubing while micromanom-
eter readings are being made at remote test  holes, partially
neutralizing any sub-slab depressurization that the vacuum
may be producing (EB92); and e)  local inhomogeneities at the
point where the vacuum nozzle is inserted (Hi92). Despite this
frequent over-prediction, with proper interpretation, this diag-
nostic testing still can provide useful guidance regarding ASD
system design. Efforts are continuing to develop better guid-
ance on how to interpret the results of this testing, to account
for its tendency to over-predict pipe requirements.

Again, the quantitative approach for suction field extension
measurements needs to be considered only in cases where
sub-slab communication is poor or uneven.

The basic setup for quantitative suction field extension testing
using a vacuum cleaner is illustrated in Figure 9.

 •  Equipment and materials required

   -   The more quantitative suction field extension measure-
      ments require the same equipment and materials as
      required for the qualitative testing (Section 3.3.1), ex-
      cept that the chemical smoke stick is no longer an
      option for determining sub-slab depressurizations. The
      digital micromanometer,  which permits quantitative
      determination of the suctions, is required.
                                                       88

-------
  A Magnehelic® gauge, with a scale extending above
  1.5 in. WG, is now mandatory to monitor the sub-slab
  depressurization at the baseline test hole, as illustrated
  in Figure 9. (The suctions produced in me baseline test
  hole when a 90-watt fan is being simulated—often 0.2
  to 1.5 in.  WG—are within the  upper scale of some
  micromanometers. However, there may sometimes be
  excursions to much higher suctions (up to 80 in. WG),
  and sometimes higher-suction fans will be simulated,
  perhaps  25 in. WG or higher. These suctions would
  damage  many micromanometers,  so that the micro-
  manometer is not recommended for use at this loca-
  tion.)

•  In addition, the vacuum cleaner must be equipped with
  a speed  controller or other means of flow adjustment,
  so that flows can be adjusted to achieve the desired
  suction in the baseline test hole. In Figure 9, this flow
  adjustment is illustrated as being achieved by a valve
  allowing adjustment of the amount of ambient  air
  bleeding into the vacuum intake nozzle.

 Test procedure

-   Select the location for one or more 1.25-in. vacuum
   cleaner suction holes.

   --  The criteria for site selection are similar to those for
      the more  qualitative approach. If the qualitative
      suction field testing (Section 3.3.1) was conducted
      initially, the same suction hole can be used for this
      testing.

   —  Since the poor communication will likely prevent
      the vacuum cleaner suction from extending to por-
      tions of the slab remote from  the suction hole, it
      may be desirable to have a second suction hole in a
      section of the slab remote from the first one. Testing
      at the  second  hole can serve to confirm the mea-
      surements made at the first hole,  or to  suggest
      whether communication varies in different parts of
      the slab.

-  Select the locations for the 1/4- to 1/2-in. suction test
   holes.

   - A baseline test hole must be drilled about 8 to 12 in.
      away from each vacuum suction hole. The ability to
      interpret the test results will depend upon the main-
      tenance of the proper sub-slab depressurization at
      this baseline test hole. The vacuum cleaner must be
      operated with the depressurization at that hole be-
      ing maintained at the level which a SSD fan would
      be able to maintain in a sub-slab pit beneath a SSD
      suction pipe.

   ~ Several remote test holes must be drilled. Ideally,
      about three such test holes would be drilled—one
      about 3 ft from the suction hole, another about 10 ft,
      and a third as far from suction  hole as possible—in
      each of two  or three directions away from  the
      suction hole. The distance of the most remote hole
        can be selected based upon experience regarding
        how far away suction from the vacuum might be
        expected to be detected. If the qualitative testing
        (Section 3.3.1) was conducted initially in the house,
        the test holes drilled for that testing can be utilized
        in the quantitative testing, if convenient.

     ~ The actual locations where remote test holes can
        practically be sited will be determined by the finish
        on the slab, and the location of sub-slab utility lines
        and other obstructions. Figure 10 illustrates a pos-
        sible scenario for locating suction and test holes on
        a fully finished slab, by drilling the holes in the
        corners of closets, under carpeting, and in available
        unfinished areas.

     Drill the suction hole and the  test holes, to the extent
     that these holes were not already drilled during prior
     qualitative suction field testing.

     - (Optional) If a grab sample is to be taken to deter-
        mine sub-slab radon concentrations at the test holes,
        that sample would be taken at this time, after a brief
        vacuuming  to  remove  dust.  Preferably, the hole
        should remain closed with rope caulk for awhile
        before sampling to permit re-equilibration of sub-slab
                            Note: Test holes indicated
                                  by small circles
    Baseline
    Test Hole
  II
Suction
Hole  _
Figure 10.  Example of siting vacuum cleaner suction  hole and
          sub-slab suction test holes for quantitative suction field
          extension testing, when slab is fully finished.
                                                      89

-------
   concentrations. See the related discussion in Sec-
   tion 3.3.1.

— Vacuum the holes  carefully to remove dust and
   drilling debris that may block the holes.

— Inspect the nature of the sub-slab material, possibly
   using the drill bit, to aid in the interpretation of the
   results, as discussed in Section 3.3.1.

— Seal the vacuum cleaner nozzle into the suction
   hole,  using rope caulk or putty  to close the gap
   between the nozzle and  the top of the slab,  trie
   nozzle should extend down about half the thickness
   of the slab.

-- Insert the sample tube from the Magnehelic gauge
   into the baseline test hole, extending the tube down
   about half the thickness of the slab. Seal the sample
   tube into the hole using rope caulk to close the gap
   between the tubing and the top of the  slab.

- Temporarily close the test holes with rope caulk.

Place the house  in the condition that will  be used
throughout this testing. It is recommended  that this
condition be with exhaustappliances operating, to simu-
late challenging conditions. If no  sub-slab depressur-
kations are measured under these conditions, the appli-
ances might be turned off (and perhaps windows opened)
to determine performance under less challenging con-
ditions (see Section 3.3.1).

With  the vacuum cleaner off, measure  the  sub-slab
pressurization (or depressurization) at each test hole
with the micromanometer, to determine the "baseline"
pressure  difference across the slab under the weather
and appliance conditions existing  during the  test. See
procedure for this measurement described in Section
3.3.1. This baseline measurement will indicate how the
existing conditions compare against the expected maxi-
mum slab pressure differential, and may thus  aid in
interpreting the results with the vacuum operating (see
Section 3.5.4).

Turn  on the vacuum cleaner, and  adjust the speed
controller so that the sub-slab depressurization mea-
sured by the Magnehelic gauge at the baseline test hole
is approximately that which is expected to exist in the
sub-slab  suction  pit beneath a SSD pipe when the
intended SSD fan is operating. This step is critical in
achieving quantitative results.

— With  the standard 90-watt ASD fans in relatively
   low-flow conditions, this suction will be roughly
   1.5 in. WG.

— In higher-flow cases, with the 50- or 90-watt fans,
   the suction can range from roughly 0.25 to  over 1
   in. WG, depending on system flow.
—  With the high-suction/low flow fans sometimes used
   in very poor communication cases, this suction can
   be several inches of water.

—  If the vacuum is unable to develop these desired
   suctions, and if there are no nearby slab openings
   through which air might be short-circuiting into the
   vacuum, this result suggests that sub-slab flows are
   high. While the vacuum will thus  not be able to
   quantitatively simulate the ASD system, the result
   suggests that sub-slab communication is good and
   quantitative diagnostics may thus not be necessary.

~  Note: If an ASD fan mounted on  a 4-in. pipe is
   being used to develop the suction field, rather than a
   vacuum cleaner, then the suction field will auto-
   matically reflect that which would be generated by
   a SSD system. In that case, no baseline test hole
   would be needed, and no effort will be required to
   control  the depressurization  there.  This  would be
   the case if: a) a sub-slab suction pit has been exca-
   vated under the 4-in. pipe; and b) the diagnostic fan
   system has been designed to simulate the suction
   losses in mitigation system piping. If the diagnostic
   ASD fan is in fact the permanent fan mounted on
   the  initial pipe  of a  SSD  system, i.e., if
   "during-installation" diagnostics are  being per-
   formed, as discussed earlier, then  system suction
   losses will be simulated by definition. If a portable
   ASD fan test stand is being used, piping losses can
   be simulated by reducing fan power, or by adding
   piping with elbows upstream or downstream of the
   fan.

(Optional) If the vacuum cleaner is fitted with a pilot
tube or other flow measurement device,  the flows in the
vacuum nozzle should be recorded as  the suctions at
the baseline hole are adjusted.  These  flow measure-
ments would comprise the sub-slab flow diagnostics
discussed in Section 3.5.1. From the sub-slab depres-
surization data at the baseline hole, and from the flow
data, a mitigator can judge what kind of fan will be
needed and roughly what suction it will maintain in the
suction pit given observed flows.  This rough assess-
ment can be used to select the sub-slab depressuriza-
tion at the  baseline hole to be used for these suction
field extension diagnostics. (As a minimum, if there is
no pilot tube, qualitatively assess vacuum flows through
the sound of the motor or the feel of the exhaust.)

Wilh ihe micromanometer, measure Ihe depressuriza-
tion created beneath the slab at  the remote tesl holes,
with the vacuum cleaner adjusted to maintain the proper
depressurization at the baseline hole as  selected above.

—  With the rope caulk closing the other test  holes,
   insert sample tube from the  micromanometer into
   each tes"t hole in turn, to a deplh  aboul half the
   thickness of ihe slab. Carefully seal gap belween
   tubing and slab with rope caulk or putty. Failure to
   effectively seal this penetration could partially neu-
   tralize the sub-slab suction, significantly impacting
                                                  90

-------
      results when the sub-slab depressurizations are mar-
      ginal to begin with.

   —  If communication is poor, it may take several min-
      utes  with  the  vacuum cleaner operating for the
      suction field to become established. Where commu-
      nication is marginal but not poor, the suction field
      may become established within 30 seconds or less.
      If initial measurements show unacceptably low de-
      pressurizations induced by the vacuum, and if these
      measurements were made before (he vacuum had
      operated for several minutes, they should be re-
      peated after several minutes.

   —  Observe the reading  on  the micromanometer at
      each test hole. Record the range and/or average of
      the readings. Compare these values with the values
      measured before the vacuum cleaner was turned on.

   —  Check the zero on the micromanometer often, pref-
      erably before each reading in cases where the mea-
      sured depressurizations are very low.

   -  Periodically re-check the Magnehelic gauge to en-
      sure that proper depressurizations are being main-
      tained at the baseline test hole.

-   After testing is complete, permanently close all of the
   slab holes drilled for this testing using hydraulic ce-
   ment or other non-shrinking cement.

 Interpretation of results

-   The measured depressurizations can  be used to plot
   lines of constant sub-slab depressurization on the house
   floor plan. See Figure 11.

   —  Since the number of test holes will be limited, and
      communication can vary in different directions away
      from the suction hole, the actual curves that can be
      drawn in practice will be more approximate and
      irregular than those shown in Figure 11.

   —  As discussed previously, because the depressuriza-
      tion being maintained by the vacuum cleaner at the
      baseline test hole was  about the same as that which
      would be maintained  by  a  SSD fan in  a sub-slab
      suction pit, the lines of constant depressurization in
      Figure 11 theoretically should be about the same as
      if a SSD pipe had been operating at the location of
      the vacuum nozzle.

-  By this interpretation, the effective suction radius of a
   SSD pipe would be the distance from the baseline test
   hole to that constant-depressurization line  where
   sub-slab depressurization has fallen to the "minimum
   acceptable" level.

   -  If the house was depressurized during the diagnos-
      tic testing—if the  house  was closed, the weather
      was cold, and exhaust appliances were operating—
      the isobars represent depressurizations achieved
Figure 11.  Example of graphical interpretation of the results from
          quantitative  suction field  extension measurements.
          (Adapted from Reference Fo90)
        under challenging conditions. Under these condi-
        tions, the "minimum acceptable" sub-slab depres-
        surization may be the lowest suction measurable
        with the micromanometer (about 0.001 in. WG),
        since, after installation, the ASD system will rarely
        see house depressurizations greater than those ex-
        perienced during the testing.

        If the house was being depressurized by operation
        of exhaust appliances during the diagnostic testing,
        but if the weather was mild, the "minimum accept-
        able" sub-slab depressurization should be  great
        enough to account for the increased house depres-
        surization that might be expected due to  thermal
        effects in cold weather. Thermally induced depres-
        surization will depend upon the local climate; in a
        two-story house in cold climates, the depressuriza-
        tion in  the basement  could be roughly 0.015 in.
        WG. In theory, ideally, the "minimum acceptable"
        sub-slab depressurization in mild weather should be
        about this level in cold climates.

        As discussed in Section 3.5.4, review of the pres-
        sure differentials measured across the slab before
        the vacuum cleaner was turned on can provide some
        guidance regarding whether the goal under these
                                                     91

-------
   conditions should in fact be to achieve 0.015 in.
   WG sub-slab depressurization with the vacuum on.
   Were the exhaust appliances depressurizing the
   house during mild weather to a degree approaching
   0.025 to 0.035 in. WG, the conservatively estimated
   cold-weather maximum discussed in Sections 2.3.1b
   and 2.3.le (perhaps  due to  operation  of a
   whole-house exhaust fan that would not be operat-
   ing in cold weather)? If so,  the target "minimum
   acceptable" sub-slab depressurization  with the
   vacuum cleaner on under these conditions could be
   low (around 0.001 in. WG), since  the diagnostic
   tests were  apparently being conducted under
   near-worst-case conditions despite the mild weather.

—  If the house was not being depressurized by opera-
   tion of exhaust appliances, but if the weather was
   cold, the "minimum acceptable" sub-slab depres-
   surization should be great enough to account for the
   increased house depressurization that might be ex-
   pected due to appliance operation. As discussed in
   Section 2.3.1b, appliance-induced depressurizations
   can be as high  as 0.01 to 0.02 in. WG. In theory, it
   would thus be logical for the "minimum accept-
   able" sub-slab  depressurization to be about 0.01 to
   0.02 in. WG  under these conditions. Again, as
   discussed above, review of the pressure differen-
   tials across the slab with the vacuum cleaner off
   might provide some guidance about how closely the
   expected maximum house depressurizations  were
   being approached during the testing,  and thus
   whether the ASD system in fact needs  to be de-
   signed to accommodate an increased challenge as
   great as 0.01 to 0.02 in. WG.

—  If the suction field diagnostics were conducted when
   the weather was  mild  and without operation of
   exhaust appliances, the  "minimum  acceptable"
   sub-slab depressurization would nominally equal
   the conservative maximum estimated house depres-
   surization resulting  from combined thermal and
   appliance operation effects (0.025 to 0.035 in. WG).

In practice, SSD systems tend to perform effectively
with fewer pipes than would be predicted based upon
the effective suction radius derived from these diagnos-
tics. Thus, the "minimum acceptable" sub-slab depres-
surization used in  this analysis can probably be on the
low end of the values discussed above.

In addition, as  discussed in Section 2.3.le (Climatic
conditions), this conservative maximum basement de-
pressurization range due to combined thermal and ap-
pliance effects (0.025 to 0.035 in. WG) would be less
in mild climates, or in cases where the house was leaky
or did not  have some of the more common exhaust
appliances (such as central furnace fans, clothes driers,
and attic fans). The upper end of this conservative
range assumes that  major depressurizing appliances
will be operating in  cold weather; but in fact, among
these appliances, whole-house and attic fans will not be
operated then, and clothes driers will be operated only
intermittently. Although this depressurization range may
      be conservatively high for many houses, it is used here
      as a conservative but reasonable design tool which can
      be useful as long as it is properly understood.

   -  Superimposing this effective suction radius onto a house
      floor plan, the number and location of SSD suction
      pipes can then be selected in an effort to ensure that all
      (or most) of the slab achieves the desired sub-slab
      depressurization.

      ~  Particular attention should be placed on ensuring
         that the proper depressurization is achieved around
         the slab perimeter, and near any other locations
         where significant  soil gas entry might be antici-
         pated. Maintaining good depressurization is prob-
         ably somewhat less important in the center of un-
         broken slabs.

3.4    Procedures  for  Radon Grab
Sampling and Sniffing
In some cases, grab sampling or sniffing, to determine radon
concentrations inside or near potential soil gas entry routes,
may help identify the  most important entry routes and may
thus aid in the design of the ASD system. Examples of where
such radon measurements might help include:

a)  Identification of regions of unusually high radon concen-
    trations beneath slabs. In selecting the locations for SSD
    pipes, pipe location might be biased toward regions where
    high sub-slab radon concentrations are near slab open-
    ings.

b)  Identification  of hollow block foundation walls having
    unusually high radon levels inside the cavities. A wall
    having a high concentration might warrant a BWD com-
    ponent being added to the SSD system.

c)  In houses having fieldstone foundation walls, determina-
    tion of radon concentrations in the chinks between the
    stones. High radon levels would suggest that the wall is
    an important entry route, and that SSD treating the slab
    alone may not be sufficient.

d)  Identification  of radon  levels in  floor drains or other
    openings, to assess their potential importance as entry
    routes.

e)  Identification  of radon levels in crawl spaces adjoining
    basements or slabs  on grade, to assess  the potential
    importance of the crawl space as an entry route compared
    to the adjoining wing.

Grab sampling/sniffing can be considered in any case where a
mitigator's prior experience has suggested that it might be
cost-effective, reducing the overall costs to the homeowner.
Where suction field extension testing is  being  conducted
(usually where communication is  unknown or poor), grab
sampling/sniffing beneath the  slab is a relatively easy addi-
tional measurement to make, if the mitigator has  the proper
equipment.
                                                 92

-------
A distinction is drawn between "grab sampling" and "sniff-
ing." Both of these approaches involve drawing a gas sample
over a very  short time period—no more than five minutes,
often less. The distinction between the two approaches is:

a)  Grab sampling is a quantitative approach. Depending
    upon the particular grab technique,  the sample  is col-
    lected,  allowed to age, or counted for a long enough
    period to permit the concentration of radon in the sample
    to be quantitatively determined with an acceptable sensi-
    tivity. For perhaps  the most common grab sampling
    approach (use of an alpha scintillation cell), this period
    can be about four hours; the sample needs to age for
    about this period to ensure equilibrium between the radon
    gas and its decay products.

b)  Sniffing is a less  accurate, semi-quantitative approach.
    Measurement accuracy and sensitivity are sacrificed in
    order to obtain essentially immediate results, enabling
    rapid identification of those sampling locations which
    were the "hottest" In these  measurements, no time is
    allowed for sample aging, and sample counting time is
    reduced to a minimum.

To the extent that such radon measurements are included in
pre-mitigation diagnostic testing, sniffing will often be the
preferred approach for commercial mitigators. The simplicity
and reduced time requirements for sniffing will be important
in keeping costs down. The improved accuracy of grab sam-
pling will generally not be  needed, in view of the anticipated
qualitative use of the results.

Several techniques are available for conducting grab sampling
and sniffing.

a)  Alpha scintillation cell. This is perhaps the most common
    technique, and is one of three grab sampling techniques
    covered in EPA's measurement protocols (EPA92d). A
    sample is drawn into a cell having a zinc sulfide phosphor
    coating  on its  interior surfaces. As the radon  and it
    progeny decay, the alpha particles released cause light
    pulses  (scintillations)  when they impact the phosphor;
    these scintillations are counted using a photomultiplier
    tube and sealer. In practice, an alpha-scintillation-based
    continuous radon monitor is often used for these mea-
    surements. For quantitative grab sampling, the  sample
    must be aged for a sufficient period  after collection
    (usually about four hours) to ensure that the radon and its
    decay products are in equilibrium. The calibration of the
    scintillation cell (i.e., the number  of counts for each
    picocurie  per liter) assumes that  the alpha  particles are
    being released by radon and progeny in equilibrium. For
    sniffing, the sample from a given sampling location is
    pumped through the scintillation  cell continuously, and
    the counting is completed in a very brief period (usually
    within about 5 minutes or less, as discussed later). In the
    sniffing case, it is recognized that the sample is nowhere
    near equilibrium, and that the counts per minute recorded
    by the device thus cannot be quantitatively related to a
    radon concentration (except, in specific cases, via sophis-
    ticated calculations).
b)  Activated charcoal. In another technique covered in EPA's
    protocols (EPA92d), a grab sample from a given location
    is drawn through a charcoal-filled cartridge for about an
    hour. The  cartridge is  then analyzed by placing it on a
    sodium  iodide gamma scintillation system or a germa-
    nium gamma detector. This technique does not appear to
    be as widely used as is alpha scintillation. Also, because
    of the required sampling duration, it is not applicable for
    sniffing. Accordingly, it is not addressed further here.

c)  Pulsed ion chamber. One continuous radon monitor on
    the market operates on the principle of detecting, using an
    electrometer, ion pulses generated when radon gas de-
    cays. In this device, radon decay products are constantly
    removed from the measurement chamber, and thus nomi-
    nally do not contribute to the electrometer reading. As a
    result, there is no need to age the sample in order to
    achieve equilibrium, as is  necessary with alpha scintilla-
    tion cells. The manufacturer offers a pumped
    (flow-through) variation of this monitor with  an im-
    proved electrometer that can be used for grab sampling
    and sniffing (Fe92). With the pump continuously draw-
    ing sample through the chamber, this device reportedly
    has adequate sensitivity to provide readings after 2 to 20
    minutes of sampling. While the required duration will
    depend upon the radon concentration, it may generally be
    viewed  that longer sampling/measurement periods will
    tend to give the accuracy and sensitivity associated with
    quantitative grab samples, while shorter periods will tend
    toward  less accurate,  more semi-quantitative sniffing
    samples. Although the pulsed ion chamber approach is
    not currently included in EPA's grab sampling protocols,
    it  is a  convenient method for  grab sampling, and is
    utilized by a number of investigators.

d)  lonization meter.  Another device which has been mar-
    keted to provide immediate readings of radon concentra-
    tions is  based upon measurement of the total ionization
    existing in the air (IL90). This meter is thus distinguished
    from the pulsed ion chamber above, which measures ion
    pulses caused by radon decay. The accuracy and sensitiv-
    ity of the ionization meter for radon grab sampling appli-
    cations has not been tested by EPA, and experience with
    this device is very limited in the U. S. Accordingly, the
    ionization meter is not discussed further here.

In interpreting  the results from radon grab sampling/sniffing,
the mitigator must recognize that the real contribution from a
potential entry route  will  depend, not only on the radon
concentration at/in the opening, but also  on the flow rate of
soil gas into the house through the opening. Thus, the radon
levels alone do not necessarily determine how important a
potential entry route may be. An opening with a less elevated
radon level but a high flow (such as an uncapped hollow-block
wall) can be a more important contributor to indoor radon
than one with a high radon concentration but low flow (such
as a hairline slab crack).  As another example, sub-slab radon
concentrations may be relatively high in the central portion of
an uncracked slab, but almost none of this sub-slab radon will
be entering  the house, since convective flow is zero and
diffusion through the unbroken concrete is negligible. By
contrast, the sub-slab radon concentrations near the slab pe-
                                                        93

-------
rimeter may be lower, but flows into the house through the
wall/floor joint and the block foundation walls can be rela-
tively significant, often making the slab perimeter an impor-
tant location for SSD pipes despite somewhat lower sub-slab
radon concentrations there.

In research  studies, more elaborate diagnostic procedures
have been developed which consider potential flow as well as
radon levels at slab test holes, in order to obtain a better
measure  of radon  entry potential around  the  slab (Tu90).
However, in commercial mitigation installations, efforts to
determine flows will probably almost never be cost-effective.
The logical approach is to try to "guesstimate" whether flows
into the house are likely to be relatively high or low through
the entry route being sampled.  For sub-slab measurements
through test holes in the slab, are there significant slab open-
ings near the test hole? For concentrations inside block foun-
dation walls, the focus should probably be on whether these
levels are significantly higher than is observed  in the house,
and whether one wall in the house seems more elevated than
the other walls. The qualitative nature of this interpretation
underscores  the statement above, that the more qualitative
sniffer approach will likely be adequate for these measure-
ments in many  cases, and that the additional time and effort
required for more accurate grab sampling may often not be
warranted commercially.

The radon measurement procedures being discussed here are
for diagnostic purposes only, to aid in ASD system design. As
presented here, these procedures are not intended to determine
the indoor radon concentrations in the house.

The applications of grab sampling/sniffing discussed above
focus on its possible utility as apre-mitigation diagnostic test.
It may be more important as a post-mitigation diagnostic, for
trouble-shooting in cases where the initial mitigation installa-
tion is not performing as expected.

The equipment and procedures listed below address the scin-
tillation cell and the pulsed ion chamber approaches. Addi-
tional discussion of the alpha scintillation cell approach can
be found in References EPA88b, Fo90, and EPA92d.

 *  Equipment and materials required (alpha scintil-
    lation cell approach)

   -  Alpha scintillation cells.  Either single-valve (evacu-
      ated) cells or double-valve (flow-through) cells can be
      used for grab sampling; flow-through cells are required
      for sniffing. Commonly,  cells having a volume  of
      about 100 to 300 mL are used for these diagnostics.
      Cells ranging from 100 mL to 2 L are available; the
      larger cells provide greater sensitivity, unnecessary for
      the purposes of mitigation diagnostics.

   -  For flow-through cells, a battery-operated pump to
      draw gas sample into  the  scintillation  cell.  A
      hand-operated suction bulb could also be used to draw
      the gas sample. If a continuous radon monitor is used
      for this testing, as is usually the case, the pump associ-
      ated with the monitor will be used.
-  For evacuated cells, a pump capable of evacuating the
   cells to at least 25 in. of mercury before use.

-  Portable photomultiplier tube and sealer to count the
   scintillations, with a digital display or printer to indi-
   cate counts per unit time. Most commonly, a continu-
   ous radon monitor fitted with a scintillation cell will be
   used in the  field;  this monitor will incorporate the
   sample pump, the photomultiplier tube, and the sealer.

-  Flexible sample tubing, to fit through the test hole and
   draw the sample into the scintillation cell.

-  0.8-micron filter assembly, to be mounted in sample
   tubing  upstream of scintillation cell, to prevent dust
   (e.g., from the sub-slab)  from contaminating the cell.
   The filter will also remove radon decay products, which
   is important in  interpreting results when short aging
   times are used.

-  Rope caulk, to seal the gap between the sample tubing
   and the face of the slab.  Duct tape can also be useful
   when the  tubing is inserted through large openings,
   e.g., in a block wall.

 Test procedure (alpha scintillation cell approach)

-  Prior to visiting house, purge the scintillation cells with
   outdoor air and allow to age overnight. Because these
   samples are for the purpose of mitigation diagnostics
   rather than for  determining indoor radon concentra-
   tions, it is not necessary to purge the cells with aged air
   or nitrogen as specified in EPA's indoor measurement
   protocols (EPA92d). Perform a background count over
   a period of 2 minutes (Fo90) to 10 minutes (EPA92d)
   with the portable photomultiplier tube/sealer, to  con-
   firm that the background has been reduced to an ac-
   ceptably low level. The background counts per minute
   should be less than 10% of the expected sample counts
   per minute. Cells which will be used for high-radon
   samples (e.g., from beneath the slab)  can tolerate a
   higher background count than those that may be  used
   for lower-level samples.

-  For sub-slab samples, gas samples  will normally be
   drawn through 1/4- to 1/2-in. diameter test holes drilled
   through slab, often in connection with suction  field
   extension testing.

   --  After the test hole is drilled, clean the dust out of the
      hole, operating the vacuum cleaner as briefly as
      possible (only a few seconds) in order to minimize
      any artificial reduction in the sub-slab radon level.

   ~  Ideally, the hole should be temporarily closed with
      rope caulk or duct tape for 15 to 30 minutes, to
      permit the sub-slab concentrations to recover from
      any dilution resulting from the drilling and vacuum-
      ing  process  (Fo90). In practice, a delay this  long
      may not always be practical. In such cases, a delay
      of only a few minutes  may be sufficient—espe-
      cially with flow-through cells, which draw a large
                                                        94

-------
      sample volume—in view of the qualitative way in
      which the results will generally be interpreted, as
      long as all of the test holes are treated in the same
      way.

   -  Insert sample tube into test hole, and use rope caulk
      to seal the gap between the tubing and the top of the
      slab. The measured radon  concentration  will be
      significantly reduced if the gap is not effectively
      sealed, because significant  amounts of house air
      will be drawn down through the hole, diluting the
      soil gas sample.

-  For samples from inside a block foundation wall, gas
   samples will normally be drawn through pre-existing
   holes through the blocks, e.g., around utility penetra-
   tions such as water/sewer lines and electrical junction
   boxes.

   ~  After the sample tube has  been  inserted  through
      such wall openings, an effort should be made to
      close the remainder of the opening using rope caulk,
      duct tape, or plastic sheeting (depending on the size
      of the  hole), to  reduce the amount of house air
      drawn in with the sample.

-  Likewise,  samples drawn from inside other openings
   (such as floor drains and chinks in  stone foundation
   walls) would be drawn by inserting the sampling tube
   into the opening and then closing the opening with duct
   tape or plastic sheeting as possible, to reduce dilution
   of the sample.

   —  Leaving the opening closed in this manner for some
      period of time before drawing the sample would
      give the radon  concentrations in the opening  an
      opportunity to rise toward their maximum levels.

-  Samples of air from large regions (such as crawl spaces)
   would be obtained simply by drawing the sample from
   a central position within that region.

-  The sample train will consist of:

   —  A short length of sample tubing, leading from the
      test hole to the 0.8-micron filter;

   —  A second length of tubing, leading from the filter to
      one port of the scintillation cell;

   — For flow-through cells, a third length of tubing,
      leading from the second port of the scintillation cell
      to the sample pump (or hand suction bulb).

-  The preferred sampling and counting procedure for
   grab sampling is as follows:

   —  If using an evacuated cell, open the valve for 10 to
       15 seconds to allow the sample to fill the cell.
      (Note: One possible advantage of evacuated cells is
      that they draw only one-third to one-tenth the sample
      volume drawn by comparably sized flow-through
cells, thus minimizing any impact of the sampling
process on the radon concentrations in the region
being sampled.)

If using a flow-through cell, operate the pump suffi-
ciently long to draw at least three cell volumes of
sample gas through the cell. For indoor radon mea-
surements, EPA recommends 10 cell volumes
(EPA92d). This should ensure that the sample gas
has effectively displaced almost all of the previous
(outdoor air) sample. Depending upon the volume
of the cell and  the  flow rate of the pump, the
sampling time required could range from a less than
a minute, to several minutes.

Close the cell ,valve(s) immediately after sampling.
Allow the cell to sit for four hours, to permit the
radon progeny to come  into equilibrium with the
radon gas.

Place the cell in the photomultiplier/scaler scintilla-
tion counter, if necessary. (If a continuous radon
monitor is being used, the cell will already be in the
counter.) If the cell is being placed in  a separate
counter, allow a 2-minute delay after the cell is
placed in the counter before commencing counting,
to avoid counting spurious scintillations caused when
the window of the cell is exposed to bright ambient
light.

Count for as long as necessary to obtain the number
of counts needed to provide the desired sensitivity,
in view of background counts.  Counting time will
depend upon the radon concentration in the sample,
and hence the counts registered per minute. For
radon mitigation diagnostics, some sources feel that
only 100 counts is sufficient for semi-quantitative
purposes; for typical cell efficiencies (counts/minute
recorded per pCi/L in sample), 100 counts would be
obtained within a fraction of a minute (for samples
well over  20 pCi/L) to perhaps 10 minutes  (for
samples below 4 pCi/L), depending on concentra-
tion. EPA's protocols for indoor radon measure-
ments (EPA92d) recommend at least 1,000 counts
for good statistics; this number of counts would be
obtained within about 15 seconds to an hour. The
time over which the counts are measured must be
measured, if the counter is not part of a continuous
radon monitor that counts for specified periods of
time.

Convert the resulting counts into the radon concen-
tration, using the calculational procedure indicated
in Figure 12 (adapted from EPA92d).

After counting, purge the cell with  outdoor air (or
with aged air or nitrogen) to minimize the buildup
of the cell background. Cells which were exposed
to particularly high-radon samples may have to sit
unused for a day before the background counts per
minute has returned to acceptably low values.
                                                     95

-------
   1,  Divide the total number of observed scintillation counts
      by the total counting time, to  determine the  average
      counts per minute (cpm).

   2.  Calculate the radon concentration in picocuries per liter
      using the following formula:

   Radon concentration (pCi/L) =

  (cpm from sample)-(background cpm)  	1	  C_
      Cefl efficiency (cpm per pCi/L)      Cell volume (L)  A

      In this equation, C and A are correction factors account-
      ing for decay of radon and its decay products in the
      sample during the aging  and counting processes, ob-
      tained as discussed in 3 and 4 below. In radon mitigation
      diagnostics, where aging and counting times  may be
      reduced and where inaccuracies in the measurement are
      being accepted, the ratio C/A will generally be so close to
      1 that this term can be ignored.
  3.
C is a factor accounting for the radioactive decay of the
equilibrium radon and decay products in  the sample
during counting.
      C-(1,000063)N where N*
      the sample was counted
                      ; number of minutes over which
   C = (1.00378)N where N = number of hours over which sample
   was counted.

   The figure in the second equation is saying that the equilibrium
   cpm  at the beginning of a 1-hour counting  period will be
   0.378% greater than the average cpm over the entire period.

   For example, if the sample was counted for 30 minutes (0.5
   hour):

   C = (1.000063)30 = (1.00378)0-5 = 1.00189.

4.  A is a factor accounting for the decay of the  originally present
   radon and decay products in the sample during the time that
   the sample is aging, prior to counting.

   A = (0.99987)N where N = number of minutes over  which
   sample was aged, or

   A = (0.99248)N where N = number of hours over which sample
   was aged.

   The figure in the second equation is saying that, after one hour
   of aging, 99.248% of the radon still remains.
Figure 12. Procedure for calculating sample radon concentrations from measured counts when using the alpha scintillation cell technique for
          grab sampling.
      A modified sampling and counting procedure for grab
      sampling, intended to reduce the sample aging time at
      the expense of accuracy, is as follows.

      — Obtain the sample as in the preferred procedure
         above.

      — Allow the scintillation cell to  sit for only 5 to 10
         minutes after the end of sampling, rather than 4
         hours. This period permits reasonable grow-in of
         the first radon decay product, polonium-218—one
         of the  two alpha-emitting progeny—so that the
         sample is beyond the initial period of rapid alpha
         growth. It also provides time for essentially com-
         plete decay of any thoron in the sample along with
         the first thoron decay product, both of which are
         alpha emitters.

      -- The sample is then counted as above.

      — The number of scintillation counts obtained by this
         approach can be converted to a reasonably accurate
         radon  concentration, despite the non-equilibrium
         conditions existing during counting, by appropriate
         mathematical treatment of the non-equilibrium situ-
         ation (Sc92). If the approach in Figure 12 is used,
         which assumes equilibrium, the result must be con-
         sidered only semi-quantitative.
                                                            The sampling and counting procedure for sniffing is as
                                                            foUows (EPA88b,Fo90):

                                                            ~ Because the scintillation cell must be used in a
                                                               flow-through mode with frequent counting, from a
                                                               practical  standpoint, this procedure  requires  an
                                                               alpha-scintillation-based continuous radon monitor
                                                               equipped with a flow-through cell. Evacuated cells
                                                               and passive cells cannot be used in this application.

                                                            ~ Operate the sampling pump continuously, so that a
                                                               gas sample is being continuously drawn through the
                                                               scintillation cell during counting.

                                                            — Set the counter to make counts over a fairly brief
                                                               interval, just long enough to give a reasonable num-
                                                               ber of counts for the radon concentrations  in the
                                                               samples being  drawn. Commonly, a 30-second
                                                               counting interval is used. A shorter interval could
                                                               be used for very high-concentration samples, and a
                                                               longer interval may be useful for low concentra-
                                                               tions.

                                                            — After beginning to draw a sample from the selected
                                                               sniffing location, allow the system to cycle through
                                                               a number of counting intervals, until the number of
                                                               counts per interval becomes acceptably steady. Com-
                                                               monly,  ten intervals are used. When the counts per
                                                               interval have become reasonably steady, the last
                                                               several intervals might  be averaged  to  give the
                                                          96

-------
      value for that sampling location.  Others  suggest
      simply taking the counts for the last interval.

   —  The results must be left in the form of number of
      counts, and treated as a qualitative/semi-quantitative
      number. There is no theoretical basis for converting
      these counts to an  actual radon concentration, al-
      though some investigators have developed an em-
      pirical correlation between sniffing counts and ac-
      tual  radon  levels  for their particular equipment
      (EPA88b).

   ~  Since the background counts in the  cell will be
      impacted by  the samples drawn from the initial
      sniffing locations, it is recommended that the cell
      periodically be purged with outdoor air. During
      purging, the monitor would be taken outdoors for a
      few minutes, the pump operated continuously, and
      the  counts  per  interval observed  until the back-
      ground counts have dropped to an acceptably low
      value.

      If the previous sample had a particularly high radon
      concentration, the background created by this sample
      may be sufficiently high that it  will not return to a
      suitable value in a short period of time. In that case,
      this  cell should not  be used for  another sample
      having a potentially  lower radon  level, until  the
      background from the previous  sample has had an
      adequate chance to decay.

 Equipment and procedure  (pulsed  ion chamber
 approach)

-  In place of the alpha scintillation cells and photomulti-
   plier tube/sealer required  for the alpha scintillation
   approach, the pulsed ion chamber approach requires an
   ionization chamber with an electrometer system to
   detect and count the ion  pulses generated when radon
   atoms release  alpha particles. Radon  decay products
   are continuously removed from the chamber electro-
   statically, and thus nominally do not contribute to the
   ion pulses by their decay. In practice, the ion chamber,
   the electrometer system, and the pump required to draw
   samples through the chamber in this application, are all
   contained in a continuous radon monitor (Fe92).

-  The pulsed ion  chamber requires the same type of
   sample train as that listed previously for the scintilla-
   tion cell approach: flexible sample tubing; a filter to be
   placed in the tubing between the sampling location and
   the monitor; and rope caulk or other material to seal the
   gap between the sample tubing and the slab or wall
   opening though which the sample is being drawn.

-  During sampling, the sampling pump operates continu-
   ously, drawing a sample through the ionization cham-
   ber. Ion pulses are counted as the sample is being
   drawn. Sampling must continue until an adequate num-
   ber of pulses have been recorded to provide the desired
   accuracy and sensitivity. Since radon decay products
   play no role in the  counts, there is no need to be
   concerned about time required for sample aging, as
   with scintillation cells.

-   The necessary counting duration will depend upon the
   radon concentration in the sample, and hence the ion
   pulse rate. The manufacturer recommends a counting
   time of either 2 or 20 minutes (Fe92).

-   With this device, a grab sample may be viewed as one
   where the concentration and counting time are such
   that the number of counts are sufficient to give reason-
   able accuracy. A sniffing sample may be viewed as one
   where the number of counts is  so low that the results
   must be considered semi-quantitative.

 Interpretation of results (scintillation cell or pulsed
 ion chamber approach)

-   If the sub-slab radon concentration in one part of the
   slab is significantly greater than that in other parts (by a
   factor of three or more), and if there are potential entry
   routes in that part of the slab, then a SSD suction pipe
   should probably be placed in that part of the slab (if
   communication is not good), or at least biased toward
   that part (if communication is relatively good).

   —  If there are not slab openings in the vicinity of the
      "hot spot," SSD pipe location should usually be
      biased toward the entry routes  preferentially  over
      the high-radon areas.

   ~  Note that radon distributions under a slab can some-
      times be influenced by weather conditions, such as
      winds and precipitation.

-   If radon concentrations inside a hollow-block founda-
   tion wall are unusually high, then the mitigator might
   be prepared to supplement the SSD or DTD system
   with a BWD component treating that wall.

   —  Weather  conditions, especially  wind velocity, can
      influence radon concentrations inside block walls.

-  If high radon levels are found in the joints of fieldstone
   foundation walls, this result suggests that the wall may
   be an important entry route. Unless the native soil is
   fairly permeable, a S SD system may not be expected to
   treat the soil on the exterior face of the wall, and thus
   may not treat wall-related entry. Especially if this result
   is found in combination with poor sub-slab communi-
   cation, some mitigation technique other than ASD (e.g.,
   basement pressurization or ventilation) should be con-
   sidered.

-  If high radon levels are found in a floor drain, the floor
   drain is probably not trapped. The drain  should be
   trapped, both to prevent radon entry through this route
   and to avoid potential leakage of house air do wn through
   the drain into an ASD system.

-  If high radon concentrations are found in the air in a
   crawl space adjoining a basement, the mitigation sys-
                                                     97

-------
      tern may need to include an SMD leg in the crawl space
      to supplement the SSD or DTD leg in the basement.
 face), to define "low" flows. These measurements will enable
 interpretation of observed flows during sub-slab testing.
3.5    Procedures for Other Types of
Diagnostic Tests
As discussed in  Section 3.1.2, there are a variety of other
diagnostic tests that a mitigator may sometimes choose to
perform, in addition to the visual survey (Section 3.2), suction
field extension measurements (Section 3.3), and grab sam-
pling/sniffing (Section 3.4). These other tests will have a
practical benefit, and be cost-effective, less often than will
those discussed in Sections 3.2 through 3.4. They will gener-
ally be warranted as pre-mitigation tests only when particular
conditions exist.

These other diagnostic tests are discussed below.

3.5.7  Sub-Slab Flows

This diagnostic test involves measurement of the flows gener-
ated in the vacuum cleaner, as an add-on to the suction field
extension testing discussed in Section 3.3. Commonly, this is
done using a pilot tube. Other options are measurement of the
pressure drop across a calibrated flow orifice in the vacuum
cleaner nozzle, or some type of anemometer.

Sub-slab flow measurements can be used to aid in interpreting
the suction  field  extension results,  and  in  selecting  an
appropriately-sized fan for an ASD system.

 •  Simple sub-slab flow measurement

In ils simplest form, sub-slab flow testing involves recording
the flow in the vacuum as the suction field extension test is
being made, i.e., at the one vacuum cleaner speed setting used
for the suction field testing. In the qualitative  approach for
suction field extension measurements, the flow measurement
would  thus be made at maximum speed in the vacuum (see
Section 3.3.1). Under these conditions, the measured flows in
the vacuum cleaner would not represent flows that could be
expected in the SSD system, but would be only a qualitative
indicator of whether sub-slab flows are high or low. In the
quantitative approach, it would be made at the vacuum speed
that was selected to maintain the sub-slab depressurization at
the baseline test hole at the suction expected to be maintained
by the ASD fan in the SSD suction pit (Section 3.3.2). Under
these conditions, the vacuum flows should nominally simulate
the ultimate SSD flows.

Such flow measurements are so  easy to make that they should
be considered any time pre-mitigation suction field extension
diagnostics are being done. As a minimum, the flows should
always be qualitatively estimated from the sound of the vacuum
motor or the feel of the exhaust (Br92, Bro92).

If a mitigator is unfamiliar with  the flow characteristics of the
vacuum cleaner being used, flows in the nozzle should be
measured: a) with the nozzle in free air, to show the highest
expected flows; and b) with the  open end of the nozzle closed
fairly tightly (e.g., by pressing  it against some suitable sur-
 Relatively high flows in the vacuum during sub-slab testing
 would tend to confirm observed good sub-slab communica-
 tion. With good flows and good communication, the typical
 50- or 90-watt centrifugal in-line tubular fan will generally be
 the appropriate choice for the ASD system, because these fans
 generate adequate suction at relatively high flows (1-2 in. WG
 at zero flow, 125-270 cfm at zero static pressure). High flows
 might also be revealing short-circuiting of air into the vacuum
 from some  nearby slab opening, especially if they are ob-
 served in conjunction with limited suction field extension. If
 inspection suggests that short-circuiting is  the cause of the
 high flows, then sealing of the leakage points may be an
 important component of the system installation;  or, to the
. extent that complete sealing is not possible, the selected fan
 must be able to provide the necessary suctions while handling
 this leakage flow.

 Relatively low flows in the vacuum cleaner during sub-slab
 testing would tend to confirm observed poor sub-slab commu-
 nication. Depending  upon how low the flows are, the 50- or
 90-watt in-line tubular fans may still be a reasonable choice.
 However, at very low flows, a much higher-suction fan de-
 signed for low-flow operation may need to be considered, as
 discussed in Section 4.4.2. If flows are low but communica-
 tion is good, this could be indicating a very tight slab and
 native soil with a good aggregate bed under the slab. Again,
 the 50- to 90-watt fans could be a good choice, to handle any
 increase in the flows that will occur if leaks develop in the
 slab.

 Thus,  the simple flow measurement can possibly provide
 some insights to aid in interpreting the suction field results,
 and may help confirm the selection of the system fan.

  •   More extensive sub-slab flow measurement

 The more extensive sub-slab flow measurement involves op-
 eration of the diagnostic vacuum cleaner at a range of speeds,
 rather than  at only a single speed as in the "simple" case.
 Operation over a range of vacuum suctions and flows will
 enable determination of the effect of vacuum speed on nozzle
 flow.  This information  allows the  calculation of a
 suction-vs.-flow "performance curve" for the sub-slab mate-
 rial, which can then nominally be used to aid in selecting an
 ASD fan having the optimum performance characteristics
 (EPA88b,Fo90).

 For this testing  to be useful, the vacuum  cleaner must be
 operated  with the sub-slab depressurizations at the baseline
 test hole being carefully adjusted (as discussed for the quanti-
 tative suction field extension testing in Section 3.3.2), so that
 the vacuum cleaner will nominally reproduce the flows that a
 SSD system would  create. Otherwise, the vacuum cleaner
 suction-vs.-flow curve would be meaningless for system de-
 sign. (If an ASD fan mounted on a 4-in. pipe is being used to
 generate the suction, instead of a vacuum cleaner, representa-
 tive flows will almost automatically result, if the fan test stand
 is operated considering the likely suction losses in the piping
 of an actual system.)
                                                       98

-------
There is some question regarding how often such extensive
(multi-suction) sub-slab flow measurements will in fact be
useful, beyond the simple (one-suction) flow measurement
discussed previously. As discussed in Section 3.1.2, the selec-
tion of the fan for a given installation will most commonly be
made based upon experience in the geographical area, perhaps
with some guidance from the simple flow measurement. The
50- or 90-watt centrifugal tubular fans (or some appropriate
equivalent) will commonly be used whenever there is any
reasonable flow and communication. Very high-suction units,
discussed in Section 4.4.2, may be selected when the flows are
unusually low and the communication is unusually poor. It is
not clear how often the more extensive sub-slab flow  mea-
surements will be sufficiently accurate to permit meaningful
fine-tuning of fan selection, beyond the relatively gross  com-
parison just indicated. That is, it is not clear how often the
multi-suction flow results could permit more rigorous selec-
tion of the particular fan brand and model having the optimum
performance curve for a specific house.

Some mitigators have found the results from the more exten-
sive flow measurements to provide useful insights in aiding
ASD design. A common procedure for conducting such  mea-
surements is given below.

The  equipment and materials required for this measurement
are essentially the same as those listed in Section 3.3.2 for the
quantitative suction field extension measurements. However,
for the flow measurements, it is now required that a pitot tube
(or some other device) be available for measuring the nozzle
flows.

The vacuum cleaner is operated to generate a range of sub-slab
depressurizations in the baseline test hole, 8 to 12 in. from the
suction hole. The baseline test hole is the hole where depres-
surizations are being maintained at a level similar to that
which would be generated in the sub-slab pit under a SSD
suction pipe by an ASD fan. The suction at the baseline hole
must be varied through several values within the range that
the expected ASD fan would produce in the pit These values
would vary, depending on the particular fan envisioned.

For example, a 50-watt tubular fan might typically generate
pit suctions between 0.1 and 0.75 in. WG, depending on flow;
a 90-watt tubular fan might generate suctions between 0.25
and 2 in. WG. Thus, to help decide between a 50- and 90-watt
fan,  reasonable baseline suction values might be approxi-
mately 0.2, 0.75, 1, and 2 in. WG. The high-suction fans
might generate suctions between 0.1 and perhaps 40 in. WG,
depending on the specific fan and the flow rate. For these fans,
reasonable baseline suction values might be 2, 10, and 20 in.
WG. (If one  of these high-suction fans is not generating at
least a couple inches of water suction, the flows are probably
too high to warrant a fan of this type.)

At each of these baseline sub-slab depressurizations, the flow
in the vacuum nozzle is measured.

The results (baseline sub-slab depressurization. vs. flow in the
vacuum cleaner) are then plotted on a curve, with suction as
the ordinate and flow as the abscissa. This curve, which will
show flow increasing as suction increases, might be referred
to as a "sub-slab performance curve." The basic approach for
using these flow data to select a suitable fan is to compare this
sub-slab performance curve with alternative/an performance
curves (which will show flow decreasing as suction increases).
The fan that would be selected would be the one for which the
fan performance curve intersected the sub-slab performance
curve at a point about mid-range on the fan curve. The basis
for this selection is that that fan would be operating at a
comfortable point, and should be able to  handle increases or
decreases in flow that might result over time, e.g., as the
underlying soil dried  out or became moist  during different
seasons, or as small cracks developed in the slab.

In fact, there is some inaccuracy in directly comparing the
sub-slab  and fan performance curves.. The sub-slab perfor-
mance curve is based upon suction under the slab; the fan
performance curve is based upon suction at the fan. There are
two reasons why sub-slab depressurization at the baseline
hole does not equal suction at the fan: 1) suction losses as the
gas in the pit accelerates  up to SSD pipe velocity; and 2)
suction losses in the SSD piping, between the slab and the fan.
Accordingly, the sub-slab performance curve should be modi-
fied to account for these  factors, and the fan performance
curves then compared against the modified sub-slab perfor-
mance curve, if this interpretation is to be rigorous.

Using  standard equations for suction losses  at abrupt en-
trances (Ca60), it can be calculated that the suction losses due
to Item 1 above will usually be small at the velocities typical
in SSD pipes; assuming that  1,500 ft/min is  the highest pipe
velocity that can be tolerated due to flow noise in the piping,
the maximum suction loss due to pit gas acceleration would
be on the order of 0.05 in. WG, depending on pipe diameter.
As  a result, this first factor can usually be neglected.

Thus, only the suction losses due to piping  friction (Item 2
above) usually need to be considered. To address this, the
mitigator would have to estimate what the suction losses in
the envisioned system piping will be as a function of flow
rate, using the procedures discussed in  Section 4.6.1. The
sub-slab depressurization vs. vacuum flow performance curve
would then have to be corrected accordingly. For example,
assume that the vacuum cleaner flow testing showed a flow of
30  cfm at a baseline sub-slab suction of 0.75 in. WG, and a
flow of 40 cfm at  1.5 in. WG. And assume that the piping
suction loss  calculations suggested that piping losses would
be about 0.08 in. WG at 30 cfm and 0.14 in. WG at 40 cfm.
Under these conditions, the  modified sub-slab performance
curve would be plotted using a flow of 30 cfm at 0.75 + 0.08
= 0.83 in. WG, and a flow of 40 cfm at 1.5 + 0.14 = 1.64 in.
WG. That is to say, the fan would have to be able to maintain
a suction at the fan of 0.83 in. WG when the flow is 30 cfm, in
order to maintain a suction in the suction pit of 0.75 in. WG at
that flow.

In some cases, when the flows are low, the suction losses due
to pipe friction might be low enough to be neglected, just as
the losses due to pit acceleration. In those cases, the measured
sub-slab  performance curve could be used without modifica-
tion.
                                                        99

-------
The decision regarding whether or not the more extensive
sub-slab flow measurements are a cost-effective diagnostic
test rest with the individual mitigator, and will be based upon
the specific conditions that the mitigator encounters. How-
ever, it is expected that most mitigators will find this diagnos-
tic test unnecessary most of the time.

3.5.2 Well Water Radon Analysis

Radon released from the surrounding  soil and rock will dis-
solve to a certain degree in the underground aquifers from
which  wells  draw.  If a  house is served  by a private or
community well, and if the well water is not aerated before
use in the house, a significant fraction of the dissolved radon
will  be released into  the air when the water is used. The
release of the waterborne radon will be especially great when
the water is aerated during use (e.g., when sprayed through a
shower nozzle or in a dishwasher), and when the water is
heated. The waterborne radon will thus  contribute to the
airborne radon levels in the house.

Water that is drawn from reservoirs and rivers, and water that
has been treated by municipal water authorities will generally
have very low radon concentrations. Thus, only water that is
drawn directly from  wells (without any intermediate aeration
step) may be of concern.

Based  upon typical water usage rates  in  a  house, typical
natural ventilation rates, and typical house volumes, a rule of
thumb is that 10,000 pCi/L  of radon in  well water will
contribute approximately 1 pCi/L to the indoor air concentra-
tions, on the average over time, and on the average throughout
the house (Bru83). By  this rule of thumb, 40,000 pCi/L in the
water would  contribute an average airborne  level equal to
EPA's original  action  level of 4 pCi/L.  Of course,
water-induced airborne concentrations would be greatest at
the time that the water was being used, in the immediate
vicinity of where it was being used. While the exact contribu-
tion of the  well water to the airborne levels will vary from
house to house, because water usage rates, natural ventilation
rates, and other parameters will vary from the typical values
assumed, it has generally been found that the 10,000:1 rule of
thumb is reasonably good (Be84, Fi91).

Any radon that is being contributed to the indoor  air by the
well water will not be impacted by an ASD system, which can
address only that portion of the radon entering the house with
soil gas. In the large majority of cases  (although not always),
the soil gas contribution to indoor levels will be much greater
than that from well water. However, in geographical  areas
where elevated radon  concentrations  in the well  water are
common, radon from  water can sometimes complicate the
ability to achieve indoor levels below 4 pCi/L with  ASD
alone. In those cases, a water treatment step could be required
in addition to ASD. In cases where it is desired to reduce the
indoor concentration to levels well below 4 pCi/L, the radon
contribution from well  water will become increasingly impor-
tant.

Mitigators working in areas where elevated water concentra-
tions are sometimes encountered  will be  well advised to
confirm the water radon levels. This knowledge will permit
the mitigator to provide the best guidance to the homeowner
regarding the reductions that might be anticipated with the
ASD system. It will also save the mitigator from offering a
guarantee to achieve 4 pCi/L (or some lower level) with ASD
when in fact the water source might be sufficient to prevent
that level from being achieved without water treatment.

In some cases, the homeowner might have already obtained a
well water radon measurement through contacts with state or
local agencies. But where the homeowner has not done this,
the mitigator may be called upon to provide  guidance on how
the homeowner  can get the water measurement made, or
perhaps to arrange for the measurement himself.

The following discussion describes how to proceed in obtain-
ing an analysis  of the radon concentrations in water. As
emphasized  in the preceding discussion, such analyses are
necessary only if: a) the house is served by a well; and b)
experience in the area shows that well water radon concentra-
tions can sometimes be high; and c) the homeowner has not
already had an analysis completed.

 •  Screening using a gamma meter

Before incurring the cost  of sending a water sample  to a
laboratory for analysis, a simple, qualitative screening test can
be conducted using a gamma meter to assess whether water
radon concentrations might be high.

If radon levels in the water are sufficiently elevated, radon
decay products and lead-210 deposited in the plumbing sys-
tem will result in gamma radiation in the plumbing signifi-
cantly above background levels. A gamma reading might be
taken against the well water tank inside  the  house.  One
complication in using the water tank is that, if other radionu-
clides such as uranium or radium are present in the water,
these elements will also contribute to the gamma radiation at
the tank. It would be preferred to  make the gamma measure-
ments at some other location where a mass of water stands,
but where uranium and radium will not be deposited.  One
location that meets these criteria is a commode.

If the gamma levels measured at a commode or elsewhere in
the plumbing are significantly above the background levels
measured in the house remote from the plumbing, this  sug-
gests that well water radon concentrations may be elevated. In
this case, it would be advisable to have a water sample sent to
a laboratory for quantitative analysis.

 •  Analysis by a laboratory

The analysis of the water sample must be conducted by a
qualified laboratory. Thus, the role of the mitigator would be
to ensure that a good sample is properly drawn and provided
to the  lab. In some cases, a state agency may operate a
laboratory which can conduct the  analysis. Or, the state may
be able to indicate laboratories in the area  that  can conduct
such analyses, if the mitigator is not already  aware  of a
suitable lab. Thus, the logical first step  would be a contact
with the appropriate state agency  listed in Section 15 of this
document. If the state cannot conduct the analyses or identify
candidate laboratories, various laboratories offer analytical
                                                       100

-------
services by mail. These laboratories advertise in the trade
literature.

Whichever laboratory is selected for the analysis will provide
a vial in which to collect the sample. The lab will also provide
instructions for collecting the sample, which will  then be
returned to the lab. The person collecting the sample should
follow the instructions from the specific laboratory that he/she
is dealing with.

Depending upon the particular lab, the house water  sample
might be injected (in known quantity) directly by the sample
collector into a "cocktail" of organic compounds. The com-
pounds in the scintillation cocktail emit photons of energy
upon excitation by the alpha particles released by the radon,
thus permitting the alpha emissions to be counted in the lab.
Sometimes a known quantity of water sample is injected into
a vial containing mineral oil, which extracts the radon out of
the water and which is subsequently mixed with the cocktail
at the laboratory. Alternatively, the sample could be collected
by  filling  an empty sample vial, with the injection of the
sample into the cocktail being handled later at the lab.

The typical sampling procedures discussed below addresses
both sampling methods.

    -  Identify a tap in the house plumbing (indoors or out-
      doors) where a water sample can be drawn, prior to any
      carbon filter that may be present, and preferably prior
      to any water softener. This tap must be a cold water tap,
      so that the water sample will not have been heated (and
      thus potentially de-gassed) in the water heater.

    -  Before sampling,  allow the water to run from the tap
      for a period of time to flush the "old" water out of that
      branch of the plumbing. This step is particularly impor-
      tant at taps (such as outside spigots) which may not
      have been used in days or weeks;  some significant
      portion of the radon originally present in the old water
      may have decayed away.

    -  If the sample is being collected in an empty vial pro-
      vided by the lab (which will most commonly be the
      case):

      — Place a rubber adapter over the selected tap, con-
         nected to a length of tubing which directs the water
         to  the sampling vial.  The adapter prevents any
         aeration of the sample  as it leaves the tap, even if
         the tap has an aerator. (If there is any doubt, remove
         the aerator.)

      — Allow the water to  flow, via the length of tubing,
         into the open  sample vial. The end of the tubing
         should be placed well below the water surface in the
         vial, to avoid aeration. The water is allowed to flow,
         overflowing at the mouth of the vial, until  the vial
         has been flushed a number of times.

      - The vial is then immediately capped, using care to
         avoid any air bubble in the vial.
      —  Alternatively, instead of the rubber adapter, fill a
         bowl or bucket with water continuously running
         from the tap, with the tap outlet (including any
         aerator) below the water  surface in the  bucket.
         Water will be continuously overflowing the top of
         the bucket. Submerge the capped vial in the bucket;
         remove the cap beneath the tap, allowing the vial to
         fill; then re-cap the vial under water, making sure
         that no air is trapped.

      —  The vial is then immediately delivered or mailed to
         the lab for analysis.

   -  If the sample is to be injected directly into the cocktail
      (or into mineral oil) by the sample collector:

      —  Where a rubber adapter with tubing has been placed
         over the tap, allow the water to flow, via the length
         of tubing, into a suitable small open container. The
         end of the tubing is placed well below the  water
         surface in the container, to avoid aeration where the
         water leaves the tubing. The  water is allowed to
         flow until the container has been flushed a number
         of times.

      —  Using a syringe, a known quantity of water  in the
         container (as specified by the lab) is withdrawn, and
         is injected into the capped vial containing the cock-
         tail. The vial is then vigorously shaken, to mix the
         water sample and the  cocktail. The sample with-
         drawn from the container using the syringe should
         be withdrawn well below the surface of the water in
         the container.

      —  Alternatively, instead of filling a container, bend
         the tubing upward and turn the tap on  slowly, so
         that the water slowly overflows the upward-directed
         end of the tubing. Insert the syringe needle down
         into the open end of the tubing, and withdraw the
         water sample from well below the surface.  If the
         water sample is being drawn from an outside spigot,
         a short length of hose  screwed onto  the spigot
         would serve this same purpose.

      —  The vial is then immediately delivered or mailed to
         the lab for analysis.

3.5.3  Measurements to  Determine
Significance of Building Materials as a
Radon Source
On infrequent occasions, materials used in the construction of
the building may be important contributors to indoor radon
levels. This will usually be the case only where stone with an
unusually high natural radium content has been used as aggre-
gate in the concrete,  or perhaps for  structures such as fire-
places. It can also occur if, e.g., uranium mill tailings have
been used in the aggregate or as fill around the house.

In most areas of the country, mitigators will rarely see  a case
where building materials are making any significant contribu-
tion to  indoor radon. In those areas,  pre-mitigation tests
                                                        101

-------
addressing building materials will not be warranted. But in
those areas  where  building materials are occasionally  an
important source, a mitigator may be well advised to make a
gamma survey (and/or perhaps a flux measurement) wherever
conditions look suspicious, before guaranteeing that 4 pCi/L
will be achieved with an ASD system alone.

In worst-case situations, removal  and replacement of the
elevated building materials or coating them with some appro-
priate radon barrier may be required in order to reduce the
house below 4 pCi/L. It is doubtful whether such steps would
be warranted in cases other than, e.g., uranium mill tailing
contamination; these steps would thus be beyond the scope of
the typical mitigator. And the treatment of this building mate-
rial source would be required for the purpose of protecting the
occupants from gamma radiation, as well as from radon.

Two approaches can be considered for estimating the relative
importance of building materials. The simplest approach is
the use of a meter to determine gamma radiation at various
locations around the house.  Materials having a high concen-
tration of radium (and also, necessarily, of  other radionu-
clides) will  also have high gamma emissions. Thus,  high
gamma radiation is evidence of possibly high potential for
radon release. The second approach for estimating  building
materials effects is the "flux  measurement," where a container
is sealed over the surface of, e.g., the slab, and the increase in
radon concentration over time is measured.

Of the two approaches, the gamma survey is the easiest to
conduct, and permits numerous, rapid measurements around
the house to locate hot spots. Thus, mitigators in areas where
building materials may be a common problem should prob-
ably obtain a gamma meter to permit such surveys, as dis-
cussed later. The flux measurement may provide a more
quantitative estimate of the potential for radon emission from
the building surface. However, the flux  measurement takes
more time to conduct, and the number of test locations is thus
limited. The flux measurement approach is probably  most
applicable in cases where the results from a previous gamma
survey indicate that the building surface is "hot," and a more
quantitative  estimate is desired of how much radon may in
fact be being released from the materials.

 •  Gamma survey

Gamma radiation is naturally present in  the  environment
around any house, as a result of naturally-occurring radionu-
clidcs in the surrounding soil and rock, and as a result of
cosmic radiation. Natural outdoor ground-level gamma read-
ings vary around the country, but a typical range is 5 to 20
miero-Roentgens per hour (|xR/hr).

Commonly, if there are not elevated levels of radionuclides in
the concrete aggregate or in other stone-based building mate-
rials, the  gamma radiation levels inside  a  house will  be
slightly lower than those outdoors, because  the house shell is
providing some shielding from the natural radiation. But if
gamma levels at specific locations indoors are higher than the
levels outdoors, this would suggest that the building materials
in the immediate vicinity of the high indoor readings have a
higher radionuclide content than does the native soil and rock
around the house.

The gamma survey thus consists of making gamma measure-
ments at several locations indoors, and comparing these read-
ings against those made at several locations outdoors.

A hand-held scintillometer that measures ionizing radiation in
units of (oR/hr should be used. Such a "micro-R" meter is the
most convenient and commonly-available instrument to pro-
vide the results required here.

The micro-R meter must be calibrated (i.e., the number of
meter counts/sec per |aR/hr must be determined). Commonly,
these instruments are calibrated in a laboratory using a known
cesium-137 source. Laboratory calibrations using aradium-226
source are not recommended; meters calibrated using radium
appear to give readings in the field that are high by a factor of
about two,  due to differences in the gamma energy of  this
radionuclide relative to the spectrum of gamma-ray energies
normally found in nature. A more scientifically rigorous cali-
bration is possible, by calibrating the meter against a pressur-
ized ion chamber. Alternatively, specially-prepared radioac-
tive test pads could be used which better represent the normal
natural spectrum of gamma-ray energies. These more rigorous
calibrations will generally not be conveniently available to a
mitigator, and they are not really warranted in any event. The
improved accuracy that these better calibrations would pro-
vide is not necessary, given the comparative manner in which
the results are used (indoor vs. outdoor levels).

Readings with the micro-R meter should be made at multiple
indoor locations around the lowest level of the house, making
sure that readings are made near all major structures contain-
ing earth-based building materials. These structures would
include the slab, the foundation walls, and fireplace  struc-
tures. At each test location, a reading should be taken: a) with
the meter flush against the building surface;  and b) with the
meter 3  ft away from the surface. The gamma radiation
emitted by  a surface will drop significantly as  the meter is
moved away; repeating the measurements  flush with  the
surface and 3 ft away will help confirm the extent to which the
observed radiation is in fact coming from the surface rather
than from other sources.

Readings with the micro-R meter should also be made at
several locations outdoors, around the perimeter of the house.
Again, at each location, readings should be taken flush against
the ground  and also at a height of 3 ft.

The indoor vs. outdoor readings are then compared. In gen-
eral, the  indoor readings should be about the  same as  the
outdoor readings, or perhaps one or  two |oR/hr lower. If
indoor levels  are slightly higher (by, say, one or two nR/hr),
this result would suggest that the material used in construction
is slightly more elevated than the native soil and rock, but the
building materials are not contributing significantly to indoor
radon levels. However, if any significant number of the indoor
readings are significantly greater than the outdoor background
(by,  say, 10  to  20 |oR/hr or more), suggesting  that some
significant  amount of the building material is distinctly el-
evated relative to background, then the mitigator should be
                                                       102

-------
alerted that building materials might be contributors to the
indoor radon levels. One mitigator uses a guideline that stone
foundation walls  can  be expected  to be important radon
sources when gamma readings flush against the wall are twice
the background level (Bro92).

If it thus appears that building materials might be a source, the
mitigator has two options.

 1. If the mitigator's experience enables a reasonable judge-
    ment from the micro-R readings regarding what the build-
    ing material contribution may be, i.e., if it is reasonably
    clear that building materials are not the sole or predomi-
    nant source, and that some soil gas treatment step is thus
    still  necessary, then one option is  for the mitigator  to
    advise the homeowner of this  situation, warning that
    some residual radon levels will remain after installation
    of the ASD system, due to the building materials.

 2. If the building material contribution may be a major
    component of the indoor radon levels, it may be advisable
    to conduct a flux  measurement, to  better quantify how
    much radon is in fact being released from the building
    materials.

It is emphasized that a structure must be of relatively signifi-
cant size, and/or have distinctly elevated gamma readings,  to
pose a significant threat  in terms or radon emissions. An
isolated stone fireplace with slightly elevated gamma readings
will probably not be a significant source. Rock collections,  or
collections of radium-dial clocks, will almost never be signifi-
cant sources, despite elevated localized gamma readings.

In addition to revealing the possible contributions of building
materials to indoor radon, gamma surveys can also provide
other information  that can be helpful in system selection and
design. As mentioned in Section 3.5.2, a high gamma reading
at a commode inside the house (resulting from accumulated
radon  decay products) will suggest whether there may be
elevated radon levels in the well water. In special cases where
high-radium strata exist in the underlying  soil,  hot  spots
identified by the gamma meter around the slab, may suggest
where the strata may be surfacing  around the foundation,
information which may or may not be useful in ASD design.

 «  Surface flux measurements

Flux measurements attempt to determine the actual rate  at
which radon is being released  from  a  solid surface,  in
picoCuries per unit area of the surface per unit time. Since the
actual release of radon is being measured, rather than  using
gamma radiation as a surrogate, flux measurements would be
expected to provide a more rigorous estimate than would a
gamma survey of the actual radon contribution from a build-
ing surface.

As indicated previously, flux measurements are not suitable
as a routine pre-mitigation test for potential building material
contributions to indoor radon. If a mitigator is working in  an
area where potential building  material contributions  are a
common concern, the gamma survey  approach  should be
utilized. Flux measurements will be cost-effective under only
certain circumstances, discussed earlier.

A procedure for flux measurements has been discussed else-
where  (EPA88b); due  to  the  limited applicability of this
measurement, the full procedure will not be repeated here. In
summary, a container of known volume having one open end
of known area is sealed, open end down, onto the surface from
which the flux is to be determined. A radon grab sample is
withdrawn from the sealed container at the beginning of the
test, and than again after some period, typically 30 minutes to
an hour. From the measured increase in radon concentration
inside the container over this time, a radon flux (e.g., in units
of pCi/ft2/hr) can be calculated.

From typical uncracked concrete surfaces, the radon emana-
tion (due to naturally occurring radium in the aggregate, about
1 pCi per gram of concrete) is roughly 10 to 40 pQ/ft?/hr.
Measured levels significantly above that range would suggest
that the concrete is  an  above-average contributor to indoor
radon levels.

It is emphasized that the flux container must be sealed over a
solid, uncracked section of the building surface. The objective
of the test is to determine the emanation of radon from radium
in the building material (or the diffusion of radon through the
solid, unbroken material). If the flux container is sealed over
an opening, such as  a crack in the slab or such as the porous
surface of a block foundation wall, then the radon that appears
in the  container will be the combined result  of multiple
mechanisms:  1)  emanation from  the material (or diffusion
through the solid material); and 2) diffusive movement (or
convective flow) through the opening.

Measurement of the diff usive/convective flux of radon through
openings has sometimes been utilized (often in research set-
tings) in an effort to understand which entry routes may be the
most important among alternative candidates. Such a use of
flux measurements, over openings,  is  not discussed  here.
From a mitigator's standpoint, this information is either  un-
necessary as a pre-mitigation tool for ASD design, or can be
estimated from more direct and simpler measurements, such
as direct grab sampling/sniffing beneath the slab or inside
block walls. In any event, the thrust of the flux measurements
being discussed  in this section is to determine the possible
contribution of building materials, and  location  of the flux
container over an opening would invalidate the test for that
purpose.

3.5.4 Pressure Differential
Measurements Across the House  Shell

Pressure differential measurements across the house shell can
be made at two locations:

   -  Above grade, across the walls between the house inte-
      rior and the outdoors.

   -  Below grade, across the slab (i.e., between the  house
      and the sub-slab region). In this context, the sub-slab
      pressure measurements would be made without  any
      suction being drawn on the sub-slab region by a vacuum
                                                       103

-------
      cleaner or fan,  for the purpose of determining the
      natural depressurization of the house relative to the
      sub-slab.

Pressure differential measurements across the shell can be
considered for at least two different purposes, in connection
with ASD systems:

   -  To permit more rigorous interpretation of the sub-slab
      suction field extension results when designing an ASD
      system. For this purpose, below-grade measurements
      (across the slab) would be the most meaningful; they
      would also be easily conducted in conjunction with the
      suction field extension testing.

   -  To indicate whether the house may be prone to com-
      bustion appliance backdrafting, which could be exacer-
      bated by an ASD system. For this purpose, above-grade
      measurements would be appropriate.

Pressure differential diagnostics could also be considered in
conjunction  with blower door operation, in the  selection/
design of basement pressurization systems. This testing could
suggest how much supply flow is required through the blower
door (and hence how large a system fan would be required) in
order to adequately pressurize  a basement  relative to the
sub-slab. An analogous measurement could aid in the design
of crawl-space depressurization  systems, although in that
case, the question would be how large a fan would be needed
to depressurize the crawl space relative to the living area.

Uses of pressure  differential measurements to aid in
interpreting suction field extension test results. There
are two ways in which pressure measurements across the shell
might be used to aid in interpreting sub-slab suction field
extension test results.

   -  Sub-slab suction field extension measurements some-
      times have to be made when the weather is not very
      cold,  or without all exhaust appliances  operating. In
      such cases, pressure measurements across  the shell,
      conducted in parallel with  the sub-slab suction field
      testing, would indicate how the driving force existing
      during the diagnostics compares against the estimated
      maximum driving force that might be estimated to exist
      during cold weather with the exhaust appliances oper-
      ating.  The pressure differential measurements might
      show that the existing driving force during the diagnos-
      tics was much less than this estimated maximum (per-
      haps 0.025 to 0.035 in. WG, according to the conserva-
      tive rule of thumb discussed in Sections 2.3.1b and
      2.3.1e). In that case, the pressure measurements would
      reveal what sub-slab depressurizations the diagnostic
      vacuum cleaner (or fan test stand) would have to estab-
      lish during the diagnostics in order to compensate for
      the higher driving forces that would be expected later.

      For this purpose, the preferred pressure differential to
      measure is that across the slab, between the house and
      the sub-slab region (with the sub-slab vacuum off); that
      is the differential against which the ASD system will
      have to compete. The pressure differential between
      indoors and outdoors above grade will normally tend to
      be slightly greater than that across the slab, and could
      thus tend to give a slightly elevated indication of what
      the actual driving force is. Not only is the differential
      across the slab  the preferred measurement for  this
      purpose, but it is easy to obtain if holes have already
      been drilled through the slab in conjunction with suc-
      tion field extension tests.

   -  Pressure differential measurements across the shell can
      be made under a variety of conditions with different
      exhaust appliances operating, or perhaps at different
      weather conditions, in an  effort to identify the most
      challenging house depressurization driving forces in a
      particular house. The mitigator might  then elect to
      ensure that any suction field extension  tests are con-
      ducted under those most challenging conditions. Or,
      when interpreting the suction field extension results to
      ensure the ability of the system to handle the worst-case
      driving  force, the mitigator can then use  the  actual
      expected maximum depressurization for that house,
      rather than the rule of thumb values for thermal and
      appliance effects (0.025 to 0.035 in. WG). Again, the
      preferred pressure differential to measure would be that
      across the slab, with the sub-slab vacuum off.

In practice,  whenever a mitigator plans to conduct sub-slab
suction field extension diagnostics anyway, it makes sense to
include pressure measurements across the slab with the vacuum
off. These results would then be used to aid in data interpreta-
tion, as discussed in the preceding paragraphs. Such measure-
ments with the vacuum off are already included in the proto-
cols described in Sections 3.3.1  and 3.3.2 for suction field
extension diagnostics. On this basis, pressure differential mea-
surements across the house shell (slab) would be conducted
any time suction field extension measurements are felt to be
warranted, i.e., primarily when sub-slab communication is
unknown or poor.

Where suction field extension diagnostics are conducted, they
should be made with the exhaust appliances operating, simu-
lating worst-case conditions as closely as possible. If worst-case
house depressurizations cannot be fully simulated, then the
slab pressure measurements with the vacuum off can be used
to make any necessary corrections in the suction field results
based upon the rules of thumb. For example, if the diagnostics
are not performed during mild weather—at a time when the
pressure differential across the slab was about zero with the
vacuum off—it would be assumed that the sub-slab depressur-
izations measured with the diagnostic vacuum cleaner operat-
ing should be at least about 0.015 in. WG, to account for the
increased driving force  expected from thermal  effects during
cold weather.  (Alternatively,  if the suction field  extension
diagnostics  are performed during mild weather,  the  house
could be deliberately depressurized using a blower door dur-
ing the testing to simulate cold-weather depressurizations;
this is not commonly done.)

Most mitigators never  make  pre-mitigation indoor-outdoor
pressure  differential measurements above grade  as part of
sub-slab diagnostics.
                                                        104

-------
Uses of pressure differential measurements to assess
the risk of backdrafting. One potential concern with ASD
systems is whether the air drawn out of the house by the
system may be sufficient to cause backdrafting of combustion
appliances, possibly exposing the occupants to carbon mon-
oxide. Backdrafting as a result of ASD operation is most
likely to occur when: a) the house is particularly tight, as in
very cold  climates or in other climates where the house has
been intensively tightened for energy efficiency; b) the flue is
partially blocked or improperly designed, reducing the draft;
c) other major depressurizing appliances are also operating; or
d) the ASD  system is exhausting  a  significant amount of
house air,  as with BWD systems. ASD systems will generally
not cause backdrafting except in cases where the house draft
was marginal to begin with.  Many mitigators will encounter
ASD-induced backdrafting only infrequently, except in very
cold climates (Wi91,  Ang92, Fit92). However, the conse-
quences are of backdrafting are so severe that mitigators must
always be alert to this threat.

There are several options for testing for backdrafting. Because
mitigators often conduct backdrafting tests as a/jarf-mitigation
diagnostic, at which time they are required by EPA's stan-
dards (EPA91b), these options are discussed in Section 11.5
of this  document However, since one of  the test options
involves pressure measurements across the house shell above
grade, and since a mitigator might choose to perform  this
testing prior to mitigation, to assess whether selection of ASD
for a given house might potentially contribute to backdrafting
problems, this issue is addressed briefly here.

Measurement of  the pressure differential across the shell
above grade as an indicator of possible backdrafting assumes
that: a) the draft in the flues of furnaces, water heaters, etc.,
can be expected  to be a predictable  value relative to the
outdoor pressure; and b) backdrafting should not occur as
long as the depressurization of the house (relative to outdoors)
is less than that value. The draft in a well-connected, unob-
structed flue  during cold  weather may typically be about
0.028 in. WG (Br92). Thus, for example, two draft protocols
suggest that,  to be conservative, house depressurization by
0.020 in.  WG or more, as measured across the shell above
grade, should be viewed as a threat that backdrafting may
occur in traditional gas-fired furnaces (CMHC88, TEC92).

However, in fact, the draft can sometimes be well below 0.028
in. WG (sometimes falling to about 0.01 in. WG or less) when
the weather is mild, when the flue is obstructed, or when the
flue is not properly connected to the appliance (Fit92, Ne92).
Thus, it will be wise to supplement any pressure differential
measurements across the shell with a more positive determi-
nation of whether combustion  product  spillage is in  fact
occurring  (e.g.,  using chemical smoke tracer around draft
hoods). See Section 11.5.

Procedure for pressure  differential measurements
across the house shell. The procedure for making pressure
differential measurements across the house  shell  at the  slab
has been discussed in connection with sub-slab suction field
extension  testing, in Sections 3.3.1 and 3.3.2.
The  procedures for making pressure  differential measure-
ments across the house shell above grade have been described
in other documents (Fo90). Because of the limited applicabil-
ity of these measurements for ASD design, only a summary of
the procedure will be presented here.

The  equipment required for the measurements is a digital
micromanometer, sensitive to + 0.001 in.  WG, the same
device required for the sub-slab suction field extension mea-
surements described in Section 3.3. Also required are one or
two lengths of flexible (but not collapsible) tubing that can be
connected to the micromanometer ports.

One length of tubing must be long enough  to run from the
reference port of the manometer to the outdoors, through a
crack in a door or window that can be closed over the tubing
without pinching it. The micromanometer itself (or the end of
the sample tube connected to its second port) would be at the
indoor location where house depressurization (or pressuriza-
tion) is to be measured. This indoor location will depend upon
the specific objectives of the testing. For  example,  if the
objective is to assess the possible risk of backdrafting, the
indoor location will be near where the combustion appliances
are located.

It is usually difficult to obtain a reliable measurement of the
pressure differential across the house shell above grade when
any wind is blowing. Winds will tend to pressurize the upwind
side of the house, and depressurize the downwind side. These
pressure effects will vary  as the wind velocity varies. In
addition, wind blowing directly into  the open end of the
sample tube outdoors will create a dynamic effect on the
pressure readings obtained by the micromanometer,  a dy-
namic  effect that will fluctuate as the wind varies.

Several approaches have been suggested in an effort to reduce
the fluctuations in the pressure measurements created by the
dynamic effects of winds. In concept, one approach would be
to ensure that the end of the outdoor sample tube is always
oriented perpendicular to the wind direction, so that the wind
is always blowing perpendicular to the  opening and the mea-
sured pressure is the static pressure. This is impractical, since
the wind direction is not steady and is  difficult to monitor at
the precise location of the sample tube. A more practical
approach is to place some diffuser material over the open end
of the tube,  to reduce flow-induced  effects. Fritted glass
(Fo90), cotton (Fo90), and open-cell foams such as used in air
diffusers for aquariums (Tu92), have been suggested for this
purpose. Yet another approach would be to mount the end of
the tubing between two parallel plates (Br92), or to mount it in
a section of pipe perpendicular to the tube and open at both
ends. [One mitigator uses a soda can for this purpose (Bro92)].
This approach would ensure that air flow is always perpen-
dicular to the opening in the tube, and hence that it is the static
pressure being measured.

In addition to the concern about avoiding wind-induced dy-
namic  pressure effects so that a static  pressure can  be mea-
sured,  there is also concern about the  variation in the static
pressure. The static pressure on a given side of the house will
change as the wind velocity  varies. In addition, the static
pressure will be different on different  sides  of the house, so
                                                        105

-------
that the pressure differential result will vary depending upon
where the outdoor measurement is made.

In selecting the measurement location, the effects of the house
on the outdoor static pressure can be avoided by extending the
outdoor sample  tube a distance  away from the  house. As
another option,  one  could make individual measurements
across the shell on all four sides of the house, and then average
these in an effort to "average out" the effect of location around
the house. Some investigators have extended sample tubing to
all  four sides of the house,  manifolding all of the tubes
together to provide a single integrated pressure reading; this
approach will not normally be warranted during commercial
diagnostic testing; in addition,  some investigators suggest that
this approach will not in fact provide an effective  integration
(Br92).

The house conditions during the measurements are selected
based upon the specific  test objectives. These  conditions
include: exterior doors and windows open vs. closed; interior
doors open vs. closed; and exhaust appliances on vs.  off.
When the objective is to  assess the driving force for radon
entry: the exterior doors and  windows are normally closed;
interior doors  are often open, but tests are sometimes also
repeated with  the doors closed to assess depressurization in
particular rooms that may often be closed off by the occu-
pants; and exhaust appliances are often on, but tests may be
repeated with  the appliances  off. When the objective is to
assess the threat of backdrafting (e.g., based on the procedures
in References CMHC88 and  Ne92): the exterior doors  and
windows are normally closed; the interior doors are often
selectively closed in a pattern to maximize depressurization in
rooms having combustion appliances; and exhaust appliances
are normally on, again with the intent of creating a worst-case
situation.

3.5.5 Blower Door Measurements

A blower door is a calibrated fan which can be mounted in a
doorway of a house with the remainder of the doorway sealed,
blowing out of or into the house. The fan can be adjusted to
operate at different flow rates (up to about 3,000 cfm) to
maintain alternative selected degrees of depressurization or
pressurization within  the house, typically in the range of 0.06
to 0.24 in. WG. The air flow rates through the fan at these
different depressurizations are measured.

From these results, the "effective leakage area" through the
house shell (as defined by a Lawrence Berkeley Laboratory
infiltration model) or the "equivalent leakage area"  (as de-
fined by the National Research Council of Canada) can be
calculated. The effective leakage area is the nominal area of
the openings through the  shell that would have  to exist in
order to explain the blower door flows that would be predicted
when the measured results are extrapolated down to 0.016 in.
WG |f these openings were combined into a single bell-mouthed
nozzle. The equivalent leakage area is the nominal area of
shell  openings when the  results are extrapolated down to
0.040 in. WG, if the openings were combined into a single
round, sharp-edged orifice.
In concept, this approach is generally analogous to the stan-
dard orifice flow calculation. If an orifice of known diameter
is mounted in a pipe, and if the pressure drop created by gas or
liquid flow across this orifice is measured, the flow velocity of
the gas or liquid in the pipe can be calculated. In the case of
the blower door, the flow rate  is measured for a known
pressure drop,  allowing the orifice size (i.e., the effective or
equivalent leakage areas) to be calculated.

It is emphasized that the blower door calculations assume that
all of the openings to be combined into a single bell-mouthed
nozzle or into a single sharp-edged orifice, whereas the actual
openings are in fact numerous small gaps  of varying configu-
rations with much different flow characteristics than a nozzle
or a round orifice. As a result,  the calculated effective or
equivalent leakage areas are not the actual combined areas of
all of the openings. Nevertheless, the effective or equivalent
leakage areas can still be useful numbers which allow one
house to be compared with others in terms of leakiness.

Depending upon the manner  in  which the blower door is
operated, it can be used to determine the effective leakage
area for the entire house, or for an individual  story of the
house.

Commonly, the effective leakage  area is used, in conjunction
with a mathematical model and various assumptions, to esti-
mate the average natural ventilation rate of a house. It should
be understood that any such calculation of the ventilation rate
from blower door data is only an estimate.  The blower door
determines effective leakage area, not ventilation rate. The
actual ventilation rate at any point in time will depend heavily
on weather conditions, house characteristics, nature and loca-
tion of the openings, and homeowner activities;  to the extent
that these features are incorporated into  the models used to
estimate ventilation rate from blower door data, they can be
incorporated in only a gross, averaged manner.

As a rough rule of thumb, the average natural ventilation rate
of a house (expressed in cfm) is estimated to be approximately
one-twentieth of the blower door flow rate (in cfm) when the
blower door is depressurizing (or pressurizing) the house by
0.20 in. WG.  While this simple correlation gives  a rough
indication of average natural ventilation rate,  possibly ad-
equate for many field mitigation purposes, it can be in error by
a factor of 2 or more in some cases.

The effective leakage area of a house, or the ventilation rate
calculated from blower door results, will rarely, if ever, be
information that can aid in ASD design.  Thus,  mitigators
planning to install an ASD system in a house will essentially
never conduct  a blower door diagnostic test for ASD design
purposes. Blower door testing will  be warranted primarily
when the features of a particular house require a mitigator to
assess specific alternative mitigation options other than ASD.

Blower  door tests are most likely to be considered when
basement pressurizaiion, crawl-space depressurization, or house
ventilation  are among the  mitigation  techniques being as-
sessed. The blower door result would indicate the tightness of
the basement (or crawl space), providing  a direct measure of
how large a fan would be required to maintain the desired
                                                       106

-------
pressurization of the basement (or depressurization of the
crawl space). Or, if house ventilation were being considered,
the blower door would provide an estimate of the average
natural ventilation rate of the house, permitting a calculation
of how many cfm of fresh ah" would be required to increase
the ventilation rate adequately to get the desired radon reduc-
tion.
                                     '/*
Procedures for operating blower doors and for converting the
results to effective leakage areas and estimated natural venti-
lation rates have been described elsewhere (SheSO, ASTM87,
TEC87, EPA88b,  Tu88b). The standard procedure is pre-
sented in ASTM E 779-87, "Standard Test Method for Deter-
mining Air Leakage Rate by Fan Pressurization" (ASTM87).
Because blower door testing in not usually warranted for ASD
design, no discnssion of blower door procedures will be
presented here.

3.5.6   Tracer Gas  Testing

Three types of tracer gases have been used for various pur-
poses associated with radon mitigation:  halogenated hydro-
carbons,  commonly used as refrigerants; PFTs; and SF6. Of
these, only halogenated hydrocarbons would be convenient
for practical use by a commercial mitigator for pre-mitigation
testing in connection with ASD systems, and even they would
likely have limited value. PFTs might be used infrequently for
pre-mitigation ventilation rate measurements, in cases where
a house ventilation technique is being considered for mitiga-
tion. Both halogenated hydrocarbons and PFTs might also
infrequently find some use as a post-mitigation diagnostic tool
for trouble-shooting ASD systems that are not performing as
expected. The use of SF6 (usually to determine house ventila-
tion rate) requires such elaborate equipment, and is so labor
intensive, that SF6 would not be considered outside of a
research setting.

 •  Halogenated hydrocarbons

The advantage of halogenated hydrocarbons, marketed under
such trade names as Freon® and Genetron®, is that both the
gases and the detectors are relatively  inexpensive, simple to
use, and readily available through the refrigeration and air
conditioning industry.

The primary disadvantage of  chlorinated refrigerants is that
they are destructive to the Earth's stratospheric ozone layer.
For this reason, the original chlorofluorocarbon (CFC) refrig-
erants (such as  R-12)  and the so-called "transition"
hydrochlorofluorocarbon (HCFC) refrigerants (such as R-22),
are being largely phased out of production over a 10- to
25-year period; in accordance with the Clean Air Act Amend-
ments of 1990 (Public Law 101-549, Title VI). Venting of
these substances by the refrigeration industry is now banned
by those Amendments. While the Amendments do not explic-
itly ban the use of small  quantities of CFCs and HCFCs as
tracer gases for radon mitigation diagnostics, EPA recom-
mends that these gases never be used for pitigation diagnos-
tics.

Only the chlorine-free advanced refrigerants—hydrofluoro-
carbons (HFCs), such as  R-134a—are not damaging to the
ozone layer. The venting of these gases by the refrigeration
industry is not currently banned. Therefore, to the extent that a
mitigator might wish to conduct this type of diagnostic test-
ing, HFCs should be used.

Refrigerant gases can be obtained in conveniently sized com-
pressed cylinders. Durable hand-held detectors, designed to
pinpoint the source of refrigerant leaks, can be obtained for
about $200 to $300. These  detectors,  which are usually
battery-powered and portable, are sensitive to refrigerant con-
centrations as low as 5 to 50 ppm. Many of these detectors
give an audible signal when trace levels of the refrigerant are
detected, with the signal varying in intensity depending upon
the concentration present in the air. The output is thus qualita-
tive, which is usually sufficient for radon diagnostic purposes.

It must be ensured that any  detector that is purchased is
capable of detecting the specific refrigerant that is going to be
used as the tracer. The older, CFC-based detectors, which
function by detecting chlorine, will not detect HFCs.

Most mitigators will find testing using halogenated hydrocar-
bons as a tracer gas  to be unnecessary as a pre-mitigation
diagnostic in most cases. Especially where sub-slab commu-
nication is good, sufficient information for ASD design will
usually be obtained from prior experience, from the visual
survey, and from suction field extension testing (if needed).
Testing using HFC tracers will not be needed.

However, there might occasionally be situations where inno-
vative  utilization of HFC tracer testing prior to mitigation
might be cost-effective. Among the examples, indicated in
Section 3.1, are:

   -  Injection of the tracer  beneath  an adjoining slab on
      grade, and detection of the tracer in the exhaust from a
      vacuum cleaner drawing suction beneath the adjoining
      basement slab, in an effort to assess whether the suc-
      tion field from a basement SSD system would extend
      beneath the adjoining slab on grade.

   -  Injection of the tracer  into one drain tile  entering a
      sump, and detection of the tracer in the exhaust from a
      vacuum cleaner drawing suction on a second drain tile
      entering that sump, in an effort to determine whether
      the tiles form a contiguous loop.

Halogenated hydrocarbons have also occasionally been used
for post-mitigation diagnostic tests, as discussed in Section
11. For example, HFCs injected into ASD exhaust piping can
be used to detect where exhaust gases may be re-entraining
into the house.

 •  Perfluorocarbon tracer gases (PFTs)

In the most common application of PFTs, one or more
2-in.-long vials which emit a selected PFT gas are deployed in
selected locations around the house. These emitters passively
release any one of several alternative (slightly different) per-
fluorocarbon compounds at a steady rate, dependent on tem-
perature. After the PFT concentrations in the house have been
given a chance to reach steady state, PFT detectors are de-
                                                       107

-------
ployed at selected locations. These detectors adsorb any of the
several PFT gases that are being used.

The emitters and the detectors are purchased from a desig-
nated laboratory. After the detectors have been exposed for a
selected period of time (commonly two days to one year), the
detector is returned to the laboratory for analysis. Based upon
information supplied by the tester regarding the test condi-
tions, the laboratory report calculates the ventilation rate
(where the emitter and detector are placed in the same zone)
or the flow rate of air between zones (when the emitter and
detector are placed in different zones).

Because multiple PFT compounds are available, it is possible
to emit different compounds simultaneously in different zones
in the house (e.g., with one  compound emitted in the base-
ment, one in the living area, and one in an adjoining crawl
space). Detectors would be deployed in all zones. From the
results, it would be possible to calculate not only the total
ventilation rate of the house, but also the rate of air movement
between the different zones within the house.

It is important to recognize  that the ventilation rates deter-
mined by PFTs will be integrated (i.e., averaged) over the
entire period that the detector was deployed. Ventilation rates
in houses can vary significantly over time, e.g.,  as  wind
patterns change, as windows are opened, etc. The PFT result
will average all of these effects. It is also important to recog-
nize that, since ventilation rates and interzonal flow rates vary
with environmental and house operating conditions, measure-
ments made under one set of conditions will not address what
the rates would be under different conditions.

To the extent that PFTs might have a practical use in commer-
cial pre-mitigation diagnostics, it would be to determine aver-
age house  ventilation rates  as  part of designing  a house
ventilation system. PFTs are relatively easy to use, but they do
take time to deploy and retrieve, they are relatively costly, and
there is commonly a delay in  obtaining analytical results back
from  the laboratory. Thus, although PFTs provide  a  more
rigorous measure of house ventilation rates than do the esti-
mates obtained from blower door  testing, many mitigators
needing to make ventilation rate measurements will probably
find it more convenient to use the blower door.

In research projects, PFTs have sometimes been used for
post-mitigation measurements in ASD systems. For example,
some investigators have released PFTs inside the house and
measured the amounts appearing in the ASD exhaust gas.
These tests were conducted to estimate the amount of indoor
air being exhausted by the ASD system, and hence the system
heating and cooling penalty (Bo91, Fi91). It is doubtful that a
mitigator would often have  occasion to conduct such tests
commercially. In one project (Fi91), aPFT emitter was placed
inside ASD exhaust piping to measure the amount of exhaust
re-entrained into the house. However, this approach is experi-
mental, and would likely not be practical for trouble-shooting
a commercial installation.
The procedures for utilizing PFTs for ventilation measure-
ments are discussed elsewhere (Di86). Because PFTs will
rarely be used in conjunction with commercial ASD installa-
tions, the procedures are not discussed in further detail here.

 •  Sulfur hexafluoride tracer gas (SFJ

With proper apparatus, SF6 tracer gas can be used to monitor
house ventilation patterns on a continuous basis, thus provid-
ing more comprehensive ventilation information than is pos-
sible with the time-integrated results obtained using PFTs.
However, the labor and the equipment required to use SF6 are
so extensive  that  testing  using this tracer gas  cannot  be
afforded outside of a research setting.

A commercial mitigator will never have the need for such
continuous ventilation results. As discussed above, a mitiga-
tor considering an ASD system will generally not require
information on house ventilation rates at all, and thus would
not need to conduct even PFT testing, let  alone the more
elaborate SF6 testing. Even a mitigator considering a house
ventilation  approach  would consider, at  most,  the
time-integrated PFT measurements (or, more likely, the more
approximate blower door approach to estimating house venti-
lation rates).

Although SF6 will never be used in the design and installation
of a commercial radon mitigation system, a brief review of
how it is used is presented here as background, since mitiga-
tors will occasionally see references to tests using SF6.

SF6 is injected into the house from a compressed gas cylinder.
Concentrations in the house air are then measured either using
a gas chromatograph located in the house, or by collecting
house air samples for chromatograpnic analysis at a remote
laboratory. The SF6 is utilized by one of three techniques:

   -  the dilution technique, where the tracer gas is initially
      brought up to a uniform concentration inside the house,
      and where the ventilation rate is then determined by
      observing the drop-off in the concentration over time
      (generally several hours). A standard method for using
      the dilution technique has been published (ASTM83).

   -  the steady state injection rate technique,  where the
      tracer  gas is continuously fed into  the house at a
      constant rate, and  the ventilation rate is determined by
      observing the concentration in the house air that is
      being maintained by this constant feed.

   -  the constant concentration technique, where the flow of
      tracer gas into  the house is continually adjusted as
      necessary in order to maintain a constant concentration
      of the gas in the house. The ventilation rate is deter-
      mined from the flows of tracer that are required.
                                                        108

-------
                                             Section 4
               Design  and  Installation  of Active  Sub-Slab
                              Depressurization Systems
The discussion in this section draws heavily from the detailed
review of available data on active SSD systems, presented in
Section 2.3.1.

4.1    Selection  of the Number of
Suction Pipes
The number of SSD pipes required will depend upon sub-slab
communication. In this document, three categories of commu-
nication are considered: good, marginal, and poor.

The objective in selecting the number of suction pipes is to
maintain adequate sub-slab depressurization everywhere. As
discussed in Sections 2.3.1b, 2.3.le, and 3.3.2, ideally,  the
SSD system should be maintaining the sub-slab depressuriza-
tion at: any measurable value (about 0.001 in. WG), if depres-
surization is measured during worst-case conditions of cold
weather and exhaust appliance operation; about 0.015 in. WG,
if measured during mild weather but with appliances operat-
ing; about 0.01-0.02 in. WG, if measured during cold weather
without appliances operating; and about 0.025-0.035 in. WG,
if measured during mild weather without appliances operat-
ing. As has  been discussed previously, these are  probably
conservative figures.

4.1.1  Houses Having Good Sub-Slab
Communication

As discussed jn Section 2.3.1a (House size) and 2.3.1c (Num-
ber of suction pipes), one or two suction pipes will generally
be sufficient for a SSD system to treat essentially any residen-
tial basement slab or slab on grade when sub-slab communi-
cation is good. This has been demonstrated on houses with
slabs as large as 2,700 ft2 (and on schools as large as 50,000
ft2). One pipe can be sufficient even when  the house has
hollow-block foundation walls. When a good layer of aggre-
gate is present, one or two pipes can sometimes be sufficient
even in houses where there are sub-slab obstructions such as
interior footings or forced-air supply ducts, which  could be
expected to reduce or interrupt the communication.

Where multiple separate slabs are present in a given house—
e.g., when a slab on grade adjoins a basement—it may be
advisable to install at least one suction pipe beneath each slab.
Installing a suction pipe beneath the adjoining slab on grade
(often from inside the basement), in addition to installing a
pipe through the basement slab, frequently appears to improve
radon reductions throughout the house. However, it is  not
always necessary for the purpose of reducing  concentrations
below 4 pCi/L when basement sub-slab communication is
good. The possibility of avoiding the need for a pipe beneath
the adjoining slab may be improved by installing the base-
ment suction pipe near the  stem  wall separating the  two
wings.

Sometimes a second slab will exist on the same level as
another, with a footing/foundation wall in between. This
situation can  occur, for example,  when  a  subsequent
slab-on-grade  addition has  been added  on to an older
slab-on-grade house, or when the house (as initially con-
structed) was completely bisected by a load-bearing founda-
tion wall resting on footings beneath the slab. When commu-
nication is good beneath  both slabs, it may or may not be
necessary to install a SSD suction pipe beneath each slab. If
local experience does not provide guidance on this matter, the
mitigator should perform  diagnostics, or should be prepared
to add a suction pipe in the second slab if treatment of the first
slab proves inadequate by itself.

The decision to install a second (or additional) suction pipe in
a given slab in a given good-communication house will be
determined by a mitigator's general practices, and by experi-
ence with houses in that geographical area. Where there are a
number of factors present that suggest the possible need for a
second pipe, a mitigator  will often find it cost-effective to
simply install the second pipe at the  outset, or to perform
daring-mitigation suction field diagnostics, as discussed in
the introductory portion of Section 3.3, rather than to risk a
call-back if an  initial one-pipe system does not perform ad-
equately.  Factors which, in combination, might suggest the
need for a second pipe include:

   -  a particularly large  slab;

   -   house or geological features which suggest that there
     will be significant air leakage  into  the system from
      inside the house or outdoors, thus increasing air flow
      and reducing suction field extension, such as:

      —  significant slab openings which cannot be  closed
         (e.g., a perimeter channel drain partially concealed
         behind wall finish, or a hollow-block fireplace struc-
         ture which penetrates the slab);

      —  block foundation walls which extend above grade
         in slab-on-grade houses, which can allow outdoor
         air to flow through the  block into suction pipes
         placed near the perimeter, and
                                                     109

-------
      --  highly permeable native soils,  which can allow
         outdoor air to flow through the soil into the system.

   -  experience in the geographical area which suggests that
      even a reasonably good aggregate layer can sometimes
      be uneven or interrupted, or which suggests that certain
      obstructions that are observed or suspected (such as
      forced-air ducts or grade beams/thickened slabs) can
      significantly degrade SSD  performance, despite the
      presence of good aggregate;

   -  hollow-block foundation walls;

   -  a high source term (i.e., soil gas radon concentrations
      greater than about 2,000 pCi/L), when present in com-
      bination  with some of the  other factors listed here,
      since failure to adequately treat even a small fraction of
      the entry routes can be important when soil gas concen-
      trations are very elevated;

   -  the presence of appliances that might be expected to
      create significant house depressurization, which might
      overwhelm the SSD system if sub-slab depressuriza-
      tions are only marginal.

Clearly, some of the above factors will be more important
than others, depending upon the situation. The combinations
of these  factors which will warrant a second SSD pipe in
good-aggregate cases may vary by geographical area.

The procedures for deciding whether a  house has "good
communication" have been discussed in Sections 3.2 and 3.3.

In general,  houses having an uninterrupted bed of aggregate
beneath the slab will have good communication. The commu-
nication will be especially likely to be good if the aggregate is
clean, coarse stone (i.e., without a lot of fine material to block
the voids between the larger stones); however, communica-
tion will  probably be reasonably good even if the aggregate
has not been washed and if some fine material is present As
discussed in Section 3.2, the presence of aggregate can some-
times be determined by visual inspection, if it is visible
through slab openings, or if the  homeowner observed the
house during construction and can confirm its presence. The
likely presence of aggregate can also often be inferred  from
local building codes, and from experience with other houses
in the area.

Good communication can sometimes exist,  at least under
parts of the slab, even when aggregate is not present.  Such
communication can be provided by: a) permeable native soils
underlying  the slab (such as well-drained gravel  soils); b)
subsidence of fill material beneath the slab, usually around the
perimeter where excavations for the footings were backfilled
during construction, leaving an air gap between the soil and
the bottom  of the slab; and c) insulation board or other such
material placed beneath the slab before it was poured, if this
material is itself porous, or if air gaps have developed between
the material and the slab or soil.

In the absence of aggregate, the mere presence of one of  these
other three features will not necessarily guarantee  sufficient
communication to permit effective treatment of the slab with a
single SSD pipe, unless vacuum cleaner suction field exten-
sion testing confirms that adequate suction field extension can
indeed be achieved. Highly permeable native soils can some-
times result in such high air flows from outdoors that multiple
suction pipes will be needed (and perhaps that the fan should
be reversed to operate the system in pressure, as discussed in
Section 2.4.1). Subsidence of backfilled soil cannot be relied
upon to be sufficiently complete to permit effective extension
of the suction field from a single SSD pipe around the entire
perimeter, if communication is otherwise poor.  Experience
with insulation board beneath the slab is too limited to permit
definitive statements about its possible  role in improving
communication in otherwise-marginal cases.

The presence of a good aggregate layer will not necessarily
ensure good communication (i.e., good extension  of the mea-
surable suction field) where there are interruptions in the
aggregate,  e.g., by forced-air  supply  ducts, or  by interior
footings or grade beams.  However, as discussed in Section
2.3. la (Sub-slab obstructions), one SSD pipe can often still be
sufficient, despite these obstructions, if there is otherwise a
good layer of aggregate  everywhere. The  success with a
one-pipe system in such cases will depend on the ability of the
suction field to extend through or under the obstruction.

Where the nature of the underlying aggregate is unknown, or
where the aggregate layer is known to be  incomplete, the
determination of whether or not the communication is  rela-
tively good can be made using the qualitative suction  field
extension measurement procedure described in Section 3.3.1.
Alternatively, a mitigator could decide to proceed with the
installation of the  first SSD suction pipe, and measure the
sub-slab suction field produced by this pipe in deciding upon
the nature of the communication and any need for additional
pipes (the dHrmg-mitigation diagnostic approach discussed in
the introductory portion of Section 3.3).

4.1.2  Houses Having Marginal or Poor
Sub-Slab Communication

Where sub-slab communication is  marginal or  poor, more
SSD suction pipes will be required, as discussed in Section
2.3.1a.

In houses having marginal communication, one suction pipe
per 350 to 750 ft2 of slab area has typically been  found to be
required, corresponding to 2 to 4 SSD pipes in a typically
sized basement slab. In cases where communication beneath
the basement slab is this marginal, it will  be increasingly
important that any adjoining slab on grade wing  also be
directly treated with additional SSD pipes.

In some  extreme cases, sub-slab communication will be so
poor that even 2 to 4 SSD pipes will prove to be  inadequate.
For most mitigators,  such truly poor-communication houses
will prove to  be only  a  few percent of the local market,
although in some regions of the country, such houses may be
encountered more frequently. These houses commonly have
their foundations built into bedrock, or into a very tight (and
sometimes moist) sand, or into wet clay, with no aggregate.
Mitigators have reported installing as many as 8 to 11 suction
                                                       110

-------
pipes in such houses, with 20 pipes installed in one house
being studied under a research project (corresponding to about
1 pipe per 100 ft2).

Various steps  can be considered to reduce the  number of
suction pipes in poor-communication cases, as discussed later.
These steps include excavation of pits beneath the slab where
the pipes penetrate, use of very high-suction/low-flow fans, or
perhaps the use of high-pressure air or water jets beneath the
slab to drill channels between the soil and the underside of the
slab (in an effort to improve communication). But it is clear
that, whatever is done, there will  be a significantly increased
labor effort and/or materials cost involved in installing SSD
systems in such houses, and there will likely continue to be a
significant number of suction pipes, even if the number can be
reduced by the steps just listed. In houses having such poor
communication, it would be appropriate to give consideration
to mitigation approaches other than ASD, prior to  selecting
that technology.

As with houses having good communication, houses having
marginal or poor communication  may have some characteris-
tics that will become apparent during the visual inspection.
These can include: a lack of aggregate visible through slab
openings; observations by the homeowner that no aggregate
was placed during construction; knowledge that building codes
or practices at the time of construction did not include use of
aggregate; familiarity with the local geology; and experience
with other similar houses in the area

In the absence of such guidance from the visual inspection,
the fact that the communication is not good cannot be deter-
mined until  the qualitative suction field extension measure-
ment is conducted, as described in Section 3.3.1. These quali-
tative diagnostics would identify whether the communication
is good or not good. However, if communication is not good,
the qualitative  diagnostics might not  reveal whether it is
marginal or poor. The vacuum cleaner suction might fail to
extend to the remote test holes  used in the qualitative ap-
proach, whether the communication is marginal or whether it
is poor, providing no basis for distinguishing between the two.
And the qualitative tests would provide no definitive guidance
regarding the number of suction pipes required.

Thus, if the qualitative suction field extension testing shows
that sub-slab communication is not good, the mitigator has
several options:

   -  If local experience suggests that the communication is
      probably marginal, rather than poor, proceed to install
      a two- to four-pipe SSD  system  without further
      pre-mitigation diagnostics.  Post-mitigation suction field
      diagnostics would be advisable to ensure that an ad-
      equate suction field has been established. The subse-
      quent retrofit of additional pipes into the system might
      be  necessary if the initial two- to four-pipe system
      proves to be inadequate in reducing radon  levels. It
      might be advisable to install the system with T fittings
      at appropriate locations to  simplify the retrofit of addi-
      tional pipes if needed.
   -  Install a one-pipe (or perhaps two-pipe) SSD system,
      with provisions to add additional suction pipes. Before
      putting the finish on the installation, operate the system
      fan and measure the sub-slab suction field developed.
      Add the number of additional suction pipes needed to
      achieve the required depressurizations, as indicated in
      the introductory portion of Section 4.1. This is the
      dHnng-mitigation diagnostic approach referred to ear-
      lier, and will provide the most definitive design guid-
      ance. It is, by definition, a quantitative diagnostic ap-
      proach, because the system flows and depressuriza-
      tions will automatically be equal to  the flows and
      suctions developed by the SSD system.

   -  Proceed with quantitative suction field extension mea-
      surements using a vacuum cleaner, as described in
      Section 3.3.2, to determine how many suction pipes are
      needed before the system is installed.

Many (but not all) investigators have observed that quantita-
tive suction field extension diagnostics with a vacuum cleaner
tend to over-predict the number of suction pipes needed, as
discussed at the beginning of Section 3.3.2, when the proper
flows are maintained in the vacuum. As  indicated in that
earlier discussion, some of the reasons for the over-prediction
may procedural errors in conducting the diagnostics or inher-
ent limitations on the ability to reproduce SSD sub-slab
depressurizations  using a vacuum cleaner. Other possible
explanations  why fewer-than-predicted suction  pipes can
achieve good radon reductions are that: a) the higher-flow
SSD system is creating a sub-slab ventilation component that
the vacuum cleaner cannot reproduce; and b) the goal depres-
surizations (up to 0.025-0.035 in. WG) are conservative.

If sub-slab suction field measurements are made using any of
the three options above, and if "effective radii" (as in Figure
11) are drawn, either figuratively or literally, the interpreta-
tion of those results to determine the number of suction pipes
must be done in conjunction with the determination of pipe
location, discussed in Section 4.2.2. Given  that the objective
is to achieve proper coverage of the slab with a suction field,
there are two alternatives for deciding upon what proper
coverage entails.

   -  Complete coverage of the slab. If it is desired to ensure
      that the selected depressurization is maintained be-
      neath all points of the slab, one would have to identify
      practical sites for pipes on the house floor plan. Pipes
      would be distributed among those sites as necessary
      such that the effective radii overlapped each other and
      the entire house perimeter, theoretically maintaining
      the desired depressurization  essentially everywhere
      under the slab. This approach is desired any time that:
      a) there are important entry routes in the interior of the
      slab, not just at the perimeter; and/or b) it is not
      desirable to locate the pipes near the perimeter for one
      reason or another, e.g., due to local concerns regarding
      air leakage into the system through perimeter block
      foundation walls. Due in part to air leakage  experi-
      enced with perimeter SSD pipes in slab-on-grade houses
      with very poor communication in Florida, this is the
                                                        111

-------
      approach recommended in Reference Fo90 for such
      houses.

   -  Coverage of perimeter only. Alternatively, one might
      decide to place the pipes near the perimeter, since the
      most common major entry routes (the wall/floor joint
      and, if present, the hollow-block foundation wall) will
      be located  there.  Also,  the sub-slab communication
      will commonly be somewhat better immediately beside
      the perimeter, due to the excavation and backfill that
      had to occur in that region when the footings were
      poured during construction. In this case, the effective
      radius might be used to select the pipe spacing around
      the perimeter, to ensure that the desired sub-slab de-
      pressurization is maintained at all points around the
      perimeter. By this approach, if the effective radii did
      not extend  sufficiently far to adequately depressurize
      the entire interior portion of the slab, it might be
      decided to leave that interior portion not fully treated,
      with the assumption that there are no major entry
      routes there.

Many mitigators and researchers focus on the perimeter treat-
ment approach in many instances (Sc88, Tu91b, Gad92, K192,
Mes92).

4.2   Selection of Suction  Pipe
Location

4.2.1  Houses Having Good Sub-Slab
Communication

When sub-slab communication is good, the location of the one
or two suction pipes can be fairly flexible.

One primary consideration in location selection will be the
convenience of the home-owner. This criterion will generally
dictate that the pipes be located in unfinished space, such as
the unfinished portions of  basements, or in utility rooms or
workshops in otherwise-finished basements and slabs on grade.
If no unfinished space is available, the pipes can be located
inside closets. The pipes should be located so that they do not
interfere with occupant traffic patterns, i.e., they should gen-
erally be  near other already-existing obstructions, or near
walls. If the homeowner has plans to finish a portion of  a
currently unfinished basement, or otherwise has preferences
regarding pipe location, these factors would also influence
site selection.

One-pipe SSD systems do  not require that the single pipe be
centrally located. If necessary, the pipe can usually be at one
end of the house, near the perimeter wall. Where two pipes are
used, it would be appropriate to space them as uniformly as
possible. For example, if half of a basement is unfinished, it
would generally be logical to locate both pipes in the unfin-
ished portion (one pipe at  one end of the house, the second
near the  central wall dividing the finished and unfinished
portions). But if  only  one room  is unfinished, it will be
necessary to locate the second pipe in a finished portion of the
slab.
The routing of the exhaust piping will often play an important
role in locating the suction pipe. For example, if the exhaust
piping is to be routed from a basement into an adjoining
slab-on-grade garage and  then through the garage roof, it
would be logical to insert the suction pipe through the base-
ment slab at a convenient location near the wall adjoining the
garage, in order to reduce the length of the horizontal piping
run. Or, if the exhaust is to be routed up through an existing
utility chase, the suction pipe might be located near that chase.

Factors associated with the slab could play a role in locating
the pipes. If an unfinished sump pit is present, for example, a
pit with no drain tiles or sump pump, hence not providing an
opportunity for sump/DTD, and if this pit is to be used as a
ready-made hole through the slab for a SSD suction pipe, this
would clearly dictate suction location. Observed or reported
sub-slab utility lines, or areas having heating coils built into
the slab, could  rule out certain parts of the slab, to avoid
damaging these utilities when drilling through the slab. How-
ever, some mitigators recommend deliberately locating suc-
tion pipes near sub-slab utility lines when there are interior
footings or other sub-slab obstructions (Bro92, K192); the
channel excavated for the utility line may penetrate through
the obstruction, thus providing an avenue for the suction field
to extend through the obstruction.

If a given house has multiple slabs, each having good commu-
nication, it may be desirable (although not always necessary)
to locate a suction pipe in each slab, as discussed in Section
4.1.1. One such case would be a basement house having an
adjoining slab on grade. Location of the basement suction
pipe near to the stem wall between the two wings may reduce
the likelihood that a separate pipe will be needed to treat the
adjoining slab.

Where more than one suction pipe is planned, it would be
desirable to locate the pipes so that they can most conve-
niently be manifolded together, if possible, so that they could
be connected to a single fan.

The desirability of locating  a suction pipe in a garage depends
upon  how  well the garage slab communicates with  the
living-area slab. For example, in basement houses where the
garage is On the  same level as the basement, the garage slab
and the living-area slab may be integral;  such houses are
effectively  walk-out  basements, with the garage being the
grade-level portion of the basement. In such cases, the suction
pipe can be placed in the garage with no penalty. But as
another example, slab-on-grade houses  having  attached
slab-on-grade garages commonly (by code) have (he garage
slab offset several inches below the living-area slab, some-
times with a footing/stem wall between the two slabs. Even
when both slabs were poured at the same time, the communi-
cation between the two may be uncertain.  In these cases, a
suction pipe in  the  garage may not be fully effective in
treating the living area unless the pipe is installed to draw
suction beneath the living-area slab, as discussed in Section
4.5.6.

It would generally be good practice to locate the suction pipes
toward suspected important entry routes to help ensure effec-
tive treatment of those routes, when this can be done without
                                                       112

-------
resulting in undue air leakage into the system. While this is
generally good practice,  it is not always necessary  in
good-communication cases, since the suction field will tend to
extend to that entry route even  when the suction pipe is
remote. For example, in a walk-out basement, the front wall
may be completely below grade while the rear wall is com-
pletely above grade; it would be logical in this case to locate
the suction pipe toward the below-grade wall. Or if there is a
block fireplace structure in the middle  of the slab which
penetrates the slab and rests on footings underneath, it might
be appropriate to bias the pipe toward that structure (although
not necessarily to place the pipe immediately beside the
structure).

However, where the entry route is a major opening, it would
be best to try to close this opening rather than attempting to
treat it, without closure, by placing a SSD suction pipe nearby.
For example,  locating a suction  pipe near to  an  unclosed
perimeter channel drain could result in substantial house air
leakage into the system, potentially preventing the suction
field from effectively extending to more remote points under
the slab while also increasing the house heating/cooling pen-
alty.

Suction pipes are commonly installed vertically down through
the slab from inside the house, such as shown in Figure 1.
Alternatively,  the pipes  can be inserted horizontally through
the foundation wall from outdoors, just below slab level, such
as illustrated in Figure 2. Vertical interior pipes will typically
be preferred for basement slabs, due to the depth of excava-
tion that would be necessary to get below slab level from
outdoors. However, where  the basement  slab is very highly
finished, horizontal extension of the pipe beneath the footing
from outdoors is sometimes preferred for aesthetic reasons
(K189). In slab-on-grade houses, where the slab is near to
grade level, vertical pipes inside the house are often still
preferred, especially in one-story houses with attics. In these
cases, penetration of the SSD pipe up through the ceiling and
into the attic provides a  convenient exit route for the exhaust
piping which is more aesthetic than the exterior fan and stack
that a penetration from outdoors could necessitate.

Horizontal penetration of the SSD pipe from  outdoors will
most commonly be applied in slab-on-grade houses where it is
not desired to  take an exhaust stack up through the house and
through the roof. Good examples are houses with a flat roof
and no attic (a feature characteristic of the Southwest), and
houses with a cathedral  ceiling. But as indicated above, pen-
etration from  outdoors can be used in any slab-on-grade or
basement house where,  due to interior finish, it is desired to
keep all piping outdoors.

If the suction pipes are to be installed horizontally  from
outdoors, the location  of  the  wall penetrations  would  be
selected based upon:

    -  Aesthetics. Any horizontal suction pipe that is to be
       above grade should preferably be in the rear of the
       house, away from the street.
   -  Accessibility. The presence of driveways, patios, land-
      scaping, etc., immediately beside the house would com-
      plicate access to the foundation wall.

   -  The possible presence of sub-slab utility lines in the
      vicinity.

   -  The desired spacing between suction pipes, if there are
      to be more than one, and the possible need  to  treat
      multiple slabs, if present.

Horizontal penetration of the suction pipe through the founda-
tion wall is also widely used in cases where a slab-on-grade
wing adjoins a basement. In these cases, suction pipe(s) to
treat the adjoining slab on  grade  are inserted  horizontally
through the stem  wall from  inside  the basement, just below
the adjoining slab. This piping is then commonly manifolded
together with the  vertical interior suction pipe(s) penetrating
the basement slab, and connected to a single fan. Treating the
adjoining slab with a horizontal penetration from inside the
basement is usually simpler, less expensive, and more aes-
thetic than the options of installing a vertical pipe inside the
finished living  area on the upper slab, or of installing a
horizontal pipe from outdoors. It also facilitates tying together
the piping treating the upper  and lower slabs into a single-fan
system.

Many mitigators prefer locating suction pipes near the slab
perimeter. The perimeter is often the most aesthetic location
and the most convenient for the occupant. In many cases, it
may also be  the best location from the  standpoint of radon
reduction; the suction field may extend effectively around the
perimeter, treating this region having the highest radon entry
potential. (This  latter point  may be less crucial  in
good-communication cases,  where  the suction field is likely
to extend effectively around the perimeter regardless of pipe
location.)

A concern has been raised  that location of pipes near the
perimeter of slabs at grade (slab-on-grade houses, or the
grade-level side of walk-out basements) could permit substan-
tial amounts  of outdoor air to flow into the system  through
block foundation  walls or through the soil  beneath  shallow
footings, interfering with suction field extension (Fo90, K192).
This could be  a  potential problem either  with horizontal/
exterior pipes, which will almost always be near the perimeter
of a grade-level slab, or with vertical/interior pipes  that are
located near the perimeter. However, as discussed in Section
2.3. Ic (Location of suction pipes), experience suggests that
when sub-slab communication is good, perimeter location of
SSD pipes in such slabs does not generally appear to present a
sufficiently severe air leakage problem  to  seriously impact
SSD design or performance,  even with block foundations that
extend above grade. Perhaps  leakage would become more of a
problem in cases where the underlying native soil is highly
permeable (e.g., well-drained, gravel soils). The problem with
perimeter location in such  houses appears to arise when
communication is not good.

It would thus  appear  that  mitigators do not have to feel
constrained against perimeter placement of SSD pipes in
grade-level slabs  having good communication. However, for
                                                         113

-------
safety's sake, it may still be desirable to locate the pipes away
from grade-level perimeters, especially with block founda-
tions and shallow footings, if another location  is available
which is also convenient.

4.2.2  Houses Having Marginal or Poor
Sub-Slab Communication

Location of the SSD suction pipes will become more impor-
tant when communication is marginal or poor, because favor-
able sub-slab characteristics can no longer be relied upon to
extend the suction field to the places where it is needed most.
Thus, it is more important that the pipes be located near the
major soil gas entry routes. (However, if the entry route is a
significant opening that cannot be at least partially sealed, the
suction pipe should not be so near that excessive air leakage
into the SSD system results.)

In particular, major entry routes include the perimeter wall/
floor joint, the wall/floor joint where any interior load-bearing
wall  (or fireplace structure) penetrates the slab, and
hollow-block foundation walls, where present. Thus, except
perhaps in some cases involving grade-level slabs, discussed
later, it would appear generally desirable to locate the suction
pipes near the foundation walls, where the radon entry poten-
tial is usually the greatest. The pipes should be about 6 in.
away from the walls, perhaps a little further in some cases,
just far enough away to avoid the footings and to provide
space for operating any equipment (such as a coring drill)
required to prepare the hole through the slab. Location of the
pipes in this manner reflects  emphasis on the approach re-
ferred to as Coverage of perimeter only, discussed at the end
of Section 4.1.2.

Locating  the pipes  near  the  foundation  walls has another
advantage, in addition to being near the major entry routes. In
poor-communication houses, the sub-slab communication is
likely to be best around the perimeter. Even if much of the
slab is poured on undisturbed, impermeable native soil or
bedrock, some excavation and back-filling would have had to
have taken place when the  footings  were poured. This
back-filled soil is likely to be more permeable than the undis-
turbed soil, and may even have subsided, creating an air gap
between the soil and the underside of the slab. Thus, a suction
pipe located in this region may be able to treat a much greater
length of the perimeter than would be predicted by the effec-
tive suction radius determined from quantitative suction field
extension diagnostics conducted in the middle of the slab.

In houses with marginal communication where  two suction
pipes are felt to be needed, one logical configuration would be
to place the one pipe near the middle of each of two opposing
walls, if these pipes can then conveniently be  manifolded
together. Such manifolding of remote pipes will be most
convenient in certain cases, such as: vertical/interior pipes in
unfinished basements; vertical/interior pipes in slabs on grade
where horizontal runs to connect the pipes  can be made in the
attic; or horizontal/exterior pipes where  the pipes can be
connected via a buried loop  of piping  around the  house
perimeter outdoors. Where such manifolding is not conve-
nient, the mitigator might wish to try pipes  on two adjacent
walls. In  houses where four pipes are needed,  one logical
configuration would be to place one pipe near the middle of
each perimeter wall, in an effort to best distribute the suction.

In some cases, there may be important entry routes toward the
interior of the slab, such as fireplace structures which pen-
etrate the slab, extensive slab cracking, etc. In these cases, it
may be desirable to  place one or more of  the pipes at  an
interior location, near these entry routes. Likewise, if there is
a load-bearing interior foundation wall which penetrates the
slab and rests on footings, this interior foundation wall may
require a pipe nearby, just as in the case of the perimeter
foundation walls.

In the few percent of houses where sub-slab communication is
truly poor, rather than just marginal, one pipe beside each
perimeter  wall  (and  beside major  interior  routes), as just
discussed, may be inadequate. In these cases, suction pipes
can be spaced around the perimeter at the intervals required to
maintain the desired sub-slab depressurizations (as summa-
rized in the introductory portion of Section 4.1) at the perim-
eter, consistent with the Coverage of perimeter only approach
in Section 4.1.2.

If quantitative suction field diagnostics have been performed,
the pipes can be positioned so that the effective suction radii
(as discussed under Interpretation of results in Section 3.3.2)
overlap around the perimeter. This approach assumes that the
sub-slab communication around the perimeter is the same as it
is at the location where the suction field extension diagnostics
were conducted, which may well have been  toward the slab
interior; this is probably a conservative assumption. Or, the
mitigator could utilize the daring-mitigation diagnostic  ap-
proach, installing the first few SSD pipes, measuring perim-
eter sub-slab depressurizations with the system fan operating,
and then continuing to add pipes at the necessary  locations
until the desired depressurizations are achieved everywhere
around the perimeter.

By the above method, no attempt is made to ensure that the
suction also covers the entire interior of the slab (the Com-
plete coverage of the stab approach in Section 4.1.2), unless
there are apparent significant entry routes there.

But where the slab  is above grade,  with  a hollow-block
foundation/stem wall and/or shallow footings, it may some-
times be desirable to locate the suction pipes away from the
perimeter  when  communication is poor. In  such cases,  the
Complete  coverage of the slab approach may be necessary.
Where this approach is necessary, the pipe locations can again
be  selected based either on quantitative  diagnostics  or
during-mitigation diagnostics, as above, except that now the
objective is to achieve adequate depressurization everywhere.

The possible need to locate pipes away from  the perimeter in
poor-communication slab-on-grade houses with shallow block
foundations is based upon experience in slab-on-grade houses
in Florida (Fo90, Fo92), and is supported by some experience
in Arizona and southern California (K192).  In some of the
Florida cases, perimeter locations provided  lesser radon re-
ductions than did pipes more toward the interior, suggesting
that the benefits of  perimeter placement were being out-
weighed by the disadvantage of increased outdoor air leakage
                                                       114

-------
into the system through the stem wall or through the soil under
the footings. However, perimeter placement will not always
be ruled out in such houses. In New Mexico, where communi-
cation was also very poor, with slabs above grade and block
foundations, good radon reductions were achieved with pe-
rimeter pipes (Tu91b), The more favorable results apparent in
New Mexico may be due, in part, to the fact that the source
term was much lower there than in Florida.

If it were desired to locate the suction pipe toward the interior
when inserting the pipe horizontally through the foundation
wall from outdoors, it would be necessary to auger horizon-
tally beneath the slab for some distance. This would add to the
complexity and cost, and would increase the risk of hitting
sub-slab utility lines. In most cases where the suction  pipe
must be located toward the interior, it will  probably be pre-
ferred to use vertical/interior suction pipes.

In some cases where sub-slab communication  is poor  or
uneven, the slab may in effect be divided into segments,  with
a suction field in one segment not effectively extending into
adjoining segments. In these cases, it would be desired  to
install at least one suction pipe in each segment. The qualita-
tive and quantitative suction field extension diagnostics de-
scribed in Section 3.3.1 and 3.3.2 might provide some indica-
tion of the presence and location of these segments. However,
it is likely that the number of test holes (and the single vacuum
cleaner suction point) will be inadequate to  map the sub-slab
characteristics sufficiently completely. The suggestion given
previously, that suction pipes be placed beside two to four of
the perimeter foundation walls (and near any rhajor unclosed
interior entry routes), should generally help ensure that all
such sub-slab  segments  will be treated. However, in cases
where  this does not occur, it will be necessary to conduct
post-mitigation diagnostics, as discussed in Section 11,  to
identify the areas of the slab not being adequately treated.

In addition to the above considerations regarding the place-
ment of SSD suction pipes in marginal- or poor-communication
houses, the  constraints which  pertain  to  placement  in
good-communication houses also apply. As discussed in Sec-
tion 4-2.1, among these constraints are:

   -  Location of the pipes in unfinished space or in closets.

   -  Location of pipes so that  they do not interfere  with
      occupant traffic patterns (which should be achieved if
      they are near the perimeter walls, as discussed above).

   -  Location of pipes to facilitate the manifolding of the
      multiple pipes together, if possible, and to facilitate the
      routing of the system exhaust.

   -  Location of pipes away from sub-slab utility lines.
       [Some mitigators recommend location of the pipes
      near such lines, since improved communication  may
      be found in the sub-slab trench that was excavated for
      these lines, and since the lines may penetrate sub-slab
      obstructions such as interior footings (Bro92, K192)].

   -  Location of pipes in garages only in cases where it is
      known  that these pipes will  communicate with the
      region beneath the basement or living-area slab, or
      when the pipes are installed in a manner which will
      treat the livable-area slab (Section 4.5.6).

   -  Insertion of pipes either  vertically from indoors, or
      horizontally from outdoors.

In houses having multiple slabs, the need to have suction
pipes beneath each slab is increased when communication is
marginal or poor. Quite possibly, multiple  pipes  will be
needed for each slab.

The discussion in Sections 4.1.2 and 4.2.2 focuses on careful
selection of the number and location of suction pipes as a
means for successful application of SSD to  marginal- and
poor-communication houses. Other approaches can also be
considered in an effort to reduce the number of suction pipes
required. These other approaches include:

   -  Methods for  improving the performance of a  given
      suction pipe,  given the marginal or poor communica-
      tion. Such methods include: excavating a larger pit
      beneath the slab at the point where the pipe penetrates,
      to further reduce pressure losses and to intersect addi-
      tional fissures/permeable  strata  not in direct contact
      with the underside of the slab (see Section 4.5.1, Exca-
      vating a pit beneath the slab); using a higher-suction
      fan, capable of developing suctions as  great as 25-40
      in. WG under the low-flow conditions  commonly ob-
      served in  poor-communication houses (see Sections
      4A2 and 4.4.3); or possibly mounting  two mitigation
      fans in series, to increase suction. More comprehensive
      pre-mitigation suction field extension measurements,
      to more completely map the sub-slab region and thus
      permit more  informed selection of pipe number and
      location, might also be considered as a method in this
      category.

   -  Methods for improving the communication. One devel-
      opmental method that has been explored experimen-
      tally has been the use of high-pressure air or water jets
      beneath the slab, in an effort to drill channels between
      the soil and the underside of the slab. (Note that water
      jets should never be used where the underlying soil is
      an expansive clay.)

It is emphasized that each of the above methods will involve
some effort and cost, which will at least partially offset any
sayings that might be achieved by reducing  the number of
suction pipes or facilitating their placement. But  in very
poor-communication cases, one or more of these measures
may be  needed, in addition to  multiple pipes, in order to
practically achieve adequate radon reductions.

4.3     Selection of Suction Pipe Type
and Diameter

4.3.1   Type of Suction  Pipe

Rigid, non-perforated poly vinyl chloride (PVC) piping is the
standard piping used in the industry, being readily available
and having the structural integrity suitable for this application.
                                                       115

-------
In some cases, rigid, non-perforated polyethylene (PE) piping
or acrylonitrile butadiene styrene  (ABS) piping have also
been reported to have been used, having an appearance and
physical characteristics generally similar to PVC piping. Where
PE or ABS piping is used, it must be noted that the PVC
cement used to join PVC piping and fittings can be ineffective
on these materials. The fittings must be of the same material
as the piping, and the appropriate cleaning solvents and adhe-
sives  for that material must be  used to join the pipes and
fittings.

In cases where the suction pipes are being inserted horizon-
tally from  outdoors, some mitigators  have reported using
flexible corrugated polyethylene or polypropylene piping (simi-
lar to drain tiles, but not perforated) for horizontal runs below
grade outdoors, due  to ease of installation in this  situation
(K189, K192). This flexible non-perforated piping has been
used:  to extend horizontally through the foundation wall in
slabs  on grade, as in Figure  2; as a manifold connecting
multiple horizontal suction pipes of the type shown in Figure
2; and, in basement houses, to extend beneath the footing into
the sub-slab region from outdoors.

In cases where an exhaust stack extends above the eave the
outside of the house, many mitigators 3- by 4-in. aluminum
downspouting (or mock PVC downspouting having a 3.5-in.
square cross-section) for the exterior stack, for aesthetic rea-
sons. Regular 2- by 3-in. aluminum downspouting has some-
times  been used also, but creates a substantial back-pressure.

Other types of ducting are not advised  for SSD  systems.
Flexible ducting and flexible clothes drier hose should never
be used in any part of the system. This type of hose is subject
to being torn, and can sag, creating a site for accumulation of
condensed  moisture. Also, it can be difficult to obtain suffi-
ciently gas-tight joints with this hose.

Many mitigators use thin-walled PVC piping, rather than
thicker-walled Schedule 40 piping. The thin-walled piping
provides significant savings in materials  costs (He91c), al-
though the amount of this savings will vary around the coun-
try depending upon  the local market  Labor time may be
reduced somewhat because the thin-walled piping is lighter
(and thus easier to handle), and much easier to cut Also, it is
somewhat flexible, allowing it to be flexed slightly to simplify
iis installation (e.g., in aligning it simultaneously through a
cored hole in the slab and through a hole in the ceiling above).
The quality of thin-walled piping and fittings can vary signifi-
cantly (Bro92); mitigators using thin-walled piping should
locate a source of high-quality  material.

The Schedule 40 piping has the advantage of greater structural
integrity, and should be considered in cases where the piping
may be subject to physical impacts, or where the increased
rigidity of the heavier pipe is desired. While many mitigators
use thin-walled pipe extensively, considering it to be suffi-
ciently strong and durable for this application, others refuse to
use anything other than  Schedule 40  due to  its increased
durability.  Some fittings, such as the 6-to-4-in. bushing dis-
cussed later as an option for supporting the suction pipe at the
slab, are available only in Schedule 40. In some locations,
Schedule 40 is now required for mitigation systems in specific
applications, such as new construction (K192).

At least two types of Schedule 40 PVC piping are available:
pressure-rated (solid PVC); and foam-core, suitable for waste,
drain, and vent use (K192). The Schedule 40 foam-core piping
and fittings are much less expensive than the pressure-rated
material, provide better durability than the thin-walled piping,
and offer some of the same advantages as the thin-walled in
terms of being easier to handle and cut than the pressure-rated
Schedule 40.

Thin-walled piping does not fit properly into  Schedule 40
fittings (joints,  elbows, tees), and vice-versa (Fo90, Bro92,
Mes92). The mitigator should not attempt to mix PVC piping
and fittings of different weights.

Resistance to ultraviolet (UV) radiation over years is a general
concern for PVC piping located outdoors (e.g., in exterior
stacks). Eventually, the pipe exposed to sunlight will become
brittle and crack if not properly protected. This  concern is
greatest for thin-walled  piping. Any  length of PVC piping
located outdoors should  be painted, regardless of weight, or
coated with a UV  protectant. Due to UV degradation, no
significant length of thin-walled piping should be used out-
doors even with a coating.

4.3.2  Diameter of Suction Pipe

Two factors are considered in selecting the diameter of the
piping used in SSD systems.

   -  First, the pipe should be of sufficiently large diameter
      such that there is  not excessive suction loss resulting
      from air flow through the piping. The suction loss for a
      given length of piping,  which results from friction
      between the gas and the pipe wall, increases as the air
      velocity increases. For a given air flow rate (in cfm),
      velocity increases  as pipe diameter decreases. Thus, the
      smaller the pipe diameter, the  more of a given fan's
      suction capacity will be consumed in moving air/soil
      gas through the piping, and the less will be available to
      maintain suction beneath the slab.

      Likewise, the pipe should be sufficiently large so that
      the velocity is low enough to avoid excessive  flow
      noise inside the pipe and noise where the exhaust jet is
      released. Mitigators report that flow noise and exhaust
      jet noise tends to become objectionable when it gets as
      high as 1,000-1,500 ft/min, which would correspond to
      roughly 90-130 cfm in a 4-in. diameter pipe. The flow
      noise can become a problem at even lower velocities
      when the pipe is routed through a bedroom closet
      Noise problems will be the worst at fittings, such as
      elbows and size reducers.

   -  Second, the pipe should be of sufficiently small diam-
      eter to reduce aesthetic impact and to facilitate installa-
      tion.

Typically, flows in SSD systems are in the range of 20 to  100+
cfm when operating with one of the commonly utilized 50- to
                                                        116

-------
90-watt centrifugal in-line tubular fans mounted in 4-in. pip-
ing, although individual installations will be  encountered
outside that flow range.

Figure 13 shows the suction loss per 100 ft of piping as a
function of the flow rate, with pipe diameter as a parameter,
calculated using standard fluid dynamic equations as pre-
sented in  Reference Ca60. This calculation  assumes
smooth-walled pipe having average wall friction. The flexible
corrugated polyethylene or polypropylene piping, with a cor-
rugated wall, would have much higher suction losses (about
1.8 times as high, according to Reference K192).

In practice, most mitigators usually  use 4-in. diameter PVC,
PE, or ABS piping in SSD systems, except in cases where the
mitigator routinely encounters very low-flow situations, or
where space limitations (such as the need to fit the pipe inside
a stud wall) dictate the use of smaller-diameter piping. Four-in.
piping represents a reasonable compromise between the two
factors listed above. It is sufficiently wide to limit suction
losses and flow noise to reasonable levels even in cases where
actual flows are at the upper end of the commonly observed
range for SSD systems, as discussed below. It is also wide
enough to keep flow velocities below the value at which flow
noise is reported to become a problem, except perhaps at the
highest SSD flow rates. It is sufficiently small to not be too
obtrusive; and it resembles some of the sewer piping which
    0.01
                              10
                         Air Flow (CFM)
                                                    100
 Figure 13.  Suction loss per 100 linear ft of straight piping, as a
           function of gas volumetric flow rate and pipe diameter, for
           circular pipe having smooth walls. (Derived from Refer-
         •  ence Ca60.)
may also be visible in some basements. It is readily available
as either thin-walled piping  or as  Schedule 40, and it is
reasonably easy to handle.

The detailed procedure for estimating system suction loss is
discussed later, in Section 4.6.1. For the purposes here, it is
sufficient to assume that the equivalent length of piping for an
example system  is about 70 ft (consisting of about 35 ft of
straight piping total upstream and downstream of the fan, plus
two mitered 90° elbows adding a flow resistance equivalent to
about a 35-ft length of straight 4-in. piping). As shown in
Figure 13, in the typical range of SSD volumetric flow rates
observed in reasonably good-communication houses with stan-
dard 50- and 90-watt tubular fans  (about 20 to 100 cfm),
friction losses are roughly 0.03  to 0.6 in. WG per 100 ft of
straight 4-in. pipe. For the example piping configuration with
70 linear ft, the total loss would  be in the range of 0.02 to 0.4
in. WG. These losses can be handled by the 90-watt in-line
fans, which develop 1 to 1.5 in. WG static pressure or more at
those flow rates. Except at  the highest flows (which  the
50-watt fans probably would not generate), the friction losses
can also be handled by the 50-watt in-line fans, which develop
suctions of 0.4 to over 0.75  in. WG at these flows. These
figures  thus confirm the reasonableness of selecting 4-in.
diameter piping in the typical case.

Three-in. diameter piping is useful for any portion of the
piping run that must be located inside stud walls. The 2-in. by
4-in. studs will  not provide a gap between the sheets of
sheetrock large enough to accommodate a 4-in. diameter pipe,
but 3-in. piping  will fit conveniently. The 3-in. piping will
have a greater suction loss than  the 4-in. piping, as indicated
in Figure 13. And the velocity at which flow noise is expected
to become objectionable occurs  at a flow rate of about 50-75
cfm with 3-in. piping, a flow range which might commonly be
achieved.

Using the procedure discussed  in Section 4.6.1, a mitigator
would have to calculate the impact of a given length of 3-in.
piping in any particular  system under consideration. From
Figure 13, with  3-in. piping, the friction losses at 20 to  100
cfm range from roughly 0.15 to 2.5 in. WG per 100 ft of pipe.
(In fact, with 3-in. piping, the flow range would be lower than
20-100  cfm, due to the  increased suction loss.) Using  the
example system considered previously for the 4-in. case (35 ft
of straight piping plus two elbows, which now has a resistance
equivalent to 60 ft of straight 3-in. piping), the total loss in the
system  would be 0.09 to  1.5  in WG.  At the lower flows
(giving 0.09 in.  WG suction loss), both the 50- and 90-watt
fans (which develop 0.75 and  1.5 in. WG,  respectively, at
those flows) could easily tolerate  the entire system being
installed with 3-in.  piping. However, at  the higher flows,
where the two fans sustain suctions of about 0.4 and 1  in. WG,
respectively, the 1.5 in. WG suction loss resulting when the
entire system consists of 3-in. pipe could not be tolerated. As
a result, use of 3-in.  piping throughout would cause both the
system flows and the sub-slab depressurization to be  reduced
with either the 50- or 90-watt fan. The impact on flows  and
sub-slab depressurizations at the higher flows would be less if
a 110-  to  140-watt in-line radial blower, recently  on  the
market (Ra92), were used (see Section 4.4.2).
                                                        117

-------
Rather than constructing the entire example system out of
3-in. pipe, one might use 4-in. pipe for much of the system,
and, say, use 20 ft of straight 3-in. pipe (and a 4-to-3-in.
reducer) for the segment that extends up between the stud
walls. In this case, the suction loss for the total system over
tlie range of 20-100 cfm would be about 0.1-0.8 in. WG. Most
of this range could be tolerated by the 90-watt fan, except
perhaps at the highest flows, and part of it could be tolerated
by the 50-watt fan.

It must be recognized that if the suction loss becomes too high
for the fan to handle at the assumed flows, the effect will be
that the system flows will drop, and the fan will operate at a
different point on its performance curve. As a result of this
drop in flows, sub-slab depressurizations will also drop. Since
SSD systems in good-communication houses are sometimes
over-designed anyway, this drop in flows and sub-slab de-
pressurizations may not be serious in terms of radon reduction
performance.

In view of the above discussion, it is believed that if the
sub-slab communication  is  good and if a 90-watt fan is
utilized, a segment of 3-in. piping in the system (or sometimes
even use of 3-in. piping throughout the  system)  will  be
acceptable unless the flows are in the upper part of the range
and/or  the segment of 3-in.  piping is long. That is, the in-
creased suction loss created by this smaller piping may  be
accommodated by the 90-watt fan, so long as the sub-slab
communication is good enough such that high sub-slab de-
pressurizations do not have to be maintained at the suction
hole.
Three-in. piping is slightly less expensive than is 4-in. piping
of the same weight, and is a little easier to handle, because a
worker's hands will easily fit around its circumference (He91b,
He91c). However, 3-in. piping and fittings are less readily
available in some areas. Also, a mitigator who stocks prima-
rily 4-in. piping will encounter some additional time require-
ments  in planning for the use of 3-in.  piping in a single
installation, and in procuring the proper amount of the smaller
Pipe-

In poor-communication cases where flow is very low—on the
order of about 10 cfm or less per suction pipe—1.5- to 2-in.
diameter piping has sometimes been used, to facilitate instal-
lation and improve system aesthetics. At such low flows,
system suction losses and flow noise should not be a problem.
However, as discussed below, if flows became as high as
25-35 cfm,  suction losses could become unacceptable for
many fans, and flow velocities would surpass the 1,000-1,500
ft/min value at which noise can become objectionable.

From Figure 13, the friction loss in 2-in. piping at a flow rate
of 10 cfm is roughly 0.3 in. WG per 100 ft. For the example
system considered above (35 ft of piping plus  two elbows,
now having a flow resistance equivalent to 52 linear ft of 2-in.
piping), the system loss (assuming the entire system consists
of 2-in. piping)  would be 0.15  in. WG. This loss can  be
handled by the standard 90-watt centrifugal tubular fans (which
can develop about 1.5 in. WG static pressure at 10 cfm), the
110- to 140-watt in-line radial blowers (which develop 2.5 to
4 in. WG at this flow), and the high-pressure/low-flow fans
(capable of about 20 to over 40 in. WG at this flow). However,
at 35 cfm, the friction loss in the 2-in. piping rises to about 3
in. WG per 100 ft, or about 1.5 in. WG for the 52-ft example
system. The loss in the example system is about equal to the
total static pressure mat can be developed by the  90-watt
in-line fans at 35 cfm; it is also greater than the static pressure
that can be developed by many of the high-pressure/low-flow
fans at this flow rate, which is higher than many low-flow fans
are designed to handle. The loss could be  handled by the
140-watt in-line radial blower,  which reportedly generates
almost 3.5 in. WG at 35 cfm (Ra92).

Thus, with essentially all of the 90-watt and higher-powered
fans, 2-in. piping can be considered for individual SSD suc-
tion pipes in poor-communication houses, if desired for con-
venience or aesthetic reasons, when flows in the riser are in
the range of 10-20 cfm and less. Where flows are sufficiently
low and the fan suction sufficient great, analysis analogous to
that in the preceding paragraph indicates that 1.5-in. piping
can also be considered.

Systems in houses having poor communication often consist
of a number of suction pipes connected to a manifold pipe,
which in turn leads to a single fan. As indicated above, 1.5- to
2-in. piping may be satisfactory for the individual suction
pipes, which may be contributing less than 10 cfm apiece to
the system. However, the manifold pipe, which will carry the
combined flow from all of the suction pipes, should probably
be 4-in. piping, if the combined flow is greater than roughly
25-35 cfm (depending upon the fan/blower performance curve
and the length  of piping).

As implied in  the preceding discussion concerning the ex-
ample piping configuration, suction losses in the system for a
given pipe diameter will depend not only on  the length of
straight piping, but on the number and  type of fittings in the
network (e.g., elbows, tees, 6-to-4-in. reducers). These fittings
create a  flow  resistance/suction loss equivalent to  certain
lengths of straight piping. Depending upon the number and
type of fittings, the fittings can sometimes be a more impor-
tant contributor to suction loss than is the straight piping. The
method of using Figure 13 to calculate the total suction loss in
different piping systems is presented in more detail in Section
4.6.1, in connection with design of the piping network.

If flows have  been measured during pre-mitigation suction
field extension testing (as described in Section 3.3.1 and
3.3.2), or if more extensive sub-slab flow measurements have
been made (as described in Section 3.5.1), the mitigator will
have an early indication of whether the system is likely to be
moderate- to high-flow (suggesting the need for 3- to 4-in.
piping), or if it is likely to be very low-flow (in which case
1.5- to 2-in. piping can be considered, if desired). In some
cases, the likelihood that a house may present a very low-flow
case will be apparent to a mitigator prior to any diagnostic
testing, based upon the visual survey and upon experience in
the geographical area.

4.4    Selection of the Suction  Fan

The fan must be able to maintain an appropriate suction in the
piping (and hence beneath the slab) at the flows encountered
                                                       118

-------
in the system. Thus, the appropriate fan for a given system
will depend upon the flow characteristics of the sub-slab and
the difficulty expected in achieving an adequate suction field
distribution.

Tubular fans. While many commercially available fans and
blowers might be considered for use in mitigation systems, the
most widely utilized fans to date are centrifugal in-line fans,
typically rated to draw 50 to 90 watts maximum. These fans
are referred to in this document as tubular fans, consistent
with the terminology used in the heating, ventilating, and air
conditioning industry (ASHRAE88). Because of their wide-
spread use, in-line tubular fens are the type of fan illustrated in
the various figures in this document.

Where sub-slab communication is good,  the smaller 50- to
70-watt tubular fans will sometimes give adequate sub-slab
depressurizations, if flows are not real high. However, in
general, the 90-watt fans are recommended for use in any
house having good or marginal sub-slab communication, be-
cause of their ability to generate  relatively high suctions
(generally 1 to over 1.5 in. WG) over the full range of flows
observed in SSD systems (20 to 100 cfm). The 90-watt tubular
fans have also been applied successfully in houses having
poor sub-slab communication, although  the flows in such
houses are often so low that the fan would not be operating at
an optimum point on its performance curve.

High-suction/low-flow blowers. In poor-communication
houses, where system flows can be very low, and a high
suction might aid in extending the suction field beneath the
slab, high-suction regenerative or high-speed centrifugal blow-
ers have been used in some cases. High-suction blowers that
have been specifically designed for radon mitigation applica-
tions can develop suctions as high as 5 to 50 in. WG at the low
flow rates commonly encountered in poor-communication
systems (around 20 cfm) (Pe90, Ra92). The purchase price of
these high-suction blowers is much  greater than that of the
standard  centrifugal in-line tubular fans, and their power
consumption  is much higher (about 200-300 watts maxi-
mum). These high-suction  blowers  are not configured for
in-line use, and thus have a different appearance from the fans
illustrated in this document.

Success with the use of these blowers has been reported in a
number of installations (Cra91, Py92, Zu92). According to the
manufacturers, a fair number of these blowers have been sold
(Zu92). However, a rigorous comparison has not been con-
ducted of the effectiveness of the high-suction blowers vs. the
90-watt tubular fans in poor-communication installations.

A potential problem with some of these high-suction blowers
is that they are designed for low flows, with maximum flows
often no higher than 25-50 cfm. If flow increases after instal-
lation (e.g., due to development of new air leaks through the
slab), the blower may be overwhelmed and suction may drop
to low values. Some users also report that these blowers can
be noisier than the tubular fans.

In-ttne radial blowers. A series of in-line radial blowers
designed for radon mitigation applications has recently come
on the market, offering a performance curve intermediate
between the 90-watt tubular fans and the high-suction/low-flow
blowers. Unlike the high-suction blowers, these new blowers
reportedly can handle the higher flows sometimes observed in
SSD systems, being able to maintain suctions of almost 1 in.
WG at flows of 100 cfm, similar to the 90-watt fans. But these
blowers can reportedly develop suctions of over 2 to 4 in. WG
as flows drop, compared to the 1.5+ in. WG maximum with
the 90-watt fans. These in-line radial blowers could thus be a
reasonable choice in marginal- or poor-communication houses.

However, the radial blowers use somewhat more power than
the 90-watt fans (95 to 140 watts maximum), and they are
somewhat more expensive to purchase (although the least
powerful radial model is comparable in price to the 90-watt
fans). Their performance (in terms  of reliability, durability,
and noise levels) has not been as widely demonstrated as it has
for the tubular fans, although the manufacturer claims that a
large number have been successfully installed to date.

The radial blowers have an appearance similar to that of the
in-line tubular fans.

Basis for fan/blower selection.  In the large majority of
cases, the fan will be selected based upon experience in other
houses in the area. If sub-slab aggregate is observed during
the visual inspection (Section 3.2) or is known to be present,
communication will be good. In this case, flows are more
likely to be toward the middle or upper portion of the 20-100+
cfm SSD flow range (although this is by no means assured),
and a 50- to 90-watt tubular fan will usually be the appropriate
choice.

If no aggregate is present based upon the visual inspection, the
fan selection must be based on experience. A 90-watt fan will
probably be a reasonable choice in any case. Unless the native
soil is highly permeable or air leakage is otherwise expected
to be high, a 95-  to 140-watt radial blower could also be a
reasonable choice, if these prove successful in practice. A
high-suction/low-flow  blower would not be a good choice
based solely on visual inspection (without further, sub-slab
diagnostics),  unless local experience has demonstrated that
flows in such cases are consistently very low.

If suction field extension diagnostics have been conducted
(Sections  3.3.1 and 3.3.2), and if flows were qualitatively
judged or quantitatively measured during these diagnostics,
the flow and communication results could help in fan selec-
tion.  Moderate flows  and excellent communication might
suggest that a 50- to 70-watt fan would be satisfactory. But in
most cases, especially when flows are high or communication
is not excellent, a 90-watt fan would be suggested. If flows are
moderate  and communication is marginal or poor, a 110- to
140-watt  in-line radial blower  might  be suitable. Only if
quantitative flow  measurements during quantitative suction
field testing show flows below 10-20 cfm should a high-suction/
low-flow blower be considered.

Where pre-mitigation  sub-slab flow measurements are con-
ducted to  aid in fan selection, the qualitative flow estimation
approach  (see Section 3.3.1, Test procedure) or the simple
quantitative approach (see Section 3.5.1, Simple sub-slab flow
measurement) will usually be sufficient, if flow results are
                                                       119

-------
needed at all. These approaches will reveal whether sub-slab
flows are generally high, moderate, or low. The more compre-
hensive, quantitative flow measurement approach discussed
in Section 3.5.1 (Mere extensive sub-slab flow measurements)
will probably be of value infrequently, if ever, to most mitiga-
tors. The performance curve for each of the individual fans on
the market covers a fairly broad range; thus, it is doubtful that
the more extensive flow testing would often provide informa-
tion that would  aid in any meaningful fine-tuning of fan
selection (i.e., enabling the mitigator to pinpoint the preferred
brand or model of fan for a specific house), beyond what can
be done with the simpler information from the suction field
testing.

The one case in which the more extensive pre-mitigation flow
measurements might be of value is the  case where sub-slab
flows are very low, and the mitigator is trying to decide  on
whether to utilize one of the high-suction/low-flow blowers.
The performance curves of these blowers  can vary signifi-
cantly from one another, and from those of the in-line tubular
fans and radial blowers. A high-suction/low-flow blower can
be overwhelmed if flows  are only slightly higher than ex-
pected. Thus, when one of these blowers is being considered,
it could be well worth the investment to take the time to make
a more extensive, quantitative pre-mitigation flow measure-
ment to ensure that such a blower is indeed the best choice.
Fortunately, the industrial vacuum cleaners often  used to
conduct the diagnostics  commonly have  the  performance
characteristics needed to  simulate the high-suction blowers
reasonably well.

Another question that a mitigator might like to address with
extensive, quantitative sub-slab flow diagnostics is whether a
SO-  to 70-watt tubular fan  might be sufficient  in
good-communication cases, avoiding the need for a 90-watt
fan. Unfortunately, that question often cannot be effectively
answered by this test, if a vacuum cleaner is being used for the
pre-mitigation diagnostics. Flows  in good-communication
houses will often be too high to permit  a vacuum cleaner to
reproduce SSD flows. To address  this  question with these
diagnostics, the mitigator would have to use portable ASD fan
test stands, as discussed in the introductory text in Section 3.3,
utilizing 50-, 70-, and/or 90-watt fans to accurately reproduce
the flows and suctions that such fans would generate in that
particular house.

4.4,1  Centrifugal In-Line Tubular Fans

The 50- to 90-watt  centrifugal in-line tubular  fans have be-
come the standard fans used in the industry, for a variety of
reasons.

   -   Their performance curve—i.e., the suctions developed
      at different flows—are in the range needed for SSD
      applications.  In particular, the 90-watt  fans provide
      reasonably high suctions over the full range of SSD
      flows.

   -   Their in-line configuration makes them convenient for
      installation in the system piping.
   -  Their purchase price is reasonable, ranging from roughly
      $90 for the 50- to 70-watt fans (with 4- to 5-in. cou-
      plings), to roughly $100 for the standard 90-watt fan
      with 6-in. couplings,  to roughly $130-$150 for the
      larger, 100-watt fans with 6- to 8-in. couplings, based
      upon 1991 costs (He91b, He91c).

   -  They are quiet (about 2 to 4 sones), if mounted prop-
      erly to avoid vibration.

Table 1 lists some of the tubular fans that are being marketed
for radon mitigation, from three manufacturers.  This list is
provided as an illustration of the types of fans available, and is
not to be construed as a complete listing of fan manufacturers,
or of the models offered by the manufacturers listed. To put
the capabilities of these fans into some perspective, the listing
is subdivided into six general categories, according to ascend-
ing power requirements or fitting diameter.

Of the six categories of fans listed in Table  1, the Category 4
fans (the 90-watt fans having 6-in. fittings) are the ones which
appear to be the most universally applicable. They should
provide ample suction and flow in essentially all good- and
marginal-communication houses. In some  cases, they  do a
credible job even in poor-communication houses.

The performance curve for each fan is summarized by the
flow-vs.-suction columns on the right-hand side  of Table 1.
The 90-watt fans are rated as being able to  move 270 cfm at
zero static pressure, i.e., the condition that would exist if a
free-standing fan with no upstream or downstream couplings
or piping were operating at full power  in free air,  with
essentially ambient (0 in. WG) pressure at both the inlet and
outlet. When this fan is mounted in a piping network with the
resistance of the sub-slab and the suction piping and couplings
on the suction side, and with the resistance of any exhaust
piping and stack on the pressure side, flows  will decrease
significantly below 270 cfm, reflecting the extent of the
suction at the inlet and of the back-pressure at the outlet.
Simply installing a 6-in.-to-4-in. coupling on each side of the
90-watt fan, as would be necessary to install this fan, with its
6-in. fittings, into a network of 4-in.  piping, would create
enough resistance to reduce the flow to roughly 180 cfm
(Br92).

In good- and marginal-communication houses, the 90-watt
fans typically draw flows ranging from below 20 to over 100
cfm, depending upon the resistance  of the sub-slab and the
suction loss  in the piping network. In accordance with their
performance curve, the fans can maintain a suction of about
1.0 in. WG at the upper end of this flow range; but in fact, the
suctions in the SSD system piping near the slab  penetration
will sometimes be below 0.5 in. WG at 100 cfm due to friction
losses in the piping. (See the discussion in Sections 4.3.2 and
4.6.1.) At the lower end  of the flow range, the fans can
maintain a suction somewhat greater than  1.5 in. WG. (Al-
though Table 1  indicates a suction >  1.0 in. WG at zero flow
for the Category 4 fans, based upon manufacturers' literature,
these fans have been widely demonstrated to generate some-
thing close to 1.7 in. WG at no flow.) At such low flows,
friction losses in the piping will be much  reduced, and the
suction being maintained in the piping will be close to the
                                                       120

-------
Table 1.   Examples of In-Line Tubular Fans Which Have Been Marketed for Radon Mitigation1*
Fan
Category
1

2

3



4



5

6



Fan
I.D3
K4
R100
F/FR100
K4XL
F125
K5
T1
R125
F/FR150
K6
T2
R150
F/FR160
K6XL
F/FR200
K8
T3A
R200
Max.
Watts
-
50
70
-
70
-
-
50
90
-
--
90
100
--
100
—
—
125
Horse-
power
1/40
--
..
1/20
1/30
1/40
1/40
--
1/20
1/20
1/20
-
1/15
1/15
1/15
1/15
1/15
--
Fitting
• Diameter
(in.)
4
4
4
4
5
5
5
5
6
6
6
6
6
6
8
8
8
8
RPM
2800
2900
2500
2450
2500
2800
2800
2900
2500
2150
2150
2350
2150
2150
2150
2150
2150
2800
Flow
@ 0 in. WG
122
124
160
179
205
158
158
146 .
270
270
270
270
361
360
410
410
410
541
(cfm)
@ 1 in. WG
--
--
60
50
-
-
-
--
110
110
110
102
122
122
135
135
135
324
Max.
Suction
(in. WG)
>0.75
>0.75
>1.0
>1.5
—
>0.75
>0.75
>0.75
>1.0
>1.0
>1.0
>1.0
>1.0
>1.0
>1.0
>1.0
>1.0
>2.0
  This listing is intended as an example of the types of fans which have been utilized in, or marketed for, radon mitigation systems. This list is
  not intended to be a comprehensive compilation of fans suitable for this application, nor does inclusion of a fan on this list signify endorsement
  by EPA.                      .  '
  All fan specifications presented here are taken directly from the manufacturers' literature.
  The fan identifiers refer to the following manufacturers: F and FR = Fantech; K = Kanalflakt; R = Rosenberg; T = Kanalflakt.
maximum which can be generated by the fan. In general, the
poorer the communication, the lower the flow and hence the
greater the suction.

In houses having good communication, the 50- to 70-watt fans
(Categories 1,2, and 3 in the table) can be sufficient to reduce
indoor levels below 4 pCi/L, unless flows  are very high.
Alternatively, the 90-watt fans could be turned down to oper-
ate at reduced power; this step would probably prolong fan
life, so long as the  fan was not turned down so far that the
reduced air flow could not adequately cool the motor. How-
ever, as discussed in Section 2.3.1, use of a  smaller fan, or
reducing the fan power, can result in increases in the residual
indoor radon level, even if levels remain below 4 pCi/L.

There  are several incentives to  use  a smaller  fan. These
include reductions in the system installation and operating
costs and, on a national scale, reduced consumption of natural
resources (coal, oil, gas) and  reduced pollution from the
generation of electrical power to operate the larger fans. In
terms of the installation cost, the savings in the purchase price
of the smaller fans, relative to the 90-watt fans is fairly small,
about $5 to $10 in 1991 dollars. Use of a 50-watt fan will
reduce operating  costs (due to fan electricity and the house
heating/cooling penalty) by roughly $70 per year relative to
the 90-watt fan, or about $5.50 per month, in an "average"
climate (Washington, D. C.) (He91b, He91c).  The actual
operating cost savings can vary significantly, depending on,
e.g., the local climate and fuel costs. Each mitigator and
homeowner will need to make their own decision regarding
the tradeoffs between the increase in health risk due to in-
creased radon exposure likely with the smaller fan, versus the
somewhat reduced costs and reduced environmental impact of
the smaller fan.

Another potential advantage of using a smaller fan is reduced
risk of backdrafting combustion appliances,  due to the re-
duced amount of air drawn out of the house. Backdrafting will
be a serious situation if it occurs and should be checked as part
of post-mitigation diagnostics (Section 11.5) or perhaps dur-
ing pre-mitigation diagnostics (Section 3.5.4). However, as
discussed elsewhere, backdrafting usually occurs in conjunc-
tion with SSD systems only when the house was in a backdraft
condition or had only a marginal draft prior to mitigation. This
issue will need to be addressed regardless of the  SSD  fan
selection. Thus, it is doubtful that concern about backdrafting
will often influence the choice between a 90-watt fan and a
50- to 70-watt fan.

Based upon experience, in very few cases should the 90-watt
fans be inadequate to handle the volume of air withdrawn
from the sub-slab  in residential SSD installations. Flows
greater than perhaps 150-180 cfm combined with suctions less
than perhaps 0.25-0.35 in. WG in the system piping near the
slab penetration  (and  confirmed by  low depressurizations
measured beneath the slab) would suggest that the 90-watt fan
                                                         121

-------
might be inadequate. If flows are so high from a residential
installation that the 90-watt fan cannot move enough air to
maintain adequate sub-slab depressurizations, the first step
would be to look for (and close) major unclosed slab openings
which may be permitting large amounts of house air to enter
the system.  The system piping might also be checked for
leaks. If the high flows are found to be due to an unavoidable
cause,  such  as highly permeable native soil, or major slab
openings that are not practical to close, such as a perimeter
channel drain concealed behind wall finish, then options other
than a larger fan might be considered next These options
could include addition of more suction pipes, or conversion of
the system to a sub-slab pressurization system in the case of
highly permeable native soils.

If the decision is made to  remain  with  a SSD system, the
90-watt fan may be replaced with a 100-watt fan. As shown in
Table 1, the 100-watt units with 8-in. couplings can move
about 410 cfm or more at zero static pressure, compared to
270 cfm for the 90-watt fans, and can develop suctions at zero
flow which are at least comparable  to those with the 90-watt
fans. These fans have a purchase price about $30-$50 more
than the 90-watt fans (1991 prices), and will have a somewhat
greater operating cost, due  to the modest increase in power
requirements and to the increased amount of treated air that is
likely to be withdrawn from the house. Alternatively, a second
90-watt fan could nominally be mounted in parallel with the
first, a  step which  would likely increase flows  more than
would a switch to a 100-watt fan. Use of a second 90-watt fan
would also result in a greater increase in purchase price (a
$100 increase rather than a  $30-850 increase), and a greater
increase in  operating cost  due to electricity and heating/
cooling penalty.

It is re-emphasized that the 100-watt high-flow fans would be
considered only in cases where flows were unusually high. If
the  problem  is  that  flows are  very  low,  as  in
poor-communication houses, the use of 100-watt fans  (in an
attempt to better  extend the suction field)  would not be a
suitable response. Rather, in low-flow cases, the appropriate
option to consider would be use of higher-suction blowers, as
discussed in Sections 4.4.2 and 4.4.3.  (Alternatively, two
90-watt fans could be nominally be mounted in series to
roughly double the suction,  although these fans might not be
operating at an optimum point on their performance curves.)

In assessing the actual operating costs of SSD systems,  it
should be noted that the power ratings presented in Table 1 are
the maximum power that the fan would draw, if the fan were
operating at maximum flow and zero suction. In most cases,
the fan will be  operating  at an intermediate point  on its
performance curve, and will be drawing less power. For
example, in  one study where fan power drawn by SSD fans
was measured, the 90-watt  fans were in fact drawing  60-65
watts (Tu91c); in a second study, they drew 59-72  watts,
averaging 68 watts (Bo91).

One other consideration in selecting a fan will be whether the
fan is rated for outdoor use. Of the general exhaust configura-
tions discussed in Section 4.6, two have the fan located in an
enclosed space: the case where the exhaust stack rises up
through the interior of the house, with the fan in the attic; and
the case where the piping from a basement SSD system is
routed out through the basement band joist into an adjoining
slab-on-grade garage, with the fan in the garage. For these
configurations, the fan would not have to be rated for outdoor
use.

However, two other exhaust configurations envision the fan
being outdoors: the case where the piping from  a basement
system is routed out through the band joist to outdoors, with
the fan mounted on the piping outdoors; and the case where
the suction pipes penetrate beneath the slab horizontally from
outdoors, with the fan necessarily mounted on a riser from the
horizontal pipe(s)  outdoors. In these cases, if the fan were not
rated for outdoor use, it would have to be suitably enclosed.
As discussed in Section 4.6, EPA's interim mitigation stan-
dards rule against location of the fans within liveable space
(such as inside basements) due to concerns about leaks of
high-radon  exhaust gas from the pressure side  of the fan.
Thus, when a fan is not rated for outdoor use, locating it inside
the basement is not considered an appropriate approach for
addressing this issue.
Many of the fans listed in Table 1 are not currently UL listed
for outdoor applications, due to concerns about whether the
electrical components are adequately weather-proof. Some of
these fans have been incorporated in numerous outdoor instal-
lations over the years, and have demonstrated reasonable
durability during that time period, as discussed in Section
2.3.1. However, occasional failures have been reported in fans
not rated for outdoor use due to rainwater entering the electri-
cal box on  the side  of the fans  (Bro92). This has been
observed when the coupling is not tight where the wire enters
the box; it has also been observed when the electrical switch
on the outside of the house is higher than the fan, so that
rainwater will run down the wire into the  coupling. One
possible solution to this problem that has been used in practice
is to drill a hole near the base of the fan's electrical box, so
that the water will  drain out and not flood the box (Bro92).
Infrequent cases have also been encountered where the fan's
electrical box has flooded by condensate inside the piping,
when the wiring leading from the box to the fan motor inside
the housing has not been adequately sealed where it enters the
box from the interior of the fan (Mes92).
At the time of this writing, only one of the fan manufacturers
listed in the table is marketing fans  UL  listed for outdoor
mounting; others are currently working on models that will be
rated for outdoor use.
One other consideration in selecting a fan is that the fan
housing should preferably be integral, and otherwise designed
to reduce leakage of high-radon exhaust gas on the pressure
side. Even though fans in attics, crawl spaces, and garages are
outside the living area, it is generally desirable to eliminate
leakage of exhaust gases  in  those areas.  In  general, fans
having integral plastic bodies minimize leakage; this includes
all of the fans in Table 1 except for the Kanalflakt K series
fans. Where there are leakage points around the fan housing,
these should generally be caulked before installation.
                                                       122

-------
4.4.2  In-Line Radial Blowers

In-line radial blowers designed specifically for radon mitiga-
tion applications have recently been introduced onto the mar-
ket. From the manufacturer's performance data, these blowers
should be able to produce suctions comparable to those of the
90-watt tubular fans at flows toward the upper end of the
typical SSD flow range (i.e.,  about 1 in. WG at 100 cfm).
Thus, the radial blowers should perform as well as the stan-
dard tubular fans at relatively high-flow conditions.

But these radial blowers offer the advantage of much higher
suctions (over 2 to 4 in. WG)  at the low flows, compared to
just over 1.5 in. WG with the  90-watt fans.  Thus, these new
blowers would be expected to provide higher suctions over
much of the SSD flow range. This feature could help extend
the suction field in marginal- and poor-communication houses
using fewer suction points than would  be needed  with a
standard fan; it could also accommodate the piping friction
loss where small-diameter piping must be used.

While these in-line radial blowers provide lower suctions than
do  the high-suction/low-flow  blowers discussed in Section
4.4.3 at" low flows (e.g., 10-20 cfm), they offer a potentially
significant advantage in that  they can handle much higher
flows (over 100 cfm). Flows this high would overwhelm the
high-suction/low-flow units, reducing their suctions  to low
levels. Thus, the in-line radial blowers could be considered in
marginal-communication cases where flows  would  be too
high for the high-suction units, and will be better able to
handle increases  in  system  flows  over  time  in
poor-communication houses if, e.g., new air leaks appear in
the slab after installation.

The performance characteristics of three in-line radial blow-
ers are summarized in  Table 2.  This list is intended for
illustration of the types of units available;  it is not intended as
a comprehensive compilation of blowers having these charac-
teristics, nor does inclusion of a blower  on this list  signify
endorsement by EPA. The particular blowers included on the
list are being marketed specifically for radon mitigation. All
of the data in the table were  provided by the manufacturer
(Ra92, Zu92).

Because these blowers are relatively new  on the market, they
have been less widely demonstrated in the field than have the
tubular fans. Thus, at this time, it is not possible to rigorously
compare the actual radon reductions achieved in the field
compared to reductions with the 90-watt fans, or to rigorously
compare the noise levels associated with the blowers vs. the
fans. Nor can anything be said regarding the durability of the
radial blowers. However, the manufacturer claims that a large
number of the radial blowers have been successfully installed
to date, and that they are about as quiet as the tubular fans
(Zu92).

The GP201 has a capital cost (according to 1993 prices) about
the same as do the 90-watt tubular fans in Section  4.4.1
(generally a little below $100). The GP201 also draws about
the same amount of power (95 watts maximum, compared to
90 watts maximum for the tubular fan), so that it should have
a similar cost for electricity. The other two blowers in Table 2
have a somewhat higher capital cost (about $110 to $160)
than do the tubular fans; they also draw more power (110 to
140 watts maximum), thus suggesting a somewhat higher cost
for electricity.  The radial blowers might be considered when
any increase in installation and  operating  costs would be
offset by cost savings resulting from reductions in the number
of suction points that  would otherwise be needed with the
90-watt tubular fans.

Because these blowers are designed for in-line mounting, they
can be mounted on the piping in the same manner as the
in-line tubular fans discussed in  Section 4.4.1. The blower
couplings are 3 in. diameter; they would thus be mounted in 3
in. piping, or, if 4-in. piping is used, would require a 4- to 3-in.
adapter coupling.

4.4.3  High-Suction/Low-Flow Blowers

The 90-watt centrifugal in-line tubular fans have sometimes
given fair radon reductions in houses having poor sub-slab
communication and low flow (10-20 cfm or less) when there
have been sufficient SSD suction points, although the fans
were  operating at the low-flow  end of their performance
curve. Better reduction performance (or comparable reduc-
tions with fewer suction points) might be achieved by using
the in-line radial blowers, if these prove successful hi practice.
The in-line radial blowers will provide higher suctions, but,
like the tubular fans, are designed to operate routinely at flows
higher than those seen in true poor-communication houses.

Where flows are in fact as low as 10 to 20 cfm and less, better
performance (with fewer suction  points) and longer fan life
Table 2.   Performance Characteristics of Some In-Line Radial Blowers Being Marketed for Radon Reduction Systems
Brand
Name
DynaVac


Fan
I,D
GP201
GP301
GP501
Max.
Watts
95
110
140
Fitting
Diameter
(in.)
3
3
3
RPM
3000
3000
3000
Flow (cfm)
@ 0 in. WG @
-130
-140
-140

1 in. WG
77
92
95
Max.
Suction
(in. WG)
2.0
2.6
>4.0
                                                       123

-------
might be achieved by using high-suction/low-flow blowers,
designed to operate routinely at these very low flows. The
very high suctions in these high-suction/low-flow units (18 to
50 in. WG maximum) might extend more effectively beneath
the slab with a reduced number of suction points, compared to
the 1.5+ in. WG maximum achievable with the 90-watt tubu-
lar fans and the 2.5-4 in. WG maximum with the in-line radial
blowers.

The performance characteristics of two brands of high-suction/
low-flow blowers are summarized in  Table 3, based upon
manufacturer literature (Pe90, Ra92). This list is intended for
illustration of the types of fans available; it is not intended as
a comprehensive  compilation  of blowers suitable for  this
application,  nor does inclusion  of a fan on this list signify
endorsement by EPA. The particular blowers included on the
list are being marketed specifically for radon mitigation, and
have reportedly been soundproofed for acceptably quiet op-
eration.

The Pelican S-3 blower is a centrifugal regenerative blower,
specifically configured and soundproofed for residential ra-
don mitigation use. Although it is being used for mitigation
installations being made  by its manufacturer, it is not cur-
rently being marketed for use by other mitigators. TheDynaVac
units are high-speed centrifugal blowers.

The radon reduction performance of these blowers has not
been extensively compared against that of a 90-watt tubular
fan (or a 110- to  140-watt in-line radial blower) in a given
house. Some mitigators working in poor-communication houses
report that the use of one of these high-suction blowers has
reduced the required number of suction points by one or more,
compared to use of a standard fan (Bro92, K192). However, it
is not possible at this time to provide  specific guidance
regarding the benefits of using  one of these blowers rather
than one of the other units under specific conditions.

It is re-emphasized that  these  high-suction/low-flow units
generally cannot handle flows above the low end of the SSD
flow range. As shown in Table 3, all but one of the units falls
to zero suction once the flow increases to about 25-50 cfm. As
a result, if one of these blowers is installed in a system initially
having a very low flow, and if the flow subsequently increases
after installation, for example, due to new openings created in
the slab or due to a drying and increased permeability in the
underlying soil, these low-flow units could be overwhelmed,
and the radon reduction performance could be seriously de-
graded.

The  manufacturers  claim  that these high-suctionAow-flow
blowers are acceptably quiet if mounted properly to avoid
vibration, although exact figures are not available to compare
against the 2-4 sones  for the tubular fans. The units are
enclosed within a soundproofed casing. In practice, some
mitigators report that with one or two of the blowers in Table
3, exhaust mufflers and careful siting of  the fan can be
required to reduce the blower noise to levels that homeowners
do not find objectionable.

None of these units can be supported by the PVC piping; both
must be mounted  on (or suspended from) rafters, joists, the
side of the house, or other members. They are thus somewhat
less convenient than the in-line tubular fans and in-line radial
blowers, but the piping can be readily designed to account for
the configuration of these units.

A main disadvantage of .the high-suction/low-flow blowers is
that they are significantly more expensive than the tubular
fans. An individual Pelican  unit retailed for $695 in  1991
(Pe90), compared to about $100 for the 90-watt tubular fan;
individual DynaVac blowers retail for $630-$660 in  1992
(Ra92). At near-zero flow, these blowers would draw a maxi-
mum  of 200-300 watts, although, at higher flows, the power
consumption  would  be well below that. This maximum con-
sumption is roughly 100-200 watts more than the maximum
consumption of the 90-watt tubular fan, which would increase
the annual cost of electricity to run the fan  by $70-$ 140, if
electricity costs $0.08 per kWh. These blowers would most
often be considered in cases where these increased installation
and operating costs are offset by savings resulting from reduc-
tions in the number of suction points required.

Because there has been less experience with these high-suction
fans, there is less information on their durability.

In addition to the two brands of high-suction fans listed above,
there are a variety of industrial blowers on the market that can
operate at low flows and generate high suctions. But these
other blowers have not been  designed with residential radon
mitigation applications in mind. Of particular concern, many
are too noisy for  residential applications. In addition, their
fittings may not be  configured for easy connection to PVC
piping, and the fan  casing itself may not be configured for
Table 3.   Performance Characteristics of Some High-Suction/Low-Flow Blowers Suitable for Radon Reduction Systems
Fan
Manufacturer
Pelican Environmental Corp
DynaVac


Fan
1. D.
S-3
HS2000
HS3000
HS5000
Max.
Watts
210
270
195
320

@ 0 in. WG
27
110
44
53
Flow (cfm)
@ 10 in. WG
22
72
37
47
Max. Suction
@0cfm
(in. WG)
26
18
34
50
                                                       124

-------
convenient mounting in a house. The housings around some
of these blowers may not be leak-tight, so that caulking of the
housing leaks would be required. Like the fans listed above,
these other blowers have a higher purchase price and a higher
power requirement (150-250 watts) than do the tubular fans.

4.5    Installation of Suction Pipes
Beneath the Slab

4.5.1  Vertical Pipe Installed Down
Through Drilled Hole Indoors

Probably the most common method of installing SSD suction
pipes is to drill a 4.5- to 6-in. diameter hole through the
basement slab or the slab on grade inside the house, and to
mount the pipe vertically through that hole. This method is
illustrated genericalry in Figure 1, although there are a variety
of acceptable methods for actually installing the pipe in the
hole, as discussed later.

Selecting the specific hole location. The general location
of each suction hole will have been selected during  system
design based upon a number of criteria, described in Section
4.2. During the actual installation, described  here, a more
specific identification will be required of exactly where the
hole is to be drilled.

Several factors will contribute to this selection. For one thing,
the hole will have to be far enough from the foundation wall
and any other obstructions (at least 6 in.) to avoid the footing
(which commonly extends out 4 in.)  and to provide the
workers with sufficient clearance to operate the drilling equip-
ment. In addition, sometimes a SSD suction pipe in a base-
ment is to be connected by a horizontal run to another suction
pipe, or is to be routed to a remote point where it will exit from
the basement In these cases, if the basement ceiling is unfin-
ished,  it will often  be optimal to locate the suction pipe
directly beneath the space between two joists supporting the
floor above, so that the pipe can easily be connected to the
horizontal run between the joists. Or, if there is a suspended
ceiling, the mitigator may wish to locate the pipe centrally
under a ceiling panel for easy penetration of the ceiling. If the
pipe will extend straight up through the ceiling as part of an
interior stack, the mitigator may wish to line up the slab hole
with the hole in the ceiling above; finish on the floor above, or
obstructions in the attic, might influence the precise location
of the ceiling hole.

Since the slab hole is often an inch or two larger in diameter
than is the pipe, there will be some flexibility for shifting the
pipe an inch or so one way or another, in order to get the pipe
straight and precisely lined up with any overhead connections.

In selecting the specific hole location,  the installers should
also consider some of the criteria that should have been used
to select the general location during system design. In particu-
lar, the installers should be alert to the possible location of
sub-slab (on in-slab) utility lines, and to the availability of a
convenient route for connecting  the suction pipe to other
suction pipes and to the exhaust point.
Drilling the hole through the slab. The hole that is made
through the slab will usually be between 4.5 and 6 in. in
diameter when the 4-in. piping (which is  4.5  in. outside
diameter) is being used. The desired diameter of the slab hole
for 4-in. piping will depend upon how the suction pipe is to be
mounted in the slab, as discussed later. When smaller-diameter
piping is to be used, the slab hole could nominally be smaller,
although it will still have to be large enough to accommodate
a worker's hand or an auger if a pit is to be excavated beneath
the slab, as discussed later.

To make the hole through the slab, many mitigators use a
rotary hammer drill, an electric power drill which includes a
hammering motion. The rotary hammer drill is used to drill
holes, typically 1/2 in. diameter, through the slab around the
circumference  of the desired suction hole. A chisel bit is then
used to break out the concrete in the center.

Another option for drilling the slab holes is a coring drill with
a 4.5- to 6-in. bit. A coring drill uses a hollow bit the size of
the desired hole to remove a plug of concrete of that diameter,
leaving  a slab  hole having a smooth perimeter. Coring drills
can be operated dry using a carbide bit or wet using a diamond
bit. With wet drills, where water is used to cool the bit during
drilling, a  dike must be used to capture the water during
drilling. Dikes have been fabricated out of sand (Sc88), tem-
pered masonite (Ba92), and 1/8-in. thick flexible plastic sheet-
ing (Ba92). Also, a wet vacuum cleaner or some other ap-
proach must be used to remove the resulting slurry. Although
wet coring drills are quieter than dry drills, the mess associ-
ated with wet drills can preclude their use in heavily finished
areas.

Compared  to the coring drill approach, the rotary hammer
approach takes longer and leaves a hole having a more ragged
perimeter. However, coring drills are much larger and more
expensive than rotary hammers; two operators  may be re-
quired; and, with the wet coring drills, care must be taken to
capture  and dispose of the slurry. Accordingly, many mitiga-
tors commonly use rotary hammers ,for residential applica-
tions, and rent  a coring drill for those cases where many holes
will be required (e.g., in the large slabs often encountered in
schools and large buildings). Mitigators installing large num-
bers of systems may find it cost-effective to purchase a coring
drill directly.

It is recommended that the installers attempt to control the
drill so that it penetrates no deeper than the bottom of the slab,
insofar as possible. The concern is to avoid damage to utility
lines  that may unexpectedly be beneath the slab despite the
efforts that were made to locate the hole where no lines would
be present. (Stopping the drill at the bottom of the slab would
not avoid damage to utilities embedded in the slab.) If a rotary
hammer drill is being used, the mitigator may wish to probe
carefully beneath the slab with the drill bit after making the
first  holes around the  core circumference,  pushing the bit
somewhat deeper, to try to identify the presence and nature of
any sub-slab obstructions (such as a utility line or a footing)
before drilling further  holes. After the concrete core is re-
moved, the fill should be inspected. If this inspection shows
that  no lines  are present and if the sub-slab material  is
                                                        125

-------
hard-packed, the drill might then be used to penetrate deeper
into the hard sub-slab material, to aid in excavation of a pit
should be ventilated, to reduce the concentration of dust and
radon.
Another option to avoid damage to sub-slab or in-slab utility
lines is a safety cut-off switch that has been developed by one
mitigator  (Bro92). This switch would open, and shut off
power to the drill, if the drill bit touches anything metal in or
beneath the slab, such as metal water lines, gas lines, and
electrical conduits. The circuit for such a safety switch could
be assembled in a box;  the power cord from the drill would
plug into the box, and a cord from the box would plug into an
electrical outlet in the house. Two leads would extend from
this box; one would be clipped to the drill, and the other would
be clipped to a water pipe or other metal pipe extending
through the slab,  creating a low-voltage circuit Since all
metal utilities effectively interconnect, connecting  the one
clip to any accessible metal line entering the sub-slab will
effectively include all metal utilities in this circuit If the drill
bit touches any metal line in or under the slab, the low-voltage
circuit will be closed, activating a relay which will open a
switch in the wiring providing power to the drill, thus shutting
off the drill.

It is recommended that a visual indicator (such as a light bulb)
also be included in the  low-voltage circuit, so that the light
comes on when the drill bit touches metal (Bro92). This light
may flash before the relay switch is activated, alerting the
workers that a poblern  exists before damage is done to the
utility line.

Mitigators would have to construct this cut-off switch them-
selves. A circuit diagram for this switch is included in Refer-
ence EPA89c. This safety cut-off switch would not work if the
sub-slab utilities were water or sewer lines made of plastic.

Some mitigators also use a ground fault interrupter (GFI) in
the power cord to the drill, to protect the workers. The GFI
would shut off power to the drill if there were a short-circuit in
the drill. Thus, a GFI could be helpful when working under
wet conditions, when short-circuits might be created by the
presence of water. The GFI would also shut off the drill if the
drill cut into a sub-slab electrical conduit, creating an electri-
cal surge  in the drill; in this  case, the GFI  might reduce
damage to the conduit,  but would not prevent damage alto-
gether, and would not necessarily prevent electrical shock to
the worker. For this application, GFIs are mounted on a short
length of  cord which is plugged in between the electrical
outlet and the drill.

Except  with  the wet coring drill, some dust is generated
during the drilling process. Continuous operation of a vacuum
cleaner immediately beside the drill bit is generally helpful in
reducing the amount of dust in the air around the installers,
and the amount deposited in the house. The vacuum cleaner
should exhaust outdoors; vacuums depending upon bags to
capture the dust should be located outdoors. If a dry coring
drill is being used in a finished area,  temporary dust curtains
have  been suggested to isolate the  drilling  area from the
finished part of the house, although many mitigators find this
unnecessary. In cases where the drilling process generates
significant amounts of dust and noise, the workers should be
provided with respiratory and ear protection. The work area
Excavating a pit beneath the slab. As discussed in Section
2.3.1, excavation of a pit directly under the slab hole will
often aid in the distribution of the suction field beneath the
slab.  The pit will reduce  the suction loss  encountered as
sub-slab gases remote from the pipe (initially moving at only
a few feet per minute or less)  accelerate to pipe velocity
(perhaps 100 to 1000 fl/min). The loss is much less if this
acceleration occurs in the free space of a pit rather than in the
pore space within soil or gravel.

Stated in another way,  the pit significantly increases the
surface area of the exposed soil  face through which the soil
gas flows. Increasing this surface area reduces the maximum
velocity of the soil gas through the soil pores, and hence the
suction loss. To illustrate this point, with a pit of sufficient
size,  the gas beneath  the slab will  never be moving at a
velocity as high as it will be after it enters the suction pipe.
But if the pipe were resting directly on the soil, so that the
exposed soil face were equal to the cross-sectional area of the
pipe, the gas under  the slab near  the pipe entrance would
necessarily have to be moving at a greater velocity than that in
the pipe, since the void space in the soil is much less than the
open area in the pipe.

Another benefit of excavating a pit is that the pit will poten-
tially intersect fissures or permeable strata at some distance
from the suction hole,  or at some depth beneath the slab. In
addition, the pit will help prevent sub-slab water from block-
ing the suction pipe.

From a practical standpoint, such pits are probably not really
necessary when there is  an aggregate layer under the  slab.
Because of the good  communication  in such houses, the
system could still perform well despite sustaining the in-
creased suction loss resulting from the lack of a pit. In
addition, the suction loss might be less in the relatively large
pores between pieces of aggregate (compared to the loss that
would take place in the tiny pores between hard-packed soil
particles). However, pits are usually relatively easy to dig
when aggregate is present, to the extent that some mitigators
are not able to separate out the labor hours/costs required to
dig the pit from the total labor/costs required to drill the hole
(He91b, He91c).

Thus, it is  recommended  that a pit always be excavated
beneath the cored hole. Where there is good aggregate, this is
done because it is easy to do, it might help, and there is no
reason not to do it When there is good aggregate, a relatively
small pit will probably be sufficient. Where there is not good
aggregate, the pit is excavated because it can be important in
improving the performance of the system; in this case, a larger
pit may be warranted. In general,  the more packed the soil
(and the harder it is to excavate  the pit), the more important
die pit probably is.

The pit is often excavated largely by hand, by reaching
through the 4.5- to 6-in. hole in the slab. If it is clear that there
are no utility lines buried in the soil under the hole, the rotary
hammer drill or coring drill that was used to make the hole
                                                         126

-------
through the slab might be used to help break up hard-packed
soil immediately under the slab, to simplify the excavation. A
variety of power and hand tools can also be used to loosen the
hard-packed soil. Some mitigators insert an auger or some
similar tool through the slab hole to break up the soil. These
augers can be driven by a power drill, possibly fitted to drill at
a right angle. The shaft on the auger can be about 2 ft long.
Many mitigators find it convenient to use the industrial vacuum
cleaner to remove the loosened soil through the slab opening,
rather than digging it out with a trowel.

A hole roughly 1 to 3 ft in diameter and 4 to 18 in. deep can be
reasonably excavated in this manner. Thus, a pit as large  as
about 10-14 ft3 (about 75-100 gal) could nominally be dug.
This is much larger than is necessary based on fluid flow
considerations  (Br92), and pits  are significantly smaller  in
most cases (commonly 1-3 ft3, or roughly 5-25 gal). While the
appropriate size and shape of the pit will depend upon the
particular house and geological conditions, some mitigators
report often finding  it preferable to increase pit diameter
rather than to increase depth (Bro92).

Usually, this pit is left as an empty void. In some cases where
there is no aggregate under the slab, some mitigators suggest
placing some gravel in the bottom of the pit to help prevent
the sides from caving in  as  a  result of sub-slab moisture
(Ba92):

Because of the high radon concentrations that can exist under
the slab, the worker digging through the open slab hole may
sometimes need to wear a respirator, depending upon sub-slab
levels and the  time it takes to excavate the pit. Radon mea-
surements at the slab by one mitigator in a region having only
moderate sub-slab radon concentrations and good aggregate
(so that excavating a sub-slab  pit took only 10 minutes),
suggested that no significant additional radon exposure oc-
curred during pit excavation (Mes92). If the vacuum cleaner is
used to remove dirt, it should be exhausted outdoors.

Mounting suction pipes through slab. A variety of ac-
ceptable methods exist for mounting the suction pipe in the
slab. Figure 14 illustrates several methods for the case where
the weight of the pipe is not being supported at the slab;  in
these cases, the piping will have to be supported overhead in
some manner,  as discussed below. Figure 15 shows several
methods for mounting the suction pipe in cases where the pipe
weight is being supported at its slab penetration.

In essentially all cases, the bottom of the pipe ends somewhat
above the underside  of the slab, or roughly  even with the
underside. This  is to prevent the pipe opening becoming
plugged with soil over time if the pit caves in, or blocked by
water if the pit floods during wet periods. The pipe must not
rest on the bottom of the pit; this would restrict air flow into
the pipe, and could result in the pipe becoming blocked by
cave-ins and water. In the one example shown where the pipe
is supported by the bottom of the pit (Figure 15E), notches
which extend up to the underside of the slab have been cut in
the pipe to avoid these problems (An92).

Another common feature in all of these options is that the gap
between the outer circumference of the pipe and the concrete
        Suction
        Pipe
                                    Flowable
                                    Urethane Caulk
                      5" to 6"
                      OVp.)
      A) Use of backer rod and f lowable caulk.
             Note: When using expanding
             foam, care must be taken
             that the pipe opening does
             not become blocked
        Pipe
Flowable
Urethane Caulk

Expanding
Urethane Foam
   B) Use of expanding foam and f lowable caulk.
             Note: Mortar provides
             additional support
             against lateral
             movement of the pipe.
                                    Flowable
                                    Urethane Caulk
       Suction
       Pipe —
       Slab
! Mortar

Backer Rod or
Expanding Foam
      C) Use of mortar and f lowable caulk.
Figure 14.  Alternative approaches for installing the SSD suction pipe
          in the slab, in cases where the weight of the suction pipe
          is not supported at the slab penetration.
                                                       127

-------
Non-Flowable (or
Ftowable) Urethane
Caulk Forced into
Irregularities
              Slab
                          41/2"
                                     Suction Pipe
                                     (41/2"O.D.)
            A) Drill slab hole exactly 41/2" dia.,
               so that outer diameter of pipe
               lodges tightly against irregular
               sides of hole.
                   Urethane Caulk
                   Forced into
                   Irregularities
                            Slab
                                          5"
Suction Pipe
(41/2" O.D.)

Straight Fitting
(5" O.D.)
                            B) Drill slab hole exactly 5" dia.,
                               so that outer diameter of
                               straight fitting lodges tightly
                               against irregular sides of hole.
  Non-Flowable
  Urethane Caulk
            Slab
                          5-6"
Suction Pipe

PVC Flange
           C) Drill slab hole 5-6" dia., mount
              suction pipe in PVC flange, so
              that lip of flange rests on top
              of slab.
                                                        Suction Pipe


                                                   Non-Flowable
                                                   Urethane Caulk
                            Slab
  6X4" Dia, PVC
  Bushing
  (Available in
  Schedule 40)
                           D) Drill slab hole about 6" dia.,
                              mount suction pipe in 6"X4" PVC
                              bushing, so that lip on 6" end
                              of bushing rests oh top of slab.
 Flowable
 Urethane Caulk

Backer Rod or
Urethane Foam
            E) Suction pipe allowed to rest on
               bottom of pit.
                                                    Notch Cut in
                                                    Sides of Pipe
                                                    Near Bottom
Rguro 15.  Alternative approaches for installing the SSD suction pipe in the slab, in cases where the weight of the suction pipe is supported at
          the slab penetration.
                                                      128

-------
must be effectively caulked closed. Since this slab opening is
right at the point where suction is being drawn, this opening
would have a particularly detrimental effect on suction field
distribution if left unsealed. Urethane caulks are recommended
for this application, because they bond better to the concrete
than silicone caulks. If the gap around the perimeter is fairly
narrow (e.g., less than about 1/4 in.  wide),  a gun-grade
(non-flowable) polyurethane caulk is probably the best choice;
this caulk should be forced down into the gap, in order to
maximize the pipe and concrete surface area that it contacts,
to improve bonding.  If the gap is wider than 1/4 in.,  it will
likely be necessary to force some appropriate material (such
as backer rod) down into the gap (see Section 4.7) to provide
support for the caulk. In this case, a flowable urethane caulk
can be used; the flowable caulks work their way into irregu-
larities in the surfaces, providing a better seal.

In the details in Figure 14, the caulk, foam, or mortar must be
applied after the entire piping network is in place. Otherwise,
the vertical pipe may shift somewhat during the remainder of
the installation process, causing the seal to be broken at the
slab.

Figure 14A shows one common approach when the piping is
not supported at the slab (Bro92). This is the specific method
illustrated in Figure  1. The 4.5-in. O.D. suction pipe is in-
serted  into a 5- to 6-in. diameter slab hole. Backer rod (a
closed-cell foam which is sold in rope form) is stuffed down
into the gap between the pipe and the slab, and the channel
above the backer rod is flooded with flowable urethane to a
depth of less  than 1 in.  (consistent with  the  instructions
provided by the caulk manufacturers). One possible disadvan-
tage of this approach is that the caulk and backer rod do not
provide much support against lateral movement of the pipe.

Figure 14B is similar to 14A, except that the support for the
flowable caulk is now provided by expandable foam rather
than backer rod (An92). The  foam may provide some addi-
tional  support against lateral movement. In this case,  the
bottom of the pipe is shown slightly below the underside of
the slab, in an effort to avoid plugging of the  pipe by the
expanding foam.

Figure 14C shows a method for providing  further support
against lateral movement (An92). In this case, a layer of
mortar is placed between the foam and the caulk. Mitigators
using this technique report that the flowable caulk can be
applied directly on top of the wet mortar, without waiting for
the mortar to set (An92). Applying the caulk while the mortar
is still wet prevents the caulk from adhering to the mortar, but
apparently does not prevent adequate bonding to the side of
the pipe and to the concrete slab.

A variety of approaches can  be used to support the  piping
weight at the slab penetration, some of which are shown in
Figure 15. One method is to drill the slab hole to be just about
equal to the 4.5-in.  O.D.  of  the suction pipe, as shown in
Figure 15A (Mes92); the pipe is forced down into the  hole
until it lodges tightly against the sides of the hole. This is
possible in part because the face of the concrete inside the
hole is irregular, if the hole was prepared using a rotary
hammer drill. A  similar approach is shown in Figure  15B
(Mes92); however, in this case, a straight fitting is cemented
onto the end of the pipe, so that the hole diameter can now be
5 in. instead of 4.5 hi. (somewhat facilitating the excavation of
a sub-slab pit).  In both of these cases, the residual  gap
between the pipe and the slab will probably have to be caulked
with non-flowable urethane caulk, unless the fit is indeed tight
enough that flowable caulk can be used.

Another method for supporting the suction pipe is illustrated
in Figure 15C (Fo90). In this case, the end of the suction pipe
is cemented into a PVC flange having an inside diameter of 4
in. The outer diameter of the flange, about 5 in., is slightly
smaller than the slab hole. The lip of the flange, about 6 in. in
diameter, rests on the top of the slab as shown in the figure,
supporting the pipe. A non-flowable caulk must be forced into
the gap between the outer circumference of the flange and the
concrete, and  a  good bead of the caulk should be placed
between the lip of the flange and the top of the slab, before the
flange is pressed down into the hole. The suction pipe should
be effectively cemented to the interior of the flange, to prevent
leaks at the seam between the pipe and the flange.

Another effective approach is to use a 6- by 4-in. bushing, as
shown in Figure 15D (K192), a fitting which is available in
Schedule 40. At one end, this bushing is designed to fit around
the 4.5-in. O.D. of a 4-in. suction pipe; at this end, the bushing
itself has an O.D. of about 5 in. At the other end, the bushing
is designed to fit inside a 6-in. pipe; the main extension of the
bushing will thus be about 6 in. O.D. at this end, but it has a lip
which is about 6.5 in. diameter, equivalent to the O.D. of 6-in.
pipe. The 4-in. suction pipe is cemented into the smaller end
of the bushing; the 6-in. end of the bushing is placed down
into  the slab hole, which should be about 6 in. diameter; and
the 6.5-in. wide lip rests on the top of the slab, supporting the
piping. Non-flowable urethane caulk is  forced between the
sides of the bushing and the hole, and between the underside
of the Up and the top of the slab.

In Figure 15E, the weight of the pipe is supported by allowing
the pipe to rest against the bottom of the pit (An92). To avoid
flow restriction and to avoid pluggage of the pipe with soil or
water, notches are cut in the side of the pipe, extending at least
as high as the underside of the slab. In the figure, the gap
between the pipe and the slab is sealed with backer rod and
flowable urethane caulk.

Another option  that has been used (Mes92) is to partially
embed one or more large masonry nails (spikes) laterally into
the side-wall of the slab hole, at a depth perhaps halfway
between the top and underside of the slab. A portion of each
spike extends into the hole opening. The pipe then rests oh the
exposed portions of the spikes.

In the cases shown in Figure 14,  where the  weight of the
piping is not supported at the slab, some means for supporting
the pipe will be required to prevent it from dropping all the
way down to the bottom of the hole. This will be done using
hangers and/or strapping, or perhaps pipe clamps. The exact
manner in which hangers, strapping, and/or clamps will be
used to support the suction pipe will depend upon the configu-
ration of the remainder of the piping network.
                                                        129

-------
Discussion of the configuration and support of the overall
piping network is presented in Section 4.6.3. The discussion
here addresses only the immediate support of the individual
vertical suction pipe, in cases where it is not supported at the
slab penetration.

Figure 16 illustrates some alternative methods for supporting
the suction pipe in the case where the vertical suction pipe
connects to a horizontal piping run. There are two categories
considered for this case: a) the horizontal run is perpendicular
to the overhead floor joists, below the joists; and b) the run is
parallel to the joists, up between the joists.

Figure 17 shows some alternative methods for supporting the
suction pipe in the case where the pipe extends straight
upward through the overhead flooring.

In all cases, the support is provided in a manner that attempts
to reduce or avoid the transmission of vibration noise from the
pipe to the joists or flooring.  In cases where  there is a
horizontal run of piping, supports should be provided every 4
to 10 ft along the run; one of these supports should be near the
vertical pipe.

One common support material is plastic strapping, perhaps 1
in. width. Figure 16A shows two possible methods for using
strapping to support a horizontal pipe running below the joist.
The ends of the strap are nailed into  the joist. It is important
that the pipe be suspended a short distance below the joist, as
shown, rather than being held against  the joist, to avoid
transmission  of vibration. Vibration  is not transmitted well
through the strapping itself. Where the pipe is parallel to and
up between the joists, the upper inset in Figure  16B shows a
common way to suspend the piping  between the joists. The
rectangle labelled "Strapping loop" in the lower diagram of
Figure 16B is attempting to depict such strapping suspending
the pipe between the joists. (Such a suspending strapping loop
is also what is being depicted by the rectangle labelled "Strap-
ping" in Figure 1 and subsequent figures.)

As shown in Figure 17A, strapping can also be used to support
vertical pipes. However, in this case, the strapping must be
attached  to the pipe by screws in order to prevent the pipe
from slipping down through  the  strapping. (When screws
must be inserted into pipes, it may be advisable to use Sched-
ule 40 piping, since the thin-walled pipe may be prone to
crack.)

For horizontal runs, hangers can also be used to support the
pipe, as illustrated in Figures 16A and 16B. One common type
of hanger is a plastic J-hook, shown in two of the insets. The
method of using the hook will depend upon whether the pipe
3s parallel or perpendicular to the joists.

Where the pipe extends  straight up through  the overhead
flooring, the weight of the piping  can  also  be supported
against the flooring. As shown  in Figure 17B, some mitigators
cement a straight fitting, or pipe coupling, into the pipe at just
the right position; the Up of the fitting rests on top of the
overhead floor, providing the support. Alternatively, as shown
in Figure 17C, a two-piece pipe clamp can be clamped tightly
around the pipe at that location, resting on the flooring. In this
case, it may be desirable to place some type of cushion (such
as closed-cell insulation) between the clamp and the floor to
reduce the transmission of vibration noise to the flooring.

Figures  16 and 17 show the suction pipe extending straight
upward from the slab. Where the suction pipe is installed near
a perimeter wall, another option used by some mitigators for
supporting  the pipe involves offsetting the pipe so that the
riser is flush against the wall (Bro92, K192). To accomplish
this, a vertical pipe stub about a foot long would be installed in
the slab, and two 45° bends would be attached to this stub to
provide the desired offset.  In this case, the pipe would be
supported by attachment against the wall, using one of the
approaches described in Section 4.6.5 for attaching an exterior
stack against the  side of a  house (see Exhaust piping from
exterior fans - support of stack against house).

4.5.2  Vertical Pipe Installed Down
Through Large Hole Indoors (Large
Sub-Slab Pit)
When a 4.5- to 6-in. diameter hole is cored through the slab, as
described in Section 4.5.1, the size of the sub-slab pit that can
be excavated is determined by the length of an arm, or the
length of an auger that can be inserted through the hole. The
largest feasible pit excavated through a slab hole would thus
be roughly 18 in. radius and 18 in. deep, or about 10-14 ft3
maximum (about 75-100 gal). Many pits are probably much
smaller than this,  due to the practical difficulties in carrying
out the excavation.

In terms of reducing suction loss due to acceleration of  soil
gas up to pipe velocity, a 10- to 15-gal pit is probably larger
than is necessary, even for the highest pipe velocities, based
on fluid flow calculations (Br92). Thus, from this standpoint,
there is  no  incentive to make the pit any larger. However, a
larger pit would potentially help intersect fissures or perme-
able strata more remote from the suction hole. On this basis,
some investigators have postulated that a larger pit might be
useful  to help  extend  the suction  field  in  very
poor-communication houses.

Excavation of a larger pit would require that a larger hole be
made through the slab, using a jackhammer or some other
means. From a practical and structural standpoint, the largest
such slab hole that might be considered would be about 3 ft by
3 ft. This hole size would permit the excavation to extend to a
greater radius than 18 in. and a greater depth. Such a large
hole would increase the cost of the installation, by about $200
per hole (He91b, He91c). With slabs that are tiled or other-
wise finished, it would increase the aesthetic impact as well.

The data presented in Section 2.3.1 show that increases in pit
size can sometimes, although not always, improve the mea-
sured suction field extension in houses having poor sub-slab
communication. Most of the pits studied were smaller than
those that would be achieved by creating a 3 x 3-ft hole in the
slab, preventing a rigorous  assessment of the practical ben-
efits of making a pit  that large. The largest pit studied in
Reference Py90 (24 in. square and 18  in. deep)  did  not
significantly increase suction field extension relative to smaller
                                                       130

-------
       Floor
                  Joist
         rT~
Plastic   V^
J-Hook-/
                               Hanging
                               Clamp
                    Continuous Loop
                    of Plastic Strapping
                    Around Pipe
         Overhead
         Flooring —
90° Elbow Fitting
Tightly Cemented
to Adjoining Piping
          Floor
          Joist
              Suction
              Pipe —
             Sealant
                     ~\
Pipe Hanger Attached to
Overhead Floor Joist,
Firmly Supporting
Vertical Suction Pipe.
(Repeated Every 4-10 Ft.
Along Horizontal Piping
Run.) See Insets Above.
                A) Horizontal piping run is
                   perpendicular to floor
                   joists.
                                                                       Floor
                                 Joist
                                                                    Continuous Loop of
                                                                    Plastic Strapping
                                    Bracket
                                                                                     Plastic
                                                                                     J-Hook
                                                                Overhead
                                                                Flooring —
                                                                 Floor
                                                                 Joist
                                                               Suction
                                                               Pipe 	
                                   Sealant
                                                                                      _r
                                                                                  <•— Strapping Loop
                                                                                     Attached to the
                                                                                     Two Adjoining
                                                                                     Floor Joists.
                                                                                     See Inset Above.
                                    B) Horizontal piping run is
                                       parallel to the floor joists.
Figure 16.  Some alternative approaches for supporting the weight of a suction pipe when there is a horizontal piping run, in cases where the
          pipe 'is not supported at the slab penetration.
                                                      131

-------
 Straight Fitting,
 to Join Two
 Sections of Piping
Overhead
Flooring-
                                  Hole Through
                                  Overhead Flooring
                                  Accommodates Pipe
 Floor
 Joist
Continuous Loop/
of Plastic
Strapping
                        \
         Sealant
                 ~\
                                  Screws to Prevent
                                  Vertical Movement
                                  of Pipe
                                  Suction
                                  Pipe
             A) Pipe supported by loop of
                strapping from floor Joists.
                                                        Straight Fitting,
                                                        with Bottom Lip
                                                        Resting on Flooring •
B) Pipe supported by lip of
   straight fitting resting
   on top of overhead floor.
                                                                                              Clamp
                                                                                              (Two Pieces)

                                                                                              Bolt
                                                                                Top View
                                                          Clamp Around Pipe,
                                                          Resting on Overhead
                                                          Flooring, so that
                                                          Weight of Pipe Is
                                                          Borne by Floor
                                                                                           Cushion
                                                                                           J
                                                                    C) Pipe supported by 2-piece
                                                                       clamp resting on top of
                                                                       overhead floor.
Figure 17.  Some alternative approaches for supporting the weight of a suction pipe when the pipe rises directly up through the overhead
          fioorina with no horizontal run.
                                                     132

-------
pits (20 in. square by 16 in. deep) of the size that might be
excavated through a 6-in. diameter slab hole.

In summary, there is no experimental basis for believing that a
large pit (beyond that which can be excavated through a 6-in.
slab hole) will be helpful in improving SSD performance in
poor-communication houses. The only theoretical basis for
considering large pits in residential cases, i.e., the hope of
intersecting remote fissures or permeable strata, will be very
site-dependent; the potential for actually intersecting fissures/
strata by making a large pit has not been demonstrated.

As a  result, preparation of  a  large opening in the slab to
excavate a large pit will not commonly reflect good practice.
Consequently, this approach is addressed here only briefly,
insofar as it differs from the approach in Section 4.5.1. This
approach  is considered  here because some  mitigators may
occasionally face geological situations where it may be appli-
cable.

Making the hole through the slab. If the pit is to be left
open, rather than being  filled  with aggregate — so that the
restored slab will be suspended over an open cavity — the pit
should probably be no more than about 4 x 4 ft, or perhaps
somewhat larger, for structural reasons. (If a much larger pit is
planned, the mitigator should check with a structural expert to
ensure that the suspended slab will have adequate strength.)
The mitigator will have to decide how large the slab hole
should be to enable excavation of a pit of the desired size. A
slab hole of about 2 x 2 ft to 3 x 3 ft might be reasonable to
provide the needed access.

A jackhammer or an electro-pneumatic roto-stop hammer
would be logical tools for creating such a hole in the slab. The
jackhammer or electro-pneumatic could be rented for  this
purpose. Alternatively, if the mitigator has a coring drill, the
large slab hole could be created by drilling multiple adjacent
6-in. circular holes. It is recommended that the hole be pre-
pared with the sides at an angle (i.e., with the hole dimensions
being smaller at the bottom  of the slab than at the top). The
resulting slant in the vertical face of the old concrete will help
support the weight of the new concrete which will be poured
to restore the slab.

As discussed in Section 4.5.1, the installers should use care to
avoid rupturing any unexpected utility lines beneath the slab.
Use of a safety cut-off  switch would help avoid damage to
utilities.

Excavating a pit. Once the hole has been created in the slab,
the excavation of the pit can be accomplished using a variety
of tools. If the underlying soil is hard-packed or is rock, the
jackhammer (or other appropriate power equipment)  can be
used to break up the soil.

The pit would be excavated out under the slab to the desired
dimensions (e.g., 4 x 4 ft). If the pit is not at least 3 x 3 ft, there
is no point in making the large hole through the slab,  since a
pit approximately 3 x 3 ft can often be excavated through a
6-in.  slab hole.  If it is known that a permeable stratum
underlies  the surface soil at some reasonable depth, an effort
might be made to penetrate down to that stratum.
It is suggested that the pit be left as an open cavity when the
slab is restored. This will maximize the benefits of the pit,
providing the greatest reduction in suction loss resulting from
acceleration of the soil gas to pipe velocity through the pit.
The other option, back-filling the pit with clean, coarse aggre-
gate, would simplify the restoration of the slab, but would
somewhat increase the suction loss.

Mounting suction pipes through the slab. Some type of
plywood or sheet metal form must be prepared to permit new
concrete to be poured, to restore the slab. While a variety of
approaches are possible for preparing this form, one approach
that has been used  with some  success in this application
(Cr92a) is to prepare a piece of plywood slightly larger in
cross-section than the slab hole, with a4.5-in. hole cut gener-
ally near the center to accommodate the suction pipe.  This
plywood is then cut in half, so that it can be fit down through
the slab hole. Masonry nails or spikes are partially driven
laterally into the side-wall of the slab hole at intervals around
its perimeter. The two sections of the plywood form are then
pulled up tightly against the underside of the slab, by attach-
ing wire to the plywood and pulling this wire tight around the
protruding nails.

Another approach for supporting the plywood form would be
to cut the plywood so that it is slightly larger than the bottom
dimensions of the slab hole, but smaller than the dimensions
at the top of the slab (Bro92). The plywood will thus lodge
against the sloped sides of the slab hole near the bottom of the
slab. With this approach, the  re-poured segment of slab will
necessarily be slightly less than 4 in. thick.

The suction pipe is placed vertically down through the 4.5-in.
hole in  the plywood sheet The pipe is  inserted such that its
lower end is just below the sheet and the underside of the slab
(i.e., such that the pipe ends in the open space within the hole,
and is not resting on the bottom of the  hole).  The pipe is
supported in this suspended position by hangers, strapping, or
clamps  attached to the joists or flooring overhead, using one
of the methods illustrated in Figures 16 and 17.

Fresh concrete is then poured on top of the plywood, restoring
the slab, with the pit remaining as an open cavity underneath.
While the concrete is soft, a groove perhaps 1/2-in. deep is
tooled around the perimeter of the slab hole, where the new
concrete meets  the  original  slab; a groove is also tooled
around  the circumference of  the suction pipe, where it pen-
etrates the new concrete. When the concrete has set, these
grooves will be flooded with flowable urethane caulk, in an
effort to keep these seams gas-tight. A follow-up visit to the
house will be required to conduct this subsequent caulking
step.

Over time, the plywood or sheet metal support may decay or
rust away. This is not a problem. After the concrete covering
the hole is dry, its weight will be borne by the slanted sides of
the original slab.

As one alternative to this approach, the pit could be back-filled
with clean, coarse aggregate. The suction pipe would then
                                                        133

-------
penetrate into this aggregate, to a depth just below the under-
side of the slab. A layer of plastic sheeting, or perhaps of
roofing material, would be placed over the top of the aggre-
gate, to prevent the wet concrete from settling down into the
aggregate and plugging the pore space. The slab would then
be restored, as above, with grooves tooled to permit subse-
quent caulking.

4.5.3 Vertical Pipe Installed Into Unused
Sump Pit Indoors

On occasion, a mitigator may encounter a basement having a
sump pit having no drain tiles and never containing any water.
Such unused sump pits may be encountered, for example, in
cases where codes require installation of a sump but where the
particular lot involved has no drainage problem, so that the
builder or homeowner did not install drain tiles or a  sump
pump.

On those occasions where unused sump pits are encountered,
they may be used as ready-made holes through the slab for a
SSD system, avoiding the need for drilling a hole with a rotary
hammer or a core drill. The pit would be capped with a
gas-tight cover, and a suction pipe would be installed through
this cover, as discussed in connection with sump/DTD sys-
tems In Section 5. If drain tiles were present in the sump, this
would be a sump/DTD system; but since no tiles are present in
this case, it is considered a SSD system.

It must be ensured that the unused sump pit is communicating
with the sub-slab region. For example, if the walls and bottom
of the sump are formed by a plastic sump liner having no
penetrations, or by poured concrete, the SSD pipe would be
drawing suction on a closed box; the suction would not be
effectively extending under the slab. In such cases, it would
be necessary to ensure that a number of holes were drilled
through the walls of the sump lining, at a level immediately
beneath the slab.

Cost analysis indicates that the presence of such an unused
sump pit will not reduce the installation cost of a SSD system
(He91b, He91c). The savings in labor achieved by avoiding
the need to drill a hole through the slab will be approximately
offset by the increase in labor required in sealing a cover over
the sump. In fact, the installation cost will be somewhat
higher with the sump pit, by an amount equal to the materials
cost for the sump cover. However, where such a pit exists, it
still  will generally make sense to use the pit as  the SSD
suction hole.  Even if the suction pipe were to be installed
through the slab at another location instead, the sump pit
would still have to be capped, to prevent radon entry and to
reduce the amount of house air that would flow into the SSD
system through an uncapped sump.

4.5A Horizontal Pipe Installed Through
Foundation Wall from Outdoors

In some cases, it may be desirable to insert the suction pipes
horizontally through the foundation wall from outdoors, such
as illustrated in Figure 2. This approach will reduce the
aesthetic impact indoors, but can increase the impact outdoors
(especially when compared with interior suction pipes having
an exhaust stack rising inside the house). As discussed in
Section 4.2.1, installation of the pipes horizontally from out-
doors is most likely to be attractive when: a) the slab is near
grade level, so that less excavation is required to get below the
slab outdoors; b) interior finish is extensive, so that the cost,
complexity, and aesthetic  impact of interior pipes would be
significant; and/or c) the house is not amenable to an interior
stack with a fan in the attic (including houses having flat roofs
or cathedral ceilings with no attic), so that the aesthetic impact
of an exterior fan and stack would be unavoidable even with
interior suction pipes.

Exterior horizontal pipes would  most commonly be consid-
ered in slab-on-grade houses, or basement houses having
slabs only moderately below grade. However, cases have been
reported where exterior penetrations  have been  used in full
basement houses, due to extensive finish in the basement and
due to homeowner preference (K189).

Selecting the specific hole location. The criteria for se-
lecting the specific location for the suction hole on the exte-
rior face of the foundation wall  will be similar to those for
selecting the specific slab locations for vertical interior pipes.
There must be adequate space beside  the house to permit
excavation and use of the drilling equipment (e.g., away from
bushes, patios, and chimneys).  The location  must permit
convenient mounting of the fan and routing of the exhaust
stack. For example, the pipe should not be mounted directly
under windows through which the fan and stack would be
visible from indoors; it should be located so that the piping
will be least obtrusive outdoors.

The exact depth beneath the slab at which the hole is drilled
may require judgement. For example, if the foundation stem
wall for a slab-on-grade house is hollow block capped with a
solid L block, as shown in Figure 2,  it  could be desirable to
drill through the course of hollow block immediately below
the L block. But the hole should not be too far below the slab,
since it is desired to access any aggregate layer immediately
beneath the slab.

In basement houses, where the foundation wall generally does
not extend below  the slab, the suction pipe will usually
penetrate beneath  the  footing. The pipe can also be routed
beneath the footing in slab-on-grade houses when the footing
is very shallow (in mild climates).

As  with vertical interior pipes,  installers using  horizontal
exterior pipes must be alert to sub-slab utility lines or forced-air
ducts beside the foundation wall. If it is planned to auger some
distance horizontally beneath the slab — so that the SSD pipe
will extend beneath the slab instead  of ending immediately
beside the foundation wall ~ attention must be paid to utility
lines that might be beneath the slab within that distance.

Drilling the hole through the foundation wall. If piping
having a 4-in. inside diameter will be used, a hole 4.5 to 6 in.
diameter must be drilled horizontally through the foundation
wall of slab-on-grade houses. Some excavation  will usually
be necessary to expose the wall beneath  the slab, and working
space immediately beside the wall may thus be limited. Espe-
cially where the foundation wall is block, and is hence rela-
                                                       134

-------
lively easy to drill through, the rotary hammer drill  will
usually be the appropriate tool; a series of 1/2-in. holes would
be drilled around a 4.5- to 6-in. circumference, and the interior
chiseled out).

Even where the wall is poured concrete 8 in. thick, the rotary
hammer drill may still be preferred. To use a coring drill on
poured concrete walls, it would be necessary to clear a suffi-
cient area to accommodate the drill, and to  block  up or
otherwise support the  drill  in a horizontal  position while
drilling the hole. The coring drill bits are often only 6 in. deep;
consequently, it would be necessary to drill into the wall to a
depth of 6 in., chip out the still-attached core,  then drill the
remaining 2 in. through the 8-in. wall. As a result,  it will
usually not be convenient to try to operate a coring  drill
sideways  to prepare  the hole, even when a coring drill is
available.

If it is  desired to extend the hole  horizontally  for some
distance beneath the slab, an auger can be used to carefully
extend the hole horizontally through the sub-slab fill. If the
hole is to extend several feet, powerful augers can be rented or
purchased. For shorter distances (2 ft or less), a smaller auger
bit that can be driven by a power drill could be used (the same
auger bit  that can be used to excavate  sub-slab pits).  The
smaller auger bit could be driven by a right-angle drill if there
is not sufficient clearance to accommodate the shaft  of the
auger straight  on. The  drill  must have a torque adjustment
feature, so that the drill  does not begin spinning and injure the
operator if the auger binds.

Such augering  has  been found to  be desirable in
poor-communication slab-on-grade houses with block foun-
dations in Florida (Fo90, Fo92) and in Arizona and southern
California (K192). In those houses, the apparent problem was
air leakage into the system through the block foundation walls
when the suction point was near the wall.  Such leakage
appears clearly to be less of a problem in good-communication
houses and in houses with poured concrete foundations. Even
in some poor-communication slabs on grade with block foun-
dations, it will sometimes be possible avoid  the need for
augering, based upon the experience in New Mexico (Tu91b).

If augering is performed, and if the hole through the founda-
tion wall is not immediately below the slab, e.g., if it has been
lowered in order to miss a course of solid L blocks, care must
be taken that the augered hole be horizontal. If the hole is
angled upwards, toward the bottom of the slab, a low spot will
be created in the piping near the exterior face of the founda-
tion wall.  Condensed moisture could accumulate at that point,
potentially blocking flow.

To avoid damaging sub-slab utility lines, it would be desirable
to initially drill only to the interior face of the foundation wall.
Sometimes forced-air ducts are  located immediately  beside
the foundation wall. It could be helpful to use a safety  cut-off
switch and perhaps  a  GFI  (as  discussed in Section 4.5.1,
Drilling the hole  through the slab) with the rotary hammer,
coring drill (if used), or auger.

If the horizontal pipe is being installed from outside a base-
ment house, the foundation wall usually will not  extend
beneath the slab; the slab will be resting on the footings. As a
result, the effort described above to drill through the founda-
tion wall will be unnecessary. In this case, excavation through
the soil under the footing will be required, by hand or using an
auger.

Excavating a pit beneath the slab. If the suction pipe is to
end near the interior face of the foundation wall (i.e., if the
hole has not been extended with an auger),  a sub-slab pit
should be excavated inside the foundation wall. As discussed
in Section 4.5.1, this should be done regardless of whether or
not there is aggregate under the slab. Possible pit sizes, and
the procedures for excavating the pit are generally the same as
those discussed in that earlier section for vertical/interior
pipes.

The pit  should always extend upward to the underside of the
slab, as illustrated in Figure 2, ensuring that the suction pipe is
communicating with the layer of aggregate (if present) be-
neath the slab. This is particularly important in cases where
the hole has been drilled through the wall some distance
below slab level. Even where there is no aggregate, there may
be a region of improved communication immediately under
the slab,  due  to subsidence of  the  soil or to an air gap
associated with sub-slab insulation, if present.

Especially where communication is poor, the pit should also
extend as far toward the interior of the slab as possible.

Investigators in Florida recommend that the pit not extend to
the left and right, exposing the interior face of the foundation
wall, when the walls are constructed of  hollow block (Fo90).
Air leakage into perimeter suction pipes through block walls
was a significant problem in Florida, and the concern was that
exposure of the interior face of the block wall would exacer-
bate this problem. Leakage would be much less of a concern
with poured concrete foundation walls.

The pit should extend downward below the wall hole. This
will reduce the risk of the pipe becoming blocked by soil or
water if the soil walls of the pit shift over time, or if water
collects in the pit due to condensation in the pipe or due to wet
seasons.

If the hole is extended using an auger, a pit might be created at
the end of the  augered hole by manipulation of the auger,
angling it up and down, and side to side.

Mounting  suction pipes  through the watt. When this
configuration is used in slab-on-grade houses, the suction pipe
would be mounted horizontally through the foundation wall
from outdoors. Unless an extended hole has been augered, the
pipe should terminate within a few inches of the interior face
of the wall, within the open space of  the excavated pit. In
basement houses, the pipe is mounted horizontally under the
footing, again terminating in the sub-slab pit.

The suction pipe must not be sloped upward toward the
underside of the slab. The greatest risk of this  occurring is
when the hole through the wall has  been made at some
distance below slab level. Such a slope would create a low
point in the piping at the exterior face  of the wall where
                                                        135

-------
condensate would accumulate, blocking gas flow. Where the
horizontal pipe is  at a depth beneath  the aggregate layer,
communication between the pipe and the aggregate must be
provided by extending the suction pit upward, rather than by
sloping the  pipe upward. In  fact, sloping  the pipe slightly
downward beneath the slab to facilitate condensate drainage
may be desirable, if this is possible.

The horizontal pipe should be well supported on the exterior
of the wall by soil underneath. Unless this pipe is being
connected by a horizontal run to other suction points around
the house perimeter, the outer end of this pipe will be directed
90° upwards to a fan and an  exterior stack above grade, as
discussed in Section 4.6.5. Even if this fan and stack are
supported by brackets or strapping, some of the weight will
likely be borne by the end of the horizontal pipe. If this end of
the pipe is not adequately supported underneath, the torque
applied by the fan/stack weight could cause the end of the pipe
beneath the slab to shift upwards, creating the low point for
condensate accumulation discussed in  the preceding para-
graph.

The gap between the outer circumference of the horizontal
pipe and the foundation wall  must be closed, to reduce the
amount of ambient air that would be drawn down through the
(usually shallow) soil into the  system. Since this annular
opening will be on the vertical face of a  wall, flowable
urethane caulk will not be applicable, since it would just drain
away. Gun-grade (non-flowable) polyurethane caulk could be
one choice. The wall and hole surfaces would be cleaned as
appropriate (e.g., the drilling dust vacuumed up), to remove
loose dust and dirt that would hinder bonding of the caulk to
the wall surfaces. The caulk should be worked into the hori-
zontal gap to a reasonable depth, to increase the contact area
for bonding.

Expanding closed-cell foams are perhaps the most common
choice to seal this gap. Foams  could be particularly appealing
where the foundation wall is hollow block. Properly injected,
the foam could seal not only the gap at the exterior face of the
wall, but also the gap between the interior hollow-block core
and the pipe, and the gap at the interior face of the wall (under
the slab). In this way, the foam would block  not only the
ambient air flowing down through the soil along the exterior
face of the wall, but also ambient and indoor air that might be
drawn down through the block cores.  The foam might be
injected into the cavity through  the gap between the suction
pipe and the exterior face of the wall  (Bro92), or through
small specially  drilled holes in  the exterior face above and
below  the pipe (K192). When using foam to seal the entire
block cavity, some type of support may have  to be placed
inside the block below the pipe in order to prevent the foam
from dropping down into the cavities below.

If foam is used, care must be exercised to ensure that the foam
does not expand into the sub-slab pit in  a  manner which
blocks the suction pipe. Such  blockage of the pipe has occa-
sionally been observed when  the pipe ends immediately be-
side the interior face wall (under the slab). To be sure, the
foam should be injected before the vertical riser is connected
to the horizontal pipe, so  that  the horizontal pipe can  be
inspected and rodded out, to detect and clear away any foam
that may have expanded into the sub-slab end of the pipe.

The use  of foam to close this gap for block walls has the
advantage of reducing air leakage into the system through the
block cores, but at the same  time, necessarily reduces any
block-wall depressurization component of this SSD system.
Use of foam in this manner  has generally appeared to be
advantageous, or at least not detrimental, in terms of the radon
reduction performance of the system; i.e., any BWD compo-
nent was either insignificant or unnecessary. In only one study
house (a large slab on grade with good aggregate and sub-slab
forced-air supply ducts) did the BWD component appear to be
important, with the use of foam creating a measurable deterio-
ration in  system performance (House 1 in He91a).

In addition to gun-grade polyurethane caulk and expanding
foam, cement or quick-setting mortar have also sometimes
been used to seal the gap around the suction pipe (An92,
Bro92). One  mitigator reports that termites and ants have
sometimes been observed to chew through expanding foam in
some regions, so that cement might be preferred at least for
the outer seal for this below-grade application in such regions
(Bro92).

In cases where the hole has been extended with an auger, and
a horizontal pipe is being inserted for some distance beneath
the slab,  it will be particularly important to rod out the pipe
after it has been inserted. The end of the pipe beneath the slab
may have become plugged with soil as the pipe was being
pushed through the augered hole. This problem can be re-
duced by inserting the pipe only part way into the augered
hole (K192), although it must  be inserted far enough so that
the open end  of the pipe is  reasonably isolated from the
interior face of the block wall (thus avoiding the excessive air
leakage into the system through the wall that the augering step
was intended to avoid).

After  the suction pipe has been installed and the exterior
excavation around the pipe is being re-filled, it would be
advisable to pack the back-filled dirt in an effort to reduce the
possible leakage of ambient air down through loose fill and
into the system.

Rigid PVC, PE, or  ABS  piping is a reasonable choice for
exterior suction  pipes, just as for pipes  installed indoors.
However, one mitigator (K189, K192) reports success using
flexible corrugated polyethylene and polypropylene piping
for the  horizontal below-grade penetrations  through
slab-on-grade foundation walls, beneath basement footings,
and beneath slab-on-grade footings in cases where the foot-
ings are  shallow. This corrugated piping is similar to the
material commonly used for drain tiles, except it is not perfo-
rated. Use of the flexible piping simplifies installation. It can
be considered for use in this exterior application because it is
supported horizontally by the  underlying soil, potentially
avoiding sagging; and, being below grade, there is  less con-
cern regarding leakage through the difficult-to-seal seams
between piping segments, and regarding damage to the flex-
ible material. This flexible piping is then connected to rigid
PVC piping to serve as the riser on which the fan is mounted
above grade.
                                                       136

-------
Where non-perforated flexible piping is used for  exterior
suction pipes, it must be buried at least  18 in. below grade
(K192) to avoid excessive ambient air leakage into the system
through the unsealed seams between piping segments. Also,
care must be taken to lay this flexible piping on a reasonably
flat-bottomed channel, since it can sag, providing low sites for
condensate accumulation.

4.5.5  Horizontal Pipe Installed Through
Foundation Wall from Inside Basement

Perhaps the most common use of horizontal suction pipes is to
treat adjoining slabs on grade in basement houses having such
adjoining slabs. The pipes are inserted horizontally  through
the stem wall separating the basement and the adjoining slab,
at a level just below the adjoining slab. Treating the adjoining
slab in this manner avoids the need to install pipes in the
living area on the adjoining slab (which will usually be highly
finished), and greatly simplifies connection of the SSD piping
treating the slab on grade into the suction pipes treating the
basement slab.

A typical configuration for installing a horizontal suction pipe
beneath an adjoining  slab on grade from inside the basement
and for connecting the horizontal pipe into a basement SSD
system is illustrated in Figure 18.

The methods for installing horizontal suction pipes  through
the stem wall from inside the basement are similar  to those
described in Section  4.5.4, for the case where the pipes are
installed from outdoors. There are  only a few differences. In
selecting the specific location for  drilling through the stem
wall, some attention will now be required to the nature of the
basement finish. Assuming that the horizontal pipe  is to be
connected to a vertical riser from the basement slab,  the hole
in the stem wall must be located near a basement riser (or in a
position to otherwise facilitate connection to a riser).

As with exterior horizontal penetrations, arotary hammer drill
may be the best approach for making the hole through the
stem wall, especially if the stem wall is hollow block. If the
wall is poured concrete and if it is desired  to make the hole by
supporting a coring drill sideways, sufficient space  must be
available for using the drill in this position and for supporting
it in this manner from below or  above.  See Section 4.5.4,
Drilling the hole through the foundation wall.

The pit excavated beneath the adjoining  slab should extend
upward to the underside of that slab. It should extend  outward
and downward enough to  prevent the horizontal pipe from
becoming blocked with soil if the sides of the pit shift over
time.

With  the configuration shown in  Figure 18, the horizontal
pipe could be sloped in either direction to avoid condensate
buildup in the pipe.  Any condensate in  the horizontal pipe
could drain either to the pit beneath the slab on grade,  or to the
vertical riser in the basement.

Care is needed in sealing the horizontal pipe into the stem wall
to reduce the amount of basement air drawn into the SSD
system  around this penetration. Thus, whether the wall is
poured concrete or hollow block, the gap in the face of the
wall toward the basement (the right-hand face in Figure 18)
must always be carefully sealed.

When the stem wall is block, mitigators report that significant
amounts of air can be drawn into the system via the hollow
blocks, even when care has been taken to seal the gap around
the penetration in the face of the block inside the basement
(Bro92, Mes92). This leakage, which adds a BWD component
to the system, will reduce the suction that can be maintained
in the SSD piping. The reduced suction can be a problem in
many cases, especially when sub-slab communication is mar-
ginal or poor and when maximum suction is thus needed in the
piping. As a result, it is often important that the gap also be
sealed  at the "outer" face of  the block; the outer face is
defined here as the face against the soil under the adjoining
slab, or the left-hand face in Figure  18. Sealing of the outer
face will reduce the amount of air from inside the leaky block
cavities being drawn into the pit and thus into the system.

If the hole through the "inner" face of the stem wall (inside the
basement) is sufficiently large, this could facilitate injecting
foam around the gap in  the  outer face by inserting a wand
from inside the basement,'before the inner face is  sealed
(Mes92). The foam might seal not only the gap in the outer
face, but also the block cavity around the pipe to further insure
that  air leakage  via  the block wall will be reduced. One
mitigator recommends drilling small holes through the inner
face of the blocks just above and below the suction pipe, and
injecting foam  into  the block  cavity through these holes
(K192). To aid in foaming the block cavity, it may be neces-
sary to stuff some type of supporting material into the void
beneath the pipe, to prevent the foam from  dropping down
into the voids below as it is being injected. Care must be taken
to ensure that the foam does not expand into the pit beneath
the adjoining slab in a manner which blocks the suction pipe.

The  below-grade horizontal pipes discussed in Section 4,5.4
are supported by the soil beneath the pipe. When the horizon-
tal pipe is inside  the basement, it may be supported at its
connection with the vertical riser when the horizontal pipe is
as short as suggested in Figure 18. If the horizontal pipe is not
near a riser, and must run for some horizontal distance before
connecting to  a riser, this horizontal run will have to be
supported by hangers or strapping as discussed in Section
4.5.1 (Figure 16) and Section 4.6.3.

4.5.6  Suction Pipe Installed from Inside
Garage

In houses having  an adjoining garage, it  would be very
convenient to be able to install the mitigation system in the
garage.  Especially when the garage is  a one-story slab on
grade with its own roof, placement of the suction pipes and
the fan  entirely in the garage would eliminate the aesthetic
impact in both the living area and outdoors, and would com-
monly result in  the  entire system being remote from the
bedroom wing.

In basement houses having walk-out basements with a garage
slab on the same level as the basement, the garage slab will
often be integral with the slab of the adjoining basement In
                                                       137

-------
                                                                     Note: Careful sealing of pipe
                                                                     penetration through outer face
                                                                     of block can be important to
                                                                     prevent excessive leakage of
                                                                     air into the system via the
                                                                     block wall.
             Expanding Urethane
             Foam or other Sealant
                                                     -T—n
                                                      •'-.  '•  <\
                                                                         Hole through Face of
                                                                         Block Inside Basement
                                                                         Must Be Sufficiently
                                                                         Large to Provide Access
                                                                         to Outer Face for Proper
                                                                         Sealing of Penetration
                                                                                Basement
             .  SLAB
                                                                             Open Hole (6"
                                                                             to 18" Radius)
Figure 18. A typical configuration for treating an adjoining slab on grade using a horizontal suction pipe penetrating the stem wall from inside
          the basement
these cases, if communication is good, vertical suction pipes
inside the garage may well treat the entire slab. In basement
houses having an adjoining slab-on-grade garage, suction on
the garage slab will likely have only limited impact on the
living area, and the garage will thus be of no use for locating
suction pipes. (However, in this latter case, the garage could
still provide an excellent location for possible routing of the
exhaust stack and mounting  of the fan, eliminating the need
for a stack inside the living area or outdoors.)

Of primary concern in this section are slab-on-grade houses
having an adjoining slab-on-grade garage. In such houses, the
living-area slab will essentially always be a few inches higher
than the garage slab, for code reasons. In some cases, there
will be a foundation wall (supporting the living-area slab)
separating the region under the garage slab from the region
under the living-area slab. In other cases, there will not be a
separating foundation wall, but there will still be the step
between the slabs; the concrete for both slabs may have been
poured at the same time, with  a form having been used to
create the step.

In either case, the communication between the garage slab and
the living-area slab will be uncertain and will vary from house
to house. This will always be true in houses having generally
poor sub-slab communication. It can also be true in  cases
where  there is aggregate beneath the living-area slab,  espe-
cially when there is a separating foundation wall. Therefore, it
                                                        138

-------
is uncertain whether a vertical suction pipe installed down
through the garage slab will effectively treat the living area.
Sub-slab suction field extension testing, together with experi-
ence with other houses in the area, would be necessary to
determine whether the house might be adequately treated with
vertical pipes in the garage.

On  occasion, slab-on-grade houses have been encountered
where the step down to the garage slab occurs in the garage
several feet away from the  frame wall which separates the
garage from the living area. That is, the living-area slab seems
to extend several feet into the  garage. The potential for
success with a vertical suction pipe in the garage is increased
in such cases, if the pipe is installed in that portion of the
living-area slab which  extends into the garage. But in most
cases, the step  down to the garage slab occurs right at the
frame wall, or within less than 1 ft of the wall; in these cases,
a vertical pipe through the living-area slab within the garage is
not practical.

Where vertical pipes in the garage will not adequately treat the
adjoining living area, the garage can still be considered as a
possible site for installing a suction pipe beneath the living-area
slab on grade, inserting the  pipe horizontally or at an angle
from inside the garage. The approach for installing such pipes
from inside the garage is discussed in the remainder of this
section.

In some cases when there is a foundation wall between the
living-area and garage slabs, steps as tall as 12 in. have been
observed between the two slabs. A step this tall would provide
a clearance of roughly 8 in. between the bottom of the
4-in.-thick living-area  slab and the top of the garage slab,
enough to install a 4-in. diameter suction pipe. In these cases,
a horizontal pipe could be installed through the foundation
wall beneath the living-area slab, using the same consider-
ations as apply when the horizontal pipes are installed from
outdoors (see Section 4.5.4). The only difference would be
that, in these cases, no excavation would be needed to expose
the  exterior face of the foundation wall beneath the living-area
slab.

But in most slab-on-grade houses having adjoining  garages,
there will be a much smaller drop (perhaps only 4 in.) between
the  living-area  and garage slabs. One logical approach for
inserting a suction pipe beneath the living area from inside the
garage is  to  install the pipe at an angle. This approach is
illustrated in Figure 19, adapted from Reference Fo90.

Whether a mitigator chooses to angle a pipe from inside the
garage in slab-on-grade houses, or whether the pipe is instead
installed vertically in the  living area, for example,  inside a
closet or in a utility room, will  depend upon a variety of
factors. These factors include the availability of pipe locations
inside the living area, the required pipe location to achieve the
desired suction field distribution under the slab, the resulting
pipe routing required if the pipe is in the living area, and the
preferences of the mitigator and homeowner. Location of the
pipe in the garage becomes more preferred when the house
has multiple stories and/or no attic, complicating pipe routing
and fan placement with indoor suction pipes.
                             To Fan in
                             Garage Attic
House Slab -i
r
X
*


                                           Suction Pipe

                                             45° Elbow
                                           Foam and/or
                                           Urethane
                                           Caulk
                                           Garage Slab
 Pit
Foundation Stem Wall
(May Sometimes Be
Present) .
Figure 19. One possible method for angling a suction pipe beneath
          the living-area slab of a slab-on-grade house, from inside
          an adjoining garage (adapted from Reference Fo90).
Selecting the specific hole location. The criteria for se-
lecting the specific hole location along the junction between
the garage and the living area will be generally similar to
those discussed in Section 4.5.4, for penetrations from out-
doors. The location will be selected to avoid obstructions in
the garage, and with consideration to the anticipated routing
of the exhaust piping within the garage.

Preparing the hole. If the pipe is indeed inserted at a 45°
angle right at the junction of the two slabs, as shown in Figure
19, the hole will have to penetrate concrete perhaps 8 in. thick
or more. If a foundation stem wall is present, and if it were
poured  concrete, the lower portion of the hole could  be
penetrating a layer of concrete even thicker still.

A rotary hammer drill could be used to prepare this hole, in
the manner described in earlier sections. Or, a coring drill
could be supported at a 45° angle.  Since coring drill bits are
often only 6 in. deep, it could be necessary to core the first 6
in., chip out the still-attached core of concrete, and then core
the remainder of the distance.

Excavating a pit. The pit beneath the living-area slab will be
excavated by hand through the diagonal hole in the concrete.
The methods for this excavation  will be  similar to those
discussed in Section 4.5.1. The pit should extend upward as
necessary to reach the underside of the living-area slab, inter-
secting  any aggregate layer. When aggregate is not present
under the living-area slab, the importance of this pit increases:
                                                         139

-------
The pit should be sufficiently deep and wide such that shifting
of the dirt sides of the pit over time or entry of water into the
pit during wet seasons will not block the suction pipe.

Mounting the suction pipe. The suction pipe is inserted
through the angled hole, as shown in Figure 19. The end of the
pipe beneath the slab should be in the open space within the
pit, above the bottom of the pit.  The gap between  the outer
circumference of the pipe and the sides of the hole must be
sealed with non-flowablepolyurethane caulk or with expand-
ing closed-cell foam. This angled section of pipe will be held
in this position by the hangars/strapping which will be secur-
ing the piping run overhead in the garage.

4.6    Design/Installation of the
Piping Network and  Fan
After the individual suction pipes are installed by one of the
methods discussed in Section 4.5, they will need to be con-
nected to a fan which will exhaust outdoors. There are alterna-
tive methods for configuring the  overall piping  network and
fan/exhaust system, and there are a variety of approaches for
the detailed design and installation  of the piping and fan/
exhaust system. Sections 4.6.2 and 4.6.3 address the design
and installation of the piping network leading from the indi-
vidual suction pipes to the fan; Sections 4.6.4 and 4.6.5 cover
the mounting of the fan and the design and installation of the
exhaust piping/stack on the exhaust side of the fan.

When there is more than one SSD suction pipe, these pipes
will generally be joined together by a horizontal run of PVC
piping. The resulting single manifold pipe must be routed to a
single fan. Usually, the amount of soil gas flow drawn by SSD
systems is sufficiently low  such that it is not necessary to use
more than one fan for the entire system. Multiple fans will be
warranted only  in cases where it  is not practical to  manifold
the suction pipes  together  for one reason or another, and it
decided to install two or more separate systems (with different
pipes routed to different fans) rather than trying to connect the
pipes.

For the purposes of this discussion, there are two fundamental
types of exhaust configuration:

 1. Interior stack, A representative version of this configura-
    tion is illustrated in Figure 20. When the suction pipes are
    inside the house, the piping can be routed up through the
    interior of  the  house, usually through closets or other
    inconspicuous areas on the  floors overhead, or via an
    existing utility chase. If there is an attic, the fan  should be
    mounted in the attic, exhausting through the roof. Mount-
    ing the fan in the attic protects the fan from the weather
    (especially  important when the fan is not UL rated for
    exterior use), and improves aesthetics outdoors.  If there is
    no attic, the fan could be on the roof.

 2. Exterior  stack.  One representative version of  this con-
    figuration is shown in Figure 21. In basement houses, the
    exhaust piping penetrates the basement band joist or
    foundation  wall just above (or below) grade. It  is recom-
    mended that the fan be mounted on this piping outdoors
    near grade, and that a  stack on the discharge side of the
    fan be installed to release the exhaust above the eaves.
    This approach usually minimizes the aesthetic impact
    inside the house, but increases the impact outdoors due to
    the stack extending up outside the house.

One common variation of the exterior stack configuration,
which can be used when a basement house has an adjoining
slab-on-grade garage, is to route the piping through the base-
ment band joist into the garage. The discharge stack is then
directed up inside the garage through the garage roof, assum-
ing that there is no living area above the garage. In this
approach, referred to here as a "garage stack," the fan could be
mounted in the garage attic, if present, or near the garage slab
at the base of the stack.

Mitigators  have tried a variety  of other variations to the
exterior stack configuration, with the primary intent of mini-
mizing  the aesthetic impact of  the exterior stack. In one
variation, the exhaust piping penetrating the basement shell is
extended horizontally below grade for  some distance away
from the house, where the exhaust pipe then comes above
grade at a less  visible location. In  another  variation, the
exhaust stack has" been deleted altogether, with the fan dis-
charging immediately beside the house at grade;  in some
cases, the fan has even been placed inside the basement to
further reduce the impact outdoors. Elimination of the exhaust
stack, and location of the fan inside the basement, are incon-
sistent with EPA's interim mitigation standards (EPA91b).

Where the suction pipe has been installed horizontally through
the foundation wall from outdoors, as discussed in Section
4.5.4 above,  the exhaust configuration will almost always
involve an  exterior stack. Of course, the exact configuration
would necessarily look different from that illustrated in Figure
21 for vertical indoor suction pipes.

4.6.1  Suction Loss in the Piping Network

Calculation of the suction loss in the piping network was
discussed previously in Section 4.3.2, in connection with the
selection of the suction pipe diameter. It is also referred to in
Section 4.4, in connection with selection  of the appropriate
fan. In this current section, this calculation is  covered more
definitively, in connection with the overall design of the
piping network (such as illustrated in Figures 20 and 21) in
conjunction with the selection of pipe diameter and fan per-
formance.

The suction loss in the piping network will depend upon: a)
the velocity in the pipe, which in turn depends upon the
volumetric flow rate (in cfm) and the pipe diameter; and b) the
length of piping and the number  and nature of the  flow
obstructions,  such as fittings. (It will also depend upon the
pipe  wall friction, which would be higher for  corrugated
flexible piping than for smooth-walled  PVC; only
smooth-walled pipe is being considered in this discussion.)
The system fan has to be selected to provide the needed
sub-slab depressurizations after compensating for the suction
loss in the piping.

In practice, most mitigators will have little occasion to per-
form piping loss calculations in residential applications. The
                                                       140

-------
       Notes:
       2.
Horizontal piping run in attic
shown only for illustration.
Stack could have penetrated
straight upward through
roof, if there were sufficient
headroom to install fan
at that location.

Electrical wiring to
fan illustrated in
later figure.
                                                                 Exhaust
                                               Horizontal
                                               Piping Sloped
                                               Down Toward
                                               Stack
                                                                                 Strapping (or other
                                                                                 Support) to Support
                                                                                 Piping and Fan
                                                             Insulation May Be
                                                             Needed on Attic Piping
                                                             in Very Cold Climates
                                               PVC T-Fitting,
                                               Cemented Tightly
                                               to Adjoining
                                               Piping
                                                                 r
                                                                         m
Strapping (or other
Support) to Support
Piping Weight (Every
4 -10 Ft.). See Earlier
Figure.
                                               Joist
                                                          Horizontal Piping
                                                          Sloped to Drain
                                                          Condensate toward
                                                          One of the Suction
                                                          Pipes
                                                   Suction Indicator
                                                   (or other Failure
                                                   Indicator/Alarm)
Figure 20.  One representative SSD piping configuration illustrating an interior exhaust stack.


                                                      141

-------
                                                                     Notes:
            Exhaust
 Strapping
        4" Round-to-3"X4"
        Rectangular
        Adapter  ~~

          4" DIa. Round
          PVC 45° Elbow
Standard Bracket to
Attach Downspout to
Side of House
         One Bracket As
         Close to Overhang
         As Possible
         Downspout Used
         As Stack
                      1. Figure depicts one configuration
                         utilizing 3"X4" metal downspout
                         as the exterior stack. Other
                         possible configurations using
                         PVC piping as the stack are
                         shown in a later figure.

                      2. Figure depicts stack being
                         supported at rain gutter by
                         strapping attached through
                         outside lip of gutter. Other
                         methods of support at the
                         gutter, and against the fascia
                         in the absence of a gutter, are
                         shown in a later figure.

                      3. Electrical wiring to fan
                         illustrated in later figure.
                                                     PVC T-Fitting,
                                                     Cemented Tightly
                                                     to Adjoining Piping
                                    l— Strapping (or Other
                                       Support) to Support
                                       Piping Weight (Every
                                       4-10 Ft.). See Earlier
                                       Figure.
                                                                      Horizontal Piping
                                                                      Sloped to Drain
                                                                      Condensate Toward
                                                                      One of the Suction
                                                                      Pipes
                                                            Suction Indicator
                                                            (or Other Failure
                                                            Indicator/Alarm)
Figure 21.  One representative SSD piping configuration illustrating an exterior exhaust stack.

                                                     142

-------
most common use will be in cases where it is necessary to
determine whether 2- or 3-in. diameter piping can be substi-
tuted for 4-in. piping in selected cases without encountering
undue suction losses. In some cases, where sub-slab commu-
nication is marginal  or where the piping network is particu-
larly long or convoluted, it may also be helpful in fan selec-
tion (e.g., to assess whether a higher-suction radial blower of
the type discussed in Section 4.4.2 might be necessary). And
when interpreting the more extensive sub-slab flow diagnos-
tics described in Section 3.5.1 to aid in selecting a fan, suction
loss calculations would be required to convert the sub-slab
depressurizations measured  at the baseline test hole  to the
suctions at the  fan inlet (so that a desired  fan performance
curve could be plotted).

On the other hand, it is unlikely that a mitigator will often
have much  choice regarding the length  of piping and the
number and type of fittings. The piping configuration will
probably often  be determined by other factors: the need to
locate suction points  in order to achieve an adequate sub-slab
suction field distribution; the location and nature of finish
inside and around the house; and the practical routing of the
exhaust stack. Assuming that the suction pipe location and the
piping configuration  have been selected intelligently to begin
with, there will probably not be much flexibility for reducing
the length or number of fittings if calculations suggest that
suction losses may be significant.

And, in general, there will  probably not often be a major
concern with suction losses in the piping network in residen-
tial applications, if 4-in. piping has been used and if one of the
90-watt tubular fans (or equivalent) has  been used,  as dis-
cussed later.

Thus, the mitigator should locate the suction points and design
the system piping in an effort to reduce the length of piping
(upstream and  downstream  of  the fan)  and to  reduce the
number of fittings, consistent with the need to achieve ad-
equate suction  field  extension and to have and acceptably
aesthetic system. With this approach, suction loss calculations
of the type discussed below will usually be limited to cases
where small-diameter piping or where long or convoluted
lengths of piping are being considered.
Calculation procedure. The suction losses in the piping
will result from two factors: 1) friction between the air and the
walls of the pipe, calculated as loss per unit length of straight
pipe; and 2) turbulence created by flow obstructions in the
piping system, namely, fittings such as elbows, tees, and size
reducers.

The losses  in straight pipe due to wall friction can be calcu-
lated based upon standard fluid flow considerations, depend-
ing upon the friction resistance of the pipe wall. The losses per
100 ft of smooth-walled pipe having average wall friction
were presented  in Figure  13  in Section  4.3.2.  The figure
presents the suction loss as a function of volumetric flow rate
and pipe diameter, i.e., as a function of gas velocity.

Suction losses associated with flow obstructions are handled
by expressing that loss in terms of the loss in an equivalent
length  of straight pipe. For many fittings, a L/D ratio can be
obtained from the literature, that is, the  length of straight
piping (L) that would give the same friction loss as mat caused
by the fitting, divided by the diameter of the piping (D). The
losses associated with various fittings are presented in Table
4, derived  from Reference Ca60. The figures  in  the  table
include both the L/D ratio for the fitting, and the equivalent
length of straight piping that would give the same suction
losses as a  function of pipe diameter.

Note that the suction loss in a 90° elbow  can vary dramati-
cally, depending upon whether the elbow makes a smooth vs.
a sharp turn. The internal dimensions of many commonly
available PVC elbows appear to be a cross between a smooth
turn (on the outer circumference) and a mitered turn (on the
inner circumference). Thus, the actual L/D ratio for the  el-
bows commonly used in mitigation applications will probably
be somewhere between 9 and 65.

Other common obstructions include size expanders or reduc-
ers, for example, the 4- to 6-in. couplings used to connect fans
having 6-in. unions onto 4-in. piping. The suction gains due to
such expanders and the losses due to reducers, which depend
upon velocity, can be calculated based upon tables and equa-
tions in Reference Ca60.

Once the piping  network has been designed for a given SSD
installation, such as illustrated in Figures 20  and 21, the
Table 4.   Number of Feet of Straight Pipe Required to Create the Same Suction Loss As Created by the Flow Obstruction in One Fitting
Type of Fitting
Tee
90° elbow
- smooth round curve
- mitered (sharp turn)
45° elbow
- smooth round curve
- 3-piece elbow
L/D
Ratio
-60*

9
65

4.5
6
Equivalent Feet of Pipe, for Different Pipe
Pipe Diameter 2 in.
10ft

1.5ft
11ft

0.8ft
1ft
3 in.
15ft

2ft
16ft

1ft
1.5ft
Diameters
4 in.
20ft

3ft
22ft

1.5ft
2ft
  This L/D ratio, for gas flow around the 90° bend within a tee fitting, varies slightly with gas velocity. The ratio for flow straight through the tee
  would be less than that for the flow making the turn from the branch.
                                                         143

-------
approximate suction loss that will occur in the network can be
calculated as follows:

 1. Convert the flow obstructions in the anticipated fittings in
    the total piping network (upstream and downstream of
    the fan) into an equivalent length of straight piping, using
    Table 4;

 2. Add this equivalent length to the total length of straight
    piping anticipated (upstream and downstream of the fan);

 3. Obtain from Figure 13 the approximate suction loss per
    100 ft of straight piping, for the anticipated pipe diameter
    and system volumetric flow.

 4. Multiply the  total number of feet derived in 2) above,
    divided by 100, times the loss per 100 ft obtained in 3).
    This will be the total estimated suction loss in the piping
    network at the projected air flow.

If there are several 90° elbows or tees in the piping network, it
would not be unusual for these fittings to  make  a larger
contribution to the total suction loss than that from the straight
piping.

Note that this calculation procedure addresses the piping and
fittings on the discharge (pressure) side of the fan, as well as
those  on the suction side. Such piping and fittings on the
pressure  side could result, e.g., the exterior stack shown in
Figure 21. The flow resistance on the pressure side will create
a back pressure that will reduce flows, and will reduce the
fan's ability to develop suction. Thus, flow resistance on the
pressure side must be considered on the same basis as resis-
tance  on the suction side, for the purposes of calculating the
suction that can be established.

Interpretation of results.  The total piping suction loss
calculated in 4) should be subtracted from the suction that can
be maintained by  the planned fan at the flow rate anticipated
in the system, obtained from the fan performance curve. This
difference will be the suction that the fan will maintain in the
suction pipe at the point where the pipe penetrates the slab. It
will also roughly equal the suction existing immediately under
the slab in the sub-slab pit; at typical flows, there will be only
a small suction loss resulting from the acceleration of the gas
in the pit to pipe velocity.  (See the calculation procedure for
suction losses due to "abrupt entrances" in Reference Ca60.)

If the calculated suction in the pipe near the slab is unaccept-
ably low, then steps may need to be taken to improve it. If
small-diameter piping has been planned, it may be necessary
to switch to larger-diameter piping, at least in some parts of
the system (such as the manifold pipe to which the individual
suction pipes connect). One could also switch to a fan which
can develop higher-suction at the anticipated flows, such as
those  discussed in Sections 4.4.2 and 4.4.3, or could mount
two standard tubular fans in series. One could also  consider
reducing the length of piping or number of fittings, although,
in many cases, it is doubtful that there would be sufficient
flexibility in adjusting the piping  configuration to achieve
substantial reductions in suction losses by this method.
 It is not possible to rigorously define the minimum acceptable
 suction in the pipe near the slab that would trigger design
 changes.  The real measure will be how well this suction
 extends beneath the slab. This will vary from case to case,
 depending upon communication and flow. As a rough rule of
 thumb based upon experience, with a 90-watt in-line duct fan
 in houses having reasonably good communication, the pipe
 suction near the slab should probably not fall much below 0.5
 in. WG; however, where communication is good, lower suc-
 tions may be acceptable. In poorer-communication cases,  a
 higher suction would be desirable. (A higher suction would
 also  be  expected;  since flows are often quite low in
 poor-communication houses, large pressure drops through the
 piping would be less likely.)

 The use  of Figure 13, in accordance with Step  3) above,
 requires some estimate, prior to  system installation, of what
 SSD  system  flows will be. If no pre-mitigation diagnostics
 have been performed other than a visual  inspection, this
 estimate will be very rough, based upon the mitigator's expe-
dience with other similar houses in the area. If  a simple
 (qualitative or  semi-quantitative) measurement of sub-slab
 flows has been made using a vacuum cleaner prior to mitiga-
 tion, as discussed in Section 3.5.1, the estimate of SSD flows
 would be somewhat more reliable, but still fairly rough. Only
 if the more extensive, quantitative,  pre-mitigation sub-slab
 flow measurements have been  made with the diagnostic
 vacuum cleaner and only if the piping system suction losses
 have been calculated based upon the suction-vs.-flow charac-
 teristics of the sub-slab region as discussed in Section 3.5.1
 will the flow estimates in the piping be reasonably accurate.

 Evaluation  of need for suction loss calculations. At the
 beginning of Section 4.6.1, it was stated that there will prob-
 ably not often be a major concern with suction losses in the
 piping network in residential applications, if 4-in. piping has
 been used and if one of the 90-watt tubular fans (or equiva-
 lent)  has been  used. The calculation procedure described
 above can be used to support this statement.

 Suppose that sub-slab communication is good and that, as  a
 result, flows are at the high end of the SSD range (100+ cfm).
 At those flows, losses in the 4-in. piping would be 0.6 in. WG
 per 100 ft, from Figure 13 in Section 4.3.2. The 90-watt fans
 would maintain about 1 in. WG at their inlet at 100 cfm. Thus,
 the fans could tolerate a piping length (including the equiva-
 lent length contributed by fitting  resistances) of 100 to 150 ft
 while still maintaining several tenths of an inch WG under the
 slab,  which   may  well  be  sufficient   suction  in
 good-communication cases. A piping length of 120 ft would
 correspond to, for example, 40 linear ft of straight 4-in. piping
 plus four 90° elbows or tees; this could often be sufficient for
 a system  in a two-story basement house requiring some hori-
 zontal run in the basement or attic in addition to the vertical
 rise up to the roofline.

 In fact, with  an extended length  of piping, flows might drop
 below 100 cfm, and suction correspondingly increase in ac-
 cordance with  the fan performance curve. But again, with
 good communication, some reduction in flow will often not be
 a problem, unless the high flows  are resulting from excessive
 leakage through a major unclosed entry route in the slab, or
                                                        144

-------
through highly permeable soil. In those cases, more suction
points or a larger fan might be warranted.

Suppose, on the other hand, that sub-slab communication is
poor, and that flows were at the low end of the SSD range
(about 20 cfm). At this flow, the 90-watt fans develop suc-
tions at the fan of perhaps 1.7 in. WG, and, from Figure 13,
the suction loss in 4-in. piping is roughly 0.03 in. WG per 100
ft. At such low suction losses, the fan would still be maintain-
ing about 1.5 in. WG suction under the slab even if the piping
run were 300 ft equivalent. It is difficult to imagine a piping
configuration so long and convoluted as to have an equivalent
length greater than this, even with multiple suction points and
numerous elbows and tees. Consequently, piping length/con-
figuration becomes almost irrelevant with 4-in. piping when
flows are so low. The only question in a case such as this is
whether more suction pipes or a higher suction  fan may be
needed due to the poor communication.

In summary, whether communication is good or poor, it is
doubtful that a mitigator will often encounter a piping con-
figuration so long or convoluted in residential applications
that  it would be justified to increase SSD  piping diameter
above 4 in., when a 90-watt tubular fan is being used. Accord-
ingly, when the intent is to install an SSD system with 4-in.
piping with  such a fan, piping  suction loss calculations will
usually not be necessary.

4.6.2  Considerations in Pipe Routing
Between Suction Points and Fan

If there is more than one SSD suction pipe, the multiple pipes
will  usually be connected by a horizontal run of piping. This
piping run will then continue to a point where it can penetrate
the shell  of the living area and lead to a single fan.  Two
representative routings are illustrated in Figures 20 and 21.

This piping should be routed so that it does not interfere with
normal traffic patterns in the house, or with the adjustment,
operation, or maintenance of any mechanical equipment. (The
one  exception to this can be the sump  pump in sump/DTD
systems, discussed in Section 5.) It should also be as aestheti-
cally acceptable as possible.

Connection of multiple pipes in basements having un-
finished or suspended ceilings. In an unfinished or par-
tially finished basement, vertical interior suction pipes will
normally extend upward to the basement ceiling, to the level
of the perimeter band joist which rests on top of the founda-
tion wall. If a vertical suction pipe is to be connected to one or
more other suction pipes elsewhere in the basement, a hori-
zontal pipe is. run across an unfinished ceiling (or above a
suspended ceiling)  to permit the suction pipes to be manifolded
together.

If one suction point is by the front or rear wall of the house
and  if a second pipe is needed, it would be logical to try to
locate the second pipe directly opposite the first on the oppo-
site wall (between  the same two overhead floor joists), if this
location would provide the needed suction field distribution.
In this case, the horizontal pipe joining the two points would
be parallel to the floor joists (which usually run from front to
rear), and the horizontal piping run could be up between the
joists, making it as unobtrusive as possible. This is the con-
figuration illustrated in Figures 20 and 21.

If the two suction points are on opposite ends of the house, so
that the connecting horizontal run is perpendicular to the floor
joists, the horizontal pipe will necessarily have to be below
the joists. Four-in. horizontal pipe running perpendicular to
the floor joists  would  never be run through cored holes
through  the joists, not  only because of installation effort
involved, but also because of structural concerns, as discussed
in Section 4.6.3 below. If there is a support beam running
from  one end of the house to the other, providing support to
the floor joists in the middle of the basement, the horizontal
piping can logically be run beside  this  support beam, just
below (and perpendicular to) the floor joists. If there are a
number of suction pipes along the length of the basement,
horizontal legs from the various pipes can tap into the central
horizontal pipe along its length.

Selection of where  the piping exits  the basement  At
some convenient location, the collection pipe must be routed
up through the house to a fan in the attic, as in Figure 20. Or,
the pipe must be  routed through  the band joist to a fan
outdoors or in the garage, as in Figure 21.  (Occasionally, a
mitigator may choose  to penetrate the foundation wall below
the band joist.)

Where the exhaust piping will rise through the house, the
location for this interior stack will be selected to provide the
most convenient and aesthetic route (e.g., up through a utility
chase, or through closets on the floors above). It would be nice
to avoid routing the pipes through bedroom closets, if this
were possible, to avoid the risk that the flow noise in the stack
might disturb sensitive occupants.

If the piping is to exit  through the band joist, the location for
this penetration  would be  selected based on a number of
criteria. Existing constraints may prevent all of these criteria
from being simultaneously met in the optimum manner. These
criteria include:

    -  Achieving the shortest and least visible horizontal run
      inside the house, avoiding indoor obstructions.

    -  Minimizing  the visual impact of the fan and stack
      outside the house; this will often mean that the penetra-
      tion should be on the rear (or perhaps side) of the
      house, preferably where the fan at the base of the stack
      might be concealed behind shrubbery.

    -  Avoiding obstructions on the exterior face of the house
      that would interfere with extension of the exterior stack
      up to the eave; thus, the piping cannot penetrate di-
      rectly beneath any windows (through which the stack
      could be seen from inside the house), and especially
      not directly below overhung bay windows.

    -  Ensuring  that the final discharge point (usually just
      above the eave  directly above the penetration) will be
      at least  10 ft away from upstairs windows, skylights, or
      other openings in the house shell, consistent withEPA's
                                                        145

-------
      interim radon mitigation standards (EPA91b), as dis-
      cussed in Section 4.6.4.

   -  Avoiding location  of the fan, the stack, or the final
      discharge point near bedrooms, if possible. The noise
      associated with the fan and stack, although often subtle,
      can occasionally be objectionable  to sensitive occu-
      pants, especially if generated near a bedroom window.
      The noise results from: fan noise heard at the fan, and
      fan noise transmitted inside the exhaust pipe and heard
      at the discharge point; stack vibration transmitted from
      the fan, or vibration in the house siding to which the
      stack is attached; flow noise in  the stack, if the gas
      velocity is high enough; and turbulence hi the exhaust
      escaping the pipe. Noise can be a particularly important
      consideration when using one of the high-suction/
      low-flow fans discussed in Section 4.4.3, since these
      tend to be noisier than the tubular fans.

   -  Locating the penetration relatively near to an electrical
      junction box inside the house, to facilitate connection
      of the fan wiring into the house circuitry, discussed in
      Section 4.6.5.

Where the fan and stack are inside an adjoining garage, many
of the above  criteria will be more easily met The fan and
stack will not be visible outside the house, avoiding concerns
about reducing the visual impact outdoors. There should be
increased flexibility in routing the stack around any obstruc-
tions in  the garage. The piping can be routed horizontally in
the garage attic if necessary to ensure that the exhaust is at
least 10 ft from any openings in the living area. The fan and
stack will likely be away from the bedroom wing, unless there
are bedrooms above the garage.

Pipe routing considerations when  there is  only one
SSD suction point.  If there is only one suction pipe pen-
etrating the slab, the only horizontal run needed in the system
would be the run over to the band joist penetration or to the
location where the interior stack is to rise through the house.
The location of the suction point would normally be selected
in an effort to minimize this horizontal run. Where the stack is
to extend up through the house, it could be possible to locate
the suction point directly under the path of the stack, so that no
horizontal run would exist at all.

Diameter of the piping network. In houses where there are
4-in. diameter suction pipes, the horizontal  piping which
connects the individual suction pipes will generally also be
4-in. diameter piping. As discussed in Section 4.6.1, flows in
SSD systems should almost never be high enough to warrant
the use  of  larger-diameter piping  for the central horizontal
pipe. Where such nigh flows are encountered in a SSD sys-
tem, the problem is probably one of leakage through  slab
openings or piping joints, suggesting a need for sealing rather
than for larger piping.

In poor-communication houses where a number of 2-in. or
3-in. diameter vertical suction pipes have been installed, it
may be advisable to use 4-in. piping for the horizontal collec-
tor, depending upon the actual flows and the fan used. Al-
though flows in individual  suction pipes will be low, the
combined flows from all pipes could be sufficiently high to
create an unacceptable suction loss in a 2-in. diameter hori-
zontal collection pipe.

Requirement that horizontal piping runs be sloped. At
all locations in the system, the horizontal piping must slope
slightly downward towards the vertical suction pipes, so that
condensed moisture can drain away. There must be  no low
points in the system piping where condensate can accumulate
(unless provisions are made for drainage, as discussed later).
Soil gas has  a high moisture content. This moisture will
condense inside the pipes whenever the piping run is exposed
to a temperature below that of the soil gas, which will occur
during cold weather in piping runs which are outdoors or in
unheated areas such as  attics. Ensuring that all horizontal
piping is sloped back toward the vertical suction pipes will
permit any condensed moisture to flow back down beneath
the slab through these risers.

If the horizontal piping were sloped down away from the
risers, the condensate could accumulate at the low end of the
piping. As a minimum, such  accumulation would reduce the
effective diameter of the piping, thus reducing the flow and
increasing the suction loss.  This could potentially  reduce
system performance, even if the  slope appears to be only
slight. In the extreme case, if the slope is great enough, the
water could block flow  entirely, thus rendering the  system
completely ineffective.

The degree of slope towards suction pipes can be fairly slight.
Some mitigators use slopes as great as 1 in. per 4 to 8 linear ft
when the  horizontal run is relatively short, consistent with
sewer codes. However, this degree of slope could sometimes
be inconvenient in long  horizontal runs of 4-in. piping; for
example, a 30-ft run perpendicular to the floor joists would
terminate at a suction pipe 6  in. below the joists if the slope
had to be  1 in. per 5 ft. In fact, since the flows involve only
condensate, the more rigorous slopes specified in sewer codes
are probably not necessary for this application. Some mitiga-
tors use more gradual slopes,  ranging from 1 in. per 10 ft to 1
in. per 40 ft, with one mitigator proposing 1 in. per 100 ft.

Approaches when  the piping must be routed over  or
under obstructions. A low point in  the  horizontal piping
can be created if the piping has to jut up or down in order to
pass over or under some utility piping or  other obstruction
along the  basement ceiling. If the piping forms  an inverted
"U" to pass over the obstruction, the horizontal segment on
the far  side of the obstruction could become a low point,
unless there is a vertical suction  pipe on  each  side  of the
inverted "U" to drain the moisture away. If the piping forms a
"U" to pass under  the obstruction, this "U" creates a trap
which could accumulate water over time. The placement of
the vertical risers and the routing of any horizontal piping
should be selected in an effort to avoid such low points in the
piping.

Figure 22  illustrates two approaches for addressing this issue
where potential low points  are unavoidable. If the piping
forms a "U" over the obstruction, as in Figure 22A, it would
be optimal to place a vertical suction pipe on each side of the
"U," with the horizontal run on either side of the "U" sloping
                                                        146

-------
downward toward the appropriate riser. If the piping forms a
"U" beneath the obstruction, as in Figure 22B, the  most
fail-safe approach would be to locate a suction pipe directly
under the "U" as shown.

If it is not practical to locate vertical pipes as suggested in
Figures 22 A arid B, a less preferred option would be to drill a
hole at least 1/2 in. diameter through the bottom of the piping
at the low point. Flexible tubing of this  diameter should be
firmly mounted into the hole, leading to some location where
the condensate can drain:  for  example, a nearby vertical
suction pipe, a  floor drain, or the voids inside a  nearby
hollow-block wall. Tubing at least 1/2 in.  diameter is required
to reduce the risk of plugging. Care must be taken to avoid
kinks in the tubing. Also, drain tubes should not be located in
unheated areas where the tube might become plugged with
ice. This design may result in some suction loss in the system,
due  to air being drawn into the pipe through the  tubing;
however, unless the system is marginal  to begin with, this
small air leak should not noticeably impact system  perfor-
mance.

Vertical mounting of fans. Whether the piping is routed
out through the band joist or up inside the  house, the piping
should always turn upward prior to the point at which the fan
                                 /— Flooring
    >Suction Pipe on
      Either Side of Hump
          Sewer Pipe
          (Cross-Section) or
          other Obstruction
A)  Piping run is routed over the obstruction; suction
    pipes on both sides of hump permit condensate
    drainage to sub-slab.
Strapping-,
            .Utility Pipe (or
             other Obstruction)
 Floor Joist
         ^7
^
                             — Suction Pipe Installed
                                at Bottom of Dip
B) Piping run is routed under the obstruction; suction
   pipe at low point in dip permits condensate drainage
   to sub-slab with no risk of pluggage.
Figure 22. Two of the alternative methods for providing proper drain-
          age when a horizontal piping run must be routed over or
          under obstructions, creating low points in the piping.
is mounted, so that the fan will be mounted vertically. Vertical
fan mounting will ensure that condensate will drain out of the
fan and not accumulate in the fan housing.

Pipe routing considerations for basements with fin-
ished ceilings.  The preceding discussion  focused on base-
ment houses which were partially or largely unfinished, where
any required horizontal piping  could be run across an unfin-
ished ceiling or above a suspended ceiling. Where the ceiling
and walls are finished  with sheetrock and/or panelling, the
piping installation may be more complicated and more expen-
sive, in order to reduce the aesthetic impact.

One option would be to install the horizontal piping runs in
the basement, as before, but providing any additional finish
that would be required (including framing around the piping,
as necessary) in order to make the installation aesthetically
acceptable. A second option would be to take each suction
pipe straight up through the house (e.g., through closets on the
floors above) into the attic, avoiding any horizontal runs in the
basement; if there is more than one pipe, the horizontal runs to
connect the pipes would be located in the attic. See the later
discussion of attic piping in slab-on-grade houses.

A third option for handling extensively-finished basements
would be to take each pipe out through the  band joist, and to
connect each pipe to  its  own fan  outdoors, avoiding the
complications involved in connecting the pipes.  As a fourth
option, the pipes could be taken out below  the band joist, so
that they penetrated the shell below grade level, in which case
the multiple pipes might be connected via a horizontal loop of
pipe buried around the perimeter outdoors. In  this case, a
vertical riser for the  fan would be tapped into the loop  at a
location suitable for an exterior fan and stack (see Selection of
where the piping exits  the basement above). A  fifth option
would be  to install the suction pipes horizontally from out-
doors, discussed later.

Pipe routing  considerations for slabs on grade. In
slab-on-grade houses having an attic, especially houses hav-
ing only one story, it will commonly be most convenient to
install vertical suction pipes inside the house, extending up
through the ceiling into the attic. Horizontal piping runs can
be made in the attic relatively easily, to connect the different
suction pipes (if there  is more than one), and to route the
piping to a point where the fan  can be mounted (on a vertical
segment of piping inside the attic) such that the fan exhaust
then penetrates straight up through the roof on the rear slope.

Another option would be to install the suction pipes horizon-
tally from outdoors.

Routing  considerations  with attic piping runs. With
attic piping runs in either basement or slab-on-grade houses,
the routing of the horizontal piping should be selected to
avoid any  obstructions in the attic, and to be out of the way of
expected traffic patterns, to the extent possible.  Again, care
must always be taken that the horizontal piping slope slightly
downward toward the riser pipes with ho low points, to permit
drainage of condensate.
                                                        147

-------
If there is only one riser pipe entering the attic from the living
area, as in Figure 20, the only horizontal run needed in the
attic would be that required to route the piping to the point in
the altic where exhaust pipe can penetrate up through the roof.
This will usually also be the point at which the fan is mounted,
In the manner illustrated in the figure. This generally should
be somewhere away from  the bedroom wing in an effort to
avoid fan noise there, especially when one of the high-suction/
low-flow fans is to be used. This location should also be on
the rear slope of the roof, so that the exhaust pipe will not be
visible from the street. Other considerations in fan location
include: ensuring that the exhaust point is at least 10 ft away
from openings in  the house shell, such as skylights or win-
dows in adjoining wings, in accordance with EPA's interim
standards (EPA91b); adequate headroom in the attic to mount
the fan; access for maintenance; and convenient access to a
power supply. In climates having heavy snowfall, it has also
been suggested that the exhaust pipe penetrate the roof slope
at a point sufficiently close to the peak so that substantial ice
dams will not build up behind the pipe, causing damage to the
roof or shingles (K192). In any climate, it is desirable to
penetrate as far up on the roof slope as possible to reduce re-
entrainment.

If the single riser pipe can penetrate into the attic  directly
below the roof penetration point, there would not have to be
any horizontal attic run at all.

Where there is more than one vertical riser pipe entering the
attic, these pipes must be  connected by a horizontal run in
some logical pattern. If there are a number of suction pipes
entering along the length of the attic, a horizontal collection
pipe can be located down the center of the attic, with horizon-
tal legs from the various suction pipes tapping  into the hori-
EOntal pipe along its length. The horizontal collection pipe
should be mounted at an elevation slightly above that at which
the risers end, so that the horizontal legs tapping into the
collection pipe will be sloped downward toward the risers. At
some convenient location, the collection pipe must be routed
to the point in the attic  where  the fan will be mounted and
where the fan exhaust can penetrate  through  the roof. In
systems having relatively high flows, connection of the fan
into the horizontal collection pipe at a central location be-
tween the vertical suction pipes will help ensure comparable
suctions in the individual suction pipes.

As with the basement piping, 4-in. diameter pipe will often be
desirable for the horizontal collection pipe (depending upon
flows and fan capability), even in cases where flows are low
enough  such that 2- or 3-in. piping is adequate for the indi-
vidual suction pipes.

Pipe routing considerations with horizontal penetra-
tion of suction pipes from outdoors. In those cases where
the suction pipe penetrates horizontally through the founda-
tion wall from outdoors, it will be common for the piping to
rise straight up from the horizontal suction pipe, with the fan
mounted vertically on this riser, in the manner indicated in
Figure 2, Where there are  multiple horizontal suction pipes,
they may  be connected by a  horizontal length or  loop of
piping buried around the perimeter of the house. One repre-
sentative configuration for  such a horizontal length of piping
is illustrated by the three-dimensional inset in Figure 2. The
riser and fan would then tap into this loop at an acceptable
location (see Selection of-where the piping exits the basement
above). To avoid condensate buildup, this connecting loop of
piping must be buried at an elevation slightly above that at
which the horizontal suction pipes penetrate through the wall,
so that the  "horizontal"  suction pipes are in fact sloped
slightly downward. If flexible piping is used for any portion of
the below-grade horizontal run, care must be taken that the
floor of the trench is flat, so that no part of the piping sags to
form a low point for condensate buildup.

4.6.3 Considerations in Pipe Installation
Between Suction Points and Fan

A number of factors should be considered when the piping
runs are installed.

 •  The rigid PVC, PE, or ABS piping is commonly obtained
    in 10-ft lengths, which must be cut to the desired length
    or spliced together with straight fittings (or couplings)
    where runs longer than 10 ft are needed. Other PVC
    fittings commonly  used in the piping network are tees,
    90° elbows, 45° and 22.5° bends, Y fittings,  and size
    reducers or expanders (e.g.,  2-in.-to-4-in. adapters, for
    connecting a 2-in. diameter suction pipe to a tee in a 4-in.
    diameter central collection pipe).

 •  The fittings must be of the same material (PVC, PE, or
    ABS) and of the same weight (thin-walled or Schedule
    40) as the piping.

 •  All piping and fittings  must be  carefully cemented to-
    gether with cement formulated for the particular material
    used (PVC, PE, or ABS), to ensure a permanent and
    gas-tight connection.  Simple pressure fits will not pro-
    vide a sufficiently gas-tight seal. While simple pressure
    fits might sometimes look tight, a significant fraction of
    the fan capacity could be consumed in drawing house air
    into the piping through the leaky joints. In addition, some
    simple pressure fits could become disconnected due to
    wear and tear over time.

 •  As emphasized previously, all horizontal piping runs
    should be  slightly sloped (1 in. drop for every 4 to 100
    linear ft),  so that condensed moisture will  drain back
    down into the sub-slab region via the suction pipes. Low
    points in the piping, where condensate could accumulate,
    should be  avoided. If low points are unavoidable, a
    vertical suction pipe  should be installed under the low
    point, if possible, as  shown in Figure 22; otherwise, a
    drain tube at least 1/2 in. diameter should be installed.

 •  The piping network must be properly supported. Hori-
    zontal runs should be supported every 4 to 10 ft. Several
    alternative methods for providing such support to hori-
    zontal runs are shown in Figure 16, using pipe hangers or
    plastic strapping attached to the overhead floor joists in
    basements. See also the discussion associated with these
    figures, toward the end of Section 4.5.1. Methods for
    supporting horizontal runs in attics, attaching the hangers
    or strapping to braces  and rafters in roof trusses, are
                                                       148

-------
illustrated later in Section 4.6.4. The piping should never
be supported against any other utility piping, ducting, or
mechanical equipment that may subsequently be moved
during servicing or relocated.

Vertical runs can be supported from wooden members of
the house using methods such as those illustrated in
Figure 17. Vertical suction pipes can also be supported at
the slab, by a method such as those shown in Figure 15.

Even the simplest piping network should be supported at,
at least, two locations. One of these locations can be at
the slab. One of the support locations should be near the
vertical suction pipe, to ensure that it does not drop into
the sub-slab pit, and one should be near the fan  (see
Sections 4.6.4 and 4.6.5).

Strapping  can effectively support the weight of the pip-
ing, but will usually not provide significant support against
lateral movement of the piping. Where strapping is the
primary support, a mitigator may wish to include some
measures  to reduce lateral  movement. Lateral support
will usually  be provided by the slab penetration, and by
the penetration through the band joist (exterior stacks) or
overhead flooring (interior stacks). Caulking the seams
around the penetration through band  joist for exterior
stacks, required in any event to prevent water entry, may
further reduce lateral movement of the pipe at that loca-
tion. Likewise, caulking of the seam around the penetra-
tion through the overhead flooring with interior stacks
will provide lateral support as well as improving appear-
ance (Bro92).

In supporting the pipes from  wooden members in the
basement  and attic, the pipes  may have to be isolated
from the wooden member to reduce the transmission of
vibration to  the wooden members, thus reducing noise.
The hangers and strapping shown in the various figures
ensure that the piping does not contact the wood. Where
there is "hard" contact with the wood, as with the pipe
clamp resting on the flooring in Figure 17, some cushion-
ing material should be placed between the clamp and the
flooring.

As discussed in Sections 4.6.4 and 4.6.5, fans are usually
mounted vertically on the rigid  piping using flexible
couplings, e.g., in the form of 4-to-6-in. adapters in cases
where the 90-watt fans (with 6-in. diameter couplings)
are connected to 4-in. piping.  These flexible PVC cou-
plings,  which have the appearance of rubber, greatly
reduce the transmission of fan vibration to the piping,
thus reducing noise. The flexible couplings  must be
clamped tightly to both the fan and the piping to avoid
leaks.

In very cold climates, any piping runs in  unheated but
protected space (such as attics and vented crawl spaces)
may need to be wrapped with insulation. Such insulation
will reduce the amount of condensation inside the pipe,
and, in particular, will reduce the risk that this condensate
will freeze and plug the pipe with ice.  Pipe insulation in
 attics and vented crawl spaces appears less important in
 climates that are only moderately cold.

 In cold climates,  some mitigators find it  important to
 insulate exterior piping (outdoor stacks),  to avoid ice
 blockage of the exterior stack during cold  weather. For
 such exterior use, one mitigator (Wi92) reports using the
 black closed-cell foam insulation commonly used around
 the refrigerant lines running to outdoor air conditioning
 compressors. This tubular insulation can be obtained in
 diameters  that fit around the  outside of PVC piping.
 Another approach that has been suggested for insulating
 outdoor piping involves use of Schedule 40 PVC piping
 having a foam core; this piping, which is not pressure
 rated, has foam in the center with layers of rigid PVC on
 the inner and outer surfaces. In the extreme, a mitigator
 could enclose the PVC stack using framing or aluminum
 chases, as discussed later, so that regular fiberglass insu-
 lation (not suitable for outdoor use) could  then be used
 around the piping.

 In an effort to avoid the need for insulation, some mitiga-
 tors try to place exterior stacks on the southern side of the
 house, to increase sun exposure (K192). Where aluminum
 downspouting is used as the exterior stack for appearance
 purposes, insulation of the stack would be somewhat
 more difficult, and would likely defeat the key purpose of
 using the downspouting, namely, to make the stack unob-
 trusive.

 Some mitigators also consider insulating certain indoor
 piping under the following conditions:

-  Interior stacks rising through bedroom closets, to re-
   duce noise. Such insulation becomes increasingly im-
   portant when the flow velocity in the pipe reaches
   about 1,000 to 1,500 ft/min, at which velocity increas-
   ing numbers of occupants  find the noise objectionable.

-  To reduce possible "sweating" on  the outside of the
   pipes during hot, humid weather, in cases where such
   condensate on the outside of the pipes could cause
   damage to  house finish.  Insulation for  this purpose
   would be important for piping running through finished
   areas or above suspended  ceilings.

 Where a suspended ("drop-down" or "false") ceiling is
 present, location  of horizontal piping runs  above the
 suspended ceiling is a relatively simple method for in-
 stalling the piping which greatly reduces the visual im-
 pact of the system.

 Where pipes are installed in finished space, or outside the
 house, it will sometimes be necessary to enclose the pipes
 for aesthetic purposes, depending upon homeowner pref-
 erences. One option is to box the pipes in using furring
 strips  and sheetrock, painted to match the pre-existing
 finish.  One  mitiffator (Jo91) has reported  using
 pre-fabricated tl    sided aluminum chases with flanges,
 obtained in 10-1   igths, which can  be attached to the
 wall enclosing    piping.  These pre-fabricated metal
                                                    149

-------
chases are relatively quick to install, and avoid the need
for carpentry.

Where PVC, PE, or ABS piping extends outdoors, if it is
not boxed in, some mitigators paint the pipe (for appear-
ance purposes and to protect it from UV radiation), or
coal it with a UV protectant Only UV-protected Sched-
ule 40 piping (or piping which is inherently UV-resistant)
should be used outdoors.

Holes for the piping through wooden members such as
band joists  are commonly made using a hole saw (a
circular blade that fits onto a power drill). Holes through
sub-flooring can be made with a standard hand saw, or
with a hole saw.

Where a pipe penetrates a load-bearing wooden member,
this penetration must be made in a manner which does not
seriously reduce the structural integrity of the member.

At least 2 in. should always remain between the hole and
both the top and bottom  of the member, even in the case
of band joists (which are supported underneath by the
foundation wall). With members which are not supported
underneath  (floor joists and rafters), an additional re-
quirement is that the hole not have a diameter greater than
one-third  the height of the member. With such unsup-
ported members, the distance between the bottom of the
hole and the bottom of the member should be greater than
2 in., if possible; the lower part of the member will be in
tension, and thus will be more likely to fail than the upper
part, which will be in compression.

These guidelines mean that a 4 in. diameter pipe can be
inserted through a 2- by 10-in. band joist (or even a 2- by
8-in. band joist, when encountered); at least 2 in. will
remain at the top and bottom, assuming the penetration is
near the center of the band joist. But a 4-in. pipe cannot
be installed through a 2- by 10-in. floor joist, because 4
in. is  greater than one-third of the 10-in. height of the
joist.  A 2-in. pipe could be inserted through the floor
joist, since it would constitute less than one-third of the
joist height.

In addition, floor joists should not be notched at the
bottom to accommodate the piping, since this may unac-
ceptably reduce the strength of the joist. For example, if a
2-in. notch is cut into the bottom of a 2- by 10-in. floor
joist, the entire joist will be to the strength of a 2- by 8-in.
joist.

When a hole is being cut through the basement band joist
to route the piping outdoors near grade level, care must be
taken to extend the hole neatly through the exterior finish
(usually siding or brick veneer) on the outside of the band
joist. Many mitigators make this hole just large enough to
accommodate  the pipe (typically about 4.5 in. outside
diameter).

When the exterior finish is brick veneer, a good approach
would be to drill two pilot holes from inside the house out
through the veneer, using a rotary drill with a small bit.
This will show outdoors where the hole needs to be in
order to properly line up with the band joist and piping
indoors. The hole through the wall is then made from
outdoors,  to  minimize damage to the bricks. First,  a
hammer and chisel are used to remove the necessary
bricks and expose the sheathing; then, a hole saw can be
used to make the hole through the sheathing and the band
joist.  Following installation  of the piping, the original
bricks can be cut as necessary and re-pointed.

When the exterior finish is  wood, vinyl, or aluminum
siding, a similar procedure is used. Two pilot holes are
drilled from indoors to identify the necessary location of
the hole outdoors. The hole is then drilled with a hole saw
from  outdoors, to ensure that the hole is located in the
center of a strip of siding, and to minimize visible damage
to the siding;  also, access is usually better from outdoors,
permitting a neater job. The  hole should probably be in
the center of a strip of siding, if possible, for the purposes
of convenience and  neatness.  To accomplish this, the
hole may need to be shifted upward or downward some-
what  compared  to the position indicated  by the pilot
holes; in that case, it would no longer be exactly in the
center of the band joist. (However, in no case should the
hole be shifted to an extent  that the space between the
hole and either the top or bottom of the band joist is less
than 2 in.)

In cases where the exterior finish covering the band joist
is particularly decorative or difficult to cut (e.g., a stone
veneer), some mitigators elect to penetrate the basement
foundation wall below the band joist (and below the point
at which the finish ends on the exterior). This often means
that the penetration is below grade level. Some mitigators
consider penetration of the foundation wall below the
band joist for aesthetic reasons even in cases where the
exterior finish is regular siding or brick veneer. Although
drilling through the block or concrete wall below grade is
more complicated than sawing through the band joist, this
approach avoids the complexity and aesthetic impact of
trying to deal with and restore the exterior finish. When a
hole is made through the foundation wall below grade,
careful sealing is required in order to avoid water prob-
lems indoors.

Where the piping through the basement band joist enters
an adjoining  slab-on-grade garage, the finish on the ga-
rage side  of the band joist will usually  be fire-rated
sheetrock. In this case, the hole through the joist is made
in the same manner as that described above, for the case
of exterior siding.

In mounting  the piping horizontally through the hole
through the band joist, many mitigators simply insert an
integral section of the rigid piping through the hole, as
suggested in Figure 1.

One mitigator (K192) has sometimes found it advisable to
place a straight 4- by 4-in. flexible coupling (analogous to
the 4- by 6-in. couplings used for mounting the fans) over
the section of pipe where it penetrates the joist, to reduce
the transmission of vibration to the band joist. (In this
                                                    150

-------
 case, the hole through the band joist may have to be
 somewhat larger than 4.5 in.) Other mitigators report that
 transmission to vibration to the band joist is usually not a
 problem.

 Following mounting of the pipe through the hole, the
 penetration must be finished in an appropriate manner
 (e.g., the brick veneer must be re-pointed), and the gap
 between the pipe and the wall must be caulked well from
 outside (e.g., using non-flowable urethane caulk). This is
 mandatory to prevent rainwater that runs down the out-
 side of the pipe from entering the house and damaging the
 band joist and other wooden members.

 If the gap is closed with mortar during the repair of brick
 veneer, rather than being closed with caulk, a non-shrink
 mortar should be used, so that it does not pull away from
 the pipe or foundation during curing.

 The frame wall between the living space and an adjoining
 slab-on-grade garage is considered a fire wall according
 to  most building codes, and is covered with fire-rated
 sheetrock. Where SSD exhaust piping penetrates through
 a basement band joist (and through the sheetrock on the
 garage side of the joist) into the adjoining garage, this
 would be considered a breach of the  fixe wall. The
 concern is that, in the event of a fire in the garage, the
 PVC pipe would melt, leaving  a 4-in. diameter opening
 through the wall  which could nominally facilitate the
 spread of the fire into the living area.

 According to codes, an appropriate fire break needs to be
 installed in the piping at that penetration. Three types of
 fire breaks can be considered:

-  The fusible linkage type. Such a fire breaks consist of a
   short segment of metal pipe containing a spring-loaded
   metal damper  held open by a  fusible linkage. This
   segment of metal pipe would fit in the hole through the
   band joist, with PVC piping connecting on either side.
   In the event of a fire, the fusible linkage would melt,
   allowing the damper to close, closing the hole through
   the joist. When such a unit is installed, care must be
   taken to ensure that the damper is in fact properly open,
   since  inadvertent closure of  the damper would block
   the exhaust pipe, rendering the SSD system ineffective.

-  The intumescent wrap type. Strips of intumescent ma-
   terial  are wrapped around the outside of the PVC pipe
   in the annular gap between the pipe and band joist. Or,
   if this gap is too small to accommodate the strips, the
   strips are placed at the seam  between  the pipe and the
   joist, enclosed within a metal collar that will force the
   materiallnto the joist hole if the pipe melts. In the event
   of a fire, if the pipe melts,  the intumescent material
   would expand into the hole, sealing the hole.

-  The framing-in approach. If the SSD  stack within the
   garage is  completely enclosed with framing and
   fire-rated sheetrock,  including the point at which the
   pipe penetrates into the garage from the basement, this
      penetration would no longer be considered a breach in
      the fire wall.

 •  Where a pipe penetrates an overhead ceiling into the story
    above, care must be taken to locate the overhead hole to
    avoid obstructions on the floor above. Before any hole is
    made through a ceiling, it should be verified that there are
    not radiant heating coils in the ceiling, and that there is
    not electrical wiring, forced-air ducting, or other utilities
    above a suspended or sheetrock ceiling.

    If the ceiling hole is to be lined up with a hole in the slab
    directly below, special care will be needed in siting the
    two holes. The piping can be manipulated to fit through
    the two holes by slipping the base down to the bottom of
    the sub-slab pit, the sliding it upward through the ceiling
    hole. The use of thin-walled  piping  (which will flex
    slightly) may simplify mounting a length of piping through
    the two holes.

    •  Where a ceiling penetration is being considered, it
       should be recognized that the ceiling will sometimes
       be considered a fire break, similar to the wall between
       the living area and the garage. The ceiling  may be
       considered a fire break in multi-family housing, or in
       large buildings such as schools, though not usually in
       single-family houses. The ceiling may sometimes be
       considered a fire break in residential garages. Where
       the ceiling is considered a fire break, one of the three
       approaches indicated above for the garage fire wall
       should be implemented whenever that ceiling is pen-
       etrated by a PVC, PE, or ABS pipe.

4.6.4  Design and Installation  of the  Fan
and Exhaust  Piping-—Interior Stacks

The preceding two sections addressed the routing and installa-
tion of the piping between the suction pipe and the fan. The
next two sections address the mounting of the fan onto this
piping network, and the design and installation of the piping
leading from the fan to the ultimate discharge point.

As discussed in the introduction to Section  4.6, two basic
exhaust configurations are considered for the purposes of this
document: an interior stack, with the stack rising through the
living  area to a fan usually mounted in the attic; and  an
exterior  (or garage) stack, with the fan and exhaust stack
mounted outdoors (or in the garage). Section 4.6.4 addresses
the first  of these cases; exterior and garage stacks are ad-
dressed in Section 4.6.5.

Overall design considerations for all stack configura-
tions. There are two general guidelines that are considered in
the design of the exhaust system. Both of these guidelines
result from the fact that there can be high radon concentrations
in the exhaust (often more than 100 pCi/L and  occasionally
greater than 1,000 pCi/L).

  1. The fan  should be  outside the livable envelope of the
    house.
                                                    151

-------
 2. The exhaust  should be  released at a point where
    re-entrainment of the exhaust back into the house will be
    minimized, and where exposure of persons in the yard
    and in neighboring houses will be minimized.

Regarding the first guideline above, the fan should be outside
the livable envelope due to a concern that leaks might develop
over the years in the fan housing or in the pipe couplings or
fillings on the pressure (discharge) side of the fan. Such leaks
could result due to aging  of the cement sealing the joints, or
due  to wear and tear on the piping over the  years. If
pressure-side leaks develop and if the pressure-side piping is
inside the livable space, some of the potentially high-radon
exhaust gas would be forced into the livable space. Such leaks
could go undiscovered for years. While the amount of leakage
could be small in  many cases, there have been some cases
where indoor pressure-side leaks have caused a distinct in-
crease in indoor radon levels  compared  to  post-mitigation
levels with the leaks sealed.

No cases have been identified where indoor pressure-side
leaks in an otherwise properly installed SSD system were
sufficiently severe to cause indoor levels to  increase above
their pre-mitigation values. However, this nominally could
happen if the leakage was significant enough.

There are no definitive data on the frequency or severity of
such pressure-side leaks in the exhaust piping. Nor are there
data on how alternative methods for  careful sealing  and
supporting of the exhaust piping might reduce the risk of such
leaks. In concept, it should be possible to design the exhaust
piping to reduce this risk. However, in the absence of any data
on the effectiveness and practical durability of such measures
and in view of the problems that could potentially arise if the
fan, in fact, wound up blowing  sub-slab radon into the house,
EPA's Interim Radon Mitigation Standards (EPA91b) conser-
vatively specify that the  fan should be outside the livable
envelope.

To be outside the livable envelope, the fan should be in the
attic (for the stack configurations discussed in this section), or
outdoors or in an adjoining garage (for the stack configura-
tions discussed in Section 4.6.5). House  air flow patterns
typically involve air movement from the living area up into
tire attic and out through vents and leakage points around the
roof. Thus,  any  pressure-side leakage  associated  with
attic-mounted fans would generally  tend to be carried away
from the living area by this natural flow. Attached garages are
sufficiently well isolated  from the living space, such that if
there were any leakage from  garage-mounted fans, only a
small traction of this leakage would circulate into the living
area. Mounting the fan in the attic of the garage would further
reduce such circulation, although such attic mounting prob-
ably will not usually be necessary.

EPA's interim mitigation standards  specify that fans not be
mounted in crawl spaces, since air flow patterns would tend to
draw any  pressure-side leakage from  crawl-space-mounted
fans up into the living area. Crawl-space-mounted fans would
be of greatest concern in  cases where the crawl space is not
ventilated, and especially  where the crawl space is effectively
conditioned space (e.g., where  the crawl space is completely
open to an adjoining basement). Where the crawl space is not
conditioned, and has foundation vents around the perimeter,
any leakage would be diluted by outdoor air.

Regarding  the second  guideline listed above—minimizing
exhaust re-entrainment and exposure to persons outdoors—
EPA recommends that the exhaust be released vertically
above the house (or garage) eave wherever possible. EPA's
Interim Radon Mitigation Standards (EPA91b) currently
specify that the discharge point be at least 10  ft: above grade
level; away from any window, door, or other opening (such as
an operable skylight); away from any private or public access;
and away from any  opening in  an adjacent building. It is
anticipated that EPA's final mitigation standards will explic-
itly require that  the exhaust be discharged above the eave,
preferably the highest  eave. Design of the exhaust in this
manner should help ensure that there is significant dilution of
the exhaust by outdoor air before it can be re-entrained into
the house, or before it can reach persons in the yard or in
neighboring houses.  The  interior or exterior stack that is
required to discharge the exhaust in this manner can add about
$100 or more to the SSD installation cost (compared to the
alternative  of exhausting immediately beside the house at
grade level) (He91b, He91c), and can sometimes create an
aesthetic impact.
                   !'''  I M
Very little data exist quantifying the effects of SSD exhaust
re-entrainment, as a function of the key variables (including
exhaust configuration, exhaust location, exhaust velocity/mo-
mentum, house characteristics, and weather conditions). Docu-
mented cases where exhaust re-entrainment in fact caused a
significant deterioration in SSD performance all involve cases
where: the exhaust was directed downward toward the soil
immediately beside the house, as with a drier vent or with a
90° downward elbow on the fan exhaust (Mi87); or where the
exhaust was directed straight upward immediately beside the
house, or horizontally parallel to the house, especially when
there were doors or windows nearby (Fi91). The very limited
data which are available directly comparing above-eave and
grade-level exhausts suggest that grade-level exhausts, di-
rected  horizontal to  grade and aimed  90° away from the
house, can result in re-entrainment no more severe than that
experienced when the exhaust is discharged vertically above
the eave (Fi91). Of course, the relative effects of the grade-level
vs. above-eave exhaust might vary for differing weather con-
ditions, exhaust velocities, etc.

Also, grade-level exhaust would not necessarily prevent expo-
sure of persons in the yard immediately beside the exhaust
Limited data suggest that the radon in the exhaust will often
be substantially diluted within a few feet of the discharge
location.

In summary, exhaust vertically above the eave, or use of the
10-ft criteria currently  specified  in EPA's  standards, repre-
sents a conservative effort to help  minimize re-entrainment
and exposure of persons outdoors. This conservative approach
is currently preferred because data are not presently available
to quantify those conditions under which grade-level exhaust
will  result in acceptably  low re-entrainment and  outdoor
exposure.  The effects of a given amount of re-entrainment
will be less when the exhaust concentration is relatively low
                                                        152

-------
(e.g., below about 100 pCi/L), suggesting that measures to
reduce re-entrainment might potentially be less crucial at such
relatively low exhaust levels.

General considerations with interior stacks.  One repre-
sentative example of an interior stack is illustrated in Figure
20.

The decision to install an interior stack instead of an exterior
or garage stack will depend upon the  specific situation in a
given house, and upon the preferences of the mitigator and
homeowner. An interior stack will most likely be preferred
when: a) there is a convenient utility chase leading from the
basement to the attic through which an interior stack might be
routed; b) in the absence of a chase, there is some other
convenient route for directing an interior stack up through the
floor(s) above the basement into the attic; c) there is an attic
where the fan can be located;  and/or d) there is no adjoining
garage,  or it is impractical to route the stack through the
garage, so that a garage stack is not an option.

Where it is practical to route the exhaust piping up through the
house, and where there is an attic in  which the  fan can be
mounted, the  interior stack/attic fan  exhaust configuration
offers several advantages.

   -  The aesthetic impact outside the house  is essentially
      eliminated. The only part of the system visible out-
      doors will be a foot or two of the 4-in. diameter exhaust
      pipe, which will normally be on the rear slope of the
      roof (away from the street), and which will have the
      general appearance of a plumbing vent.

   -  The fan in the attic will be protected from the weather
      (particularly important for fans not rated for outdoor
      use) while being outside the livable envelope of the
      house. Likewise, all electrical wiring for the fan will be
      indoors, somewhat simplifying the wiring. At the same
      time, the fan and wiring will  be out of sight from
      indoors as well as outdoors.

   -  The discharge will be well above the eave (and clearly
      more than 10  ft above grade), in accordance with
      EPA's interim standards, since the exhaust  can pen-
      etrate through the roof at a point toward the peak. The
      discharge point will almost automatically be at least 10
      ft away from windows, doors, private or public access
      routes, and adjacent buildings, in accordance with EPA's
      standards. And it should be relatively easy to ensure
      that  the discharge is at least 10 ft away  from other
      openings through  the house  shell, such as  operable
      skylights, air intakes, and gable or soffit vents.

The primary disadvantage of the interior stack/attic fan con-
figuration is that a route must be identified by which the
piping can extend up inside the house. Where the house is a
one-story slab on grade (so that the pipes can penetrate the
ceiling directly into the attic), or where there is an accessible
utility chase providing a  direct path to the attic,  routing the
piping up inside the house will not be complicated. But where
such a convenient path does not exist, the piping will have to
be routed through, e.g., closets on the floor(s) above the slab.
In such cases, if there is an adjoining slab-on-grade garage, a
mitigator may wish to consider a possibly more convenient
option of routing the piping into the adjoining garage, since
garage stacks offer many of the same advantages as interior
stacks.

A cost analysis (He91b, He91c) has shown that installing the
stack up through closets in the house can sometimes be no
more expensive than installing the stack exterior to the house
or in an adjoining garage. The cost impact of the interior stack
will depend upon the exact characteristics of the house, and
the experience of the mitigator's crew with interior stacks.

Selecting the  location of interior stacks.  Interior stacks
will be extended up through the house through a utility chase,
if available, or through closets or other inconspicuous areas in
the floors above the slab, as discussed  in  Section 4.6.2.
Routing is easiest when there is an accessible chase, or when
the house is a one-story slab-on-grade so that  the pipes
penetrating the  living-area ceiling enter the attic  directly.

Selecting the  location of fans for interior stacks. With
the interior stacks, the fan is usually mounted in the attic. As
discussed under Overall  design  considerations  above, fans
mounted in attics are considered outside the livable envelope
of the house, in accordance with EPA's interim standard. Any
leakage of exhaust from the pressure side of  the fan should
flow outdoors in response to normal flow patterns up through
the house, rather than being drawn down into the living area.

Within the attic, the fan is commonly mounted directly below
the point at which the exhaust pipe will penetrate up through
the roof, as shown in Figure 20. On this basis, the criteria for
selecting the specific location for the fan within the attic are
discussed in Section 4.6.2 (see Routing considerations with
attic piping runs). These criteria include criteria associated
with the fan (e.g., access to fan in attic) and criteria associated
with the exhaust penetration (e.g., penetration  on rear slope).
Occasional situations may  be encountered where it is pre-
ferred to mount the fan at  some location other than that at
which the exhaust pipe penetrates, e.g.,  to provide easier
access to the fan for maintenance, or to simplify the wiring.

Considerations  in designing and installing the piping network
in the attic, leading to the fan, are discussed in Section 4.6.2
(see Routing considerations with attic piping runs) and Sec-
tion 4.6.3. In most cases, some horizontal run  will be needed
to direct the piping to the point where the fan will be mounted
and/or where the stack will penetrate the roof. A horizontal
run will always be required when multiple risers penetrate
into the attic and must be manifolded together.  Even where
only one riser enters the attic, there may still have to be a
horizontal run in the attic if the vertical piping leading up to
the attic cannot be placed directly below the roof penetration.

Where there is  not an attic or where any attic  is inaccessible
(e.g., where there is a cathedral ceiling or a flat roof), the fan
for any interior stack would have to be installed on the roof.
Roof mounting of the fan would eliminate some of the advan-
tages of this exhaust configuration. The aesthetic impact
outdoors would be increased, since the fan would  now  be
visible. In addition, when the roof is flat, penetration of the
                                                        153

-------
stack through the roof can create water leakage problems over
time unless the penetration is sealed very carefully. Where
there is no attic, the interior stack may sometimes still be a
viable option, but an exterior stack could be preferable.

Criteria for selecting fans for interior stacks. The selec-
tion of the appropriate fan to achieve adequate SSD perfor-
mance has been  discussed in  Section 4.4. In  the figures and
discussion in this section, the focus is on the in-line tubular
fans covered in Section 4.4.1. If another type of fan were used,
the details regarding the mounting and support of the fan
would have to be adjusted accordingly, for that other type of
fan.

With interior stacks having the fan mounted in the attic, it is
not necessary for the fan to be UL-rated for exterior use.

If the fan were mounted on the roof, the fan would no longer
be protected, and it would have to be UL-rated for outdoor
use. Or, a protective housing would have  to be  installed
around the fan. Where a roof-mounted fan is required, a fan
model specifically designed for wall or roof mounting could
be considered, although the in-line models discussed in Sec-
tions 4.4.1 and 4.4.2 could still be practical.

Mounting fans in attics. A detailed diagram illustrating a
representative technique for mounting a  fan in the attic is
presented in Figure 23.

The fan should always be mounted vertically. Vertical mount-
ing of the fan is crucial, so that condensed moisture  (and any
precipitation that enters the stack) will drain down through the
fan and down to  the sub-slab region via the system piping. If
the fan were mounted horizontally, the condensed  moisture
could accumulate in the  lower portion of the fan  housing.
Such accumulated moisture would interfere with the rotation
of the fan blades, significantly reducing fan performance and
shortening fan lifetime.

If only one horizontal pipe is running over to the fan  location,
as in Figure 20, a 90° elbow is installed on the end of the
horizontal pipe. The fan is then mounted on a vertical segment
of piping extending up from this elbow. If horizontal pipes are
running to the fan from two directions, as in Figure  23, a tee
fitting would be used. If there is no horizontal run, the fan
would be mounted directly on the vertical riser penetrating the
ceiling into the attic.

The fan must be mounted on the vertical pipe with an air-tight
coupling.  Commonly, flexible PVC  couplings, such as those
marketed by Fernco, Indiana Seal, and Uniseal. These flexible
couplings should help reduce  the vibration transmitted to the
piping  by the fan, compared  to what would be expected if
rigid PVC couplings were used. But even with these flexible
couplings, some  vibration can still be transmitted (Bro92).

Figure 23  (and the other figures in this document) illustrate an
in-line  tubular fan having 6-in. diameter couplings, such as
the 90-watt "Category 4" fans listed previously in  Table 1,
being connected to 4-in. diameter piping. Thus, a4-in.-to-6-in.
flexible coupling is shown both below the fan, and also above,
where the 6-in. fan outlet is  reduced back to 4-in. exhaust
piping that penetrates the roof. If one of the smaller fans listed
in Table 1 were being used, the couplings would be a straight
4-in. couplings or a 4- to 5-in. couplings. If one of the radial
blowers discussed in Section 4.4.2 were being used along with
3-in. suction piping, 3- to 3-in. couplings would be used.

The flexible couplings are clamped tightly to the fan and to
the pipe using standard circular hose clamps, to avoid leakage.
Leakage at the outlet coupling of the fan would force some
high-radon exhaust gas out beside the house (or in the garage
attic). Leakage at the inlet coupling would draw outdoor air
into the system, increasing system flow and potentially reduc-
ing the suction that could be maintained in the SSD pipes. The
hose clamps must be tightened sufficiently to prevent leakage;
however, severe over-tightening could cause distortion of the
pipe or fan, potentially increasing leakage (Bro92).

If there is uncertainty regarding the air-tightness of the fit,
tests could be conducted with a chemical smoke stick to detect
leaks. If leakage is occurring, the coupling should be adjusted;
if adjustment does not eliminate the leakage, the seams be-
tween the flexible coupling and the piping or fan could be
caulked. Any such caulking should be done with an  easily
removable caulk, such as silicone caulk, rather than polyure-
thane, so that the fan can be easily removed later for mainte-
nance. Caulking should not normally be necessary.

Subsequent removal  of the fan for maintenance or replace-
ment  would involve loosening the flexible  couplings  and
sliding the  fan up or down as necessary to remove the unit.
The short length of stack above the fan in Figure 23 may be
withdrawn  down through the flashing as part of this proce-
dure. If the stack is thus removed and re-installed as part of the
fan maintenance, care should be taken to ensure that any seal
between the stack and the top of the flashing remains tight
when the stack is re-installed.

Many of the plastic-bodied in-line duct fans on the market
today have integral, essentially air-tight housings, so that
there  will not be leakage into or  out  of  the fan  housing.
However, some fan models (especially those with metal bod-
ies) have seams in the housings which will permit exhaust to
leak out (or attic air to leak in) if not caulked properly. Any
such leaks in the fan housing must be detected and effectively
caulked.

The mounting configuration illustrated in  Figure 23  repre-
sents the in-line tubular fans discussed in Section 4.4.1. It also
represents  the in-line radial blowers  discussed in Section
4.4.2. Other types of fans may have to be mounted in different
ways. For example, if they were to be mounted in an attic, the
DynaVac HS series of high-suction/low-flow fans  listed in
Table 3 (Section 4.4.3) would have to be bolted into vertical
wooden members nailed into, or part of, the roof truss. The
Pelican high-suction blower can be suspended from a horizon-
tal brace in the truss, using a mounting intended to reduce the
transmission of vibration to the brace.

Support for fans in attics. The fan and piping  must be
adequately supported.
                                                        154

-------
   Notes:
   2.
   3.
Figure depicts fan hard-wired into
existing 110-volt house circuit. Other
options are illustrated in a later figure.
Wire should be stapled against wooden
member every 18 in.

Figure depicts end of each horizontal
run being supported near fan by strapping
attached to brace in roof truss. Appropriate
method of support in  a given house will be
determined by exact configuration of piping
and wooden members. Some alternative methods
of support are suggested in a previous and
a later figure.

When strapping does not support fan weight
directly, some step may be necessary to
prevent fan assembly from slipping down
on vertical pipe below.
                              Flashing, Properly
                              Blended in with
                              Existing Shingles;
                              Sealed If Necessary
                                                          Exhaust
                                                                                110-Volt Wiring

                                                                                 Electrical Switch
                                                                                 (Accessible and
                                                                                 within Sight of Fan)
                                                              Existing 110-Volt
                                                              Junction Box
                                                                                          Wooden Support
                                                                                          for Wiring, If
                                                                                          Needed1

                                                                                            To Other
                                                                                            Riser, If Any
         Stack Rising from
         Suction Pipe(s)
         through Slab Below
                                          Horizontal Runs
                                          Sloped Down Toward
                                          Pipe Rising from
                                          Living Area, for
                                          Condensate Drainage
1— Insulation May
   Be Needed On
   Attic Piping
   in Very Cold
   Climates
Figure 23. A representative method for mounting a. fan in the attic above an interior stack.

                                                     155

-------
When there is a horizontal piping run in the attic supporting a
fan at one end, the weight of the fan could cause the piping run
to sag at that end. This would create a low point where
condensed moisture could accumulate. One way of avoiding
such sagging is to support each horizontal piping run at a point
near the fan, using hangers or strapping attaching the pipe to
members in the roof truss. Figures 20 and 23 show strapping
attaching the horizontal piping to a horizontal roof brace.
Each horizontal  run is supported in Figure 23. Other ap-
proaches using hangers and strapping can also be considered,
adapting some of the alternatives  shown in Figure 16  for
supporting  basement runs from overhead floor joists; the
difference would be that, in this case, the supporting members
would be components of the roof truss instead of floor joists.

If the horizontal  run were parallel to the attic floor joists, it
could be down between the joists, suspended by a strapping
loop analogous to the top diagram in Figure 16b.

Some investigators consider supporting the horizontal piping
directly on the attic floor joists, or by placing some  type of
solid support between the tee fitting in Figure 23 and the
underlying joist, so that the piping and fan weight is borne by
the joist. If this approach were used, it would be important that
sufficient padding be placed between the piping and the joists,
to prevent fan vibration from being transmitted to the joists.

Support could also be provided directly at the fan,  supple-
menting or possibly replacing any support at the horizontal
piping. The fan could be supported by strapping looped tightly
around the  bottom flexible coupling, and attached to a roof
brace or rafter. Two of the three diagrams in Figure 24 are
examples this approach; although this figure is for the case
where there is no horizontal run in the attic, those two ap-
proaches for supporting the fan would also  be applicable
when a horizontal run is present. The fan could also  be
supported directly using a fan mounting bracket, attaching the
fan to wooden members in the roof truss. Use of a  bracket
could result in transmission of vibration to the truss.

However the support is provided, it is important that  support
of each horizontal run be provided near the fan. And (or), the
fan itself should be supported directly. It may also be advis-
able to  support each horizontal run in the attic at the end
remote from the fan (although  such remote support is not
shown in Figure 20), depending upon how the interior stack is
supported on the stories below.

In some cases, there may be no horizontal piping run in the
attic --  i.e., the  riser penetrating  the ceiling into the attic
extends straight up through the roof. Figure 24 illustrates three
possible alternatives for providing support in the attic for
straight vertical  runs. Two of the options  show  strapping
looped around the flexible  coupling beneath the fan and
attached to either a brace or a rafter in the truss.  The third
option involves strapping looped around the pipe at the level
of the attic floor joists, and attached to the two adjacent joists;
this option is analogous to that illustrated in Figure 17A, for a
vertical pipe in a basement extending up between the over-
head floor joists. For this third option, a screw would have to
attach the strapping to the pipe, to prevent the pipe from
slipping down through the strapping (Br92). Schedule  40
piping would probably be required in this case, to prevent the
screw from causing a split in the plastic.

A fan mounting bracket, attaching the fan to the roof truss,
would be another alternative for use with straight vertical
runs.

As discussed in Section 4.6.3, any piping network should be
supported at at least one other location, in addition to near the
fan. If there is a horizontal run on a story below (e.g.,  in a
basement), support should be provided on each end of that
basement run, near the suction pipe on one end and near the
interior stack on the other, as in Figure 20. When the interior
stack extends straight up through the house from the suction
point in the slab, it is increasingly important that the system
also  be supported on the stories below, so that the  entire
weight of the stack is not suspended from the connection in
the attic. The most positive support for such an interior stack
would be at the slab, using one of the approaches in Figure 15
or equivalent. Another option would be to support the stack at
a floor penetration, e.g., using one of the approaches in Figure
17.

The weight of the fan is supported directly only in the first two
cases in Figure 24, where the strapping is looped around the
lower coupling on the fan; it would also be directly supported
if a fan mounting  bracket were used. But in the third case in
Figure 24, and in Figure 23, the support is being applied to the
piping leading to the fan; these configurations are relying on
the piping to support the weight of the fan. Where the fan is
thus not supported directly, there may be a concern that, over
time, the weight of the fan may cause it and the lower flexible
coupling to slide down over the vertical pipe below, especially
in hot climates. Such slippage could result in the vertical pipe
under the fan being shoved upward into the fan. This could
interfere with the rotation of the rotor, reducing system perfor-
mance and possibly damaging the fan. At least one mitigator
reports having encountered this problem, perhaps because the
lower coupling was not sufficiently tightened (K192); others
have not experienced it (Mes92).

To address this possible problem, one mitigator (K192) sug-
gests installing a screw into the pipe immediately below the
lower coupling, to prevent this coupling from sliding down
the pipe in cases where the fan is not directly supported. Such
a screw under the coupling is shown in Figure 23 and the third
option in Figure 24. Another option  would be to make the
vertical pipe beneath the fan short enough so that the lower
coupling could be supported by the lip of the tee fitting (or 90°
elbow) beneath.

Wiring fans in attics. The fan must be connected to a power
source in a manner consistent with the National Electrical
Code and any additional local regulations.

A common approach is to hard-wire the fan into the  house
circuitry in the manner indicated in Figure 23. Since the
in-line duct fans commonly used with SSD systems draw less
than 1 ampere of electrical current, they can usually be wired
into any convenient existing 110-volt circuit (e.g., a nearby
electrical junction box, as shown in the figure). But where the
addition of the mitigation fan causes the total connected load
                                                        156

-------
Strapping Looped
Tightly Around
Bottom Coupling,
Attached to Brace
Two Ends of Strap \
Attached Together  \
to Rafter or Brace  —^

               Staple
Strapping Looped
Around Bottom
Coupling, and the
Two Ends of the
Strap Stapled
Together
      Note: When the piping forms
      a straight vertical run, as
      depicted here, with no
      horizontal run, it would be
      advisable to ensure that the
      piping is also supported at the
      slab penetration or elsewhere
      in the livable area, as well as
      in the attic.
  Strapping Looped
  Around Pipe, Screwed
  into Pipe; Ends of
  Strap Attached to
  Adjoining Ceiling
  Joists
          Joist
                                                                                           Screw
                                     Strapping
                                     Screw
Figure 24.  Some options for supporting a fan and exhaust piping in an attic in cases where there is no horizontal piping run in the attic.


                                                     157

-------
on the circuit to exceed 80% of that circuit's rated capacity, a
separate, dedicated circuit should be installed to power to fan.

The National Electrical Code requires that a switch (to turn
the fan power on and off) be placed in the circuit within sight
of the fan. Ideally, this switch would be relatively close to the
attic access, to facilitate access by the occupant

The wiring connects to the fan inside the  electrical box
mounted on the side of the fan. The wiring leads from the fan,
through the on/off switch, to the junction box. Where codes
require  that the wiring be stapled every 18  in.,  a vertical
wooden support may have to be installed in some cases such
as lhat illustrated in Figure 23, where the wiring drops from
the truss brace to a pre-existing junction box mounted on an
attic floor joist.  No suspended length  of wire should be
installed within arm's length of an attic access door, to avoid
the risk that someone may inadvertently grab  the wire when
entering or exiting the attic (Bro92).

Because all of the wiring is in the attic, the wiring and switch
can be rated for interior use.

In some locations, codes may permit the  fan to simply be
plugged in to a nearby electrical outlet,  as shown in Figure
25A. In this case, the plug may be construed as the switch
within sight of the fan. The electrical cord will usually have to
be no longer than 6 ft for this approach  to be acceptable. In
some areas, it may never be acceptable  to simply plug in a
continuously operating appliance such as  a mitigation fan,
regardless of cord length.

If the fan cord is plugged into an existing electrical outlet in
the attic, it must be ensured that the power to that outlet will
not be interrupted. In particular, the power to outlets incorpo-
rated into attic light fixtures may go off when the attic light is
switched off.

When fans are installed by mitigators, codes generally require
that the installation of 110-volt wiring  be conducted by a
licensed electrician, if this wiring consists  of anything more
than simply plugging a 6-ft cord into a nearby outlet. To
simplify the installation in cases where the distance is longer
than 6 ft, one vendor is marketing the system illustrated in
Figure 25B, where the wiring is only 24 volts. Wiring of that
low voltage can  generally be installed without a licensed
electrician, A transformer in the wiring  where it is plugged
into the house electrical outlet steps the system voltage down
from 110 to 24 volts. In this product, the transformer box also
includes an ammeter which measures the current being drawn
the fan, thus serving as a failure  indicator. The wiring laid
between the transformer and the fan is thus 24 volts. A
transformer within the tubular fan steps the voltage back up to
110 volts, for normal fan operation. Because the wiring lead-
ing to the fan is  only 24 volts, the vendor indicates  that a
switch is no longer needed within sight of the fan; the "switch"
is now  the 110-volt plug within sight  of the transformer/
ammeter (in the basement in Figure 25B).

Exhaust piping from fans mounted in attic. An exhaust
pipe on  the pressure side of the fan must be  installed up
through the roof.  Assuming that the fan is mounted directly
beneath the point at which the roof will be penetrated, this
exhaust piping will consist of a length of straight pipe several
feet long, as shown in Figure 23.

The hole through the roof can be cut with a hole saw or a
reciprocating saw. Because the water-tightness of the roof
will have to be restored with flashing and sealant, it is desir-
able that this hole be reasonably small and neat. The hole
should be  only  slightly  larger than the pipe, to facilitate
installation and to allow for any expansion or contraction, and
should be cut as close to vertical as possible.

After the length of exhaust pipe is installed on the fan outlet,
extending up through the roof, steps must be taken to prevent
leaks around the penetration. Flashing must be installed which
fits snugly around the pipe, and which must be carefully fitted
under the existing shingles above and beside the penetration.
Some  shingles will have to be loosened  or removed and
replaced to install the flashing, a step which must be carried
out carefully in order to avoid tearing the shingles.

While caulking around  the flashing is nominally not required
if the flashing is properly installed, it is advisable to  caulk
around the perimeter of the flashing where it contacts the
adjacent shingles after  the flashing has been nailed in place.
Caulking the perimeter will help ensure against leakage dur-
ing heavy rains (Bro92). Some mitigators also recommend
caulking the heads of the nails holding the flashing in place,
and placing a bead of caulk at the seam between the flashing
and the exhaust pipe, at the top of the flashing (Mes92).

Where the penetration has been made through a flat roof, the
seam between the pipe and the roof must be thoroughly sealed
with liberal quantities of roofing tar.

The exhaust pipe should extend above the roof by a foot or
two. The primary concern in selecting stack height is to ensure
that the discharge point will not become blocked by drifts of
snow or  leaves. Taller  stacks might be expected to improve
exhaust dispersion,  helping the exhaust jet to penetrate the
boundary layer of air  flowing over the roofline. However,
there are no data demonstrating a need to make the stack taller
than 1 to 2 ft. It is likely that any reduction in re-entrainment
resulting from making the stack taller than 1 to 2 ft would be
offset by the increased  aesthetic impact the taller stack.

Where the  stack above the fan is only a few feet long, it is
usually not necessary  to provide separate support for the
stack. But where this stack is much more than about 3 or 4 ft
long, one  mitigator recommends supporting it separately,
usually with strapping connected to the members in the roof
truss (K192).

Caps on the exhaust stack. Some mitigators install a vent
cap on top of the stack, primarily to prevent precipitation from
entering  the stack. The concern is that excessive amounts of
rainwater being  channeled to the sub-slab  region via the
suction pipe might reduce sub-slab communication near the
suction pipe, thus reducing system  performance. However,
there is some question  whether rainwater entry into the pipe
will indeed be excessive in many climates. For example, a
1-in. rainfall would nominally introduce 0.05 gal of water into
                                                        158

-------
                                                                      24-to-110-Volt
                                                                      Transformer Built
                                                                      into Modified Fan
                                                                     24-Volt Cable
                                                                     (18 Gauge)
                                     110-Volt Electrical
                                     Cord, No Longer
                                     Than 6 Ft.
                                        Existing 110-
                                        Vott Electrical
                                        Outlet, Within
                                        6 Ft. of Fan
           A) Fan plugged into accessible
              existing 110-volt electrical
              outlet with cord no longer
              than 6 ft.
                                                                             24-Volt Cable, No
                                                                             Longer Than 50 Ft.
                                                                                    Control Box (110-to-
                                                                                    24-Volt Transformer
                                                                                    With Ammeter Built In,
                                                                                    or With Other Failure
                                                                                    Indicator/Alarm).
                                                                                    110-Volt Electrical
                                                                                    Cord, No Longer
                                                                                    Than 6 R.

                                                                                    Existing 110-Volt
                                                                                    Electrical Outlet
                                                                       B) Transformer used to step
                                                                          house voltage down to 24
                                                                          volts, so that cable between
                                                                          outlet and fan can be 24-volt
                                                                          cable, simplifying installation.
Figure 25.  Some alternative approaches that can sometimes be considered for wiring SSD fans mounted in attics.
a 4-in. diameter stack, an amount which will usually be small
compared to the condensate expected from the soil gas. And,
at least where aggregate is present beneath the slab, it may be
expected that any water will often drain away and not interfere
with suction field extension through the aggregate layer.

A  vent cap would also prevent leaves, other debris,  and
animals from entering and potentially plugging the stack. The
momentum of the air  leaving the stack should generally
prevent light debris such as leaves from entering the stack,
without a cap. While many mitigators report occasional prob-
lems due  to birds, bats,  squirrels, bees, or other animals
entering the stack  when there is  no cap, these experiences
appear to be infrequent (Bro92, K192, Mes92).
                                                       159

-------
A number of mitigators feel that vent caps are unnecessary
(An92, Bro92, Mes92, KI92). Caps can even be undesirable,
because many cap designs will increase back-pressure in the
system, contribute to ice buildup at the discharge point during
cold weather, and potentially reduce dispersion of the exhaust.
For these reasons, vent caps have not been included in any of
the preceding figures.
                       Two possible designs for vent caps are shown in Figure 26.
                       The design in Figure 26A, which focuses on preventing the
                       entry of rainwater into the piping, is not commonly used in
                       residential radon mitigation installations, but is widely used in
                       various other commercial applications. This design would not
                       create any back-pressure in the piping; however, neither would
                       it help prevent the entry of debris into the stack, if that is a
                       serious concern.
         Rain Cap

         PVC Stack
Supports
Attaching
Rain Cap
to Stack
                     Top View
Rounded
PVC Cover
                                                             PVC Grille
                                                                                         PVC Collar, to
                                                                                         Attach to Top of
                                                                                         4-ln. Diam. Stack
      Rain Cap —-s.     \
                    T
      RainWater—sJ

                     W
      PVC Stack
                             Rain Hits Sides of
                          \  Cap, Drips Out through
                             Annulus between Cap
                           \ and Stack
                                                                        3-Dimensional View
                                                                          .Vent Cap
                                                                 Exhaust
            A) Annular rain cap (section of
               duct having somewhat larger
               diameter than PVC stack).
                                      B) Rounded PVC cap
Figure 26.  Two possible configurations for vent caps on SSD stacks.
                                                      160

-------
Figure 26B shows a cap design more widely marketed for
radon mitigation applications. The cap itself resembles a
section of PVC piping sliced in half down its axis, and fitted at
its ends with PVC grilles. This cap is held in place on top of
the stack by a circular PVC collar at its base, which slips down
over the exhaust pipe like a fitting and can be cemented in
place. This cap should effectively prevent the entry of both
precipitation and debris but would be expected to create some
back pressure and potentially reduce dispersion.

There is also an  aluminum cap on the market. The cap has a
perforated cylindrical wall, somewhat larger in diameter than
the 4-in. exhaust pipe, with a solid plate on top. A circular
aluminum collar at its base fits tightly inside the PVC exhaust
pipe.

Another design  sometimes utilized involves placing one or
two 90° elbows on top of the stack, creating an inverted "L" or
"U"  so  that the  discharge is  directed horizontally  or down
toward  the roof. The inverted U, in particular, should be
effective at preventing entry of precipitation and debris, but
again,  would create back pressure and potentially reduce
dispersion.

One mitigator (K192) reports placing hardware cloth  screen
(1/4-in. mesh) over the stack  opening, primarily to keep
animals out. The screen is bent to fit inside the pipe, where it
will initially remain in place by pressure fit; the screen is then
secured in place by sheet metal screws. This screen should not
significantly influence system back pressure or reduce disper-
sion.

One of the potential problems  with some  cap designs is
increased back pressure on the fan. These caps serve as flow
obstructions on the pressure side of the fan, potentially reduc-
ing the suction that the system can maintain beneath the slab.
In particular, the cap in Figure 26B, the aluminum cap, and
the inverted L and U, might be expected to offer this disad-
vantage. However,  this back pressure may in fact not be a
serious problem  except in cases where flows are high, if the
fan is properly selected. One group of investigators (C191) has
found that the back pressure created by the cap in Figure 26B
will be about 0.002 to 0.02 in. WG at flows of 20 to 70 cfm.
This pressure loss will usually be insignificant compared to
the suctions typically maintained in the suction piping by the
in-line tubular fans (typically several tenths of an inch of
water, up to more than 1.5 in. WG). An inverted U, with two
elbows, would have a higher pressure loss, 0.01 to 0.2 in. WG
over the typical SSD flow range (20 to 100 cfm). At the higher
flows, this loss might occasionally be important in marginal
systems.

Another potential disadvantage of the cap in Figure 26B, the
aluminum cap,  and the inverted L is that they deflect the
upward-flowing  stream, causing it to discharge horizontally
instead of vertically upward, and reducing its momentum. The
inverted U would cause the exhaust to discharge vertically
downward, and would further reduce exhaust momentum. By
changing  the direction and reducing the momentum  of the
exhaust, these caps will reduce the potential for the exhaust
gas "jet" to penetrate the boundary layer created by air move-
ment over the roofline of the house, thus hindering dispersion
of the exhaust outside of the building wake. Intuitively, the
exhaust will probably still be  diluted sufficiently well that
re-entrainment should not be a serious problem; however,
there are no data or calculations to confirm this intuition.

Another concern regarding any rain cap is that it may contrib-
ute to the accumulation of ice at the discharge  during cold
weather, restricting flow and creating significant back pres-
sure. By providing more cold surface area and reducing the
exhaust momentum, a cap could add to the tendency of water
droplets in the exhaust to collect and freeze.

In summary, there are no data confirming that rain caps are
really necessary; from the discussion above, it would seem
that they should not be necessary in many cases. On the other
hand, neither are there data confirming that rain caps will do
any harm; perhaps the greatest risk would be one of increased
ice accumulation during cold  weather. Under these  condi-
tions, it is left to the discretion  of the mitigator regarding
whether or not to install a cap. At least one fan manufacturer
reportedly now requires a rain cap on the stack, as part of the
fan warranty, to keep precipitation out of the fan.

Mufflers on exhaust stack.  Sometimes sufficient noise is
emitted at the point where the exhaust is discharged, that some
homeowners find the sound  objectionable. This  noise could
result from: a) fan noise which is transmitted inside the stack;
and b) turbulence in the exhausted gas as it escapes. Fan noise
inside the pipe is generally  higher for the higher-powered
fans, such as the high-suction/low-flow units in Section 4.4.3.
The noise from gas turbulence would be expected to increase
as exhaust velocity increases, and to be greater when there is
an obstruction (such as  a fitting or vent cap) at or near the
exhaust point

Mufflers are marketed which can be installed in SSD exhaust
stacks downstream of the fan to reduce this noise. Mufflers
are commonly 6- to 12-in. long sections of pipe lined on the
inside  with a sound-deadening material (such  as acoustic
foam)  which "captures" some of the sound waves being
transmitted inside the piping due to the fan. Unless the muffler
increases the diameter of the exhaust stack at its discharge
point, thus reducing exhaust velocity, it is not clear whether it
will significantly reduce the  noise associated with the turbu-
lence in the escaping gas. It is suspected that much of the
noise associated with the exhaust may often be fan noise.

In the case of interior stacks with attic fans, the muffler, if
needed, would logically be installed in the attic above the fan.
Alternatively, it  could be installed outdoors after the piping
has penetrated the roof.

Mufflers are generally  required on the exhausts from the
high-suction/low-flow fans. They are not often used with the
in-line tubular fans in Section 4.4.1. However, even with the
tubular fans, situations have occasionally been encountered
where the sound from the exhaust point — although subtle - is
objectionable to  sensitive occupants. Mitigators will have to
decide on the value of a muffler on a case-by-case basis.
                                                        161

-------
Since mufflers are not commonly used with the in-line tubular
fans,  they have not been  included  in  the figures in this
document.

Considerations when fan must be on the roof. All of the
preceding discussion has assumed that the fan will be inside
an attic. Where an accessible attic exists, that will generally be
the preferred location for the fan. Where there is no attic, or
where the attic is inaccessible, the  interior stack exhaust
configuration would  necessitate that the fan be on the roof.
This situation would take away some of the advantages of the
interior stack.

If the fan must be on the roof, this will impact the design and
installation approach described above in several ways.

 1. If there is no attic, any horizontal piping runs will have to
    be inside the house or  on the roof, which can be  more
    complicated and/or  less aesthetic than locating them in
    the attic.

 2. If the roof is flat or is over a cathedral ceiling, making the
    hole in the roof for the pipe penetration  may be  more
    complicated. With flat  roofs, careful sealing around the
    penetration will be necessary, since flat roofs create an
    increased  threat  of water leakage. With cathedral ceil-
    ings, the slope may be  steeper; moreover, since the roof
    penetration may be in finished living area, the conditions
    under which the hole is drilled can be more exacting.

 3. The  fan will have to be rated for exterior use. A fan
    designed for roof or side-wall mounting may be consid-
    ered, rather than an in-line tubular fan.

 4. The  wiring  that  is outdoors  will  have  to  be
    sunlight-resistant, and  otherwise  rated for exterior use.
    Likewise, the switchbox will have to be wet-proof, rated
    for exterior use, assuming that it is outdoors (e.g., on the
    roof) in order to be within sight  of the fan. The proce-
    dures for wiring exterior fans are discussed in Section
    4.6.5.

4.6.5 Design and Installation of the Fan
and Exhaust Piping—Exterior and
Garage Stacks
General considerations with exterior and garage stacks.
One representative example of an exterior fan and stack is
illustrated in Figure 21. If a slab-on-grade garage adjoined the
basement at the point where the piping exits the basement, the
fan and stack would be inside the garage, with the  stack
penetrating the garage roof  (assuming that there is no living
area above the garage).

The decision to install an exterior or garage stack instead of an
interior stack will depend  upon the  specific  situation in a
given house, and upon the  preferences of the mitigator and
homeowner. An exterior stack will most likely be preferred
when: a) there is not a convenient utility chase leading from
the basement to the attic through which an interior stack might
be routed; b) there is not an acceptable route for directing an
interior stack up through the floor(s) above the basement into
the attic; c) there is no attic; d) there is no adjoining garage, or
it is impractical to route the stack through the garage for one
reason or another; or e) the suction pipes penetrate the founda-
tion  wall horizontally from outdoors. A garage stack will
often be preferred when: a) an adjoining garage exists; b) the
suction pipes penetrate the slab indoors (rather than horizon-
tally from outdoors); and c)  the location of the suction pipes,
the presence of living area  on the overhead floors,  or other
factors make it more convenient to route the piping through
the wall adjoining the garage than to install and exterior or
interior  stack.

When the fan and the stack are outdoors, this exhaust configu-
ration creates a greater aesthetic impact outdoors, compared
to the interior stack/attic fan configuration. However, exterior
stacks can reduce the aesthetic impact indoors, especially in
complex or highly finished houses,  and may sometimes be
preferred by the homeowner and mitigator. The exterior stack
will often be less expensive than, or comparable in price to, an
interior  stack, unless the house is particularly amenable to an
interior  stack  (one-story slabs on grade and houses  with
accessible utility chases) (He91b, He91c). The cost  increase
associated with an interior stack appears to be the least when
the mitigator's crew is experienced with interior stacks. An-
other potential concern with exterior stacks is an increased
risk restriction or pluggage due to the freezing of condensate
during cold weather; such freezing problems have been re-
ported in cold climates.

Where there is an adjoining garage, locating the fan and stack
in the garage can offer the advantages of both  the interior
stack and the exterior stack. Like the interior stack, the garage
stack will essentially eliminate the aesthetic impact outdoors;
it will keep the fan and wiring in a protected location; and it
will  usually avoid the risk of stack  freezing during  cold
weather. Like the exterior stack, the garage stack will avoid
aesthetic impacts inside the living area. A garage stack may be
roughly $100 more expensive than an exterior stack because
of the firebreak that is often  needed where the pipe penetrates
from the basement into the  garage and because of the costs
associated with penetrating the garage roof (a cost which the
exterior stack avoids unless it penetrates the roof overhang,
also  discussed later) (He91b, He91c). A garage stack may be
either more or less expensive than an interior stack, depending
upon the particular house characteristics and the experience of
the mitigator with interior stacks.

Selecting the location of exterior or garage stacks. The
criteria for deciding where the exterior or garage stack will be
located have been discussed in Section 4.6.2 (see Selection of
where the piping exits the basement).

An additional consideration  in locating exterior stacks is that
the risk of the stack becoming plugged with frozen condensate
during cold weather can be reduced if the stack can be located
on the south side of the house, where it will be exposed to
more sunlight (K192).

Selecting the location of exterior or garage fans. Where
there is an exterior stack, EPA's standard specifies that the fan
must be outdoors (EPA91b), as emphasized previously. With
garage stacks,  the fan must be in the garage. Location of the
                                                       162

-------
fans outside the livable area is intended to reduce the risk of
increases in radon levels if leaks develop over the years in the
fan housing or piping on the pressure side of the fan.

Exterior fans are commonly mounted near grade level imme-
diately outside the foundation wall, at the base of the exhaust
stack, as shown in Figures 1 and 21. This location will reduce
the visual impact of the fan,  compared to the alternative of
mounting it higher in the stack, and will improve the ability to
potentially conceal the fan behind shrubbery. Location near
grade also facilitates access to the fan for maintenance, and
may simplify the electrical wiring. Also, in view of the fact
that aluminum downspouting is sometimes used for exterior
stacks for aesthetic reasons, attempting to mount the fan on
top of the artificial downspout would represent an added
complication.

One potential disadvantage of having the fan at the base of
exterior stacks is that, in cold climates, ice that has formed
inside the necessarily uninsulated stack during the winter may
become dislodged and fall down into the fan during thaws,
preventing proper system operation and potentially damaging
the fan. One mitigator who  has  occasionally observed this
problem  reports  that although the fallen ice did  sometimes
prevent rotation of the fan rotor and hence interfere with
system performance, it did  not  seem  to damage the fans
(K192).

With garage stacks, one potential disadvantage of having the
fan near the slab is that any pressure-side leaks would release
high-radon exhaust near the living area. However, unless the
garage is conditioned space, the garage is usually sufficiently
well isolated from the living area (and the amount of exhaust
released  will usually  be sufficiently small) such  that this
concern should not necessitate that the fan be moved up to the
garage attic, if placement in the attic is otherwise not desired.

If the garage does have an accessible floored attic, location of
the fan in the garage attic could be considered. If the fan is
located in a garage attic, selection of the exact location within
the attic  would be made as  discussed in Section 4.6.2 (see
Routing considerations with  attic piping runs) and Section
4.6.4.

Criteria for selecting  exterior and garage fans.  The
selection  of the appropriate  fan  to achieve adequate SSD
performance has been discussed in Section 4.4. In the figures
and discussion in this section, the focus  is on  the in-line
tubular fans covered in Section 4.4.1. If another type of fan
were used, the details regarding the mounting and support of
the fan would have to be adjusted accordingly, for that other
type of fan.

With exterior fans and stacks, the fan should be UL-rated for
exterior use. This will mean that the wiring  box and connec-
tion on  the side of  the fan will have  to  be properly
weather-proofed. With fans  mounted in garages, rating for
exterior use is unnecessary.

Mounting exterior or garage fans. In accordance with the
discussion in Sections 4.6.2 and 4.6.3, the system piping will
have  been installed through the basement band joist and
exterior finish to the outdoors (for exterior stacks), or through
the foundation wall below the exterior finish. With garage
stacks, the piping will have been installed through the base-
ment band joist and the fire-rated sheetrock in the garage;
also, the fan wiring could be somewhat different in garages
because it will be in an enclosed space, as discussed later.

A detailed diagram illustrating a representative technique for
mounting an exterior fan is presented in Figure 27. This figure
is for the case where the suction pipes are installed inside a
basement, and  is a detailed enlargement  of  a portion the
system shown in  Figure 21. If the suction pipes penetrated
horizontally from outdoors, of course, the portion of the figure
below the fan would look as shown in Figure 2. If the fan and
stack were in a garage, the pipe from the basement would be
penetrating fire-rated sheetrock on the exterior face of the
wall, rather than brick veneer as shown.

As  discussed earlier in Section 4.6.4 (see Mounting fans in
attics), the fans should be mounted vertically, so that conden-
sate and rainwater will drain out of the fan. As shown in
Figure 27, a 90° elbow is usually installed on  the end of the
horizontal pipe penetrating through the house shell. The fan is
then mounted on  a vertical segment of piping extending up
from this elbow.

For convenience in supporting the fan, some mitigators have
suggested mounting exterior fans horizontally,  directly on the
pipe penetrating the house wall. By this approach, a drain hole
perhaps  1/4 to 1/2 in. diameter would be drilled at the low
point in the fan housing, on the underside of the fan, to permit
water to drain out. Based upon simple orifice  calculations, a
hole of this size should result in a maximum leakage into the
system of only a few cfm of air, and thus  should not create
excessive suction loss with the in-line tubular fans. However,
much experience suggests that, even though the  soil gas
passing through the fan will be at a temperature well above
freezing even during the winter, the drain hole can still ice
shut in outdoor fans during extended cold weather. If the drain
hole becomes plugged with ice, condensate will accumulate in
the fan housing. In addition, horizontal mounting will void the
warranty of some fans. Thus, vertical mounting is recom-
mended.

The fan is commonly mounted on the vertical pipe using
flexible PVC couplings, to help reduce the vibration transmit-
ted to the piping by the fan. See the discussion under Mount-
ing fans in attics in Section 4.6.4. The flexible couplings are
clamped tightly to the fan and to the pipe using circular hose
clamps, to avoid leakage. As in other figures in  this document,
Figure 27 illustrates a 4-in.-to-6-in. coupling, connecting a
tubular fan with 6-in. couplings onto 4-in. piping. The cou-
pling size would vary if the fan or pipe size varied.

Figure 27 shows  a short segment of 4-in.  piping below the
lower flexible coupling (between the coupling and the 90°
elbow), and another such segment above the upper coupling.
These segments should be sufficiently long so  that, when the
couplings are loosened, the couplings can be slid up or down
as necessary to remove the fan for repair or replacement. With
interior stacks, the length of the pipe stub under the attic fan
will sometimes be of  less concern since, at  least  with the
                                                        163

-------
     Female End Up at
     Seams Where Segments
     of Downspout Are
     Joined, to Reduce
     Risk of Ice Buildup
Notes:
1.
     Caulk
     3"X4" Rectangular
     Metal Downspout

     Shim Between
     Bracket and
     House, If Needed"

     Standard Downspout
     Bracket, to Attach
     Downspout to Side
     of House
     4" Round-to-3X4"
     Rectangular
     Adapter

     Regular 45° Fittings
     (4" Dla. Round PVC)
     to Offset Stack,
     Against House'
     Fan Rated for
     Exterior Use,
     or Enclosed

     110-VoIt Wiring,
     Rated for
     Exterior Use
             Wet-Proof Switch
             Box, Near Fan1
2.
For clarity, switch box is shown below
pipe penetration through wall. More
commonly, switch box will be up beside
fan or pipe penetration. Wiring from
switch to interior junction box may
penetrate house through separate
hole through band joist, as shown
here, or may enter via same
penetration as suction pipe, outside
the pipe.

Other configurations for the fittings
and piping immediately above the fan
are illustrated later.
3. Other options for attaching the stack
   to the house are presented later.

4. Shim between bracket and house may be
   needed if piping configuration or
   irregularities on face of house force
   stack to be too far away from house.
   Male end up at downspout seams would
   create a lip inside the stack against
   which condensate running down
   interior might collect and freeze in
   cold climates.
                                                                              Existing 110-Volt
                                                                              Junction Box
Figure 27. A representative method for mounting an exterior fan at the base of an exterior stack.


                                                    164

-------
configuration shown in Figure 23, the short length of stack
above the fan could be removed  with the fan if necessary.
However, with exterior stacks, which will be firmly attached
to the side of the house, it is important that the fan be able to
be removed without movement in the piping either above or
below.

If the particular fan model being used  does  not  have an
essentially air-tight housing, any leaks in the fan housing must
be detected and effectively caulked.

As one variation of the mounting configuration in Figure 27,
an ABS  plastic enclosure is being marketed which can be
mounted around exterior fans to  improve appearance  and
provide some protection. This housing screws against the side
of the house. It encloses the fan and its couplings, the elbow
and the wall penetration below the fan, and the fittings above
the fan (which serve to offset the stack against the side of the
house). The commercially available enclosure comes with a
"transition box" which offsets the stack  against the house,
taking the place of the fittings above the fan; otherwise, the
mounting of the fan inside the enclosure is essentially identi-
cal to that illustrated in the figure.

The mounting configuration illustrated in Figure 27 repre-
sents the in-line tubular fans and radial blowers discussed in
Sections 4.4.1  and 4.4.2. The DynaVac HS series of
high-suction/low-flow fans listed  in Table 3 (Section 4.4.3)
bolt directly onto the side of  the house when used with
exterior stacks. The Pelican high-suction fan is not designed
for use with exterior stacks.

Support for exterior and garage fans. The weight of the
fan must be adequately supported. The weight of the stack
above the fan must also be supported separately; support for
the stack is discussed later (see Exhaust piping from exterior
fans below).

The horizontal pipe extending through the wall below the fan
will be supported at its penetration through the band joist, and
by hangers or strapping inside the basement This pipe should
adequately support the weight of an in-line fan, if the pipe
extends out from  the house only as far as is required to
provide clearance between the fan housing and the house.

Many mitigators find that no further support ifor exterior or
garage fans is needed in many cases. However, for additional
support,  some mitigators have  sometimes attached the fan
housing directly to the side of the house using a fan mounting
bracket. If this is done, the bracket must be attached tightly to
the house, to reduce the transmission of vibration; if there is
wood siding, the bracket should be screwed into studs in the
wall. As another way of providing additional support, some
type of support could be placed between the ground and the
elbow beneath the fan, so that the fan weight is borne by the
ground. Some mitigators have sometimes accomplished this
by installing a T-fitting under the fan, instead of the 90° elbow
in Figure 27, with one leg directed upward and one down-
ward; the fan is then mounted on the  upward leg, while a
length of rigid pipe is installed into the downward leg and is
embedded  in the  ground under  the fan.  On occasion, as
another alternative for providing additional support, the fan
has been attached to a post embedded in ground beside the
fan.

Since these additional support features for the  fan are not
usually necessary, they are not shown in Figure 27.

Wiring exterior and garage fans. Considerations involved
in the electrical wiring for fans have been discussed in Section
4.6.4 (see Wiring fans in attics).

When the fan is inside a garage, the wiring considerations will
be very similar to those for attic fans. The fan can be hard-wired
into a convenient existing house circuit with an on/off switch
in sight of the fan; since, the wiring is inside the garage, the
wiring and switch can be rated for interior use. Where codes
permit, the fan may be plugged into a nearby electrical outlet
using a cord no longer than 6 ft.

Where the fan is outdoors, more care in the wiring is required.
See Figure 27. To be within sight of the fan, the on/off switch
will  often have to be mounted outdoors on the side of the
house near the fan, as in the figure. In some cases, the switch
is located indoors, in sight of the fan. Commonly, an outdoor
switchbox will be mounted at an accessible location close to
the electrical box on the fan, so that a short length of outdoor
electrical conduit can connect the fan to the exterior switchbox
with a minimum run of the conduit along the side of the house.
Runs of conduit along the face of the house should be stapled
every 18 in.

Wiring from an outdoor switchbox penetrates the house shell
and is hard-wired into the house circuitry (e.g., at an existing
junction box in the basement, as in Figure 27). Li the figure,
the wiring penetrates the wall through a separate hole drilled
especially for the wiring. Alternatively, if there is a gap
around the 4-in. hole that has been drilled for the exhaust pipe,
the wire might enter the house through this gap between the
outside of the pipe and the wall. Such a gap may exist, for
example, when  the hole has  been made  through a  block
foundation wall with a rotary hammer drill, and is thus slightly
irregular. The National Electrical Code prohibits running
electrical wiring inside air ducts;  thus, it would be against
code to route the wiring into the house inside the SSD piping,
drilling a small hole in the pipe just outside and just inside the
house to accommodate the wire.

Any wiring or conduit used outdoors must be rated for exte-
rior  use. Likewise, if the switchbox is mounted outdoors, it
must be weather-proof and rated for exterior use.

Exhaust piping from exterior fans—selection of stack
material. When an exterior fan is mounted near grade level
as in Figures 21  and 27, an exhaust stack must extend verti-
cally above the fan, if the exhaust is to be at least 10 ft above
grade in accordance with EPA's interim standards (EPA91b).
This exterior stack generally extends to or above the eaves.
There are numerous options in designing exterior stacks.

The exterior stack can be PVC, PE, or ABS piping of round
cross-section, like the remainder of the piping in the system.
The diameter of the stack will commonly be the same as that
in the rest of the system. The back-pressure created by this
                                                        165

-------
stack will have the same impact on sub-slab suctions as would
an equivalent length of piping on the suction side of the fan.
Use of round piping for the stack has the greatest aesthetic
impact among the alternatives, since it will clearly look like a
stack. Also, UV-resistant Schedule 40 pipe must be used for
exterior stacks, to reduce embrittlement over time. Painting
the stack would help address both of these concerns.

Instead of round pipe, some mitigators commonly used alumi-
num rain gutter downspouting for the stack, so that the stack
resembles a downspout. Where this is done, it is advisable to
use downspouting of 3- by 4-in. cross-section, rather than the
2- by 3-in. downspouts common in residential applications.
The larger downspouting has a cross-sectional area similar to
that  of  4-in, diameter piping, thus  reducing system
back-pressure and flow noise compared to the smaller down-
spout. Aluminum downspout stacks can be noisier than PVC
stacks due to vibration, unless care is taken during installa-
tion, as discussed later (Bro92, K192). Aluminum downspout
stacks may also be more prone to condensate freezing during
cold weather (K192); use of the downspouting with the larger
cross section should reduce the risk of complete pluggage
with ice.

As another alternative, PVC piping of rectangular cross-section
is being marketed.  This rectangular piping  is  intended  to
resemble a downspout once painted to match the house.

Exhaust piping from exterior fans—offset against
house. Whatever material is used for the stack, the piping
immediately above the fan is almost always offset toward the
house, so that the  stack will be against  the  house. This
improves the appearance and facilitates supporting the stack.
Even with this offset, the stack may have to stand at least a
fraction of an inch away from the house, due to the dimen-
sions of the fittings and obstructions on the face of the house.
For example, if the fan is mounted beside a brick veneer but
the story above has wood siding (which will protrude out from
the veneer), the stack will have to stand off from the veneer by
a fraction of an inch in order to clear the siding above.

There is a wide variety of methods for achieving  this offset, a
few of which are discussed here.

In Figure 27, a pair of regular 45° fittings of 4-in. diameter
PVC (or PE or ABS) are installed above the fan, with short
lengths of straight 4-in. PVC piping in between as necessary
to mount the fittings. Since the figure assumes that the stack is
3- by 4-in. aluminum downspouting, a PVC adapter fitting is
shown above the upper 45° elbow, to serve as the transition
between  the  round  piping and the downspout.  The PVC
fittings and piping would be tightly cemented together. The
bottom of the downspout would be fit inside the 3- by 4-in.
opening in the PVC  adapter, being bent or crimped as neces-
sary to accomplish a good fit; the seam between the PVC and
the aluminum would then be effectively caulked.

Some of the  other alternatives for achieving the  offset are
illustrated in Figure 28. The first option is almost identical to
that in Figure 27 (a pair of 45° fittings and an adapter with a
downspout stack). The difference is that, in this case, the 45°
elbows are "street" fittings rather than regular fittings. With
street finings, one end of the fitting is narrowed so that its
outside diameter becomes the same diameter as the outside
diameter of the straight lengths of piping; normally, the fit-
tings are about 1/2 in. wider. The use of street fittings enables
the two 45° fittings to be cemented together and to the adapter
above without the need to insert a short length of straight pipe
(perhaps 4 in. long) in between.

The second option in Figure 28 differs from that in Figure 27
in that two regular 90° fittings are used above the fan, instead
of 45° fittings. The use of 90° fittings requires that the fan be
slightly farther away from the house.

The third option differs from the second in that the adapter to
convert from the round fittings to the 3- by 4-in. downspout-
ing is omitted. The rectangular aluminum downspout is crimped
as necessary to  fit directly into the round fitting, and the
residual gap is carefully sealed with caulk. When the bottom
of the 3- by 4-in. downspouting is properly bent, it reportedly
fits reasonably well into a round 4.5-in. i.d. fitting.

The fourth option in Figure 28 differs from the others in that is
shows the use of 4-in. round PVC piping as the stack, rather
than 3- by 4-in. downspouting.

Exhaust piping from exterior fans—connection of stack
segments. Where PVC  (or PE or ABS) piping of round or
rectangular cross-section is used as the exterior stack, the
segments of piping are connected with straight fittings/cou-
plings. The piping and fittings are cemented tightly together,
as is done with piping runs'  inside the house. Assuming that
the stack is above  the fan, and hence is operating under
pressure, cementing the joints  is necessary to prevent the
escape of high-radon exhaust gas  immediately beside the
house through leaks between piping segments. It  also im-
proves the physical integrity of the stack.

When aluminum downspout is used as the stack, some mitiga-
tors always or sometimes seal the seams between segments
using urethane caulk (Bro92). Some mitigators rarely caulk
the seams, relying on a tight pressure fit between the segments
(K192). Others sometimes caulk seams,  when the seams are
near openings in the house shell (Mes92). Downspout stacks
may be inherently somewhat leakier than PVC stacks, be-
cause the joints between segments cannot be cemented and
can be difficult to effectively and permanently seal (Bro92).

In cold climates, downspout stacks may have an increased risk
that condensate will freeze inside the stack. To reduce the risk
of freezing, one mitigator recommends that segments of down-
spouting be joined in a manner such that the wide end of lower
segment joins  with  the  narrow  end of  the upper segment,
rather than, vice-versa. See Figure 27. If this were reversed, so
that  the  narrow end of the lower segment were  pointing
upward inside the upper segment, a lip would be created;
condensate running down inside the stack could collect on this
lip, possibly increasing the risk of freezing.

Exhaust piping from  exterior fans—support of stack
against house.  The exterior stack must be effectively sup-
ported against the side of the  house,  to prevent  it from
dropping and to maintain its integrity.
                                                       166

-------
     3"x4" Rectangular
     Metal Downspout

      4" Round-to-3x4"
      Rectangular
      Adapter

  45° Street
  Fittings (4" Dia.
  PVC) Avoiding
  Need for
  Segments of
  Straight Pipe
  Between the
  Fittings
 3"x4" Rectangular
 Metal Downspout
 Forced Directly into
 4" Round Adapter;
 Gaps Caulked Well

    Regular 90°
    Fittings (4"
    Dia. Round PVC)
    3"x4" Rectangular
    Metal Downspout

    4" Round-to-3x4"
    Rectangular
    Adapter

Regular 90
Fittings (4"
Dia. Round
PVC)
    4" Dia. Round
    PVC Stack	
  Regular 45°
  Fittings (4"
  Dia. Round PVC)
Figure 28. Some of the alternative methods for offsetting exterior stacks against the house.

                                                   . 167

-------
A wide variety of approaches have been used for supporting
the stack, depending upon the whether the stack is PVC or
aluminum downspouting. Some of the methods that different
mitigators have reported using are shown in Figure 29.

Whatever support method is used, several such supports are
likely to be needed, at intervals along the stack run. At least
one support will be needed for each stack segment; some-
times, a support may be warranted at each end of a segment. A
support will definitely be required toward the upper end of the
top segment, close to the point where the stack extends
outward around the roof overhang; routing the stack around
the overhang is discussed later. Some mitigators recommend
that the lowest support be some distance above the fan, in an
effort to reduce the amount of fan vibration transmitted to the
siding (K192).

Minimizing the transmission of vibration noise to the house is
a possible concern any time a rigid support is used (i.e., hi all
of the cases  in Figure 29 except for the strapping in Figure
29D). It is also of concern when the stack is in direct contact
against the house; this may be particularly common when the
stack is downspouting. Transmission of vibration noise seems
to be worse with downspout stacks than with PVC stacks
(Bro92). In addition to placing the lowest support a sufficient
distance above the fan, other approaches that have sometimes
been used in an effort to reduce vibration noise include:

a)  Placing a cushioning material between the stack and the
    rigid support (Figures 29A, 29B, and 29F). If there is a
    shim between the stack and the house to cause the stack
    to stand off from the house as discussed below and if the
    stack is drawn tightly against the shim,  a cushioning
    material might be placed between the stack and the shim.

b)  Placing  a cushioning material between the support and
    the house  (Figures 29C and 29E), or between the shim
    and the house. This may be somewhat less effective than
    a) above, since, once vibration has been transmitted to the
    support, it may still get through to the siding via the
    screws which will necessarily penetrate any cushion.

c)  Using strapping, as in Figure 29D. Strapping will be less
    likely to transmit vibration, in cases where it can be used
    without drawing the stack tightly against the house.

d)  Avoiding direct contact between the stack and the house,
    except via cushioned supports and shims. However, di-
    rect contact can improve the integrity and appearance of
    the stack, and is especially common for downspout stacks.
    Supports such as those in Figures 29A and 29D usually
    result in direct contact, with the stack drawn tightly
    against the house.

e)  Supporting the stack tightly against a surface that is less
    likely to vibrate, such  as brick veneer rather than wood
    siding, where this is an option.

In cases where the side of the house is basically flat, as with
uninterrupted brick veneer, the stack may often be installed
flush against the house, depending upon the type of support
used. Placing the stack flush against the house will be most
common with downspout stacks, one possible reason why
downspout stacks sometimes seem to be noisier than PVC
stacks. In these cases, Figure 21 and the other figures may be
somewhat misleading, in that they suggest that the downspout
stack is a couple of inches away from the veneer.

In other cases, there may be obstructions along the wall which
will require that the stack stand off from the wall by a fraction
of an inch or more. One example would be a two-story house
with veneer on the lower story and wood siding on the upper;
the siding would extend a fraction of an inch out from the
veneer. In such cases, the stack may have to stand off from the
veneer in order to clear the siding. Some stack supports, such
as those in Figures 29C and 29E, inherently create a stand-off.
Other supports, such as those in Figures 29A, 29D, and 29F,
tend to draw  the stack tigMy against the house. With such
supports, a shim of appropriate thickness could be used to
create  the necessary  stand-off. Shims would commonly be
pieces  of treated wood which are inserted between the house
and the support. Shims are shown in Figures 29B and 29F.

It will commonly be appropriate for a shim to extend laterally
all the  way behind the stack, from the right-hand connection
between  a given support and  the house,  to  the left-hand
connection. In such cases, the support will usually draw the
stack tightly against the shim. Figures 29B and 29F do not
show the shim extending in this manner. Rather, they reflect a
case where there are  two small shims for each support, one
located on each side of the stack where the bracket or clamp
attaches to the house.

The support methods shown in Figure 29 reflect some of the
more common alternatives.  Figure 29A  is a standard metal
bracket that fits around downspouts, of the type commonly
used to attach the downspouting from rain gutters. Figure 29B
is the same type of bracket, but with a shim  to serve as a
stand-off. Figure 29C is a standard downspout bracket of a
different design, which fits entirely behind the downspout
rather than wrapping around it. Figure 29D shows a continu-
ous loop of plastic strapping, wrapped around  the stack and
drawing it tightly against the house. Figure 29E represents a
two-piece metal pipe clamp; the two sections bolt together on
the outside of the stack, and are held by a metal channel which
screws into the side of the house. Figure 29F shows standard
galvanized metal strapping which extends tightly around the
stack and screws into the wall on each side; in this example, a
shim is shown to provide a stand-off.

There are a variety of other possible approaches. With down-
spout stacks which are flush against the house,  one mitigator
sometimes screws the downspout directly into the house,
inserting a screw through the rear of the downspout from
inside the downspout, near the joint between segments (Bro92).
Another option would be to screw a galvanized metal strap
vertically up the side of the house, directly behind the stack;
the stack could then be attached  to  this strap, e.g., using
stainless steel hose clamps tightened around the stack and the
metal strap (Bro92).

Where the support is being attached to siding, galvanized
screws suitable for outdoor use would be an appropriate
choice. Where  the support is  being  attached to  veneer, a
                                                       168

-------
 Bracket
                          Cushion Between Bracket
                          and Downspout to Reduce
                          Vibration Noise?
                                      Screw into
                                      Side of
                                      House
                                     3"X4"
                                     Downspout
                                                      Cushion?
                                                              Shim, If
                                                              Necessary
        A) Standard sheet metal bracket
           around 3"X4" metal downspout
           stack.
                              B) Standard sheet metal bracket
                                around 3"X4" downspout with shim
                                between bracket and house to
                                permit stack to stand off from
                                house as necessary to clear
                                irregularities on face of house.
                                     Downspout
        C) Standard sheet metal bracket
           on back of 3"X4" metal downspout
           stack.
                                                           One End of
                                                           Strapping
                                                           Attached to
                                                           Side of House
                                                           on Each Side
                                                           of Stack

                                                           Continuous
                                                           Loop of
                                                           Plastic
                                                           Strapping
                                                             D) Plastic strapping
Bolt
Attaching
Two Clamp
Pieces
 One Piece of
 Two-Piece
 Metal!
 Clamp
4" Dia.
Round
PVC Stack
                                                         Cushion?-
                                        Cushion?
-Metal Channel into
 Which Pipe Clamp
 Locks. Screwed into
 Side of House
        E) Two-piece metal clamp around
           circular stack.
                              F) Metal clamp or strapping
                                around circular stack.
                                                         Shim, If
                                                         Necessary for
                                                         Stand-Off
                                                                                         Metal Clamp
                                                                                         or Strapping
Figure 29. Some alternative methods for supporting an exterior stack against the side of a house.


                                                   169

-------
masonry screw or nail could be used. Or a lag bolt could be
used, with a hole being drilled in the brick to accommodate
the lag, and the screw then screwed into the lag.

Exhaust piping from exterior fans—routing stack
around or through roof overhang. Where an exterior
stack reaches  the roof overhang,  there are two options in
configuring the stack: to angle the stack outward around the
overhang and rain gutter, extending the stack above the eave
outside the gutter; or to penetrate straight up  through the
overhang. These options are illustrated in Figure 30, for the
case where the stack is 4-in. diameter PVC piping.

Routing the stack out and around the overhang is the easiest
option for exterior stacks. It will always be the choice when
downspouting is used for the stack. Penetration of the over-
hang is the more expensive  option, but  may  offer better
appearance when the stack is PVC piping. Penetration of the
overhang may not be advisable in climates  where snowfall is
heavy; ice accumulation behind the stack  during the winter
may have sufficient weight to damage the overhang.

Figure 30A shows one approach  for supporting the stack
when it is routed around the overhang. Some of the alterna-
tives are shown in Figure 31.

Whatever option is used for  routing the  stack around the
overhang, the best discharge approach will be to discharge
vertically upward above the eave,  as illustrated in all of the
options in  Figures 30A and  31. Discharging upward  will
probably best accomplish the goal of reducing re-entrainment
back into the house. Some mitigators delete the last vertical
length of stack shown in the figures, discharging horizontally
(or nearly horizontally) away from the house near soffit level.
This horizontal soffit-level  exhaust would meet EPA's in-
terim guideline requiring discharge at least 10 ft above grade
(EPA91b); however, it is unclear that it could be as effective
as the vertical-upward exhaust in helping avoid entrainment
of the exhaust gas in any air circulation zone beside the house
beneath the eave. While the horizontal soffit-level exhaust
directed away from the house can meet EPA's standard, it
would seem that — since the cost and aesthetic impact of
installing an exterior stack have been incurred — it would be
advisable to install the last vertical segment of stack that will
help ensure that the desired benefits of that exterior stack are
in fact realized.

Each of the options in Figures 30A and 31 show the vertical
exhaust at  a point several  inches above the bottom roof
shingle (and 6 to 12 in. above the lip of the rain gutter). This
height is necessary to ensure that water overflowing the rain
gutter does not enter the top of the stack, and that snow,
leaves, or other  debris accumulating on the roof or in the
gutter do not enter or block the stack opening.

When the stack is of PVC, one approach (illustrated in Figure
30A) is to bring the stack up just below the soffit, then to
direct it horizontally outward with a 90° elbow. This horizon-
tal run would have to be at a sufficient distance below the
soffit to clear the fascia, which usually extends lower than
does the rain gutter. Once this horizontal  run has  extended
beyond the rain gutter, a second 90°  elbow directs the stack
    Horizontal Run Offset Far
    Enough Below Soffit to
    Clear Fascia Board and
    Rain Gutter; Sloped Down
    Toward Side of House.
      Metal Clamps
            4" Dia.
            PVC Pipe
      Straight Fitting   j—li
      (as Necessary)   ^	'
                             777.
        A) PVC stack extends horizontally
           under soffit and around rain
           gutter
Note: Penetration of the overhang
may not be advisable in regions
having heavy snowfall, where ice
accumulation on the roof behind
the stack might have sufficient
weight to damage overhang.
         Metal Clamp -
        B) PVC stack penetrates soffit
           and roof

Figure 30.  Options when an exterior stack reaches the roof over-
          hang.
                                                       170

-------
Screw Attaching
Strapping to
Gutter
  Plastic
  Strapping
   Screw
                           Shingles -
         3x4 In.
         Downspout
         Stack
  Nylon Tie
  or Plastic
  Strapping
  4" Dia.
  PVC Stack
Metal Clamp
(Example) -
i— Shim, If
   Necessary
        A) Plastic strapping looped through slit
           in outer lip of rain gutter, or screwed
           onto lip of gutter.
      B) Nylon tie or plastic strapping nailed
         to roof under shingle.
        Note: This approach may
        increase vibration noise
        by causing vibration in
        rain gutter.
       Screw
                                       3x4 In. Metal
                                       Downspout
Shim to Offset
Vertical Exhaust
Pipe as Necessary
to Clear Shingle
Overhang
                                                          Metal
                                                          Clamp
                                                                                                     Metal
                                                                                                     Clamp
                                                                     Shim to Offset
                                                                     Horizontal Run as
                                                                     Necessary to Clear
                                                                     Fascia Board
        C) Metal downspout stack screwed
           directly into lip of rain gutter.
      D) Stack attached directly to fascia
         with clamp, when there is no gutter.
Figure 31.  Some alternative methods for supporting exterior stacks where they are routed around a roof overhang.


                                                     171

-------
upward. As usual, the horizontal run should be sloped slightly
downward toward the house to prevent accumulation of con-
dcnsate or rainwater at low spots.

To support this extension of the stack away from the house, it
is advisable to support the vertical riser against the house as
close to the soffit as possible. In Figure 30A, this is shown
being accomplished using a clamp (e.g., as in Figure 29E) or
galvanized metal strapping (as in Figure 29F) attached to the
veneer just below the first elbow. For added security, Figure
30A also shows a second clamp or metal strap supporting the
horizontal run from the soffit overhead. Because the horizon-
tal run must be offset below  the soffit in order to clear the
fascia, a shim is shown between the horizontal pipe and the
soffit. In practice,  this offsetting shim would extend all the
way from the soffit to the pipe; in this regard, the depiction in
Figure 30A is misleading.

The final vertical segment of the stack extending above the
gutter is also commonly supported from the roof or gutter. In
Figure 30A,  this is accomplished by a continuous loop of
plastic strapping that loops tightly around the stack and is
screwed into  (or inserted through a slit in) the lip of the rain
gutter. Another screw through the strapping into the PVC pipe
prevents the pipe from slipping down through the strapping
loop over time. The strapping (as well as the downspouting)
can be painted to match the existing house finish.

Figure 3 IB shows another possible approach for supporting a
PVC stack  near the gutter. In this case, nylon tie material
(perhaps 3/8 in. wide) or plastic strapping is looped tightly
around the stack and is nailed into the roof beneath the lower
shingle.

If there is no rain gutter, as in Figure 3 ID, the stack can be
attached directly to the fascia board (in this example, using a
clamp or galvanized metal strapping). Since the lower shingle
can extend out from the  fascia by an inch or more, the stack
will have to stand off from the fascia by that distance. For this
reason, a shim is shown in the figure to create the stand-off.
(In practice, the shim would extend all the way from the fascia
to the pipe.) Alternatively, one could notch the lower shingle
to accommodate the stack, avoiding the need for a stand-off
shim. However, a mitigator is probably well advised not to do
anything (such as notching a shingle) which might cause (or
be perceived to cause) subsequent water problems on the roof,
unless it is absolutely necessary.

When downspouting is used as the stack, the vertical down-
spout against the house will angle outward and upward away
from the house at a point a couple of feet below the soffit. It
then bends back to the vertical as soon as it has cleared the
rain gutter. This situation is depicted in Figures 21,31A, and
31C. As with the PVC stack, the downspout stack should be
attached tightly to the house just below the point at which it
bends away from the house. The final vertical segment should
also be attached to the rain gutter or roof. For example, it can
be attached to the gutter using a tight loop of plastic strapping
screwed into (or inserted through a slit in) the lip of the gutter,
as illustrated in Figures 21 and 31 A. This is  the  same ap-
proach shown for PVC pipe in Figure 30A. As other options,
it could be attached using methods analogous to those shown
for PVC pipe in Figures 3 IB  and 3 ID. Another possible
alternative, screwing the downspout directly into the gutter, is
illustrated in Figure 31C. This latter alternative could increase
the amount of vibration noise transmitted to the gutter.

Where an exterior stack penetrates straight up through the
roof overhang (Figure 30B), the stack is supported against the
side of the house, up to the soffit The stack must be posi-
tioned so that it will penetrate the soffit and roof in between
the joists and rafters. The hole through the soffit and the hole
through the roof can be cut with a hole saw or a reciprocating
saw, with much care taken to align these two holes.

The procedure for installing the stack through the overhang
penetration, and for restoring the water-tightness of the roof,
is the same as that described previously for installing the
exhaust from an attic-mounted fan  through the roof.  See
Exhaust piping from fans mounted in attic in Section 4.6.4.
Some caulking and finishing around the hole through the
soffit would be desirable to improve the appearance. As with
the attic-mounted fan, the exhaust stack should extend above
the roof by a foot or two.

Exhaust piping from exterior fans—finish around the
stack. Where the exterior exhaust stack is downspouting, the
final  finish will usually involve painting the downspouting
and the downspout supports to match the existing rain gutter
downspouts.

Where the exterior stack is Schedule 40 PVC piping, painting
the stack (or coating it with a UV-protectant) will be even
more important. Not only will the painting improve appear-
ance, but it will provide some protection from UV damage.
Another method for protecting the exterior stacks and improv-
ing their appearance is to box them in, and  to paint the
resulting enclosure to match the house. The stacks could be
framed and boxed  in using exterior wood.  One mitigator
(Jo91) reports  using pre-fabricated  three-sided aluminum
chases with flanges on the open side, obtained in 10-ft lengths.
These chases can be attached to the side of the house covering
the stack, avoiding the need for carpentry.

Considerations when SSD pipes penetrate horizontally
from outdoors. The preceding  discussion of exterior stacks
under Section 4.6.5 has generally assumed the case where the
SSD suction pipes have been installed vertically down through
a basement slab from indoors, and where the piping has then
penetrated the basement band joist to an exterior fan  and
stack.

Where the SSD suction pipes penetrate horizontally through
the foundation wall from outside the house as in Figure 2, the
fan and the stack will necessarily be outdoors. This situation is
most  likely to be encountered in slab-on-grade houses or in
walk-out basements.

The detailed considerations in selecting, mounting, and wir-
ing the fan, and in designing the exterior stack, are exactly the
same for horizontal exterior SSD pipes as those described
above for vertical interior SSD pipes. The only difference is
that the exterior fan will be mounted on a riser extending
upward from the horizontal exterior suction pipe as in Figure
                                                       172

-------
2, rather than being mounted on a segment of pipe extending
out through the foundation wall as in Figure 27.

Exhaust piping from garage fans—general consider-
ations. When the stack is inside a garage, the considerations
in designing and installing the exhaust piping above the fan
will differ somewhat from those discussed above for exterior
stacks.

As  discussed previously in Section 4.6.5 (see Selecting the
location of exterior or garage fans), fans for garage stacks
may be located either near the garage slab, or in the garage
attic. The fan location will have some impact on stack design.
If the fan is located near the slab, the basic exhaust configura-
tion will be similar to that shown in Figure 21  for exterior
stacks. If the fan is in the garage attic, the configuration would
be similar to that described in Section 4.6.4 for fans in house
attics. Location of the fans in the attic could be particularly
convenient when the garage has an accessible floored attic.

Exhaust piping from garage fans—selection of stack
material. Whether the garage fan is near the slab or in the
attic, exhaust piping inside a garage will usually always be
PVC (or PE or ABS) of round cross section. Because the stack
will not be visible outdoors, there is no longer an incentive to
use downspouting as  the stack material, or to go  to the
expense of using PVC of rectangular cross section.

Exhaust piping from garage fans—offset against ga-
rage wall. If the fan is in the attic, the piping penetrating into
the garage can be routed upward immediately  beside the
garage wall, so that the stack riser can be almost flush against
the wall. Since the fan is in the attic, there is no need to
provide clearance for a fan below the stack riser. Hence, the
need to offset the piping back against the house (or,  in this
case, the garage wall), so important for exterior fans, is not an
issue.

If the fan is near the slab, so that clearance for the fan must be
provided at the base of the riser,  it may or  may not be
necessary to offset the stack against the garage wall. Such an
offset  may no longer be necessary  from the standpoint of
appearance, since the stack is inside the garage. Nor may it be
necessary from the standpoint of supporting the stack, if the
stack can be effectively supported from members in the attic
roof truss. But an offset against the garage wall may still be
desired in some cases, to get the stack out of the way of the
occupant, or to enable support of the pipe by attachment to the
wall studs behind the fire-rated sheetrock.

Exhaust piping from garage fans—connection of stack
segments. As with exterior stacks and the rest of the piping
network, the joints between the segments of stack and the
fittings must be cemented tightly. If the fan is near the slab,
leaks in the stack riser would result in radon-containing
exhaust being forced into the garage. If the fan is in the attic,
leaks would result in garage air leaking into stack, reducing
the suction being maintained at the slab.

Exhaust piping from garage fans—support of stack.
The support of exterior stacks against the side of the house
was discussed above. When the stack is in the garage — and in
cases where the garage stack is nearly flush against the garage
wall — the stack could be supported against the garage wall
(for example, using adaptations of the methods illustrated in
Figure 29). In this  case, the strapping, brackets, or clamps
around the PVC stack would need to be screwed into the wall
studs, or perhaps attached to the sheetrock using lag bolts.

However, the stack can also be supported overhead from attic
floor joists or from braces and rafters in the roof truss in  the
garage attic.  Depending upon how the stack and exhaust
piping are routed in the garage  attic, the support could be
accomplished using strapping, hangers, or clamps, using  ad-
aptations of the methods illustrated in Figures 16,17,20, and
23, and discussed in Section 4.6.4 (see Support for fans in
attics). When a garage stack is supported in this manner, and
especially when the fan is in the  attic, the configuration will
resemble an interior stack in most respects, and the guidance
in Section 4.6.4 would apply.

Exhaust piping from garage fans—piping runs in ga-
rage attic. Whether or not the fan is in the garage attic,  the
exhaust piping may have to make a horizontal run in the attic
so that the exhaust point through the garage roof will be at an
appropriate location. The criteria  for selecting the location of
the roof penetration have been discussed in Section 4.6.2 (see
Routing considerations with attic  piping runs) and in Section
4.6.4 (see Selecting the location of fans for interior stacks).

In this regard, one particular consideration for garage stacks is
that garages are  commonly one-story structures attached to
two-story living areas. Since the garage stack will be immedi-
ately beside the wall between the garage and the living area,
penetration of the stack straight  up through the garage roof
with no horizontal run would result in its discharging immedi-
ately beside the second story of the adjoining living wing. As
a result, the exhaust piping will require some horizontal run in
the garage attic in order to locate the discharge point at least
10 ft from any windows, attic gable vents, etc., in the upstairs
riving area overlooking the garage roof.

On the other hand, it is also common to encounter cases where
the garage roof is at the same level as the roof of the living
area In these cases, it will be more common to encounter
situations where  the stack can extend straight up through  the
roof directly above the  point where it enters the garage,
without any horizontal run in the  garage attic. This would be
possible when the penetration from the basement into  the
garage can be made at a point where a straight vertical stack
would result in an exhaust location that would meet the other
criteria discussed in Section 4.6.2 (e.g., the exhaust will be on
the rear slope of the roof).

Once the appropriate point has been identified for extending
the exhaust pipe through the garage roof, the procedure for
installing the attic piping, the fan (if it is in the garage attic),
and the exhaust piping through the roof would be the same as
that described in Section 4.6.4 for the interior stack/attic  fan
case.

Exhaust piping from  garage fans—finish around the
stack. Because the piping associated with garage fans will be
                                                        173

-------
inside the garage, the need to paint the garage stack, or to
frame it in for the sake of appearance, will usually be greatly
reduced compared to the case of exterior stacks.

As discussed toward die end of Section 4.6.3, it may some-
times be decided to frame in garage stacks as a means for
meeting code requirements for a fire break where the plastic
piping has penetrated the firewall between the basement and
the garage. Framing in the garage stack would be an alterna-
tive to use of a fusible linkage or intumescent wrap type of fire
break.

Caps and mufflers on the exhaust stack. The consider-
ations regarding the decision to install a cap or muffler on an
exterior or garage stack are essentially the same  as those
discussed in Section 4.6.4 for interior stacks.

Caps are not usually thought to be necessary. With exterior
stacks, where the stack is discharging near the eave, there may
be an increased risk that the cap, in deflecting the upward
exhaust jet near the eave, could cause the exhaust to be swept
back down under the overhang and re-entrained into the
house. The caps would also tend to defeat a key objective of
stacks fabricated from downspouting, namely, to make the
stacks inconspicuous.

Where a downspout is used as a stack, one should not attempt
to simulate a cap by bending the downspout into a "U" at the
eave, so the exhaust is directed at back downward beside the
eave. This configuration could be directing the exhaust down
into the recirculation zone beside the house, increasing the
risk of re-entrainment.

Mufflers will usually be needed for the high-suction/low-flow
fans, but are not often  needed with the in-line tubular fans.

Since most caps and all mufflers on the market are designed
for use with round PVC piping, some adaptations would be
necessary if they were to  be installed in an exterior stack
fabricated from downspouting.

Remote exhaust—a variation of the  exterior stack.
Some mitigators and homeowners are  concerned about the
aesthetic impact created by an exterior stack immediately
beside the house, as in Figure 21. Where there is no garage,
and where the exhaust system must be outdoors, one option
that has  been considered to avoid this impact has been to
extend the exhaust pipe horizontally below grade to  some
point remote from the house, perhaps 10 ft or more away
(E188, Wi91, Bro92). At  this remote location, the buried
horizontal piping is routed above grade and exhausted. When
using this approach, some mitigators mount the fan where the
piping comes above grade; others mount the fan in the buried
section of the piping, inside a buried enclosure.

Assuming the case of a basement with SSD suction pipes
installed through the slab vertically indoors, the exhaust pip-
ing from these suction pipes would penetrate the house shell
below grade. This will usually mean penetrating the basement
foundation wall below the band joist. A trench is dug from the
point of the wall penetration to the point in the yard where the
discharge point is to be installed; this location may be behind
shrubbery, or in some other inconspicuous location. By EPA's
mitigation standards, this remote location should be at least 10
ft from any public or private access routes, or from openings
in adjacent buildings.

The horizontal run of piping between the penetration and the
exhaust location is buried in this trench. If the fan is to be
mounted above grade, one logical configuration would be to
install a tee fitting on the remote end of this piping, with one
leg oriented  upwards and one downwards. A vertical riser
would be installed on the upper leg, with the fan mounted
vertically on this riser above grade. A stub of piping on the
lower leg of the tee would extend down into a bed of gravel, in
a pit excavated at that location. The purpose of this  dry well
below the fan is drain away any condensate or rainwater, and
is especially  important in cases where the contours of the lot
necessitate that the pipe be sloped downward away from the
house.  One possible uncertainty with this configuration is
whether the dry well may become flooded during wet periods.

If the contour of the lot is such that the buried horizontal pipe
is sloped downward toward  the house,  the condensate and
rainwater would drain back into the SSD  riser in the base-
ment In that case, the dry well below the fan would be less
necessary.

Another option that may sometimes exist for handling drain-
age when the pipe is sloped away from the house, is to extend
the buried piping run to the point  where it reaches grade level.
In this case, the system would become identical to the remote
portion  of a DTD/remote discharge system as  depicted in
Figure  4:  a  trap is placed between the fan  riser  and the
discharge line to prevent outdoor air from being drawn through
the discharge line by the fan, and any condensate or rainwater
drains away  through the above-grade discharge. Li this case,
the dry well would be unnecessary.

The fan is secured vertically on the riser using flexible PVC
couplings, perhaps a couple of feet above grade. By EPA's
current standards, it would not be acceptable to simply allow
the fan to discharge vertically at that height. Rather, in order
to reduce exposure by persons in  the  yard nearby to the
potentially high-radon exhaust,  a stack would  have to be
appended to the fan discharge, so that the exhaust point would
be at least 10 ft above grade. For structural stability, such a
stack would have to be supported, e.g., with posts installed for
this purpose, or perhaps with guy wires. Such a stack would
often eliminate the aesthetic benefits of moving the exhaust
away from the house.

If the stack were not installed, contrary to  standards, the fan
exhaust would have to be protected in order to prevent debris
and animals from dropping into the blades, or to keep children
from getting their fingers caught in the blades. Some fan
manufacturers  sell grilles that could be installed on the ex-
haust side of the fan for this purpose. Alternatively, a short
length of PVC piping might be installed on the exhaust side of
the fan, perhaps with a hardware cloth screen on top.

While such remote location of the fan  and exhaust system
would improve aesthetics if a 10-ft stack were not required,
there are some potential disadvantages to this design. Even if
                                                       174

-------
no 10-ft stack is needed on the remote fan, this configuration
will be more expensive than the case where the exterior fan
rises immediately beside the house, due to  the trenching
required to bury the horizontal run away from the house, and
due to the added effort involved in coring through the founda-
tion wall below grade  rather than sawing through the band
joist (He91b, He91c). The electrical wiring will be compli-
cated slightly, since a conduit will now have to be laid out to
the remote fan.  There will be some suction loss in the length
of buried piping, although this will generally not be a serious
problem  when  90-watt fans are being  used, and it will be
comparable to the back-pressure loss that would be created by
a stack above the eave.

The preceding discussion assumes the remote fan is installed
above  grade. Some mitigators  install  the fan  in a buried
enclosure, for aesthetic reasons. The enclosure is accessible
via an access door in the top of the enclosure at grade. To
reduce the size  of the excavation for the fan enclosure, some
firms mount the fan horizontally; the buried horizontal piping
run penetrates the enclosure on each side, connecting to the
fan. A hole is drilled on the underside of the fan housing, and
gravel is installed below the open-bottomed enclosure, to
allow condensate entering  the  fan housing to  drain away.
Others make the enclosure larger and deeper, so that the fan
can be mounted vertically. With  proper excavation, any of the
approaches just discussed with  a remote fan mounted  verti-
cally above grade could be installed with the fan vertically
below grade.

With any system having the fan  below grade, great care must
be taken to ensure that the piping and the dry well beneath the
fan drain properly. Cases have been reported where careless
installation resulted in the ultimate flooding of the entire fan
enclosure and underground piping run.  Also, the lifetime of
fans mounted in potentially humid underground enclosures
has not been demonstrated.

Because of these concerns, it is anticipated that EPA's final
mitigation  standards will prohibit  mounting of fans below
ground.

Requirement  that fans be outside livable envelope of
house. As emphasized at  the  outset of  Section 4.6.4 (see
Overall design  considerations for  all stack configurations)
and elsewhere  in  the preceding discussion, EPA's interim
mitigation  standards require that SSD fans be outside the
livable envelope of the house. Although extensive data are not
available  documenting  the  frequency and effects of
pressure-side exhaust leaks when fans are indoors, the poten-
tial risk  of such leaks is considered great enough that it
warrants the aesthetic  impacts  and other complications in-
volved in placing the fan outdoors  when an exterior stack is
used.

Mounting a fan indoors is contrary to EPA's standards. If an
installer were to consider mounting a fan indoors because it
was truly impractical to mount the fan outdoors, that installer
should consider special precautions in an effort to reduce the
risk of exhaust leakage into the  livable area. It must be
recognized that there is no evidence that any such precautions
would compensate for the  risks incurred if the fan were
mounted indoors. Such precautions thus could never be con-
sidered as a substitute for abiding by the standards.

Among the precautions that might be considered could be the
following: a) using a fan having an integral housing, to avoid
leakage through seams in the housing; b) using Schedule 40
(heavy gauge) PVC, PE, or ABS for all system piping; c)
installing only  a  minimal length of exhaust piping on the
pressure side of the fan inside the house; d) firmly supporting
the interior piping against both vertical and lateral movement,
especially on the pressure side of the fan, so that any physical
stresses that might be imposed on  the fan or piping by the
occupants would be unable to create any flexing of this piping
network  that would loosen or break piping seals; and e)
carefully caulking all seams in the piping on the pressure side
of the fan would, even though they have  been carefully
cemented and  may appear  to be gas-tight at the time of
installation. These seams include those in the flexible PVC
couplings connecting the fan to the piping (even though the
hose clamps appear to have produced a gas-tight seal), and
seams between PVC piping  and fittings (even though these
joints have been cemented).

Requirement that exhaust be discharged at least 10ft
above grade. As emphasized at the outset of Section 4.6.4
(see Overall design considerations for all stack configura-
tions)  and elsewhere,  EPA's  interim mitigation  standards
require that SSD exhausts be  released at  least  10 ft above
grade, among other requirements.  For exterior stacks,  this
requirement generally  means  that  a stack must extend up
beside the house, usually to a point above  the eave. It is
anticipated that EPA's final  standards will explicitly require
that the exhaust be discharged above the eave, preferably the
highest eave. The objective  of this requirement is  to reduce
the risk of exhaust re-entraining back into the house or expos-
ing persons outdoors. While there is not an extensive data
base defining the extent of re-entrainment or exposure with
different exhaust configurations, the health risks that could
result are considered sufficiently great that they warrant the
aesthetic impact and cost of installing an exterior stack to
avoid the increased risks that might be anticipated if the
exhaust discharged at grade level.

Exhausting at grade level immediately outside the house, i.e.,
installing a system similar to that in Figure 21, but eliminating
the exterior stack up to the eave, is contrary to EPA's stan-
dards. If an installer were to consider exhausting at grade level
because it was truly impractical to install an interior or exte-
rior stack, that installer should consider special precautions in
an effort to reduce the risk of exhaust re-entrainment and
exposure to persons outdoors. It must be recognized that there
is no evidence  that any such precautions would compensate
for the risks incurred if the exterior stack were eliminated.
Such precautions thus could never be considered as a substi-
tute for abiding by the standards.

Results to date suggest that if the  exhaust were released at
grade, least re-entrainment is likely  to occur if the exhaust
were directed horizontal to grade, 90° away from the house.
Limited  results suggest that exhaust horizontally at grade,
directed  90° away from the house, can sometimes result in
re-entrainment  no more severe than  that resulting when the
                                                        175

-------
exhaust is vertical above the eave, at least under the condi-
tions tested (K91).

Intuitively, to reduce re-entrainment, it would seem desirable
for the  grade-level  exhaust to be at least some minimum
distance (e.g., at least 10 ft) from openings in the house shell
(primarily windows and doors), if at all possible. To reduce
exposure of persons outside the house, the other EPA exhaust
standards should be met: the discharge location should be at
least 10 ft from public or private access, and from openings in
adjacent buildings. In particular, the exhaust should not be
near patios or under decks.

One approach for directing the exhaust away  from the house
at grade would be  to replace the exterior stack shown in
Figure 21 with the following. A short length of vertical pipe
would be attached to the coupling above the fan, to provide a
connection for a 90° elbow. The elbow would be cemented to
the top of this short vertical pipe, with the outlet of the elbow
being horizontal to grade, directed  away from the wall of the
house. A short  length of horizontal  pipe cemented into the
outlet of this elbow could be important, so that the exhaust
gases develop a clear momentum away from the house before
being discharged. If the gases were discharged directly from
the horizontal end of the elbow, without the length of piping,
there would be a turbulence resulting from their deflection in
the elbow that could cause them to  mix more rapidly with the
outdoor air, rather than being expelled away from the house in
a clearly defined jet.

If there is not going to be a stack above the fan, another option
would be to install the fan vertically downwards, rather than
upwards as in all of the preceding figures. In this case, the 90°
elbow that is attached to the end of the horizontal pipe through
the wall would be directed down  instead of  up. Downward
mounting might reduce the visibility of the fan in cases where
the exhaust piping penetrates the foundation wall at a suffi-
cient height. Again, it is recommended that the exhaust from
the fan be re-directed horizontally away from the house by an
elbow on the outlet  of the fan, as discussed in the preceding
paragraph.

A grade-level exhaust must not be directed parallel to the side
of the house, immediately beside  the house.  Specifically, it
must  not be directed downward  toward grade beside  the
foundation, or upward beneath the eave along the face of the
house, or horizontal to grade parallel to  the house. In many
installations where re-entrainment has been confirmed to have
been a problem in the past, the exhaust has been parallel to the
side of the house. In one commonly encountered configura-
tion which should always be avoided, the grade-level exhaust
is directed down toward grade by a dryer vent mounted on the
side of the house, connected to the end of the piping penetrat-
ing the foundation from inside the basement

4.7    Slab Sealing  in Conjunction
with  SSD  Systems
House air will leak into the SSD system through any unclosed
openings in the slab. Such leakage can interfere with  the
extension of the suction field  beneath the slab (potentially
reducing radon reduction performance). It will also increase
the house heating and cooling penalty; heated or air-conditioned
air will be drawn out of the house by the system through the
unclosed cracks, and will be replaced by infiltration of out-
door air. If sufficient house air is drawn into the system via
unclosed slab openings, the SSD system could also contribute
to depressurization of the basement or house, thus potentially
contributing to back-drafting of combustion appliances.

For the purposes of this discussion, slab openings are subdi-
vided into three categories: "major" openings,  which should
always be sealed as part of SSD installation unless sealing is
truly impractical; "intermediate" openings, which should gen-
erally be sealed, if this is reasonably possible; and hairline
cracks, for which  sealing can be helpful but is often not
required.

Major openings. The sealing of major slab openings can be
very important in achieving good performance and reducing
operating costs. "Major" entry routes are defined here as those
which  are relatively large and distinct, such as perimeter
channel drains, large openings through the slab around utility
penetrations, floor drains connecting to the soil or the sub-slab
region, and sump pits. Such major  routes should always be
sealed. In  cases where such major routes are inaccessible —
such as a perimeter channel drain concealed behind floor and
wall finish ~ a conscious decision may sometimes be made to
accept the penalties of leaving the opening unsealed, rather
than incurring a cost that the homeowner would find unac-
ceptable in  trying  to seal this opening. However, in  such
cases, the  homeowner should be consulted on  this decision,
and advised of the reductions in performance and the increase
in operating costs that may result.

Intermediate openings. In addition to the  "major" slab
openings discussed above, there will often be smaller, but still
potentially important, openings of intermediate size. Examples
of such "intermediate" openings include, e.g., the wall/floor
joint when there is a modest but distinctly visible gap (perhaps
1/32 to 1/16 in. wide or  greater), partially open expansion
joints,  and small openings through the slab  (e.g., around
individual utility pipes penetrating the slab). Such intermedi-
ate openings should routinely be sealed closed. However,
where such intermediate openings are inaccessible, e.g., where
a  1/32- to  1/16-in.  wide  wall/floor joint disappears behind
wall finish in a finished part of the basement, the inability to
seal this opening will be less of a concern, compared to major
openings.

Hairline cracks. Hairline cracks include, for example, set-
tling cracks or wall/floor joints which are tight, with no (or
very little) separation. Given their length, the total leakage
area represented by hairline cracks could be meaningful de-
spite their narrow width. However, such  cracks can be diffi-
cult to seal properly and effectively. Simply applying a bead
of caulk on top of this crack in a basement may be helpful, but
it  is unclear how effective this  sealing would be, or how
durable. To seal such cracks effectively, a channel might have
to be routed in the concrete  along the crack, to  provide
adequate surface area for adhesion of the sealant (see Section
4  in Reference EPA88a);  such  a procedure would be
time-consuming, and could significantly  increase the cost of
the installation.
                                                       176

-------
Experience has indicated that a properly designed S SD system
can usually handle the leakage from hairline cracks without a
significant reduction in system performance. Therefore, since
sealing the cracks is difficult and is often not necessary for
good system performance, special efforts to seal  hairline
cracks will often not be warranted. Attempts to  seal hairline
cracks will likely be most warranted in cases where sub-slab
communication is very poor and the cracks are near the SSD
suction  pipes.  Any  leakage through the  cracks  in
poor-communication cases might reduce the performance of a
marginal SSD system, which would be having a hard time
extending a suction field even in the absence of such leakage.

Because caulking of hairline cracks may reduce the heating/
cooling penalty and possibly improve system performance,
mitigators who routinely apply  a bead of caulk to accessible
wall/floor joints, even when the crack is only a hairline and
there is no gap, should continue to do so. Homeowners who
are installing their own systems (and who are thus not con-
cerned about the labor costs of applying the caulk) may wish
to do this. However, it must be recognized that the seal may
not be durable or fully effective, and that it may or may not be
possible to detect the benefits of such caulking.

General comments on seating. In the discussion here, the
word "seal"  is not used in the sense of a  true air-tight seal.
While such a true seal would be desirable,  it is often difficult
to achieve such a true seal, and almost impossible to maintain
it permanently as the foundation shifts over the years. In the
context here, where the objective is to significantly reduce
house air flow into the SSD system, a seal is considered
adequate if it closes the opening (and reduces  the air flow
through it) to a large degree, even if not 100%. Thus, for
example, if an opening through the slab is  sealed with caulk,
and if the bond between the caulk and the surrounding con-
crete breaks  over time (due, e.g., to foundation shifting), the
caulk may still be blocking a large amount of the air flow
through the opening.

Detailed procedures for sealing slab openings are discussed in
Section 4 of Reference EPASSa. The discussion below re-
views key features for sealing some of the openings of par-
ticular concern from the standpoint of SSD performance.

4.7.1  Perimeter Channel Drains

Perimeter channel drains, also known as canal drains  or
"French" drains,  are 1- to 2-in.  wide gaps  around the perim-
eter of basements (i.e., 1- to 2-in. wide wall/floor joints). A
perimeter channel drain is often intended  primarily  to drain
water that enters the basement through the porous face of
block foundation walls during  wet weather; in some cases,
holes are drilled through the face of the lower course of block,
at a level below the top of the slab, to facilitate this water flow.
The drains will also handle water entering  from on top  of the
slab, e.g., due to a clothes washer overflow. Perimeter channel
drains can also  be specifically designed to handle  water
entering from beneath the slab,  although many are not.

The water entering the perimeter channel drain is commonly
routed to the region beneath the slab. In some cases, there are
drain tiles under the slab,  draining to an existing sump in the
basement or to a remote discharge; much of the water from the
perimeter channel drain would be expected to enter these tiles
and drain to the sump or discharge. In other cases, there will
be no drain tiles, in which case the sub-slab aggregate will be
serving as a dry well.  In  still other cases, the perimeter
channel drain may drain, via an underground  channel of
aggregate, to a dry well installed remote from the house.
Where the basement is finished, the perimeter channel drain
will be concealed behind the wall finish.

Perimeter channel drains can be expensive to seal properly,
even when they are accessible. Accordingly, some mitigators
have  reported installing SSD systems without  sealing  the
perimeter channel drain, and obtaining reasonably good radon
reductions, at least when sub-slab communication is reason-
ably good. To avoid undue leakage of  house air into  the
system, the SSD suction pipes  were installed  toward  the
center of the slab, away from the perimeter. No studies have
been conducted to evaluate the effect of leaving the perimeter
channel drain unsealed. But since the opening created by the
perimeter channel drain is so large, so widely distributed, and
so strategically located, some penalties might be expected to
result from this approach. Indoor radon concentrations might
be expected to rise during cold weather; the ability of the SSD
treatment to extend to block foundation walls would seem to
be greatly reduced; and the heating/cooling penalty would
intuitively be greater than average.

Thus, serious consideration  should always be given to prop-
erly sealing perimeter channel drains, except in  those cases
where they are concealed behind wall finish (so that the cost
of sealing could be prohibitive) or where  water  drainage
problems or code violations might be created by sealing. A
decision not to seal part or all of the perimeter channel drain
should be made with consultation from the homeowner. The
mitigator should be prepared to seal the channel drain, or to
otherwise modify the system to compensate for air leakage
through the drain, if an initial  mitigation system does  not
achieve adequate radon reductions due to  the open drain.

In some locales, perimeter channel  drains are required by
code. In some cases, the drains have an important water
drainage function, as discussed above. Thus, any steps to seal
the drains must be taken with care to properly maintain the
drainage function.

Where the perimeter channel drain is known to have no water
drainage function either from above or below the slab and
where codes do not require that the drain be left open, the 1- to
2-in. wide cap can simply be mortared closed. While mortar-
ing the drain closed may be the simplest approach, it can be a
risk for mitigators. The mitigator may be considered liable for
any subsequent drainage problems that might be attributed to
the closure of this supposedly non-functioning drain.

When the perimeter channel drain has a water drainage func-
tion,  it can be sealed in a manner which maintains that
function. See Figure 32A, and Section 4 of Reference EPA88a.
The objective of this sealing approach is to install a caulk seal
in the channel at a level beneath the top  of the slab, so that
there is still: a channel beneath the seal (to  handle water
entering through  weep holes at the base of the block wall, or
                                                       177

-------
                               Backer Rod
                               Drainage Channel Above Seal (for Water
                               Entering Drain from Above the Slab)

                               Layer of Flowable Urethane Caulk (>^1/2" Deep)
                                         Slab  .   B

                                     *   •   * A'   '   A
                   Note: Sealing a perimeter channel
                   drain in this manner may necessitate
                   the installation of a sump, if the
                   drain is used to collect water
                   entering from above. If water flow
                   from above is heavy, sealing the
                   channel drain may not be advisable
                                        Drainage Channel Below Seal (for Water
                                        Entering Drain from Below the Slab)

                                       Weep Hole (if Present): Must Be Below Seal
Figure 32A.  Cross-section of a sealed perimeter channel drain, illustrating water drainage channels both above and below the seal.
otherwise entering from  beneath the slab); and a channel
above the seal (to handle water entering through the block
face above the slab, or otherwise entering from on top of the
slab).

By this approach, backer rod of a diameter slightly greater
than the width of the perimeter channel drain (or some other
appropriate material) must be stuffed  down  into  the gap
between the slab and the foundation wall, to support a layer of
sealant. Backer rod is a compressible, closed-cell polyethyl-
ene foam material which  is formed into cords  of alternative
diameters. The top of the backer rod must be at least 1/2 to  1
in. below the top of the slab; the bottom of the backer rod must
be at least 1 to 2 in. above the bottom of the slab (or the top of
the footing). If weep holes have been drilled through the face
of the block wall at a level below the top of the slab, to allow
water inside  the block cavities to  flow into the perimeter
channel drain, the bottom of the backer rod must be above
those holes, so that they remain open.

The top of the perimeter channel drain is then closed with an
appropriate sealant, supported by the backer rod. The sides of
the perimeter channel drain must be clean (and preferably
should be dry) when the sealant is applied, to ensure a good
adhesive bond  between the sealant and the concrete. The
preferred sealant is flowable urethane caulk, since this caulk
will effectively fill in the irregular space between the top of
the backer rod and the slab, and will adhere well to  the
concrete. The urethane caulk layer should not be more than
about 1/2 in. deep, since it may not cure properly if it is much
deeper than this; see the instructions from the manufacturer.
The top of the urethane caulk sealing the drain must be below
the top of the slab, by perhaps 1/2 in. or more, creating a
channel on top of the seal.

This sealing approach prevents house air from flowing down
through the perimeter channel drain into the sub-slab region.
But at the same time, it permits water flowing through  the
weep holes beneath the seal to flow into the gap under  the
backer rod, and from there into the sub-slab aggregate (or to
wherever the perimeter drain was originally designed to direct
the water). Also, any water entering from above the slab has a
channel in which to collect, on top of the caulk seal.

If a significant amount of water is expected to enter the drain
from above the slab,  the channel above the caulk seal will
have to be designed in a manner which will allow this water to
drain away. One approach could be to leave gaps in the seal at
intervals around the perimeter, so that the water in the upper
channel can drain down through those gaps into the sub-slab
                                                       178

-------
region as originally intended. These gaps should be located in
particular along those walls where water flow from above the
channel is expected to be heaviest. Such gaps in the  seal
would, of course, result in house air leakage into the SSD
system through those gaps (and, potentially, radon entry into
the house through the gaps). However, since most of the
perimeter channel drain would be sealed, the seal would be
significantly reducing the amount of air flow. Where a gap
was left in the seal, the channel under the seal should be sealed
on the sides of that gap, so that the sub-seal channel remains
sealed off from the basement.

Another option if  there is significant  water entering from
above the slab would be to utilize an existing sump, if present,
or to retrofit a  new sump into one corner of the basement to
handle this water. See Figure 32B. This figure assumes that
drain tiles had  been installed under the  slab during construc-
tion to collect water from the perimeter drain, and that the tiles
drain to an existing sump. Use of a sump may not always be
satisfactory in cases where water enters the top  channel at
points remote from the sump, since the water may not reliably
flow through the shallow channel over extended distances to
the point at which the sump is located, without overflowing
onto the slab at some other location.
    Whether an existing sump is available, or whether a new sump
    has to be installed, a channel would have to be routed in the
    top of the concrete slab, directing the water in the top channel
    of the sealed perimeter drain into the sump, as shown in the
    figure. Especially in cases such as that in Figure 32B, where
    the existing or new sump connects to the sub-slab region, the
    sump must be capped so that it is not a major unsealed slab
    opening (see Section 4.7.4). Because the top channel in the
    sealed perimeter channel  drain will be routinely directing
    water into the sump from above, the sump cover will have to
    be fitted with a trap so that the water drainage function of the
    sump  can continue  while the cover remains air-tight. Sump
    covers are  discussed in detail in Section 5.5.

    The preceding discussion concerning the capping of the sump
    assumes the case where it  has been decided hot  to draw
    suction directly on the sump, and where a SSD suction pipe
    through the slab remote from the sump is the preferred ap-
    proach. If the sump is an  existing  sump with drain tiles
    emptying into it, as in Figure 32B, one might commonly elect
    to install sump/DTD instead of SSD, as discussed in Section
    5. The steps for sealing the sump and the perimeter channel
    drain would be essentially the same if sump/DTD were the
    intended mitigation approach, as discussed in Section 5.7.
   Waterless Trap
   in Sump Cover
  Sump Cover, Resting
  on Top Lip of Sump
  Liner
Water from Channel
Above Perimeter Drain
Seal (Via Channel
on Top of Slab
Channel Routed
   in Top of Slab
Channel
Above
Drain Seal

 Perimeter
 Drain
 Sealant
                                                                                                        Backer
                                                                                                        Rod
     Sump (Existing
     or Retrofit)
                                                                          Existing Drain
                                                                          Tile, If Present
Figure 32B.  Method for sealing the sump and the slab channel leading to the sump in cases where perimeter channel drains empty into a
            sump in the basement.
                                                       179

-------
If a sump were not already present, and if a sump had to be
retrofit in order to handle water from on top of the sealed
perimeter channel drain, this would add to the SSD installa-
tion cost In some cases, the perimeter channel drain will have
initially been designed to route the water into the sub-slab
aggregate with the aggregate serving as a dry well (i.e., with
no drain tiles to handle the water); in other cases, the channel
drain will route the water into a gravel channel leading to a
remote dry well. In such cases, the sump installed as part of
the channel drain sealing process could do the same, and it
would not  be necessary to install a sump pump and water
discharge line in the sump.

In many cases, a portion of the perimeter channel drain may
be inaccessible, for example, concealed behind wall finish. In
such cases, it may be decided to seal the accessible portion of
the perimeter channel drain (in  the unfinished part of the
basement),  but to leave the inaccessible portion (in the fin-
ished part)  unsealed, at least initially, due to the  costs in-
volved in removing and  replacing the finish. When this is
done, the exposed open ends of the sub-seal channel must be
caulked shut where the seal ends, so that house air cannot leak
into the sub-seal channel. When a portion of a channel drain is
scaled, the  mitigator should understand how that particular
channel drain functions (e.g., using aggregate as dry well, or
utilizing sub-slab  drain tiles), to ensure that drainage prob-
lems will not be created.

4.7.2  Other Major Holes through Slab

In addition to perimeter channel drains, there will be other
cases where major holes exist through the slab. One common
example is  the opening, commonly about 2 ft square, often
left in the slab around the water and sewer lines that come up
through the slab under bathtubs and shower stalls that are
installed  on slabs. This opening is intended to provide the
plumber some flexibility when installing the bathtub. (There
may also be openings around the water and sewer penetrations
under sinks and commodes in the bathroom, but these will
generally be smaller than that around the bathtub plumbing.)
Some other examples include unused sump holes, or a hole at
the site of some former structure in the basement. An extreme
example of a large hole through the slab would be the case
where some portion of the slab is missing, leaving bare earth
exposed. This latter situation could occur, e.g., when small
wings to the basement (such as root cellars or greenhouses)
are left with bare-earth floors.

Slab holes around bathtub plumbing. Where a bathroom
has been installed on a basement slab or a slab on grade, the
plumbing for the bathtub (and the slab opening around this
plumbing)  is commonly  accessible through an access door
through the frame wall at the head of the tub. This  slab hole
could be mortared shut, with the water and sewer lines pen-
etrating the new mortar. However, this could potentially inter-
fere with any subsequent plumbing work that may have to be
done on the system. A simpler approach might be to close the
slab  opening using an expandable closed-cell polyurethane
foam. The foam  expands as it  hardens and cures, tightly
closing the opening. If access to the plumbing below the slab
is subsequently needed, portions of the foam can be cut away
(e.g., with a utility knife) as necessary to expose the plumb-
ing. After the plumbing repairs have been made, any holes cut
in the hardened foam can be filled with new foam, or with
caulk or other suitable sealant.

In some cases, no access door will be provided for the bathtub
plumbing, and the major slab opening under the bathtub will
not be as conveniently accessible. In such cases, a decision
must be  made regarding whether to leave the bathtub slab
opening unclosed, or to incur the expense of creating and then
restoring an opening in the wall to allow the sealing to be
accomplished. Some mitigators routinely make the hole through
the wall and seal the slab. One mitigator reports installing an
air vent  grille (such as those used with forced-air heating
systems)  in the wall hole after the slab is sealed, as a conve-
nient and neat approach for restoring the wall hole (Ba92).

Other mitigators recommend postponing efforts to seal such
inaccessible bathtub slab openings (K192). Where sub-slab
communication is reasonably good and radon levels are not
real high, it may sometimes be advisable to initially install the
SSD system without closing the inaccessible bathtub opening,
with the intent that that sealing step could be taken later if the
initial SSD installation did not achieve adequate performance.

Miscellaneous major slab holes. Where holes of any
significant size exist through the slab, and where these holes
serve no  function, the holes should be closed.

If the soil is not level beneath the slab at the location of the
slab hole and if the  hole is  sufficiently large, it may be
necessary to fill in the sub-slab region up to the underside of
the slab,  using some suitable fill material. To reduce subse-
quent settling of this fill, and cracking of the patch that will be
applied on top, any such fill should be compacted.

Especially where the  hole is in an exposed location where
there will be foot traffic, the hole will best be closed using
concrete  or non-shrink mortar. Use of concrete or mortar may
also provide the best appearance, blending in with the remain-
der of the slab. The sides of the slab around the perimeter of
the opening should be cleaned, in an effort to  improve the
bond between the  new and the old slab. New concrete or
mortar is poured to fill the hole, and the perimeter of the patch
(between the new and old concrete)  is  tooled to provide  a
channel about 1/2-in. deep. When the new concrete or mortar
has hardened, this channel should be flooded with flowable
urethane  caulk, in and effort to plug any  crack that may form
as the new mortar shrinks and settles.

In some cases, when the hole is in a remote or less accessible
location,  one might elect to close it using expanding foam, as
discussed previously  for the  case of slab openings  under
bathtubs.

Sections of missing slab. From time to time, entire sections
of basement slab have been found to be missing in  certain
houses.  For example, the basement  may  have a  small
earthen-floored wing which is used as a root cellar. Or, some
significant portion of the basement floor may have been left
unpaved, for one reason or another.
                                                       180

-------
In such cases, the best course of action will be determined by
the specific circumstances. Depending upon the size and use
of the unpaved area, it may be recommended to the home-
owner that a slab be poured (with aggregate beneath) for that
section.

In other cases, perhaps the best approach would be to treat it
as an unpaved crawl space wing adjoining the slab.  In this
case, plastic sheeting would be sealed over the exposed soil. If
SSD on the paved portion of the basement proved to be
insufficient to adequately reduce radon levels, the SSD sys-
tem might be supplemented by a sub-membrane depressuriza-
tion under this plastic sheeting (see Section 8).

4.7.3  Untrapped Floor Drains Connected
to the Soil
Floor drains  in a basement  slab  can present a significant
problem, if they: a) drain to soil (e.g., via perforated drain tiles
beneath the slab, or via pipes to  a  septic system); and b) are
not trapped.

When the drain empties through perforated piping that passes
through the sub-slab region and when the drain is not trapped,
house air can be  drawn  into the  SSD  system through the
untrapped drain and through the perforations in the drain  tiles.
The length of tiles leading from the drain could be "intercept-
ing" the sub-slab suction field, providing a source of leakage
air that could prevent the suction field from extending  further
beneath the slab.

Due to the amount of air that can move through the floor drain
and drain tile, the SSD system may not be able to adequately
depressurize the soil surrounding the drain  tile. Thus, the
untrapped floor drain may continue to serve as a significant
radon entry route into the house.
drain grille. These units seal the existing drain, and direct all
water through the throat of the Dranjer unit.

Although designs of the different models vary slightly, the
principle of each is that, at the bottom of the throat,
downward-flowing water encounters an impoundment which
it must overflow before entering the sewer line. This im-
poundment may be a small bowl in which the base of the
throat is submerged, as in Figure 32B or in the DR-2; or, in the
"J-series" unit, the base of the throat may curve upward in the
form of a simple P trap (see Figure 5 in Reference EPA88a).
The opening at the top of the impoundment is covered by a
weighted ring  (as in Figure 32B) or a ball, which floats to
permit the water to overflow.  However, when no water is
flowing, the ring or ball seats into the opening. The seating of
the ring or ball prevents gas movement upward or downward
through the drain; i.e., soil gas cannot move upward into the
house,  and house air cannot be drawn down through the drain
by the  SSD system. The floating ring or ball provides a seal
even in cases where there is no water in the trap, hence the
term "waterless."

If the floor drain is never used, an alternative approach would
be to simply mortar the drain opening shut. This approach is
not recommended unless it is certain that the drain will indeed
never be needed, and is not required by code. The drain could
also be plugged with a rubber stopper, which could be re-
moved on any occasion when the drain were needed.

Sometimes when a floor drain drains to a septic system, there
will be a clean-out plug in the drain line downstream of the
trap. If this plug were missing, a situation which has some-
times been observed, soil gas could still flow up into the
house,  bypassing the trap, even when the floor drain trap is
full of water. If this clean-out drain plug is missing, it must be
replaced (e.g., with a rubber stopper).
One technique to determine whether a particular floor drain is   4.7.4  Sumps
untrapped would be to remove the top grille and inspect using
a flashlight during the visual inspection of the house (Section
3.2). Other techniques include chemical smoke testing at the
drain during the visual inspection, to  assess whether air is
flowing up through the drain; and radon grab sampling in the
drain (Section 3.4). It can be difficult to independently deter-
mine whether the drain connects to drain tiles.
To address both the potential disruption of the suction field
and the continued radon entry through the floor drain, the
floor drain must be trapped. Even if the floor drain connects to
a septic system by a solid pipe, rather than the perforated
piping connection to the soil considered previously, trapping
the drain will have the benefit of preventing radon entry into
the house from the septic system.

If the floor drain already contains a trap, it must be ensured
that this trap remains full of water. If the floor drain does not
contain a trap, one can often be retrofit. Dranjer Corp. sells
several models of plastic "waterless" traps for various appli-
cations, such as the "F-series" trap shown in the sump cover in
Figure 32B.  A slightly modified version  of that trap, the
DR-2, has been designed to fit into many standard floor drains
ranging from 2 to 8 in. in diameter, beneath the existing floor
Sumps not only provide a major opening through the slab, but
also, if drain tiles empty into the sump, as is often the case,
they can connect widely to the aggregate and soil beneath the
slab and around the foundation. As such, they can be major
soil gas entry routes, and they can provide a major leakage
path for house air or outdoor air to leak down into a SSD
system.
In many cases where a sump is present and where drain tiles
empty into the sump, a sump/DTD system will be installed
rather than a SSD system.  In this case, the sump would
necessarily be sealed with an air-tight cover, in the manner
described later in Section 5.5, as an inherent part of the sump/
DTD system. Where there are no drain tiles entering the sump
and no sump pump, the sump pit may be used as a ready-made
hole through the slab for a SSD pipe, in which case the sump
pit would also be fitted with an air-tight cover as part of the
installation of the suction pipe.

In those cases where a sump exists and where SSD suction
pipes will be installed independent of the sump, the sump
opening must still be fitted with an air-tight cover as shown in
Figure 3 and described in Section 5.5.1. Otherwise, a large
                                                       181

-------
amount of basement air can flow through the sump and drain
tiles into the SSD system. If water will enter the sump from on
top of the slab, a waterless trap must be installed through the
sump cover.  If a sump pump is present and if it is not a
submersible pump, it must be replaced with a submersible
pump.

4.7.5 Intermediate Openings Through
the Slab

Slab openings defined here as being "intermediate" should be
sealed whenever they are accessible. They are distinguished
from the "major" openings discussed previously in that if the
intermediate openings are inaccessible and inconvenient to
seal, it will be less warranted to incur increased efforts and
costs attempting to seal these openings.

Wall/floor joints wider than a hairline  crack (but not
perimeter cftannel drains). If the wall/floor joint is wider
than a hairline crack (e.g., wider than roughly 1/32 to 1/16
in.), it should be caulked wherever accessible. Because of the
nature of this crack, gun-grade (non-flowable) urethane caulk
will generally be the best selection. Flowable caulk would
tend to puddle on top of the slab, if the crack is narrow; or, it
may disappear down under the slab, if the crack is wider but is
still too narrow to allow some backer-rod type of support to be
stuffed down before caulking.

The wall/floor joint should be wire-brushed to remove depos-
its and loose concrete, and should be vacuumed before caulk-
ing, to try to make the surface as clean as reasonably possible.
The bead of gun-grade  caulk should then be worked down
into the crack to the extent possible.

Even  where the entire wall/floor joint is  readily accessible,
caulking the joint in this manner can add more than $100 to a
mitigator's installation cost, depending upon the basement
size, the  amount of preparation, and  labor rates (He91b,
Hc91c).

Gaps around utility pipes penetrating the slab.  When
utility pipes (e.g., water, sewer, and fuel lines) penetrate a
slab, a crack  or gap  will usually exist between the concrete
and the pipe perimeter. Sometimes, the concrete  will have
been poured flush against the pipe,  and only a hairline crack
will exist around the perimeter. Other times, a gap will be left
around the utility line, often created by a sleeve of metal or
packing around the pipe. This gap may be intended to provide
subsequent flexibility in mounting a fixture (such as a sink or
commode) on top of the pipe, or to protect the pipe when the
concrete is poured.

If the gap around the perimeter is more than a hairline (e.g.,
wider than about 1/32 to 1/16 in.) but not wide enough to stuff
down  some  supporting material,  it should caulked with
gun-grade urethane caulk, as discussed previously for wall/
floor joints. If it is wide enough to accommodate some sup-
porting material, this support should be provided,  and flow-
able urethane caulk should be used. If there is still some
packing material around the perimeter of the pipe, between
the pipe and the concrete, the top layer of this packing should
be scraped away to  provide a channel for the caulk. In all
cases, the surfaces should be vacuumed and cleaned to pro-
vide a reasonably good surface for adhesion of the caulk.

The utility lines usually having the biggest gaps (especially
under commodes) will be inaccessible without removing the
fixture.  Unless communication  is poor, SSD  systems are
usually  able to handle air leakage through such gaps suffi-
ciently well. It will probably not be cost-effective in most
cases to remove the fixture and seal the gap, in an effort to
improve SSD performance. (As discussed previously, these
gaps around individual pipes are distinguished from the large
holes commonly left around the plumbing for bathtubs, which
are considered "major" openings.)

Expansion joints. Sometimes an expansion joint will be
installed in a residential slab while the concrete  slab is being
poured.  Expansion joints are strips of asphalt-impregnated
compressible  fibrous material about  1/2 in. wide. In some
regions, they are installed around the perimeter of the slab, at
the wall/floor joint, serving  as a buffer between the slab and
the foundation wall. In other cases, they are installed in the
middle of the slab, across the width of the slab (perpendicular
to the front and rear walls), often at points where there is a
discontinuity  in one of the walls. They are referred to as
expansion joints because they will compress if the slab ever
expanded, thus avoiding cracking from  that cause. (In fact,
wet concrete will shrink after being poured, and will usually
never again reach its "wet" dimensions, except under unusual
circumstances.) These joints also serve to isolate one slab
segment from another, and may thus sometimes help control
slab cracking by permitting independent movement of the
segments.

Because the concrete will  tend  to shrink  away  from the
expansion joint material as it cures, a gap may exist between
the material and the adjoining concrete, enabling air leakage
down through the joint  into the  SSD system. This leakage
could inhibit the extension of the suction field generated by a
SSD pipe from extending across the joint into a neighboring
segment of the slab. Leakage through perimeter  expansion
joints could inhibit the suction field from extending to treat
the block foundation walls. Even where there  is  no visible
gap, the expansion joint material will probably not be creating
an air-tight seal.

Sealing  an expansion joint will be most important when: a)
the suction pipe is installed  near the joint, since  nearby leaks
have the greatest impact on suction field extension; b) sub-slab
communication is not good, since it is under those conditions
that marginal suction field extensions can least tolerate leak-
age; or c) the gap is wide. Thus, for  example, one situation
where sealing of expansion joints would likely  be important
would be the case where there is a perimeter joint, and where
poor sub-slab communication encourages location of the suc-
tion pipes around  the perimeter.  Where communication is
good, sealing these joints may be less crucial; however, even
with good communication, sealing could potentially improve
system radon reductions and reduce heat loss. A number of
mitigators report that they routinely seal the accessible por-
tions of expansion joints, especially perimeter joints (Br92,
Bro92, K192).
                                                        182

-------
One method for sealing an expansion joint would be to route
or chip out the top 1/2 in. of expansion joint material, creating
a channel along the length of the joint. After the debris is
vacuumed up, this channel could then be filled with urethane
caulk. Gun-grade (non-flowable) urethane caulk may be the
best choice in some cases, since flowable caulk can disappear
down into the porous material and any adjacent gaps (Bro92).
Chipping out the top of the expansion material can be a
time-consuming process, because the material is very pliable.
In no case should a channel be created simply by compressing
the joint material down, since the material will eventually
return to its original shape and dislodge any caulk that had
been applied on top of it (K192).                   /

To avoid the effort involved in chipping out the top portion of
the expansion material, the existing material might simply be
trimmed down to the level of the slab using a wire brush or
possibly a utility  knife  (if it extends above the slab). The
surfaces would have to be vacuumed to remove debris that
might reduce caulk adhesion. Gun-grade urethane would then
be spread on top of the existing material without chipping, and
worked down into any visible gap. This approach has  report-
edly worked well with perimeter expansion joints (K192).
With the perimeter joint, a generous bead of gun-grade caulk
is applied over the joint, and is forced down into any gaps
around  the material and attached to the slab and the wall
beside the joint. Although the caulk does not bond well to the
joint material itself, it will bond to the slab and wall, encasing
the entire joint.

Sometimes expansion joints will be inaccessible. Joints around
the perimeter can be concealed behind wall finish. Interior
joints, perpendicular to the front  and rear walls, will some-
times be concealed under frame  walls. It will often  not be
worthwhile to disrupt the finish attempting to seal the expan-
sion joint, although, as discussed previously, the final deci-
sion will depend  on the location of the suction  pipes, the
sub-slab communication, and the width of the gap. If an initial
mitigation system is not performing adequately, apparently as
the result of an unsealed inaccessible expansion joint, the
mitigator will need to consider taking steps to seal the joint or
making other system  modifications to compensate for the
leakage through the unsealed joint.

Other, small slab holes. Occasionally, other small holes
will be found through the slab. Such holes might result, for
example, where some previously existing slab penetration
may have been removed. Where such holes are observed, they
should be mortared, foamed, and/or caulked closed.

4.7.6  Openings in Block Foundation
 Walls

In addition to the openings in the slab, there will also be
openings in the foundation walls. Where the walls are hollow
block,  such air leakage paths into the wall  could  further
reduce the ability of the SSD system to develop a suction field
in the walls. The major wall openings which would be consid-
ered first for closure  would be:  the open voids  in the top
course of block, if there is not a course of solid cap blocks on
top; gaps around utility line penetrations through the wall; and
missing mortar or defects in the blocks.
As discussed in Sections 2.1,2.2.1, and 2.3.1, SSD seems to
"treat" block foundation walls, in large part, by intercepting
soil gas before it enters the void network. It is not clear that
SSD often treats the walls by establishing a measurable de-
pressurization or flow inside the wall. Moreover, even when
SSD pipes are located near block walls, the depressurizations/
flows that would be maintained in the wall by a stand-alone
SSD system appear to be very low. Under these conditions, it
is unclear how often any practical degree of wall sealing (e.g.,
sealing of top voids) would have a significant effect, given
that significant  wall openings will still remain (such as the
block pores  and mortar joint cracks) which will still provide
substantial cumulative leakage area.

Therefore, in cases  where sub-slab communication is fairly
good and where a stand-alone SSD system (with no  BWD
component)  is expected to do  a fairly good job in a given
house, it is unclear whether significant improvement in SSD
performance will often be achieved by an extensive effort to
seal openings in the block foundation  wall. Wall sealing will
most likely  be  of significant  value with stand-alone SSD
systems in cases where sub-slab communication is marginal
or poor, and/or where the block walls  appear to be a particu-
larly important entry route. Wall sealing can also be important
in cases where  a BWD component is going to be added to
supplement a SSD system, as discussed in Section 7. A BWD
component is most likely to be added for the same reasons just
stated, namely,  poor sub-slab communication and an impor-
tant wall radon source.

When sub-slab communication is good and a stand-alone SSD
system is planned, sealing of the most important wall open-
ings might help to reduce soil gas entry through the  walls,
even if it does not aid in depressurization of the walls by the
SSD system.

If wall sealing is attempted, the most important wall-related
opening to close would be the open voids in the top course of
block, if there is no course of cap  block.  The procedures for
sealing the top voids and other wall openings are discussed in
Section 7.7.1.

4.8     Gauges/Alarms  and Labelling

Even where a first-class job has been done in installing a SSD
system, the  system  will not continue  to provide low indoor
radon levels over the long term unless it is properly  main-
tained by the occupant.  Gauges or alarms are required to alert
the occupant when the suctions or flows in the system piping
fall to unacceptable levels. Labelling advises the homeowner
who the installer was (in the event that service is needed), and
which switches and circuit breakers control power to the fan.
Labelling also indicates which pipes and other elements are
components of the system, so that they are recognized by new
occupants who may move into the house in the future, and by
maintenance personnel. Labelling thus should reduce the risk
of the system being inadvertently  turned off or damaged by
future owners or by service personnel.
                                                       183

-------
4.8.1  Gauges and Alarms

EPA's interim mitigation  standards (EPA91b) require that
some type of mechanism or device be installed to monitor
system performance and warn of system failure. This warning
device must be plainly visible and easy to interpret.

Warning devices are mandatory because there is no other way,
other than frequent radon measurements, for homeowners to
ensure that the system is continuing to operate properly. It is
not sufficient to rely upon the sound of the fan, to reveal if it
has stopped operating (or if the bearings are about to fail).
Experience has indicated that the electrolytic capacitor in the
fan circuitry can  fail, causing a dramatic reduction  in fan
performance; however, the fan can continue to operate at this
reduced performance for an extended period, even though the
fan would sound as if it  were operating normally  (Fi91).
Moreover, the sound of the fan operation would not reveal the
development of system leaks, and reduced suction resulting
from increased air flow. And finally, fan operation can be so
quiet that a homeowner not paying particular attention to the
system might not detect that the system was not operating,
even if the fan stopped.

Pressure gauges. Many mitigators install mechanical  gauges
which measure suction in the SSD piping upstream of the fan.
Such gauges display the system suction, often in quantitative
units of  measure.  In some  devices, the reading  is
semi-quantitative, as discussed below. As the term is used
here, a "gauge" does not inherently incorporate a  visual or
auditory alarm  to alert the occupant  when suctions move
outside the acceptable range;  thus, gauges  rely upon the
occupant to take the initiative to read them. However, a gauge
could be used in conjunction with an alarm, discussed later
under Pressure-activated alarms.

Pressure gauges are mounted on the piping leading to  the fan
as one of the last steps in the installation  process,  as the
mitigator is confirming that proper suctions are being estab-
lished the system. The gauge must be marked to indicate the
expected range of suctions that the system should maintain
during typical operation, so that the occupant can readily tell
when the system is outside its normal operating range.

The exact values for the  upper  and lower  suctions  which
define the "typical" range for a given system will depend
upon the following factors: the  performance curve  of the
system fan; the flow characteristics of the sub-slab region; and
the nature of the SSD piping network (i.e., the suction loss in
the piping). This last factor can also be taken to include where
the gauge is mounted on the piping network, since a gauge
mounted near the fan may measure a much higher suction
than a gauge mounted near the slab remote from the fan. See
Sections 4.4 and 4.6.1 for  further discussion of the suctions
that may be expected with different fans, flow rates, and
piping configurations. The actual "normal operating range" to
be  marked on the gauge  for a given installation will  be
determined in large part by the post-mitigation system suction
and flow measurements that the mitigator makes in the  system
piping immediately following installation (see Section 11.3).
The occupant must be advised that a decrease in the suction
below the minimum level could indicate a problem such as
fan failure or leaks in the system piping, that must be cor-
rected if the system  is to  continue  to reduce radon levels
adequately. Likewise, an increase above the estimated maxi-
mum could indicate a problem such as blockage of the piping.

Pressure gauges should be mounted  in the piping relatively
near to the point where the suction pipe penetrates the slab,
since this will tend to provide the best measure of the suction
that is being maintained in the sub-slab pit. If the gauge were
mounted near the fan and if the fan were remote from the slab
penetration, the gauge readings would be strongly influenced
by the fan and could tend to remain  high. For example, if a
leak developed near the slab penetration, this  leak could
significantly reduce sub-slab suction.  However, the increased
system flows resulting from this leak  would increase the
pressure drop through the piping, so that the suction near the
fan might be affected much more modestly. Suction near the
fan could remain within the normal operating range marked
on the gauge, even though sub-slab suctions may potentially
have dropped below the acceptable level.

While some types of gauges can be mounted directly on the
suction piping, some gauges are sometimes or always mounted
on a wall a short distance from the SSD suction piping, and
are connected to the piping by a length of 1/8-in.  diameter
flexible tubing. The gauge should be mounted at a location
such that the entire length of this  tubing remains inside the
basement or living area. Some mitigators have reported that
when this narrow tubing extends outside the living envelope.
e.g., to outdoor piping associated with an exterior stack or to
piping in an attic, moisture can condense inside the tubing
(An92, K192). This condensate will block the  connection
between the gauge and the suction pipe, causing the gauge to
give a reading of zero suction, incorrectly suggesting a prob-
lem with the SSD system.

Several types of pressure gauges are in common use, having
the sensitivity required  to measure suctions in the range
commonly seen in SSD piping.

 •  Magnehelic®gauge. Dwyer Magnehelic" gauges, which
    commonly cost about $40-$55 (He91b, He91c), are the
    most expensive gauge. However, they are easy to read
    and can be mounted directly on the suction pipe if. de-
    sired.  Whether mounted on the pipe  or on a nearby wall,
    it is connected to the pipe by narrow tubing which ex-
    tends  from a fitting on the gauge to a hole in  the pipe.
    Magnehelic gauges can measure suctions up to 0.5 to 2
    in. WG, and even up to 50 in. WG, depending upon the
    particular gauge; the most sensitive gauge has a sensitiv-
    ity of about ± 0.01 in. WG.

 •  Curved inclined manometer. Somewhat more  sensitive
    than the U-tube manometer at low suctions, the inclined
    manometer has the disadvantage that it cannot be mounted
    directly on the SSD pipe. It must be mounted on a flat
    surface (such as a wall) nearby,  and is connected to the
    pipe by tubing. An inclined manometer may also have to
    be re-zeroed periodically (Bro92). Inclined manometers
    commonly sell for about $18-$25. One typical inclined
                                                       184

-------
    manometer on the market can measure suctions up to 3 in.
    WG, with a sensitivity of roughly 0.02 to 0.1 in. WG at
    suctions below 1 in. WG.

 •  U-tube manometer. The U-tube manometer is the least
    expensive device, costing $8-$15. It can  be  mounted
    directly on the vertical SSD suction pipe, although mount-
    ing on a nearby wall would provide a more secure loca-,
    tion. In either case, a length of flexible tubing would
    connect the manometer to a hole in the pipe. One typical
    U-tube manometer on the market can measure suctions
    up to 4 in. WG, with a sensitivity of 0.1 in. WG.

 •  Floating-ball device. One vendor markets a device which
    attaches directly to the side of the suction pipe. A stream
    of house air is drawn up through the device into a small
    hole drilled in  the side of  the SSD  pipe. When the
    homeowner places a finger over the inlet hole for  house
    air flow into the device, the suction in the pipe causes a
    ball within the device to float upward,  either to the
    "green" zone (fan operating properly) or the "red" zone
    (maintenance required). This device, which is less quanti-
    tative than the  others, markets for about $10-$20. A
    number of mitigators have reported problems  with this
    device, due to difficulties in ensuring proper movement
    of the floating ball and difficulties in seeing the ball.

Ammeter (measuring current to fan). One vendor mar-
kets a gauge which, rather than measuring system suction,
measures the current being drawn by the fan as a surrogate for
suction. This gauge utilizes the principle that, based upon the
fan performance curve and the fan motor performance, the
power being drawn by the motor increase as the air flow being
moved by the fan increases.

This qualitative gauge, which is mounted on a wall, is placed
in the wiring leading to the fan.  A needle on the  gauge
indicates whether the fan is drawing current within an accept-
able (green) range, or whether current has  dropped or risen
into an unacceptable (red) range. The gauge does not provide
a quantitative measure of the actual amperes being drawn.

The acceptable value of the current is set  by the  mitigator
immediately after installation, based upon the current being
drawn by the fan after the system has been adjusted to give
satisfactory performance.

This ammeter gauge has been illustrated previously in Section
4.6.4, in connection  with Figure 25B.

Pressure-activated alarms. In addition to suction gauges,
discussed above, there are also alarms on the market which
provide a visual and/or and audible signal  that comes on  if
suction drops  below a pre-set value. These alarms are also
referred to as pressure switches. Unlike gauges, alarms do not
provide a continuous measure of the suction that exists in the
piping.  Rather, they only indicate when the suction falls
below the acceptable minimum set by the mitigator. Also
unlike gauges, alarms/switches only indicate when suction
has fallen unacceptably (e.g., due to fan failure or develop-
ment of a piping leak); they will not indicate when suction has
risen unacceptably (e.g., due to pipe blockage).
Alarms may be used in combination with one of the gauges
listed above, or may stand alone with  no gauge. Gauges
(including gauges combined with a visual or audible alarm)
are preferred over alarms alone. The gauge provides a con-
tinuous, often quantitative measurement of how system suc-
tion may be changing. Many alarms cost more than the most
expensive gauges.

Visual and audible alarms require a power source. They must
be plugged or wired into a house circuit different from the one
into which the fan is wired, so that failure  of the fan circuit
(e.g., tripping of the circuit breaker) would not simultaneously
disable the alarm. The alarm should be plugged into a circuit
without a switch that can be  turned off by the occupant.
Alarms should  automatically reset  when power is restored
after service or after a power supply failure. According to the
current draft of EPA's  final Radon Mitigation Standards,
battery-powered alarms would be unacceptable unless they
are equipped with a low-power warning feature, since the
occupant might forget to consistently replace the batteries at
the required intervals over the years.

Where a gauge has not been installed along  with the alarm, it
is recommended that the alarm have a light (e.g., a green light)
which remains  on continuously while the fan  is operating
properly, confirming that the alarm is functional. Audible
alarms should be wired so that they can be  turned off by the
homeowner after indicating a problem with the SSD system.

Like pressure gauges, pressure-activated alarms  should be
mounted on the suction pipe toward its penetration through
the slab.

Location of gauges and alarms. Gauges  and visual alarms
should be mounted in a location frequented by the occupants
on a regular basis. They should not be mounted in closets
unless this is unavoidable. Where the gauge or alarm must be
placed in such a location, the gauge or visual alarm should be
in combination with an audible alarm.

As discussed previously for gauges and alarms which measure
system suction, the gauges or alarm should be mounted near
the point where the SSD suction pipe penetrates the slab.

4.8.2 System Labelling

In accordance  with EPA's interim mitigation  standards
(EPA91b),  all components  of the  SSD system  should be
labelled, to avoid inadvertent disabling of the system by future
owners/occupants of the house or by service personnel.

A label should be posted at a central location on the mitigation
system, or on the electric panel or other prominent location,
including a system description, and including the name and
telephone number to contact if a problem arises with the
system. It is also recommended in the current draft of EPA's
final Radon Mitigation Standards that this label also include
the date of installation and an advisory to re-test the house for
radon  every 2  years. This  label should be legible from a
distance of at least 2 ft.
                                                       185

-------
Any electric switch controlling power to the SSD fan must be  Visible portions of the system, including the fan and interior
labelled. If the  fan or any electrically  powered alarm is  piping, should be labelled at at least one location on each story
plugged into an outlet, this outlet is considered a switch and  of  the  house.  The labels  should read, "Radon  Reduction
should be labelled. Also, the circuit breaker switch or the fuse  System."
should be labelled at the breaker or fuse  box, indicating the
circuit into which the fan has been wired, and the circuit into
which any electrically powered alarm has been wired.  Such
labelling should help prevent homeowners from inadvertently
turning off the fan or the alarm at the box.
                                                        186

-------
                                            Section 5
              Design  and Installation of Active  Drain-Tile
      Depressurization S>ystems  (Sump  Depressurization)
Where a sump with drain tiles exists in a basement house,
sump/DTD will  commonly  (but not always) be the  ASD
variation selected for installation, rather than SSD. The drain
tiles provide an in-place network to aid in distribution of the
suction field, usually around the slab perimeter. Distribution
of the suction around the perimeter should help ensure effec-
tive treatment of the wall/floor joint, often one of the most
important radon entry routes. In addition, if the suction pipe is
installed in the sump, the sump pit provides a ready-made hole
through the slab for the suction pipe.

One typical configuration of a sump/DTD system is illustrated
in Figure 3.

Commonly, the drain tiles emptying into sumps form a com-
plete or partial interior loop, inside the footings beneath the
slab. Sometimes, sumps will have exterior drain tiles, forming
a complete or partial loop around the outside of the footings.
If sub-slab communication is good, sump/DTD  will  often
perform very well: a) with interior tiles, even if the tiles form
only a partial loop; or b) with exterior tiles, if the tiles form a
complete or largely complete loop (i.e., around at least three
sides of the house). If communication is marginal or  poor,
sump/DTD will likely perform best as a stand-alone system
with interior tiles forming a complete loop. With marginal/
poor communication, sump/DTD may also perform well with
nearly complete exterior loops, if there is not a major soil gas
entry route toward the interior of the slab (such as  a block
fireplace structure penetrating the slab).

Under conditions other than those listed above, sump/DTD
may sometimes  have to be supplemented by SSD suction
pipes. These other conditions include: a) marginal/poor com-
munication, partial  interior  loop;  b) good communication,
partial exterior loop; and c) marginal/poor communication,
partial exterior loop. However, good  performance has some-
times been reported with DTD/remote discharge even under
some of these less favorable conditions. One mitigator reports
achieving reasonable performance with partial exterior loops
around basements having no sub-slab aggregate, apparently
due to reasonably good permeability in the underlying soil
(K192).

Occasionally, sump pits will be encountered that do not have
drain tiles emptying into them. Where such sump pits open to
the sub-slab fill, these sumps can be capped and suction drawn
on them as discussed  in Section 5.5.2, utilizing them as a
ready-made hole through the slab. However,  as discussed in
Section 4.5.3, such an installation would actually be a varia-
tion of SSD, and would not be sump/DTD.

Many of the details involved in designing and installing sump/
DTD systems are similar those for SSD systems, discussed in
Section 4. Thus, the discussion in Section 5 will often refer to
the previous section. The discussion in this section will focus
on those  design and installation features unique to  sump/
DTD.

The discussion in this section draws heavily from the detailed
review of available data on active sump/DTD systems, pre-
sented in Section 2.3.2.

5.1   Selection of the Number of
Suction Pipes: Need  for a SSD
Component to Sump/DTD System

Often, sump/DTD systems will not need to be supplemented
by a SSD component. That is, suction on the drain tiles alone
will be sufficient, and it will not be necessary to  install
additional suction pipes into the sub-slab region at locations
remote from the drain tiles. In these cases, one pipe, drawing
suction on the drain tiles at the sump or at any other conve-
nient location, will generally be sufficient. The tiles should
effectively distribute the suction along their entire length.

More than one pipe could be needed to draw suction on the
drain tiles in cases where a portion of the drain tile loop is
isolated from the remainder of the loop, as a result of physical
damage or of silting. In such cases, a separate pipe could be
needed to treat each isolated segment, if treatment of only one
segment fails to provide sufficient radon reductions, hi prac-
tice, it would usually be difficult to ascertain that some
particular segment is thus isolated.

If suction on the drain tiles creates flows so high that adequate
suction cannot be maintained in the piping or under the slab,
the answer will almost never be to install a  second suction
pipe into the drain tile loop (and/or additional fan capacity) in
an effort to handle these flows. Rather, the answer will usually
be to conduct additional sealing as necessary to reduce air
leakage into the system, as discussed in Section 5.7. In sum-
mary, there will rarely be occasions in practice when a second
suction pipe will be tapped into the drain tile network.
                                                    187

-------
One of the more common situations under which a SSD
suction pipe will be required in addition to the sump/DTD
suction pipe, will be the case where the SSD pipe is needed to
treat a slab-on-grade wing which adjoins the basement with
the sump. The drain tiles entering the sump often loop only
around the basement footings, so that suction on the sump
alone may  not directly treat the adjoining slab.  (There are
cases, usually with exterior loops, where the tiles will extend
around the slab on grade as well as the basement; in that case,
suction on  the tiles at one location should treat both slabs
directly.)

Where the tiles loop  only  around  the  basement,  then, as
discussed in Section 4.1, it is generally advisable to install a
SSD pipe under the adjoining slab on grade, since that step
usually appears to improve radon reductions, especially in the
living area on the adjoining slab. Where the tiles loop only
around the  basement, installing the SSD pipe under the ad-
joining slab is most important in cases where the communica-
tion beneath the basement slab is marginal or poor, or where
the radon source term is high. In those cases, it will sometimes
not be possible to reduce the living area on the adjoining slab
below 4  pCi/L without a  pipe  treating that slab  directly.
Where communication beneath the basement slab is good and
the source term not particularly high, sump/DTD in the base-
ment alone is more likely to reduce the entire house below 4
pCJ/L; however, even in this case, a SSD pipe beneath the
adjoining slab will usually still provide additional reductions.
The SSD pipe beneath the adjoining slab will  usually be
installed from inside  the basement, as discussed in Section
4.5.5, and manifolded into the sump/DTD piping, leading to a
single fan.

Sometimes, SSD pipes will be needed to help treat the base-
ment slab,  supplementing the sump/DTD system. As listed
previously, such supplemental basement SSD pipes will most
likely (but will not always) be needed under the following sets
of conditions: a) marginal/poor sub-slab communication, par-
tial interior loop (i.e., on less than three sides of the house); b)
good sub-slab communication, partial exterior loop, low per-
meability in native soil; and c) marginal/poor communication,
partial exterior loop, low  soil permeability. Supplemental
SSD pipes  may sometimes also be necessary with complete
exterior loops, if communication is marginal/poor, if the soil
permeability is low, and if there is a major entry route toward
the slab interior. The pre-mitigation  diagnostics  can some-
times help to foresee the need for supplemental SSD pipes, by
suggesting the nature of the drain tile loop or by identifying
the sub-slab communication.

Inspection of the sump during the visual survey, together with
knowledge of construction practices in the area, with  any
available construction plans, and with observations by the
homeowner during construction, can help suggest the nature
of the tiles. For example, two tiles entering the sump at 90°
angles, parallel to the perimeter walls at the corner where the
sump is located, would suggest (but not prove) the presence of
a complete interior loop. A single tile penetrating  into the
sump parallel to a foundation wall would suggest that the tiles
are inside the footing, but could be suggesting that a complete
loop may be less likely. A single tile penetrating the sump
perpendicular to the adjoining foundation wall could suggest
an exterior loop. Where the ultimate direction of the tiles is
ambiguous, the tiles might be probed by inserting a plumber's
snake from inside the sump, to determine changes in direction
near the sump. It is difficult to confirm the extent of the loop
(i.e., complete or partial) unless the homeowner observed the
installation of the tiles when the house was under construc-
tion, or unless this is known from as-built construction draw-
ings or from established practices of the particular builder.

If the basement sub-slab communication is not apparent from
observed aggregate under the slab (e.g., around the tile pen-
etrations through  the sump wall) or from experience  with
construction practices  in the area, qualitative suction  field
extension testing with the diagnostic vacuum cleaner could be
conducted, as discussed in Section 3.3.1. The problem is that,
unless it is known where the tile loop may be incomplete, this
communication information will not give definitive guidance
regarding how many supplemental SSD pipes are needed, and
exactly  where they should  be installed. Accordingly, rather
than doing pre-mitigation suction field extension measure-
ments, it may often be most cost-effective to simply install the
sump/DTD system,  and to then determine the appropriate
number  and location of supplemental SSD pipes (if any turn
out to be necessary) from sub-slab suction field measurements
with the system operating.

If any SSD pipes turn out to be needed in the basement  slab,
supplementing the sump/DTD system, the SSD component of
the system would be designed and installed as described in
Section  4. Any  SSD pipes  would commonly be manifolded
into the  sump/DTD system piping, leading to a single fan.

5.2    Selection of DTD Suction Pipe
Location:  At  Sump or  Remote
The pipe to draw suction on the drain tiles can be installed into
the drain tile loop  at any location around the loop. Often, the
suction pipe is installed remote from the capped sump, as in
Figure 3. Sometimes, the pipe is installed at the sump, through
the sump cover.

Some mitigators prefer to install the suction pipe into the tiles
remote from the sump, for several reasons. A primary reason
is to simplify subsequent maintenance on the sump pump or
inspection of the  sump if water drainage problems occur
(Sh91, Br92, Bro92, K192). Having the suction pipe away
from the sump cover not only simplifies the removal of the
sump cover, but it also reduces the risk that the pipe penetra-
tion through the cover will not be properly resealed by the
pump repairman or homeowner after the maintenance is done.

A related advantage of installing the suction pipe at a  remote
location is that any subsequent leaks that develop in the sump
cover would have a less significant effect on sump/DTD
system performance. The sump cover is the location where
leaks are most prone to occur. As mentioned above, leaks
could develop during subsequent sump maintenance, if the
owner or service personnel do not reseal the cover or some of
the cover penetrations adequately after inspections or repairs
are completed. Leaks could also develop if any of the seals
break over time, or if a water trap through the cover dries out.
On this basis, it would theoretically be desirable to tap into the
                                                      188

-------
drain tiles as far away from the sump as possible. House air
leaking into the sump would then suffer a significant suction
loss in flowing through the drain tile loop over to the suction
pipe, especially in view of the wall friction created by the
corrugations in the flexible corrugated drain tiles. As a result,
the suction at the sump would be reduced (reducing the air
flow through the sump leaks), and sufficient suctions might be
maintained at other locations in the loop such that the system
could still be effective.

Another reason for tapping into the tiles at a location remote
from the sump could be to facilitate routing of the  exhaust
stack. For example, if the sump were on the opposite end of
the basement from an adjoining slab-on-grade  garage, one
might elect to tap into the tiles at the opposite end of the
basement, beside  the wall adjoining the garage, so  that the
exhaust piping could conveniently be routed up through the
garage, with the fan in the garage and the exhaust penetrating
the garage roof.

Where an exterior loop of tiles connects into the sump, a
further advantage to remote installation of the suction pipe is
that all of the piping will then be outside the house. This can
sometimes be desirable for aesthetic reasons.

Of course, tapping into the tiles remote from the sump re-
quires that the configuration of the tiles be known (e.g., it is
known that they  form a complete  interior loop).  This is
necessary so that the mitigator is able to core a hole  through
the slab at a remote location, with reasonable certainty that the
tiles  will be approximately underneath.

If the DTD suction pipe is installed through the slab to tap into
the drain tiles remote from the sump, the sump must still be
capped with an air-tight cover. Otherwise, basement air would
be drawn down into the sump and through the tiles, dramati-
cally reducing system suctions and performance, increasing
the house heating/cooling penalty, and increasing the risk of
combustion appliance backdrafting.

The prior discussion focuses on installation of the suction pipe
remote from the sump. However, many mitigators routinely
install the pipe through the sump cover. Even mitigators who
commonly install  the pipe remote from the sump sometimes
find  it desirable to install the pipe through the cover for one
reason or another. Certainly, installation through the  cover is
simpler than remote installation, since  it avoids the need to
core through the slab or excavate outdoors to expose the tiles
at a  remote location. Installation of the suction pipe through
the sump is a reasonable as well as common approach. How-
ever, as discussed in Section 5.5.2, this installation must be
done in a manner which facilitates subsequent access to the
sump for maintenance.

The  preceding discussion has addressed the location of the
suction pipe which is used to draw suction on the drain tiles
(the  DTD suction pipe). In those cases where supplemental
SSD suction pipes also turn out to be required, the location of
the SSD pipe(s) would be selected as discussed in Section 5.1.
5.3    Selection  of Suction  Pipe Type
and  Diameter
The considerations in selecting the type of suction pipe (usu-
ally either lightweight or Schedule 40 PVC, PE, or ABS
piping) are the same for sump/DTD systems as those dis-
cussed in Section 4.3.1 for SSD systems. Where the suction
pipe taps into an exterior  drain tile loop remote from the
sump, in which case the piping will necessarily be outdoors,
the piping should be Schedule 40, coated with paint or a UV
protectant, for improved UV resistance.

Regarding the diameter of the piping, 4-in. diameter piping is
typically used in sump/DTD systems. Sump/DTD systems
tend to have flow rates somewhat higher than those in SSD
systems (commonly  50-150 cfm with  the  90-watt in-line
tubular fans, compared to 20-100+ cfm in SSD systems). The
higher flows in sump/DTD systems probably result from more
air leakage into the system through the wall/floor joint and
through block foundation walls, since suction is being drawn
immediately beside the entire perimeter. Perhaps there is also
some leakage associated with the sump cover. Where the tiles
are outside the footings, some of the air might also be outdoor
air drawn down through the soil from grade, especially when
some portion of the tiles are at a relatively shallow depth.

Because of these higher flows, 4-in. diameter piping will often
be preferred  for sump/DTD systems. Even at the lowest
commonly-observed flow rate of 50 cfm, the  flow velocity in
3-in. piping would be about 1,000 ft/min, the value at which
flow noise can start to become objectionable. Even with the
4-in. piping, flow noise could become objectionable at 90 to
130 cfm, in the middle to upper flow range for sump/DTD.

The example system considered in  Section 4.3.2  had an
equivalent piping length of 70 ft, consisting of 35 linear ft of
straight piping and two mitered 90°  elbows. Referring to the
friction loss curves in Figure 13, the higher flows seen in
sump/DTD systems (100 to 150 cfm) would result in a friction
loss of 0.4 to 1.0 in. WG if this example system consisted of
4-in. piping, a loss that could be handled by the 90-watt in-line
tubular fans discussed in Section 4.4.1. However, if the piping
were 3 in. diameter, the suction loss would be 1.6 to 4 in. WG,
a loss  that would challenge even the in-line radial blowers
discussed in Section 4.4.2. If a length of 3-in. piping had to be
used with a sump/DTD system in order  to fit the piping into
existing space constraints (e.g., inside stud walls), it would
seem desirable to limit the length of the narrower piping to the
extent possible.

But having said this, it must be recognized  that if the 3-in.
piping were used in the example system, and  if suction losses
as calculated above became too high for the fan to handle, the
system flows would drop, and the fan would operate at a
different point on its performance curve. Suctions in the drain
tiles would thus also necessarily drop. The lower flows would
also mean lower velocities and hence reduced flow noise. In
view  of the common effectiveness of sump/DTD systems,
especially when sub-slab aggregate is present, such a reduc-
tion in flows and suction may not always result in a serious
degradation in  system performance. Thus, the use of 3-in.
piping may not in fact always be a serious problem. Neverthe-
                                                       189

-------
less, it would still seem desirable to use 4-in. piping where
possible.

In view of the above discussion, the use of 6-in. piping in an
attempt to reduce suction loss is probably not warranted in
most cases. Where the sub-slab characteristics  are such that
sump/DTD flows are toward the lower end of the range, the
standard in-line tubular fans will be able to handle the suction
loss in 4-in. piping. Where flows tend toward the upper end of
the range, any reduction in system flows and in  drain tile
suctions resulting from the use of 4-in. rather than 6-in. piping
will probably often not create a serious degradation in system
performance, even with the standard fans. Six-inch piping
might be considered in occasional cases where flow noise in
narrower pipe becomes objectionable.

5.4   Selection of Suction  Fan

Since the flows from sump/DTD systems tend to be relatively
high, the 90-watt in-line  tubular  fans  discussed in Section
4.4,1 will usually be  the best choice. At the flows typically
maintained by sump/DTD systems with  the  90-watt fans
(50-150 cfm), these fans can commonly establish suctions of
about 0.75-1.0 in. WG in the system piping, based upon the
published fan performance curves. As discussed previously,
this suction will usually be sufficient to handle friction losses
that will occur in 4-in. diameter piping at those flow rates.

Some mitigators find that, under best-case conditions (a com-
plete interior drain tile loop and good sub-slab communica-
tion), the smaller 50-watt in-line tubular fans are sufficient for
sump/DTD systems. These fans will draw smaller flows and
will maintain somewhat lower system  suctions. Because of
their lower flows,  they will also  sustain  less friction  loss
through the piping, which will  partially compensate for the
lower system suctions. In areas where effective long-term
performance has been demonstrated with  the 50-watt fans,
these fans can be considered for  best-case sump/DTD sys-
tems. However, in other cases, selection of a 90-watt fan will
provide the best assurance that effective performance will be
maintained over the range of conditions that the  system is
likely to encounter over time.

The in-line radial blowers discussed in Section 4.4.2 could
also be considered for use on sump/DTD systems. However,
at the relatively high flows in these systems, the higher
suction capabilities of the radial blowers would not be uti-
lized, and they would be an unnecessary investment.

Because of the relatively high flows usually encountered with
sump/DTD systems, the high-suction/low-flow fans discussed
in Section 4.4.3 for SSD systems in low-flow houses  will
essentially never be suitable for sump/DTD installations.

5.5   installation of  a Suction Pipe
into the Drain Tile Loop
As indicated previously, suction can be drawn on the drain
tiles in one of two alternative ways. The suction pipe can be
installed into the tiles at some point remote from the sump; or,
the pipe can be installed through the sump cover.
5.5.1  Suction Drawn on Tiles Remote
from Sump

Where the likely configuration of the drain  tiles can be
estimated  fairly reliably (e.g.,  based  upon common  house
construction characteristics in the area), it can be beneficial to
draw suction on the tiles at some location remote from the
sump. While installation of the pipe remote from the sump
will make the installation slightly more difficult, remote in-
stallation offers the benefits of facilitating subsequent sump
pump inspection and maintenance and of sometimes facilitat-
ing the routing of the exhaust piping,  among other possible
benefits.

Interior tile loops—connecting the remote suction pipe
(indirect approach). Where the tiles are thought to form a
complete loop inside the footings, beneath the basement slab,
a hole through the slab is created through the slab near the
foundation wall at the point at which it is desired to install the
suction pipe. The hole  is made directly over the estimated
position of the drain tile. Since the tile will usually not turn out
to be exactly beneath the slab hole, some exploratory excava-
tion by hand through the hole will be required in order to
locate the tile.

Commonly, the hole through the slab will be  a  5- to 6-in.
diameter hole prepared using a rotary hammer,  or using  a
coring drill, as described in Section 4.5.1 (see Drilling the
hole through the slab) for vertical interior SSD pipes. When
the drain tile is located by excavation beneath this hole, a hole
through the wall of the tile is made, as illustrated in Figure 3.
This procedure results in a hand-excavated pit beneath the
slab, as discussed in Section 4.5.1, with the tile loop now
opening into this pit. The pit should be excavated to some
depth below the drain tile, and to some distance to the right
and left of the hole that has been created in the tile wall, in an
effort to  avoid the tile becoming silted  shut or otherwise
plugged by sub-slab fill or soil entering the hole that has been
created in the tile (e.g., due to settling or water movement).
The suction pipe is then sealed into this slab hole, exactly as
described in Section 4.5.1 for a vertical SSD pipe (see Mount-
ing suction pipes through the slab, and Figures 14 and 15).
The only difference is that, now, this pipe is in fact a DTD
pipe (at least in part), since its suction is extending through the
sub-slab pit into the drain tile loop.

Interior tile loops—connecting the remote suction pipe
(direct approach). With the configuration described above
and shown in Figure 3, the suction pipe does not physically
connect to the drain tile. Depending upon  how accurately (or
inaccurately) the mitigator guessed the location  of the tile
when the slab hole was drilled, the suction pipe may, in fact,
be several inches away from the drain tile. Where the drain
tiles are inside the footings, and especially where sub-slab
communication is good, a physical connection is probably not
necessary. However, where sub-slab communication is poor,
it may become more important to  connect the suction pipe
directly into the tiles, to ensure most effective distribution of
the suction through the tiles.

If a better connection is desired,  the suction  pipe can be
inserted directly into the drain tiles. One approach  for accom-
                                                      190

-------
plishing this would be to install a PVC "T" fitting into the
drain tiles with the leg extending upward. The suction pipe
would then be cemented into this upward leg.

To install a T, a second hole may have to be made through the
slab directly over the tile, once the tile has been located by the
initial exploratory excavation, if the first hole was not directly
over the tile. This second hole would have to be 6 in. or
greater in diameter,  large enough to  accommodate the  T
fitting. A pit would be excavated beneath this second slab hole
as necessary to expose the tile. A small section of the tile
would be cut out, so that the fitting can be inserted into the
tile. The T fitting would then be fit into this hole in an inverted
fashion, so that the two ends of  the top of the  T can be
connected to the two opened ends of the drain tile, and so that
the leg of the T is protruding straight upward through the slab
hole. Each of the cut ends of the tile would connected to the
top of the T. If the drain tiles are rigid perforated ABS or PVC,
the ends of the T might be connected using straight flexible
black PVC couplings, similar to those described in Section
4.6.4 for  mounting fans.  Hose clamps might be used to
connect the couplings to the T and  the drain tiles. If the drain
tiles are flexible corrugated black polyethylene or polypropy-
lene, the tile might be connected to the ends of the rigid T
using hose clamps, or using screws and urethane caulk.

When the drain tile is the flexible corrugated material, one
mitigator suggests fabricating a saddle out of a flexible corru-
gated T fitting, rather than using a rigid PVC T (K192). This
saddle would be cemented over a hole on the top of the drain
tile. This approach is further discussed in Section 6.5.

Once the suction pipe has been connected to the drain tiles
with the rigid or flexible T fitting, the slab penetration must be
sealed. If the slab hole was only  about 6 in. diameter, the
sealing  might be accomplished by stuffing some  support
material (such as backer rod or foam) into the annulus be-
tween the pipe and  the side of  the  slab,  and filling the
remaining channel above the support with urethane caulk and/
or mortar, analogous  to the options shown in Figure 14. If a
hole much larger than that had to be made in the slab, it could
be  desirable (and the suction pipe would have more lateral
support) if the pit were filled with  crushed rock and the slab
restored with mortar. Channels would  be tooled around the
outer perimeter of the new mortar patch, and at the juncture
between the mortar  and the  suction pipe; these channels
would be filled with flowable urethane after the mortar had
set.

Interior tile loops—supporting the remote suction pipe.
When the remote suction pipe is connected to an interior loop
using the so-called indirect approach above, the pipe may be
supported at the slab, if the pipe is mounted using one of the
techniques illustrated in Figure 15 (or an equivalent tech-
nique). If the pipe is not supported at the slab, it can be
supported overhead using, e.g., one of the methods illustrated
in Figure 16 or 17.

Since an interior drain tile loop will usually be within about 6
in.  of a perimeter foundation wall, a mitigator could also
choose to offset the suction pipe against the wall by connect-
ing a pair of 45° elbows to the pipe where it comes up out of
the slab. The riser could then be attached to the perimeter wall
using,  e.g., one of the  methods shown in Figure 29 for
attaching an exterior stack to the side of a house.

When the remote suction pipe is connected to an interior loop
using the so-called direct approach above,  the pipe will be
supported beneath the slab if an inverted rigid PVC T fitting is
used. The T will be resting on the soil beneath. If a flexible
corrugated saddle fitting is used, according to the second
direct option above, the suction pipe would not be supported
beneath the slab, and would have to be supported overhead in
some manner.

Exterior tile loops—connecting the remote suction pipe.
The above discussion focuses on the case where the tiles are
beneath the slab, inside the footings. If the tiles draining into
the sump were outside the footings, it would be possible to tap
into  the tiles outdoors. This outdoor connection  would be
made in a manner similar to that described in Section 6.5 (and
illustrated in Figure 4) for the DTD/remote discharge case.
The  difference  would be that, in the sump/DTD case, the
connection would necessarily always be in the tile loop itself.
Unlike the DTD/remote discharge example in Figure 4, there
would be no discharge  line to  a dry well or above-grade
discharge point.

Capping the sump. Even when the suction is drawn on the
tiles remote from the sump, the sump should be capped with a
cover that is largely air-tight, to avoid excessive flow of house
air into the sump/DTD  system. The sump cover would be
installed  as described in Section 5.5.2 (and as illustrated  in
Figures 3 and 33), except that the PVC suction pipe would not
penetrate the sump cover. As discussed in Section 5.5.2, this
cover must be attached over the sump in a manner which will
facilitate its removal and replacement during sump mainte-
nance.

As part of capping the sump, a check valve should be inserted
in the water discharge line running from the sump pump.
Otherwise, outdoor air could be drawn through the water pipe
and pump into the sump/DTD system, reducing the suction
field that can be established by the system.

Efforts to carefully seal the  sump become somewhat less
critical when suction is drawn remote from the sump, since
leaks at the sump will have a less dramatic effect on sump/
DTD performance under that condition. However, significant
sump leaks could still degrade system performance. Thus, the
mitigator is well advised to seal the sump carefully, even in
cases where suction is drawn remote from the sump.

5.5.2  Suction Drawn on  Capped Sump

The other common method for drawing suction on the drain
tiles is to insert the suction pipe through the air-tight cover
over the sump. When suction is drawn directly on the sump in
this manner, it is increasingly important that the sump cover
remain largely  air-tight; leaks in the cover could result  in
substantial short-circuiting of house air into the system, re-
ducing system performance.
                                                       191

-------
Design of sump covers—general Figure 3 illustrates one
common sump cover design, a flat circular disk which rests on
top of the slab around the sump hole. While that figure shows
the suction being drawn remote from the sump, the same
cover design could be considered when suction is drawn at the
sump. Some other sump covers are illustrated in Figure 33;
these covers are all  illustrated for the case where suction is
being drawn at the sump.

Pre-fabricated sump covers of the type illustrated in Figures 3
and 33  can  be  purchased. These pre-fabricated covers are
circular disks of a diameter adequate to rest on the lip of the
sump liner (if this lip is exposed), or on top of the slab around
the sump hole. They are commonly fabricated out of materials
such as molded polyurethane, molded polypropylene, molded
ABS, or 1/4- to 1/2-in. thick sheets of hard clear plastic or
Lucite.

Mitigators can fabricate covers such as those in Figures 3,
33 A, and 33B out of sheets of hard plastic or Lucite, or from
sheet metal. It  is recommended that the covers  be  strong
enough to support the weight of a person without breaking, to
reduce the risk of damage to the cover by the house occupants
and by service personnel.

Where suction is to be drawn at the sump, the sump covers
must have at least two to three penetrations: one (about 4.5 in.
diameter) for the suction pipe; one (1 to 2 in. diameter) for the
water discharge line from the sump pump; and one for the
electrical connection to the sump pump. Sometimes the pump
discharge  line and the electrical connection pass  through a
single hole, reducing the number of holes from three to two.
When suction is to be drawn on the drain tiles remote from the
sump, as discussed in  Section 5.5.1, the hole for the suction
pipe would not be needed.

Another opening would be required for a waterless trap as
shown in Figures 33A and 33B, if water is expected from on
top of the slab (discussed later). Additional openings would be
required if specific drain lines (such as  an air conditioner
condensate drain) were to be installed through the cover. Such
additional  openings should be minimized, to avoid complica-
tions  with maintaining the cover's seal over time (e.g., as
cover removal and  resealing are required to permit pump
maintenance). As discussed later,  all of these penetrations
must be effectively sealed when the cover is installed.

Fabrication of the cover out of some transparent material such
as Lucite or Plexiglass will facilitate subsequent viewing of
the pump and the water level in the sump by the home-owner.
A number of non-transparent covers on the market, such as
tliat in Figure 33C, include a clear pump viewing window in
the cover to permit  such inspection. Such a window some-
times also serves as an access door, through which service
personnel  could gain  limited access to the pump (e.g., for
flipping the switch on the pump) without removing the sump
cover and  thus without breaking the seals around the cover
perimeter and around the various cover penetrations (Mes91,
Bro92).  Such a door would be screwed onto the cover with a
gasket to provide a seal.
Design of sump covers to accommodate water from on
top of the slab. When water is expected to enter the sump
from on top of the slab (e.g., from air conditioner condensate
drains or sink overflows), provisions must be made to permit
this water to pass through the cover from on top. Where this
water is  from a specific source such as an air conditioner
condensate drain, one option would be to install a permanent
penetration through the cover for this drain line. However, for
cases where the water will be flowing over the top of the slab
(such as  from an overflowing sink), the best approach is to
install a trap through the cover.

If the water flow is frequent enough such that a standard "P"
trap will  remain full of water, such a standard water trap can
be installed. However, in cases where the flow will be infre-
quent or uncertain, such  a trap could dry out, unless the
homeowner consistently added water to the trap. It would be
unwise to rely upon the homeowner remembering to continu-
ally add water over the years. If the trap dried out, this would
create an opening through the cover which could dramatically
reduce the effectiveness of the cover's seal, and thus signifi-
cantly reduce the effectiveness  of the sump/DTD  system,
especially when suction is being drawn at the sump. Accord-
ingly, as a general procedure,  it would be advisable in such
cases to install a "waterless" trap through the cover (e.g., as
illustrated in Figure 32B). As discussed in Section 4.7.3, when
water is not flowing through the waterless traps, a weighted
ball or ring seats in the trap opening, ensuring that the trap is
closed even if the water dries out.

Where water is expected to enter the sump from on top of the
slab, it would be preferable for the cover to be recessed below
the top of the slab, if possible, thus forming a depression over
the sump which will aid  in the collection  of water at that
location.  Alternatives for recessing the cover in that manner
are illustrated in Figures 33A and 33B.

In some cases, there is a sump liner insert forming the walls
(and perhaps the bottom) of the sump, and the top of this liner
ends at a level below the top of the slab. In this configuration,
the liner  creates a lip around the circumference of the sump
hole just below the top of the slab. This lip provides an ideal
support for a flat circular cover of the type described above, as
illustrated in Figure 33A. Where such a lip does not exist,
some investigators have suggested that  one  be created by
attaching strips of wood to the inside of the liner just below
the top of the sump (Br92).

Alternatively, when there is no lip, the cover can be designed
to be recessed while still being supported by the top of the
slab. A design to provide a recessed cover in such cases is
illustrated in Figure 33B. A few pre-fabricated molded plastic
covers have been marketed in the general configuration shown
in Figure 33B. Alternatively, such a cover might be fabricated
by the mitigator out of sheet metal.

If the cover cannot conveniently be recessed as shown in
Figure 33A or 33B, a flat cover on top of the slab (as in Figure
3, but with a trap) should often work satisfactorily, if the
thickness of the cover material is not great enough to create a
dam.
                                                       192

-------
                                          Flexible Coupling,
                                          Joining Main Pipe
                                          and Pipe Stub
           Main Suction Pipe
—Wr-
                                       r
             Waterless Trap — .

              Recessed
              Flat Cover
              Screw
    Gasket or
    Silicone Caulk
     Water Overflow
     from Trap
    Submersible
    Pump
Pipe Stub,
Sealed into
Cover
                                                      _A
             Grommets or
             Urethane Caulk
                            /
                                                      Check
Water
Discharge
Line

pjpe
Union
                                 Note: Flexible coupling in suction
                                 pipe and union in water discharge
                                 pipe are to facilitate subsequent
                                 removal and replacement of sump
                                 cover for sump/pump maintenance.
                                                    Sump
                                                    Liner
                A) Sump liner forms a lip just below slab surface;
                   lip used to support a flat circular cover.
   Recessed Sheet
   Metal Cover
  Gasket or
  Silicone
  Caulk
    Bolt
                                      Molded
                                      Plastic Cover

                          Clear Acrylic
                          Viewing Window,
                          Cemented or Screwed
                          into Cover

                             Gasket or
                             Silicone
                             Caulk
                                Bolt
           B) Recessed cover designed to
              be supported by top of slab.
                                    C) Raised cover typical of some
                                       commercially available covers.
Figure 33. Some alternative sump cover designs (illustrated for the case where suction is being drawn at the sump).


                                                    193

-------
Attachment of the sump cover. The cover must be at-
tached to the slab, or to the lip of the sump liner, in a manner
which will facilitate its subsequent removal and replacement
for sump inspection and sump pump maintenance.

To facilitate cover removal, it is recommended that the cover
be bolted tightly to the slab or Up, with a foam, rubber, or
caulk gasket between the cover and the concrete or liner to
help achieve air-tightness. When caulk is used as the gasket
instead of foam or rubber, silicone caulk should be selected
instead of urethane caulk for this particular application. Sili-
cone caulk bonds  less well to  the concrete.  This  has the
disadvantage of giving a less durable seal, but permits reason-
ably convenient removal of the sump cover.

Urethane bonds so well that the cover would  be extremely
difficult to remove. Thus, it is not recommended that urethane
caulk be used to bond the cover to the slab or liner lip.

Where the cover rests on the lip of the sump liner, as in Figure
33A, appropriate screws or bolts should be used to attach the
cover to  the liner lip, with a  suitable gasket in between to
provide an air-tight seal. When silicone caulk is used as the
gasket, a continuous, generous bead of caulk should be placed
around the entire perimeter of the Up.  With such recessed
covers, provisions  must be made so that the cover can be
grasped for removal during subsequent maintenance.

When the cover rests on top of the slab, as in Figures 3,33B,
and 33C, masonry bolts or anchors would be used to secure
the cover to the slab, with a suitable gasket between the cover
and the slab.

Need for submersible sump pump.  The pump in capped
sumps should be a submersible pump. The motors of pedestal
pumps, not designed for submersible operation,  are subject to
rusting if enclosed within a capped sump. Thus, the existing
sump pump is not a submersible pump, it should be replaced
with a submersible pump as part of the installation.

Installation of cover around pump connections.  The
water discharge line from the sump pump and the electrical
connection to the pump must be disconnected, and fit through
the corresponding holes in the cover when the cover is in-
Stalled. The remaining gaps between the water line or the
electrical cord and the cover must be sealed. This seal may be
accomplished with gun-grade  caulk. Alternatively, one ven-
dor markets a line of rubber grommets which will fit tightly
around the water line or the electrical cord, seaUng the hole
through the cover.

A check valve must be installed in the water discharge line
from the sump pump. When the pump is not operating and this
line is thus empty, the discharge line would constitute a major
source of air leakage into the sump if the check valve were not
installed, reducing the suction that the sump/DTD system
would be able to maintain in  the sump. Without the valve,
outdoor air would flow from the water discharge point, down
through the pipe and the sump pump and into the sump, under
the depiessurization created by the system.
Check valves can be designed to be threaded onto metal water
discharge lines, cemented onto PVC Unes, or attached to
either metal or PVC lines using flexible coupUngs which are
small versions of the coupUngs used to attach mitigation fans
to the system piping. Many check  valves can be mounted
either horizontally (as in Figures 3 and 33) or vertically;
others can be mounted only verticaUy. Where the check valve
is connected by small flexible coupUngs, it can also be used as
the point at which the water discharge line can be discon-
nected when the cover must be removed for maintenance. See
Design of the system to facilitate future maintenance below.

When the check valve is cemented onto the water discharge
line, it cannot be used  to disconnect the Une  during sump
maintenance. In these cases, a piping union should also be
added to the discharge line near the sump (K192). See Figure
33. The discharge Une is cut at that point, and rejoined by the
union nut. When this union is loosened, the water discharge
line can be disconnected.

Connecting and supporting the sump  suction pipe.
When suction is to be drawn directly on the sump, the 4-in.
PVC, PE, or ABS  suction pipe is installed vertically down
through the sump cover, as shown in Figures 3 and 33.

The gap between the outer circumference of the pipe and the
hole through the cover must be effectively sealed.  Again,
caulk can be used to seal this gap, or, alternatively, a rubber
grommet capable of fitting around the 4-in.  pipe can be
purchased. One advantage claimed for the grommet is that it
wiU flex with the cover if someone steps on the cover, and will
not break like a caulk seal might under those circumstances
(K192).  Another claimed advantage  is that the  grommet en-
sures that the pipe does not contact the cover  at any point,
preventing pipe vibrations from being transmitted to the cover,
potentially reducing noise.

In many cases, the weight of the piping will not be adequately
supported at the sump cover. Thus, the weight of this pipe
must be supported overhead, e.g., by hangars or strapping
connected to the overhead floor joists, as discussed for SSD
risers in Section 4.5.1 (and as illustrated in Figures 16 and 17).
Since the sump wiU almost always be near a perimeter foun-
dation wall, another option would be to offset the riser against
the foundation wall, and to attach it to the waU, as discussed in
Section 5.5.1 (see Interior tile loops - supporting the remote
suction pipe).

To faciUtate subsequent removal of the sump cover for pump
maintenance, the PVC suction pipe should be cut near the
cover penetration, and rejoined using a flexible PVC coupling
with hose clamps. Thus,  only a  stub of 4-in. PVC piping
would  in  fact be sealed into the cover.  This situation is
illustrated in Figure 33. When it was desired to remove the
cover, the hose clamps would be loosened and the coupling
moved, so that the pipe stub could be removed with the cover.
The alternative to this approach would be to break the seal
where the pipe penetrated  the cover each time the cover
needed to be removed, so that the cover could be removed
while the entire suction pipe remained in place.  One possible
concern regarding this approach is that the personnel servic-
                                                       194

-------
ing the sump pump might not get the pipe penetration through
the cover properly reseated when the cover was replaced.

Design of the, system to facilitate future maintenance.
It is re-emphasized that the effectiveness of the sump/DTD
system will depend heavily on achieving and maintaining an
effective seal of the sump with the cover. The caulk seals and
the rubber grommets (if used) must remain intact, and any
water trap (or waterless trap) through the cover must remain
air-tight (i.e., water must remain in the trap, or the weighted
ball or ring in the waterless trap must continue to seat prop-
erly). The occupant must be clearly informed (and the label-
ling on the system must clearly indicate) that these seals must
be re-established if the cover is ever removed for subsequent
maintenance of the sump pump.

The issue of installing the system in a manner to facilitate
subsequent removal of the sump cover has been raised several
times during the preceding discussion. In summary, with  a
properly installed system having the suction pipe through the
cover, the removal might proceed as follows: 1)  the sump
cover would be unbolted around its perimeter (and, if mere
were a silicone caulk seal, that seal would be broken); 2) the
flexible coupling connecting the suction pipe stub through the
cover would be loosened and slid up or down (or, if such  a
break in the suction pipe is not included, the seal at the pipe
penetration through the cover would be broken); 3) the water
discharge line would be disconnected, either at the  check
valve (if the valve is connected using flexible couplings) or at
the point where the pipe union is installed; and 4) the cover
and the sump pump would then be removed from the sump as
a unit, with the 4-in. PVC pipe stub, the water discharge line,
and the pump electrical cord remaining sealed at their penetra-
tions through the cover. If the pipe, water line, and electrical
penetrations had been sealed using  rubber grommets, the
breakage of the seal where the suction pipe penetrates the
cover (the alternative in step 2 above) would consist of simply
loosening the grommet.

5.6    Design/Installation of the
Piping Network and Fan
The design and installation of the piping network and fan for a
sump/DTD system would be exactly as described in Section
4.6 for a SSD system.

5.7    Slab Sealing in  Conjunction
with Sump/DTD Systems
The same types of slab sealing, discussed in Section 4.7 in
connection with SSD systems, would also be important for
sump/DTD systems.

Perimeter channel drains.  Sealing of perimeter channel
drains, as discussed in Section 4.7.1, would be of particular
importance for sump/DTD systems, in cases where the drain
tiles form an interior loop around the basement perimeter.
Because the suction  would be being directed immediately
beside the foundation wall by the tiles, such  systems would
exacerbate the leakage of house air down through unsealed
perimeter channel drains into the system.
Wall/floor joint. For the same reason, sealing of the wall/
floor joint where accessible (in cases where the gap is wider
than, e.g., 1/32 to 1/16 in.) may be more important for sump/
DTD systems than for SSD systems. See Section 4.7.5.

Sumps. As emphasized previously in Section 5, the sealing
of sumps  is  an automatic and mandatory part of installing
sump/DTD systems. The specific procedures for sealing sumps
have been described in Section 5.5.2.

Floor drains. In some cases, floor drains may drain directly
into the sump or into the drain tile loop. Where the floor drain
drains directly into the sump, this will usually be apparent by
virtue of a pipe from the drain penetrating the sump wall.
Where the drain connects to the drain tile loop remote from
the sump, this will be less apparent. Where floor drains drain
into the sump/ drain tile system, it is particularly important
that the floor drain be trapped. Otherwise, a substantial amount
of basement air will be drawn into the system through the
open drain, dramatically reducing  system performance and
increasing the house heating/cooling penalty.

Where the floor drain does not contain a trap, or where the
trap does  not remain full of water, a waterless trap  can be
installed, as  discussed in Section 4.7.3. One alternative ap-
proach that some mitigators have  used in cases where the
drain empties directly into the sump, has  been to create a
"trap" inside the sump. By this approach, an elbow is fitted
onto the floor  drain pipe entering the sump, directing it
downwards  toward the bottom of the sump. There,  it is
immersed in a weighted plastic container full of water. One
problem  with this approach is that, if the pump  does not
regularly  fill with water to a level above the top  of the
container, the container could dry  out over time, rendering
this in-sump trap ineffective.

Other drains. On occasion, other drains may also empty into
the sump  or into the  drain tile loop. Any such other drains
must be identified, and addressed in some appropriate man-
ner.

Where the drain tile loop is outside the footings, basement
window well drains (and sometimes even rain gutter down-
spouts) have occasionally been observed to empty into the
drain tile loop. Large amounts of outdoor air would be drawn
into the system if these features were not addressed in some
manner. In the case of window well drains, they might be
trapped in a manner  similar to floor drains, or sealed  alto-
gether if they never receive any significant amounts of water.
(One concern with the use of waterless traps in such exterior
drains is that there is  an increased risk that debris may enter
the trap, preventing the weighted ball or ring from  seating
properly.)

On those infrequent occasions where rain gutters are found to
empty into the tiles, one option would be to re-route the runoff
from the gutters so that the gutter penetration into the tiles can
be sealed off. However, rather than run the risk of modifying
the house drainage patterns, the mitigator would probably be
best served in such cases by abandoning sump/DTD as the
mitigation approach,  capping the sump, and using SSD in-
stead.
                                                      195

-------
5.8   Gauges/Alarms and Labelling

The considerations discussed in Section 4.8, concerning gauges/
alarms and labelling of SSD systems, also apply to sump/
DTD systems.

The system labelling should include clear instructions that the
sump cover must be carefully resealed if it is ever removed for
pump maintenance or for any other reason.
                                                  196

-------
                                            Section 6
                        Design and Installation of Active
                    Drain-Tile  Depressurization Systems
                      (Above-Grade/Dry-Well  Discharge)
Sometimes, rather than draining into a sump inside a base-
ment, drain tiles will direct collected water to an above-grade
discharge at a location on the lot remote from the house. Use
of above-grade discharge requires that the lot be sufficiently
sloped such that an suitable low point is available. Above-grade
discharge is sometimes referred to as draining to daylight, or
draining to an outfall. Or, rather than draining above grade,
the water may be directed to an underground dry well away
from the house, if the underlying soil is sufficiently porous.

A variation of above-grade discharge is sometimes encoun-
tered in farming communities. In this variation, the drain tiles
around the house empty into an extensive existing network of
field tiles that had originally been installed to drain the fields.
The outfall from the field tiles might be a great distance from
the house.

Tiles with above-grade or dry-well discharge are most com-
mon in basement houses. However, they will sometimes also
be encountered around the footings of other substructure types
on sloped lots.

Most often, the drain tiles draining to an above-grade dis-
charge or dry well will form a partial or complete exterior
loop, on the outside of the footings. Where the tiles form a
complete loop, a length of tile (referred to here as the  "dis-
charge line") taps into this loop on the downhill side of the
house, and extends away from the house (usually with a
downward slope) to the point where it emerges above grade or
ends in a dry well. The water collected in the loop around the
foundation is discharged at that point One common configu-
ration for tiles with above-grade  discharges, often encoun-
tered in houses  having walk-out basements,  has die tiles
forming a U around the three sides of the house that are below
grade (the uphill side of the house, and the two ends); each leg
of this U (one on each end of the house) then comes above
grade on the downhill side of the house.

Less commonly, these tiles may be interior to the footings,
beneath the slab.  In this case, the discharge line must pass
through or beneath the footings, to extend to the discharge
point remote from the house.

Where tiles having an above-grade or dry-well discharge exist
in basement houses,  suction on these tiles can be a viable
mitigation approach. Figure 4 illustrates a typical DTD/re-
mote discharge system in a basement house.
The tiles will aid in distribution of the suction around the
basement footings, near the major soil gas entry routes (the
wall/floor joint and the block foundation wall). However,
where these tiles are exterior tiles  (which is often the case
when there is a remote discharge), it can be important that the
tiles form a complete or nearly complete loop (i.e., on at least
three sides of the house). With exterior tiles, if there are major
entry routes (such as a block fireplace structure penetrating
the slab) toward the slab interior, it can also be important that
sub-slab communication be reasonably good. Where the exte-
rior loop is only partial, or where there are major interior entry
routes, there is an increased chance that a DTD/remote dis-
charge mitigation system may need to be supplemented by
SSD pipes. See the discussion in the introduction to Section 5,
regarding the cases under which sump/DTD systems with
exterior loops may need to be supplemented by SSD.

While exterior drain tiles with remote discharges sometimes
also are present in slab-on-grade and crawl-space houses, the
data are limited regarding whether DTD/remote  discharge
would be an effective mitigation approach in those substruc-
ture types. In slab-on-grade houses, the tiles may be near the
surface (depending upon the depth of the footings), so that
large amounts of outdoor air could  sometimes be drawn into
the system, potentially reducing system effectiveness.  In
crawl-space houses, the suction on the tiles might not be
expected to significantly reduce the amount of soil gas enter-
ing the crawl space. For these reasons, the following discus-
sion focuses on basement houses.

One mitigator who has tested DTD/remote discharge in these
other substructure types (K192) reports that the technique has
worked well in slab-on-grade houses in Colorado, because the
footings are at a sufficient depth (30 in.) to reduce entrainment
of ambient air into the system. In crawl-space houses, DTD/
remote discharge with a membrane over the crawl-space floor
is consistently less effective than traditional individual-pipe
SMD.

Many of the details involved in designing and installing DTD/
remote discharge systems are similar to those for SSD and
sump/DTD systems, discussed in Sections 4 and 5. The dis-
                                                    197

-------
cussion in this section will focus on those design/installation
features unique to DTD/ remote discharge.

The discussion in this section draws heavily from the detailed
review of available data on  active DTD/remote discharge
systems, presented in Section 2.3.3.

6.1    Selection of the  Number of
Suction Pipes: Need for a SSD
Component to  the DTD/Remote
Discharge System

If the drain tiles form a complete (or at least three-quarters
complete) exterior loop  and if the sub-slab communication is
reasonably good, one suction pipe, drawing suction on the
drain tiles at a convenient location, will  normally be suffi-
cient. Even if sub-slab communication is only marginal, one
pipe depressurizing a complete exterior loop will probably be
sufficient, if there is no  major soil gas entry route toward the
interior of the slab, remote from the perimeter.

If the tiles form an interior loop, one suction pipe tapping into
the tiles will usually be sufficient if: a) the loop is a complete
loop, whether communication is good or marginal; or b) the
loop is only partial, if communication is good. A second
suction pipe tapping into the drain tiles would be needed only
if one segment of tiles was isolated from the remainder of the
loop by silting or physical damage, in which case a second
pipe would be needed  to treat the isolated segment. Such
isolation of a particular segment could be difficult to ascertain
in practice.

Under the conditions in the preceding two paragraphs, addi-
tional SSD pipes will often not be necessary to supplement the
DTD/remote discharge  system,  unless the tiles have been
silted or damaged.

If the tiles are outside the footings and form only a partial
exterior loop (i.e.,  are beside fewer than three  sides of the
house), or if sub-slab communication is marginal or poor and
there is a significant entry route  toward the slab interior, then
the chances are increased that an exterior DTD/remote  dis-
charge system would have to be supplemented by SSD pipes.
Likewise, if the tiles are inside the footings, SSD pipes may be
necessary if the interior loop is only partial and if, in addition,
the communication is marginal or poor.

However, one mitigator has  reported  sometimes achieving
good performance with partial exterior loops without crushed
rock beneath the slab (K192). In these cases, the slab is resting
directly on graded native soil (a decomposed granite). (Crushed
rock is present in the trench around the footings containing the
drain tiles.) This good performance may be due to: relatively
good permeability in the native soil underlying the foundation
in some cases, enabling the suction field to extend beneath the
footings into the sub-slab region, and around the perimeter to
locations  where tiles are absent; and the tendency of the  tiles
to intersect below-grade trenches via which utility lines pen-
etrate the foundation, providing a relatively high-permeability
route for the suction to penetrate the foundation wall.
The extent of the drain tile loop, and whether it is exterior or
interior, can be difficult to ascertain unless the homeowner
observed the tiles being installed during construction of the
house, unless there are reliable as-built drawings, or unless a
confirmation can be obtained directly from the builder. Exca-
vation at several locations around the  foundation during the
pre-mitigation  visual  survey can reveal whether the loop is
outside the footings, and whether it appears to be complete.
However, if the footings are very deep, such excavation could
be a time-consuming process. The presence of a discharge line
emerging above grade on each end of the house would tend to
confirm that the tiles form a U around three sides of the house.
Adjoining wings (e.g., a slab on grade adjoining a basement)
increase the likelihood that the tile loop is discontinuous, and
that the adjoining wing may require treatment with  supple-
mental SSD pipes. However, in  some cases, an exterior drain
tile loop will extend  around adjoining slab-on-grade living
areas as well as the basement.

If the sub-slab  communication is not apparent from observa-
tions during the  visual inspection or  from experience with
house construction characteristics in the  region, qualitative
suction  field extension testing  can be conducted with  the
diagnostic vacuum cleaner. Together with information on the
nature of the tile loop, the pre-mitigation suction field data
would suggest the likelihood that SSD pipes will be necessary
as a supplement to the DTD/remote discharge  system.  But
unless the location and extent of the tile loop can be estimated
reasonably  well, the  vacuum cleaner diagnostics may  not
suggest  specifically the number of supplemental SSD pipes
that will be needed, or where they should be placed.

In summary, rather than excavating around the house to assess
the extent  of  the drain  tile loop, and  rather than doing
pre-mitigation  suction field extension  measurements, it may
often be more  cost-effective to  simply proceed to install the
DTD/remote discharge system.  The appropriate number and
location of supplemental SSD  pipes (if any turn out to be
needed) can then be determined from sub-slab suction field
measurements and indoor radon measurements after the DTD
system is operating. One mitigator with substantial experience
installing DTD/remote discharge systems reports that, at least
under the conditions he experiences, supplemental SSD pipes
are generally not needed (K192).

If the likelihood appears high that supplemental SSD pipes are
going to be needed, then the mitigator might opt to abandon
the option of installing a DTD/remote  discharge system, and
to instead focus  directly on a  SSD system. This could be
particularly advisable in cases where the house has  a base-
ment that is not heavily finished, so that installation of a SSD
system would not be especially complicated.

If any SSD pipes turn out to be needed,  supplementing the
DTD/remote discharge system, the SSD  component of the
system would be designed and installed as described in Sec-
tion 4. The SSD pipes might  be manifolded to the same
exterior fan and exhaust stack that is serving the DTD  system.
(Since the drain tiles  will most often be exterior tiles when
there is a remote discharge,  the fan and exhaust for a DTD/
remote discharge system will almost  always be outdoors.)
Alternatively, it may sometimes be more convenient to install
                                                      198

-------
a separate fan and exhaust stack for the SSD pipes, so that the
SSD system  will be  completely  independent of the  DTD
system.

6.2    Selection of DTD Suction Pipe
Location
Since the suction will  extend readily through the entire drain
tile network, the suction pipe may be connected to the drain
tiles at any convenient location. With the exterior tiles most
commonly encountered when there is remote discharge, the
fan will almost always be mounted on this vertical suction
pipe at the point where it  extends above grade level, as
illustrated in Figure 4. Many of the criteria used in selecting
the suction pipe location will be based upon the advantages
and disadvantages of having the fan at that location.

The suction pipe is  most commonly installed in the loop of
tiles immediately beside the footings, for several reasons.
However, it could  also be  installed in  the  discharge line
leading away from the house to the water discharge location.

Figure 4  depicts the  suction  pipe tapping  into the single
discharge line near the house.  This location is shown in the
figure only for ease of illustration; it is not a recommendation
that the suction pipe  should necessarily be located at that
point.

Location of suction pipe in loop beside the house. There
are several reasons  why it is common  to install the suction
pipe into the tiles immediately beside the footings, rather than
in a discharge line more remote from the house. First, location
of the vertical suction pipe immediately beside the house
greatly facilitates mounting  a  stack against  the side of the
house, exhausting above the  eave, consistent with EPA's
interim standards (EPA91b).

Second, if the suction pipe is mounted in the discharge line
remote from the house, there can be a significant suction loss
between the suction point and the house at the flows encoun-
tered  (usually 50 to  150 cfm  with a  90-watt fan); this is
especially true if the  drain tiles  are the flexible corrugated
type, which offer about 1.8 times the wall friction of smooth
pipe (K192). Since DTD with exterior loops can be dependent
upon maintaining very good suction in the exterior tiles, if this
suction is to extend beneath the foundation into the sub-slab,
this suction loss has been observed to be significant.

Other factors contributing to suction loss between remote fans
and the house are: damage to the discharge line while it was
exposed during construction; blockage of the line by,  e.g.,
rodent nests; and leakage of outdoor air into the line, since the
discharge  line will be close to grade at locations remote from
the house. These factors have been found to have a significant
effect on  system performance  when the  suction pipe is in-
stalled in the discharge line remote from the house (K192).

A secondary consideration in remote location of the fan is that
the electrical connection will have to extend  some distance
back to the house.  However,  with the use of underground
conduit, this is not a major problem.
When the suction pipe is installed in the loop  immediately
beside the house (or in the discharge line near the house), the
pipe should be installed at a location where the fan and the
exhaust stack (which will generally rise straight upward above
the fan) can  aesthetically be installed. The criteria used to
select this location were discussed  in  Sections 4.6.2 (see
Selection of where the piping exits the basement) and 4.6.5
(see Selecting the location of exterior or garage  stacks). This
location would be at the rear of the house, and at a point where
the stack will  not be rising in front of windows. It may also be
helpful to locate the connection (and the fan) away from the
bedroom wing, if possible, to reduce the risk of the occupants
being disturbed by fan or exhaust noise.  If the  DTD/remote
discharge  system is going  to be supplemented by SSD pipes,
the DTD suction pipe location may be selected to simplify
manifolding together with the SSD piping.

Other factors  can also be considered in selecting the location
in the loop around the house. First, a central location would
generally  be  preferred. For example, if  the tiles form a U
around three sides of the house, it might be preferred to locate
the suction pipe toward the bottom of the U, rather than at one
end, in order  to achieve even distribution of the suction field.

Second, the mitigator may not wish to  choose the  location
where the footings and tiles are the deepest below  grade, in
order to avoid excessive excavation to expose the tiles at that
point. However, neither should a particularly shallow location
be selected, since that could result in greater short-circuiting
of outdoor air into the system through the  soil. It is recom-
mended that the suction pipe be installed at a point where the
tiles are at least 3 ft below grade.

Location of suction pipe in discharge line remote from
the house. Figure  4 is  attempting to depict the case where
there is a complete tile loop around the foundation, and where
the suction pipe taps into the single discharge  line leading
from this  loop to the outfall. In other cases, the tiles could
form a U around three sides of a walk-out basement, with each
leg of the U  coming  above grade (on opposite ends of the
house). Where the tiles form a  U, the suction pipe could be
located in either one of the two  discharge lines, if it  were
desired to tap into a discharge line remote from  the footings.

Several potential advantages may be apparent if the suction
point were located in the discharge  line remote  from the
house. However, these apparent advantages may often not be
sufficient  to justify  remote location. One possible advantage
of remote location  is that locating  the  suction pipe in the
discharge line near the outfall will generally reduce the amount
of excavation that will be required in basement houses. The
discharge  line near its outfall will be at a shallower depth than
will the loop  beside the foundation,  which will be at footing
depth. There  will  have  to be some excavation around the
discharge  line in any event, even  if the suction pipe is in the
loop beside the foundation, in order to install a check valve in
the discharge line (see Figure 4 and Sections 6.5  and 6.7). For
efficiency, it  could  be reasonable to install the  suction pipe
and the check valve into the drain tiles at the same excavation.
This savings  in excavation effort can be offset by  reduced
system performance, as discussed above.
                                                        199

-------
A second advantage of remote location of the suction pipe is
that the fan would be away from the house, reducing the fan
and exhaust noise in the living space. A third possible advan-
tage, if an exhaust stack were not required, would be that the
suction pipe could be located such that the fan is mounted
behind shrubbery a couple feet above grade away from the
house, which could provide a better appearance than would be
the case if the fan were immediately beside the house. Such
remote location of the exhaust could also sometimes reduce
the amount of exhaust re-entrainment back into the house.
However, EPA's interim mitigation standards (EPA91b) cur-
rently require that a 10-ft tall stack be installed above the fan
in such remote installations, to avoid exposure of persons in
the yard; thus, it would not be possible to realize any aesthetic
benefits from hiding the fan behind shrubbery remote from
the house.

If the discharge line consists of perforated piping, the suction
pipe should be installed at a point where the line is at least 3 ft
below grade, to reduce outdoor air short-circuiting into the
system through the soil.

Considerations when the tile loop is inside the footings.
On occasion, the drain tile loop may be inside the footings,
rather than exterior. In such cases, connecting the suction pipe
to the tile  loop  (rather than to the  discharge line) would
require that a hole be cored through the basement slab in order
to access the tile. Under these conditions, the system would
take on the appearance of the left-hand side of Figure 3, where
a sump/DTD suction pipe is tapping into  interior tiles at a
point remote from the sump.

Tapping into the interior loop beneath the slab would make it
possible for the system piping to be indoors. This situation
could reduce the aesthetic impact outdoors, and may be pre-
ferred in specific cases if the basement is partially unfinished
and if there is a route for the exhaust stack either up inside the
house or through an adjoining garage. But if the fan and stack
will have to be outside the house in any event, due to interior
finish and/or homeowner preferences, it will be easiest to
simply tap into the discharge line outside the house. Installing
the suction pipe in the discharge line in these cases will avoid
any aesthetic impact indoors, one of the general advantages of
DTD/remote discharge systems.

6.3    Selection of Suction  Pipe Type
and Diameter
The considerations in selecting the type of suction pipe, and
the piping diameter, are the same for DTD/remote discharge
systems as those discussed in Section  5.3 for  sump/DTD
systems.

The flows from DTD/remote discharge systems appear to be
commonly about 50-150 cfm with the 90-watt fans, the same
range observed from sump/DTD systems. Accordingly, as
discussed in Section 5.3,4 in. diameter piping will normally
be preferred to avoid undue friction loss and flow noise. Four
in. piping is also of a convenient size for connecting to the
existing drain tiles, which are usually 3 or 4 in. diameter.
The length of piping associated with DTD/remote discharge
systems may be relatively short when the suction pipe is
immediately beside the house. The piping system illustrated
in Figure 4 would have an equivalent length (including the
equivalent length of the fittings for friction loss purposes) of
only about 35 ft, even for a two-story house. At the upper end
of the observed flow range (100 to 150 cfm), the total friction
loss that would be expected in  this length of 4-in.  piping
would be about 0.2 to 0.5 in. WG, based on Figure 13. This
suction loss can generally be handled by the 90-watt tubular
fans. Hence, use of 6-in.  piping to reduce the suction loss is
probably not warranted in most cases, although it may occa-
sionally be desired to reduce flow noise.

At the middle to upper end  of the flow range, the suction
losses (0.8 to  2 in. WG) and flow noise could nominally
become unacceptable if 3-in. piping were used for the 35-ft
system. As discussed in Section 5.3 in connection with sump/
DTD systems,  if 3-in. piping  were in fact used, flows would
drop,  and the fan would operate at a different point on its
performance curve; hence,  suction losses and flow noise
would in fact be less. Suction in the drain tiles would also be
reduced. In the case of sump/DTD systems, especially when
there is a complete interior loop with good aggregate, this
reduction in tile suction that would result from the use of 3-in.
pipe may sometimes be tolerated without a dramatic reduction
in performance. But with DTD/remote discharge systems,
where the tiles are commonly exterior to the footings and
where communication to interior entry routes is less well
assured, it is less clear  that the reduced suction could be
tolerated. Accordingly, with DTD/remote discharge systems,
it would seem generally more desirable not to reduce the pipe
diameter below 4 in.

6.4    Selection of the  Suction  Fan

Since the flows  from DTD/remote discharge systems are
usually similar to those from sump/DTD systems, the consid-
erations in selecting  a fan are  generally similar to those
discussed for the sump system in Section 5.4.

At the flows typical in these systems with 90-watt  in-line
tubular fans (50-150 cfm), these fans can commonly establish
suctions of about 0.75-1.0 in. WG in the system piping, based
upon the fan performance curves. These suctions should usu-
ally be sufficient to handle the friction losses in 4-in. diameter
piping, and to  achieve good system performance. Thus, the
90-watt fans would generally appear to be a good choice for
DTD/remote discharge systems.

With sump/DTD systems, the smaller, 50-watt tubular fans
discussed in Section 4.4.1 will sometimes give adequate per-
formance under favorable conditions (with the tiles forming a
complete loop inside the footings, and with good sub-slab
communication). These smaller fans are not recommended for
DTD/remote discharge systems. Remote discharge systems
commonly involve exterior drain tiles. Exterior tiles may be
expected to face a greater challenge in extending a suction
field under the footings and through block foundation walls to
treat interior slab entry routes. Thus, it is considered advisable
to routinely use the larger fan in DTD/remote discharge
installations.
                                                      200

-------
The in-line radial blowers discussed in Section 4.4.2 could
also be considered for use on DTD/remote discharge systems.
For systems having flows around 100 cfm and higher, the
additional expense of a radial blower would not be warranted,
since  the blower would not be maintaining suctions higher
that those which would be produced by a 90-watt tubular fan.
However, for lower-flow systems, the radial blowers could
establish a significantly higher suction (2 to 3 in. WG at 50
cfm, compared to about 1.25 in. WG with a 90-watt tubular
fan at that flow). The increased suction may sometimes be
useful in achieving better performance with exterior drain
tiles.

As with sump/DTD systems, the relatively high flows charac-
teristic  of DTD/remote discharge systems mean  that the
high-suction, low-flow fans, discussed in Section 4.4.3 for
SSD installations in poor-communication houses, would never
be applicable in this case.

6.5    Installation  of a Suction Pipe
into the Drain Tiles
In most cases with DTD/remote discharge systems, the suc-
tion pipe will be installed into an exterior drain tile loop or
into a discharge line outside the house. The discussion here
thus focuses on that case.

For those limited cases where there is an interior loop and the
mitigator elects to tap into this loop through the basement
slab, inside the house, the reader is referred to the discussion
in Section 5.5.1, for the case where a sump/DTD suction pipe
is installed into an interior loop remote from the sump. See
Interior tile loops - connecting the remote suction pipe (indi-
rect approach) and Interior tile loops - connecting the remote
suction pipe (direct approach). Installation of the suction pipe
inside the basement is not discussed any further here.

Finding the remote outfalls). Where the tiles discharge
above grade remote from the house, the first step will gener-
ally be to locate the outfall(s) for the discharge line(s). The
location of the outfall(s) will need to be known so that a check
valve can be installed, as discussed later, regardless of where
the suction pipe is to be installed.

Locating the outfaU(s) will  usually be done during the
pre-mitigation visual survey. This is not always easy, since
outfalls will sometimes be  covered by dirt and leaves. The
mitigator should be alert to  the possible presence of two
outfalls, as would be the case when the tiles form a U around
three  sides of the house and discharge at each end.

If the  drain tiles empty into an extensive network of field tiles,
finding  the distant outfall from the field tiles might be ex-
tremely difficult, and is probably not necessary.

Finding the exterior drain tiles or discharge lines. To
install the pipe, an excavation is made to expose the drain tile
at the selected location.

If the pipe  is to be installed in the exterior loop around the
foundation, the excavation would be immediately beside the
house. In this case, finding the tile should be relatively easy; it
is known that the tiles will be right beside the footing. The
only question would be whether tiles are present beside the
footing at the particular location where it is desired to install
the suction pipe, if the loop is not complete.  Where the
basement has an adjoining slab-on-grade or crawl-space liv-
ing area, tiles will not always be present beside the adjoining
wing; where there is an adjoining slab-on-grade garage, tiles
will generally not be present by the garage.

If, instead, the pipe is to be installed in the discharge line,
some exploratory excavation may be needed to find the dis-
charge line at the desired distance from the house. Visually
tracing the path, of the discharge line between the outfall and
the house will suggest the general path of the discharge line,
but may not reveal the precise location of the line.

Maintaining integrity of gravel trench and  cover. The
drain tile loop by the footings will commonly be laid in a
trench filled with gravel, as illustrated in Figure 4. Sometimes
this trench may be covered by some material  to  prevent silt
from working its way into the gravel. If present, this material
may be geotechnical  cloth in  some cases, although  such
materials as roofing paper, straw, and newspaper have some-
times been encountered.

During the excavation to expose the tiles, care should be taken
so  that the gravel  trench and  any cover material  can  be
properly restored around the tiles after the pipe installation is
completed. For example, sections of the cover material should
not be damaged or disposed of.

Pipe installation  using rigid T-fitting. One option for
installing the  suction pipe into the drain tiles would be to use
a rigid PVC, PE, or ABS T-fitting. This is the approach which
is illustrated in Figure 4, for the case where the  discharge line
consists of flexible black corrugated high-density polyethyl-
ene or polypropylene, which is common in many areas.

Where a rigid T-fitting is to be used, the drain tile is severed at
the point where it is exposed, in order  to accommodate the
T-fitting. This fitting (usually 4 in.  diameter, as discussed in
Section 6.3) is installed with the T inverted, with the leg of the
T extending upward. The ends of the top of the  inverted T are
connected to either end of the severed drain tile.

When the drain tile is 4-in. diameter flexible corrugated tile as
in Figure 4, one option for making this connection would be to
force each severed end of the tile into one of the  openings in
the top of the T, as  illustrated in Figure 4. In the figure, the
tiles are shown as being attached to the T fitting using a bead
of urethane caulk between the  tile and the fitting,  and by
screws through the fitting into the tile.

Careful caulking of the tile/fitting seam may not always be
necessary. If  suction leaks into the gravel trench around the
footings, this will still be directing the suction  where it is
desired. However, it is advisable that some steps be taken to
ensure that the tile  and fitting do not subsequently become
disconnected. Disconnection might sometimes result in ex-
cessive ambient air leakage into the system, and could provide
an opening that could permit gravel or silt to partially block
the tiles.
                                                       201

-------
In some areas, the flexible corrugated tiles are 3 in. rather than
4 in. diameter. Three-in. tiles would fit loosely but satisfacto-
rily into the ends of the 4-in. T-fitting, for the type of installa-
tion illustrated in Figure 4.

In some cases, the drain tiles will be rigid perforated PVC or
ABS pipe, or baked clay. To install a rigid T-fitting onto rigid
PVC pipe, one option could nominally be to cement the PVC
T-fitting directly onto the severed ends of the PVC drain  tile
(or to cement an ABS T-fitting onto ABS  drain tile). How-
ever, it may be more practical to connect the ends of the T to
the rigid plastic or clay tiles using a flexible PVC coupling,
similar to those typically used to attach ASD fans  to  the
system piping. In this case, the coupling would likely be a
straight 4-in. x 4-in. coupling,  rather than  the 6-in. x 4-in.
couplings shown in Figure 4 and elsewhere for attaching fans.
The couplings could be clamped to the tile and to the T using
hose clamps. These flexible couplings are commonly used for
underground sewer pipe connections, and  are advertised as
being resistant to soil chemicals and roots.

However the rigid T is installed, a vertical section of PVC
piping  is then cemented into the leg of  the T, extending
upward to above grade level. Usually, the suction fan will be
mounted vertically, directly on this riser, a couple feet above
grade level, as shown in Figure 4. The T-fitting will rest on the
soil beneath the tiles, and will thus support the weight of the
overhead piping and fan.

Pipe installation directly into tiles without rigid T. It is
also possible to insert the suction pipe directly into the drain
tiles without severing them and installing a rigid T-fitting.

A variety of approaches might be envisioned for accomplish-
ing this. The following approach has been used often by one
mitigator who commonly encounters the flexible corrugated
drain tiles (K192).

A flexible corrugated T-fitting is used. The 4-in. PVC suction
pipe—long enough to extend above grade—is mounted on the
upward leg of the corrugated fitting. This is done by twisting
the corrugated leg of the fitting  up into the PVC pipe with a
liberal amount of urethane caulk; several sheet metal screws
are then inserted through the wall of the PVC pipe and into the
corrugated  material inside, to create a firm joint. Where  the
corrugated  drain tile is 4 in. diameter, it may be helpful to
cement a 4-in. PVC straight coupling onto  the bottom of the
PVC riser, the larger inside diameter of the coupling (4.5 in.)
may facilitate a comfortable fit of the corrugated tile up into
the PVC pipe.

About one-third of the horizontal bottom of the inverted T
fitting is cut off, creating a horizontal "saddle" which can be
snapped over the existing corrugated drain tile.

A liberal amount of urethane caulk is spread over the interior
of the corrugated T-fitting saddle, and this saddle is snapped
down over the drain tile with the PVC pipe extending upward.
Sheet metal screws are used to firmly connect the saddle to the
drain tile while the caulk cures.  A hole saw blade of suitable
diameter (3 to 4 in.), mounted on a shaft long enough to
extend down through the PVC riser, is then used from above
to cut a hole in the top of the drain tile directly beneath the
riser. The corrugated plug cut from the tile will come up with
the hole saw when the saw is withdrawn.

As an alternative to using the hole saw, one could consider
cutting a hole in the top of the drain tile before the corrugated
saddle is attached.

Unlike the case where a rigid  PVC T-fitting  was used to
connect the suction pipe, this flexible corrugated T saddle will
not as effectively support the weight of the piping and fan
overhead. Thus, in this case, it may be desirable to support the
fan using a bracket attached to the side of the house, or using
some other means.

Installation of a check valve at the outfall. Where the
discharge line discharges to an above-grade outfall on the lot,
a check valve (or reverse flow valve) must be installed some-
where in the discharge line. The check valve permits water in
the discharge line to  flow  out  into the outfall, but would
prevent outdoor  air from flowing up the line in the reverse
direction in response to the suction being developed by the
mitigation fan.

The  check valve is mandatory.  Failure to install this valve
would result in a substantial reduction in system radon reduc-
tion  performance. Suctions would drop dramatically, due to
the large amount of outdoor air flowing  into the system.
Where there are two outfalls, as where the tiles form a U
around three sides of the  house, a check valve must be
installed  in each  outfall.

Usually, the check valve is installed at the outfall, as in Figure
4. Where the suction pipe is also installed in the discharge
line, rather than in the foundation loop, the check valve must
be between the suction pipe and the outfall; it must not be
between the suction pipe and the house.

Where the discharge line empties into  an underground dry
well rather than an above-grade outfall, a check valve may not
be needed. If the dry well is near grade level and is configured
such that large amounts outdoor air might be drawn into the
system through the gravel bed, a check valve could be desk-
able.

Among the commercially available check valves that can be
used in this application are units which are ordinarily used to
prevent sewer line waste water  and gases from backing up
into  drains which empty into the sewer line. These valves are
available in 4-in. PVC fittings. The conventional sewer line
check valves contain a hinged, spring-loaded gate in the pipe
which is  normally held closed, but which would open toward
the outfall when the amount of water becomes sufficiently
great on the side of the gate toward the house. The gate cannot
swing open in the other direction, so that the suction devel-
oped by the fan could never cause the gate to open toward the
house, drawing in outdoor air.

The  check valve illustrated in Figure 4 is not a conventional
sewer line valve, but is a valve being marketed specifically for
used in DTD/remote discharge systems. It is not spring loaded.
A trap door mounted on a 45° angle is held closed by gravity
                                                        202

-------
when no water is present, but is sufficiently buoyant to lift
when there is water, allowing the water to flow to the outfall.
This valve requires less water than do conventional sewer
valves to cause it to open, ensuring efficient discharge of any
water in the tiles.

In the second edition of this  technical  guidance manual
(EPA88a), a water trap rather than a check valve is shown as
being installed in the discharge line to prevent air flow into the
system. The water trap is no longer the recommended ap-
proach for accomplishing this  objective. Check valves are
easier to install,  usually requiring less excavation. Check
valves have the additional advantage that they will not dry out
during extended periods of dry weather, as a water trap could
unless the homeowner was careful to add water. If a water trap
dried out, the performance of the system would deteriorate
significantly. And unlike a check valve, which could  be
installed at or near grade level, a water trap would have to be
installed at a sufficient depth such that the water in the trap
would not freeze during  extended periods of  cold weather;
this could necessitate additional excavation. Water traps could
also become partially blocked by  dirt and  debris over the
years.

Although the check valve could be installed anywhere in the
discharge line between the outfall and the suction location,
they will usually most conveniently  be installed at the outfall.
To install the check valve in this  case, the corrugated tile
would have to be exposed at the outfall. Then, possibly, a
short length of the existing tile would have to be removed to
make room for the valve fixture.

Check valves are commonly mounted in a segment of 4-in.
PVC piping. Thus, once  the tiles have been exposed at the
outfall, the pipe segment containing the valve would be  in-
stalled onto the drain tiles in the same manner described
above when a rigid T is employed to install the suction pipe
(see Pipe installation using rigid T-fitting).

When the drain tiles are flexible corrugated material, as in
Figure 4, the end of the corrugated drain tile is twisted into the
PVC pipe segment containing  the check valve. A liberal
amount of urethane caulk is used  at this joint, to hold the
check valve fixture in place and to prevent outdoor air leakage
through the seam (which will necessarily be at or near grade).
Sheet metal screws  through the PVC into the corrugated
material provide  further physical support. Where the drain
tiles are 4-in. corrugated material, a 4-in. PVC straight cou-
pling (4.5 in. i.d.) can be cemented onto the end of this check
valve pipe segment as in Figure 4, to facilitate insertion of the
4-in. drain tiles. Where 3-in. corrugated tiles are present, the
tiles would easily fit directly into the 4-in. piping without the
coupling.

If the discharge line is rigid PVC pipe, the check valve could
be cemented directly onto the end of the pipe. Or, if the line is
rigid PVC or ABS pipe, or if it is clay tile, the check valve
could be connected  using suitable flexible PVC couplings
with hose clamps.

Considerations  when the tiles empty into, field tiles. The
preceding discussion regarding the installation of a check
valve assumes the case where the drain tiles from the house
discharge at an outfall relatively near to the house. On those
occasions in farming communities when the tiles empty into a
network  of field tiles, the outfall from the  field tiles  can
sometimes be a mile or more away from the house.

In such cases, it is probably neither necessary nor advisable to
try to trap the entire field tile network at its outfall, even if the
outfall could be located. When the outfall is that distant,  it is
not clear how much ambient air will really be drawn from the
outfall. Moreover, it might not be advisable to risk restricting
the out-flow from the extensive network, especially when the
field tiles are draining fields owned by more than one land-
owner.

Experience with DTD/remote discharge systems in such cases
indicates that a large amount of air will commonly be drawn
into the DTD system from the field tile network (Wi92). If the
mitigator finds it necessary to install a check valve to achieve
acceptable performance in these cases, the check valve should
be installed in the discharge line near the house, between the
house and the point at which the house tiles connect into the
field tiles. It has been reported that installation of a check
valve may not always be necessary (Wi92).

Restoration  and backfilling. After the suction pipe has
been installed in the tile, the gravel bed around the tiles should
be restored, as should any cover which was present on top of
the gravel. All excavations must be filled in, to re-cover the
tiles.

6.6     Design/Installation of the
Piping Network and Fan
When the suction pipe is installed in an exterior drain tile loop
immediately beside the foundation outdoors, or when it is in
the discharge line near to  the foundation, the system piping
will be installed hi the general manner described toward the
end of Section 4.6.3 for the case of SSD suction pipes in-
stalled horizontally through the foundation wall from  out-
doors (see Pipe routing considerations with horizontal pen-
etration of suction pipes from outdoors). The fan and exhaust
stack (which would necessarily be an exterior stack) would be
installed as discussed in Section 4.6.5.

The fan is mounted vertically on the riser a couple feet above
grade level, as shown in Figure 4. Assuming that an exterior
stack is planned to direct the fan exhaust above the eave, this
stack would rise above the fan, and would be attached to the
side of the house.

If the suction pipe were installed in the discharge line at some
distance from the house, the issues involved in mounting the
fan and designing the exhaust would be similar to those
discussed toward the end of Section 4.6.5 (seeRemote exhaust
- a variation  of the exterior stack).  Of course, when the
suction pipe is extending up from a drain tile discharge line as
is the case here, the concerns expressed in Section 4.6.5 about
ensuring proper drainage for any condensate or rainwater are
automatically taken care of. As discussed in Section 4.6.5,
location  of the fan remote from the house  would make  it
                                                        203

-------
difficult or impractical to install an exhaust stack at least 10 ft
high, in accordance with EPA's current standards.

In those limited  cases where  the suction pipe tops into an
interior tile loop inside the basement, the design and installa-
tion of the piping network, and of the fan and exhaust system,
would be exactly  the same as described in Section 4.6 for SSD
systems, where the SSD suction pipes are inside the basement.

6.7    Slab  Sealing in Conjunction
with  DTD/Remote Discharge
Systems

The same types of slab sealing, discussed in Sections 4.7 and
5.7 in connection with SSD and sump/DTD systems, would
also be important for DTD/remote discharge systems. Since
the drain tiles will often be exterior in remote discharge cases,
it may be even more important for these openings to be sealed,
since the extension of the exterior suction field  under the
footings and through a block foundation wall into the sub-slab
region  will likely be weaker than in the SSD and sump/DTD
cases.

Since the drain tiles direct the suction immediately beside the
perimeter foundation  wall, it  becomes of increased impor-
tance that the following openings be sealed: perimeter channel
drains (as described in Section 4.7,1); wall/floor joints, where
accessible and when wider than about 1/32 to 1/16 in. (Section
4.7.5);  and perimeter expansion joints (Section 4.7.5).

Where floor drains drain directly into the drain tile network,
these floor drains must be trapped if the DTD/remote dis-
charge system  is to be able to maintain suction in the tiles.
Failure to trap  these drains could result in a dramatic reduc-
tion in  system performance, due to  basement air leakage into
the system via the open floor drain. Untrapped floor drains
can be trapped by installing a waterless trap, as discussed  in
Section 4.7.3, or  by plugging the drain if it is never used.

With the exterior drain tile loops usually encountered  in
remote discharge cases, it is not uncommon to find a variety
of other drains emptying into the tiles. Particular examples
which  have been observed on occasion include drains from
the bottom of basement window wells, and rain gutter down-
spouts. Any such other drains  must be identified, and would
have to be addressed  in some appropriate manner to prevent
excessive air leakage  into the system.

As discussed in Section 5.7, in connection  with sump/DTD
systems, such drains can be impractical to address satisfacto-
rily. Untrapped window well drains might be  trapped with a
waterless trap, similar to untrapped floor drains; however, the
large amount of debris expected in such exterior drains might
quickly render the waterless traps  ineffective. Or, window
well drains might be  sealed altogether if they never receive
any significant  amounts of water, however, this could result in
water problems later, if in fact water did enter the window
wells. One mitigator having experience with DTD systems on
exterior drain tile loops (K192) reports that these systems will
sometimes still give reasonable performance even if connect-
ing window well drains are not trapped or sealed.
Where rain gutter downspouts empty into the exterior tiles,
perhaps the only viable option (if DTD/remote discharge is to
be made to work effectively) is to re-route the runoff from the
gutters to grade level at some appropriate location on the  lot,
so that the gutter penetration into the tiles can be sealed off
altogether. However, rather than run the risk of modifying the
house drainage problems and  of facing a possible liability
issue later, the mitigator would probably be best served in
such cases by leaving the rain gutters alone, abandoning DTD/
remote discharge as the mitigation approach, and using SSD
instead.

The drain tile discharge  line  is  one unique example of a
"drain" opening to the drain tiles, which could enable exces-
sive air leakage into the DTD/remote discharge system if not
addressed. As discussed in Section 6.6, a check valve must be
installed in the discharge line to prevent air leakage into  the
system.

6.8    Gauges/Alarms and Labelling

The considerations discussed in Section 4.8, concerning gauges/
alarms and labelling of SSD systems, also generally apply to
DTD/remote discharge systems. However, some special con-
siderations also apply for the gauges and alarms, since the  fan
and piping for DTD/remote discharge systems will commonly
be entirely outdoors.

Some of the gauges and alarms discussed in Section 4.8.1  are
not weather-proofed for use outdoors. Thus,  some of  the
devices designed for direct mounting on the suction pipe may
not be applicable on the  exterior  piping unless protected by
some form of enclosure. Such an  enclosure could reduce  the
occupants' ability  to see or hear the gauge/alarm.

Often it may be decided  to mount a pressure gauge indoors,
and to connect the  gauge to the exterior suction pipe by 1/8-in.
diameter flexible tubing that will penetrate the house shell. As
indicated in Section 4.8.1 (see Pressure gauges), some miti-
gators  have reported  that moisture can condense or freeze
inside the tubing when the tubing extends outdoors, causing
the gauge to give erroneously low readings. A mitigator may
try to reduce this problem  by  insulating the tubing and/or
using  tubing of a  larger  diameter. However, it would  seem
that the most important step in such cases would be to advise
the occupant of this risk and of how to correct it when  the
problem occurs.

Many alarms/pressure switches are designed to be mounted
directly on the suction pipe. Such alarms having 110-volt
wiring designed to plug into a house outlet may  have  to be
hard-wired into the house circuitry, via exterior rated conduit
extending through the house shell.  Simply plugging  these
devices into an outdoor outlet, or routing the cord indoors and
plugging it into an indoor outlet, will probably not be accept-
able by code in most locations. As indicated in Section 4.8,
the alarm must be wired into a circuit different from the one
handling the fan.

In the particular version  of the ammeter gauge discussed in
connection with Figure 25B in Section 4.6.4, where the con-
duit from the exterior fan would be 24 volts, and where this
                                                       204

-------
wiring would connect to an ammeter and  110-  to 24-volt  centrally located label describing  the system would more
transformer  inside the house,  the 110-volt cord from the  likely be on the electrical panel or some other prominent
transformer could plug into an indoor outlet. Measuring cur-  location, rather than on the system, since the system will be
rent to the fan instead of pressure would avoid the risk of  entirely outdoors. If labels on the system piping are required
condensation and freezing in the outdoor segment of flexible  only on exposed interior piping, in accordance with the cur-
tubing, discussed above.                                    rent draft of EPA's final Radon  Mitigation Standards, such
                                                         labels would not be needed on DTD/remote discharge piping
The labelling requirements indicated in Section 4.8.2 would  that is entirely outdoors.
remain the same for DTD/remote discharge  systems. The
                                                        205

-------

-------
                                             Section 7
                        Design  and Installation of Active
                    Block-Wall Depressurization Systems
In the large majority of cases, SSD (or, where appropriate,
DTD or SMD) will be the first variation of the active soil
depressurization technology to be considered in a given house.
BWD will usually be considered only as a supplement to SSD
in houses having block foundation walls, in cases where SSD
alone does not appear to be adequately reducing radon entry
through one or more of the foundation walls. It appears that a
BWD component is most likely to be required to supplement a
SSD system when sub-slab  communication is marginal  or
poor, and the SSD suction  field is  thus less well able  to
prevent soil gas entry into the wall void network. However,
supplemental BWD has sometimes been found to be benefi-
cial even in cases where the SSD system was maintaining
good suction beneath the slab.

BWD has sometimes  been found to be very effective as a
stand-alone method, without a SSD component. However,  its
performance as a stand-alone method has been inconsistent.

Two variations of the BWD approach are considered: the
"individual-pipe" approach;  and the "baseboard duct" ap-
proach.

Individual-pipe approach. In the individual-pipe approach,
an individual suction pipe is inserted into the block cavity at
one or more locations around the foundation perimeter. This is
the variation illustrated in Figure 5, for the case where BWD
is being used as a stand-alone technique in a basement house.

The individual-pipe approach has been used in stand-alone
BWD installations. It is also the approach commonly used
when a BWD component is supplementing a SSD system.
When used in conjunction with SSD, it is relatively easy to
connect an individual BWD suction pipe to a SSD pipe
located near the wall needing  treatment, as illustrated  in
Figure  34. Thus, suction can be drawn on the walls via the
same piping network that is being  installed for  the SSD
system.

The individual-pipe variation has been tested almost solely in
basement houses. A few tests have been conducted in slab-on-
grade houses, with the individual BWD pipes penetrating the
cavities in the block stem wall at or below grade, horizontally
from outdoors. As  discussed in Section 2.3.4, the results from
slab-on-grade houses are insufficient to enable guidance re-
garding when such a BWD suction pipe is preferable in such
houses over the option of penetrating the pipe all the way
through the wall, making it a SSD pipe. BWD has not been
tested in crawl-space houses, either as a stand-alone method
or as a component supplementing SMD.

Because there is little or no experience in slab-on-grade and
crawl-space houses, the discussion of individual-pipe BWD
systems in this section focuses on basement houses.

Baseboard duct approach. The second BWD variation,
referred to as the baseboard duct approach, is illustrated in
Figure 35. In this approach, a series of holes is drilled into the
void network around the perimeter of the basement, just above
slab level. These holes (along with the wall/floor joint) are
then enclosed within  a plenum ("baseboard") sealed  to the
slab and wall around  the  perimeter, and this plenum is con-
nected to  a fan. In commercial installations, this plenum
commonly consists of pre-fabricated sections of  commer-
cially available channel drain.

Compared to the individual-pipe approach, the baseboard duct
approach should intuitively provide better distribution of the
suction around the walls,  because suction is not being drawn
only at individual points. In addition,  the baseboard duct
approach is more likely to develop depressurization beneath
the  slab—especially in cases where the  wall/floor  joint is a
perimeter channel drain—because the duct will be drawing
some (albeit low-level) suction on the sub-slab region directly
through that crack, providing a SSD component to the BWD
system.

The baseboard duct approach is most commonly used for
stand-alone BWD systems, rather than  for adding a BWD
component to a SSD system. Installation of baseboard ducting
will usually be more difficult than installing individual suc-
tion pipes into one or two  walls beside SSD risers as in Figure
34. Thus, where the objective is to tap a BWD component into
SSD suction piping,  the  individual-pipe BWD approach is
usually employed.

Where a stand-alone  BWD system is envisioned,  the base-
board duct and individual-pipe approaches are more likely to
be comparable in installation effort. In the stand-alone case,
the  increased effort required with the  baseboard duct ap-
                                                     207

-------
                                         1-2" Dia.
                                         PVC Pipe

                                         4" Dia.
                                         PVC Pipe
                                     — PVC
                                        Adapter
         A) Reduced-diameter suction pipe
            installed in wall in order to
            reduce air flow out of wall,
            preventing BWD flows from
            overwhelming SSD system.
                                       4" Dia.
                                       PVC Pipe
                                        PVC Ball
                                        Valve
                                                             Note: Metal Dampers May Not
                                                             Always Reliably Restrict Flows.

                                                                                             4" Dia.
                                                                                             PVC Pipe
                            Metal
                            Damper
                            Installed
                            Inside
                            PVC Pipe
B) Metal damper installed in
   PVC pipe penetrating wall,
   to enable adjustment,
   restriction of air flow
   out of wall.
          C) PVC ball valve installed to
             adjust, restrict wall flows
Figure 34. Some approaches for connecting individual BWD suction pipes into a SSD system.


                                                    208

-------
     Flexible
     Coupling
      Exhaust Fan
      (Rated for
      Exterior Use
      or Enclosed)
                       Exhaust Option 2

                       Exhaust Released
                       Above Eave
Note: Closure of top block voids can be
very important to avoid degradation of
BWD performance and increased
heating/cooling penalty caused by excessive
leakage of house air into the system.
    Exhaust Option 1
     To Exhaust Fan
     Mounted in Attic
                     Close Top Voids

                    Top Void

                     Brick Veneer
                    Concrete Block

                    Drilled Hole
                                                                        Straight Fitting

                                                                       Strapping (or Other Support)
                      Soil
                      Gas
                                                                            Joist
                      Close Major Mortar Cracks
                      and Holes in Walls
                      Basement Air Through Block
                      Pores, Unclosed Cracks,
                      and Holes

                      Suction Pipe Tightly Sealed
                      into Baseboard Duct
                      Baseboard Duct Tightly Sealed
                      Against Floor and Wall

                      Opening in Pipe

                      Sealant Around Entire Seam
                      Where Pipe Penetrates Duct
Figure 35. Block-wall depressurization (BWD) using the baseboard-duct approach.


                                                    209

-------
proach—to drill holes around the perimeter and to seal the
baseboard duct onto the wall and slab—may be offset by the
increased effort required with the individual-pipe approach to
manifold together the large number of individual pipes that
may be needed around the basement

Because of the nature of the baseboard duct installation, this
BWD variation is applicable only in basement houses.

Depressurization  vs. pressurization.  The discussion in
Section 7 addresses the case where the block-wall treatment
systems are operated in suction. Where block-wall treatment
is used in combination with SSD, operation in suction will be
the usual approach.

However, where block-wall treatment is used as a stand-alone
method, the amount of air withdrawn from the basement by
the system may sometimes create sufficient basement depres-
surization to cause back-drafting  of combustion appliances in
the basement. In such cases, the mitigator may wish to con-
sider operation of the block-wall system to pressurize, rather
than depressurize, the walls. Block-wall pressurization is dis-
cussed in Section 9.

The discussion in this section draws heavily from the detailed
review of available data on BWD systems, presented in Sec-
tion 2.3.4.

7.1    Selection of the Number of
Suction Pipes

7.1.1  Individual-Pipe Variation

Number of suction pipes per wall. With the individual-
pipe variation of BWD, each block wall that must be treated
will probably require one or two  suction pipes.

Suction drawn in the block cavities of one wall (e.g., the front
wall) will  usually extend only weakly, if at all, around the
corners into the adjacent end walls, for two reasons. First, the
flows from a  wall are usually fairly high, because of the
extensive leakage area inherent in block walls; for this reason,
the suction and flow field inside a wall will  not be able to
extend very far. Second, the mason who laid the block during
construction could have applied the mortar and laid the blocks
in a manner which inhibits  the extension of the suction and
flow field around the comer.

This same concern exists when there is a discontinuity in one
wall (i.e., a pair of corners in a wall which causes the wall to
be comprised  of two parallel segments offset from  each
other). In such cases, at least one suction pipe may be needed
in each wall segment that is to be treated.

Because the block walls are so leaky, one often cannot rely
upon a wall cavity suction field  extension test using a diag-
nostic vacuum cleaner (analogous to sub-slab suction field
extension  testing) to predict, prior to installation, when a
given wall will need more than one suction pipe. The high air
flows from the wall will often overwhelm the vacuum cleaner,
with the result that little apparent extension of the  vacuum's
suction field may be apparent, even in cases where only one
BWD suction pipe may turn out to be sufficient. As a rule of
thumb, in EPA's installations in Pennsylvania (He87, Sc88), a
second suction pipe was generally added whenever a wall was
longer than about 25 ft.

The need for a second pipe in a given wall will be  greater
when that wall may be subject to particularly high air leakage.
Features which could suggest high leakage include open top
voids which cannot be effectively closed, exterior brick ve-
neer (creating an air gap as illustrated in Figure 5), and a
fireplace structure (which could conceal substantial leakage
routes that cannot be sealed).

Number of walls that must be treated. Having determined
the number of suction pipes required for each wall that must
be treated, the total number of suction pipes in a BWD system
will then depend upon the number of walls requiring treat-
ment. This issue is discussed more fully in Section 7.2.1 (see
Selection of the walls that must be treated).

When the BWD system is supplementing a SSD system, one
or two walls will usually be treated. At one to two pipes per
wall, the total number of supplemental BWD pipes would thus
be one to four, most or all of which will usually be connected
to a nearby SSD pipe with a T  fitting as in Figure 34.

When an individual-pipe BWD system is installed as a stand-
alone measure, all block walls  which penetrate the slab may
need to be treated. This would include the usual four perim-
eter walls, any discontinuities in the walls, any interior block
walls and support structures, and, sometimes, the foundation
walls for a crawl-space wing  that adjoins the basement. A
stand-alone BWD system could thus have a total of four to ten
suction pipes.

In a limited number of cases where one or two walls were the
primary radon entry routes into the house, stand-alone BWD
systems have reportedly been  successful treating only those
walls (Mes92). Thus, in  best cases, one or two pipes have
occasionally been sufficient in  stand-alone systems.

When an individual-pipe BWD system is installed as a supple-
ment to a SSD system, the number of SSD suction pipes will
be selected as discussed in Section 4.1. Usually, the number of
BWD pipes will then be selected as necessary to complement
the SSD system. In some cases,  a tradeoff may be possible.
The mitigator may choose to install an additional SSD pipe in
an effort to prevent soil gas entry into a particular wall, rather
than installing a BWD pipe in that wall, or vice-versa.

7.1.2  Baseboard Duct Variation

Vertical suction pipes. Usually, only one vertical suction
pipe penetrates the baseboard plenum, as illustrated in Figure
35, if the plenum forms a largely continuous loop around the
basement perimeter. The baseboard is supposed to distribute
the suction from this pipe to the walls  around the entire
perimeter.

If the baseboard duct cannot be  continuous due to obstructions
or finish, and if the duct must be installed as two or more
isolated segments,  then a suction pipe must penetrate each
                                                      210

-------
segment. The Auction pipes from the various isolated seg-
ments might be manifolded together, and directed to a single
fan.  Or, the  pipe  from one or more of the isolated duct
segments might be directed to a second fan, as discussed in
Sections 7.4 and 7.6.

Even in cases where the baseboard duct is continuous, without
isolated segments,  it may sometimes be desirable to have a
second suction pipe penetrating the duct at a location remote
from the first pipe. Because of the high flows expected in the
duct, and the relatively small cross-section inside many com-
mercially available pre-fabricated ducts, the suction loss in-
side the duct can be substantial. As a result, the suction inside
the duct at a point remote from a suction pipe may be quite
low  (0.002 to 0.05 in. WG, or less). A second suction pipe,
tapped into the duct on the opposite side of the basement from
the first, could  help increase the suction at the remote loca-
tion, even if the second pipe is manifolded into the same fan as
the first. If the second pipe is connected to a second fan, the
improvement in suction would be greater.

Baseboard ducting. The preceding discussion concerning
the number of suction pipes has focused on the rigid PVC, PE,
or ABS pipe  that penetrates the baseboard duct and leads to
the fan (the vertical pipe labelled "Suction pipe" in Figure 35).
However, the baseboard duct itself might be considered a
"suction pipe,"  in the sense that it directs suction around the
wall/floor joint.

Where there are no isolated segments of duct, the duct forms
only one "suction  pipe." Isolated segments can be avoided
when the duct can form a continuous loop (an "O") or a nearly
continuous loop (a "C," interrupted in only one location, e.g.,
where there is a door or a fireplace structure in the basement
wall). To avoid isolated segments, there would also have to be
no isolated interior block walls/structures penetrating the slab,
requiring a separate segment of duct.

Where isolated segments of duct are necessary, each segment
might be considered a separate "suction pipe." As discussed
previously, each such segment will also  require a separate
vertical PVC suction pipe.

7.2    Selection of Suction Pipe
Location

7.2.1   Individual-Pipe Variation

Lateral position on the wall. If a single individual BWD
suction pipe is being installed into a wall, it would be gener-
ally logical to locate it near the horizontal center of the wall,
midway between the two ends. However, if constraints (such
as basement  finish) prevent it from being installed in the
center,  it should be acceptable to locate if off center. Like-
wise, if two suction pipes are being installed into a wall, they
would logically be placed about equidistant from each other
and from the  ends  of  the wall,  unless constraints dictate
otherwise.

If one section  of the wall seemed to contain higher radon
concentrations  than the remainder of the wall, it would seem
logical to locate a suction pipe in that section.
Where there is a discontinuity in a wall, one BWD suction
pipe would be installed on each side of the discontinuity, to
ensure that each wall segment would be treated. If there is a
fireplace structure in a wall, potentially serving as a major
unclosable leakage source, a suction pipe in that wall might be
located near the fireplace structure.

If the BWD pipe is being installed as a supplement to a SSD
system in a basement, the BWD pipe(s) in a given wall would
normally be installed immediately adjacent to a vertical SSD
suction pipe through the slab beside the wall, if possible, so
that the BWD pipe protruding horizontally out of the wall can
easily be T'd into the vertical SSD pipe.

Where possible, the exact location of a suction pipe would be
selected to minimize the aesthetic impact (e.g., not in finished
areas), and to minimize interference with house usage (e.g.,
not blocking a usual traffic route).

Vertical position on the wall. In terms of the vertical
positioning of the BWD suction pipe on the wall, the suction
pipe should be located as near to the slab as possible, prefer-
ably in the first or second block above the slab. Placement
close to the basement slab should increase the possibility that
the wall-cavity suction field  will extend into the sub-slab
region,  treating the wall/floor joint and other slab-related
entry routes away from the wall. That is, it will maximize any
SSD component that the BWD pipes might provide. In addi-
tion, location of the penetrations close to the slab should help
ensure the best depressurization of the cavities in the footing
region, where most of the soil gas probably enters the void
network.

Also, location of the individual BWD suction pipes near the
basement slab will reduce the height to  which soil gas is
drawn up into the wall. This will potentially  reduce the
amount of radon that might then be drawn into the house if the
low suctions  being maintained in  the voids by  the BWD
system  were temporarily overwhelmed by homeowner or
weather effects.

The BWD pipes use the wall cavities as a soil gas collector,
drawing the gas up into the walls and out through the  suction
pipes. But the suctions established inside the wall voids by the
system are low, lower than fhe 0.01 to 0.15 in. WG measured
in the suction pipes. These  low suctions are particularly
subject to being overwhelmed by the basement depressuriza-
tions that can be created by, e.g., changes in temperature and
turning on exhaust fans such as in clothes driers. If the BWD
pipes are high on  the wall and are,  in essence,  filling the
cavities with diluted soil gas, an overwhelming of the system
could draw this soil gas into the house.

Selection of the watts that must be treated—BWD as a
supplement to SSD. In addition to the location of the pipes
installed in a particular wall, another issue is which walls are
selected to receive pipes.

Where an individual-pipe BWD component is being installed
as a supplement to a SSD system, commonly only one or two
of the block walls are treated. The usual situation is that the
SSD system is taking care of most of the slab-related entry
                                                        211

-------
routes, and there are only one or two walls where the sub-slab
depressurization is failing to adequately prevent soil gas entry
into the wall cavities.

BWD components are usually added to SSD systems either: a)
In cases where prior experience in similar houses (and perhaps
pre-mitigation testing) suggests, prior to mitigation, that such
a component is needed; or b) in cases where an initial stand-
alone SSD system has already been installed and has failed to
achieve adequate radon reductions. Marginal or poor sub-slab
communication, and very high radon concentrations in the
soil gas, are the conditions most likely to create a need for a
BWD component.

In the former  situation, where the possible need for the
supplemental BWD pipes is recognized prior to mitigation,
the prior experience may  suggest which walls could need
treatment. These could include, e.g., the wall beside an adjoin-
ing slab-on-grade garage, or the wall the deepest below grade
in a walk-out basement In some cases, selected pre-mitiga-
tion measurements (radon grab samples drawn from inside the
block cavities) might aid in the selection of a wall to receive a
BWD pipe. Walls having pre-mitigation radon concentrations
greater than some value would be selected for such direct
treatment, because that wall thus appears to be a particularly
important entry route.

But there can be a problem with selecting walls based upon
radon measurements in the cavities prior to  mitigation. The
radon level in the wall is not necessarily a good indicator of
whether the SSD system will adequately prevent entry of that
radon into the wall; the SSD system might take care of that
wall without a BWD component. Only in the unusual case
where radon levels inside the wall are higher than those under
tlie slab, can one confidently conclude from such pre-mitiga-
tion measurements that there is an important source of radon
into that wall in addition to the sub-slab region. Where that is
the case, treatment of the sub-slab alone will probably not
adequately treat that wall.

In the case where an initial SSD system is not doing the job,
radon grab sampling inside the wall cavities with the SSD
system operating  would indicate which walls are not being
adequately treated by SSD. These would be the walls selected
to receive BWD pipes.

Selection of the walls that must be treated—BWD as a
stand-alone system. Where individual-pipe BWD  is in-
tended as a stand-alone mitigation option, it will sometimes be
necessary to treat every block wall which penetrates the slab.
This situation will arise when all of the walls are important
entry routes, and where stand-alone BWD has been selected
rather than SSD, e.g., because of poor sub-slab communica-
tion, or because it is otherwise expected that SSD will not
adequately treat the walls. In this situation, the need to treat
each wall results because the high flows through the walls will
reduce the extension of the suction and flow fields from one
wall to the next, and because the suction field will  not always
reliably turn the comers.

In other cases, stand-alone BWD will be selected because one
or two of the walls appear to be the primary entry route for the
radon into the house. In these apparently limited cases, it can
be sufficient to treat only those specific walls (Mes92).

In cases where all walls are important entry routes, it will be
necessary to treat:

   -  Each of the four perimeter foundation walls. Even  a
      perimeter wall that is largely above grade (such as the
      above-grade wall in a walk-out basement) should gen-
      erally be treated; it can still be serving as a conduit for
      soil gas from the  footing up into the basement, even
      though there is no soil in contact with the above-grade
      portion of its exterior face.

   -  Each segment of each perimeter wall, in cases where a
      discontinuity in the wall divides it into segments.

   -  Any interior block wall which penetrates the slab, e.g.,
      a load-bearing wall perpendicular to the front and rear
      walls, dividing the basement into sections.

   -  Any interior block structure, such as a structure hous-
      ing a fireplace in  the basement, or supporting a fire-
      place on the floor  above.

Where there is a slab-on-grade or crawl-space wing adjoining
the basement, the block foundation walls associated with that
adjoining wing may also need to be treated, especially in cases
where the void network  in the walls for  the adjoining wing
does not  link with the network in the basement walls.

Indoor vs. outdoor locations. In most cases in basement
houses, the individual BWD suction pipes are installed into
the wall from inside the  basement, as illustrated in Figure 5.
Indoor installation is  generally simpler, at least when the
basement is not significantly finished, and it minimizes the
piping visible from outdoors. Where the BWD pipes are
supplementing an interior SSD system, the BWD pipes will
always be indoors, so that they can easily be tied into the SSD
piping. The BWD pipes treating interior walls and structures
will necessarily always have to be indoors.

However, when abasement in heavily finished (and when any
SSD pipes  are also installed from outdoors), the BWD pipes
for the perimeter foundation walls can be installed into the
blocks from outdoors. Exterior installation would mean drill-
ing into the block cores from outdoors rather than indoors. It
would require excavating  beside the exterior face of the
basement wall to expose the exterior face of the blocks at the
location where the BWD pipe is to be installed. If the pipe is
to penetrate the wall near the footing, this could mean an
excavation as deep as 5 to 8 ft. The individual exterior BWD
pipes installed in this manner around the perimeter of the
house could be extended vertically upward to near grade level,
and manifolded together via a horizontal loop of piping buried
just below grade level around the perimeter. Any exterior SSD
pipes (penetrating horizontally through the  foundation wall
below slab level) could also be connected to this same piping
network.

Other considerations for pipe location in basements.
When individual BWD pipes are being installed from inside a
                                                       212

-------
basement, some of the other considerations in selecting pipe
location will be similar to those discussed in Section 4.2.1
regarding  the  location of interior  SSD pipes. These other
considerations include: location of the pipes in unfinished or
inconspicuous areas where possible; location of the pipes in
an effort to minimize interference with occupant traffic pat-
terns (which should not usually be a problem) or with occu-
pant usage of the space; location of the pipes to facilitate
manifolding of the multiple pipes  together, in cases where
they cannot be tapped into an immediately adjacent SSD pipe;
and location of the pipes away from any utility lines which
may occasionally be inside the block cores (such as electrical
conduits leading to electrical outlets in the block wall).

Considerations for slabs on grades. Where individual
B WD pipes are installed in a slab-on-grade house, they will be
installed horizontally into the cores of the foundation stem
wall from outdoors, near grade. In essentially all cases where
BWD pipes have been installed in a slab-on-grade house, they
have been installed in conjunction with SSD pipes also being
installed horizontally through the foundation wall from out-
doors. In  such cases, it would appear logical to install  the
BWD pipes into the block cores just below the slab, at  the
same level as the SSD pipes. Such BWD pipes might be
expected to have the greatest benefit on foundation stem walls
which are largely below grade. In cases where the stem wall
comes above  grade because the  slab  is above  grade,  the
amount of outdoor air drawn through the exterior  face of the
block will probably be increased.

7.2.2 Baseboard Duct Variation

Vertical suction pipes. Where a single vertical PVC suction
pipe penetrates a continuous baseboard duct, this suction pipe
can be located beside the foundation wall a^where around
the basement perimeter. The location of this suction pipe can
be selected for convenient routing of the exhaust piping—i.e.,
directly below a  convenient route for an interior stack up
through the living area above, or directly beside a convenient
point for  penetration through the band joist  to an exterior
stack (e.g., at the rear of the house). It can be selected in an
unfinished part of the basement.

Sometimes it is desirable to install a second  (and maybe a
third) suction pipe into a continuous baseboard duct, because
the high flows from the walls can sometimes prevent  the
suction field from one pipe from effectively extending through
the duct around the entire perimeter. A second suction pipe
should be located approximately on the opposite  side of the
basement  from the first pipe. The exact location would be
selected to avoid basement finish and other obstructions, and
to facilitate the routing of the piping. If this second pipe is to
be manifolded together with the first, and both directed to a
single fan, the locations of the two pipes would be selected to
facilitate this manifolding. If the second pipe is to be routed to
its own fan, its location would be selected to facilitate  the
routing of a second stack up through the house, or the routing
of the piping through the band joist to a second exterior stack.

Where the baseboard ducting is interrupted, so that there are
two or more isolated segments of ducting around the perim-
eter, there will have to be a vertical PVC suction pipe located
in each of these isolated segments. Within  the flexibility
provided by the locations and extents of the segments, these
suction pipes should be installed at locations in the segments
that facilitate manifolding  the  various pipes together, and
directing them to one or two fans.

Baseboard ducting. Ideally, the baseboard ducting itself
should be installed over the wall/floor joint around the entire
basement perimeter, without interruption. In addition to being
thus installed on all four perimeter walls and  around any
discontinuities in these walls, it should also be  installed on
any interior block wall,  and around any interior  block struc-
ture (such as a fireplace  structure), that penetrate the slab.

Complete coverage of the entire perimeter may not be neces-
sary for adequate radon reduction, as  long  as there is a
significant segment of ducting along each block wall to ensure
that each wall receives some treatment. Among  the obstruc-
tions that can cause interruptions in the ducting are: doorways
through the block wall  (e.g., in walk-out basements); fire-
places built into the wall; heavy interior finish or framing; and
furnaces, stairways, shower stalls, and other obstructions in-
stalled flush against (or very near to) the block wall.

The need for complete coverage of the perimeter by  the
baseboard duct might be questioned.  The individual-pipe
BWD approach applies suction only to isolated points along
the walls, not continuously around the perimeter. However,
most testing of the baseboard approach  to date  has  utilized
nearly complete coverage. And, in the limited testing by EPA
of very high-radon houses in Pennsylvania (Sc88), even with
nearly complete coverage, these houses were  not reliably
reduced below 4 pCi/L.

Many commercially installed baseboard duct BWD systems
are installed for the combined purpose of radon reduction and
basement water control (E188). Commercially available "base-
board ducting" has been on the market  for basement water
control for many years, prior to the current concern about
radon. Where the ducting has a water control as well as a
radon control purpose, the need for ducting to be installed in a
contiguous length may be mandatory, so that the collected
water can flow through  the duct, around the perimeter, to a
sump for discharge.

With interior block walls, where both faces of the wall  are
accessible, installing the baseboard duct on  just one face
might sometimes be sufficient. If the interior wall separates a
finished portion of the basement from an unfinished area, the
duct would conveniently be installed on the unfinished side.

By the very nature of the system, the baseboard  duct is
installed over the wall/floor joint inside the basement. The
suction on the wall—through holes drilled through the wall a
few inches above the slab, enclosed within the duct—is thus
applied at the base of the wall. Applying the  suction at this
location: a) increases the potential that the BWD system will
develop a suction field under the slab, creating a SSD compo-
nent; b) increases the suction in the block cavities near  the
footing, where much  of the soil gas likely enters the void
network; and c) reduces the height to which the system will
draw soil gas up into the wall, as discussed in Section 7.2.1.
                                                        213

-------
Where the baseboard duct is also used for water drainage, and
where a sump is thus also enclosed within the system, suction
on this sump increases the SSD component of the system.

7.3    Selection of Suction Pipe Type
and Diameter

7.3.1  Individual-Pipe Variation

Type of piping. As with SSD and DTD systems, the type of
piping typically used for individual-pipe BWD systems .is
rigid thin-walled or Schedule 40 PVC, PE, or ABS piping.

Pipe diameter—BWD as a supplement to SSD.  The
selection of the pipe diameter is influenced by the very high
air flows that BWD pipes will draw out of the walls. In the
large majority of cases,  the substantial friction loss in the
piping, combined with the high flows, will result in quite low
suctions in the block voids, even when the 90-watt tubular in-
line fans are employed.

In the most common case, where individual BWD pipes are
installed as a supplement to a SSD system and are connected
to nearby vertical SSD suction pipes, reducing the diameter of
the BWD pipe is one common option for restricting flows out
of the walls. See Figure 34A,  and the discussion in Section
7.5.1 regarding the need to restrict flows out of the walls in
combined  SSD+BWD systems. In the typical case illustrated
in Figure  34A, the  vertical SSD pipe is  the usual 4 in.
diameter, but the horizontal BWD pipe is 2 in. diameter.

However, as shown in Figures  34B and 34C, 4-in. piping can
also  be  used for the BWD leg in  combined SSD+BWD
systems. Where 4-in. piping is used, a damper or  valve is
commonly installed in the BWD pipe to restrict the flow from
the wall.

Whichever of the approaches in Figure 34 is utilized, flows
from the walls are usually restricted adequately such that there
is no need to use piping larger than 4 in. for the remainder of
the combined system. See Section 4.3.2 for discussion of the
selection of pipe sizes for SSD systems.

Pipe diameter—BWD as a stand-alone system. When
individual-pipe BWD is being used as a stand-alone method
as in Figure 5, with no SSD component, restricting air flow
from the walls is no longer an issue. For stand-alone  systems,
radon reduction performance will likely be improved if the
wall flows are increased as necessary to achieve adequate
suction field distribution in the void network. Thus, piping of
adequately large diameter should be selected.

EPA's limited experience with stand-alone BWD  systems
indicates that, with a 90-watt fan and 4-in. diameter individual
suction pipes in each wall, the flow in each pipe will typically
be no more than about 50 cfm, sometimes less. At that flow
rate, 4-in.  piping can be used for the individual suction pipes
into the walls without suffering an unacceptable suction loss,
especially since any one suction pipe will probably extend no
more than 10 or 20 ft from the wall to a central trunk line
(collector) running down  the center of the basement. From
Figure 13, friction losses between the wall entry point and the
trunk line would be on the order of 0.05 in. WG or less under
these circumstances.

However, that central collector pipe will carry the combined
flow from all of the individual wall pipes, and can thus see
particularly high flows. With 4-in. piping, the maximum flow
that would be generated by a 90-watt "270 cfm" tubular fan is
about 180 cfm, as discussed in  Section 4.4.1. With 6-in.
piping, flows as high as 200 cfm have been observed in BWD
systems.

At 180 cfm, the central collector pipe could sustain a friction
loss of about 0.75 in. WG down a 30- to 40-ft-long basement
if the collector were simply a straight length of 4-in. piping
(from Figure 13). This loss would cause the fan to operate at a
different point on its performance curve, significantly reduc-
ing the total system flows well below 180 cfm. As a result, the
suctions that would be being maintained in the walls would
decrease significantly.

Thus, in  the EPA installations, 6-in.  diameter  piping  was
consistently used for this high-flow central collector. With a
straight 6-in. collector 30 to 40 ft long, the friction loss would
be only about 0.1 in. WG at 200 cfm. Accordingly, Figure 5
shows a 4-in. diameter wall suction pipe connecting to a 6-in.
trunk line.

7.3.2  Baseboard Duct Variation

Vertical suction pipes. Commonly, the vertical suction pipe
that penetrates the baseboard duct and extends up to the fan is
PVC (or comparable) piping.

Some early research installations used sheet metal ducting
extending up the basement wall for this purpose, in order to
increase  the  cross-section and  thus reduce suction  losses.
However, such sheet metal ducting is more difficult to  install
and is  difficult to seal  effectively. It does not provide a
sufficient reduction in friction losses to compensate for these
complications. Thus, use of PVC piping is now recommended.

Where a single vertical suction pipe is used to draw suction on
the baseboard duct at one location around the perimeter,
experience suggests that flows in this pipe can be roughly 100
cfm  or more. (Exact flows will  depend upon a variety of
variables.) This flow is lower than the values that are  some-
times seen in the trunk line of individual-pipe BWD installa-
tions (up to 200 cfm), because of the friction losses encoun-
tered inside the baseboard duct. At  100 cfm, 4-in. diameter
piping can probably be used, especially if the total piping run
is only about 30 ft (e.g., extends relatively straight up from the
basement slab to the exhaust point at the roof). In this case, the
total suction loss would be on the order of 0.2  in. WG, an
amount that could usually be handled by a 90-watt fan.

Where a  second or third suction pipe tap into the baseboard
duct, 4-in. piping may still be sufficient, depending upon how
the piping is routed. If the piping connects directly to a second
fan, essentially  identical to the first pipe, then 4-in. piping
could again be adequate. If the piping from the second suction
pipe is going to be directed across the basement to manifold
into the first fan, 4-in. piping might again be sufficient for the
                                                      214

-------
roughly 20- to 40-ft piping run across the basement, assuming
again a flow of about 100 cfm or less. But where the first and
second pipes come together, the combined flow could become
significantly greater than 100 cfm. Thus, 6-in. piping could be
warranted downstream of the junction, including for the ex-
haust stack.

Because the high flows could sometimes warrant 6-in. diam-
eter piping, Figure 35 shows all of the PVC piping as 6-in.
piping.

Baseboard ducting. In most commercial BWD/baseboard
duct installations, the base-board ducting consists of commer-
cially  available, pre-fabricated  ducting composed of vinyl,
ABS, or similar material. This commercial ducting, which has
been marketed for years for basement water control purposes,
is commonly sealed to  the wall and  slab  using an epoxy
adhesive. Figure 35 illustrates one version of such commer-
cial ducting.

The commercial ducting shown in Figure 35 has a bottom that
curves back toward the wall, and is cemented directly on top
of  the wall/floor joint.  This configuration creates  a vinyl
channel inside the ducting, which can be helpful in channel-
ling any water which enters the duct. Commercial channel
drains for water control  would normally be considered for
houses which do not already have a perimeter channel (or
canal) drain in the form of a 1- to 2-in. wide wall/floor joint.
The particular commercial drain shown in the figure would be
a reasonable choice in cases where there is not a pre-existing
perimeter canal.

However, when a baseboard duct system is being considered
for radon mitigation, there may be cases where the baseboard
approach may be of interest in part because a perimeter canal
drain does already exist. In such cases, a commercial channel
drain having the cross-sectional configuration in Figure 35
would cover the canal, interfering with drainage into the canal
and with extension of the baseboard duct suction field into the
sub-slab region. Thus, where there is a pre-existing perimeter
canal drain, one may wish to use a commercial channel drain
that attaches to the slab a couple inches away from the wall, in
order to leave the canal open.

In some custom installations, the baseboard ducting has been
fabricated by the mitigator using sheet metal. The sheet metal
duct would be attached to the wall and slab using masonry
screws and caulk. The metal duct can be fabricated, to create
either a rectangular or a triangular cross section when attached
to the wall. Fabrication of the ducting out of sheet metal has
the advantage of enabling a much larger cross-sectional area
inside the  duct than is available with the relatively small
commercial ducting, thus reducing  friction losses. However,
custom fabrication and installation  of sheet metal ducting is
more difficult  and time-consuming than is the use of the
commercially available  ducting, and hence will usually be
much more expensive. In addition, because it is larger and is
not pre-fabricated, the  metal ducting will have  a greater
aesthetic impact.

Commercially available baseboard ducting is typically fairly
small. It is commonly only 4 to 5 in. tall. As illustrated by the
example in Figure 35, it has a cross section that is not quite
rectangular, creating an enclosure extending an average of
perhaps only 1 in. out from the wall. Thus, when attached to
the slab and wall, this ducting will have a cross sectional area
of 4 to 5 in2 (equivalent to a round pipe having a diameter of
roughly 2.5 in.). Given the high flows that will come from the
walls, the suction loss inside this duct will be substantial; the
ducting was designed to channel water and to be aesthetic, not
to move air. The friction loss will significantly reduce the
suction that can be maintained in the duct, and the amount of
air that can be drawn out of the walls, unless one or more
additional suction pipes are installed into the duct. However,
since this low suction will be distributed around the entire
perimeter by the ducting, and since the wall/floor joint, a
major entry route, will be  largely or entirely enclosed and
depressurized, albeit to only a low level, these features may
help compensate for the low level of the suction.

Custom installations have utilized sheet metal ducts having
much larger cross sections, from 12 to 36 in2 (Sc88). The cross
sectional area in these ducts would be equivalent to that in a
round pipe having a diameter of 4 to 6.75 in., greatly reducing
suction loss compared to that in the commercial ducting.
However,  the larger  ducts have a much greater aesthetic
impact, as well as being more difficult to fabricate and install.
And even with the larger baseboard ducting, the pressures
inside the ducting (with the system operated to pressurize the
wall) were only 0.002 to 0.38 in. WG (and flows were 0.2 to
96 cfm), depending upon proximity to a fan (Sc88, Fi91).

7.4    Selection of the Suction  Fan

With either the individual-pipe or the baseboard duct varia-
tions of the BWD approach, the high air flows expected from
the walls will generally dictate that the fan be at least compa-
rable to the 90-watt in-line tubular fans listed  in Table 1.
These  fans can move up to 270 cfm at zero static pressure,
although their practical maximum in BWD systems will prob-
ably be about 180 to 200+  cfm, depending upon the size of
piping used. These fans will  often be able to maintain ad-
equate suctions  at the wall suction hole, at the  air flows
encountered and with the suction losses encountered in the
piping.

As discussed in Section 4.4, the 90-watt in-line tubular fans
(Section 4.4.1)  are generally recommended for stand-alone
SSD systems having good to poor sub-slab communication.
Higher-suction blowers (Sections 4.4.2 and 4.4.3) can be an
appropriate alternative when communication is  marginal to
poor, and the smaller (50- to 70-watt) in-line tubular fans
(Section 4.4.1) will sometimes be suitable when communica-
tion is good but flows are not excessively high.

When the SSD system is supplemented with a BWD compo-
nent, the flows from the combined SSD+BWD  system will
tend to be higher than those from stand-alone SSD systems.
This can be true despite the fact that steps have been token to
restrict air flows out of the walls, as illustrated in Figure 34.
Because of these higher flows, the 50- to 70-watt tubular fans
will generally not be a good choice for such combined sys-
tems. Likewise, flows will very likely be too high for the
high-suction/low-flow fans (Section 4.4.3). And flows will
                                                        215

-------
likely be high enough such that the radial blowers (Section
4.4.2) would not be generating the higher suctions which is
the primary incentive for their use. Accordingly, the 90-watt
tubular fan will probably often be the logical choice for a
combined SSD+BWD system.

With stand-alone BWD systems, total flows from the systems
(when 90-watt fans are used) are usually on the order of 100 to
200 cfm, depending upon the pipe size. Suction in the system
piping immediately beside the walls is usually 0.1 to 0.4 in.
WG or less, sometimes significantly less. Under these condi-
tions, it is recommended that the fan always have a capability
at least equal to that of the 90-watt in-line fans. In some cases,
where the BWD system includes only a single fan, one of the
high-flow 100-watt fans listed in Table 1 (with 6-in. or even
8-in. couplings) might be considered.

In some of EPA's research installations with the baseboard
duct BWD variation, it has been found to be desirable to add a
second 90-watt fan, in order to maintain adequate suctions and
flows in the duct around the entire perimeter. An additional
benefit of the second fan was that where there were multiple
isolated segments of baseboard ducting, it was sometimes
easier logistically to route the suction pipes from the segments
on one end of the house to a second fan, rather than attempting
to manifold all pipes to a single fan. This second benefit could
also sometimes warrant a second fan in individual-pipe BWD
installations,  when it is not practical to run a single 6-in.
collector pipe down the  center of the basement to direct the
combined flow from all of the individual wall pipes to a single
fan. In any case, one or more of the suction pipes (from one
side of the house) would be directed to one of the fans, and the
remaining suction pipes (from  the other side) would be di-
rected to the second fan.

7.5    installation of Suction Pipes
Into the Block Walls

7.5.7  Individual-Pipe Variation

A hole of the same diameter as the  BWD suction pipe is
prepared through one face of a block at the location where the
individual pipe is to be installed. This hole would expose the
cavity inside the particular block, and would not penetrate all
the way through to the other side of the wall. This hole would
be drilled through the inside face of the block wall, inside the
basement, in cases  where the BWD pipes are to enter a
basement wall from indoors. Or, it would be drilled through
the outer face of the block, from outdoors, if the pipe is  to
enter a basement wall or a slab-on-grade foundation stem wall
from outside the house.

The hole would be positioned on the selected block so that it
penetrated the face at the location of a cavity. That is, it would
not at the location of one of the block's interior cross mem-
bers. To judge the proper location,  one must identify whether
the blocks in that wall have two interior cavities, or three.
Often, the top of some blocks will be exposed at some point
around the basement, revealing how many cavities the blocks
contain.
Commonly, a rotary hammer is used to prepare the hole into
the block, by drilling a series of 1/2-in. holes in a circular
pattern having the diameter of the BWD suction pipe. Because
the block face over the cavity is thin, and because the con-
crete/aggregate mix used to fabricate the blocks is less dense
than is solid concrete, drilling holes through the face of blocks
with a rotary hammer is much simpler than is drilling such
holes through a 4-in. thick concrete slab, as discussed in
Section 4.5.1.

The BWD suction pipe is inserted horizontally, partway into
the block cavity. After this pipe is supported by connection
into the remainder of the piping network (as illustrated in
Figure 5 or 34), the gap between the block face and the pipe,
around the pipe circumference, must be  sealed. Gun-grade
(non-flowable) polyurethane caulk should be worked into the
gap to form a good seal. If this gap is not sealed, air will leak
through the gap, reducing  the effectiveness of the BWD
system^ Because this gap will be  immediately beside the
suction pipe, it will potentially have a greater impact than will
other unsealed openings in the wall.

When individual BWD pipes are connected into a SSD  sys-
tem, it is usually necessary to restrict the air flows out of the
wall, using a technique such as those illustrated in Figure 34.
The reason for restricting the wall flow into the SSD+BWD
system is that high flows from the BWD pipes would signifi-
cantly reduce  the suction that could be  maintained in the
system, including the SSD pipes. Thus, failure to restrict the
wall pipes  would significantly reduce the  treatment of the
sub-slab by the SSD system.

Of course, restricting the BWD pipes in this manner dramati-
cally reduces the treatment of the walls. As discussed in
Section 2.3.4, tests of such combined SSD+BWD systems
have commonly shown that the optimum approach in these
cases is to restrict the BWD flows, sacrificing some part of the
wall treatment in order to maintain effective sub-slab treat-
ment

One of the simplest, least expensive, and more widely utilized
methods for restricting wall flow is to use smaller diameter
piping for the leg extending into the wall. Figure 34A illus-
trates this approach, for the case where 2-in. pipe has been
used for  the wall and where the vertical SSD pipe is 4 in.
diameter. For this configuration, 4-in. T fitting is installed into
the SSD riser adjacent to the wall hole, with the leg of the T
pointed horizontally toward the wall. A 4- to 2-in. adaptor at
the end of the T leg narrows down to accommodate the 2-in.
pipe that then extends horizontally into the wall cavity. Be-
cause of the substantial suction loss in even such  a short
length of 2-in. piping, the use of the  small-diameter piping
significantly reduces the flow out of the wall into the system.

Figure 34B shows the option of using 4-in. pipe for the wall
leg, and of installing a damper in the leg to allow restriction of
flows from the wall. Dampers are relatively inexpensive, and
would enable adjustment of wall flows after installation in an
effort to optimize system performance. However, some miti-
gators have reported that dampers  are not always reliable
(K192).
                                                       216

-------
Use of a PVC ball valve, as illustrated in Figure 34C, also
permits adjustment of flows after installation, and offers the
further advantage of being more reliable than the use of a
damper. However, PVC valves are relatively expensive.

7.5.2  Baseboard Duct Variation

Vertical suction pipes—where  no sump is present.
Some manufacturers of channel drain market a  specially
designed segment of baseboard ducting, designed to easily
accommodate a vertical PVC suction pipe. Such a segment
would be installed into the baseboard loop at the one or more
locations where a vertical PVC suction pipe is to be installed
into the baseboard ducting. The suction pipe would then be
cemented into the fitting in this segment, providing an air-tight
seal. Where such connection segments are available, this
would be the preferred approach for connecting the suction
pipe into the baseboard.

Alternatively, as illustrated in Figure 35, a hole could be cut in
the baseboard ducting to accommodate the suction pipe. The
pipe is installed vertically, almost flush against the foundation
wall, fitting into the hole cut in the ducting. At the base of the
vertical pipe, against the slab, a section of the pipe within the
ducting would be cut away, providing an opening between the
suction pipe and the interior of the baseboard ducting.

By this latter approach, the junction between the PVC suction
pipe and the baseboard ducting would then have to be sealed
very well, with gun-grade polyurethane caulk or other appro-
priate sealant The base of the vertical pipe, where it rests on
the slab, would have to be effectively sealed against the slab
with polyurethane caulk. Alternatively, an end cap could be
cemented onto the base of this pipe. And the seam between
the pipe and the baseboard duct must be sealed.

Unlike vertical SSD pipes, where the pipe penetration through
the slab tends to anchor the pipe against lateral movement
caused by physical impact, the vertical suction pipe in base-
board duct BWD systems is resting on top of the slab, and has
only the caulk bead at its base to prevent lateral movement.
This pipe should be attached firmly to the wall using brackets,
in order to reduce the risk that it will be jarred and the seals at
its base broken.

Vertical suction pipes—where a sump is present. The
preceding discussion addresses the case where the baseboard
duct does not connect to a sump. Where there is a sump, the
vertical suction pipe can be installed in the capped sump,
rather than as described above.

In many cases, commercial baseboard duct systems  are in-
stalled for the combined purposes of basement water control
and of radon reduction. The need for water control may often
be a factor in the decision to install a baseboard duct BWD
system in the first place.  Where water control is also an
objective, a sump and sump pump will commonly have to be
retrofit into the house as part of the system,  to handle the
collected water. (The need for water control and for a retrofit
sump will usually not exist in houses having a pre-existing
perimeter channel (canal) drain.)
An added concern is that when holes are drilled through the
interior face of the block wall near the slab, basements that did
not originally have a serious water problem may begin to have
one. Water flowing through the block cavities during wet
periods, which previously might have drained largely to the
sub-slab region via the base of the wall, might now flow into
the baseboard ducting through the holes. To avoid claims that
the mitigation system has negatively impacted the drainage
features of the  house, it might sometimes be advisable to
install a sump and sump pump system as an integral part of the
baseboard duct system, to handle any such drainage in cases
where water problems might occur.

Figure 36 illustrates an approach used by one vendor in cases
where combined water control and radon mitigation are to be
achieved with a baseboard duct system and a retrofit sump
(E188). In this configuration, a single PVC pipe serves both as
the conduit for collected water from the baseboard  ducting to
the sump, and also as the pipe by which suction is drawn on
both the baseboard ducting and the sump. To install this pipe
and the new sump, of course, a section of the slab would have
to be removed and then restored after installation.  The sump
must be fitted with an air-tight cover, as discussed  in Section
5.5.2.

In other cases, the baseboard ducting  might connect to the
sump by a configuration other than that shown in Figure 36. In
such other cases, it might sometimes be preferred for the
suction pipe to penetrate the sump cover, in a manner such as
illustrated in Figure 33 and discussed in Section 5.5.2.

An installation such as the one in Figure 36 has a major SSD
component by virtue of the suction on the sump.  This SSD
component can be niore important than the BWD component.

Baseboard ducting. Prior to the installation of the base-
board  ducting, holes of about 1/2 in. diameter are drilled
through the interior face of the blocks, into the cavities, at all
locations where the ducting is to be installed.  These holes
would usually be in the first course of block above the slab,
within 4 in. of the slab, so that the will be enclosed within the
baseboard ducting. These holes permit the suction from the
depressurization system to extend into the void network uni-
formly around the perimeter. In EPA's testing of  baseboard
duct BWD systems (Sc88), as well as in commercial systems
(E188), such holes have usually been drilled into every cavity
of every block.

No attempt should be made to seal the wall/floor joint at
locations where the baseboard ducting  will be installed over
the joint. Leaving the joint open will facilitate the extension of
the ducting suction down into the sub-slab region,  increasing
any SSD component of the system. Where the baseboard does
not cover the entire length of the wall/floor joint,  and where
that joint is a 1- to 2-in. wide perimeter channel drain, the
channel drain opening beneath the baseboard must be sealed
in some appropriate manner at the point where the baseboard
ends, so that basement air is not drawn into the baseboard duct
through the open perimeter channel drain beneath  the end of
the baseboard.
                                                       217

-------
                             Baseboard
                             Duct
       Concrete
       Block-
       Drilled
       Hole
        PVC Suction
        Pipe

        Water Discharge
        Line
JjL/— Opening in Suction Pipe,
  v     Inside Baseboard Duct
  Soil
  Gas
                                                                                             Screw
                                                                                                    Gasket
                        Soli Gas from Sub-
                        Slab Region Entering
                        Suction Pipe	
                               Water from Baseboard
                               Duct Entering Sump
                                            Perforations in
                                            Side of Surnp Liner,
                                            to Allow Suction
                                            Field to Extend
                                            into Sub-Slab
                                            Region (Providing
                                            SSD Component)


                                              Retrofit
                                              Sump Liner
Figure 36. One specific example of a baseboard duct BWD configuration with a major SSD component, for the case where a sump and sump
          pump are installed as part of the radon mitigation system. (Safe-Aire, Inc.)
The baseboard ducting must be attached to the slab and wall,
forming an air-tight seal over the wall holes and the wall/floor
joint. Where commercially available vinyl or ABS ducting is
used, the ducting is usually bonded to the wall using epoxy or
other adhesive. The surface of the wall and  slab should be
cleaned and abraded first, to improve adhesion. If sheet metal
ducting were used, the metal ducting would be anchored to the
wall and slab with masonry screws, through flanges incorpo-
rated into the ducting  for this purpose. The metal ducting
would  be sealed against the wall and slab with a continuous
bead of gun-grade polyurethane caulk between the flange and
the concrete. The masonry screws alone would not provide an
adequate seal. When the slab contains irregularities, special
care and additional caulking will be needed to ensure a good
seal.
         Commercially available ducting is marketed in pre-fabricated
         5-  to  10-ft lengths.  Butt-joint  fittings, having the same
         cross-sectional configuration as the ducting, are used (with
         sealant)  to join lengths of straight ducting.  Special
         pre-fabricated sections of ducting can sometimes be obtained
         to turn inside or outside corners. Where these corner sections
         do not form an integral piece, the mitered junction at the
         comer must be sealed carefully. Where the ducting  must be
         interrupted, end-cap fittings are available to seal the open end.
         As indicated above, where the baseboard ducting is being
         sealed over a perimeter channel  drain, the gap between the
         open drain and the base of the baseboard must be sealed at
         ends.
                                                       218

-------
When the lengths of baseboard ducting are custom fabricated
out of sheet metal, more care will be required in sealing the
seams using sheet metal screws and polyurethane caulk. The
butt-joint and end-cap  fittings, and the pre-mitered corner
sections, will not be available, and these junctions will have to
be formed and sealed on site.

When the baseboard is also going to  collecting  water, a
particular effort is needed to ensure that the baseboard forms a
contiguous loop (at least  a "C," if not a complete "O").
Continuous ducting is needed so that the collected water can
drain to a single sump. In these cases, it may be necessary to
move some obstructions, or to penetrate through an interior
stud wall which extends perpendicular to the block foundation
wall. If part of the basement is finished, the baseboard ducting
can sometimes still be installed, by trimming of the base of the
panelling and underlying furring strips, and by trimming any
carpeting, to accommodate the baseboard. (Where the finish
against the foundation wall is a stud wall, this approach would
not be practical.)

The holes that are being drilled through the face of the
basement wall will permit any water that ever flows through
the wall cavities to drain out into the baseboard ducting. Prior
to the drilling of these  holes, much of the water that might
have entered the cavities might have drained to the sub-slab
region. Thus, if it is thought that the house might have a
drainage problem (based upon  the contour of the lot, the
appearance of the walls, and the occupant's experience), a
sump should be installed in conjunction with the baseboard
duct system. If there is an existing sump, it could be connected
to the baseboard ducting.

If the wall/floor joint is a perimeter channel (canal) drain, it
might drain to an existing sump  via a loop of drain tiles under
the slab. See Figure 32B. In these cases, the perimeter channel
drain is enclosed within the baseboard ducting, using a base-
board duct configuration that does not interfere with drainage
into the perimeter canal (see Section 7.3.2, Baseboard duct-
ing). The sump should  be capped with an air-tight  cover, to
prevent excessive air flow through the drain  tiles  into the
baseboard duct system (as well as to reduce radon entry into
the house via the sump).

If the wall/floor joint is not a  perimeter channel drain, but
there is an existing sump, a channel could be routed through
the slab, directing water from inside the baseboard ducting to
the sump. This would be analogous to the approach illustrated
in Figure 32B, for the case where a perimeter channel drain is
sealed. If a channel were routed, it must be covered with, e.g.,
Plexiglass sheet, sheet metal, or a length of PVC piping cut in
half down its axis. In this case, it could be desirable for the
sump cover to rest on top of the slab (rather than below slab
level, as in Figure 32B), so that the sump end of the routed
channel could be enclosed under the sump cover. The seams
between the channel cover, the sump cover, and the baseboard
ducting must be sealed  well. The baseboard ducting, the slab
channel, and the sump should form one completely enclosed
unit on which suction can be drawn, with waiter flow through
the perimeter drain and into the sump unimpeded.
If there is no sump initially, one can be installed as part of the
combined radon mitigation/basement water control system.
One option for doing this was illustrated in Figure 36. Poly-
urethane or polyethylene sump crocks can be purchased for
this purpose, with the top molded (and  with bolt holes pro-
vided) to accommodate a gasketed, air-tight sump cover that
is sold with the crock.

Installation of a sump will require removal of a section of the
slab near to the foundation wall with a jackhammer, excava-
tion of a pit at that location to accommodate the sump crock,
and restoration of the slab. If a PVC pipe is installed into the
side of the sump crock to baseboard water into the sump and
to enable suction on the sump and baseboard, as illustrated in
Figure 36, a portion of the footing might also have to be
chipped away, as shown, to make room for the pipe. After the
sump crock and  the PVC pipe are mounted  in place, the
excavation is back-filled with material  that was excavated,
with a layer of crushed rock on top, and new concrete poured
to restore the slab.

While the new concrete is still soft, a groove perhaps 1/2 in.
deep is tooled around the perimeter of the slab hole, where the
new concrete meets the original slab. A groove is also tooled
around the perimeters of the sump crock and of the PVC pipe
where it penetrates the new concrete, and along the wall/floor
joint (if the slab hole extended all the way to the foundation
wall).  When the concrete  has  set,  these  grooves will be
flooded with flowable urethane caulk, in an effort to keep
these seams gas-tight.

7.6    Design/Installation  of the
Piping Network and Fan
The design and installation of the piping network and fan for
BWD systems would be essentially the same as described in
Section 4.6 for a SSD system.

In a stand-alone BWD installation utilizing the individual-pipe
variation, where multiple individual 4-in. BWD pipes  pen-
etrate into the cavities of different walls, it is advisable that
the individual 4-in. pipes from each wall connect to a  6-in.
diameter collector pipe, due to the volume of  the combined
flows. When the wall penetrations are made from inside the
basement, this 6-in. trunk line would usually run down the
center of the basement. If the penetrations are from outdoors,
the trunk line would form a loop around the house perimeter.
Where a continuous trunk line is not feasible, two trunk lines
could be used, each potentially leading to a separate fan. Some
of the individual 4-in. suction pipes would connect to one of
the trunk lines, some would connect to the other.

Sometimes, two fans will be installed in stand-alone BWD
installations, as discussed in Section 7.4. A second fan may be
installed when there is a second trunk line in an individual-pipe
installation, as discussed in the preceding paragraph.

A second fan may be installed in a baseboard duct installation
when a second suction pipe is  installed into the baseboard
ducting loop on the opposite side of the basement from the
first. A second suction pipe and fan would usually be consid-
ered for baseboard systems primarily in cases where there is
                                                       219

-------
no sump adding a major SSD component to the baseboard
system.

When two fans are used, each fan and exhaust system would
be installed as an independent unit, with all of the consider-
ations discussed in Section 4.6 applying to each one.

Because of the large air flows drawn from the block walls by
Stand-alone BWD systems, the radon concentration in the
exhaust is usually much lower than that observed from SSD
systems. This could impact decisions on design of the exhaust
piping.

The suction pipes leading to the fans will be 6 in. diameter in
some cases, as discussed in Section  7.3. As indicated in
Section 4.6.2, where a pipe penetrates the band joist (which is
supported underneath by the foundation wall), at least 2 in. of
wood must always remain between the hole and both the top
and the bottom of the joist. Thus, the 6-in. suction pipe could
be installed through a 2- by 10-in. band joist, provided the
hole is cut exactly in the center of the joist. A 6-in. pipe could
not be installed through a 2- by 8-in. band joist. And a 6-in.
pipe could never be installed through floor joists which are
not supported underneath. For such unsupported members,
tlie rule is that the diameter of the hole must not be greater
than one-third the height of the member; a 6-in. hole would be
more than one-third the height of a typical 2- by 10-in. joist
(or even of a 2- by 12-in. joist).

7.7   Wall and Slab Sealing in
Conjunction with BWD Systems
Sealing of major block wall openings  can be crucial in the
application of BWD systems. The block waUs are inherently
leaky, and significant amounts of air will leak into the BWD
system  regardless of the sealing effort that is undertaken.
However, failure to close certain major opening (such as the
open voids in the top course of block) can result in  such
excessive air leakage that the ability to the BWD system to
treat the wall can be compromised.

It is also recommended that major slab openings be sealed. A
stand-alone BWD system can sometimes extend a weak suc-
tion  field beneath the slab. Failure to close a wide wall/floor
joint could prevent any suction from reaching the sub-slab
region. Since BWD systems will not reliably treat entry routes
toward the interior of the slab, the need to seal such interior
routes in conjunction with stand-alone BWD systems will be
increased. With major slab-related entry routes, such as un-
trapped floor drains, sealing (or trapping) these major open-
ings is generally good practice in any event, to reduce poten-
tially significant soil  gas entry if not aid in suction  field
extension. And finally, with  some baseboard duct installa-
tions, sealing of sumps is  an integral part of the installation.

7.7.1   Wall Openings

Open  voids in the top course of block. One of the  most
crucial openings in block walls to close in conjunction with
BWD is the open voids in the top course of block in the wall,
in cases where the wall is not constructed with a course of
solid cap block on top.
These top voids are more accessible for closure in some cases
than in others, depending upon the construction features of the
particular house. Inability to close the open top voids will not
always render BWD completely inapplicable, but it can seri-
ously detract from the performance of the system, and can
make indoor radon levels more subject to  increases when
weather conditions or the usage of depressurizing appliances
challenge the system. It may also increase the house heating/
cooling  penalty,  and the risk of back-drafting combustion
appliances in the basement. In addition, failure to close the top
voids enables the block wall void network to continue to serve
as a chimney for soil  gas flow up into the house, at those
locations where the BWD  system is unable to adequately
depressurize the cavities.

Figures 37A and 37B illustrate two approaches for closing the
top voids, depending upon accessibility.

Where the dimensions of the block and of the sill plate are
such that one to several inches of the open void are exposed,
as in Figure 37A, there is sufficient accessibility such that the
top voids can be sealed from above with  mortar or with
expandable closed-cell, one- or two-component urethane foam.
To prevent the mortar or foam from dropping down into the
block cavities, some suitable support can be forced down into
each individual void, if necessary, leaving a depth of about 2
in. for the mortar or foam which is then applied above the
support. Where codes prevent crumpled newspaper from be-
ing used as such a support,  insulation material could be one
suitable alternative. A rapidly expanding, quick-setting foam
might be applicable without any support underneath.

If several inches  of the void are exposed, mortar can be
considered to close the top voids. Mortar may have a lower
materials cost  than  foam,  but will be  much more
time-consuming to apply. When mortar is used, it is crucial
that the mortar be forced all the way to the far face of the void
under the sill plate, to ensure that the entire void is closed. See
Figure 37A.

When only an inch or two of the top void is exposed, there is
not sufficient room to accommodate a person's hand into the
void. Mortar is thus no longer an option, because it could not
be reliably spread to the far face of the  void under the sill
plate. In these cases, expandable foams must be used.

Expandable foams are obtained in 12- to 24-oz. aerosol spray
cans (for small jobs), or in compressed cylinders up to 16 Ib.
(likely required when an entire course of blocks is to be
sealed). Foams can be applied directly from the can, or can be
injected through a hand-held wand or nozzle connected to the
container with a hose. The wand or nozzle can be inserted into
the small opening accessible under the sill plate, injecting the
foam in a manner which will effectively seal the entire void.
The foam should initially be directed toward the far face of the
void, again to ensure that the entire void is sealed.

This sealing effort must be  applied to every void in the top
course of block.

In the majority of cases, the dimensions of the block and sill
plate will be such that less than an inch of the top void will be
                                                       220

-------
                 Siding

                 Sheathing

                 Wallboard
                                Floor Joist

                                Band Joist

                                Sill Plate

                                Concrete Block

                                Foam or Mortar
                                to Close Void

                                Support: for Foam/
                                Mortar if Needed

                                Top Void
           A)  Closure of top void
               when void is reasonably
               accessible.
                              Sill Plate

                              Caulk or Foam

                              Top Void
            B) One option for closure
               of top void when a
               fraction of an inch of
               the void is exposed.
                         Brick Veneer
                             Sheathing
Foam to
Close Veneer
Gap
 Floor Joist

Drilled Access
Hole Through
Band Joist
Closure Plate

  Caulk
                                                                                               Sill Plate
         C) One option for closing gap
            between exterior brick veneer
            and interior block and sheathing.
Figure 37. Some options for closing major wall openings at the top of block foundation walls, in conjunction with BWD systems.

                                                     221

-------
exposed, if any of the void is exposed at all. In such cases, it
will be impractical or impossible to force support material
down into the cavities. Without such supporting material, it
could be more difficult to get foam to effectively close the top
voids, since large amounts of foam might drop down into the
cavities below. Where none of the void is exposed, there is an
additional complication, in that the cavity is not accessible
other than by drilling through the block face or sill plate.

Where at least a fraction of an inch of the void is exposed, a
foam nozzle could be inserted into the  void. If a rapidly
expanding, quick-setting foam is used, it may successfully
seal the entire top void, despite the lack of any support
underneath. Another option might be to drill a hole into each
cavity, through the face of the block just below the top of the
wall; some type of support for the foam could then be inserted
through this hole.

If this is  unsuccessful, efforts  could focus on sealing the
fraction-of-an-rneh opening  between the sill plate and the
foundation wall, as illustrated in Figure 37B. This approach
seals that portion of the top void that is accessible, and relies
upon the sill plate to serve  as at least a  partial cap on the
portion which is inaccessible. Expanding foam applied on top
of the wall, beside the sill plate, is one material that might be
used to seal this gap. Or, a small strip of wood might be sealed
against the sill plate and the foundation wall using beads of
polyurethane caulk.

Where none of the top void is visible, foam could be injected
into each top cavity through a hole drilled in the face of the
block. If the foam did not expand rapidly enough to avoid
dropping into the cavities below, one suggestion has been to
drill a second hole below each injection hole, then inserting
some type of support through the lower hole. Alternatively,
the seam between the sill plate and the foundation wall might
be caulked with gun-grade polyurethane caulk, in an effort to
improve the ability of the sill plate to serve as a cap for the top
voids.

Reliance on the sill plate as a cap for the top voids, in the
manner indicated in the preceding  two paragraphs, is less
effective than would be successful injection of foam into the
block cavities. The inaccessible outside seam between the sill
plate and the  block is left uncaulked, and the sill plate would
thus be expected to be a leaky cap. However, use  of the sill
plate saves a lot of time and expense. It has seemed  to be
sufficient in a few of the EPA study houses in Pennsylvania
where stand-alone BWD systems were tested (He87, Sc88);
however, in other cases, two fans were required and/or radon
levels were not reliably reduced below 4 pCi/L.

Fireplace structures. Fireplace structures incorporated into
block walls  offer the potential for large and inaccessible
openings within the structure. For example, there may be a
substantial gap between the actual fireplace insert (and the
flue) inside the structure, and the blocks which form the outer
walls of the structure. Such a gap could extend to  above the
basement, serving as a thermal and air-flow bypass to the
upper levels of the house. Thus, even when the top voids in
the remainder of the block foundation wall itself can be
scaled, large amounts of air from the basement, the upstairs,
and outdoors could leak into the BWD system via this inac-
cessible opening.

There is no practical way to seal such openings within fire-
place structures. The fireplace/chimney structure would have
to be torn down and rebuilt. Where such a structure exists,
additional BWD suction pipes near the structure, and perhaps
increased fan capacity, might be considered to compensate for
the expected leakage. Supplementing the BWD pipes with
SSD should also  be considered. The EPA  study houses in
Pennsylvania in which stand-alone BWD systems gave the
poorest performance generally had fireplace structures (He87,
Sc88).

Gap associated with brick veneer. In houses having exte-
rior brick veneer, a gap occurs between the veneer and the
sheathing and block behind the veneer. This gap is depicted in
Figure 37C. This gap could sometimes serve as an important
source of air  leakage into the BWD system, with outside air
and house air being drawn down through that gap (e.g., from
the eaves where the veneer ends,  or from house air leakage
through the sheathing on the floors above).

In an effort to seal this gap, it would be necessary to drill into
the gap at intervals (probably through  the band joist, as
illustrated in  Figure 37C), and to inject closed-cell urethane
foam through the drilled holes via  a hose/nozzle from  a
compressed cylinder.

It is not clear from available data under what conditions it will
be cost-effective to try to seal the veneer gap in this manner.
Reasonably good BWD performance has sometimes been
achieved even in brick-veneer houses  where  this gap was not
closed.

Major and intermediate holes through the block wall.
Any major openings through the block wall must be sealed.
Such major openings would include, for example, partially
missing blocks in cases where there is some major penetration
has been installed through  the wall.  Intermediate openings
should also be sealed. These would include, e.g., modest gaps
around utility line penetrations, chinks  in  the  blocks, and
places where  significant amounts of mortar have fallen out of
mortar joints.

Block pores. While the pores in the blocks are small, they
cover the entire face of the wall, and hence can add up to a
substantial  leakage area. However,  they are difficult and
expensive to seal, requiring  that the entire face of the wall be
painted or  coated. Tests have  shown that  the porosity of
blocks can vary by an order of magnitude,  depending upon
how they were manufactured (Ru91). A variety of coatings
can reduce this porosity by 95% or more, including, e.g.,
epoxy  paints and cementitious block filler. With  standard
latex paint, three  coats were found to be needed in order to
reduce air-flow through the  wall by more than 95% (Ru91).

If the basement were unfinished and the walls  thus readily
accessible, the cost of painting the walls in a typical basement
was estimated to cost about $400-$ 1,000, depending upon the
coating used, including labor and materials. This would in-
crease the cost of a BWD installation by about 20-70%. If part
                                                        222

-------
or all of the basement were finished, and if finish thus had to
be removed and replaced as part of the coating effort, the cost
could be significantly higher.

It is not clear under what conditions it is cost-effective to try
to seal the block pores, in view of the expense and potential
complications involved in coating the basement walls. In the
EPA testing of stand-alone BWD systems in Pennsylvania, it
was not apparent that sealing of the block pores was necessary
in order to get  adequate BWD performance, although no
definitive tests were carried  out in  order to quantify  the
benefits that such sealing would have provided. In only one of
the EPA study houses, which had very porous cinder block
walls,  were the  block  pores  sealed (by coating the  entire
interior face of the walls with a waterproofing paint).

7.7.2 Slab Openings

Perimeter channel drains. With stand-alone BWD sys-
tems, proper closure of perimeter channel drains can be even
more important  than it is with SSD systems (see Section
4.7.1). BWD systems will sometimes extend only a relatively
weak depressurization beneath the slab. A wide gap between
the slab and  wall will  permit so much air leakage into the
system at that point that any extension of the wall suction field
into the sub-slab region will probably be eliminated.

With the individual-pipe BWD approach, the perimeter chan-
nel drain should be closed as discussed in Section 4.7.1.

With the baseboard duct BWD approach, which is the BWD
approach most likely to be selected when a perimeter channel
drain is present, the perimeter drain will be sealed by enclo-
sure within the baseboard ducting. Any sump associated with
the perimeter channel drain will also be capped as part of the
BWD  installation, as discussed  in  Section 7.5.2 (see  Base-
board  ducting).  The perimeter channel drain would not be
caulked or mortared shut at locations where it was going to be
covered by the baseboard ducting, since the open channel
beneath the baseboard could add a significant SSD component
to the system.

If a portion of the perimeter channel drain cannot be covered
with baseboard ducting in a baseboard duct BWD installation,
judgement will have to be utilized in determining how to seal
this uncovered segment. Often, if it  cannot be covered by the
baseboard ducting, it may be too inaccessible for sealing. As a
minimum,  the opening in the slab will have to be sealed
immediately under the end of the ducting, so that basement air
will not flow through this channel into the ducting. Uncovered
sections of the perimeter channel drain could be sealed using
the techniques discussed in Section  4.7.1.

Wall/floor joint (when not a perimeter channel drain).
Where the wall/floor joint is covered by baseboard ducting, it
should always be left uncaulked, to increase the possible SSD
component that the baseboard duct BWD system might pro-
vide. However, uncovered sections  of the  wall/floor joint
should be caulked as discussed in Section 4.7.5, whenever it is
wider than a hairline crack. This might be particularly impor-
tant with stand-alone individual-pipe BWD systems, to help
the wall suction extend beneath the slab.

Other major and intermediate holes through the slab.
Other major holes through the slab, such as openings around
bathtub plumbing, sections of missing slab, and other miscel-
laneous  major openings should be sealed, as discussed in
Section 4.7.2, to aid in the distribution of the BWD suction
field to pouits under the slab remote from the walls. Likewise,
intermediate  slab openings,  such as expansion joints and
small slab holes (in addition to the wall/floor joint, discussed
previously), should be sealed whenever accessible, as dis-
cussed in Section 4.7.5.

Untrapped floor drains. Untrapped floor drains should be
addressed as discussed in Section 4.7.3. Trapping these drains
may aid in distribution of the weak  BWD suction field be-
neath the slab, and will prevent continued soil gas entry into
the house through the drain.

Sumps. Where an existing sump with drain tiles is present in
a basement, a  sump/DTD system will often be the ASD
approach selected for that house. Where a stand-alone BWD
system is selected for a house with a sump, it will commonly
be because there is a  perimeter channel  drain draining to
sub-slab drain tiles which empty into the sump. In such cases,
the baseboard duct BWD approach will often be selected, and
the sump will be enclosed as part of the BWD installation, as
discussed in Section 7.5.2.

If for any reason a sump exists that is not being incorporated
into the BWD system,  that sump should be capped with an
air-tight cover.

7.8    Gauges/Alarms and Labelling

The considerations discussed in Section 4.8, concerning gauges/
alarms and labelling for SSD systems,  also apply to BWD
systems.

Because of the high air flows from the block walls, a pressure
gauge on  a  stand-alone BWD system can register  lower
suctions than gauges on SSD or DTD systems. However, if
the gauge is located near the fan, and if there is a significant
piping run  associated with the BWD system, the gauge may
still provide a reasonably high reading. Due to potentially
high suction  losses common  in  the piping for stand-alone
BWD systems, resulting from the high flows, the fan will
have to maintain a fairly high suction in the piping immedi-
ately beside the fan in order to maintain a relatively low
suction in individual pipes near their penetration into the wall,
or in the baseboard ducting.
                                                       223

-------

-------
                                            Section 8
                       Design  and  Installation of Active
               Sub-Membrane Depressurization Systems
Where the house is built over an earthen- or gravel-floored
crawl space, and where the crawl space is largely or entirely
accessible, SMD is the applicable variation of the ASD tech-
nology. In effect, a "slab" of polyethylene sheeting is being
laid over the exposed soil floor, and the region beneath this
"slab" is being depressurized, analogous to SSD or DTD. See
Figures 6 and 7.

Depressurization beneath a membrane over the crawl-space
floor appears to be generally more effective at reducing radon
concentrations in the living  area than is forced ventilation/
depressurization of the entire crawl space, and distinctly more
effective than are natural ventilation and forced ventilation/
pressurization of the crawl space (He92). Ventilation of the
entire crawl space  is less expensive to install  than is SMD,
although the forced ventilation approaches probably involve a
greater heating/cooling cost penalty, since more treated air is
likely to be drawn or forced out of the living area.

Crawl-space ventilation approaches are apparently used  in
about one-third of the commercial installations in crawl-space
houses, rather than SMD; sealing and barriers were used  in
another one-third (Ho91). It is suspected that the large fraction
of ventilation and sealing installations are most likely associ-
ated with:  a) houses having adjoining basements, where a
SSD or DTD system in the basement is providing most of the
reductions, and where the crawl-space ventilation or sealing
component is providing only a limited additional benefit; or b)
low indoor radon levels, where crawl-space ventilation can
prove sufficient.

Ventilation approaches are particularly applicable in cases
where the crawl space is inaccessible. In these cases, SMD is
not an option, since reasonably good access to the crawl space
is required in order to install the membrane.

Less mitigation research,  and fewer commercial mitigation
installations, have  been completed in crawl-space houses,
relative to basements and slabs on grade. Thus, less definitive
guidance is possible for mitigation systems in crawl-space
houses.

The discussion in this section addresses only SMD in crawl
spaces having bare soil floors, or floors having gravel or a
plastic vapor barrier over soil. Where the crawl space has a
poured concrete slab or an unfinished concrete "wash" floor,
it will not be necessary to install a polyethylene membrane
over the floor. SSD beneath  the existing concrete will be the
appropriate ASD technology, rather than SMD. The one ex-
ception might be crawl spaces with badly cracked concrete
wash floors, in which case it may still be desirable to lay a
membrane. To design and install SSD systems in such cases,
see Section 4.

Where there is a basement or slab-on-grade wing adjoining
the crawl space, that adjoining wing will often have  to be
treated with  SSD or DTD, in addition  to (or instead of)
crawl-space SMD treatment. For such treatment of the adjoin-
ing slabs, see Sections 4  through 6. Occasionally, where
crawl-spaces have block foundation walls, it might be neces-
sary to supplement the SMD system with individual BWD
pipes. For individual-pipe BWD supplements, see Section 7.

The discussion in this section draws heavily from the detailed
review of available data on SMD systems, presented in Sec-
tion 2.3.5.

8.1     Selection of the Approach for
Distributing Suction,  and  the
Number of Suction Pipes

8.1.1  Approach for Distributing Suction

Two approaches can be considered for distributing suction
beneath the membrane of SMD systems.

   -   The mitigator can choose to insert one or more indi-
      vidual suction pipes through the membrane, in which
      case the SMD system would be analogous to SSD. This
      approach is referred to here as the "individual-pipe/
      SMD" approach. See Figure 6.

   -   Or, the mitigator can choose to install  a length (or
      perhaps a matrix) of perforated piping beneath the
      membrane to aid in the distribution of the suction. In
      this  case, the  SMD system would be analogous to
      DTD. This approach is referred to as the "sub-membrane
      pipinglSMD" approach. See Figure 7.

Another option that has been  considered, analogous to the
sub-membrane piping/SMD approach, has been to lay a strip
of porous matting beneath the membrane. The strip of matting
might be viewed as analogous to a length of perforated piping.
                                                    225

-------
 Some mitigators with significant experience using SMD sys-
 tems routinely use sub-membrane perforated piping (An92,
 K192). Others always use the individual-pipe approach, ex-
 cept perhaps in large or irregular crawl spaces (Bro92, How92,
 Sh92).

 Potential advantages of sub-membrane piping. Use of
 sub-membrane perforated piping should aid in the distribution
 Of the suction field beneath the membrane. Perforated piping
 would thus likely be most helpful when the crawl-space floor
 is a packed, impermeable soil with no imported gravel on top,
 suggesting marginal or poor sub-membrane communication.
 On the other hand, it would seem that individual suction pipes
 (with no sub-membrane perforated piping) should most likely
 be adequate in cases where the crawl-space floor either  con-
 sists  of native gravel soil, or is covered with a layer of
 imported gravel, suggesting good sub-membrane communica-
 tion.

 An additional potential  benefit  of sub-membrane piping is
 that, by distributing the suction over a wider area, it may be
 more forgiving in cases where ruptures  occur in the mem-
 brane. If a rupture occurs near an individual suction pipe, or a
 seam becomes unsealed, air short-circuiting into the indi-
 vidual pipe could  dramatically reduce the already-limited
 suction field extension created by that pipe, degrading SMD
 performance. But if such a rupture occurs near a portion of
 sub-membrane perforated piping, the degradation would intu-
 itively be expected to be less severe.

 Guidance regarding the need for sub-membrane pip-
 ing. Individual-pipe/SMD seems to be consistently effective
 in reducing living-area concentrations below 4 pCi/L in crawl
 spaces up to about 1,500 to 2,000 ft2. IMs has been the  case
 even when: a) the crawl-space floor is a fairly tight soil, and
 sub-membrane communication would thus appear to be rela-
 tively poor; and b) the membrane has not been fully sealed.

 The use of perforated piping beneath the membrane is most
 likely to be warranted when reductions  below 2 pCUL are
 needed, or when the crawl space is larger than 1 £00 to 2,000
ft*. Sub-membrane piping may be beneficial even in smaller
 houses, increasing radon reductions when sub-membrane com-
 munication is limited. Sub-membrane piping is least likely to
 be needed when there is  a layer of gravel  on the floor.

 Careful sealing of the membrane may increase the crawl space
 size that can conveniently be treated by individual-pipe/SMD
 systems, and improve their performance. In addition, multiple
 individual suction  pipes may be a suitable substitute for
 perforated piping under the membrane.

 To date, there has been no definitive direct comparison of
 JndivJdual-pipe and sub-membrane piping SMD systems
 back-to-back in a single house. Thus, it is not possible to
 quantify  the benefits of using sub-membrane piping, or to
 specify more precisely when it will be needed.

 Results  to date.  Many SMD  installations have used the
 individual-pipe approach. The large  majority  of  the
 individual-pipe installations have been in houses where the
 crawl-space floor appeared to be a packed, impermeable  soil,
with no gravel. But despite the implicit poor submembrane
communication and the failure to use perforated piping, all but
one of the EPA study houses having individual suction pipes
and having crawl-space floor areas of 1,500 ft2 and less have
been reduced below 4 pCi/L in the living area with only one
suction pipe. Only the two houses having floor areas of 2,000
ft2 and greater (and no gravel) required a second suction pipe
to get below 4 pCi/L, or were not reduced below that level,
despite  complete  sealing of the membrane in  one of the
houses.  This result  would indicate  that the  individual-pipe
approach  can  work reasonably well in  moderately sized
crawl-space houses,  even when the sub-membrane communi-
cation would appear to be marginal to poor.

On the other hand, none of the "pure" crawl-space houses in
EPA's study—i.e., none of the houses having only a crawl
space with no adjoining wing  of some other substructure
type—were reduced below 2 pCi/L with individual-pipe/SMD
installations. Among the study nouses, only those having an
adjoining  basement with an operating SSD system in the
basement were reduced below 2 pCi/L when an individual-pipe
SMD system was installed in  the  crawl-space wing. The
basement  SSD system was believed to be, or  was demon-
strated to be, largely responsible for the observed reductions.

By comparison, a  number of  "pure" crawl  spaces  with
sub-membrane perforated piping have been reduced below 2
pCi/L. This fact suggests that perhaps the  use of perforated
piping (or additional individual suction pipes through the
membrane) might have helped performance in the houses with
individual-pipe systems. However, no definitive studies have
been conducted to confirm this suggestion or to quantify any
benefits that the perforated piping might have provided.

None of the individual-pipe installations in pure crawl-space
houses involved complete sealing of the membrane among the
EPA study houses cited above. Complete sealing of the mem-
brane might have helped some of those systems achieve levels
below 2 pCi/L. Mitigators report achieving  levels below 2
pCi/L with fully  sealed individual-pipe systems (How92,
Sh92).

In contrast to individual-pipe/SMD  systems, sub-membrane
piping/SMD systems have consistently reduced living-area
concentrations below 2 pCi/L,  even in houses  as  large as
1,500 to 2,700 ft2.  This result has led to the conclusion stated
earlier, that the use  of perforated piping beneath the mem-
brane is most likely  to be required when reductions below 2
pCi/L are needed, or when the crawl space is larger than 1,500
to 2,000 ft2. Careful  sealing of the membrane may reduce the
number of individual suction pipes required in large crawl
spaces,  and/or improve the performance of individual-pipe
systems relative to sub-membrane piping systems.

Cost of sub-membrane piping. There is a modest installa-
tion cost penalty associated with the perforated piping. The
piping has a materials cost of about $0.30 to $0.40 per linear
foot, which  would  translate to  perhaps $0.45  to $0.60/ft
installed. Thus, installation of perforated piping beneath the
membrane will add roughly $10 to $30 to the installation cost
if the piping is laid as a single length down the center of the
                                                       226

-------
crawl space, and roughly $40 to $100 if the piping were laid in
a loop around the perimeter.

These costs apply if the piping is simply laid on top of the
floor beneath the membrane, or is buried in loose gravel. If it
is felt to be necessary to dig a trench for the piping in packed
soil, so that the piping does not  create such a bulge in the
membrane and is less subject to damage from foot traffic, the
increase in installation cost caused by  the sub-membrane
piping would be greater.

8.1.2  Number of Suction Pipes

Individual-pipe/SMD approach. As discussed in Section
8.1.1, one individual suction pipe through the membrane at a
central location has generally  been  sufficient to reduce
living-area concentrations below 4 pCi/L in crawl  spaces
having floor areas smaller than 1,500 to 2,000 ft2. This has
been true even in cases where sub-membrane communication
appeared to be marginal to poor, and even in cases where the
membrane has not been fully sealed.

While one  pipe has consistently  seemed to be adequate in
moderately sized houses, the data base is fairly limited. The
statement above is based on data from only 12 EPA-sponsored
individual-pipe installations in houses  of 1,500 ft2 and less
with no gravel on the floor, although the results are generally
confirmed by  mitigators  who use the individual-pipe ap-
proach (How92,  Sh92). The mitigator  might be prepared to
install one or more additional suction pipes in individual pipe/
SMD installations in crawl spaces with  no gravel, even when
the crawl space is smaller than 1,500 ft2, whenever the floor
area approaches that size.

In the EPA study houses, post-mitigation levels achieved in
the living area were 2 to 4 pCi/L, and pre-mitigation concen-
trations in the living area were almost always less than 20 to
30 pCi/L. If the goal were to achieve less than 2 pCi/L, or if
the pre-mitigation levels were particularly high, more suction
pipes through the membrane could be required.

The individual-pipe/SMD systems in the EPA study houses
often did not have the membranes fully sealed. As discussed
in Section 8.1.1 (see Results to date),  careful sealing of the
membrane  may reduce the number of pipes  required for a
crawl space of a given size. Sealing may also improve perfor-
mance, enabling the individual-pipe system to achieve indoor
levels below 2 pCi/L.

Where there is native or imported gravel on the crawl-space
floor, improving sub-membrane communication, it is far more
likely that one individual suction pipe will  provide better
reductions in large crawl spaces. Individual-pipe/SMD sys-
tems with sub-membrane gravel are analogous to SSD sys-
tems in basements having good sub-slab communication.
Where there has been imported gravel, single suction pipes
have consistently reduced living-area levels below 2 pCi/L,
although this particular set of conditions has been tested in
only two houses having small crawl spaces (about 300 ft2)
adjoining basement wings that were also being treated with
sump/DTD  (Mes90a).  Thus,  the individual-pipe/SMD ap-
proach has not received a fair test in crawl-space houses with
gravel floors, and it is impossible to provide definitive guid-
ance regarding how large a crawl space might be treated with
a single suction pipe when gravel is present

The  individual-pipe/SMD approach has been tested in two
"pure" crawl-space nouses having crawl-space floors of 2,000
ft2 and larger and having  no gravel on the floor to improve
communication. In neither of these houses  was one suction
pipe sufficient to achieve living-area levels below 4 pCi/L
(Py90).  In one of these houses (DW29, with a floor area of
2,000 ft2 and with the membrane unsealed), two suction pipes
were required to reduce indoor levels to 3 pCi/L (from a
pre-mitigation level of 16 pCi/L). In the second house (DW27,
with a floor area of 2,300 ft2 and with the membrane fully
sealed), two pipes were insufficient to reduce  indoor levels
below 5 pCi/L*(from a pre-mitigation level of 33 pCi/L).

Thus, when the individual-pipe/SMD approach is applied to
houses larger than 1,500 to 2,000 ft2 with no gravel, at least
two suction pipes might be needed. Sometimes more than two
pipes might be needed. Careful sealing of the membrane
might reduce the number of individual  pipes required, or
improve the radon reductions. At this time, more definitive
guidance regarding the number of pipes in these cases is not
possible.

Where it is desired to install an individual-pipe system in a
large house with no gravel, the mitigator  might begin by
installing one or two suction  pipes.  Any  additional  pipes
might then be added, guided by suction field measurements or
smoke visualization tests beneath the membrane, if the initial
installation did not provide adequate radon reductions. Alter-
natively, a submembrane piping/SMD system  might bs in-
stalled at the outset.

In the two study houses  having  the highest pre-mitigation
levels (88 and 160 pCi/L), individual BWD pipes also had to
be added  to  supplement  the individual-pipe/SMD  system,
treating the block foundation  walls.  This  result could be
suggesting that in cases where the source term is as high as it
likely was in these houses, the weak depressurization created
beneath the membrane by individual-pipe/SMD systems may
not adequately prevent radon entry into the block cavities.

Sometimes a crawl space will be divided into sections, with a
footing  and even a load-bearing wall separating the sections.
In this case, there would be a  separate membrane covering
each section, and each  section can be viewed  as a separate
crawl space.  Each section would require at least one indi-
vidual suction pipe, in accordance with the preceding discus-
sion.

Sub-membrane piping/SMD approach. Where a matrix
of perforated piping is laid beneath the membrane, only one
vertical PVC suction pipe is required. This single suction pipe
can penetrate the membrane and connect to the sub-membrane
piping at any convenient location.

More than one suction pipe would be required only in cases
where the  sub-membrane piping  consists  of  two or more
isolated segments, in which case a suction pipe would be
required for each segment. The sub-membrane piping will not
                                                       227

-------
commonly be present in isolated segments, except in the case
where the crawl space is divided into sections by footings and
foundation walls. In that case, each section of the crawl space
would require its own membrane and its own sub-membrane
piping system.

8.2    Selection of Suction Pipe
Location
8.2.1  Individual-Pipe/SMD Approach

Where a single individual suction pipe penetrates the mem-
brane, the penetration is usually at some central location in the
crawl space. Where two individual suction pipes penetrate the
membrane, one penetration is usually made at a central loca-
tion in each half of the floor area. Where there are more than
two pipes,  the penetrations would  logically be distributed
uniformly around the crawl space, unless sub-membrane suc-
tion field extension measurements with the first couple pipes
operating would suggest some other distribution.

Individual suction pipes should not  penetrate the membrane
immediately beside any unsealed seams in the membrane. To
be conservative, the pipes should not be located immediately
beside sealed seams. Air short-circuiting into a pipe through a
nearby seam that is not sealed, or that inadvertently is not
fully sealed, could dramatically reduce the extension  field
extension from that pipe. Such leakage would further reduce
sub-membrane depressurizations, which will usually be weak
even in the absence of unsealed seams. Accordingly, the
individual pipes  should penetrate the membrane near the
center of one of the membrane  sheets,  away from seams
between adjoining sheets and away from seams against perim-
eter foundation walls and any interior obstructions  (such as
support piers).

Where the foundation walls are hollow block, air can leak into
tire sub-membrane region through the blocks. Accordingly,
unless a specific objective is to treat the block  walls, indi-
vidual SMD pipes should generally not be immediately beside
the perimeter wall. If a suction pipe is ever placed near a wall,
the membrane must be effectively sealed to the wall, regard-
less of whether the wall is block, poured concrete, or other
material (see Section 8.7).

Where the crawl space is divided into isolated sections by a
footing and/or foundation wall, the  above criteria would be
used to locate an individual suction pipe for each section.

8.2.2  Sub-Membrane  Piping/SMD
Approach
Vertical suction pipe. The PVC suction pipe which pen-
etrates  the  membrane and connects to the sub-membrane
perforated piping can tap into the perforated piping at any
point along its length. Thus, the suction pipe can penetrate the
membrane at any convenient location over the submembrane
piping-

This location could be selected to simplify the routing of the
piping and the exhaust stack. For example,  the suction pipe
could be located near the point where the piping will penetrate
the foundation wall to an exterior stack, or at the point where
the stack can conveniently extend up through the house for an
interior stack.

Where the sub-membrane piping is not in a continuous loop, it
would be logical to locate the suction pipe toward the middle
of the perforated piping, if this is  convenient. Especially
where corrugated flexible drain tile is used as the perforated
piping, and where flow friction inside the perforated piping
can thus be relatively high, application of the suction near the
middle of the piping should help achieve a more even suction
field distribution beneath the membrane.

If there is a section of the membrane that could not be fully
sealed or is otherwise expected to be fairly leaky, the suction
pipe should tap into the perforated piping remote from that
section, if possible, to reduce the leakage of crawl-space air
into the system.

Sub-membrane perforated piping. The perforated piping
beneath the  membrane can also be considered as  suction
piping, helping to distribute the suction beneath the mem-
brane. This perforated piping is laid on or trenched into the
crawl-space floor in some suitable matrix under the mem-
brane. Different matrices have been used in different installa-
tions, with no definitive study to define which configuration
might be optimal under different circumstances.

Mitigators who use sub-membrane piping commonly install a
single, straight length of piping down the center of the crawl
space, parallel to the  long dimension of the house (An92,
Fit92, K192). This simple configuration generally appears to
perform well. More complicated patterns are usually consid-
ered only when: a) the crawl space is divided into isolated
segments, in  which case a length of perforated piping is
required in each segment; or b) the crawl space has a compli-
cated floor plan, such as an L shape, in which case the length
of perforated piping should bend and extend down each leg of
theL.

When two adjoining crawl spaces are separated from one
another by a footing, one length of sub-membrane piping will
sometimes still be used to treat both segments (An92, K192).
In these cases, the perforated piping beneath the membrane in
one crawl space will come up through the membrane, snake
over the footing, and then penetrate beneath the membrane in
the second crawl space. In these cases, the perforated piping
might take the shape of a single long length (if the two crawl
spaces are back-to-back), or a U (if the two crawl spaces are
side by side). Where this approach is used for treating adjoin-
ing crawl spaces, the short section of flexible corrugated
piping that snakes over the footing above the membranes, and
joins  the sections of perforated sub-membrane piping in the
two crawl spaces, must not be perforated. Also, its penetration
through the membranes must be well sealed.

In some  research installations,  where the crawl space was
particularly wide, two or more parallel lengths of piping have
been  used, each off center, rather than just one down the
center. These parallel lengths were then joined by a perpen-
dicular segment near the center. Multiple lengths were used in
this manner, so that no point on the floor is more than 6 to 15
                                                      228

-------
ft from a segment of piping, in an effort to ensure adequate
sub-membrane depressurization everywhere. (See Section
2.3.5b, Operation of central furnace fan, and Section 2.3.5c,
Number of suction pipes.) This research study did not define
whether the multiple lengths were in fact required, or how
wide the crawl space would have to  be  warrant using more
than one length.

If the suction field extended no more  than 15  ft in each
direction from a central length of perforated piping,  this
would suggest that any crawl space wider than 30 ft would
require more than one length of piping. However, as discussed
in Section 2.3.5c, suctions too low to be measured can extend
more than 15 ft from the suction point, even when the floor is
an impermeable soil (Br92). Thus, especially if the membrane
is well sealed, it  may be possible to treat crawl spaces wider
than 30 ft with one length of piping.

In the most extensive case, one could envision a central length
of piping, running  from one end of  the crawl space to the
other, with perpendicular legs extending out at intervals in an
effort to ensure that no portion of the floor is more than a few
feet from a segment of perforated piping.

In some research installations (Sc88,  Fi90, Gi90), the perfo-
rated piping has  formed a loop around the perimeter of the
crawl space, beside the foundation walls. In addition to treat-
ing the submembrane  region, this configuration  should pro-
vide the most effective treatment of the foundation walls.
Where the walls are hollow block, they might be an important
soil gas entry route  into the crawl space and the living area. In
addition, the soil near  the foundation wall may be a particu-
larly important radon  source, due to pressure differentials
across the wall created by wind effects.

Location of the SMD suction immediately beside the founda-
tion walls is somewhat analogous to the recommendation that
SSD suction pipes  be  located near to the walls in basement
houses where sub-slab communication is marginal. It is also
analogous to the guidance that depressurization of complete
interior drain tile loops should often effectively treat base-
ments having marginal sub-slab communication.

When the perforated piping is laid immediately beside the
foundation wall, it will be important that the membrane be
attached to  the wall with  an effective and permanent seal.
Otherwise, significant air leakage through the membrane/wall
seam could overwhelm the SMD system, preventing suction
from distributing beneath  the membrane or into the block
wall.

One possible concern with locating the piping around the
perimeter is that, in wide crawl spaces, the central portion of
the membrane might be left inadequately depressurized. Ef-
fective submembrane suction has been found in limited test-
ing to extend perhaps no more than 6 to 15 ft from the suction
pipe (see  Section 2.3.5b, Operation of central furnace fan);
central points may sometimes be more  than 15 ft from the
perimeter. However, as discussed in Section 2.3.5c (see Num-
ber of suction pipes),  suctions below the measurement limit
can sometimes extend  more than 15 ft. Also, as with SSD and
DTD systems in basements, radon entry through the central
portion of the slab/membrane may be of less concern, so long
as the membrane is providing a reasonable barrier at the
central locations.

In summary, in many cases where sub-membrane piping is
being used, a single straight length of piping down the middle
of the crawl space will be sufficient. However, in some cases,
more extensive matrices may be needed or preferred. Since
there are not definitive data comparing alternative configura-
tions under different conditions, the particular configuration
for a given installation will have to be selected by the indi-
vidual mitigator, based upon experience and upon the consid-
erations discussed above. One other consideration in selecting
the specific submembrane piping matrix is the desire to avoid
obstructions, an to be out of the way of foot traffic in the crawl
space, if possible.

8.3   Selection of Suction  Pipe Type
and Diameter

8.3.1  Individual-Pipe/SMD Approach

As with the other ASD variations,  the appropriate type of
suction piping will be rigid Schedule 40 or lightweight PVC,
PE, or ABS piping with compatible fittings.

The  total flows in individual-pipe/SMD  systems should de-
pend upon whether the membrane is completely sealed, whether
there is gravel on the floor,, and the number of suction pipes.

Limited results reported for individual-pipe systems where
the membrane was sealed, there was no gravel, and there was
only one suction pipe—i.e., for the  expected lowest-flow
case—suggest  flows in  the range of 20  to 40 cfm with the
90-watt in-line tubular  fans (Os89a,  Mes90a, How92). In
some cases, these flows represent the flows from the SMD leg
of a combined SSD+SMD system in houses where a basement
adjoins the crawl space. At these flow rates, the system piping
would have to be no larger than 4 in. diameter to avoid undue
suction losses in the piping. Segments of 3-in. piping might
not create unacceptable suction losses at these flows, if needed
to facilitate an aesthetic installation. Refer to Figure 13 and
Section 4.6.1.

Where the membrane is not completely sealed, flows  from
individual-pipe systems might be expected to be higher. Data
from two houses with the membrane not fully sealed, with no
gravel, and with only one suction pipe gave flows of 40 to 90
cfm  (Py90). Such flows would increase the need to using 4-in.
piping, avoiding sections of 3-in. piping. Since this case also
reflects the situation that could develop with the fully sealed
membrane  discussed in the preceding  paragraph, if  seals
failed or punctures occurred in the membrane over time.

Thus, the mitigator may be well served to consistently use
4-in. piping in individual-pipe/SMD systems whenever pos-
sible, to account for possible higher flows in any particular
installation. This could be desirable even when the membrane
is sealed, there is no gravel, and there is only one suction pipe.
Because the crawl-space piping will be out of sight, the use of
the larger piping should generally not create any particular
                                                      229

-------
difficulties, at least not for that portion of the piping which is
inside the crawl space.
reducing friction loss. However, flow noise in the pipe could
start to become of concern with 4-in. piping at these flows.
Where there is gravel on the crawl-space floor, the flows from
individual-pipe/SMD systems would be expected to be some-
what higher, since the grave! would facilitate the flow of air
from a larger area. In two relatively small crawl spaces (300
ft2) with gravel floors, with the membrane completely sealed
and with only one suction pipe, flows were 30 to 100 cfm with
the 90-watt fans (Mes90a). The higher flows could have
resulted from the possibility that the membrane seams in fact
might not have been sealed as well as was thought. Although
the two houses cited here had poured concrete foundation
walls, higher flows  might also be expected when gravel is
present in houses having block foundations, since the gravel
could facilitate air being drawn into the system through the
block walls.

Thus, where there is gravel on the floor, 4-in. diameter piping
should generally always be used to reduce the suction loss.
Again, flows could be increased if the membrane were not
completely  sealed (or if seals  broke), if more suction pipes
were used,  or if the crawl space were larger, further under-
scoring the justification for using piping of at least 4 in.

Use of multiple suction pipes can also increase system flows.
The highest flows reported for an individual-pipe/SMD sys-
tem—over 100 cfm—were reported for one house having two
suction pipes, without the membrane being fully sealed, but
with no gravel (Py90).  In this case, the second pipe added
about 20 cfm to the total system flow, compared to the case
where this same system was operated with  only one pipe. At
flows this high, use of 4-in.  piping  throughout would be
important.

8.3.2  Sub-Membrane Piping/SMD
Approach

Vertical suction pipe. The vertical suction pipe which taps
into the sub-membrane perforated piping and leads to the fan
should be rigid lightweight or Schedule 40 PVC, PE, or ABS,
as in the individual-pipe approach.

In houses having a single length of perforated piping down the
center of the crawl space, with the membrane completely
sealed and with no gravel on the floor, typical  system flows
appear to be 40 to 80 cfm with the 90-watt tubular fan (Bo91,
Fit92, K192). As would be expected, these flows are higher
than those from individual-pipe/SMD systems when there is
no gravel on the floor. At these flows, the piping should
generally be  no smaller than  4 in. diameter, based on the
friction losses in Figure 13 and the discussion in Section 4.6.1.

Failure to seal the membrane would be expected to increase
flows. In one house where two parallel lengths of perforated
piping were installed, where the membrane was not sealed,
and where there was no gravel on the floor, flows were 130
cfm (Fi90). At flows this high, the piping should be no smaller
than 4 in. diameter. As discussed in Section 5.3,  in connection
with sump/DTD systems, larger (6-in. diameter) piping is
probably not warranted at these flows from the standpoint of
Placing the perforated piping in a loop around the perimeter
might also be expected to  increase flows, compared to a
central straight length. In one house where the piping was
around the perimeter, with the membrane completely sealed
and with no gravel on the floor, flows were 113 cfm (Sc88).
Again, 4-in. piping would be appropriate.

Gravel on the crawl-space floor would be expected to increase
flows.  Flows have been  reported for three  houses having
gravel  on the floor  (Fi90).  All  three represent high-flow
conditions, in that, in addition to the presence of gravel, the
sub-membrane piping was laid in a more extensive matrix,
usually as a loop around the perimeter. In addition, the foun-
dation walls were hollow block, increasing the possibility of
air flow from this source. The membranes were always sealed
around the perimeter, but were not always sealed at interior
seams. Under these conditions, total flows were 187 to over
200 cfm with a 90-watt fan at full capacity. These flows are
the maximum that can be expected with the 90-watt fans,
given flow restrictions  created by the piping. The highest
flows among these three were observed in  the  one house
having a complete liner that was completely sealed, with the
piping at an interior location (not a perimeter loop); this set of
conditions would be expected to provide the least leakage of
crawl-space air into the system.

With the roughly 200 cfm flows observed in the systems on
gravel floors, the friction loss figures from Figure 13 would
suggest that 6-in. piping  could sometimes be  desirable, to
reduce both friction loss and flow  noise.  At that high flow in
4-in. pipe, a large fraction of the fan suction capacity could be
consumed by virtue of the suction  loss in the pipe (about 2 to
3 in. WG total friction loss per 100 ft of pipe), thus providing
reduced suction beneath the membrane. However, in practice,
use of 4-in. piping when there is gravel on the floor may
simply make the fan operate at a different point on its perfor-
mance curve, reducing flows and sub-membrane depressur-
izations. When there is gravel on the floor, reduced depressur-
izations may not represent a serious problem, especially if the
membrane is well sealed.

In practice, the three sub-membrane piping/SMD systems
with gravel floors utilized 4-in. PVC suction piping (Fi90).
All of  them performed very well, reducing living-area con-
centrations below 2 pCi/L, usually from pre-mitigation levels
of 15 pCi/L and higher, when the 90-watt fans were operated
at full capacity. The good results with the 4-in. suction piping
in the gravel-floored crawl spaces could be due to the fact that
the piping run was short, perhaps only 10 to 20 ft long, so that
the total loss  in the PVC piping may have been  only about
0.25 to 0.5 in. WG—a ioss readily handled by the 90-watt
fans. Another contributing explanation  could  be that with
such good sub-membrane communication and with the perfo-
rated piping to further aid in distributing the suction, little
suction is  required for such a SMD  installation  to perform
well. Careful sealing of the membrane would be particularly
important in gravel-floor cases.
                                                      230

-------
Sub-membrane  perforated  piping.   Often,  the
sub-membrane perforated piping is the same 3- to 4-in. diam-
eter flexible black corrugated piping commonly used for drain
tiles in the particular geographical area. This flexible piping is
simple to get into the crawl space, and is simple to lay. It is
particularly convenient when the matrix will involve bends,
such as when the pipe is laid in a loop around the perimeter.

Another type of piping that has been used has been rigid 4-in.
diameter perforated PVC, PE, or ABS piping. To resist crush-
ing, Schedule 40 piping might be preferred in this application.
This rigid piping is most convenient when the matrix consists
of a straight length laid down the center of the crawl space,
but, with the use of elbows and other fittings, the rigid pipe
can be formed into any matrix configuration desired.

Both the flexible corrugated piping and rigid piping should be
fairly resistant to being permanently crushed by the weight of
a person on hands and knees, even if the piping is not trenched
into the crawl-space floor. The lightweight rigid piping would
be the most subject to  damage, and might warrant being
trenched into the floor if it is near a foot traffic route. Any of
the piping could be subject to damage from heavy activities in
the crawl space, e.g., from service personnel installing a new
furnace.

Piping larger than 4 in. diameter has never been used for
sub-membrane perforated piping in SMD systems. This is true
despite the fact that the rough sides of the flexible corrugated
piping result in a  friction loss about  1.8 times  that in
smooth-walled pipe. When a straight length of piping is used,
when the membrane is fully sealed, and when there is no
gravel on the floor, good performance has been achieved
without larger piping. In these cases, flows are generally low
enough (40 to 80 cfm), the length of piping short enough, and
the system sufficiently forgiving, such that larger piping is not
needed.

In the more extreme case, where the piping forms a loop
around the perimeter of a gravel-floored crawl space, flows in
the perforated piping can be on the order of 200 cfm at some
locations, and the piping can involve over 100 linear feet. In
these cases, a significant suction loss could result through this
amount of 4-in. piping. However, the gravel which is contrib-
uting to these high  flows is also providing  excellent
sub-membrane communication which makes high suctions
unnecessary, so long as the membrane is adequately sealed.
The results show that, under these circumstances, very good
radon reductions can be achieved despite the presumably high
suction loss in the 4-in. piping. Moreover, the flows will not
be as high as 200 cfm at  all locations in the piping.

8.4    Selection  of the Suction Fan

Individual-pipe!'SMD  systems.  With  either  the
individual-pipe or the sub-membrane piping SMD approaches,
the air flows involved and the suctions desired will generally
dictate that the fan be at least equivalent to the 90-watt in-line
tubular fans  listed in Table 1 and discussed in Section 4.4.1.
These fans can move up to 270 cfm at zero static pressure,
although they are practically limited to about 180 to 200 cfm
when connected into 4-in. piping. Most of the SMD installa-
tions discussed previously have utilized such fans.

With the individual-pipe approach, the flows generated by the
90-watt fans were 20 to 40 cfm when there was no gravel on
the floor (with suctions of over 1.5 in. WG in the piping), and
30 to  100 cfm when there was gravel (with suctions of 1.0 to
1.5 in. WG).

Especially for the no-gravel case, the smaller 50-watt in-line
tubular fans could handle the flows involved. In  fact, they
would generate lower flows. In  theory, these lower flows
should be more than sufficient, considering the flows that
should nominally be developed under the membrane based
upon  fundamental principles. And, in fact, a few mitigators
report occasionally having reasonable  success with these
smaller fans in individual-pipe/SMD systems when the mem-
brane is well sealed and when communication beneath the
membrane is fairly good (Sh92).

However, in practice, the smaller fans are often not a good
choice. The smaller fans would develop considerably lower
suctions than would the 90-watt fans (probably 0.25 to 0.75
in. WG, rather than 1.0 to 1.5+ in. WG). This reduced suction
might not be sufficient to distribute suction beneath the mem-
brane when sub-membrane communication is poor. Also, in
practice, the  smaller fans could  have trouble handling the
higher flows that could develop over time, as ruptures occur
or seals break.

Thus,' in view of the limited experience with SMD systems,
and in view of the potential problems that could arise over the
long term from use of the smaller fans, it is recommended that
the 90-watt in-line fans (or equivalent) be routinely used for
individual-pipe SMD systems until data become available
demonstrating when smaller fans might be sufficient.

Where the  membrane is fully sealed, and especially where
there  is no gravel, the in-line radial blowers discussed in
Section 4.4.2 could sometimes be  a reasonable choice  for
individual-pipe systems. At the flows involved- 20 to 40 cfm
with no gravel, sometimes  well below  100 cfm even with
gravel— the radial blowers would develop distinctly higher
suctions than would the tubular fans (up to 2 to 3.5 in. WG,
compared to 1.5 in. WG). These higher suctions could aid in
distributing the suction field under the membrane.

The high-suction/low-flow fans discussed in Section 4.4.3 are
generally not a reasonable choice for SMD systems. The very
high  suctions are not required, and the flows that can be
developed (especially if membrane leaks develop over time)
will be too  high for some of the high-suction fan models.

Sub-membrane  piping/SMD systems.  With  the
sub-membrane piping/SMD approach, flows are sufficiently
high,  even when no gravel is present, it is even more apparent
that the 90-watt tubular fans will generally be the best choice.
Flows may be too high for the 50-watt tubular fans. And, since
flows are consistently above 100 cfm with the 90-watt fans,
there  would not seem to be an incentive to use a radial blower
since, at these flows, the suctions developed by the blowers
are comparable to those developed by the tubular fans.
                                                       231

-------
The flows of 100 to 200 cfm observed in these SMD systems
with 90-watt fans are too high for 50-watt tubular fans. In
practice, of course, use of a 50-watt fan in such cases would
result in lower flows and lower membrane suctions. When
gravel is present on the floor and submembrane communica-
tion is thus very good, these lower flows and suctions may in
fact be sufficient. Some mitigators have occasionally used the
smaller fans with sub-membrane piping/SMD systems in small
crawl spaces with fully sealed membranes, even when there is
no gravel (An92).

However,  experience is not sufficient to confirm that the
50-watt  fans will in  fact often be adequate. Moreover, as
ruptures in the membrane and the membrane seals occur over
time, the flows may increase to the point that the small fans
will become overwhelmed and no longer effective. Accord-
ingly, it is recommended that the 90-watt tubular fans always
be used in submembrane piping/SMD systems, until an expe-
rience based is developed demonstrating when the smaller
fans are acceptable.

With  the 200+ cfm flows observed in the sub-membrane
piping systems with gravel floors, one might even consider
using one of the larger, 100-watt fans in such installations.

However,  the sub-membrane piping  systems in the several
test houses having gravel floors were extremely effective with
the 90-watt fans despite the high flows, reducing living-area
concentrations  below 1  to 2 pCi/L. Thus, even at the high
flows, the 90-watt fans appear to be consistently distributing
adequate suction through the perforated piping to effectively
treat the house. Accordingly, the use of 100-watt fans in such
cases  would not appear to be warranted unless indoor levels
below 1 pCi/L were required.

8.5    Installation of the Membrane
and the Suction Pipes

8.5.1  Installation of the Membrane

Selection of membrane material In almost all reported
SMD installations, the membrane has consisted of polyethyl-
ene sheeting. Such sheeting is relatively easy to obtain in rolls
of appropriate dimensions (10 to 20 ft wide by perhaps 120 ft
long), and is relatively convenient to work with.

Regular polyethylene sheeting as thin as 6 mil has sometimes
been  used, although  regular 8- to 10-mil  sheeting would
provide better puncture resistance. Rather than regular poly-
ethylene,  an increasing number of mitigators are  using
high-density   cross-laminated   polyethylene.   The
cross-laminated material has  dramatically  superior tensile
strength and puncture resistance. Unlike the regular polyeth-
ylene sheeting, which can  be torn  by hand even with a
thickness of 10 mil, the high-density cross-laminated material
cannot be torn by hand, even though its thickness may be only
perhaps 4 mil. Due to  its significantly increased puncture
resistance, the cross-laminated polyethylene is recommended
despite its higher cost.

The polyethylene sheeting that is used must be stabilized to
resist UV radiation.
To provide additional protection against damage to the mem-
brane, the polyethylene should be overlain by heavier material
along expected traffic routes.  Various materials have been
used for this purpose, including roofing felt, EPDM rubber-
ized roofing membrane, and drainage mat.

Special precautions may be necessary in particular cases.
Such cases could include walk-in crawl spaces, where there
may be foot traffic over the entire crawl space floor, or crawl
spaces with very irregular floors (e.g., with sharp protruding
rocks). In such cases, special precautions could include the
use of even thicker cross-laminated material, or installation of
heavier material underneath the polyethylene,  between the
sheeting and the crawl-space floor.

In some cases, the crawl space will akeady have a plastic
vapor barrier such as Visqueen laid over some portion of the
floor, for moisture purposes. Where this plastic is in good
condition, mitigators sometimes straighten this existing plas-
tic and incorporate it into the SMD membrane. However, it is
recommended that the existing moisture control membrane be
replaced with UV-resistant cross-laminated polyethylene sheet-
ing for the SMD system.

Identification of area to  be  covered by the membrane.
Where there is sufficient  headroom throughout  the crawl
space, the entire  floor should always be covered  by the
membrane. Complete coverage  should provide the best oppor-
tunity for depressurizing the soil everywhere, intercepting the
soil gas before it can enter the  crawl space. Complete cover-
age should also reduce the leakage of crawl-space air into the
system. This is particularly true when there is gravel on the
floor, since the gravel  will facilitate air flow  through the
exposed gravel into the SMD system.  If air leakage were
sufficiently high due to incomplete coverage, this could not
only reduce system performance, but also increase the heat-
ing/cooling penalty (if treated air is drawn down from the
living area  overhead), and  perhaps  also  contribute  to
back-drafting of combustion appliances in the crawl space.

Sometimes, a portion of a crawl space might be very difficult
to access. For example, the crawl-space floor might be sloped
such that one portion has less  than 1 ft of headroom. Or, a
major obstruction such as a furnace and associated forced-air
ducting could prevent coverage of the floor directly under the
furnace.

As discussed in Section 2.3.5c (see Extent of membrane), very
limited results  to date suggest that a SMD system can some-
times still be effective even when some limited portion of the
floor is  left  uncovered. On the other hand, one mitigator
(K192) reports  that failure to provide complete coverage can
sometimes reduce SMD performance so significantly that it is
necessary either: a) to excavate the inaccessible part of the
crawl space (to provide the  headroom needed to enable com-
plete coverage); or b) to supplement the SMD  system  with
some other technique (such as a heat recovery ventilator in the
house).

In summary, sufficient experience has not yet been obtained
to definitively  demonstrate under what conditions different
amounts of floor can be left uncovered, and what the perfor-
                                                      232

-------
mance, operating cost, or other penalties might be for such
incomplete coverage. With the current data, the guidance is
that the mitigator should make a best effort to cover the entire
crawl-space floor. However, inability to cover the entire floor
should not lead the mitigator to conclude that SMD is not an
option for that house, especially in cases where the segment of
inaccessible floor is relatively limited and when there is not
gravel on the floor.

Laying the membrane.  Installation of a membrane will
require that workers spend some time  in the crawl space.
Crawl spaces can contain molds and spores  which can be
irritating and unhealthful. Workers should be provided with
appropriate respiratory protection.

The crawl-space floor must be cleared of any movable objects
that are being stored in the crawl space.  If possible, any
irregularities in the crawl space floor (such as embedded
rocks) should be cleared from the expected traffic paths (e.g.,
the path between the crawl-space entrance door and the fur-
nace or water heater), so that the membrane will be less likely
to be  ruptured by subsequent traffic.  If there is an existing
vapor barrier which is to be replaced by the SMD membrane,
the old plastic sheeting might be removed.  Some  mitigators
leave the old vapor barrier in place, and simply install the new
membrane on top of it (An92).

The 10- to 20-ft wide sheeting should be rolled out over the
floor in a flat,  smooth manner, with care by the installers to
avoid ruptures in the membrane  during installation. In some
cases, rather than bringing the entire roll of polyethylene into
the crawl space, it may  be more convenient to pre-cut the
sheeting to the desired  lengths outside, and bringing the
pre-cut sheets into the crawl space.

Adjoining sheets should be overlapped by  at least 12 in. at
seams between the sheets. Where the sheets meet the perim-
eter foundation walls, they should be cut leaving about 12 in.
excess sheeting, so that the sheeting can be extended a short
distance up the walls. This will be necessary so that the sheet
can be attached to the walls as recommended. Even if perim-
eter sealing is not initially planned, it is advisable to leave the
excess sheeting anyway, to permit subsequent perimeter seal-
ing if it is later found to be desired.

Where the sheeting comes up against an interior obstruction
such as a pier,  it should be trimmed to fit around the obstruc-
tion, again leaving perhaps 12 in. excess to enable subsequent
attachment to the sides of the obstruction. The sheets should
be laid and trimmed to fit relatively snugly around the ob-
struction. The optimum width of sheeting to use (10 vs. 20 ft)
will depend upon the dimensions, accessibility, and obstruc-
tions in a given crawl space.

Where there is a furnace or water heater in the crawl space, the
membrane should not contact the hot part of the unit. If the
unit rests upon a concrete pad, the membrane might be sealed
against the pad. If a furnace rests, e.g., on hollow blocks, the
furnace can be temporarily jacked up so that the blocks can be
moved  and the  membrane can be extended uninterrupted
under the furnace; the blocks would then be replaced.
At particularly rough or sharp points on the floor, drainage
mat, roofing material, or some other heavy material might be
placed beneath the membrane, to try to protect it from punc-
ture. Similar heavy materials should be placed on top of the
membrane as walkways along expected traffic routes, for the
same reason.

Sealing the membrane—general The available data dem-
onstrate that radon  levels in the living area can often be
reduced below  4 pCi/L even  when the membrane is not
completely sealed at the seams between the sheets, at the
perimeter wall, and around interior obstructions. However,
complete sealing is  recommended whenever possible, for a
variety of reasons:

   -  Increased indoor  radon  reductions expected with a
      given SMD system, compared to those likely with the
      same system if the membrane were not sealed;

   -  Fewer suction pipes required for individual-pipe/SMD
      systems;

   -  Reduced heating/cooling penalty in the house, due to
      less treated house air drawn down into the crawl space
      from the living area;

   -  Reduced contribution to possible backdrafting of com-
      bustion appliances in the crawl space or in an adjoining
      basement; and

   -  Increased likelihood that the membrane sheets  will
      remain in place, and will not get shifted due to  foot
      traffic over the years, opening gaps in the cover.

A disadvantage of sealing is that it will add to the installation
time and cost.  Intuitively, sealing will be  most critical for
seams which are near individual suction  pipes, or seams
which run parallel  to and immediately beside lengths of
sub-membrane perforated piping.

There are currently no definitive data demonstrating the im-
pacts  on performance if the membrane is not sealed under
various circumstances. However, the limited available  data
confirm that better performance is achieved when the mem-
brane is sealed (see Section 2.3.5c, Degree of membrane
sealing, and the discussion of individual-pipe/SMD systems
in Section 8.1). Moreover, the other  potential benefits of
sealing, listed above,  appear compelling. As a result,  it is
recommended that the membrane be completely sealed when-
ever possible.

Most mitigators who  have substantial experience installing
SMD systems indicate that they routinely seal the membrane
completely (An92, Bro92, Fit92, How92, K192, Sh92).

Where cross-laminated polyethylene has been  used as the
membrane,  gun-grade polyurethane caulk  is an  extremely
effective sealant. Where regular polyethylene sheeting has
been used, polyurethane caulk does not provide a good bond,
and some other sealant must be used.
                                                       233

-------
Sealing the membrane—seams between sheets. Seams
between adjoining sheets of membrane are usually sealed by
applying one or two continuous beads of sealant between the
sheets, in the 12-in. strip where the sheets overlap. As indi-
cated above, polyurethane caulk is a good sealant choice for
the recommended cross-laminated polyethylenes; various other
sealants have been used for regular polyethylenes.

It can be difficult to achieve an air-tight seal. To help ensure
that  the sealed seam is air-tight, some mitigators apply one
very liberal bead, using one tube of gun-grade polyurethane
caulk per 15 linear feet (K192). Other mitigators apply two
parallel beads of sealant.

After the sealant is applied, the two sheets of polyethylene are
pressed together. Some mitigators then tape the seams using
duct tape, to hold the sheets together while the sealant cures
(K192). If the tape is not used, the sheets of plastic can shift as
the workers complete  the  installation, causing  the seal to
break before the sealant cures.

Even if the mitigator decides not  to seal all of the seams
between  sheets throughout the crawl space, it is strongly
recommended that the seams be sealed for the sheet(s) through
which an individual SMD suction pipe penetrates, as well as
any seams which parallel a sub-membrane perforated pipe.
High leakage rates through unsealed seams near the suction
point could seriously reduce the extension of the suction field
to remote locations.

Sealing the membrane—perimeter foundation walls.
To seal the membrane against the perimeter foundation wall
and any  interior piers, one common approach is to simply
adhere the membrane to the wall with a bead of sealant This
is the approach illustrated  in  Figures 6 and 7, and is  the
simplest method for sealing the membrane against the wall.
There would be no other structural support for the membrane
against the wall, other than  this bead of sealant. Based upon
field experience, a bead of polyurethane caulk can bond
cross-laminated membranes to the foundation wall very se-
curely and durably, if the wall surface is reasonably flat and is
properly prepared. Some mitigators report such seals remain-
ing intact in SMD systems for at least five years (K192).

By tills approach, the  perimeter wall should first be wire
brushed at the height at which the caulk bead is to be applied
(i.e., 6 to 12-in. above the crawl-space floor), to remove any
dirt or other loose deposits.

The roughly 12 in. of excess sheeting that was left beside the
foundation wall would then be folded at a 90° angle, at the
joint between the foundation and the crawl-space floor, and
would be extended straight up the wall. That is, the membrane
would remain flat on the crawl-space floor right up to  the
wall, at which point it would then proceed straight upward,
basically flat against the wall. See Figures 6 and 7. Some slack
should be left in the vertical section of plastic rising up the
wall; this will help prevent stresses from being applied to that
seam by foot traffic in the crawl space, or by the SMD system
itself when the system fan is first turned on and the membrane
is drawn down against the soil (K192).
The membrane should remain flat against the floor right up to
the wall, rather than beginning to rise above the floor at some
distance from the wall. If some portion of the membrane is
being held suspended above the floor by the bead of caulk,
then any  weight applied to the suspended membrane by
persons or animals in the crawl space might cause the mem-
brane to rip away from the wall at that location, or might cause
the membrane itself to tear, destroying the seal.

A liberal bead of polyurethane caulk (or other sealant) would
be applied on the wall near the upper end of the membrane,
and the membrane would be pressed tightly and smoothly
against this bead. Some mitigators  report using as much as
one tube of gun-grade polyurethane caulk per 10 linear feet,
then taping the membrane against the wall using duct tape, to
hold the plastic against the wall  while  the  sealant cures
(K192).

Where there are obstructions in the wall which extend to
within 6 to 12 in. of the floor (e.g., a crawl-space access door
or a foundation vent), the membrane will need to be trimmed
to pass beneath the obstruction. The caulk bead would like-
wise extend underneath the obstruction. Where the membrane
turns a corner around the perimeter wall or around a pier, the
membrane would need to be cut and tucked appropriately to
permit a reasonably neat, sealed corner.

Current experience suggests that a bead of sealant will attach
the membrane to a flat  (block or poured concrete) perimeter
wall with a sufficiently secure and durable bond, without any
further steps to anchor the membrane against the wall. How-
ever, under some circumstances, mitigators find it desirable to
attach the membrane to a wall using a 1- by 2-in. or 1- by 4-in.
wooden furring strip nailed into the  wall.

One circumstance where a furring strip can be necessary is
when the foundation wall is irregular, e.g., when the wall is
fieldstone. In this case, one approach would be to  firmly
attach a furring strip to the wall using masonry nails. The end
of the  membrane would be caulked onto this strip, and a
second furring strip  nailed into the first, sandwiching  the
membrane between the two. The gaps between the first fur-
ring strip and the irregular wall would then be sealed using
expanding foam or caulk, as applicable (K192).

Another case where a furring strip  can be used is when the
membrane is extending all the way up the face of the wall, and
is being attached to the band joist above the wall. This will
sometimes be done when it is desired to try to increase the
treatment of a block wall, as discussed later. It might also be
an option  when the wall is irregular (e.g., fieldstone), as an
alternative to attaching  the furring strip directly to the field-
stone. One approach that has been used in this case would be
to caulk or cement the  end of the membrane onto the band
joist around the entire perimeter. The furring strip would then
be placed on top of the caulked membrane and nailed into the
band joist, sandwiching the membrane between  the furring
strip and the band joist.

Most mitigators who use furring strips simply  sandwich the
end of the membrane between two pieces of wood, as with the
approaches described above. A more rigorous approach would
                                                       234

-------
be to wrap the end of the membrane around a furring strip,
then nailing this strip to the wall or to another piece of wood
with a layer of caulk in between. However, this more elabo-
rate and time-consuming approach is probably not necessary
when the membrane has been installed in a manner which will
reduce the risk of the membrane being ripped  out of the
sandwich: that is, when the membrane rises flush against the
wall with some  slack, as  discussed above, and especially
when the membrane is cross-laminated material.

In some early SMD installations, furring strips were used to
attach the ends of the membrane directly to block or poured
concrete foundation walls, along with a bead of caulk between
the furring strip  and the wall. This was done  for increased
security, rather than  relying solely on the bead of caulk or
other sealant to anchor the membrane against the wall. The
use of furring strips in this manner, when the wall is relatively
flat, does not appear to be widely practiced at the present time,
due to the success that has been demonstrated using caulk
alone (as  discussed  above) and due  to  the cost and time
penalties involved in using a furring strip. Hie extreme case
(where the membrane is wrapped around a furring  strip) could
add about $350 to the installation cost, relative to the simplest
case (where the  membrane is attached with a caulk bead),
varying with the size of the crawl space (He91b, He91c).

Seating the membrane—configurations to treat block
foundation walls. In some cases where the foundation walls
are hollow block, it is believed that the block wall may be an
important radon  entry route into the crawl space and, from
there, into the house. In these cases, one option would be to
install an  individual BWD  suction pipe into the block walls,
analogous to the approach  shown in Figure 34 for basement
SSD systems. However, some mitigators instead  modify the
installation of the membrane  in an effort to increase the
treatment  of the block wall by the SMD system.

Several options for increasing  block  wall treatment by the
SMD system are illustrated in Figure 38.

Some mitigators try to increase the effective BWD component
by drilling a few holes through the face of the blocks below
the point  where the membrane  is sealed against the wall, as
shown in Figure 38A (K192). This should increase the flow of
air from the block cavities into the SMD system, thus increas-
ing suction in the walls, although the suction field will be
fairly weak beneath  the membrane at that point unless the
suction  is being distributed by a perimeter loop of
sub-membrane perforated piping. Another approach, which
may more positively treat the entire face of the wall and the
top voids, would be to extend the membrane all the way up the
face of the wall, as shown in Figure 38B (Sh92). In this figure,
the membrane is attached  to the band joist by being sand-
wiched between the joist and a furring strip, in  the manner
discussed previously. One  difficulty in this approach is that
there will always be some obstructions at some point along
the wall—e.g., the access door and foundation vents—which
the membrane will have to be contoured around.

Figure 38C illustrates an option used by one mitigator (An92)
for distributing the suction around the perimeter,  in an effort
to increase the treatment of the  walls. This option, analogous
to the baseboard duct BWD approach, would be an alternative
to installing a perimeter loop of sub-membrane piping as a
means  for achieving strong  suction at the perimeter. This
option is discussed further later.

Sealing the membrane—interior obstructions. Where
there are interior support piers, the membrane can be sealed
around the piers in a manner similar to that discussed previ-
ously for sealing against the perimeter foundation wall.

Where there is a concrete pad in the crawl space supporting,
e.g., a  furnace or a water heater, the membrane should be
sealed to the edges of the pad with  a bead of caulk or other
sealant.

Role of the SMD  membrane as a radon barrier. It is
underscored that the polyethylene being installed as described
above is intended as an element of a SMD system. That is, it is
serving as  a plastic "slab"  which will  enable  potentially
effective depressurization of the crawl-space soil while reduc-
ing (but not eliminating) the amount of crawl-space air being
exhausted by the SMD system.

Without operation of the SMD  fan, this membrane should
reduce (but not prevent) convective (pressure-driven) flow of
soil gas up into  the crawl space. Without the SMD fan to
create depressurization beneath the membrane, the membrane
cannot be relied upon by itself to effectively block convective
flow up from the soil.  Despite efforts to carefully seal  the
membrane, there will undoubtedly be many leakage points.

In addition, polyethylene—although a very effective convec-
tive barrier, if it were air-tight—is only a moderate barrier to
radon diffusion. As a result, radon trapped beneath the mem-
brane could continue to diffuse into  the crawl space.

Thus, installation of the membrane itself, as a stand-alone
"sealing" step without the SMD component, would likely be
of limited effectiveness, even if it is carefully sealed at all
seams. Limited experience to date  using crawl-space floor
barriers as a stand-alone mitigation approach  has shown in-
door radon reductions of only 0 to 30% (He92).

8.5.2 Installation of the Suction Pipes

Individual-pipe!SMD approach.  A hole is cut in the mem-
brane at the point where the suction pipe is to penetrate, of the
same diameter as the suction pipe. The suction pipe is inserted
vertically down through this hole, and the membrane is sealed
around the pipe in an air-tight manner.

The vertical suction pipe must be supported in some manner,
so that the open end of the pipe beneath the membrane is held
at least a couple inches above the crawl-space floor. If the
pipe were to drop so that the open end were resting directly on
the soil, air  flow into the pipe would be  severely restricted,
reducing the suction field distribution that could be main-
tained  beneath the membrane. Analogous to the case with
SSD systems (see Section 4.5.1, Mounting  suction pipes
through slab), the pipe can be supported by strapping or
hangers overhead, or it can be supported at the crawl-space
floor under the membrane.
                                                       235

-------
 Grade Level
 Outdoors
  Hollow-Block/;;;
  Foundation   "-'!'
  Wall  :.:.'.
Membrane Sealed
Against Wall with
Bead of Caulk or
Adhesive

Membrane

Hole Drilled
into Block Cavity
                                     Crawl
                                     Space
                                     Floor
        A) Holes drilled through inside face
           of block at intervals, below the
           point at which the membrane is
           sealed against the wall.
Plywood Sealed
Against Wall with
Urcthane Caulk — s
jpffsfe
£&'>$$
•,,.,•'•:•;•;!'--••)
•=-•=<«,".•';•)'
-;r;.;%f';;{
t;.«.:.v;'; V:
1 *» " * * *
' « ' B • I'",
'iK
)&%'•'*?:
.*'*'," }/ J''
"*'",' « V
•
* »*
" **
^
* *"
f »'
* **
" **
* »*
* **„
* *'.
*.
• v
^


t =•
>:
.
.
^
V
j:
— .

* , * .
4 A A ' A - •
**".*' * A * •* ' * " •* ' "
' * ' . A .. * . ^ • *
* * L * * «
5 * * ] *
A ' I A
4 ' * rf' * ' A
, * V • * **. '*
4 ' ' ' * '
* * • *
' • A '

1 	 Suction Pipe
- — Membrane Sealed
Against Plywood
or Wall
— Staple Attaching
Membrane to Plywood
^ /-Membrane
lt_
. L- Treated
• . Plywood
7-7- Sealant Where
- • Suction Pipe
•/ " . " • ". :. - v • . - ;•:.'-"• Penetrates
                                                           Furring
                                                           Strip

                                                      ••— Membrane
                                                                   B) Membrane extended upward to cover
                                                                      entire wall, sealed to band joist.
         C) Channel created around crawl-space
            perimeter using plywood; suction
            drawn on channel.
Rgura 38. Some alternative approaches for increasing block wall treatment by a SMD system.

                                                    236

-------
Several of the alternatives for installing the individual suction
pipe through the membrane are illustrated in Figure 39.

One common method for installing the individual suction pipe
to place an inverted T-fitting on the bottom end of the pipe
beneath the membrane, as illustrated in Figure  6 and in
Figures 39A and 39B. The top of the T rests on the soil floor,
supporting the pipe, and the two open ends are aimed parallel
to the floor. These three figures vary primarily according to
the steps taken to ensure that the membrane is not sucked into
the open ends of the T, blocking flow.

In Figure 6, a rigid, flat plastic or plywood plate is placed on
top  of the T, supporting the  membrane above the T for a
distance sufficient to prevent the membrane from  becoming
drawn back into  the open ends. Figure 39A is  essentially
identical to Figure 6, except in this case, the plastic "plate" is
in fact the base of plastic roof flashing units (K192), identical
to those used to flash around stack penetrations through roofs
(e.g., see Figure 23). The roof flashing is used in this applica-
tion because it is  simple and convenient. In Figure 39B, no
plate is used to specifically protect the ends of the T, but holes
are drilled through the fitting so that suction can also extend
through the sides (Br92).

Another option that has been reported has been to  install the
open end of the suction pipe in a frame of some type which
will keep the membrane away from the open  end. In Figure
39C, this "frame" is an inverted bucket with holes drilled
through its  sides (Sh92). Analogous frames have been fabri-
cated out of plywood (Mes90a). Since there is no T fitting to
support the weight of the piping in Figure 39C, the pipe would
have to be supported by strapping or hangers overhead, or by
a pipe clamp fitted tightly around the pipe and resting on the
top of the bucket  (analogous to the pipe clamp shown at the
overhead flooring  in Figure 17C).

The pipe installation approaches illustrated in Figures 6,39 A,
39B, and 39C effectively create a "pit", analogous  to the
sub-slab pits for SSD systems, by lifting the membrane above
the crawl-space floor around the pipe penetration. This "pit"
should accomplish one objective of a SSD pit, namely, reduc-
ing  suction  losses as the soil gas  accelerates (see Section
4.5.1, Excavating a pit beneath the  slab). But it will not
achieve the second objective of intercepting fissures or strata
beneath the soil surface. Figure  39D illustrates an approach
where a true pit is excavated in the crawl-space  floor, and
covered with treated plywood to maintain the cavity. This
approach has been considered  by  researchers (Py90) and
utilized by some mitigators (How92). In the specific configu-
ration shown in Figure 39D, the  weight of the piping is not
supported where it penetrates the plywood; hence, it must be
supported by hangers or strapping from the joists overhead.

As one variation of the specific configuration in Figure 39D, a
second piece of plywood is sometimes placed on  top of the
membrane  and screwed into the plywood underneath, sand-
wiching the membrane between the two pieces of plywood
(How92). In that case, the pipe penetration through the mem-
brane would be sealed using a bead of polyurethane caulk
around the perimeter of the top piece of plywood, between the
plywood and the  membrane, and by caulk around the seam
where the pipe penetrates the top piece of plywood. As a
second variation, the weight of the vertical pipe is sometimes
supported at the plywood, e.g., using a PVC flange mounted
in the hole through the plywood, analogous to the SSD pipe
configuration shown in Figure 15C (Py90, How92).

No testing has been conducted to assess any performance
benefits resulting from such a pit,  due to interception of
fissures or permeable strata.

When any of these configurations is installed, it is desirable to
keep the hole through the membrane as small as possible,
preferably no larger than the  pipe  diameter. This  should
minimize the potential for leaks developing through the mem-
brane over time near the suction point.

If one of the above configurations using a T fitting is used, the
T (and any sub-membrane plate) can be fit onto the bottom of
the suction pipe beneath the membrane after the  pipe is
inserted through the  membrane hole. Thus, the membrane
hole does not have to be enlarged to accommodate the T and
the plate. To do this, at least one of the membrane seams near
the penetration would have to be left unsealed until after the T
is installed, so that the fitting can be slid under the membrane
from the side. The hole through the membrane for the suction
pipe could be cut as  a series of slits radiating  out from the
center, leaving all of the pie-shaped pieces in place. This
could facilitate sealing the membrane against the pipe after
installation.

The perforated bucket in Figure 39C or the plywood frame
would logically be inserted under the membrane in a similar
manner. The sub-membrane sheet of plywood covering any
sub-membrane pit as in Figure 39D would be installed, with a
hole to accommodate the suction pipe, before the membrane
was laid.

Where a T fitting is being installed, some mitigators make a
slit in the membrane sufficiently large to accommodate the T
through the membrane from on top (An92). With the open end
of the T protruding up through the slit membrane, a square
piece of polyethylene sheeting  large enough  to cover the
slit—and with a hole through the middle just large enough to
accommodate the suction pipe—is sealed over the top with a
continuous bead of caulk or sealant around the perimeter,
sufficiently large to completely enclose the slit. Again, it is
particularly important that this slit be effectively sealed, since
leaks near the suction point could have a serious impact on the
distribution of the sub-membrane suction field.

Once the suction pipe has been connected to the sub-membrane
T fitting, bucket, frame,  or plywood sheet, the membrane
must be  effectively sealed around the  pipe, to  prevent air
leakage through the membrane at that location.

One possible approach for sealing the membrane against the
pipe is to use a hose clamp, as illustrated in Figures 6, 39B,
39C, and 39D. By this approach, a hose  clamp is  placed
around the vertical suction pipe above the membrane.  A
liberal, continuous bead of polyurethane caulk is placed around
the circumference of the suction pipe, probably about 6 in.
above the crawl-space floor. The membrane is then pressed
                                                       237

-------
   Suction
   Pipe
Plastic/Rubber
Roof Flashing
(Commonly Used
to Flash Around
Stack Penetrations
Through Roofs)
                                        Membrane

                                        T-Fitting
   Sealant

   Hose Clamp

   Perforations
   in T-Fitting

       Membrane
       ".  •   Crawl Space Floor
     A) Membrane sandwiched, sealed between
        two roof flashing units; base of
        flashing prevents membrane from
        being sucked into open ends of
        T-fitting
                              B) Perforations drilled in sides of
                                T-fitting, to prevent membrane
                                from blocking all air flow.
  Suction Pipe
  (Supported
  from Above) -
  Inverted Bucket
  (or Specially'
  Built Wooden
  Flange)
      Perforations
      in Side of
      Bucket
                                                         Note: A second piece of plywood is
                                                         sometimes placed on top of the
                                                         membrane and screwed into the bottom
                                                         piece, sandwiching the membrane.
Suction Pipe
(Supported
from Above)

Sealant

Hose Clamp

Membrane
                                                        Treated
                                                        Plywood
      C) Inverted perforated bucket or wooden
         ftange used to elevate membrane above
         floor and prevent membrane from being
         sucked into pipe inlet
                         D) Suction pit excavated in crawl-space floor.
Flguro 39, Some of the alternative approaches for installing individual SMD suction pipes through the membrane.

                                                   238

-------
tightly against the sides of the suction pipe, compressing the
caulk bead, and clamped into place with the hose clamp.

Where the hole has been cut in a manner which leaves a series
of pie-shaped membrane segments radiating out from  the
center of the hole, as discussed above, these segments would
be pressed against the caulked side of the pipe and would be
bound  by  the clamp. Any cuts that have been made in the
membrane around the hole to enable insertion of a T-fitting or
flange  should be effectively closed by being pressed against
the caulk and the side of the suction pipe by the hose clamp.

Where there is any doubt, additional caulk may be placed at
the seam between the hose clamp, the membrane, and the
suction pipe to ensure that this penetration is air-tight.  Air
leakage at this location could significantly decrease what will
already be a relatively weak sub-membrane suction field.

Instead of a hose clamp, Figure 39A sandwiches the mem-
brane between two roof flashing units. To seal this configura-
tion, a liberal, continuous bead of polyurethane caulk would
be placed around the entire base of the unit above the mem-
brane, between the flashing and the membrane,  as indicated
on the figure. Likewise, the seam between the upper unit and
the suction pipe would be liberally caulked.

Where the approach in  Figure 39D is supplemented by  a
second plywood sheet above the membrane, a continuous
bead of sealant is applied around the entire perimeter of the
upper plywood, between the plywood and the membrane
(How92). A liberal bead of sealant is also applied around the
seam between the upper sheet of plywood and the suction
pipe.

Except for the cases involving sheets of plywood covering  a
sub-membrane pit, the other configurations just discussed all
involve raising the membrane up several inches around the
suction pipe, as necessary to effectively clamp or otherwise
seal the membrane  around the pipe.  This  will  cause some
deformation of the membrane sheet through which the pipe is
penetrating. This deformation must not be sufficient to cause
crawl-space soil to become exposed at the seam between that
sheet and  the adjoining sheets. To avoid deformation of the
adjoining  sheets or undue stresses on the sealed seams, the
seams  between the penetrated sheet and the adjoining sheets
should not be sealed until after pipe installation has been
completed.

If a 12-in. overlap has been provided between sheets, this
should often be sufficient to prevent the soil from becoming
exposed at seams when the membrane is raised. However, to
be conservative, somewhat greater overlap might be provided
for the sheet through which the pipe will penetrate. Again to
be conservative, it would be desirable for the adjoining sheets
to overlap on top of sheet with the suction pipe, and for the
seams of those sheets to be sealed, even if the seams between
sheets are not being sealed elsewhere.

Figure 38C illustrates a unique suction pipe installation ap-
proach used by one mitigator (An92) in an effort to increase
the extension of the SMD suction to block foundation walls.
This approach, analogous to the baseboard duct BWD ap-
proach, is an alternative to installing a perimeter loop of
sub-membrane perforated piping. In this approach, a slanted
strip of treated plywood is sealed against the perimeter wall
with a bead of polyurethane caulk, creating a "baseboard
duct" around the crawl-space perimeter. The SMD membrane
extends up the plywood and is  sealed against the wall just
above the top of the plywood. The suction pipe is sealed into
this baseboard channel at some convenient point around the
perimeter. In addition to maximizing treatment of the wall,
this approach should maximize treatment of the floor immedi-
ately beside the wall; radon entry  will  likely  be highest
through this section of floor, due to wind effects against the
perimeter walls. However, because of the high flows likely to
come out of the walls,  it is possible that suction will not
extend well toward the center of the membrane.

Sub-membrane piping]SMD approach. The suction pipe
for sub-membrane piping/SMD systems consists of two ele-
ments: the perforated piping laid beneath the membrane; and
the vertical PVC suction pipe which penetrates the membrane
and connects to this perforated piping.

 •  Sub-membrane perforated piping. The perforated piping,
    which may be the flexible black corrugated piping, or
    rigid PVC piping, is laid on the crawl-space floor in the
    selected matrix, as discussed in Section 8.2.2.

    Commonly, where the floor is bare soil, the perforated
    piping will be laid on top of the soil. Where the piping is
    being laid in the path of potential future foot traffic, a
    trench might be  excavated. The trench could be filled
    with gravel, and the piping embedded in the gravel. Such
    trenching would  significantly increase the installation
    effort, and is not often done. One mitigator who has
    tested SMD systems back-to-back in the same  house,
    with the piping on top of the soil and also with the piping
    in a trench, reports that installation  of the piping in a
    trench appears to have  no  impact on radon reduction
    performance (K192).

    Where the crawl-space floor is  gravel, the perforated
    piping can be buried in the gravel relatively easily. How-
    ever, where the floor is gravel, sub-membrane piping will
    be less likely to be needed;  the individual-pipe/SMD
    approach will probably be adequate.

    Where flexible perforated piping is used in a contiguous
    pattern with no branches (such as a straight length or a
    fan or partial loop), it can easily be flexed into whatever
    pattern is desired. If a perpendicular branch were desired
     (to connect two parallel lengths of piping, or to  extend
    legs off the main trunk line), a flexible corrugated T-fitting
     could be snap-fitted into the length of flexible piping. If a
    rigid PVC T-fitting were used rather than a corrugated
     fitting, the perforated piping could be connected to the
     ends of the PVC T using screws and caulk, as discussed
     later in connection with installation of the vertical suction
     pipes.

    In many cases where there are  two parallel lengths of
     perforated piping, it will be because there are two adjoin-
     ing crawl spaces separated by a footing and/or wall. One
                                                       239

-------
 option is to connect these separate lengths using a length
 of non-perforated flexible corrugated piping that pen-
 etrates up through the membrane in one crawl space,
 snakes over the interior footing, and then descends down
 through the membrane in the second crawl space. De-
 pending on the configuration of the crawl spaces, each
 end of this non-perforated connector could connect into
 the middle of one of the parallel perforated lengths using
 a corrugated T-fitting (so that the entire matrix takes on
 the shape of an H). Alternatively,  the non-perforated
 connector might connect the ends of the perforated seg-
 ments, so that the matrix takes on a contiguous U shape.
 The penetration of the non-perforated piping through the
 membranes must be sealed well.

 Where there are parallel lengths of piping, another alter-
 native to linking these lengths beneath the membrane
 would be to install a vertical PVC suction pipe through
 the membrane into each one, and linking these risers by a
 piping network in the crawl space above the membrane.

 Where rigid perforated pipe is used for the sub-membrane
 piping, the sections of piping are joined by straight cou-
 plings, elbows, and T's as necessary to create the desired
 matrix. The fittings should be cemented onto the perfo-
 rated piping to ensure the structural integrity of the matrix
 over time.

 When corrugated or rigid fittings are inserted into the
 sub-membrane piping network, in  accordance with a
 number of the configurations just discussed, one question
 is how well the joints should be sealed. If there is some
 leakage at the joint, but if the joint is beneath a well-sealed
 membrane, it would seern that the leakage should not be a
 major problem. In assessing the need for joint sealing in a
 given installation, the factors that should be considered
 are the following:

-  How big is  the leak?  If the gap at the joint is large
   enough, the leakage at that point may become so severe
   that suction will not effectively extend to more remote
   points in the sub-membrane piping, and the system will
   take on  the characteristics of an individual-pipe/SMD
   system with an individual suction pipe penetrating the
   membrane at the leakage point.

-  How far is the leak from the vertical suction pipe? The
   nearer the leak is to the suction pipe, the greater the
   impact of the leak can be.

-  Does the leak reflect a structural weakness in the joint
   that may result in the segments of sub-membrane pip-
   ing  becoming  disconnected  over time?  The
   sub-membrane piping network may be subjected  to
   some physical stresses due to foot traffic in the crawl
   space. If the joint is sufficiently weak that the segments
   become disconnected over time, the remote segment of
   piping on the far side of the break could be cut off from
   the system.

 The joints between flexible corrugated piping and corru-
 gated snap-fittings are generally very tight, and may not
need to be cemented, screwed, or caulked together unless
the piping is expected to come under particular physical
stresses. In this case, the concern is usually not one of
leakage around the joint as  installed (as  long as  the
membrane is carefully sealed), but rather, is one of the
joint becoming separated over time. Where rigid PVC
fittings are being installed in a corrugated piping matrix,
this joint will be very subject  to becoming disconnected
unless it is firmly connected with screws or clamps and
caulk. Where rigid perforated piping is used for the entire
matrix, the joints should all be cemented, because this is
easy to do, and because the rigid piping will not flex like
the corrugated material and thus may be more prone to
having uncemented joints become disconnected as a re-
sult of physical stresses.

Where the perforated piping does not form a complete
loop, so that there are open ends, the open ends should be
capped. Capping the ends will help ensure that the suc-
tion is distributed evenly along the length of the piping. If
an end is left open, most of the air flow into  the piping
will likely enter through the open end. Under this condi-
tion, the sub-membrane piping/SMD system will tend to
take on the characteristics  of an individual-pipe/SMD
system with individual suction pipes at the locations of
the open ends.

Vertical suction pipes. A rigid non-perforated suction
pipe is installed into the sub-membrane perforated piping
network at a convenient location. Installation of the suc-
tion pipe about in the middle of the perforated piping
would help improve the uniformity of the sub-membrane
suction field. However, unless there is an unusually long
length of sub-membrane piping,  applying the suction
near the middle is probably not critical.

One approach for connecting the vertical  suction pipe
into the sub-membrane piping network is to use a rigid
PVC, PE, or ABS T-fitting or 90° elbow, as  illustrated in
the inset in Figure 7.

A leg of the T or the elbow is directed upward through a
hole in the membrane. Depending upon how the mem-
brane is subsequently to be  sealed against the pipe, one
option is to cut this hole through the membrane as a series
of slits radiating out from the center, leaving all of the
pie-shaped pieces in place. See the discussion earlier in
this section under Individual-pipe/SMD approach. The
vertical suction pipe is cemented into the upward leg of
the fitting.

The perforated piping must be firmly connected to the T
fitting or the elbow beneath the membrane, in a manner
such that it will not become disconnected over time. If
this joint became disconnected over time, there could be
significant leakage of air into the vertical suction pipe,
and  the system would  tend  to  behave more like  an
individual-pipe/SMD system with an individual suction
pipe at the location of the T.
         •  '      ' I '        , J !  '              •     '  '
If flexible perforated piping has been used, and if a T
fitting is being installed in the middle of this piping, the
                                                    240

-------
    piping is severed as necessary to accommodate the T.
    Each severed end of the perforated piping is then con-
    nected to a side of the T using the techniques discussed in
    Section 6.5 (see Pipe installation using rigid T-fltting).
    Where 4-in. corrugated piping has been used, the piping
    may be forced inside of (or around the outside of) the
    4-in. T leg, and firmly connected using screws or a hose
    clamp, as appropriate, and urethane caulk. If 3-in. corru-
    gated piping has been used, it should conveniently fit
    inside the leg of a 4-in. diameter T.

    If a rigid 90° elbow is being used instead of a T fitting, the
    elbow would be connected  to the corrugated piping as
    discussed above, but at one end rather than in the middle.

    If rigid perforated PVC piping has been used, this piping
    is cemented into each open end of the T fitting (or into the
    elbow) beneath the membrane.

    Other approaches for installing  a rigid vertical suction
    pipe into the middle of corrugated piping, rather than
    using a rigid T fitting, are discussed in Section 6.5 (see
    Pipe installation directly into tiles without rigid T).

    The membrane must be sealed tightly around the vertical
    pipe. This may be accomplished using a hose clamp, as
    illustrated in Figure 7. A procedure for  using a hose
    clamp for this purpose was discussed  earlier in  this sec-
    tion (see Individual-pipe!SMD  approach). Other ap-
    proaches can also be considered for accomplishing this
    sealing,  such as sandwiching the  membrane  between
    sealed roof flashing units (an adaptation of the approach
    illustrated in Figure 39A for the individual-pipe case).

    If there are multiple segments of perforated piping which
    have not been connected beneath the membrane, a sepa-
    rate vertical  suction pipe must be installed in each iso-
    lated segment. The resulting multiple risers will be con-
    nected by the piping network above  the membrane, as
    discussed in Section 8.6.

8.6    Design/Installation  of the
Piping Network and Fan
The design and installation of the above-membrane piping
network and the fan, for SMD systems would be essentially
the same as described in Section 4.6 for a SSD system.

Where multiple suction pipes penetrate the membrane, for
either SMD  variation, these pipes would be manifolded to-
gether, just as with SSD systems in basements. The piping
would  be hung from the floor joists under the living area
overhead. Since the crawl space is not lived-in space, there
will usually be a lot of flexibility regarding the routing of the
piping  through the  crawl  space. However, care should be
taken that the piping not block expected traffic routes (e.g., for
maintenance or subsequent replacement of a furnace located
in the crawl space).

As with other ASD variations, the exhaust piping from SMD
systems may neatly be routed up through the living area of the
house to a fan mounted in the attic, as suggested in Figures 6
and 7.

Where the exhaust stack is to be routed up the exterior of the
house  or through an  adjoining slab-on-grade garage, it is
common to  direct the SMD piping  out of the crawl space
through the band joist, just as in a basement.

For exterior stacks from systems in vented crawl spaces, some
mitigators have occasionally  taken  the exhaust piping out
through a foundation vent, if a vent existed at a convenient
location. Using a vent would save the effort of drilling through
the band joist and exterior finish. However, this can be a
building code violation, if it reduces the net vent area for the
crawl space below the area required by code. For this reason,
it is recommended that, in general, foundation vents not be
used for piping penetrations through the wall.

When there is an indoor stack, the SMD fan will generally be
in the attic. When there is an exterior stack (or a garage stack),
the SMD fan should be outdoors (or in the garage). Locating
the fan outdoors is particularly important when the crawl
space is unvented and is part of the conditioned space in the
house, which will be  the case, e.g., where  the crawl space
opens to an adjoining basement.

EPA's interim mitigation standards (EPA91b) specify that the
fans not be located in the crawl space, even if the crawl space
is  vented. The thermal stack  flows  will tend to draw the
crawl-space air up into the overhead living area. Thus, any
radon released in the crawl space due to leaks on the pressure
side of a crawl-space fan could enter the living area.

In houses with vented crawl spaces, where the fan cannot be
located in the  attic or in  an adjoining  garage, it can be
tempting to  mount the fan in  the crawl space to avoid the
aesthetic impact of mounting the fan outdoors. It would seem
that, since the crawl space is vented, any radon released from
pressure-side leaks should be diluted by the outdoor air that
will infiltrate through the foundation vents. However, limited
tracer gas data show that, due to the dynamics of crawl-space
houses, over one-half of any radon released by the fan in the
crawl space can flow  up into the living area even when the
crawl space is being naturally ventilated (Na85). As a result,
to be conservative, the interim standards require that fans not
be in crawl  spaces, despite the fact that occupants will not
normally be spending  much time in the crawl space itself.

In Section 4.6.5 (see  Requirement that fans be outside the
livable envelope of the house)* some precautions were dis-
cussed mat an installer should consider if there were no choice
but to mount a fan in a basement in violation of the standards.
One other precaution  that was considered prior to the stan-
dards for fans mounted in crawl spaces is to enclose the fan,
the fan couplings, and the pressure-side piping in a tight
enclosure fabricated from sheetrock or insulation board (An92).
A 1/2-in. hole is drilled in the pipe on the negative-pressure
side of the fan, inside the enclosure, in an effort to slightly
depressurize the enclosure.  The  concept is that  any
pressure-side leaks inside the enclosure would be drawn back
into the fan intake, preventing their release into the crawl
space. Any such enclosure would have to be tight but would
                                                        241

-------
also have to enable ready access to the fan for maintenance.
As emphasized in Section 4.6.5, there is no evidence that such
precautions would compensate for the risks incurred if the fan
were mounted in the crawl space, and the precautions thus
could never be considered as a substitute for abiding by the
standards.

8.7    Sealing in Conjunction with
SMD Systems
8.7.1  Sealing the Membrane

The most important sealing in conjunction with the installa-
tion of SMD systems will be sealing of the membrane.

As discussed  in Section 8.5.1 (see Sealing the membrane -
general), it is not possible to specify the precise conditions
under which various sealing steps will be cost-effective, or to
quantify the benefits that will result. However, available data
suggest that  better radon reduction performance will be
achieved if the membrane is completely sealed, and a number
of other benefits would be expected as well. As a result, it is
recommended that  the membrane be fully sealed whenever
possible.

Complete sealing of the membrane includes: sealing the seams
between sheets; sealing the membrane against the perimeter
foundation wall; and sealing the membrane against any inte-
rior obstructions, such as support piers and concrete pads. The
procedures for sealing these different membrane seams were
discussed in Section 8.5.1.

The most critical seams to seal would be: the seams between
sheets penetrated by individual suction pipes and adjoining
sheets; and seams running beside any sub-membrane perfo-
rated piping, including, for example, the seam between the
membrane and the perimeter foundation wall in cases where
the sub-membrane piping forms a loop around the perimeter.
The seam  between the membrane and the suction pipe, of
course, also must always be sealed well.

8.7.2 Sealing the Block Foundation Wall

Only very limited testing has been conducted to determine the
improvements in  SMD performance that might be achieved
by sealing major openings in the block foundation wall of a
crawl space.  Despite the lack of data, wall sealing might
intuitively be expected to be important in certain cases for two
reasons:
      by analogy with basement houses. In basements, SSD
      appears least able to prevent radon entry into the wall
      cavities when sub-slab communication is marginal or
      poor. When there is no gravel on the crawl-space floor,
      individual-pipe SMD systems might be expected to
      simulate poor-communication SSD systems. This anal-
      ogy would suggest that SMD systems with no gravel
      on the crawl-space floor may be failing to adequately
      treat wall-related entry.

      from experience in the two study houses having  the
      highest pre-mitigation radon levels in the living area
      (88-160 pCi/L), where addition of a BWD component
      (supplementing the SMD  system)  was necessary in
      order to reduce concentrations below 4 pCi/L (Ni89,
      Py90). Both of these houses included other complica-
      tions which prevented the tests from clearly demon-
      strating the role of the BWD component and the condi-
      tions under which a BWD component will typically be
      needed in other houses.

Open top voids, if present, would likely be the primary
openings warranting closure. The procedures described in
Section 7.7.1 and Figure 37 can be considered for closing the
top voids.

Figure 38B illustrates one approach for both sealing the entire
wall and increasing block-wall treatment by the SMD system.

As with the membrane sealing, sealing of the block founda-
tion wall (and/or adding a BWD component) are most likely
to be needed when the pre-mitigation concentrations are high,
or when  the desired post-mitigation concentration is lower
than 4 pCi/L.

8.8    Gauges/Alarms and Labelling

The considerations discussed in Section 4.8, concerning gauges/
alarms and labelling for SSD systems, also apply to  SMD
systems.

As discussed in Section 4.8.1, the gauge/alarm must be lo-
cated where it will be readily visible to the house occupant.
But much of the SMD piping will be in the crawl space, where
the house occupant may rarely go.

If the SMD exhaust piping rises up through the house via an
indoor stack, a gauge or alarm could be mounted on the stack
at some convenient location in the living area where it can be
seen or heard. If the stack is rising  through a closet in the
living area, which will commonly be the case, and if a gauge
were mounted in the closet with the pipe, rendering the gauge
less visible, the gauge might be supplemented by an audible
alarm.

Where the exhaust piping rises through an adjoining garage,
the gauge/alarm can be located in the garage.

Where the exhaust piping rises up outside the house, via an
exterior stack, a pressure gauge could be mounted at  some
convenient location in the living area. In this case, the indoor
gauge would have to be connected to the piping in the  crawl
space or  outdoors, via a line that penetrates the shell of the
living envelope. The ammeter discussed in Section  4.8.1
would be connected to the outdoor fan via the fan wiring.

Some pressure-activated alarms  will not  be  applicable for
SMD systems when the stack is outdoors, if they are designed
for direct mounting on the suction piping. Such direct mount-
ing would require either that the alarm be mounted on the
piping in the crawl space, in which case it may be too remote
to be considered readily visible or audible to the occupants; or
that the alarm be mounted on the piping outdoors, which will
not be feasible if the alarm is not designed for outdoor use.
                                                      242

-------
See the related discussion of pressure-activated alarms out-   In accordance with the discussion in Section 4.8.1 (see Pres-
doors in Section 6.8.                                       sure gauges), pressure gauges (and pressure-activated alarms)
                                                          must  tap into the suction piping at a point near  the pipe
As discussed in Sections 4.8.1 and 6.8, some mitigators report   penetration through the membrane in the crawl space. If the
that, during cold weather, moisture can condense or freeze in   gauge or alarm taps in  too close  to  the fan, it  may  see
the narrow tubing that extends from indoor pressure gauges to   relatively high suctions at the fan inlet even if there is a major
suction piping  in unconditioned crawl spaces or outdoors,   leak in the piping  near the membrane,  and even if  the
When this happens, the gauge  will give erroneously low   sub-membrane depressurizations have thus been dramatically
readings. Use of larger diameter tubing,  insulation of the   reduced. This potential  problem is especially acute with
outdoor tubing, and advice to the occupant are possible steps   pressure-activated alarms, which are often  pre-set by  the
to address this problem, as discussed in Section 6.8. Use of the   manufacturer such that they are not triggered until suctions
ammeter as the warning device would avoid this problem.     drop to extremely low levels (perhaps only 0.2 in. WG). Such
                                                          low levels can exist near the fan even if there is a total breach
                                                          in the piping near the membrane.
                                                        243

-------

-------
                                             Section  9
                              Considerations for Active
                            Soil Pressurization  Systems
For any of the active soil depressurization systems discussed
in Sections 4 through 8, the suction fan could nominally be
reversed to blow outdoor air into the soil, rather than drawing
soil gas out. As discussed in Section 2.4, active soil pressur-
ization in this manner has been found to be beneficial, relative
to soil depressurization, in cases where system flows were
high. Active soil pressurization has been found to sometimes
be beneficial in the following two situations:

 1. With sub-slab systems, as an alternative to SSD, in cases
    where the underlying soil is highly permeable, including
    well-drained gravels and highly fissured shales or lime-
    stones. Such highly permeable soils would permit greater
    than normal amounts of outdoor air to flow down through
    the soil into a  SSD system. In these cases, sub-slab
    pressurization could be an  alternative to installing a
    higher-flow fan or more suction pipes on the SSD system.

 2. With block-wall systems, as  an alternative to BWD, in
    cases where depressurization of the leaky block walls
    would draw enough air out of the basement to depressur-
    ize the basement. Such depressurization of the basement
    could increase soil gas influx and cause back-drafting of
    combustion appliances.

One typical configuration for a sub-slab pressurization system
is shown in Figure 40.

In most respects, the considerations in designing and install-
ing a sub-slab pressurization system or a block-wall pressur-
ization system will be exactly the same as those described in
Sections 4 and 7 for the corresponding depressurization sys-
tems. The primary difference between pressurization and de-
pressurization designs will be that, with pressurization: a) the
fan direction will be reversed; b) provisions will have to be
made to prevent outdoor dust and debris from blocking free
air flow through  the piping; and  c) the need for an exhaust
stack is eliminated, since the system is no longer discharging
a high-radon exhaust gas.

The discussion in this section focuses on the design and
installation differences resulting when a sub-slab or block-wall
pressurization system is being installed rather than a depres-
surization system. This discussion draws from the detailed
review of data from pressurization systems presented in Sec-
tion 2.4.
9.1    Selection of the Number of
Pressurization  Pipes

9.1.1  Sub-Slab Systems

The available data are limited from sub-slab pressurization
systems in houses amenable to pressurization. But these data
provide no solid evidence that the number of pipes required
for pressurization would necessarily be different from the one
or two pipes required for SSD systems in houses amenable to
SSD.

Information on sub-slab pressurization system configuration
are available from  six  houses amenable to this approach
(Tu87, Kn90). These installations included between one and
four pressurization pipes. These houses were only moderate in
size; their footprints generally ranged from 800 to 1,400 ft2,
with one being 1,800 ft2. The number of pipes usually corre-
sponded to one pipe per 400 to 700 ft2, although one house had
one pipe per 200 ft2,  and another house had one  pipe per
1,000+ ft2. In half of these installations, this number of pipes
was able to  reduce indoor levels  to below 2 pCi/L from
moderately elevated pre-mitigation levels (15-30 pCi/L). The
other half, having more elevated pre-mitigation levels (30-141
pCi/L), were reduced below 4 but not below 2 pCi/L.

The number of square feet per pipe is relatively low for these
pressurization systems, compared to figures as high as 1,850
to 2,700 ft2 per pipe with SSD systems in houses with good
communication. The values of 400 to 700 ft2 per pipe with
these  pressurization systems are more comparable  to  the
values of 350 to 750 ft2 per pipe encountered in SSD houses
having marginal rather than good communication.  And  the
residual indoor radon levels achieved by sub-slab pressuriza-
tion in  these houses  are no lower than those  commonly
achieved by one or two SSD pipes in houses  with good
communication.

These comparisons would make it appear that sub-slab pres-
surization systems may, on the average, require more pipes
through the  slab than will SSD systems.  This would be
intuitively reasonable. More pipes may be needed to ensure
adequate sub-slab pressures and adequate air flow into the soil
at all locations under high-flow conditions, than may be
needed  to achieve  sufficient depressurizations everywhere
under the lower-flow, more static conditions often character-
istic of SSD.
                                                     245

-------
                 Removable Cap,
                 for Pit Cleanout
                            T-Fitting
      Optional piping configuration,
      to facilitate removal of
      debris from pit
Notes:
1.   Closure of major slab openings is
    important.
2.   Fan may also be mounted at grade
    outdoors.
3.   Air filter must be accessible for
    changing or cleaning.
4.   Houses built on well-drained
    gravel soils (suitable for sub-slab
    pressurization) will not always
    have sub-slab aggregate as shown.
                                                   Sheet Metal Air
                                                   Filter Holder
           Screened Air
           Intake (Above
           Snow/Leaf Line)

                    Outdoor
                    Air
           Slope Horizontal
           Leg Down Toward
           Sub-Slab Hole
    Strapping (or
    Other Support)
                Insulation, to
                Prevent Pipe
                Sweating in
                Humidified
                Basements During
                Cold Weather
A-Collars,
Connecting
Metal Filter
Holder to
PVC Piping
                                                                                Pressurization
                                                                                Pipe
                                                                            Up Through Unclosed Openings
                                                                           Open Hole
                                                                           (As Large As
                                                                           Reasonably
                                                                           Practical)
Figure 40. Sub-slab pressurization using one typical approach.
                                                    246

-------
Another option with sub-slab pressurization, rather than add-
ing pressurization pipes, might be to use a higher-capacity
fan. For five of the six installations just cited (Tu87), the fans
were capable of generating flows of 50 to 200 cfm in the total
system, and pressures of 1.25 to 2.0 in. WG in the pipes near
the slab, in the given houses. In the sixth installation (Kn90),
the fan was a 90-watt tubular in-line fan of the type listed in
Table 1.

It must be underscored that most of the sub-slab pressuriza-
tion installations reported here are among the first such instal-
lations, and that no effort was made in the early work to
reduce the number of pressurization  pipes in these installa-
tions, or to optimize the number of pipes and the fan capacity
for general application. Therefore, it cannot be definitively
concluded from these results that sub-slab pressurization sys-
tems will in fact routinely require more pipes. Nor can an
assessment be made of the extent to which the number of
pipes might be reduced by increasing fan capacity.

9.1.2 Block-Wall Systems

Only three block-wall pressurization systems  have been re-
ported (Sc88). One of these represented the individual-pipe
variation (Figure 5), and two represented the baseboard duct
variation (Figure  35). In all three cases, the systems  were
operated with the same number of ventilation pipes (or the
same baseboard duct configuration) in both pressure and
suction, with comparable performance under both conditions.
Thus, a block-wall pressurization  system would be expected
to require the same number of pipes as a BWD system.

This conclusion is intuitively  reasonable.  Because of the
leakmess of  the block wall, BWD  systems  are  high-flow
systems just as are block-wall pressurization systems.  Both
BWD and wall pressurization thus probably have a significant
ventilation/dilution component in their mechanisms, and both
would thus be expected to require a comparable number of
pipes to distribute these flows.

9.2    Selection of Pressurization
Pipe Location
It is not apparent that the location of pressurization pipes (in
houses amenable to soil pressurization) should necessarily be
any different from that for depressurization  systems. Pipe
location for SSD suction pipes was discussed in Section 4.2;
the location for individual BWD suction pipes and for base-
board ducts was discussed in Section 7.2.

With SSD systems in houses having good sub-slab communi-
cation, Section 4.2 indicates that the one or two SSD suction
pipes might be located just about anywhere convenient. The
flows in SSD systems are low enough such that, even if a pipe
is placed at one end of a house, the suction field can extend
beneath the entire slab. With sub-slab pressurization systems
in houses underlain be highly permeable soils, location of the
pressurization pipe near one end intuitively might result in
excessive flows of air out through  the soil near that end of the
house. If that occurred, sufficient flows might not be estab-
lished at the remote end  of the house. Thus, in sub-slab
pressurization systems having only one pressurization pipe, it
might be advisable to locate that pipe at a relatively central
location, although there are no data confirming that this is in
fact required.

9.3    Selection of Pressurization
Pipe  Type and  Diameter
The type and diameter of piping that would be used for
sub-slab and block-wall pressurization systems would be the
same as that for SSD and BWD systems, discussed in Sections
4.3 and 7.3.

Even though the pressurization piping will now be handling
outdoor air rather than radon-containing soil gas, it is still
necessary to use rigid  PVC, PE, or ABS piping. Flexible
ducting and flexible clothes drier hose are still not acceptable,
due to reduced durability, difficulty in achieving sufficiently
gas-tight joints, and a tendency to sag, providing sites for
accumulation of condensate. Leakage or flow restriction re-
sulting from any of these problems with the flexible ducting
would reduce pressurization system flows, thus likely reduc-
ing performance.

The flows in a typical sub-slab pressurization system will tend
to be higher than those in a typical SSD system, since the
pressurization  approach will be used only in cases where
highly permeable native soil contributes to high flows. In the
five sub-slab pressurization systems for which flow data were
reported, flows generally ranged from 35 to 200 cfm, averag-
ing 100 cfm. This compares with the range of 20 to 100+ cfm
for SSD systems, averaging perhaps 50 cfm.

From the discussions in Sections 4.3.2 and 4.6.1,  it is clear
that, in sub-slab pressurization  systems, it will always  be
desirable to use piping of at least 4 in. diameter. Sub-slab
pressurization  flows will  essentially  always be too high to
permit  the use of 2-in.  piping, sometimes  considered in
low-flow SSD systems, and will often be too high  for use of
3-in. piping.

At the highest sub-slab pressurization flows (150 to 200 cfm),
it could seem desirable to use 6-in.  diameter piping, since
pressure losses in 4-in. piping would be about 1.5 to 2 in WG
per 100 ft at those flows (see Figure 13). However, the limited
experience with sub-slab  pressurization systems has shown
that in houses where sub-slab pressurization is preferred over
SSD, pressurization gives adequate performance with 4-in.
piping. Thus, 6-in. piping, though it would probably be help-
ful in particularly high-flow installations, does not really seem
to be necessary. Where a central trunk line down the length of
a basement connects to multiple pressurization pipes in the
basement, it might be desirable for the trunk line to be 6-in.
piping.

Flows in a typical stand-alone block-wall pressurization sys-
tem  will likely be about the same as those in  a typical
stand-alone BWD system; the leakiness of the walls should
create about the same  flows in both cases. Thus, the pipe
diameters for block-wall systems should be the same, regard-
less  of whether the system is  operating  in suction or in
pressure.
                                                       247

-------
9.4    Selection of Pressurization Fan

As just discussed, flows in sub-slab pressurization systems
can be relatively high. From the limited available data, flows
apparently can range from about 35 to 200 cfin, averaging
about  100 cfm. The actual flow will depend on the  soil
characteristics and the fan performance curve.

At the low end of this range, either a 90-watt in-line tubular
fan (Section 4.4.1) or an in-line radial blower (Section 4.4.2),
or their equivalent, could be a reasonable choice. As flows
approach the average of about 100 cfm,  the radial blowers
would no longer provide pressures any  greater than those
from the tubular fans. Thus, at flows approaching 100 cfm,
any added cost for a radial blower might be an unnecessary
investment; a 90-watt tubular fan (or equivalent) could be the
more logical choice.

At the upper end of the flow range (150 to 200 cfm),  a
100-watt tubular fan with 6- or 8-in. couplings (Section 4.4.1),
or equivalent, might be considered. In fact,.the 90-watt tubular
fans may not always be able to maintain flows this high if
there is any significant length of 4-in. piping in the system. If
the 90-watt fan were used, the system would possibly operate
at a lower flow, consistent with the fan's performance curve
and with the sub-slab flow characteristics of the house.

Sub-slab  pressurization systems may require high flows to
achieve the necessary ventilation/dilution  of the gas beneath
the slab,  and to maintain the  necessary sub-slab pressures.
Consequently, the smaller, 50- to 70-watt tubular fans  dis-
cussed in Section 4.4.1, which can move only about 120 to
160 cfm at zero static pressure, will likely be insufficient for
soil pressurization applications.  And the high-suctionAow-flow
fans discussed in Section 4.4.3, for SSD  systems in houses
with poor sub-slab communication, would never be appli-
cable.

Because of the leaMness of block walls, block-wall pressur-
ization systems should  have the  same flows typical of
stand-alone BWD systems, namely, 100 to 200 cfm. Thus, the
90- to 100-watt tubular in-line fans (or their equivalent) will
usually be the logical choice  for block-wall pressurization
systems, as discussed in Section 7.4 for BWD systems.

9.5    Installation of Pressurization
Pipes Beneath the  Slab or Into the
Block Walls

9.5.1  Sub-Slab Systems

The installation of sub-slab pressurization pipes beneath the
slab will be  essentially identical to the installation of SSD
pipes, described in Section 4.5. Almost all of the  sub-slab
pressurization installations reported to date have involved
pipes installed vertically down  through the slab from indoors,
as illustrated in Figure 40 (analogous to the configuration
described in Section 4.5.1). But in concept, any one of the six
alternative installation approaches for SSD pipes described in
Section 4.5 could also be used  for pressurization pipes.
Investigators have often observed cases with sub-slab pressur-
ization systems where dust and other outdoor debris have
been drawn into the system and deposited beneath the slab
where the pressurization pipe penetrates (Pr89, NYSEO91,
An92). This deposition has increased the back-pressure in the
piping and reduced the effectiveness of the system, reducing
the flow beneath the slab. Such deposition can create a prob-
lem despite the  presence of an air filter in the system to
remove outdoor dust from  the intake air,  as  discussed in
Section 9.6.

Excavation of a pit beneath the suction pipe, as in Figure 40,
should help reduce the impact of such deposition by distribut-
ing the dust over a larger surface area. As discussed in Section
4.5.1 (see Excavating a pit beneath the slab), one primary
purpose of such a pit with SSD systems is to reduce suction
loss as the soil  gas accelerates up to pipe velocity. With
sub-slab pressurization systems, the pit will also serve to
reduce pressure drop in an analogous manner. However, the
pit may also reduce the impact of deposited dust and debris in
hindering flow.

Some mitigators have found that the use of a sub-slab pit,
combined with the use of an intake  air filter, can  be insuffi-
cient to prevent unacceptable flow blockage by dust deposi-
tion (NYSEO91,  An92).  A  sufficiently thick  and
low-permeability layer of fine dust which has penetrated the
filter can  still deposit on the surface of the pit  over time,
restricting flow through the surrounding permeable soil.

One approach for addressing  this problem  is to  install the
pressurization pipe in the manner shown in the inset in Figure
40, to facilitate  subsequent removal of deposited material
from the pit (An92).

By this technique, a short stub of piping is mounted through
the slab, and a T fitting is cemented  onto this stub just above
the top of the slab, as shown in the inset. One end of the top of
the T is cemented onto the stub, with the other end of the top
extending upward;  the leg  of the  T  extends horizontally,
parallel  with the slab. The vertical pressurization pipe is
connected to this horizontal leg with a 90° elbow. The end of
the T that is extending upward is capped with an air-tight
flexible PVC cap, held onto the T with a hose clamp. This cap
is of the same material as the flexible couplings used to
connect the fans to the piping.

The cap can then be removed whenever the system pressure
gauge indicates that there may be an unacceptable accumula-
tion of deposited material. Access to the pit through the piping
stub permits the material to be vacuumed or otherwise cleaned
out of the pit by the homeowner or by service personnel.

This approach will increase the piping pressure loss by virtue
of the additional 90° bend and the T that are incorporated into
the piping system. However, in cases where significant depos-
its of ambient  dust are anticipated from experience, this
configuration will be far superior to the option of periodically
removing and re-installing the pressurization pipe  in order to
remove the deposits.
                                                      248

-------
9.5.2  Block-Wall Systems

The installation of pressurization pipes into the block walls,
and of baseboard ducting for wall pressurization systems, will
be identical to the installation of BWD piping, described in
Sections 7.5.1 and 7.5.2.

9.6    Design/Installation of Piping
Network and Fan
Many of the considerations in designing and installing the
piping network and the fan for soil pressurization systems will
be similar  to those for soil depressurization systems, dis-
cussed previously in Section 4.6. However, there will be some
important differences.

These differences  are illustrated in Figure 40 for a sub-slab
pressurization system. Figure 40 is to be compared with the
SSD diagram in Figure 1. While other sub-slab pressurization
configurations are possible, as discussed later, the one shown
in Figure 40 is reasonably typical. Although  figures for
block-wall pressurization systems are not shown here, sche-
matics for block-wall pressurization installations would in-
volve modifications to the BWD systems illustrated in Figures
5 and 35, similar to the modifications made in Figure 1.

Pipe routing and installation between the pressuriza-
tion points and the fan.  The requirements for pipe routing
and installation in both sub-slab and block-wall pressurization
systems are very similar to those described for SSD systems
in Sections 4.6.2 and 4.6.3, and for BWD systems in Section
7.6.

The  horizontal  piping runs must still be sloped downward
toward the vertical  piping into the slab or walls to allow
condensate drainage, and care must be taken to avoid low
points in the piping  where condensed moisture can accumu-
late.  With depressurization systems, the concern was prima-
rily  about  condensation of soil gas moisture during cold
weather, and about  rainwater that  enters the stack. With
pressurization systems, the concern is primarily about con-
densation of humidity in the outdoor air during air condition-
ing season.

As with SSD and BWD systems, the piping joints must be
carefully cemented.  Even though the piping is handling out-
door air—and even though leaks on the pressure side of the
fan would thus simply result in fresh air being blown into the
house, which might seem innocuous—such pressure-side leaks
could reduce the flows and pressures being delivered to the
soil (or into the block walls). They could also be suggesting a
problem with the physical integrity of the piping network.

Leaks on the suction side of the fan, if inside the house, would
result in house air being drawn into the system. In theory, this
might create depressurization within the house, and increase
the system heating/cooling penalty. More of the air being
blown beneath the slab (or into the walls) would be house air
rather than outdoor air. However, much of the air being blown
beneath the slab (or into the walls) is likely flowing back into
the house, through cracks and other openings in the slab and
walls. Thus, it  is  not clear whether such suction-side leaks
would really have any serious impact. Nevertheless, the miti-
gator is still advised to cement the suction-side joints care-
fully.

Some mitigators recommend insulation of all indoor piping
(An92, Br92). Such insulation is shown in Figure 40. This
insulation will reduce "sweating" on the outside of the pipes,
especially in humidified basements during cold weather, and
hence will reduce or avoid water stains in finished areas. As a
secondary benefit, it will also reduce condensation inside the
pipe during hot weather.

Because of the high flows characteristic of pressurization
systems, insulation of the piping will also serve to reduce flow
noise. The average flow of 100 cfm observed in some sub-slab
pressurizations systems to date corresponds to a flow velocity
of 1150 ft/min in 4-in. piping, approaching the velocity range
at which flow noise can start to become objectionable.

Need for a filter on the air intake. One key difference in
the design of the piping network for soil pressurization  sys-
tems is that an air filter is necessary to remove dust, pollen,
and other debris from the incoming outdoor air. One possible
configuration for such an intake filter is shown in Figure 40.

This air filter is important for sub-slab pressurization systems,
where deposited debris in the sub-slab pit has been found on
some occasions to increase system back pressure and reduce
flows. The air filter may be less critical in block-wall pressur-
ization systems (especially of the individual-pipe variation),
since the void network into which the pressurization air is
discharging is  so open that interference from  dust deposition
is intuitively less likely.

The recommended location for the filter is upstream of the
fan, just as filters in forced-air heating systems are upstream
of the central  furnace fan. This location will provide some
protection for  the fan, as well as reducing deposition in the
sub-slab pit. Upstream location of the filter in  the figure is
possible because the figure shows the fan inside the house. As
discussed later, the fan can also be located outside the house;
in that case, the filter will be downstream of the fan if the filter
is indoors.

The filters are usually housed in sheet metal boxes, similar in
some respects to the sheet metal housings for filters in forced-air
heating systems. At least one vendor markets a sheet metal
filter housing that could be installed in-line in the PVC piping,
equipped with a gasketed door for access to the filter. This is
the type of filter holder illustrated in Figure 40.  This sheet
metal box could be connected to the PVC piping using sheet
metal A-collars, as in the figure. With the A-collars properly
installed, and with the access door gasketed,  the air leakage
through joints associated with the filter housing should be
reasonably limited.

The filter shown in Figure 40 is a type which is bent into the
shape of an inverted V, Pleated paper filters are recommended
(An92, Br92). Pleated units have increased surface area and
thus achieve better removals, have better dust-holding capac-
ity (reducing frequency of replacement), and create less pres-
                                                        249

-------
sure drop. These filters are more efficient than the filters
typically used in central forced-air furnace systems.
                                                     I
Some mitigators recommend against using  washable wire
mesh filters, since these are less efficient (Br92). Others feel
that washable filters are satisfactory, since the ease of clean-
ing may encourage the occupants  to clean the filters more
regularly than would be the case if they had to buy a new
replacement each time (An92).

Instead of the gasketed-door filter housings with V-filters, as
shown in the  figure, one could also consider the slotted filter
housing design common to many forced-air heating systems,
where the filter slides into the slot. If such a housing were
used, a sheet metal cover for the slot should be included to
reduce air leakage (An92).

Periodic replacement or cleaning of the filter is critical. Other-
wise, the pressure drop across the filter will increase and
flows will decrease. Also, dust may "break through" the filter,
depositing in the fan (thus reducing fan performance) or in the
sub-slab  pit (also reducing flows). Even the pleated paper
filters, which can go the longest between replacement, must
be replaced about twice a year (An92).

For this reason, the filter must be located where it can be
easily accessed by the house occupant for filter changing or
cleaning. In the configuration shown in Figure 40, the filter
housing is up between the overhead floor joists in the base-
ment. The gasketed door in this case would likely be on the
bottom of the housing, for easy access to the filter.

General  considerations regarding fan  location and
exhaust piping design.  The primary  difference between
pressurization and  depressurization systems, in terms of the
piping network and the fan, is associated with the fan and
"exhaust" part of the system. Because the fan is blowing fresh
air beneath the slab or into the wall voids, there will be no
high-radon fan exhaust in pressurization systems, and all of
the system piping will contain outdoor air. These facts impact
the system design in two major ways:

   -  first, the fan can now be inside the living envelope, if
      desired,  because there is no longer any risk that
      high-radon gases can leak out of the pressure side of
      the fan; and

   -  second, there is no longer any need for an exhaust stack
      up through the interior or the exterior of the house. The
      "exhaust"—actually, the fresh air  intake—can be at
      grade level.

Fan location and mounting—indoor fan.  There  are
basically two options for mounting the fan: indoors  or out-
doors. In both cases, of course, the fan is mounted with its
direction reversed, relative to its direction for the depressur-
ization systems, so that outdoor air is being blown beneath the
slab (or into the block walls).

The first fan mounting option, shown in Figure 40, is to locate
the fan inside the house. In this case, the appearance of the air
intake outside the house can be a screened air intake grille, as
shown in the figure. The 4-in. diameter intake piping would
likely penetrate the band joist (or some  other convenient
location on the side of the house), and the intake grille would
be mounted at the end of this pipe. This intake grille must be
mounted at a sufficient height on the side of the house such
that it will not become blocked by snow or by accumulated
leaves or other debris.

As  discussed above (see Need for a filter on the air intake),
the indoor fan should be mounted downstream of the intake
air filter, to provide some protection from the dust and debris
entrained with the outdoor air.

The indoor fan should be mounted vertically, as shown, so
that condensed moisture can drain down into the sub-slab
region (or into the block walls). As indicated previously, the
threat of condensation inside the pipe will be greatest during
hot, humid weather when the house is air conditioned.

As depicted in the figure, vertical mounting of the fan is most
conveniently done by installing it in the vertical pipe extend-
ing down into the slab. This is easy to do when there is only
one pressurization pipe. If there were multiple pipes mani-
folded  together, the fan must be mounted  in the trunk line
upstream of the point at  which the piping leading to the
different pressurization pipes splits  off from the trunk. Loca-
tion of the fan downstream of this split would result in the fan
depressurizing the pipes upstream of the split.

If there were additional pressurization pipes beyond the one
shown  in Figure 40, the horizontal trunk line leading to the
other pipes would have to split off at a point in the vertical
pipe below the fan. Even if the fan were placed as close as
possible  to the basement  ceiling, the horizontal trunk line
would  be some distance below the  floor joists. This will
usually not be an acceptable location for the horizontal run
across the basement. Thus, where there are multiple pressur-
ization pipes, one may wish to consider other options:  mount-
ing the fan outdoors, discussed later, or mounting the fan
horizontally with provisions for water drainage out of the
housing,  an approach which could void the warranty on at
least one vendor's fans.

The soil pressurization fan should be mounted onto the system
piping with air-tight couplings, just as with depressurization
systems.  See Sections 4.6.4 and 4.6.5.  As discussed previ-
ously in  this  section (see Pipe  routing and installation be-
tween the pressurization points and the fan), the fact that the
fan will be handling outdoor air rather than soil gas does not
eliminate the need for the joints to be air-tight.

Fan location and mounting—outdoor fan. The second
fan mounting option is to mount the fan outdoors.

In this option, the portion of the system that is outdoors could
look very much like that shown in Figure 27 for the  case of
SSD with an exterior stack, except that there would be no
stack above the fan. In place of the  air intake grille shown in
Figure 40 on the side of the house, the piping would extend
out through the side of the house; an  upward-directed 90°
elbow would be mounted on the outside end of this  piping;
and the fan would be mounted vertically on a stub of pipe
                                                        250

-------
extending up from this elbow, oriented to blow outdoor air
into the piping.

Where an in-line duct fan is mounted beside the house in this
manner,  a screen must be installed on the intake, for two
reasons.  One reason is  to prevent children and pets from
reaching  into the fan blades. The other reason is to prevent
debris from entering the fan. Intake grilles for this purpose are
sold by some of the fan manufacturers.

When the fan is mounted outdoors, the air filter (which will
usually be inside the house) will be on the pressure side of the
fan, rather than on the fan inlet as in Figure 40. Thus, outdoor
fans will not have the degree of protection from outdoor dust
and pollen that  will be possible  for indoor fans  mounted
downstream of the filter. The screens on the inlets to outdoor
fans will not be effective at removing fine dust

One option that might be considered when the fan is mounted
outdoors is to mount the fan horizontally,  directly on the
horizontal pipe where it penetrates the band joist. In this case,
the fan would be mounted at the location where the screened
air intake grille is shown in Figure 40. Because the fan is
outdoors, and is  at the same temperature as the surrounding
outdoor air, the  threat of moisture condensation in the fan
housing is largely eliminated. Thus, this restriction on hori-
zontal mounting would be avoided. However, entrainment of
rainwater could  still result in some accumulation  of water
inside the housing with horizontally mounted fans, and some
provision for drainage of this water out of the housing would
be necessary (see Section 4.6.5, Mounting exterior or garage
fans). Horizontal mounting may void the warranty on some
fans.

9.7     Sealing in Conjunction with
Soil Pressurization Systems
The slab arid wall sealing suggested in conjunction with SSD
systems (in Section 4.7) and with BWD systems (Section 7.7)
are also  appropriate for the  analogous soil pressurization
systems.

With  soil depressurization systems, the sealing aids in the
distribution of the suction field beneath the slab. The sealing
also reduces the risk that soil gas will enter the house through
the sealed entry route, in the event that the depressurization is
not fully effective for that route.
With  soil pressurization systems, the sealing provides an
analogous benefit. Soil pressurization systems function in part
by pressurizing the sub-slab (or wall cavity) region, prevent-
ing soil radon from moving by convection into these regions.
Sealing will aid in distributing the pressure field, thus improv-
ing pressurization system effectiveness.

Soil pressurization systems can also be viewed as functioning
by creating relatively high  flows under the slab (or in the
walls). These high flows create the pressure field indicated in
the preceding paragraph, preventing convective soil gas flow
toward the foundation. Flows down into the soil  can  also
overcome the diffusion of radon toward the foundation. In
addition, the high flows can  be viewed as ventilating the
sub-slab or wall cavities, diluting any radon that does reach
the foundation. Since any radon that does reach the sub-slab
will likely be forced up into the house through slab cracks, the
reductions in sub-slab radon levels must more than compen-
sate for the increased flow of sub-slab gas into the house.

Accordingly, in addition to increasing sub-slab or wall cavity
pressure, the sealing of slab and wall entry routes can also be
viewed as aiding system performance by forcing more of the
flow to be directed into the soil rather than back into the
house. Directing the flow away from the house can be viewed
as helping  ensure that the diffusive movement of radon to-
ward  the foundation  is overwhelmed, and as increasing the
extent to which the air ventilating the sub-slab is sweeping the
radon away from the house rather than into the house. Sealing
may also reduce the extent to which other sub-slab "pollut-
ants"  (such as moisture, termiticides, and  fungi) may be
forced into the house by the system.

9.8    Gauges/Alarms and Labelling

The considerations discussed in Sections 4.8 and 7.8, con-
cerning gauges/alarms and labelling for SSD and BWD sys-
tems, also apply to soil pressurization systems.

Of course, for soil pressurization systems, any pressure gauges
or detectors installed on the system will be set to monitor
elevated pressure rather than suction.

In sub-slab pressurization installations where back-pressures
can repeatedly build up over time due to deposition of dust in
the suction pit, a pressure gauge on the system piping (or a
pressure-actuated alarm) will alert the occupant when the pit
must be cleaned out for continued satisfactory system perfor-
mance.
                                                       251

-------

-------
                                           Section  10
                            Considerations for Passive
                         Soil Depressurization  Systems
Passive soil depressurization systems eliminate the fan used in
ASD systems. Instead they rely upon natural thermal and
wind effects to create die desired depressurization. They may
also depend in part on the "pressure break" mechanism dis-
cussed in Section 2.5. Passive systems incorporate a fan-less
stack which usually rises up inside the house and through the
roof, with suction being created by the natural thermal stack
effect in the house and by the flow of winds over the roofline.
These naturally induced suctions are very low, compared to
the suctions that a fan could produce.

Because of these low suctions, and the potential variation in
passive performance as ambient temperatures and winds vary,
the performance of passive systems has been unpredictable.
Much of the reported testing of passive soil depressurization
installations has been in newly constructed houses, where
steps were taken during construction in an effort to improve
passive performance. These steps  included: good sub-slab
aggregate, to provide the best possible communication; a
sub-slab network of perforated piping, to facilitate distribu-
tion of the weak suction field; and a relatively tight slab, to
reduce the extent to which the weak suction field would be
further weakened by house air leakage into the system.

Passive soil depressurization systems should be considered
for retrofit into existing houses only in cases where:

 a. The required radon reductions are only moderate (no
    more than about 30 to 70%).

 b. The house appears potentially amenable to passive treat-
    ment, preferably including: good sub-slab aggregate; a
    tight slab, to reduce air leakage into the system; a poured
    concrete foundation wall,  to minimize air leakage from
    the walls; and a tight soil, to reduce air leakage through
    the soil. Existing sub-slab perforated drain tiles are poten-
    tially helpful but apparently not really necessary.

 c. The  mitigator or the house occupant are prepared to
    monitor system performance for an extended period, to
    determine whether the system is routinely overwhelmed
    under certain weather or appliance usage conditions, and
    will add a fan if needed.

 d. There is a strong preference on the part of the mitigator or
    homeowner for a passive system.
A given passive system would always provide greater radon
reductions if it were activated with a fan.

Because passive systems generate such weak suctions and low
flows, they will be potentially applicable only to SSD sys-
tems, DTD  systems with the  perforated piping inside the
footings (usually  sump/DTD systems), and SMD  systems
with a completely sealed membrane. They will not likely ever
be applicable to DTD systems having exterior drain tiles,
because the weak suction will likely not extend into the
sub-slab region. Nor will they likely ever be applicable to
BWD systems, because of the high flows required for such
systems to be effective.

The discussion in this section focuses on  the design/installa-
tion differences that would result when a passive depressur-
ization  system is being installed rather than an active system.
This discussion draws from the detailed review of the limited
data from passive systems, presented in Section 2.5.

10.1  Selection of the Number of
Suction Pipes

10.1.1    Passive SSD Systems

A passive SSD system will require more suction pipes than an
active SSD system to achieve a given radon reduction.

One to two passive SSD pipes were installed in seven base-
ment houses in Maryland having reasonably good and uni-
form sub-slab communication  (Gi90),  corresponding to one
pipe per 700 to 1,100 ft?. The one to two pipes yielded radon
reductions ranging from zero to 90%  in these houses. The
post-mitigation radon concentrations averaged between 5 and
8 pCi/L in all except one house, where they averaged about 2
pCi/L.  By comparison, when each of these systems was
activated by adding a fan, the radon reductions increased to 70
to 99%, with residual radon levels of about 1 pCi/L and less in
all houses.

This good performance with the active SSD systems is consis-
tent with that which would be expected when there is one pipe
per 700-1,100 ft2 in houses having good sub-slab communica-
tion (see Section 4.1.1). Clearly, if most of these passive SSD
systems were to reduce levels even to 4 pCi/L, the passive
systems would have to be designed with  more pipes, so that
each pipe only had to treat a smaller area. And to achieve the
reductions obtained with  the active systems (1 pCi/L), the
                                                    253

-------
added number of passive pipes could be substantial. It is
doubtful that with any practical number of suction pipes a
passive SSD system could ever match the observed perfor-
mance of the active systems in most of these houses.

In view of the  limited amount of experience  with passive
systems, it is not possible to specify more definitively how
many square feet should be treated by each passive SSD pipe
to reliably obtain indoor levels below 4 pCi/L with passive
SSD systems.

10.1.2    Passive DTD Systems

All experience  with passive  DTD systems has involved a
single vertical suction pipe connecting to the sub-slab perfo-
rated piping, usually at a sump. These passive systems pro-
vided radon reductions of 20 to 75% in two existing houses
(Gi90), and 9 to 90% (averaging 64 to 70%) in a number of
newly constructed houses (Sau91a, Sau91b). (By comparison,
activation  of each of these  installations by  adding a fan
increased reductions to above 90%, reducing all of the houses
to 1 pCi/L and less.)

There are no data to suggest whether a second passive suction
pipe connected  into the perforated piping in a given house
(e.g., on the opposite side of the basement from the first pipe)
would have improved the passive performance.

In counting the number of suction pipes in  passive DTD
systems, one also has to consider the extent of the perforated
piping beneath the slab. This perforated piping is a potentially
important contributor in distributing the passive suction.

In the retrofit installations in the existing houses in Maryland
(Gi90), the performance of the two passive  sump/DTD sys-
tems (20 to 70% reduction) was no better, on average, than
that in the seven  passive SSD installations (zero to 90%
reduction). Some of the passive SSD systems gave better
performance than did the sump/DTD systems.

Among the passive DTD installations in newly constructed
houses (Sau91a, Sau91b), the performance range (9 to 90%
reduction) was  the same as for the  passive  SSD  retrofit
installations (Gi90), but the average of the DTD systems (64
to 70%) was somewhat higher than the average of the SSD
systems (about  55%). This very limited data base may be
suggesting that  the sub-slab perforated piping was aiding
system performance to  some limited extent.  On  the other
hand, the somewhat better results with passive DTD in the
newly constructed houses—if real—may instead be resulting
from improved slab sealing implemented during construction,
rather than from the drain tiles.

In summary, the sub-slab perforated piping may or may not be
a contributor in  improving the performance of passive DTD
systems relative to passive SSD systems. Any performance
improvements achieved with the perforated piping may be
influenced by the extent and configuration of this piping
beneath the slab, discussed in Section 10.2. Where a house is
being built and where it is desired to increase the likelihood
that a passive soil depressurization system  will give adequate
performance, the builder may wish to install perforated piping
beneath the slab during construction. However, where a pas-
sive system is being considered for retrofit into an existing
house that does not have sub-slab piping to begin with, there
is no clear incentive to justify the extensive effort that would
be required (including removal and restoration of part of the
slab) to retrofit perforated piping.

 10.1.3    Passive SMD Systems

Passive SMD systems have been reported in two houses, each
of which had abasement adjoining the crawl space (Gi90). In
one  of these  cases, the  passive SMD  system was supple-
mented by a passive  SSD system in the basement. In both
cases, the sub-membrane piping/SMD  approach was used,
with a length or loop of perforated  piping placed beneath a
completely sealed membrane. One vertical suction pipe con-
nected to the sub-membrane piping.

In the house where the basement was not treated, which had
gravel on the small crawl-space floor, and where the furnace
flue served as the passive stack, this passive SMD system was
sufficient to reduce concentrations below 4 pCi/L (a reduction
of 20 to 70%). In the other house, reductions were only 0 to
30%, and levels remained significantly elevated, due to mar-
ginal and uneven communication beneath the basement slab
(perhaps in addition to any  inadequacies in the crawl-space
treatment).

One mitigator also reports achieving reductions of about 60%
or greater in crawl spaces with poured foundation walls using
a length of perforated piping  beneath  a completely sealed
membrane (K192). In these  cases, the passive stack  simply
penetrated the band joist and terminated near grade.

These limited  results  do not permit definitive conclusions
regarding  the  number of  suction  pipes or  the need  for
sub-membrane perforated piping in passive SMD systems.
Among the questions  that cannot be addressed from these
limited data are: the importance of the sub-membrane piping
under various conditions (crawl-space size, block vs.  poured
foundation  walls, gravel vs. no gravel  on the floor);  the
number of suction pipes that might have been needed on that
sub-membrane piping under  these various conditions; and the
number of suction pipes that might have been needed had the
individual-pipe/SMD approach been used. Intuitively, it would
seem that sub-membrane piping  could be important for pas-
sive SMD systems, at least  when there is no  gravel  on the
floor. Also, it is intuitively apparent that  complete and careful
sealing of the membrane will always be required to improve
passive SMD performance and to reduce the number  of suc-
tion pipes.

10.2  Selection of Suction Pipe
Location

 10.2.1     Passive SSD  Systems

There are insufficient data to enable definitive guidance re-
garding the optimum location for passive suction pipes. Intu-
itively, since the weak passive suction field may not be able to
extend very  far,  it may  be  desirable to  locate  the suction
                                                      254

-------
piping near the soil gas entry routes, such as the wall/floor
joint.

However, the pipes should not be located too near to these
routes unless the routes can be  sufficiently well  sealed to
avoid excessive air leakage into the passive system. Excessive
leakage of house air into a low-flow passive suction pipe from
a nearby  unsealed entry route could prevent the suction field
produced by that pipe from extending to other, more remote
entry routes.

With the seven passive SSD systems retrofit  into existing
houses having good sub-slab communication (Gi90), some of
the most effective systems had the one or two suction pipes
near the  perimeter. However, some systems with centrally
located pipes also  gave generally similar performance. The
data are  far too limited to show any consistent difference
between perimeter vs. central pipe location.

10.2.2    Passive DTD Systems

With the passive DTD systems, a variety of sub-slab perfo-
rated piping matrices have been tested. Some installations
have involved perforated piping loops that extended around
the entire perimeter  (Ta85, Ka89, Gi90, Sau91a,  Sau91b).
This configuration will commonly  be present in existing
houses having sumps, and  has been installed in some newly
constructed houses to enable passive DTD. Intuitively,  a
perimeter loop configuration should be the most effective at
distributing the  suction near the major entry route (the wall/
floor joint).

In some  of the newly constructed  houses that have been
tested, the perforated piping may have been laid in as a single
length extending for some distance down the center of the slab
(Ka89, Sau91a,  Sau91b).

The most extensive  sub-slab piping network that  has been
considered for passive DTD systems was tested in early work
in Canada (Vi79), and was incorporated into guidelines by the
Central Mortgage  and Housing  Corp.  of Canada for new
housing built near  uranium mining and processing sites. As
discussed in Reference EPA88a, this extensive matrix in-
volved numerous parallel lengths of 4-in. diameter perforated
piping extending down the length of the slab from end to end,
capped at each end. The pipes were laid so that no point
beneath the slab would be more than 1  to 2 ft away from a
pipe. The parallel lengths of piping were connected beneath
the slab by a 6-in. diameter manifold pipe which ran perpen-
dicular to the parallel perforated pipes (i.e., from the front of
the house to the rear) near the center of the slab. This manifold
was connected to the passive stack which rose through the
house. As discussed in Section 2.5, even this extensive piping
network  was unable to reliably  maintain 18 study nouses
below 4 pCi/L year around in passive operation (Vi79).

In summary, the limited experience with these various sub-slab
piping network configurations do not permit any conclusions
regarding whether  any one of the configurations will com-
monly give better performance than the others. With recent
experience, it is now apparent that  extensive networks are
probably  not warranted, if there is a good aggregate layer
under the slab. Intuitively, a perimeter loop of perforated
piping might be expected to enable better passive treatment of
the wall/floor joint than would a central  length of piping,
although there are no data to verify this intuition.

The preceding discussion, concerning the selection  of the
sub-slab piping configuration for a passive DTD system, is
applicable in practice only to new construction.  In new con-
struction, the piping can be laid in any desired configuration
before the slab is poured.

As discussed in  Section 10.1.2, available data  suggest that
perforated piping does not provide  a  sufficiently distinct
improvement in passive performance to warrant the extensive
effort that would be required to retrofit sub-slab piping into an
existing house where it was not present  to begin with.  In
houses having  good sub-slab communication, passive  SSD
appears to perform roughly as well as passive DTD. In houses
having marginal or poor sub-slab communication, passive soil
depressurization  should be eliminated as an option  at the
outset. In reduced-communication houses, an active system
would be the reasonable approach, rather than tearing out a
portion of the slab to lay perforated piping and gravel in the
hopes that a passive DTD system might be adequate.

With passive DTD systems, it appears that the vertical PVC
suction pipe (i.e., the passive stack) can be connected into the
perforated piping at any convenient location. However, the
connection should be remote from any suspected source of
substantial air leakage into the system, since leaks near to the
suction pipe will have the greatest impact on suction  field
extension.

Where interior drain tiles empty into a sump, the suction pipe
might be installed in  the  sump.  However, as discussed in
Section  5.2,  it  may be preferred  to locate the stack remote
from  the sump,  both to avoid increased leakage that  may
occur at the  sump, and to simplify subsequent  sump pump
maintenance.

70.2.3    Passive SMD Systems

There is no  experience with  passive individual-pipe/SMD
systems to enable guidance on where individual suction pipe(s)
should be placed. If there is gravel on the crawl-space floor,
and if the membrane  has been sealed everywhere (including
against the perimeter foundation wall), the pipe(s) can prob-
ably be located at either perimeter or central locations,  as is
apparently the case with passive SSD systems.

The limited experience with sub-membrane piping/SMD sys-
tems enables no guidance regarding whether a loop of perfo-
rated piping around the perimeter would be preferred over a
straight length down the interior, or over some other configu-
ration. As discussed for active SMD systems in Section 8.2.2
(see Sub-membrane perforated piping), special piping  con-
figurations can be needed when  the crawl space has an irregu-
lar cross section or when adjoining crawl spaces are isolated
from one another.
                                                       255

-------
10.3  Selection of Suction Pipe Type
and  Diameter
As with active systems, the  above-slab piping for passive
systems will be PVC, PE, or ABS. Perforated piping below
the slab can be rigid perforated PVC, PE, or ABS, or can be
the flexible black corrugated material commonly used for
drain tiles.

The diameter of the passive suction pipe must be selected to
minimize suction loss in the piping, since the naturally in-
duced passive suctions are so low to begin with. Usually, 4-in.
diameter piping has been used.

The very low suctions generate very low flows in the passive
systems. Reported flows have essentially always been less
than 10 cfm, and are commonly less than 1 cfm. At flows on
the order of 1 cfm, suction losses in 4-in. piping would be on
the order of 0.0001 in. WG per 100 ft of piping (extrapolating
Figure  13). The stack will often rise straight up through the
house with no  horizontal runs and no  elbows (or with very
few); as a result, the equivalent length of piping may be only
perhaps 15 to 30 ft. With this little piping and with the low
friction loss per 100 ft, the total friction loss in the passive
4-in. pining network would be on the order of 1% of the
passive suction induced in the stack by thermal effects alone
(ignoring the contribution from wind effects).

Thus, from the standpoint of friction loss, there is generally no
incentive to use piping larger than 4 in.

10.4  Selection of a Supplemental
Suction Fan
Passive soil depressurization systems,  by definition, should
operate without a fan. However, in some passive installations
which are largely able to reduce indoor levels to 4 pCi/L but
which are  occasionally overwhelmed (e.g., when the central
furnace fan comes on and depressurizes the basement), some
investigators have reported (or suggested) that a small 6-  to
10-watt booster fan in the passive piping might provide suffi-
cient marginal increases in system flows and suctions  to
compensate for the  added challenge to the system (Ta85,
Sau91b).

It must be underscored that such small fans cannot provide
adequate suctions and flows to obtain the performance associ-
ated with active systems using the standard 50- to 90-watt
in-line  tubular  fans. The 6- to 10-watt fans can not be relied
upon to make  a soil depressurization  system  function well
unless the  passive system is largely adequate by itself—i.e.,
unless only moderate reductions are needed, and unless the
passive system has demonstrated an ability to provide those
moderate reductions without the booster fan except under
certain challenges. A typical ASD installation with a 90-watt
fan will always provide greater radon reductions more reliably
titan will an installation with a 6- to 10-watt fan.

Whenever a passive system is installed, the mitigator  or
homeowner must be prepared to monitor the radon reduction
performance over a wide range of conditions, to verify that the
system is  not  consistently being overwhelmed by certain
challenges. If the passive system proves to be consistently
inadequate, the system should be activated by adding a stan-
dard 50- or 90-watt in-line suction fan. Since the passive stack
will have been installed up through the house and through the
roof, it should be relatively convenient to install an ASD fan
in the section of the stack that passes through the attic.

If it is found to be necessary to add a fan to an initially passive
system because the passive suction is occasionally over-
whelmed, it is recommended that this fan then be operated
continuously. It should not be suggested that the occupant turn
the fan on only under certain conditions (e.g., certain weather
conditions, or when certain exhaust fans are operating). Such
instructions would require the occupant to be continually alert
to the occasions when the fan should be turned on, and could
result in the fan being left off. Another concern is that, once
installed in the stack, the fan will serve as an obstruction to
flow if it is not operated, thus potentially further reducing the
performance of the passive system.

10.5  installation of Suction Pipes

The suction pipes for passive SSD, DTD, and SMD systems
would be installed in exactly the same manner as has been
described previously in Sections 4.5, 5.5,  and 8.5, for the
corresponding active systems.

The stack must be inside the house if it is to take advantage of
thermal effects. Thus, passive suction pipes will almost al-
ways be installed in heated areas. One major exception will be
passive SMD systems, where the membrane penetrations will
be in a crawl space which may often be unheated. However,
even in this case, the stack will rise through heated living
space overhead, unless a decision is made to passively  vent
through the band joist for convenience (K192), foregoing the
thermal contribution to the passive suction.

10,5.1     Passive SSD Systems

For SSD systems, the need for the stack to be indoors means
that passive SSD pipes would likely be installed only verti-
cally down through the slab indoors (see  Section 4.5.1 or
4.5.3), or horizontally through a foundation wall from inside
the basement (Section 4.5.5). Passive SSD pipes would never
be installed horizontally  through the foundation wall  from
outdoors (Section 4.5.4),  and would preferably not be in-
stalled from inside an unheated garage (Section 4.5.6), unless
one were prepared to sacrifice the thermal contribution to the
passive suction.

Because passive flows are so low, only-a small sub-slab pit
beneath the SSD suction pipe should be adequate to reduce
suction loss resulting from soil gas acceleration to pipe veloc-
ity. See Section 4.5.1, Excavating a pit beneath the slab.

10.5.2    Passive DTD Systems

In existing houses with interior drain tiles emptying into a
sump, passive DTD suction pipes may be installed into the
sub-slab perforated piping at a point remote from the sump. In
such cases, the passive suction pipe(s) will be installed as
                                                      256

-------
discussed in Section 5.5.1 for the case of interior drain tile
loops.

Alternatively, the passive suction pipe could be installed in
the sump, as discussed in Section 5.5.2.

In newly constructed  houses, where perforated piping  has
been installed beneath the slab specifically for radon reduc-
tion purposes, a PVC  T fitting may be installed into the
perforated piping to facilitate connection of the suction pipe
remote from the sump. The use of a T fitting in this manner is
analogous to the approach discussed in Section 5.5.1 for
connecting active suction pipes directly to interior drain tile
loops in existing houses [see Interior tile loops - connecting
the remote suction pipe (direct approach)].

In new houses,  the T fitting would be installed into the
perforated piping at the point at which the passive stack is to
rise up through the house, before the slab is poured. Because
the perforated piping would be completely exposed during
construction, this step would be much easier than described
for the retrofit case in Section 5.5.1, where the T had to be
installed through a hole in the slab.  A vertical stub of PVC
piping would be cemented into the upward-pointing leg of the
T, long enough  to  extend above  the top of the slab. The
perforated piping and the T would be embedded in the gravel
bed which would normally be present; and, especially if the
system were intended from the outset to operate as a passive
system, polyethylene sheeting would be laid on top of the
gravel, to reduce air leakage into the low-flow system. The
slab would then be poured, with the vertical pipe stub protrud-
ing up through the new concrete. The suction pipe would then
be cemented onto the end of the protruding pipe stub using a
straight coupling. The weight of the vertical  suction pipe
would be borne by the T beneath the slab.

 10.5.3    Passive SMD Systems

There are only limited data on passive SMD systems. How-
ever, the limited results (along with intuition) suggest that the
sub-membrane piping/SMD approach is probably the appro-
priate approach for passive systems, due to the low suctions
and flows developed by these systems, and due to the rela-
tively poor communication that might be assumed beneath the
membrane. The individual-pipe/SMD approach might best be
considered only perhaps when there is gravel on the crawl-space
floor to ensure good sub-membrane communication.

Perforated piping would be installed beneath the membrane,
and the vertical suction pipe mounted in this piping, as dis-
cussed in Section 8.5.2 for active systems (see Sub-membrane
piping/SMD approach). The membrane would be completely
sealed  over  the crawl-space floor, as described in Section
8.5.1.

If there is  gravel on the crawl-space  floor, and if the
individual-pipe/SMD approach is being considered, the indi-
vidual suction pipes would be installed as discussed in Section
8.5.2 (see Individual-pipe/SMD approach), and the mem-
brane completely sealed.
10.6  Design/Installation of Piping
Network
Some of the guidance in Section 4.6 regarding the design and
installation of the piping network and the fan for active soil
depressurization systems also applies to passive systems.
However, some of the guidance in Section  4.6  has to be
modified for passive  systems, for several reasons: a) the
passive systems have no fan; b) the passive stack will usually
need to be installed to take advantage of the phenomena that
induce natural suction; and c) passive flows are usually very
low.

Need for  stack to  rise through heated  space.  Most
importantly, the passive stack must rise up inside the house if
the system is to take advantage of the thermal stack effect.
Higher temperatures in the heated living space, relative to the
outdoors, create this natural driving force, which will depend
upon the height of the stack (i.e., the number of stories in the
house) and the  temperature difference between indoors and
outdoors.

Thus, the passive exhaust piping will essentially  always be
designed in general accordance with the interior stack con-
figuration illustrated in Figure 20, except that no fan would be
included. The exterior and garage stack configurations con-
sidered for active systems in Figure 21, where the exhaust
piping penetrates the  band joist, and where the stack  rises
above the eave outside the house or in an unheated garage, are
generally not options for passive systems.

As a result, the details  regarding the design and installation of
the passive piping network would generally be those in Sec-
tions  4.6.2, 4.6.3, and 4.6.4 which are applicable to the
interior stack case. The details in Section 4.6.5  for active
exterior and garage stacks would generally not apply.

There must be a stack if the system is to take advantage of
thermal effects. Passive venting at grade  level, with the pas-
sive exhaust pipe simply terminating outside the band joist,
would lose most of the thermal stack effect and the roofline
wind effects creating depressurization; only grade-level wind
effects  would remain. Thus,  the  passive depressurization
mechanism for system operation, which is already relatively
weak,  would be further weakened, and  the pressure break
mechanism would be  the major remaining operative mecha-
nism. See Section 2.5 for further discussion of the mecha-
nisms.

Some success has been reported achieving moderate radon
reductions with passive SMD systems venting at grade level,
as discussed earlier (K192). However, in the discussion here, it
will be assumed that there will be a passive stack that extends
up through the house.

Horizontal runs in the passive stack.  Ideally, the passive
stack would rise straight up through the house with no elbows
or horizontal runs, to  reduce the equivalent length of piping
and to thus reduce friction losses. Since passive suctions are
so low, it would seem desirable to minimize losses.
                                                       257

-------
However, passive flows will always be low. Flows are often
on the order of 1 cfm, and are rarely as great as 10 cfm. As
discussed in Section 10.3, at 1 cfm, the friction losses will be
very low, on the order of 0.0001 in. WG per 100 ft of piping.
This will commonly be very  low relative to the naturally
induced suctions, except during warm weather with no wind.

Because friction losses should be low, it is not mandatory that
the passive stack rise straight up through the house from the
point where it penetrates the slab. Within reason, elbows and
horizontal runs in heated space should be tolerable if neces-
sary. However, it would seem generally advisable to minimize
such elbows and horizontal runs, and eliminate them if pos-
sible.

It should be noted that the thermal stack effect creating the
passive driving force depends on the net height of the stack
and the temperature differential between  indoors and out-
doors.  Adding horizontal runs inside heated space will  not
decrease this driving force; it will only increase the friction
losses due to the added length of piping. In fact, to the extent
to which the horizontal run provides additional residence time
in the  stack which helps the stack gases come up to house
temperature, the horizontal run could even be beneficial.

Where horizontal runs are required in the passive piping, they
should be made in heated space, such as a heated basement,
rather than unheated space. This will contribute to bringing
the soil gas inside the pipe up to house temperature, and thus
aid the system in achieving the theoretical maximum thermal
driving force for the given  conditions of stack height and
indoor/outdoor temperature differential.

A horizontal run in, e.g., an H/meated basement would not be
detrimental to the system, assuming that the basement is at a
temperature higher than that of the soil under the house. In an
unheated basement, the soil gas inside the piping would still
experience some temperature rise, just not as much as if the
basement were heated.

Horizontal runs in unheated attics would create some limited
reduction in the thermal stack effect. That run  would help
bring the soil gas down to attic temperature, causing  the
limited column height in the attic to be at a lower temperature
during cold weather. On the other hand, during the summer,
when the attic would likely be warmer than the living space
below, this horizontal run in the attic would boost the stack
effect.

Need for pipe insulation in unheated attics. Where the
passive stack passes through an attic, which in cold weather
will be colder than the living space below, the segment of the
stack in the attic might be insulated, to maintain the tempera-
ture of the gases inside the pipe. Maintaining stack gas tem-
perature near house temperature during cold weather would
help increase the thermal stack effect, and will help avoid
moisture freezing in the pipe.

One rationale for insulating the attic portion of the stack is the
same as that discussed previously for avoiding horizontal runs
in the  attic. If the temperature of the gases inside the attic
piping can be maintained near living-space temperature, the
several feet of stack in the attic will add to the total stack
height that is  determining the driving force created by the
thermal stack effect. If, on the other hand, the soil gas inside
the attic  segment of the  stack drops essentially  to outdoor
temperature, the stack segment in the attic will add nothing to
the driving force. But  again, in the summer, the argument
would be reversed, because the hot attic would then tend to
boost the driving force if the stack were not insulated.

Passive stacks should always be insulated in the attic in cold
climates to reduce the risk that condensed moisture will freeze
inside the stack, blocking the stack. Although the soil gas
inside the pipe will be warmer with passive systems than in
active systems at the point where the stack penetrates into the
attic (since it may have risen to indoor temperature during its
residence time in the living-space piping), the gas in passive
systems will have a dramatically lower flow rate. Thus, there
is an increased chance that the gas will have dropped below
freezing at the stack exit, and there will be much less momen-
tum in the gas stream to prevent ice deposition.

Need for cap at discharge point. As indicated in Section
4.6.4 (see Caps on the exhaust stack), a cap is generally not
needed on the exhaust stacks from active systems. Any pre-
cipitation that enters  the stack will be modest relative to the
condensed soil moisture  that will be draining through the
piping anyway. And  the velocity of the exhaust will reduce
the risk of leaves or other light debris from entering the stack.

Passive stacks should also be able to handle any precipitation
that enters. However, because their exhaust velocity will be
much lower than that from active stacks (on the order of 10 ft/
min rather than perhaps 1,000 ft/min), the exhaust will be less
able to deflect light  debris. If sufficient debris  entered the
stack, it could create a serious obstruction in a  system that
may have only a marginal flow to begin with. Accordingly, a
cap may  be more important for a passive system.

A cap such as the one shown in Figure 26B, or such as one of
the others discussed in Section 4.6.4, could be  options. A
hardware cloth screen  (1/4-in. mesh), as discussed in that
section, could be a logical choice (K192). Alternatively, for
passive systems, caps can be considered which may modestly
increase  the passive  suction, such  as draft inducers (C191).
Since flows are so low, it is unlikely that any reasonable cap
design would create an  undue pressure drop unless it became
plugged with ice during cold weather.

Provision for possible  later addition  of fan. Passive
stacks, by definition, are installed without the fans discussed
in Sections 4.4, 4.6.4, and 4.6.5. However, as discussed in
Section 10.4, it may be necessary to add a fan at a later time, if
the passive suction proves inadequate.

Thus, to  the extent possible, the segment of passive stack
through the attic should be installed in a manner to simplify
subsequent addition of a fan. The stack should be located at a
point in  the attic where mere  is  sufficient headroom and
working space to permit subsequent fan installation and main-
tenance, and where the necessary electrical connections can
conveniently be made. For there to be sufficient space for the
fan itself, the stack would have to be at least 12 in. away from
                                                        258

-------
existing walls, wooden members, or utility lines in the attic,
and there would have to be at least 24 in. of vertical clearance
in the stack. Other considerations in stack and fan location are
indicated in  Section 4.6.2 (see Routing considerations with
attic piping runs).

Other guidance regarding piping installation. Much of
the other guidance given in Sections 4.6.2 and 4.6.3 concern-
ing the design and installation of the piping network for active
systems will also apply to passive soil depressurization sys-
tems.

For example, any horizontal piping runs should be sloped
downward toward the suction pipe in passive systems, to
allow condensed moisture to  drain,  and low points in the
piping  should be avoided or  drained. The piping must be
adequately supported, by hangars along horizontal runs, and
where vertical piping penetrates a ceiling into the story above.
Padding might still be inserted between the piping and any
wooden members in passive systems, as is done to reduce
vibration noise in active systems; even though a fan is not
being installed at the outset, one might be installed later.

Insulation of interior piping to reduce sweating during  hot,
humid weather will not be needed in passive systems, since
the low  gas  flows  inside the passive pipe will probably
approach house temperature fairly quickly. Also, this insula-
tion is not needed to help reduce system noise,  since there is
no fan. But this insulation might be considered subsequently,
if a fan is added at a later time.

Use of heated flues as passive stack. The prior discussion
has all focused on the most common case, where the passive
stack  consists of PVC piping that  has  been installed up
through the house.  In a couple cases, investigators have
reported using an existing furnace flue as the passive stack
(Gi90).

Use of a furnace  flue offers two potential advantages. First, it
avoids the time and cost involved in installing a separate stack
up through the house; it uses a pre-existing stack. Second,
when the furnace is operating,  the hot flue gases significantly
increase the draft up the stack and the passive suction. Where
the furnace flue has been used, four-fold and greater increases
in flow rate  (to as  high as 30 cfm) and  five- to ten-fold
increases in suction (to as high as 0.1 in.  WG) have been
reported in passive suction pipes when the furnace was oper-
ating, compared  to when it was not operating (Gi90).

Because of the high temperature of the flue, the PVC piping in
the radon mitigation system was connected to the flue using a
custom  designed length  of  sheet metal ducting, so that the
PVC piping would not exceed its safe operating temperature.

Furnace flues should not be  utilized as passive stacks without
the involvement  of a qualified specialist in heating and air
conditioning systems. The mitigator would have to ensure that
the passive installation would not endanger the occupants,
violate codes, or impact furnace performance or warranties.
Extreme care would be  required to ensure that the passive
system did not result in improper drafting of the furnace.
10.7  Sealing in Conjunction with
Passive Soil Depressurization
Systems

Because passive systems can draw only very low flows, any
leakage of house or crawl-space air into the systems could
significantly decrease the performance of what would likely
be only a marginal system even without the leakage. With
flows usually on the order of 1 cfm in the passive piping, and
with suctions typically only a few hundredths of an in. WG or
less, air leakage through slab and membrane openings would
further reduce  what  is  already  a  weak  sub-slab or
sub-membrane suction and flow field.

Accordingly, the slab sealing steps described in Section 4.7
for active systems become all the more important with passive
SSD and DTD systems. Likewise, the membrane sealing steps
described in Section 8.7 for active SMD systems become all
the more important with passive SMD systems.

10.8  Gauges/Alarms and Labelling

The very low suctions in passive systems (from below 0.01 in.
WG to a maximum of 0.1 in. WG) are at or below the low end
of the range that can be reliably measured with the manom-
eters and pressure  gauges discussed in Section 4.8.1. More
sensitive devices would be required, such as the digital micro-
manometer used for sub-slab suction  field measurements.
These micromanometers can cost $500 and more, a large
fraction of the total cost for the system.

Moreover, the suctions in these systems are likely to be highly
variable, varying as temperatures and winds change, and as
appliances are operated  in the  house. Therefore,  even  if a
sufficiently sensitive gauge were mounted  on the passive
stack, it might be expected to read below the desired range
(and set off any alarm) for some part of the time.

As a result, it does not appear that a pressure gauge is a viable
approach for passive systems. The only option would appear
to be sufficiently frequent radon measurements in  the house
during the year following installation, to confirm that the
system is not being overwhelmed by certain weather condi-
tions or the operation  of certain appliances  (such as the
forced-air furnace or a whole-house fan). This approach does
not provide the assurance that is available with active systems,
where gauges  can give  a reliable continuing  indication of
performance.

The passive systems should be labelled just as active systems
are (Section 4.8.2). Since there is no electrically powered fan
with passive systems, the electric fuse box or circuit breaker
panel would not necessarily be the most obvious location for
the centrally located label mentioned in  Section 4.8.2. Also,
there would be no electric switch that would have to be
labelled, controlling power to the fan.
                                                       259

-------

-------
                                          Section  11
              Post-Mitigation Diagnostic Test Procedures
                     for Soil Depressurization Systems
11.1  General

11.1.1    Purposes of Post-Mitigation
Diagnostics

The purposes of diagnostic testing following the installation
of a soil depressuri/ation system are to:

a)  confirm that the system is operating properly; and

b)  assess what corrective action is required, if the system is
    not operating as desired.

Testing to confirm that the system is operating properly will
always be conducted. Testing to determine necessary correc-
tive action will usually be necessary only occasionally in most
areas of the country.

11.1.2    Diagnostic Tests  that Can Be
Considered
To confirm that the soil depressurization system is operating
properly, mitigators will always: a) visually inspect the instal-
lation;  b) measure suctions  in the system piping;  and c)
conduct (or arrange for) short-term indoor radon measure-
ments, and recommend independent long-term measurements
to the homeowner. Some mitigators also conduct (or arrange
for) long-term measurements directly. EPA's interim mitiga-
tion standards also require testing for backdrafting of combus-
tion appliances (EPA91b).

Other post-mitigation diagnostics would usually be conducted
only if the testing above indicated that the system were not
operating properly. The most common additional testing when
the system is not operating as desired is measurement of
sub-slab suction field extension with the soil depressurization
system operating.

A more complete summary  of the diagnostics that can be
considered, and the cases under which they might be useful, is
given below.

  •  Visual inspection. A visual inspection of the completed
    system, to confirm that all details have been completed
    properly, should be a routine quality assurance step that
    would be completed for all installations.
Suction (and possibly flow)  in  system piping. A
suction measurement in the system piping is a simple yet
crucial measurement that should always be made to verify
that the system is operating in the anticipated range. This
measurement can provide an immediate indication of
whether certain problems have occurred during system
installation, whether there is a defect in the fan, or whether
certain complications (such as unacceptably high air leak-
age through slab openings) might exist This measure-
ment can easily be made at the same time that a pressure
gauge is being mounted on the system piping, and can be
necessary to permit proper marking of the acceptable
system operating range on the gauge to permit interpreta-
tion by the occupant.

If the suction measurement is being made using a micro-
manometer, this suction measurement can easily be supple-
mented with a system flow measurement, using a pitot
tube which comes as an attachment to the micromanom-
eter. Flow data can be an important supplement to the
suction results, for assessing system performance.

Indoor radon measurements. An indoor radon mea-
surement is the primary confirmation that the system is
accomplishing its objective. Radon measurements are
necessary to demonstrate that the  owner's/occupant's
radon exposure has been reduced as anticipated, and that
any warranty (that the system will reduce radon  concen-
trations below a given level) is being met Thus, both the
occupant and the mitigator have a stake in ensuring that
the measurement is completed properly.

EPA's interim mitigation standards  (EPA91b)  specify
that the mitigator must complete a short-term radon mea-
surement  within 30 days after the installation  is com-
pleted, or must ensure that such a measurement is com-
pleted independently  by the homeowner or by an inde-
pendent firm. The mitigator should also recommend that
the client conduct independent short- or long-term mea-
surements at least once every two years.

In view of the improved measure of exposure provided by
a  long-term measurement, the mitigator may  wish to
encourage a long-term  measurement by providing the
occupant with alpha-track detectors for this purpose.

Combustion appliance backdrafting. EPA's interim
mitigation standards require that tests be conducted fol-
                                                   261

-------
lowing  installation  of an ASD  system,  to  determine
whether the mitigation system may be contributing to
backdrafting of combustion appliances. Backdrafting of
clean-burning appliances such as furnaces and water heat-
ers may not be apparent to the occupant, and could have
serious health consequences. If it is found that the system
is responsible for causing backdrafting, the mitigator is
responsible for seeing that the problem is corrected.

ASD systems are most likely to contribute to backdrafting
in cases where the system has high flows, or where the
basement or house are particularly tight (i.e., have a low
leakage area and natural infiltration rate).

If post-mitigation measurements suggest that backdrafting
is occurring, the measurements should be repeated with
the mitigation system off, to confirm that it is indeed the
mitigation system causing the problem. Where backdraft-
ing is found, the problem will often have existed prior to
mitigation. Mtigators might consider backdrafting tests
prior to mitigation to reveal any pre-existing problems, to
avoid having to deal with this issue after mitigation.

Suction field  extension (with  the mitigation system
operating). In cases where the ASD system is not per-
forming as well as expected, the most commonly utilized
diagnostic test is a suction field extension measurement
beneath the slab, inside the block wall, or beneath the
crawl-space membrane, with the mitigation system oper-
ating. This test will identify any areas of the slab, the
wall,  or the membrane which are not being adequately
depressurized by the system, and will thus guide the
installation of additional suction pipes.

In some cases, use of a chemical smoke stick to visualize
flow  down into (or up out of) existing unsealed slab
cracks can be used to provide a rapid, qualitative indica-
tor of suction field distribution. However, the most de-
finitive test would involve quantitative sub-slab suction
measurements with a micromanometer at drilled test holes.

Radon  grab sampling and "sniffing" (with the sys-
tem operating). In certain cases where the system is not
performing adequately, radon grab sampling or "sniff-
ing" near or in potential entry routes might be conducted
as a supplement to suction field extension testing. Grab
sampling could identify those routes which still exhibit
high levels, and hence which may still not be adequately
treated.  Such results could aid in selecting the locations
of additional suction pipes.

This diagnostic may prove to be useful in certain cases
where suction field testing alone might not be expected to
indicate where additional suction pipes (or perhaps addi-
tional sealing steps) may be needed. In particular, where a
SSD,  DTD, or SMD system has been installed in houses
having block foundation  walls, suction field extension
measurements beneath the slab or membrane may not
reveal that a particular wall still requires treatment. Even
suction  measurements inside the block walls might not
reveal this, since suction fields beneath slabs or mem-
branes often will not create measurable suction  fields
throughout a wall. However, radon grab sampling could
show that one or more of the walls still contained dis-
tinctly elevated radon levels, indicating that the mitiga-
tion  system  was not intercepting the radon before it
entered the walls.

In this case, more SSD or SMD suction pipes near the
elevated walls, or the addition of a BWD component to
the system, could be  required. Or, perhaps  particular
openings in the block wall could be shown by the grab
samples to be the primary remaining entry routes into the
house, so  that sealing of those wall-related entry routes
might adequately improve performance.

Smoke flow visualization tests to check for system
leaks. In addition to their potential utility in qualitative
measurement of suction field extension to various open-
ings  in the slab, membrane, or wall, discussed above,
smoke sticks could also be used to check for leaks in the
low-pressure  and high-pressure portions of the system
piping.

Smoke flow into pipe seams on the suction side of the fan
could reveal,  e.g., if certain fittings were not adequately
sealed, or if the suction pipes were not adequately sealed
where they penetrated the slab, possibly explaining the
cause of high flows  and low suctions in the system. On
the pressure  side of the fan, smoke flows away from
seams in exterior  stacks  fabricated of rain gutter
downspouting would reveal where exhaust gases are be-
ing released immediately beside the house, possibly con-
tributing to exhaust re-entrainment.

Tests to check for re-entrainment. In cases where
radon levels  have not  been reduced as expected and
where other diagnostic tests fail to reveal the reason for
the inadequate performance, the mitigator must consider
the possibility that radon is entering the house by some
mechanism other than soil gas flow through the founda-
tion. One  other mechanism could by  re-entrainment of
the ASD exhaust gas.

Where inspection suggests that re-entrainment might be a
possible explanation, a  variety of diagnostic tests could
be considered. These could include chemical smoke visu-
alization tests, to check for leaks in  an exterior stack
beside the house, or temporary modification of the ex-
haust configuration in a manner that should significantly
reduce re-entrainment.  Tracer gas  tests  could  also be
considered, probably using a HCFC or HFC refrigerant as
the tracer (see the discussion in Section 3.5.6). It is likely
that most mitigators  would rarely, if ever, find it neces-
sary to perform tracer gas testing.

Well water  radon analysis. In addition to  exhaust
re-entrainment, another mechanism by which radon might
be entering the house (other than via soil gas flow through
the foundation) would  be as a component in the  well
water.

If suction  field extension measurements and radon grab
sampling fail to suggest why ASD performance is inad-
                                                   262

-------
    equate, a well water radon analysis might be conducted
    following installation of the ASD system. This analysis
    would enable determination of whether radon released
    into the air from the well water might be responsible for
    some significant fraction of the residual airborne concen-
    trations. This diagnostic would be considered in houses is
    served by a well which was not tested for radon prior to
    mitigation (see Section 3.5.2), in areas where elevated
    well water radon levels have sometimes been observed in
    the past.

 •  Flux or gamma measurements. If suction field exten-
    sion and grab radon measurements fail to suggest why
    ASD performance is inadequate; if the house is in an area
    where building materials have sometimes been found to
    be a significant  contributor  to indoor airborne radon
    levels;  and if flux or  gamma measurements  were not
    made prior to mitigation to assess the possible impor-
    tance of building materials  as a source (see Section
    3.5.3): then measurements might be conducted following
    installation  of the ASD system, to determine  if the re-
    sidual radon might in part be resulting from building
    materials.

The following subsections address specific procedures for
conducting  these diagnostics, along with discussion of how
the results of the diagnostic testing can be used to improve
ASD performance where necessary.

11.2  Procedures for  the Visual
Inspection
A visual inspection of the ASD during and immediately
following installation should be a routine quality assurance
step for any mitigator. Any oversights during the installation
process can be discovered at this time, and corrected prior to
their becoming apparent through elevated post-mitigation ra-
don concentrations. Some of the problems  that  might be
uncovered through the visual inspection might not have no-
ticeably impacted the post-mitigation radon measurement, but
could still have had a long-term impact in the form of reduced
system performance and reduced system durability.

The checklist of items to be reviewed during the inspection
will depend upon the particular ASD variation that has been
installed. Some of the key items to check include:

    -  Have the suction pipes been properly mounted into the
      slab, the sump, the block walls, and/or the membrane in
      accordance with the system design, and have the pen-
      etrations been properly sealed?

    -  Has  the  system  piping been properly supported by
      hangars and strapping? Are all pipe  fittings tightly
      cemented (including those associated with the exhaust
      stack), and have the fan couplings been tightly clamped?

    -  Is the piping properly sloped, and have low points been
      avoided (or fitted with a drain), to prevent accumula-
      tion  of condensed moisture which could  block air
      flow?
   -  Has the fan been mounted and wired properly? Is it
      running quietly? If it is not operating  quietly, what
      steps need to be taken to reduce noise?

   -  Is the installation neat, and does it reflect good work-
      manship? Has the work area been cleaned and restored
      adequately?

   -  Has a  fire break been  installed in cases where the
      piping penetrates a fire wall into an adjoining garage?

   -  Does the fan exhaust discharge in a manner which will
      reduce re-entrainment?

   -  In crawl-space SMD systems, has the membrane been
      installed and sealed properly?

   -  Have the slab, wall, and membrane openings which
      were identified for sealing, in fact been sealed properly
      at all locations?

   -  Has an appropriate gauge and/or alarm been installed
      oh the system piping? Has the gauge been properly
      marked to advise the occupant of the proper  system
      operating range?

For some of the steps in this inspection—e.g., for  assessing
the tightness of the pipe fittings, or the effectiveness  of slab
seals—a chemical smoke stick might be used, as discussed in
Section 11.7, to assess whether  air is being drawn into the pipe
joint or down into the apparently sealed slab opening when
the ASD system is operating.

11.3  Procedures for Suction (and
Flow) Measurements in the System
Piping

A quantitative measurement of the  suction in the  system
piping is always necessary, in order to confirm that the system
is indeed operating in the anticipated, typical range. Usually,
this measurement will be made at the point where the gauge or
alarm is to be mounted in the piping, and will be performed at
the time that the gauge or alarm is being installed. The results
from this measurement will help the mitigator mark the sys-
tem operating range on the gauge, and set the trigger value for
the alarm.

Operation at unexpectedly low suctions and high flaws.
If the system is operating at unexpectedly low suctions and
unexpectedly high flows, this could be suggesting  that there
are leaks in the system piping. Or, it could be indicating larger
than expected amounts of air short-circuiting into the system
through leaks through openings in the slab or membrane.
Thus, further smoke-stick or other diagnostics may be war-
ranted to identify the source of these leaks, and further sealing
steps could be warranted.

What constitutes "unexpectedly low  suction"  will vary, de-
pending upon the characteristics of the house and soil, the
ASD variation, the performance curve of the fan being used,
and the distance of the measurement point from the fan (i.e.,
                                                      263

-------
the amount of friction loss in the piping between the fan and
tfie measurement location). It is recommended that these
piping suction measurements be made (and that any pressure
gauge be located) reasonably near to the point at which the
suction pipe penetrates the slab, wall, or membrane, to pro-
vide the best measure of the suction actually being applied to
the soil. Making the measurement too close to the fan may
sometimes give misleadingly high suction readings.

In typical SSD and DTD systems with the 50- to 90-watt
in-line tubular fans, suctions lower than roughly 0.25 to 0.5 in.
WG  near the slab penetration would usually be "low" in
essentially any house. With 90-watt fans, and/or with tight
soils or reduced sub-slab communication, the suction might
generally be considered "low" if it fell below the upper end of
that range. With 50-watt fans, and/or with highly permeable
native soils where flows  can be expected to be  high, the
suction might not be considered "low" until it fell toward the
lower end of that range.

Of course, the  actual value at which  suctions start being
considered "unexpectedly low" in a given case will depend
upon the particular house and geographical area. For example,
pre-mitigation diagnostics or prior experience in an area might
suggest that low flows should be expected in a  particular
house, and that a 90-watt fan would typically be expected to
generate over 1 in. WG in that house. Under those conditions,
a suction below 1 in. WG might be considered "low" for that
fan in that house, and thus prompt further analysis.

With the in-line radial blowers discussed in Section 4.4.2,
"low" suction near the slab penetration in SSD and DTD
systems might be expected to be at the upper end of the 0.25 to
0.5 in. WG range cited above for the tubular fans. With the
high-suction/low-flow fans discussed in Section 4.4.3, "low"
suction could be more than 1 in. WG.

There is less experience with SMD systems. However, since
the flows are generally comparable to those in SSD systems
(see Section 8.3), the range would be expected to be about the
same as that for SSD systems. That is, with the 90-watt fans,
suctions much below roughly 0.25 to 0.5 in. WG near the
membrane penetration might be considered "low." Again, the
actual value at which suction would start being considered
"low" would depend on a variety of house, geological, and
system characteristics, as mentioned earlier.

With the high flows in BWD systems, suctions in BWD
piping near the wall penetration can be below 0.1 in. WG with
the 90-watt fans. As a result, quantification of "low" suctions
for BWD systems is not possible.

Operation at unexpectedly high suctions and low flows.
The preceding paragraphs  address the case where the system
is operating at unexpectedly low suctions and unexpectedly
high flow. If, on the other hand, the system is operating at
unexpectedly high suction and unexpectedly low flow, this
could be suggesting that the piping is blocked in some man-
ner. Perhaps a SSD pipe was installed directly over a perim-
eter  footing,  if the pipe is located near the perimeter. Or
perhaps the pipe has dropped down to the bottom of the
sub-slab pit and is embedded in the soil there.
In houses with marginal or poor sub-slab communication,
high suctions and low flows will be common. Thus, for a
suction to be unexpectedly high, it will have to be near the
maximum suction capability of the fan under communication
conditions which would lead one to expect that the fan should
wo? be near maximum suction.

Since high suctions will always be suction  near the fan's
maximum capability, it could be necessary to make a flow
measurement to supplement the suction measurement, to bet-
ter distinguish whether the high suctions and low flows are
outside of what would be an expected range. Unless the house
has poor sub-slab communication, flows below roughly 5 to
10 cfin in a SSD pipe with a 90-watt fan could be suggesting
some blockage. However, even this is not a rigorous indicator
of blockage. Flows below 5 cfm have occasionally been seen
in SSD systems with fair communication, with no pipe block-
age,  due to a combination of tight slab and tight soil which
reduces air leakage into the system.

Operation at low suctions and low flows. In some cases,
low suctions and low flows might be found simultaneously in
a particular installation. This result could be indicating that
there is a defect in the fan, in the switch or speed controller in
the fan wiring, or in the wiring itself. It could also be suggest-
ing that there is a leak in  the system piping at some point
between the measurement location and the fan.

This type of result  underscores the value of making flow
measurements as well as suction measurements during these
diagnostics. It should not be automatically assumed that, just
because suctions are low, flows must be high.

Measurement procedures. Where a pressure gauge is be-
ing installed on the system, the piping suction measurement
will commonly be made using that manometer or gauge. See
Section 4.8.1, Pressure gauges. The gauge is leveled and/or
zero  adjusted, and flexible tubing from the appropriate termi-
nal of the device is inserted through a small hole drilled in the
side of the piping for this purpose. The gap between the tubing
and the hole through the piping is made air-tight with a rubber
seal around the tubing. The suction in the piping is then read
directly from the gauge.

Where a pressure-activated alarm is being mounted on the
piping, this alarm does not provide a quantitative measure-
ment of suction. Hence, this measurement will have to be
made with a separate device. If a micromanometer has been
obtained for sub-slab suction measurements, this device (set
on the 0 to 2 in. WG scale) can conveniently be used to make
this measurement. Again, tubing from the micromanometer
would be inserted through the small hole in  the side of the
pipe, with the residual gap sealed by a rubber grommet.

Where a micromanometer is available, the flows in the piping
can also be conveniently measured, using a pilot tube attach-
ment that can be obtained with the micromanometer. The flow
results can provide important additional perspective regarding
how the system is operating, and are generally recommended.
                                                      264

-------
11.4  Procedures for Indoor Radon
Measurements

11.4.1     Initial Short-Term Radon
Measurement

Because the residual indoor radon concentration is the pri-
mary measure of the success of the system, EPA's interim
mitigation standards (EPA91b) require that a short-term radon
measurement be initiated no sooner than 24 hours after the
installation of the system, and be completed within 30 days
after installation. This measurement is intended to provide a
rapid indication of whether the mitigation system appears to
be operating properly.

According to the standards, the  mitigator must either com-
plete this short-term  measurement directly, or must ensure
that the homeowner  or a separate measurement firm com-
pletes the measurement independently. If the required mea-
surement is made independently, the results  must be provided
to the mitigator for the mitigator's records. Even if the mitiga-
tor conducts the measurement directly, he/she should recom-
mend that the homeowner have a measurement made indepen-
dently.

The short-term measurement can be  conducted using any
appropriate measurement device specified  in EPA's radon/
radon  decay product measurement protocols, and must be
completed in accordance with  these protocols (EPA92d,
EPA93). These protocols require, among other things, that a)
house be closed during the measurement and for a period of at
least 12 hours prior, and b) that the measurement be made on
the lowest lived-in level of the house.

To be conservative, a  mitigator might choose to make the
measurement in the lowest livable level (e.g., an unoccupied
basement), if this level is below the  lowest lived-in level.
Radon concentrations commonly tend to be highest on the
lowest level of the house.

Different mitigators employ different approaches in complet-
ing this measurement. Many mitigators leave a charcoal de-
tector with the occupant,  which the occupant is to deploy at
the appropriate time and then return to an independent labora-
tory for analysis after a 2- to 7-day exposure period. At least
one mitigator leaves an electret ion  chamber with the occu-
pant, which the occupant is  to deploy at the designated time
and then ship back to the mitigator for reading after 2 to 7
days'  exposure.

More  reliable deployment and retrieval of  the measurement
devices would result if the mitigator handled these steps
directly. However, this would necessitate two additional visits
to the house (one each for deployment and for retrieval),
which could be time-consuming if the house were located far
from the mitigator's offices.

At least one mitigator reports completing the short-term mea-
surement using a continuous monitor. The continuous monitor
is deployed immediately after installation, and is retrieved and
read after measurement period  of  48 hours or longer. In
interpreting the  results, the first 24 hours of readings are
disregarded, since the measurement period is not supposed to
begin until 24 hours after system activation.

Alternatively, the mitigator could arrange to have the mea-
surement completed by an independent measurement firm.
While more expensive, this approach would ensure the home-
owner that the measurement is independent of the mitigator.

11.4.2    Subsequent Radon
Measurements

Following the initial short-term measurement, any subsequent
short- or long-term measurements of the indoor radon levels
are the responsibility of the homeowner or occupant. Accord-
ing to EPA's draft mitigation standards, the mitigator should
recommend that the homeowner make a radon measurement
at least once every 2 years.

Because a long-term measurement will provide a better mea-
sure  of the occupant's  radon exposure, the  mitigator might
wish to encourage a long-term measurement by providing the
occupant with alpha-track detectors for this purpose. These
detectors could be deployed for 3 to 12 months. Locations for
the detectors would be selected with emphasis on the parts of
the house where the occupants spend most time.

11.5  Procedures for Checking
Combustion Appliance Backdrafting
EPA's interim radon mitigation standards require that a test be
made following installation of an ASD system, to determine
whether combustion appliances are backdrafting. Backdraft-
ing would occur if the house were being depressurized to the
extent that suction in the house overcame the thermal effects
that draw the products of combustion up the flues, causing the
combustion products (including some hazardous carbon mon-
oxide) to flow into the house instead of up the flue.

Backdrafting will probably not commonly be caused by ASD
systems with typical exhaust flows. Most often when back-
drafting is encountered, the backdrafting problem will have
existed prior to the installation of the mitigation  system.
However, once the system is installed, the system can worsen
the pre-existing problem, and the mitigator will have to ad-
dress this issue in some manner. Thus, some mitigators may
wish to include  backdrafting  measurements  in  the
pre-mitigation diagnostics, so that the problem can be identi-
fied and discussed with the homeowner beforehand.

Combustion appliance backdrafting would be most likely to
be created by the ASD system in the following cases where:

   -   the depressurization system flows are particularly high,
      suggesting that significant amounts of air are being
      drawn out of the areas (such as the basement or furnace
      room) where combustion appliances are located. BWD
      systems, which can have flows as high as 20Q  cfm,
      much of which has likely come from inside the house,
      thus offer a greater risk than would, for example, a SSD
      system exhausting 50 cfm.
                                                     265

-------
   -  the basement or furnace room is particularly tight, so
      that even a limited amount of exhausted air can create
      some depressurization in the areas where the appli-
      ances are located. In some cases, such tight structures
      may have had only marginal draft to begin with, and a
      slight additional basement depressurization contrib-
      uted by the ASD system might be sufficient to create
      backdrafting.

The risk of backdrafting in  a given house can be increased by
a variety of other factors associated with the design, construc-
tion, and  operation of the house and flue system (Ang92,
Kt92, Ne92). These other factors can reduce the draft in the
flue prior to mitigation, making the system  more subject to
backdrafting due to the incremental effects of the ASD sys-
tem. In  addition to the tightness of the basement or furnace
room, these other factors include: improper connection of the
flue to  the combustion  appliance; other problems in flue
design or  installation; partial blockage of the flue by debris;
and other exhaust appliances in the house (in  addition to
ASD) that can contribute to house depressurization. The risk
of backdrafting will also be greater when the flue temperature
is low, a condition that will exist when an appliance cycles on
after having been off for an extended period.

Backdrafting would not be a concern  in cases where tradi-
tional natural-draft combustion appliances are not present.
Such cases would include, e.g., houses having all-electric
space and water heating systems, or houses having advanced
high-efficiency, closed-combustion furnaces, water heaters,
and fireplaces. Also, the risk of backdrafting would be greatly
reduced with induced-draft appliances, which may have suc-
tions of 0.06 in. WG or greater in the flue.

It is re-emphasized that if the following backdrafting tests are
initially conducted with the ASD system operating, and if the
tests suggest a risk of backdrafting,  the mitigator  should
repeat the tests with the ASD system off, to assess the degree
to which the ASD system is responsible for (or is contributing
to) the problem.

11.5.1    Backdrafting Test Procedures
for Mitigators

A variety of  procedures  have  been identified  to test for
combustion  appliance backdrafting (CMHC88, NFGC88,
ASHRAE89, TEC92). Many of these procedures can be con-
ducted using equipment that many mitigators will have avail-
able (chemical smoke and a micromanometer).

Establishing "worst-case" conditions. All of the proce-
dures involve initial steps to place the house under the condi-
tions which  have the greatest likelihood of creating
backdrafting, to provide a worst-case scenario. These steps
generally include:

   -  Closing all windows and doors to the outdoors, to
      prevent draft air from being drawn in through open
      windows and doors.

   -  Closing interior doors. If the combustion appliance is
      in a basement or furnace room, closing interior doors
      will prevent the appliance from drawing draft air from
      elsewhere in the house. This is important since a base-
      ment or furnace room will commonly be tighter than
      other parts of the house.

   -  Turning on all exhaust fans, including, e.g., kitchen and
      bathroom exhaust fans, the clothes drier, any attic fan,
      and any whole-house exhaust fan. Sometimes the cen-
      tral furnace fan in forced-air systems is operated as
      well. Where the objective is to assess the effect of the
      ASD system, the ASD fan would also be operating. It is
      noted that operation of a whole-house exhaust fan with
      the windows closed in contrary to the manner in which
      such fans are normally operated;  hence, the house
      might be depressurized to an atypically great extent,
      and the backdrafting test thus might have an increased
      tendency to give a false positive result.

   -  In some cases, turning off the combustion appliances
      for awhile, to allow the  flue to cool,  thus further
      reducing the initial draft when the appliance is turned
      back on.

Worst-case conditions are established for these tests because
it is recognized that the conditions potentially causing back-
drafting will be varying  over  time. Due  to the potentially
lethal threat posed by carbon monoxide backdrafting into the
house, it is considered wise to take the most conservative
approach, which is to  utilize  the most challenging  set of
conditions.

Having established these worst-case conditions, the types of
measurements  that are then performed depend upon the spe-
cific test procedure.

Simple smoke visualization test. One of the simplest tests
(NFGC88) involves smoke visualization testing at the draft
hood in the appliance flue with the combustion appliance
operating and the flue hot, and with all of the exhaust fans
operating. (Draft hoods are also referred to as down draft
diverters.) If the smoke flow is distinctly up into the hood and
up the flue, a positive draft under these worst-case conditions
is  qualitatively demonstrated,  and backdrafting would not
seem to be a problem.

Although this test is specifically designed for gas-fired appli-
ances having draft hoods (draft diverters), a similar smoke
visualization test could be envisioned at the barometric damper
of oil-fired units, or near the combustion zones of fireplaces,
wood stoves, kerosene heaters, etc. Consistent with the guid-
ance given in Sections 3.2 and 3.3.1, this document recom-
mends that heatless chemical smoke be used, to avoid the
thermal effects and fire hazard potentially associated with the
use of a match, cigarette, or punk stick as the smoke source.

Given the ease with which this test can be conducted, mitiga-
tors may often find it convenient to begin backdraft testing
using this approach. More extensive procedures, such as those
discussed below, might then be considered if the results from
this initial testing suggested that they are warranted.
                                                       266

-------
More  comprehensive procedures—house depressur-
ization test. Two of the other procedures (CMHC88, TEC92)
include more extensive testing, although still for the most part
utilizing equipment that many mitigators will have available.

One of these procedures (CMHC88) begins with a paper
screening analysis to help assess whether the testing described
below is necessary.

The initial testing with these comprehensive procedures  in-
volves measurement of the pressure differential across  the
house shell, between the outdoors  and the .area where  the
combustion equipment if located. Such pressure differential
measurements would be made as discussed in Section 3.5.4
(see Procedure forpressure differential measurements across
the house shell), with care to  avoid interference from wind
effects. In one of these procedures  (TEC92), these pressure
measurements are to be made under three sets of conditions,
to ensure that worst-case conditions are defined: all exhaust
fans operating, central forced-air air handler off; all exhaust
fans plus the air handler on; and exhaust fans off, only  air
handler on. See the citations for further details.

According to these two procedures, the threat of backdrafting
is assessed based upon the house depressurizations observed
under the worst-case conditions. For example, 0.012 in. WG
depressurization of the house relative to outdoors is identified
as the point at which one might become concerned about
backdrafting of fireplaces having chimneys on the exterior of
the house; 0.016 in. WG for oil-fired furnaces or boilers; and
0.020 in. WG for gas-fired furnaces. These values are lower
than the suctions that would be expected in a well-designed,
properly operating  flue for each respective type appliance
during cold weather. By this approach, it is assumed that as
long as the measured depressurization in the house is less than
expected draft suction in the  flue,  backdrafting should  not
occur.

However, the house depressurization alone is not a sufficient
indicator of whether backdrafting might occur. The depressur-
ization values just cited assume that the flue is  drawing
normally during  cold weather. If the flue is improperly  in-
stalled or partially blocked, or  if the weather is mild,  the
natural draft can be lower than that assumed in identifying
those  values,  and backdrafting could occur even  if  the
worst-case house depressurization is less  than 0.012 to 0.02
in. WG.

More comprehensive procedures—smoke visualization
test. Accordingly, each of the procedures just cited (CMHC88,
TEC92) include additional tests to assess  the actual draft in
the flue under the worst-case conditions.

Both procedures include smoke visualization tests. In these
cases, the combustion appliances are all turned off to allow
the flue to cool. With exhaust fans operating, each appliance is
then turned on one at a time, with the flue being allowed to
cool between appliance tests. Backdrafting is assessed  for
several minutes for each appliance, by chemical smoke testing
at the draft hood, barometric damper, and/or combustion
zone. The flame is also observed for signs of roll-out.
If these tests indicate that backdrafting/spillage occurs  for
more than 30 seconds (TEC92) to 1 or 2 minutes (CMHC88)
after the appliance is turned on—i.e., if the flue does not heat
up sufficiently in that time to ensure proper draft—this result
suggests a potential backdrafting problem.

More comprehensive procedures—direct suction mea-
surements  in flue.  The smoke testing provides only a
qualitative indication that flue gases are (or are not) in fact
moving up the stack. Both of the cited procedures supplement
this smoke testing with quantitative measurements of the draft
in the  flue of traditional natural-draft appliances, with  the
exhaust fans operating.  This  test would provide the most
definitive  indication whether proper drafts  are being main-
tained despite the worst-case conditions.

To make the measurements, a test hole would be drilled about
two feet downstream of the draft hood or barometric damper.
The  suction in the flue  at that point (relative to the house
where  the appliance is  located)  can  be measured using a
suitable pressure gauge. Pressure measurement devices (draft
testers) are manufactured specifically for the purpose of mea-
suring  drafts in hot flues, consistent with heating industry
practice. With proper taps, considering flue temperatures, the
pressure gauges used for  sub-slab pressure measurements can
also be adapted for this use.

The acceptable suction in the  flue relative to the house will
vary depending upon the outdoor  temperature (CMHC88,
TEC92). When the outdoor temperature is below 20 °F,  the
suction in the flue (relative to the house) would be expected to
be about 0.020 in. WG or greater. But when  the outdoor
temperature is above 80 °F, a flue suction as low as 0.004 in.
WG may be all that can be expected.

More  comprehensive  procedures—carbon monoxide
(CO) measurements. The more comprehensive procedures
in References CMHC88 and TEC92  also include measure-
ments for CO, in the house and in the flue. High levels in the
house could be indicating that combustion products are enter-
ing the house, perhaps due to backdrafting. High levels in the
flue could be indicating  that the draft is insufficient to draw
adequate combustion air  into the flame zone.

Carbon monoxide levels in the living area should generally be
below  2 ppm (TEC92).  These procedures recommend that
maximum levels in the living  area be  limited to those speci-
fied  in EPA's Ambient Air Quality Standards (EPA87c): 9
ppm CO averaged over an 8-hour period, and 35 ppm aver-
aged over a 1-hour period. Levels should never exceed 200
ppm (TEC92). Carbon monoxide  levels in the flue upstream
of the draft hood or barometric damper should be no greater
than 100 ppm after 5 minutes of operation (TEC92).

Most mitigators  will not have the  equipment to make CO
measurements.

More  comprehensive procedures—general comments.
The more comprehensive procedures discussed in the preced-
ing paragraphs (CMHC88, TEC92)  have been designed pri-
marily for professionals in the field of combustion system
evaluation. The procedures have been developed to enable an
                                                       267

-------
extensive diagnosis of the combustion appliance and flue
system. The diagnosis addresses a broad range of issues, e.g.,
whether the flue is blocked  or improperly connected,  or
whether CO is entering the circulating house air through a
cracked heat exchanger. Although very important, some of
these issues extend beyond the immediate goal of the radon
mitigator, which is to ensure that a mitigation system does not
create or exacerbate a backdrafting situation.

On tliis basis, the question might be asked whether a mitigator
drawing from these more comprehensive test procedures must
in fact conduct all of the steps in these more comprehensive
procedures, if these procedures are to be used. Or, on the other
hand, might it be sufficient to carry out only selected compo-
nents of these extensive procedures? Definitive guidance on
this question is not possible at this time.

Simple measurement of flue temperature. A simple test
based on flue temperatures has been identified for combustion
systems having draft hoods (ASHRAE89).

The objective of this procedure is to ensure that the draft in the
flue  is sufficiently great such  that the amount of house air
drawn  in through the draft hood to dilute the combustion
products by at least 40%. This  determination is made using a
simple calculation based upon temperature measurements up-
stream and downstream of the draft hood, and in the house.

For this testing, all exhaust fans  are turned on, and outdoor
doors and windows are closed, as discussed previously (see
Establishing "worst-case" conditions).  The positioning  of
interior doors is not specified. The flue temperature measure-
ments are usually done with the flue hot. The citation specifies
that  the testing be  conducted when  natural infiltration is
expected to be low, i.e., when the indoor-outdoor temperature
differential  is no greater than 30 °F^and when the wind
velocity is no more than 5 mph.

While the test is fairly simple, it requires temperature mea-
surement equipment that many mitigators may not have.

11.5.2    Backup Backdrafting Test
Methods for House Occupants
Several inexpensive devices are available that can be left in
the house to alert the  occupants when backdrafting occurs.
These devices detect either CO in the room, or the tempera-
ture near the draft hood.

These devices  can be useful as a backup to the mitigator
testing options described in Section 11.5.1, alerting the occu-
pants if potential problems arise  at some point following
installation. However, they should not be considered as  a
substitute for testing by the mitigator, for two reasons. First,
the CO detection devices are generally triggered by CO
concentrations of 100 ppm and higher—a level higher than
that  of concern  in EPA regulations. Second, the devices
depend upon monitoring and replacement by the occupant;
failure  of the occupant to inspect and replace the devices
would render some of the devices useless.
Electronic carbon monoxide detectors. These monitors
operate in a manner similar to smoke detectors, sounding an
audible alarm when CO levels exceed a given concentration
for a given period of time. The higher the CO concentration,
the more quickly the alarm sounds. The alarm would typically
sound  after the detector had been exposed to 100 ppm for
perhaps an hour or longer.

These devices can plug into a wall outlet. Underwriters Labo-
ratories has  recently approved  the first of these devices.
Because they give ah audible alarm and appear to require a
minimum of occupant maintenance, these devices may be the
most reliable (although most expensive) of the alternative
backup techniques for monitoring potential backdrafting, as
long as they are not disabled by the occupant.

Passive carbon monoxide detectors. Passive CO detectors
contain a CO-sensitive dot which changes color, from a light
color to gray or black (depending on concentration), when
exposed to CO. The rate of change depends upon the CO
concentration; a concentration of 100 ppm would cause some
color change in perhaps 15 to 45 minutes, while a concentra-
tion of 400 ppm may cause a response in 2 to 4 minutes.

The device would be mounted at a convenient location in the
house.  The occupants would have to monitor the color of the
device. For this reason, the electronic devices, which attract
the occupant's attention with an audible signal, may be more
reliable.

The detector is reusable for a period of time, returning to its
original light color when the CO levels drop. However, the
dot will ultimately remain dark or bleach out, and the detector
will have to be replaced. The frequency of replacement may
be once every few months. Although these detectors are
relatively inexpensive (perhaps $3 to $4), it might be unrealis-
tic to expect the average homeowner to regularly replace them
over periods of years.

Passive temperature detectors.  One passive temperature
detector on the market uses a temperature-sensitive dot, and is
mounted beside the flue just below the draft hood. When the
flue is drafting properly, relatively cool house air is being
drawn past the device and up into the draft hood. Under these
conditions, the device is at a low temperature, and it displays
a white dot. However, if backdrafting occurs, the flow through
the hood will consist of hot combustion products flowing out
of the flue into the house. Under these conditions, the dot in
the device turns black.

As with the passive CO detector, the occupants would have to
monitor the  device. This temperature-based device is not
reusable, so that it would have to be replaced each time it was
exposed to high temperature.

11.5.3    Steps Required When
Backdrafting Is Observed

If  combustion appliance backdrafting is observed after the
mitigation system  is installed, the first step will generally be
to repeat the backdrafting measurement with the ASD system
off. A  system-off test will identify the extent to which the
                                                      268

-------
ASD system may be responsible for, or is contributing to, the
problem.

In cases where the ASD system is in fact responsible for
creating a backdrafting problem that did not exist before, one
might consider modifications to  the system. If much of the
house air being drawn into the system may be coming from
certain identifiable unsealed openings  in the slab or  wall,
sealing of those openings, if practical, might reduce house air
flow into the system sufficiently  to take care of the problem.
Or, the ASD fan might be turned down to  lower power,
drawing less air out of the house (but possibly also reducing
radon reduction performance).

However, even in the (probably  infrequent) cases where the
ASD system is the sole source of the problem, modifications
to the ASD system might not be a reliable, permanent solu-
tion. The seals to reduce air flow out of the house may fail
over time. Or, the fan operating at reduced speed may be
turned back  up to full power at  a later time by an occupant
unaware of the backdrafting threat.

A further consideration is that, in the large majority of cases
where backdrafting is observed, the problem will be that
backdrafting was occurring even  before the ASD system was
installed, or else, that the draft was marginal prior to system
installation. In these cases, the mitigation system will only be
contributing to a pre-existing condition, and no adjustments to
the ASD system could ever be  a complete solution to the
problem.

Thus, in most cases, the appropriate solution to backdrafting
will be some step (or combination of steps) to improve draft.
Such  steps could include  providing an outdoor  source of
combustion and draft air for the  combustion appliance, seal-
ing leaks in the cold-air return ducting, modifying or unblock-
ing flues, etc. In many cases, such steps will be beyond the
expertise of a radon mitigator, and will require the services of
a professional building diagnostician or a heating, ventilat-
ing, and air conditioning contractor. Mitigators and home-
owners who do not have expertise in these areas would be
well advised not to undertake these types of steps on their
own.

One step that is commonly taken to provide additional com-
bustion and draft air consists of installing an opening through
the house shell near the appliance, covered  with a grille.
Installing such an opening would be analogous to perma-
nently opening a window. The  opening should provide an
entry route for fresh air to be drawn in to provide the needs of
the appliance, reducing  basement depressurization and in-
creasing the natural infiltration rate. Installation of such an
opening would create some increase in heating and cooling
costs.

However,  even this apparently  simple step  is not always
straightforward. If the opening were on the downwind side of
the house, it is theoretically possible that rather than helping
alleviate the problem, the opening could exacerbate the prob-
lem, with the wind effects increasing basement depressuriza-
tion. Or, if ducting is going to be installed to direct the air
from  the opening toward the location of the combustion
appliances, calculations could be necessary to ensure that the
shell opening and the ducting are the correct size to provide
the amount of makeup air needed. Other problems that could
result from this  step include problems with moisture and
freezing. These possible complications are the reason why
steps to correct backdrafting problems should be left to pro-
fessionals in that field.

According to EPA's interim standards, a radon mitigator is
only responsible for correcting a "... backdrafting condition
caused by the installed mitigation system . . . ." (EPA91b).
The mitigator is not responsible for correcting a pre-existing
backdrafting problem not caused by the system. However,
once the mitigation system is installed, the mitigator may
become involved with the problem, even if the system is not
responsible.  Thus,  in very cold climates where houses are
commonly built with low infiltration rates, or in tight houses
in any climate, mitigators may wish to include backdrafting
tests prior to mitigation, so that the issue can be faced before
mitigation is undertaken.

11.6  Procedures for Suction Field
Extension  Measurements
When an ASD system is  not performing as well as expected,
one reason will often (although not always) be that the system
is not adequately depressurizing the region that it is supposed
to be depressurizing—i.e., the sub-slab, the block wall cavi-
ties, or the sub-membrane. Thus, the most valuable diagnostic
test to conduct when ASD  performance is inadequate is the
measurement of the suction field extension being maintained
with the system operating.

Even when an ASD system appears to be performing satisfac-
torily during mild weather, quantitative suction field exten-
sion measurements (Section  11.6.2 below) may  still be of
value, to suggest whether the system is maintaining sufficient
depressurization to prevent being overwhelmed during cold
weather.

 11.6.1    Qualitative Check with
 Chemical Smoke

One simple approach for obtaining a quick, qualitative assess-
ment of suction field extension is to utilize chemical smoke to
visualize flow down into (or up through) existing unsealed
openings in the slab, wall, or membrane with the ASD system
operating. Such openings could  include, e.g., the wall/floor
joint around slabs, mortar joint cracks in block walls, or
unsealed seams (if any) in the crawl-space membrane.

Slight flows through these openings can  best be observed
when only a small amount of smoke is  released near the
opening, using a flashlight to back-light the smoke. Smoke
flows down into the opening would suggest that the sub-slab
(or wall cavity or sub-membrane) is being depressurized at
that location by the system. No flow  down (or flow upward)
would suggest that the depressurization is weak or non-existent
at that location, or that the house is still depressurized relative
to the sub-slab.
                                                       269

-------
During this testing, it would be advisable to challenge the
system, to the extent possible, by turning on  any exhaust
appliances (such as the clothes drier and kitchen exhaust fan)
that might be routinely depressurizing the house under normal
living conditions. If a central furnace fan is present in the
basement, it should be turned on, also.

If there are accessible openings to enable smoke-stick testing
around the entire slab (or wall or membrane), this test would
generally indicate whether the sub-slab is being depressurized
everywhere, or where the sub-slab is being depressurized
weakly or not at all. This information could help guide the
placement of additional suction pipes. Unfortunately, suitable
existing openings may not be present or accessible  at all
locations, limiting the ability of the smoke-stick approach to
survey the entire slab. Smoke flows may be obscured by air
currents in the room. And the approach  will not provide a
quantitative measure of the depressurization being achieved,
limiting the ability to assess whether the depressurization
present at the time of measurement might likely be over-
whelmed when faced with a challenge from weather or appli-
ance usage.

11.6.2    Quantitative Measurement with
Micromanometer

Because of the limitations of the smoke-stick approach, it will
generally be necessary to make quantitative suction measure-
ments with a micromanometer whenever there is doubt about
whether the suction field is extending adequately.

Test holes would be drilled through the slab (or wall face or
membrane) to permit these suction measurements to be made
at specifically selected locations, to more reliably identify the
locations that are not being reached by the suction field. Such
non-deprcssurized locations would be sites where additional
suction pipes might be warranted.

The procedure for this post-mitigation suction field diagnostic
will  be very similar to that described in Section 3.3 for
pre-mit5gatkm diagnostics. The major  difference is that the
suction field is now being established by the mitigation fan
rather than by a vacuum cleaner.

 *  Equipment and materials required

   -  The equipment and materials needed for post-mitigation
      diagnostics will be generally the same as those listed in
      Section 3.3.1 for pre-mitigation qualitative suction field
      diagnostics.  However, since the suction  field is now
      being generated by the mitigation fan, the following
      items will not be needed:

      -  The industrial vacuum cleaner.

      —  Vacuum cleaner exhaust hose.

      --  The 1.25-in. masonry drill bit, which was to drill the
         slab hole needed to accommodate the vacuum cleaner
         nozzle.
-   The rotary hammer drill for drilling 1/4- to 1/2-in. test
   holes might not be needed, if suitable test holes already
   exist as a result of earlier, pre-mitigation suction field
   extension testing. However, even if pre-mitigation test-
   ing  was conducted,  the drill might  still be needed.
   Additional holes  might be required in different loca-
   tions for the post-mitigation testing, and previous test
   holes that were mortared shut will need to be reopened.

-   If a digital micromanometer is  used to quantitatively
   measure depressurizations, this testing will provide
   quantitative suction field extension data as discussed in
   Section 3.3.2. TheMagnehelic® gauge and the vacuum
   cleaner speed controller, identified in Section 3.3.2, are
   not necessary. The Magnehelic® gauge and the speed
   controller were needed to ensure that the vacuum cleaner
   was operated to best simulate a mitigation fan; since an
   actual mitigation fan is being used for the post-mitigation
   testing, this is no longer an issue.

 Test procedure

-   The test procedure for the post-mitigation suction field
   testing would be essentially the  same as that described
   in Section 3.3.1.  However, the  steps  in Section 3.3.1
   associated with the vacuum cleaner would no longer be
   applicable:

   — The location for the vacuum cleaner suction hole
     would no longer have to be selected. Suction will
     now be drawn at the mitigation suction pipe loca-
     tion.

   — The 1.25-in. hole through the slab for the vacuum
     nozzle no longer has to be drilled, nor the vacuum
     mounted in the slab.

-   The location for  the 1/4- to 1/2-in. suction measure-
   ment test holes would be selected in a manner similar
   to that indicated in Section  3.3.1. Some additional
   judgement may be used in selecting locations,  in an
   effort to  address regions where it is  suspected the
   suction field is not reaching.

   — For suction  measurements under slabs, the
     post-mitigation testing would likely use the same
     test holes used for any pre-mitigation diagnostics, if
     conducted. More test hole sites might be selected to
     better define the limits of the suction field exten-
     sion. For houses having multiple slabs (e.g., abase-
     ment with an adjoining slab on grade), test holes
     will likely have to be drilled in both slabs.

   — For suction measurements inside block wall cavi-
     ties, one test hole should initially be drilled into one
     block cavity of each perimeter block wall and each
     interior load-bearing block wall. These test holes
     might logically be  near the center of each wall,
     toward the slab (since suction near the footings,
     where much of the soil gas enters the void network,
     may be most important).
                                                       270

-------
   —  For suction measurements beneath SMD membranes,
      test hole location would be selected in the same
      manner as test holes for slabs.

-   As with the pre-mitigation testing, the house would be
   placed in the condition that will be used throughout this
   testing. It is recommended that any commonly used
   appliances which might tend to depressurize the house
   be left running throughout the testing, so that the test
   results will  show the suction field extension  under
   challenging conditions.

-   The sub-slab depressurization being maintained at the
   test holes  would be measured with the ASD system
   operating.

   -  With slab holes, the sample tube from the micro-
      manometer would be installed in  the test hole as
      described in Section 3.3.1.

   --  With holes in block walls,  the tubing would  be
      inserted an inch or  so into the  cavity, and sealed
      with rope caulk.

   —  With membrane holes, a couple inches of the tubing
      should be inserted through the membrane, and care-
      fully sealed with duct tape. The open end of the tube
      beneath the membrane must be horizontal (or  be
      embedded in sub-membrane gravel), so that it is not
      potentially plugged  by being embedded in imper-
      meable soil.

   ~  Record the observed range of readings on the mi-
      cromanometer, and/or the average reading.

-   If suction is not extending at all to  some of the quad-
   rants, the mitigator may wish to drill additional test
   holes closer to the suction pipe(s) and to repeat the
   measurement steps above at these new holes, to iden-
   tify how far the suction field is in fact extending.

-   All holes  that have been drilled through the slab or
   block must be permanently closed  with hydraulic ce-
   ment or  other nonshrinking  cement after testing is
   complete. Holes through crawl-space membranes should
   be closed, preferably by sealing a piece of polyethylene
   sheeting on top of the hole using urethane caulk (for
   cross-laminated polyethylenes)  or  other sealant (for
   regular polyethylenes).

 Interpretation of results

-   If no depressurization can be measured in one or more
   locations beneath the slab (or inside  the wall, or be-
   neath the membrane),  or if,  in fact, the sub-slab is
   pressurized relative to the house,  the  mitigator can
   consider installing one or more additional suction pipes
   as necessary to treat these locations.

   —  If the region not being adequately depressurized by
      the initial system is relatively limited in size, it will
      likely be cost-effective for the mitigator to proceed
         immediately to install an additional suction pipe in
         that region, without doing any further diagnostics
         first.

      —  If the region not being adequately depressurized by
         the initial system is relatively large, the mitigator
         may wish to conduct further suction field diagnos-
         tics in that region, using a vacuum cleaner to draw
         suction in the region -i.e., using the "pre-mitigation"
         diagnostics approach in Section 3.3.1 or 3.3.2, but
         limiting the test holes to sites within the untreated
         region. These vacuum cleaner  diagnostics could
         help determine how many additional suction pipes
         should be installed, and where they should be lo-
         cated within that region.

         Alternatively, the mitigator might often find it more
         cost-effective to  simply use best judgement in in-
         stalling one or more additional suction pipes, with-
         out vacuum cleaner diagnostics. If those additional
         pipes still did not reduce the indoor radon concen-
         tration to the desired level, the mitigator would then
         have to repeat the suction field extension measure-
         ments with the modified system operating, to iden-
         tify any remaining portions of the slab (or wall or
         membrane) still not being depressurized.

   -  If the suction field extension measurements are made
      during mild weather, it must be recognized that the
      increased house depressurization  that will exist during
      cold weather will reduce the sub-slab depressurizations
      observed during the mild-weather measurements. Thus,
      if a house is not reduced below  4 pCi/L during mild
      weather, and if suction field extension measurements
      show some portion of the sub-slab not being depressur-
      ized during  mild weather, the problem might be ex-
      pected to become even worse during cold weather.

Even if the newly installed mitigation  system seems to be
satisfactorily reducing indoor radon levels during mild weather,
the mitigator may  wish to  perform these  quantitative
post-mitigation suction field extension measurements in houses
where the  sub-slab communication  seems  marginal. Such
measurements would be especially logical in marginal houses
where pre-mitigation suction field measurements had been
made, and where  the test holes through the slab are thus
already existing. In these cases, post-mitigation suction field
measurements would reveal whether sub-slab depressuriza-
tions during mild weather are marginal. Such marginal sub-slab
depressurizations could be overwhelmed by the increased
challenge created by thermally induced house depressuriza-
tion during cold weather.

As discussed in  Section  3.3.2  and elsewhere,  where  the
post-mitigation suction  field testing is made during mild
weather and/or without exhaust appliances operating,  the
measured sub-slab depressurizations should be great enough
to Compensate for the increased house depressurizations an-
ticipated under the more challenging conditions. As stated in
Section 3.3.2, the conservative goal would be to maintain the
following depressurizations  everywhere under the slab: at
least 0.015 in. WG, if the measurements are made  in mild
                                                     271

-------
weather with exhaust appliances operating; 0.025 to 0.035 in.
WG, during mild weather without exhaust appliances operat-
ing; and 0.01 to 0.02 in. WG, during cold weather without
exhaust appliances operating.

In practice, it appears that these figures are conservative, and
that ASD systems can often perform well even when sub-slab
depressurizations are less. Thus, lesser depressurizations might
sometimes be adequate. However, the mitigator must recog-
nize that a system providing only  marginal sub-slab depres-
surization during less challenging conditions may be over-
whelmed during more challenging  conditions.

11.7   Procedures for Radon  Grab
Sampling and Sniffing
The procedures for post-mitigation radon grab sampling and
sniffing will be the same as those described in Section 3.4 for
pre-mitigation  measurements. The primary difference is that
the post-mitigation measurements  will be performed with a
different emphasis, for the purpose of determining why an
installed system is not providing the desired radon reductions.

Sampling inside block watts. Perhaps the best example of
how radon grab sampling or sniffing might be effectively
utilized in post-mitigation diagnostics would be in basement
or crawl-space houses with block foundation walls, where a
stand-alone  SSD, DTD, or SMD system is not providing
adequate radon reductions. If sub-slab (or sub-membrane)
suction field extension measurements suggest that the entire
sub-slab seems to be adequately depressurized, then the ques-
tion would arise regarding whether the block  walls might
continue to be radon sources despite the apparently effective
slab treatment.

Samples would be drawn from inside selected block cavities
in each of the foundation walls with the mitigation system
operating, to determine if one or more walls have distinctly
elevated radon levels. If elevated walls are found, this result
would indicate that the system is not preventing soil gas entry
into that wall, and the wall is still a source. In such cases, a
BWD component to treat the elevated walls might be added to
tlie initial system. Or, alternatively, additional SSD or SMD
suction pipes might be installed near these walls.

Sampling at unsealed openings. In general, grab sampling/
sniffing at suspected soil gas entry routes can be used to
determine whether gas in those unsealed openings is high in
radon, suggesting that that entry route is not being adequately
treated. That is, high-radon soil gas is in that opening, presum-
ably entering  the house, rather than  low-radon house air
exiting through that opening. Any openings in  the slab of
sufficient size to enable insertion of a grab  sampling tube
should have been sealed in conjunction with the installation of
the ASD system. However, to the extent that such intermedi-
ate or large unsealed slab openings still exist, grab sampling in
those  openings would suggest whether the suction field is
preventing radon entry there.

If tlie slab opening is, e.g., a crack which is too small for the
sampling tube to be inserted, a sheet of plastic can be sealed
over a length of the crack using duct tape, and the buildup of
radon beneath the plastic over a number of hours can be
measured, to assess whether radon is entering through the
crack. The plastic sheet will hinder any convective flow of
soil gas that might otherwise occur up through the crack, so
that this approach would be less reliable in revealing the
potential significance of the crack than would be suction field
measurements beneath  the slab at the crack. However, the
plastic-sheet approach may sometimes be qualitatively useful
in cases where it is not  desired  to drill test holes through the
slab at the crack location.

Where SSD  or DTD has been installed in a basement house
having poured concrete foundation walls, grab sampling in
any unsealed wall openings (e.g., around utility penetrations)
could indicate whether  radon is continuing to enter through
these openings. That is,  it would indicate whether the SSD or
DTD system were adequately treating the exterior face of the
concrete wall.

Sampling beneath the slab. Radon grab sampling or sniff-
ing beneath slabs through test holes will probably not com-
monly be useful as a post-mitigation diagnostic method. How-
ever, it might be of practical value in certain cases.

A sub-slab radon grab sample could easily be drawn through
the test hole during the post-mitigation suction field extension
measurements discussed in Section 11.6. However, the results
from a grab sample will probably not be useful if the sub-slab
is being effectively depressurized at that test hole. The sub-slab
radon concentration will commonly be of only secondary
interest as long as the ASD system is preventing  the soil gas
from entering the house at that location.

But where the sub-slab depressurization is only marginal or is
non-existent, the sub-slab radon grab sample might occasion-
ally help determine the importance of adding a suction pipe to
treat that untreated region of the slab. Higher sub-slab radon
concentrations could suggest increased importance in adding
a suction pipe in that area, if there are entry routes nearby
through which this sub-slab radon could enter the house. Note
that the radon entry potential is a function not only of the
sub-slab radon level, but of the ability of that radon to enter
the house at that location (Tu90, Tu91a).

Where the house has a combined substructure—e.g., a base-
ment having an adjoining slab-on-grade or crawl-space wing—
and where only the basement is being  directly treated (by a
SSD or DTD system), grab sampling could help determine the
extent to which the untreated adjoining wing may be contrib-
uting to any residual indoor radon level. With the basement
ASD system operating, high radon levels under the adjoining
slab on grade, or at slab openings in the slab on grade, would
suggest that radon is continuing to enter the house via the
adjoining slab.  That is, the basement ASD system is not
adequately intercepting the soil gas moving toward the adjoin-
ing slab. Likewise, high radon  levels in the  untreated crawl
space would suggest that the crawl space is continuing to be a
radon source.  In  such  cases,  a SSD  or SMD  component
treating the  adjoining wing  may have to be  added to the
basement ASD system.
                                                       272

-------
Sub-slab radon grab sampling after mitigation can also pro-
vide other information which may occasionally be of help in
diagnosing a house. Very high post-mitigation sub-slab radon
concentrations can  suggest a potentially increased diffusive
flux of radon into the house. With a very high concentration
gradient between the sub-slab and the house, radon could
diffuse up through  the open space in slab cracks and other
openings, or even  through unbroken concrete. Convective
flow is normally the predominant mechanism by which radon
enters a house; diffusion is generally expected to be a minor
contributor. However, in cases where sub-slab radon concen-
trations are very high, the diffusive component could become
important, especially when the desired indoor levels  are well
below 4 pCi/L.

A grab sample of radon in the suction piping would provide a
qualitative indication of the sub-slab radon concentration. The
levels in the piping will be diluted by indoor and outdoor air
that has been drawn into the system. If radon levels in the
piping are dramatically lower than sub-slab concentrations,
this would be an  indication that a significant  amount of
leakage is occurring.

11.8  Procedures for  Chemical
Smoke Flow Visualization Tests
General procedures. The simple procedures for chemical
smoke testing are as described earlier in Sections 3.3.1 and
11.6.1. A small amount of smoke is released near the seam or
opening of interest. A flashlight back-lighting the smoke is
sometimes helpful  in detecting the smoke patterns. Smoke
movement into the  seam or opening means that: a) the open-
ing is depressurized relative to the house; and. b) the opening
is not sealed, allowing house air to pass through it. Similarly,
smoke movement away from the opening  means  that  the
opening is pressurized relative to the house, arid that it is not
sealed, with gas  flowing through it into the house.

Chemical smoke is an inert smoke-like, fine powder. It can be
used with various types of dispensers,  which release small
puffs or a brief,  continuous stream. Chemical smoke devices
are considered safer than are, e.g., burning punk sticks or
cigarettes. In addition to the fire hazard, the smoke leaving
punk sticks and cigarettes tends to rise, since it is released at
high temperature.  These thermal  effects on the  smoke can
mask the sometimes subtle effects of the system-created air
movement that the smoke is intended to visualize.

Assessment of openings in foundation. Chemical smoke
can be useful in a number of ways. It can  be used around
unsealed cracks, openings, and  seams in slabs, block walls,
and crawl-space membranes with ASD systems operating, to
determine whether the sub-slab, wall cavity, or sub-membrane
depressurization is extending to those  locations. Similarly,
they can be used near sealed cracks, openings, and seams with
the ASD systems operating, to confirm whether in fact they
are effectively sealed (in which case there should be no air
movement into or out of these openings).

Smoke visualization can be used around potential entry routes
which may not be being treated by the ASD system, to assess
whether some form of sealing or treatment of these potential
routes might be necessary. For example, smoke tests around a
floor drain may help confirm whether or not it is effectively
trapped. Smoke tests at a hole through the poured concrete
foundation wall of a basement with a SSD system could show
whether soil gas is entering the basement through that hole.

If smoke flow is not distinctly upward over a potential entry
route, this result does not necessarily prove that radon is not
entering by convective flow through that opening. At very low
pressure differentials across the slab, the upward flow of soil
gas may be too low to be detected by the smoke. However, the
smoke test is easy to perform, and is often worth conducting
when initial ASD performance is inadequate, even though the
smoke cannot reveal very subtle flows.

Detection of piping leaks. Smoke visualization can be used
to determine whether all of the seams in the ASD system
piping have been sealed. If, e.g., a SSD suction pipe is not
adequately sealed where it penetrates the slab,  or if any fitting
joint in the PVC piping is not sealed on the suction side of the
fan, smoke will be drawn into that joint. Unlike the smoke
flows into or out of slab cracks, floor drains, etc., discussed in
the previous two paragraphs, where air movement may some-
times be subtle, the smoke flows into leaks in the ASD piping
will usually be unambiguous, due to the relatively high suc-
tions in the piping.

Similarly, smoke visualization can be used around joints in
the piping on the pressure side of the  fan,  to ensure that
exhaust gas is not leaking out at those joints. While the
pressure-side piping will almost always be outside the living
area (i.e., in the attic or out-doors), there will nevertheless be
occasions where the mitigator will be well advised to make
certain that there are no leaks there.

For example, pressure-side leaks in attic exhaust piping will
be particularly undesirable in attics where the central forced-air
furnace fan (and hence some cold-air return ducting) is in the
attic, a situation occasionally encountered. Pressure-side leaks
in garage-mounted fans are generally undesirable, since there
is almost always a door and perhaps other openings between
the attached garage and the living area. With exterior stacks
having the fan at the bottom of the stack, radon leaking from
the stack beside the house could potentially become
re-entrained through windows. This potential problem can be
of particular concern when the exterior stack is fabricated
from sections  of rain gutter downspouting,  which  is less
amenable to air-tight sealing at the joints.

Assessment ofbackdrafting. As discussed in Section 11.5.1,
smoke visualization tests at the draft hood or barometric
damper, or at the air Met to the combustion zone, is a
convenient and effective method  for assessing whether a
combustion appliance  is drafting properly  (and whether an
ASD system is impacting that draft).

11.9  Procedures to Test for Exhaust
Re-Entrainment
In some cases, an ASD  system might not be providing ad-
equate radon  reductions, but the preceding post-mitigation
diagnostics might fail to reveal the cause of the problem.
                                                      273

-------
Suction field extension testing might show the sub-slab, block
wall, or sub-membrane to be depressurized everywhere. Ra-
don grab sampling might confirm that radon is not entering
through untreated block foundation walls. In such cases, the
residual radon present in the house may be entering through
some mechanism other than soil gas flow through the slab and
walls.

One such additional mechanism could be re-entrainment of
the ASD exhaust. Other mechanisms could include entry in
well water, and release for high-radium building materials,
discussed in later sections.

If the ASD exhaust is being released above the eave, and is at
least 10 ft from openings in the house shell, re-entrainment
should usually not be a significant problem. Under these
conditions, the appropriate first steps to take when the process
of elimination suggests that re-entrainment may be occurring
are: a) a visual inspection, to confirm, e.g., that the discharge
is not near a window; and b) smoke stick tests, to confirm that
there are no leaks in the piping on the pressure side of the fan.

If these steps failed to reveal a problem, and if re-entrainment
continued to be suspected, two other approaches might be
considered. One would be to temporarily modify the exhaust,
redirecting the exhaust so that it discharges at a location so
remote that re-entrainment should be eliminated or at least
dramatically reduced. This might easily be done by clamping
non-perforated  flexible corrugated piping to the end of the
existing stack, and routing that flexible piping to  a remote
point. If this step reduced indoor radon levels, it would be
evidence that the original exhaust configuration was resulting
in re-entrainment.

Another possible approach would be to inject a tracer gas into
the exhaust piping, and then measuring for the tracer inside
the house. The general procedures associated with tracer gas
testing have already been discussed, in  Section 3.5.6. As
indicated in that earlier section, if tracer gas  testing were
considered  at all by a mitigator, it would most likely be
conducted using a HFC  refrigerant as the tracer, due to the
availability  and ease of detection of these gases. As empha-
sized previously, due to  concerns about stratospheric ozone,
any such testing should be conducted using HFC refrigerants,
rather than the CFCs or HCFCs.

11.10   Procedures for Well Water
Radon Analysis
The procedure for conducting radon analysis in well water
was  described in Section  3.5.2,  in  connection with
pre-mitigation testing.

11.11   Procedures to Determine the
Significance of Building Materials as
a Radon  Source

The procedures for gamma measurements and surface flux
measurements, to help determine whether building materials
might be a radon source, were described in Section 3.5.3, in
connection with pre-mitigation testing.
                                                     274

-------
                                          Section  12
               Operation and  Maintenance Requirements
                for Active  Soil  Depressurization Systems
Following the installation and the initial check-out of the
system by the mitigator, the homeowner or occupant will be
responsible for subsequent routine operation and maintenance
of the system.

If operating problems develop with the system, or if non-routine
maintenance is required, the owner/occupant must be in a
position to note the problem, and must ensure that any neces-
sary repairs or maintenance are completed.

12.1  Instructions Following
Installation
As a minimum, the mitigator must provide sufficient informa-
tion about the system so that the owner/occupant can detect
subsequent operating problems, and can  contact  a profes-
sional mitigator to make any necessary repairs. The mitigator
should also indicate any routine maintenance mat the owner/
occupant will be required to perform. Depending upon the
mitigator's practices and the client's inclinations, the mitiga-
tor  might also provide information  to enable the owner/
occupant to undertake some initial, simple diagnostics prior to
calling the mitigator, should operating problems develop.

EPA's interim radon mitigation standards (EPA91b) specify
that the mitigator shall provide the client with the  following
information, among other information:

   -  A description of the system, at the centrally located
      label indicated in Section 4.8.2.

   -  The name and telephone number of the mitigator, or
      other person to contact should problems develop with
      the system.

   -  A statement indicating any required maintenance by
      the owner/occupant.

   -  The method of interpreting the gauge or alarm which
      monitors system performance, in accordance with Sec-
      tion 4.8.1. This would include the expected response of
      the monitor if a problem developed with the system.

   -  A recommendation that the house be retested for radon
      at least once every 2 years.

In addition to these specifications in the interim  standards,
mitigators may wish to provide additional information to aid
the owner/occupant in proper operation and maintenance of
the system. The extent to which this additional information is
provided will be determined by the mitigator's practices and
the owner's inclinations. Among the additional information
that EPA might ultimately require, based on the current draft
of EPA's final Radon Mitigation Standards, are:

   -   Copies of contracts and mitigator warranties.

   -   The basic operating principles of the system.

   -   A description of the proper operating procedures of any
      mechanical or electrical systems installed, including
      manufacturer's operation and maintenance instructions
      and warranties.

   -   A list of appropriate actions for the owner/occupant to
      take if the system  failure warning  device indicates
      system degradation or failure.

12.2 Operating and Maintenance
Requirements

12,2.1    System Fan

The major mechanical component of an ASD system is the
system fan.

Routine  maintenance. The primary routine maintenance
for the fan is periodic inspection of the warning device that
has been installed on the ASD system, to ensure that the fan is
continuing to operate properly. In this regard, homeowners/
occupants should be encouraged to routinely check the pres-
sure gauge or pressure  alarm on the system piping, or the
ammeter in the fan wiring (see Section 4.8.1).

The client should be given clear instructions on how to read
and interpret the system warning device following installa-
tion.

If the system is equipped with a pressure gauge, the owner/
occupant  must confirm  that  system suction remains in the
proper operating range, as marked on the gauge. Mounting the
gauge in an area frequented by the occupant will help ensure
that such  checking will be done. Supplementing the gauge
with a visual or auditory alarm would attract the occupant's
attention  if the suction falls out of the acceptable range
between checks.
                                                   275

-------
If the indicator device mounted on the system is an alarm
only, with no gauge, the occupant should be encouraged to
routinely check the indicator light confirming that the alarm is
properly activated.

The owner/occupant might also periodically listen to the fan,
to ensure that it is operating. Some fans operate so quietly that
the occupant may have to listen closely to confirm that it is in
fact operating.

The fact that the fan  is operating does not mean that it is
operating properly. When electrolytic capacitors  in the fan
circuitry fail while the fan is operating, it is sometimes pos-
sible for the fan to continue operating for some time at greatly
reduced performance (Fi91). For these reasons, listening to
the fan cannot be a substitute for checking the suction gauge.
However, it may sometimes  be a convenient supplement If
the fan begins making growling noise, it could be an indica-
tion that the bearings are about to fail, which should alert the
occupant.

Ensuring that the fan is not disabled. In what might be
considered a part of routine maintenance, the owner/occupant
must ensure that the fan is not turned off when it should be
operating, or inappropriately turned down below  the power
setting established during installation.

Sometimes when the  house will  be left unoccupied for  a
period of time, the owners/occupants might wish  to turn the
fan off during their absence. If this were  done,  they must
ensure that the fan is turned back on upon their return.

Also, owners/occupants sometimes choose to turn  the fan off
during mild weather when windows are often open. However,
this practice is discouraged.  Experience has shown that the
system will not  always be turned back on every time the
windows are closed. Moreover, the open windows will not
always compensate for the ASD system being off. Thus, there
can be a resulting increase in indoor radon.

If the fan circuit is equipped with a voltage (speed) controller,
the owner/occupant must ensure that this  controller is not
inadvertently turned down to a lower speed than that at which
it was set by the mitigator. Any reduction in power to the fan
should be accompanied by careful radon measurements to
confirm that the reduced fan speed does not result in unaccept-
able increases in indoor radon. In addition to potentially
increasing indoor radon levels, operating the fan at too low a
speed could shorten fan life by increasing motor temperature,
since the lower air flows may be insufficient to provide the
necessary motor  cooling.

Fan failure and repair. Periodic inspection of  the system
warning device,  as discussed above, should alert the owner/
occupant if the fan has failed. The owner/occupant is respon-
sible for ensuring that the fan is repaired or replaced if it fails.

Commonly, repair or  replacement will be  accomplished by
contacting the mitigator who installed the  system, or some
other professional. This will be especially true when the
system is under  warranty. As specified in EPA's standards,
the client must be advised following installation regarding
what response from the warning device indicates a problem.
The client must also be provided with the name and telephone
number of the mitigator, to facilitate contact when problems
arise.

As discussed in Section 2.3, the mitigation fans on  many
commercial ASD installations have operated for a number of
years without failure. However, these systems have not yet
operated  for a  sufficiently long period to enable a more
quantitative estimate of the typical lifetime of these fans in
this application. A few fans have failed after only a year or
two. The most common cause of fan failure appears to be
failure of the electrolytic capacitor in the fan circuitry. Less
frequently, bearing failure can be a cause.

Simple steps that the owner/occupant might take prior
to contacting mitigator. When the warning device indicates
that system suction has dropped below the acceptable operat-
ing range, this will not always be the result of a problem with
the fan. In some cases, it may be due to problems with the
system piping or foundation seals, as  discussed in Section
12.2.2. Also, the situation creating the problem can sometimes
be addressed by simple corrective action on the part of the
owner/occupant.

Thus, mitigators may sometimes wish  to provide the client
with a listing of simple diagnostic tests that the  owner/occu-
pant can  quickly perform to help identify the nature of the
problem, and, in simple cases, to correct it.

Steps that the owner/occupant might take when the suction
moves outside the acceptable range could include the follow-
ing. A mitigator evaluating the problem might initially take
similar steps.

 1. If the suction has dropped  to zero, check the fan to
    determine if it is not operating.

    a. If it is not operating, has power been interrupted? Is
       the switch in the fan circuit on? If the fan is simply
       plugged into an outlet, is  the plug still in the outlet?
       Try flipping the circuit breaker for the circuit that the
       fan is on.

    b. If the fan is not operating and if the power is on, the
       fan has probably failed, or else there is a problem in
       the switch or the wiring to the fan. In this  case, the
       owner/occupant should call a professional mitigator
       to have the fan repaired or replaced. In some cases,
       repairs can be implemented on  site  (e.g., installing a
       new capacitor in some fans). In other cases, the fan
       will have to be returned to the manufacturer for repair,
       or replaced with a new fan.

       Historically, many of the manufacturers  of fans com-
       monly used in ASD systems  have offered 3-year
       warranties on the fans. Some warranties have recently
       been reduced to 1 year.  If the fan fails within the
       warranty period, the mitigator or the homeowner could
       return it to the manufacturer for repair or replacement,
       if it cannot be repaired on site. If the fan is removed
       for repair, the mitigation system will, of course, cease
                                                       276

-------
      to be operational for the duration of the repairs unless
      a temporary replacement fan can be installed.

   c.  If the suction has dropped to zero and the fan seems to
      be operating, there may be: condensation or ice block-
      ing  the  tubing between the  gauge and the suction
      piping; failure of the gauge; an interruption in the
      piping between the gauge and the fan; a blockage of
      the piping by ice or condensate; or failure of the fan.

      The owner/occupant should inspect to see if the tub-
      ing to the gauge appears blocked, to  ensure that the
      system .piping is intact, and to identify any potential
      blockage of the piping. If no breaks or blockage are
      discovered  that the owner/occupant can repair, the
      owner/occupant should call the mitigator.

      Smoke-stick testing might be performed in the hole
      through the piping where the  gauge is  mounted to see
      if there is any suction in the piping. If the smoke flow
      indicates that the piping is under suction despite the
      zero reading on the gauge, the gauge is malfunction-
      ing  or  its tubing  is blocked. If there is indeed no
      suction, the fan might be removed and checked to
      determine if the rotor is in fact turning.

2. If the suction has dropped below the acceptable range,
   but has not dropped to zero, the owner/occupant should
   do the following.

   a.  If the fan circuit has been equipped  with a voltage
      (speed) controller, check the controller to make sure
      that it has not inadvertently been changed to a lower
      setting:

   b.  Check for leaks in the system piping or in slab/wall/
      membrane seals, as discussed in Section 12.2.2.

   c.  Check  for  potential blockage of the piping down-
      stream of the fan by ice or condensed moisture, or on
      the  suction side of the fan between the fan and the
      measurement location, as discussed in Section 12.2.2.

   d.  If the above steps do not reveal an apparent reason for
      the decrease in suction, call the mitigator. One possi-
      bility may be that the fan capacitor has failed, causing
      the fan to operate at reduced suction.

3. If the suction has increased significantly above its normal
   value, this, too, could be indicating a serious problem
   with the system, although the problem probably is not
   caused by the fan.

   a.  The owner/occupant should check the system piping
      upstream of the fan (and upstream of the measurement
      location) for signs of possible blockage, as discussed
      in Section 12.2.2.
 12.2.2    Piping Network, and System
and Foundation Seals

Several aspects of the piping network, and of the seals associ-
ated with the system piping and the slab/wall/membrane, can
also contribute to reduced performance of the ASD system.
Part of the owner/occupant's operation and maintenance re-
sponsibility includes inspecting these features, and conduct-
ing (or arranging for) appropriate repairs as necessary.

Piping network (other than seals).  Depending upon the
climate and the design of the piping network, sections of the
piping can sometimes become partially or completely blocked.
This blockage can result in  either: a decrease in  system
suction, if the blockage is on the pressure side of the fan, or on
the suction side between the fan  and the measurement loca-
tion; or an increase in measured suction, if the blockage is on
the suction side upstream of the measurement location. In
either  case, the  blockage can dramatically reduce  system
performance.

Blockage on the pressure side of the fan, if it occurs, will most
commonly result from accumulation of ice in exterior stacks
during cold weather. Moisture in the soil gas will condense in
the stacks. Since exterior stacks  cannot practically be insu-
lated (except by framing them in), this condensed moisture
can sometimes freeze  along the interior walls of the stack in
particularly cold climates, despite the momentum of the ex-
haust gas and the consistently moderate temperature at which
the soil gas will enter the  stack. This ice can significantly
decrease the effective diameter of the pipe (increasing back
pressure on the pressure side of the fan, thus decreasing the
suction on the suction side of the fan). In extreme cases, the
stack might be frozen closed entirely.

In locales where ice will partially block the stack, this back
pressure should considered in the design of the system, and in
advising the occupant regarding the suctions that might be
expected on the gauge during cold weather. Often, it appears,
this partial blockage  will reduce,  but not be  fatal  to, the
performance of the system. If the stack freezes up completely,
no good solution to  this problem is apparent;  occasional
application of heat (e.g., hot water) to the exterior of the stack
may be the only option.  Clearly,  indoor  stacks would be
preferred in such climates, if stacks are required.

Blockage on the suction side of the fan can result from low
points in the horizontal piping where condensate can accumu-
late. If the suction gauge indicates increasing suctions, the
owner/occupant  should look for such potential low spots
upstream of the gauge. If low points exist and a drain tube has
been installed, the owner/occupant should check to ensure
that the tube is not plugged and that condensate is draining
freely through the tube. If serious undrained low points are
discovered in horizontal piping runs, the mitigator should be
contacted.

In some cases, increased suction  (and reduced flow) may
result not from any blockage of the piping, but from reduc-
tions in sub-slab communication, due to sub-slab moisture
during wet seasons. In the worst case, accumulation of water
in the sub-slab pit beneath SSD suction pipes could be respon-
                                                      277

-------
sible for the increase in suction, especially if the open end of
the pipe is low enough in the pit to be below water level.
Flooding of drain tiles in DTD systems could produce the
same effect.

Blockage of a SSD suction pipe could also result if the pipe
dropped into the sub-slab pit, so that the open end was now
resting on the bottom of the pit. The piping should be in-
spected for evidence whether this has occurred.

System seats. If system seals on the suction side of the fan
break over the years, this would result in an increase in system
flows (and a decrease in the measured  suction). Such seals
include:

   -  fitting joints in the piping network;

   -  the seal around the suction pipe, where it penetrates the
      slab, wall, or membrane;

   -  traps in sump covers of sump/DTD systems, such as
      illustrated in Figures 33A and 33B;

   -  the seal between the sump cover and the slab in sump/
      DTD systems, and the seal between the cover and any
      penetrations in addition to the trap; and

   -  the check valve in DTD/remote discharge systems, as
      illustrated in Figure 4.

If the owner/occupant observes a major reduction in system
suction, he/she should inspect these seals, insofar as they are
visible, to see if there any major failures are visible. Are any
of the piping joints obviously loose? Is the caulk seal around
the SSD pipe penetration through the slab still intact? Is the
trap in  the sump cover either full  of  water, or, if it is  a
waterless trap, is the weighted ring or ball seated properly?
Where there is a significant leak in the system piping, such a
leak can sometimes be detected by the  hissing sound of air
being drawn through the opening.

If visual inspection is ambiguous regarding whether some of
these seals are in fact intact, and if the owner/occupant has a
chemical smoke device (or other smoke  generator) available,
a smoke flow visualization test would quickly indicate how
well the seals are maintained.

The owner/occupant would be well advised to periodically
check these seals visually, to ensure that they are remaining
intact.

If ruptured seals are apparent from this inspection, or if the
trap in the sump cover is dry or not airtight, the owners/
occupants may be able to restore some  of these seals them-
selves. In other cases, the owner/occupant may need to call the
mitigator.

Foundation seals. If suctions in the ASD piping decrease
and flows increase, one other possible explanation could be
short-circuiting of house or outdoor air into the system through
ruptured seals in slab, wall, and membrane openings. For
example, broken caulk seals at the wall/floor joint or the
perimeter channel drain could increase  air leakage into a
basement SSD system. Broken seals between the crawl-space
membrane and the perimeter foundation wall could increase
air leakage into crawl-space SMD systems.

As with  the system seals discussed previously,  the owner/
occupant should inspect these foundation seals if a significant
decrease  in system suction is observed. In fact, the occupant
would be well advised to periodically check these seals visu-
ally in any event, to ensure that they are remaining intact. A
chemical smoke test could supplement the visual inspection.

If an occupant has a smoke device available, he/she would be
well advised to also periodically check any other seals that
were installed as part of the mitigation system, even if they do
not necessarily directly impact the performance of the ASD
system. For example, if a floor drain has been trapped with a
waterless trap, the smoke stick could be used to confirm that
the trap is continuing to be effective, by demonstrating that
gas is not flowing up into the house through the drain.

If a water trap has been installed in a floor drain or sump
cover, water should routinely be added to the trap.

Summary. If the measured system suction moves outside the
range specified on the gauge by  the mitigator, and if the
owner/occupant cannot identify and correct the cause of the
problem using the steps outlined above, the  owner/occupant
should contact the mitigator.

In some  cases, the measured system  suction may remain
within the indicated range, but may still change significantly
(in one direction or the other) from the levels that it  has
historically been maintaining. In such cases, the owner/occu-
pant would be well advised to follow the steps outlined above,
to see if the  cause  of the change can be explained and
corrected. If the changes cannot be reversed, but if the suc-
tions are  still within the indicated range, the  owner/occupant
may wish to conduct a follow-up radon measurement before
contacting the mitigator, to see if the suction  changes have in
fact degraded the system's radon reduction performance.

12.2.3     Follow-Up Indoor Radon
Measurements

EPA's interim mitigation standards (EPA91b) indicate that
the mitigator should recommend that the homeowner re-test
for radon at least once every 2 years.

Testing even more frequently might be warranted if the owner/
occupant observes any changes in the system the might imply
changes in its performance, as indicated above. Significant
changes in climate or house operation, or remodelling of the
house, could also warrant re-testing.

While EPA's  standards do not specify the type of measure-
ment technique to be used for such follow-up  testing, the
owner/occupant will probably find it least expensive and most
convenient to use either a charcoal detector or an alpha-track
detector.  These devices can be readily  purchased  locally,
deployed by the owner/occupant, and mailed directly  to a
laboratory for  analysis. Since the owner/occupant's concern
                                                       278

-------
wiH commonly be one of the family's long-term exposure to
radon, alpha-track detectors, deployed for 3 to 12 months,
may be a logical selection. If the owner/occupant is trying to
obtain a relatively quick measure of, for example, the impact
of some observed change in  system  suctions, a charcoal
detector, deployed for perhaps 2 days, may be a better choice
to address that objective.
                                                        279

-------

-------
                                            Section  13
                        Installation and  Operating Costs
                for Active  Soil  Depressurization  Systems
In estimating the costs of ASD systems, it is necessary to
consider: a) the cost of installing the system at the outset; and
b) the cost each year of operating the system after installation.

General considerations regarding  installation  costs.
Installation costs can vary significantly, depending upon a
variety of factors.

Perhaps  the most important factors are associated with the
characteristics of the house. Important house characteristics
include the substructure type, the nature of the sub-slab com-
munication (in houses with slabs), the house size, the floor
plan, and the degree of interior finish, among others. House
characteristics can impact the costs of the ASD system by
determining, among other things:

   -  the ASD variation selected;

   -  the number and location of suction pipes;

   -  the nature of the system fan;

   -  the routing of the piping network;

   -  the degree of foundation sealing required;

   -  the extent to which existing finish must be removed or
      restored;

   -  the extent of pre- and post-mitigation diagnostics re-
      quired.

Other factors influencing installation costs are associated with
the practices of the individual mitigators. Different mitigators
provide  different levels of services; for example, different
levels of pre- and post-mitigation diagnostics, different war-
ranties, and different qualities of finish  on the system. Also,
different mitigators have different labor, fringe benefit, and
overhead rates.

Because installation  costs can vary  so much between one
installation and the next, installation costs are presented in this
document as a range, incorporating the ranges of house char-
acteristics and mitigator practices. The incremental impacts of
certain key system design variables on installation costs, such
as adding suction pipes, are also shown.
In most cases, materials costs are only a small contributor to
the total installation cost. Commonly, labor costs constitute
about 70 to 80% of the total installed cost is SSD and DTD
systems (He91c). Materials costs will be a somewhat greater
portion of the total in SMD systems, due to the cost of the
membrane material.

General considerations regarding operating costs. Op-
erating costs for ASD systems include four elements: 1) the
cost of electricity to operate the system fan; 2) the heating and
cooling penalty resulting from treated house air being with-
drawn and exhausted from the house during cold  and hot
weather; 3) the cost of system repairs and maintenance; and 4)
the cost of any periodic remeasurement of indoor radon levels
that a homeowner may wish to conduct in order to ensure that
the system is continuing to perform satisfactorily.

The operating cost for a particular house will depend upon a
variety of factors. These include:

   -   The amount of electric power consumed by  the fan.
      This will be determined by: the specific  fan selected;
      and where that fan is operating on  its  performance
      curve, since  operation at less than maximum rated flow
      will result in reduced power consumption.

   -   The impact of the system on the house ventilation rate.
      This will be determined by: the amount of house air
      exhausted by the system; and the impact of that exhaust
      on the house  ventilation rate. Exhausting a certain
      amount of house air by the system will not necessarily
      increase ventilation rate by that amount; it may, in part,
      modify the  ventilation pattern without changing the
      overall rate.

   -   The local climate. An increase in the house ventilation
      rate induced by the ASD system will clearly have a
      lesser heating  and cooling penalty in a  mild climate
      than in a climate with very cold winters or very hot,
      humid summers.

   -   The efficiency of the heating and cooling  system in the
      house.

   -   The local cost of electricity and fuel.

Because operating costs can be influenced by so many vari-
ables in practice, the costs are shown as a broad range in this
                                                     281

-------
document. The assumptions that were used in deriving these
costs, and the ranges of variables covered for each variation of
the ASD process, are shown in Table 5.

Situations could arise that would be outside the range of the
individual assumptions shown in the table. For example,  a
high-suction/low flow fan might be used that would have  a
power consumption higher than the range in the table. How-
ever, it is believed that the ranges of variables included in
Table  5 result in operating cost ranges which reasonably
reflect the broad range of costs that will be encountered in
practice.
                                               13.1   Sub-Slab Depressurization
                                               Costs

                                               13.1.1    SSD Installation Costs

                                               Relatively simple systems. In one definitive study of SSD
                                               installation costs (He91b, He91c), five mitigators were asked
                                               to provide cost estimates for installing SSD systems in each of
                                               several "baseline" houses, reflecting a range of house design
                                               variables (basement vs. slab-on-grade substructure, finished
                                               vs. unfinished basement, one vs. two stories). In each house,
                                               the baseline SSD system was  defined by 12 system design
Table 5.
Assumptions Used in Estimating Annual Operating Costs for ASD Systems
  Variable
          Range
                             Comments
Ran power consumption

  - SSD, DTD, SMD    50-90 watts
  - BWD              75-180
                       watts

Fan op oration         24 hr/day,
                     365 dayfyr

tncraasa In housa ventilation rate

  • SSD              5-80 cfm
  - Swnp/DTD         10-120 cfm


  - DTD/remota disch.  5-80 cfm


  - BWD              50-320 cfm



  - SMD (vented crawl)  0-50 cfm



  • SMD (unvontod)    0-100 cfm
                     Includes cases of: 50-watt tubular fan operating at full power (maximum rated flow); 90-watt tubular fan
                     operating at reduced or full power; radial blower operating at reduced power.

                     Includes cases of: one 90-watt tubular fan operating at full or somewhat reduced power; two 90-watt
                     fans operating near full power.

                     Continuous operation assumed in all cases.
CKmata


Heating system


Cooling system


Cost of electricity


Cost of natural gas
        Los Angeles-
        Minneapolis

        Forced-air
        furnace

        Electric air
        conditioner

        $0.060-0.096
        perkWh

        $4.54-6.75
        per 10» Btu
Based upon typical SSD exhaust flows of 20-100 cfm, assuming that, typically, 20-80% of the exhaust
is treated house air (Ha89, Tu89, Bo91, CI91, R91). Assumes that the house ventilation rate is
increased by an amount equal to the amount of house air exhausted, which may not be literally true.

Based upon typical sump/DTD exhaust flows of 50-150 cfm, assuming that 20-80% of exhaust is house
air (same as SSD) when tiles are inside footings.

Based upon typical DTD/remote discharge  exhaust flows of 50-150 cfm, assuming that 10-50% of
exhaust is house air (lower percentage than for SSD, since tiles are often outside footings).

Based upon typical flows of 100-200 cfm  in stand-alone BWD systems with 90-watt tubular fan,
assuming that 50-80% of exhaust is house air (higher percentage than for SSD, since there are
numerous openings between walls and basement). Upper end of range assumes two fans.

Based upon typical SMD exhaust flows of 20-100 cfm (covering both individual-pipe and sub-membrane
piping variations), assuming that: 0-100% of the exhaust air is drawn from crawl space (Ma89b, Bo91,
Fit92); and 25-50% of this crawl-space air is treated air drawn from the living area.

Based upon SMD exhaust flows of 20-100 cfm, assuming that 0-100% of the exhaust air is drawn from
the crawl space, and that 100% of this crawl-space air is treated air.

Mild climate (Los Angeles) is 1698 heating P-days and 565 cooling infiltration F°-days (She86).
Extreme climate (Minneapolis) is 8034 heating F°-days and 1474 cooling F°-days.

Furnace bums natural gas and is 70% efficient.


Air conditioner coefficient of performance is 2.0.


Covers range of costs around the U. S. in 1987,
from Reference DOE87.

Covers range of costs around U. S. (DOE87).
                                                         282

-------
variables,  including,  e.g., the number and location  of the
suction pipes, the exhaust configuration, the degree of slab
sealing, the amount  of pre-mitigation  diagnostics, etc. In
addition to the specification of these 12 system design vari-
ables, the  effort to ensure consistency among  the estimates
included specification of the assumed travel time to the job
site, the system warranty, and the fact that applicable local
building codes were to be met.

The baseline system for each  house type was selected to
represent what might be considered a relatively simple case
for that house. For example, each baseline system  involved
only one SSD suction pipe, a relatively simple piping run, and
no slab sealing.

The five mitigators represented different major mitigation
markets around the country, and a range of approaches to the
detailed practice of radon mitigation. They also represented a
range of labor and overhead rates.

The estimated installation costs for these baseline (i.e., rela-
tively straightforward) SSD systems ranged between about
$800  and  $1700 (in 1991 dollars).  Again, this cost range
covers a range of house characteristics and a range of mitiga-
tor practices and market conditions. For this reason, this range
is felt to fairly well represent what a homeowner might expect
to pay for a relatively simple, one-pipe SSD system installed
by a commercial mitigator.

In a separate survey of 340 mitigators, the average cost of an
SSD installation was reported to be $1,135  (Ho91). This
broader mitigator survey did not compile information on the
house characteristics and the SSD system design characteris-
tics on which the reported cost was based. However,  it is clear
that this 340-mitigator  average is quite consistent with the
range of costs cited above.

More complicated systems. As emphasized previously, the
baseline cases in References He91b and He91c represent, for
the most part, relatively straightforward cases. Complications,
such as additional SSD suction pipes, more elaborate piping
runs, and  more extensive sub-slab diagnostic testing, would
serve to increase these baseline costs.

Accordingly, each of the five mitigators participating in that
study were asked to estimate the incremental changes in the
baseline costs that would occur if various alterations were
made in the twelve baseline SSD design variables.

The impacts of the SSD design variables on the installation
costs  are  discussed at length  in Reference He91c. For the
purposes here, it is sufficient to underscore that a number of
design variables can influence the baseline costs. Among the
SSD design variables having the greatest impact  on costs
among these five mitigators were:

    -  Adding SSD suction pipes. Each additional suction
      pipe could add roughly $70 to $380 to the installation
      cost, depending  upon factors such as the degree of
      house finish, the configuration of the piping run, and
      the practices of the mitigator.
   -  Sealing  the slab. The additional cost of caulking a
      perimeter wall/floor joint (when the joint is a narrow
      crack) might be roughly $30 to $360, depending heavily
      on slab  size, and depending upon mitigator practices
      (e.g., degree of surface preparation). When the perim-
      eter joint is a channel drain which is to be sealed as
      illustrated in Figure 32a, the additional cost could be
      roughly $150 to $700. These figures assume that the
      perimeter joint is accessible, and do not include any
      effort to remove and restore finish.

   -  Conducting pre-mitigation suction field extension di-
      agnostics. In cases where a separate trip to the house is
      made to conduct the  diagnostics, this step could add
      roughly $170 to $240 to the installation cost. The cost
      could vary outside this  range,  depending  upon the
      distance of the house from the mitigator's offices, and
      upon the exact nature of the diagnostics performed.

The baseline cost estimates suggested that a reasonable instal-
lation cost range for relatively straightforward systems would
be perhaps $800-$1,700,  when the system is installed by a
mitigator. From the discussion immediately above, it can be
seen that, with more complicated systems, the  upper end of
the installation cost range could increase to perhaps $2,500
or higher. Many mitigators will encounter some or all of these
complications on occasion.

The preceding discussion of installation costs assumes that the
system is installed by a professional mitigator. The use of a
mitigator is recommended, because a proficient mitigator will
have extensive experience with how systems must be de-
signed and installed in order to be effective in houses in that
area. Homeowners should not undertake mitigation installa-
tions as a "do-it-yourself project unless they have carefully
reviewed this  manual and  inspected installations  in other
houses similar to their own, and feel comfortably conversant
with the principles behind  ASD  systems.  If a homeowner
installed a SSD system on his/her own, the installation cost
would consist of the cost of materials: about $170-$480 for
the relatively straightforward system, somewhat higher for
more complicated systems.


 13.1.2    SSD Operating Costs

Fan electricity and house heating/cooling penalty. Based
upon the assumptions indicated in Table 5, the fan electricity
and house heating/cooling costs for SSD systems will often
fall within the following ranges:
  Fan electricity
  Heating penalty (cold weather)  -
  Cooling penalty (hot weather)
$ 30-75
   5-175
   2-45
$ 40-300 per year
This corresponds to about $3 to $25 per month in increased
utility bills, on average.

The lower extreme in this range assumes the least fan power
consumption and the lowest cost of electricity in Table 5,
combined with the lowest exhaust rate, the mildest climate,
                                                       283

-------
and the lowest electricity and fuel costs. And conversely, the
upper extreme assumes the combination of the maximum fan
power consumption, maximum exhaust rate, most extreme
climate, and highest electricity and fuel costs. Such combina-
tions of best-case conditions on the one hand, or worst-case
conditions on the other, will probably not frequently occur in
practice. Thus, in most cases, the actual costs of fan power
and heating/cooling will likely fall within a narrower range
than that calculated above, probably within the range of $50 to
$200.

System maintenance costs. In addition to the continuing
fan electricity and house heating/cooling costs, SSD operating
costs will also include some periodic maintenance expenses.
These will likely include occasional repair or replacement of
the fan, and occasional repair of piping and foundation seals.

With the relatively limited duration over which most SSD
systems have been operating (generally 7 years or less), there
is not a long-term statistical data base defining the average
lifetime of SSD fans, or the frequency of repairs required. The
current assumption is that the average fan will have to be
replaced every 10 years, at a total cost of roughly S150 or
more each time, including the cost of the fan and  of the
mitigator labor.

If the fan failed during the manufacturer's warranty period,
usually 1 to 3 years, there would be no charge for the repaired
or replaced fan. If the fan failed  during the mitigator's war-
ranty period, there would be no charge for the labor.

If the fan  could be repaired  on-site, e.g., by replacing the
electrolytic capacitor in  the  fan circuitry,  the cost  would
consist of the mitigator's labor. While this could vary from
site to site, it would probably be on the order of S50 (unless,
again, the  failure occurred during  the mitigator's warranty
period).

If important seals rupture over the  years (e.g., due to house
shifting over wet and dry seasons, or  to physical impacts
against the system piping), these  seals can probably often be
repaired by  the occupant an no significant expense.  If a
mitigator (or other service person) is hired to do this job, the
cost would depend upon the extent of re-sealing required.

Follow-up radon measurements. EPA recommends that
tte mitigator suggest that occupants conduct follow-up indoor
radon measurements at least  once  every 2 years following
installation. The occupant may  wish to consider measure-
ments on a more frequent interval.  Assuming that the occu-
pant uses a charcoal or alpha-track detector, the cost of each
measurement would probably be within the range of $10 to
$40.

These measurements should be made in a part of the house
where the occupants spend a lot of time. If certain members of
the family spend much time in the basement, the occupant
may wish to make simultaneous  measurements in the base-
ment and upstairs, which would increase the measurement
cost, requiring two detectors each  time instead of only one. To
make sure that periodic follow-up measurements are compa-
rable, so that any degradation in system performance can most
reliably be detected, each follow-up measurement should be
made at the same location in the house during the same time
of year, and under the same house conditions, as the preceding
follow-up measurements.

13.2  Sump/Drain-Tile
Depressurization  Costs

13.2.1    Sump/DTD Installation Costs

The study of ASD installation costs discussed in Section 13.1
(He91b, He91c) considered the incremental cost of installing
a sump/DTD system rather than a SSD system.

The mitigators participating in that study estimated that instal-
lation of a sump/DTD system would  cost about  $16 to $50
more than a SSD system, due to the need to install a cover
over the sump. This estimate assumes a relatively simple
sump cover.  It also assumes  that a submersible pump is
already present in the sump, so that the sump pump does not
have to be replaced as part of the system.

Therefore, within the accuracy of these cost estimates, the
estimated installation cost for a baseline (relatively straight-
forward) sump/DTD system would be on the same order as
that for a baseline SSD system, i.e., $800-$!,700 when in-
stalled by a mitigator, as discussed in Section 13.1.1.

The installation cost for a sump/DTD system could be some-
what higher if a more elaborate cover is installed. If the sump
does not already have a submersible sump pump, and if a
submersible pump must therefore be installed as part of the
sump/DTD system, this would further increase the cost of the
sump/DTD system  by  roughly $100-$200. In addition, any
complications on top of the baseline system—e.g., additional
slab sealing—would increase the baseline sump/DTD cost,
just as discussed for the SSD case.

Within the accuracy of these cost estimates, the cost increases
that could result from more elaborate sump covers or sump
pump replacement are felt to fall within the range of the cost
increases that could result from the other complications that
might arise. Accordingly, it is estimated that the upper end of
the  installation cost range for sump/DTD  systems would be
about $2,500  or higher, as is the case for more complicated
SSD systems.

13.2.2    Sump/DTD Operating Costs

Fan electricity and house heating/cooling penalty. As
indicated in Table  5,  sump/DTD systems are assumed to
increase the house ventilation rate to a somewhat greater
extent than do SSD systems,  thus increasing the heating/
cooling penalty.

Using the same format as that in  Section 13.1.2 for SSD
systems, the  ranges of fan  electricity and house heating/
cooling costs for sump/DTD systems are estimated to be:
                                                     284

-------
  Fan electricity
  Heating penalty (cold weather)
  Cooling penalty (hot weather)
$ 30-75
  10-250
   5-70

$ 40-400 per year
This compares with the range $40 to $300 cited previously for
SSD systems. The lower end of the range remains basically
the same for sump/DTD as for SSD because, at the very mild
climate assumed for the  best-case  condition, doubling the
minimum house air exhaust rate from 5 to 10 cfm (see Table
5) has very little impact on heating and cooling costs.

As discussed in Section 13.1.2, this full range is defined by
the best-case and worst-case combinations of conditions. In
practice, the actual costs  will likely fall within a narrower
range, probably $50 to $250.

System maintenance costs, and follow-up radon mea-
surements. The costs for system maintenance and for
follow-up monitoring would be  the same for sump/DTD
systems as for SSD systems, discussed in Section 13.1.2.


13.3  Drain-Tile Depressurization/
Remote Discharge Costs

13.3.1    DTD/Remote Discharge
Installation Costs

Based upon limited experience with DTD/remote discharge
systems, it is estimated that the typical installation costs of
these systems will likely be in the same range as  that for
relatively straightforward SSD systems, i.e., $80041,700 when
installed by a mitigator. This assumes that the cost of install-
ing a SSD pipe beneath the slab will be generally comparable
to that of excavating the drain  tile and/or the drain tile
discharge line  outside the house for the DTD/remote dis-
charge system. The effort required for this excavation will, of
course, depend upon the depth of the tiles at the point of
excavation.

For the purposes of this analysis, it is assumed that the
installation cost of a stand-alone DTD/remote discharge sys-
tem will generally not extend much above the baseline values.
Unless SSD suction pipes have to be added into the system, it
is thought to be less likely that the system will  become
sufficiently  complicated to increase the cost to  $2,500 or
higher, as was estimated for complicated SSD and sump/DTD
systems. Complications might arise with DTD/remote dis-
charge systems, such as the need to seal the slab or to excavate
to a greater depth to reach the drain tiles. However, it  is
believed that in many cases the added costs resulting from
these complications will not be sufficient to increase the costs
dramatically above the baseline range.

 73.3.2    DTD/Remote Discharge
Operating Costs

Fan electricity and house heating!'cooling, penalty. As
indicated in Table 5, DTD/remote discharge systems are
being assumed to cause the same increase in house ventilation
rate as do SSD systems. Although the total exhaust rates from
DTD/remote discharge systems are higher, the fraction of
house air in these exhausts is assumed to be lower, since the
tiles are commonly outside the footings.

On this basis, the full range of annual operating costs for fan
electricity and heating/cooling penalty is the same for DTD/
remote discharge as for SSD, i.e., $40 to $300 per year. And in
practice, the actual costs will likely fall within the narrower
range of perhaps $50 to $200 in most cases.

System maintenance costs, and follow-up radon mea-
surements. The cost for system maintenance and for follow-on
monitoring should be roughly the same for DTD/remote dis-
charge systems as for the other ASD systems discussed previ-
ously.

13.4   Block-Wall Depressurization
Costs

 13.4.1     BWD Installation  Costs

Experience with stand-alone BWD systems  is much  more
limited than is experience with the other ASD variations, so
that estimates of likely installation costs are more uncertain.

Individual-pipe variation.  In basement houses that are
relatively amenable to stand-alone individual-pipe BWD sys-
tems—i.e., where the basement is unfinished and the  walls
thus readily accessible, and where the major wall openings
(especially the top block voids) are reasonably accessible for
closure—EPA's research experience (Sc88) suggests that a
homeowner might expect to pay roughly $1,500 to $2,500 to
have the  system installed by a .mitigator. This cost range is
greater than that indicated earlier  for relatively straightfor-
ward SSD systems, because of: a) the increased number of
suction pipes that will likely be needed (at least one per wall
for the BWD system, compared to one total for the SSD
system); and b) the increased wall sealing effort that may be
required.

If the basement were partially finished, complicating both the
installation of pipes and the sealing of the walls, or if the
major wall openings were otherwise less accessible for clo-
sure, then the installation costs could be even higher (e.g.,
$3,000 or more).

Baseboard duct  variation.  Vendors of baseboard duct
BWD systems (who have had experience installing similar
systems over the years for basement water control)  indicate
that such systems can sometimes be installed for $2,000 to
$2,500 (E188). These would presumably be  the costs that
would  be obtained when the baseboard consists of plastic
channel drain which is attached using epoxy adhesive, and
when the basement is not highly finished and does not present
unusual complications. If the basement is heavily finished,
costs would be expected to be higher, due to the added costs
of, e.g.: trimming the wall and floor finish to expose the wall/
floor joint and accommodate the baseboard duct (and refinish-
ing afterwards); penetrating finished  stud walls which run
perpendicular to the block wall; and removing and replacing
                                                     285

-------
various  obstructions that block access to  the walls. Costs
might also increase if the baseboard system is also expected to
fulfil a water drainage function, and if a sump and sump pump
need to be installed in conjunction with the system. Thus, in
some cases, baseboard duct systems might be expected to cost
more than 52,500 (e.g., $3,000 or more).
13.4.2    BWD Operating Costs

Fan electricity and house heating/cooling penalty.  A
stand-alone BWD system with a single 90-watt tubular fan
can commonly have flows in the range of 100 to 200 cfm,
higher than the range for either SSD or even DTD systems.
Because the block walls are so leaky, it might be assumed that
a relatively high fraction of this exhaust (50 to 80%) is treated
house air drawn from inside the house, although there are not
definitive tracer gas measurements  to confirm that this is  in
fact the case.

On this basis, as indicated in Table 5, it is estimated that a
single 90-watt fan could increase the house ventilation rate by
50 to 160 cfm. If a second fan  is needed, as discussed  in
Sections 7.4 and 7.6, the increase  in house ventilation rate
could be 100 to 320 cfm.

Because of the high flows from stand-alone BWD systems,
only the high-flow 90-watt tubular fans are considered for this
cost estimate. Also, it is assumed that the fan will be operating
closer to its maximum flow rate (and hence maximum power
consumption) that would be the case in SSD systems. Thus,
the assumed power consumption in  Table 5 is 75 to 90 watts
for one fan, and 150 to 180 watts for two.

With a single fan, the range of annual costs for fan electricity
and for house heating/cooling, calculated using the assump-
tions in Table 5, would be:
  Fan electricity
  Heating penalty (cold weather)
  Cooling penalty (hot weather)
S 40-75
  20-350
  10-90

S 70-500 per year
This represents about $6 to $40 per month, on average. In
practice, observed costs would likely fall within a narrower
range of perhaps S100 to $400.

If two fans were used, the costs for fan electricity and for the
heating/cooling penalty would double.

System maintenance costs, and follow-up radon mea-
surements. The costs for system  maintenance and for
follow-up  radon  monitoring would be  about the same for
BWD systems  as for the other ASD variations, as discussed
previously. The one exception would be that, if  the BWD
system had two fans, the costs for fan repair or replacement
would presumably double.
13.5  Sub-Membrane
Depressurization Costs

13.5.1    SMD Installation Costs

Relatively simple systems. The study of ASD installation
costs discussed for SSD systems in Section 13.1.1 (He91b,
He91c) also considered  the costs of SMD systems in
crawl-space houses.

In this study, the "baseline"  crawl-space houses  addressed
two house design variables (small vs. large crawl spaces and
one vs. two stories). In each house, the baseline SMD system
was defined by 13 system design variables. These system
design variables  defined,  e.g., the  method of distributing
suction beneath the  membrane (individual-pipe vs.
sub-membrane piping approach), the exhaust configuration,
and the degree of membrane sealing.

The baseline SMD system for each house was selected to
represent the simplest case for that house. For example, the
baseline systems were individual-pipe systems, with the mem-
brane not sealed anywhere.

Again, estimates were obtained from each of five  mitigators
for these baseline SMD systems,  as discussed in Section
13.1.1. These mitigators represented a wide range of different
practices and markets around the U. S., and a range of labor
and overhead rates.

The estimated installation cost for the baseline (i.e., relatively
straightforward) SMD systems of the individual-pipe configu-
ration in a crawl-space house with the membrane unsealed
ranged between roughly $1,000 and $1,900 (in 1991 dollars).
These costs are somewhat higher than those cited  earlier for
SSD, due to the labor and materials costs involved in install-
ing a membrane over the crawl-space floor. The quality of the
membrane  material (i.e.,  the use of the more  expensive
cross-laminated polyethylene  rather than regular polyethyl-
ene) will have an important impact on installation cost.

As indicated, this cost range covers a range of house charac-
teristics and a range of mitigation practices and market condi-
tions. For this reason, this range is felt to fairly well represent
what  a homeowner might expect to  pay  for a simple,
individual-pipe SMD system without sealing of  the mem-
brane.

As discussed in Section 8.5.1  (see Sealing the membrane -
general), it is generally recommended that the membrane be
completely sealed. From the analysis in Reference He91b, the
cost estimated by the mitigators for sealing the membrane
everywhere using beads of caulk or other sealant ranged from
$50 to $330, depending heavily on the size of the crawl space.
As discussed in Section 8.1.1 (see Cost of sub-membrane
piping), the addition of perforated piping beneath  the mem-
brane might be expected to add another $10  to $100 to the
installation.

With  the additional cost  of  the recommended membrane
sealing, and including the option of installing sub-membrane
piping, the previous statement regarding SMD installation
                                                      286

-------
cost can be modified as follows: The estimated installation
cost for typical SMD systems of either the individual-pipe or
sub-membrane piping configuration, with the membrane fully
sealed, is generally in the range of $1,000 to $2,500.

More complicated systems. Various complications might
arise which  could increase the cost of installing a  SMD
system. Some of these factors which could  increase costs
could be: additional individual suction pipes through the
membrane; additional finish around the exhaust piping up
through the  house; large or complex crawl spaces;  more
elaborate efforts to attach the membrane to the perimeter
walls,  using a furring  strip (usually  not necessary); or
higher-quality membrane materials. In extreme cases, some
mitigators have reported excavation of parts of a crawl space,
when there is initially insufficient headroom  to provide the
access needed to install and seal the membrane everywhere.

Some of these steps, such as additional individual suction
pipes, should not cause the installation cost to exceed the
$2,500 upper value of the range cited above.  Other of these
steps might result in total costs above $2,500. Accordingly,
while it is expected that installation costs will often be within
the range of $1,000 to $2,500, installation costs could in-
crease above $2,500 for some more  complicated systems.

The costs indicated above assume that the system is installed
by a professional mitigator. As discussed in Section 13.1.1,
the use of a mitigator is  recommended. If a homeowner
installed  a SMD on his/her own, the installation cost would
consist of the cost of materials, about $2004650, depending
upon the size of the crawl space, the quality of the membrane
material used, and other variables.

 13.5.2     SMD Operating Costs

Fan electricity an house heating/cooling penalty. The
impact of the SMD system on the house heating and cooling
costs will depend upon: a) whether the crawl space is isolated
from the living area and vented; or b) whether it is open to the
living area and is not vented, in which case it would be
conditioned space. Table 5 lists both situations.

If the crawl  space is vented, it is assumed for these calcula-
tions that 25  to 50% of the crawl-space air that is exhausted by
the SMD system is conditioned air that was drawn into the
crawl space from the living area. This is a relatively arbitrary
assumption, since there are no tracer gas data confirming what
this percentage will typically be. On this basis, the increase in
the house ventilation rate is calculated assuming that 0  to
 100% of the exhaust air is drawn from the crawl space (which
has been shown by limited tracer gas data), and that 25 to 50%
of this is conditioned air. See Table 5.

With this assumption, and with the other assumptions in Table
5, the range of annual fan electricity and heating/cooling costs
for SMD systems in vented crawl spaces is as follows:
                         This represents about $2 to $20 per month, on average. In
                         practice, observed costs would likely fall within a narrower
                         range of perhaps $30 to $150.

                         If the crawl space is unvented, and if it is open to the living
                         area, then 100% of the air extracted from the crawl space by
                         the SMD system will be conditioned house air, rather than just
                         25 to 50%. With this  assumption, the range of annual fan
                         electricity and heating/cooling  costs for SMD systems in
                         unvented crawl spaces would be as follows:
                           Fan electricity
                           Heating penalty (cold weather)
                           Cooling penalty (hot weather)
                                    $ 30-75
                                       0-240
                                       0-60

                                    $ 30-375 per year
  Fan electricity
  Heating penalty (cold weather)
  Cooling penalty (hot weather)
-   $ 30-75
       0-120
       0-30

$   30-225 per year
This represents about $2 to $30 per month, on average. In
practice, observed costs would likely fall within a narrower
range of perhaps $50 to $250.

It should be noted that, when an unvented crawl space exists
which opens to the livable area as in this situation, it will often
be a crawl space opening to an adjoining basement. In such
cases, there will commonly also be a SSD or DTD system in
the basement, supplementing the  SMD system in the crawl
space. In evaluating the total operating  cost of the combined
ASD system, the operating cost of the SSD or DTD compo-
nent would also have to be considered.

System maintenance costs. One element of system mainte-
nance costs will be periodic repair or replacement of the SMD
fan. The costs for fan repair/replacement in SMD  systems
should be about the same as for SSD systems, discussed in
Section 13.1.2:

   -  roughly $150 or more for each complete replacement
      of the fan, potentially required once every 10 years, if
      the  replacement is not covered by warranty.

   -  perhaps $50, if the fan  can be repaired on site by the
      mitigator (e.g., by replacing a capacitor), again if not
      covered by warranty.

With SMD systems, there may be another major maintenance
cost: periodic repair or replacement of  the membrane. There
are insufficient historical data to  determine how frequently
this may have to be done, or the extent  of the job that will be
required.

Where the problem consists of a limited number of ruptures
caused by, e.g., traffic in the crawl space by the occupant or
by service personnel, the ruptures may sometimes be repaired
by the homeowner. Perhaps duct tape or other sealant to cover
the ruptures may be sufficient, or perhaps some section of the
membrane may have to be replaced. If a mitigator is called to
make these repairs, and if a couple hours of labor are required
(plus perhaps  some  membrane material), the cost of these
repairs to the homeowner might be expected to be on the order
of$50-$150.
                                                       287

-------
However, in some cases, it may be possible that essentially
the entire membrane may have to be replaced.  Extensive
damage to membranes by rodents over relatively brief periods
have been reported by one mitigator (Wi91). In addition, the
membrane may also ultimately deteriorate over the years due
to UV radiation, extended wear and tear, and other aging
phenomena. It cannot be estimated at this time how frequently
such complete replacements of the membrane may be neces-
sary. Removal of the old membrane and installation of a new
membrane by a mitigator might be expected to cost a home-
owner on the order of $250-$1,100, or perhaps even more,
depending upon the extent of membrane sealing required, the
size of the crawl space, and the quality of membrane material
used.

Follow-up radon measurements. The costs of follow-up
indoor radon measurements with SMD systems would be the
same as that discussed previously for other ASD variations,
namely, S10-S40 for each measurement (assuming charcoal or
alpha-track detectors), with at least one measurement made at
least once every 2 years.

13.6  Active Soil Pressurization
Costs

13.6.1     Soil Pressurization Installation
Costs
Active soil pressurization installations will differ from the
corresponding ASD  approach (i.e., from SSD or BWD) in
several ways which can influence installation cost First, no
exhaust stack will be  required for pressurization systems,
which should reduce the installation cost by roughly $50-$325
compared to the corresponding depressurization system. How-
ever, these cost savings may be partially offset by the need to
install an air filter in the inlet air piping (see Figure 40), and a
screened air intake.

In summary, a pressurization system might cost  somewhat
less to install than would the corresponding depressurization
system, at least in the prevailing case where an exhaust stack
is required for the depressurization system. However, due to
the limited experience with pressurization systems, it should
be assumed that the installation cost of a sub-slab pressuriza-
tion system will be generally in the range indicated in Section
13.1 foraSSD system (i.e., $800-$1,700 for relatively straight-
forward systems installed by a mitigator, up to $2,500 or more
for more complicated systems).

Likewise, the installation cost for a block-wall pressurization
system should be assumed to be generally in the same range as
indicated in Section 13.4 for BWD systems (i.e., $1,500-$2,500
for relatively straightforward systems, up to $3,000 or more
for more complicated systems).
            Soil Pressurization Operating
13.6.2
Costs
Fan electricity and house heating/cooling penalty. Ac-
tive sub-slab pressurization systems will be installed where
flows beneath the slab are relatively high. However, there are
no definitive data to suggest that the increased soil flows will
cause a sub-slab pressurization system to increase the ventila-
tion rate of the house any more or less  than would a SSD
system. Likewise, a block-wall pressurization system would
be expected to impact house ventilation rate in about the same
manner as would a corresponding BWD system.

As a result, there is currently no basis for expecting that the
fan electricity or the house heating/cooling penalty for pres-
surization systems would be significantly different from those
indicated earlier for the corresponding depressurization  sys-
tems. Thus,  the electricity plus heating/cooling penalty for
sub-slab pressurization systems would be expected to be in
the range of S40-S300 per year. Costs for block-wall pressur-
ization systems would be in the range of $70-$500 per year for
single-fan systems (or double that for two-fan systems).

System maintenance costs. In addition to the fan and seal
maintenance requirements discussed previously for active
depressurization systems,  active soil pressurization systems
will experience some costs associated with cleaning/replace-
ment of the filter in the air inlet, and possibly with removing
deposited dust from the pit beneath  the slab.  See Section
9.5.1.

Within the accuracy of these maintenance cost estimates, it is
assumed that the homeowner personally takes care of this
maintenance unique  to the pressurization system. It is as-
sumed  that the costs involved with replacement air filters is
small. Thus, the maintenance costs for active soil pressuriza-
tion systems are  assumed to be about the same as for the
corresponding depressurization systems.

Follow-up radon measurements. The costs of follow-up
indoor  radon measurements with active soil pressurization
systems would be the same as discussed previously for de-
pressurization systems.

13.7  Passive Soil  Depressurization
Costs

13.7.1    Passive Soil Depressurization
Installation Costs

The following discussion assumes that the passive system is
retrofit into an existing house in exactly the same manner as
an active system would be installed. That is, there would be no
additional  effort to improve the distribution of the passive
suction field by retrofitting improved sub-slab aggregate or
sub-slab perforated drain tiles beyond what existed originally.
As discussed in Section 10, without such efforts (or perhaps
even with  such efforts), passive systems  will perform much
less effectively than will an active system. Efforts to retrofit
sub-slab aggregate or drain tiles into existing houses would, of
course, dramatically increase the system costs.

Passive soil depressurization installations will differ from the
corresponding ASD systems in several ways which can influ-
ence installation cost.

   - No fan will be required for passive systems, so that the
     material cost for the fan and the cost for installing and
                                                      288

-------
      wiring the fan will be eliminated. This could reduce the
      installation cost by perhaps $200-5250.

   -   The system stack will have to rise up inside the house;
      an exterior stack (or elimination of the stack) is not an
      option. Where a mitigator would be installing an inte-
      rior stack anyway, even if the system were to be active,
      the requirement that the passive stack be indoors repre-
      sents no additional installation cost. But where the
      mitigator would otherwise prefer to install an exterior
      stack, and where there are no provisions (such as an
      existing  utility chase) to simplify installation of an
      interior stack, the requirement for an interior stack
      could increase costs by over $100 (He91b).

   -   If additional passive SSD suction pipes must be in-
      stalled through the slab  to get adequate performance
      from the passive system, installation costs would in-
      crease by $704380 per pipe added, as indicated in
      Section 13.1.1.

Based upon the above points,  it is not possible to make a
general statement regarding whether a passive system will be
less or more expensive to install than an active system. In
some cases passive systems may be less expensive; in other
cases they may be of comparable cost or more expensive. But
in all cases, they will be significantly less effective in reduc-
ing indoor radon levels.

13.7.2    Passive Soil Depressurization
Operating Costs

Fan electricity and house heating/cooling penalty. Pas-
sive soil depressurization systems will avoid the need for
electricity to operate a system  fan (assuming that a supple-
mental fan does not turn out to be required, as discussed in
Section 10). On this basis, the $30 to $75 per year included in
prior sections for the electricity to run the fan would be
eliminated altogether.

The house heating/cooling penalty would also be reduced
dramatically. Whereas the active SSD, DTD, and SMD sys-
tems are estimated to exhaust up to 120 cfm of conditioned air
from the house, passive systems may exhaust only a few cfm
total. Thus, the total heating/cooling penalty might be on the
order of only perhaps $5-$10 per year.

But again, it is emphasized that this dramatic reduction in
electricity plus heating/cooling cost—from $40-$300 per year
with an active system to $5-$10 with a passive  system—is
achieved only through a dramatic reduction in system perfor-
mance.

System maintenance costs.  With passive systems, the need
for periodic repair or replacement of the fan would be elimi-
nated.

However, the need to maintain system and foundation seals
may become even more important with passive systems. At
the low flows involved, short-circuiting  of house  air into the
system could have a significant effect on the extension of the
already-weak suction fields.

Follow-up radon  measurements. The need for follow-up
radon measurements would be increased with passive sys-
tems.

In addition to the normal follow-up measurements at least
once every 2 years, recommended by EPA for active systems,
persons installing passive systems must be prepared to make
frequent measurements over the first year of operation. These
measurements would be intended to identify the combinations
of weather conditions and appliance operating  conditions
under which the passive system is likely to be overwhelmed,
so that a supplemental fan can be installed and activated as
needed.
13.8  Summary of ASD Installation
and  Operating Costs
The ASD installation  and operating costs described previ-
ously in this section are summarized in Table 6.
                                                      289

-------
Tablo 6.
Summary of Installation and Operating Costs for ASD and Related Systems
Mitigation
System
SSO



Sump/DTD



DTD/
remote
discharge

BWD






SMO










Installation Costs
(by Mitigator) Fan Elect.+
Simple Complicated Heat/Cool
System System (per year)
$800-$1,700 to $2,500+ $40-$300
(full range);
$50-$200
(more typical)
$800-$ 1,700 to $2,500+ $40-$400
(full range);
$50-$250
(more typical)
$800-$1,700 to $1,700+ $40-$300
(full range);
$50-$200
(more typical)
$1,500-$2,500 to $3,000+ $70-$500
(full range,
if one fan);
$100-$400
(more typical).
Double if two
fans.
$1.000-$2,500 to $2,500+ $30-$225
(full range,
vented CS1);
($30-$150
typical).

$30-$375
(full range,
unvented*);
$50-250
(more typical).
Operating Costs
System
Maintenance
Repair fan (~$50)
or replace (~$1 50)
every 10 years.
Repair seals.
As above.



As above.



As above (if one
fan). Double fan
costs if two fans.




Repair/replace fan
every 10 years, as
above. Repair
membrane ($50-
$150) or replace
($250-$1,100)at
unknown intervals.





Follow-up
Monitoring
At least once
every 2 years
<5> $10-$40.

As above.



As above.



As above.






As above.









!
Active
  prassur*
  tzation
Passive
  depress-
  u rization
         Comparable to
         corresponding
         depressurization
         technology.
         May be less or
         more expensive
         than correspond-
         ing active
         systems*.
Comparable
to depress.
technology.
May be less or
more expensive
than active
systems.
Comparable
to depress.
technology.
$0 for fan
electricity
(assuming no
supplemental
fan required).
$5-$10 for
heat/cooling
penalty2.
As above for
SSD and BWD.
Fan repair/replace-
ment unnecessary.
Seal repair of in-
creased importance.
As above.
Increased
monitoring
during first
year to
understand
when system.
overwhelmed.
  Vented crawl spaces contain foundation vents permitting ventilation of the crawl space by outdoor air; the crawl space is considered
  non-conditioned space, isolated from the living area. Un vented crawl spaces are not ventilated by outdoor air, and are considered conditioned
  space; they may be open to livable space (such as adjoining basements).
  Assumes that no major effort is undertaken with passive systems to retrofit perforated piping or aggregate beneath the slabs of existing
  houses. Such retrofit steps would dramatically increase the installation cost of passive systems. Note that any reductions in installation and
  operating costs for passive systems are achieved at the expense of a major degradation in radon reductions.
                                                              290

-------
                                           Section  14
                                           References
An92
  Anderson, J. W., Quality Conservation, Spokane, WA,
  personal communication, March 1992.

Ang92
  Angell, W. J., Midwest Universities Radon Consortium,
  University of Minnesota, Minneapolis, MN, personal com-
  munication, March 31,1992.

Ar82
  Arix Corp., "Planning and Design for a Radiation Reduc-
  tion Demonstration Project, Butte, Montana," Report to
  the Montana Department of Health and Environmental
  Sciences, Appendix C, January 1982.

ASHRAE88
  American  Society of  Heating,  Refrigerating  and
  Air-Conditioning Engineers, Inc., "1988 ASHRAE Hand-
  book: Equipment," pp. 3.2-3.3, Atlanta, GA, 1988.

ASHRAE89
  American  Society of  Heating,  Refrigerating  and
  Air-Conditioning Engineers, Inc., "Ventilation for Ac-
  ceptable Indoor Air Quality: Positive Combustion Air
  Supply," ASHRAE Standard 62-1989, Appendix B, At-
  lanta, GA, 1989.

ASTM83
  American Society for  Testing  and Materials, "Standard
  Test Method for Determining Air Leakage Rate by Tracer
  Dilution," ASTM E 741-83, Philadelphia, PA, 1983.

ASTM87
  American Society for  Testing  and Materials, "Standard
  Test Method for Determining Air Leakage Rate by Fan
  Pressurization," ASTM E 779-87, Philadelphia, PA, 1987.

Ba92
  Bainbridge, R. S., Aarden Testing, Sarasota, FL, personal
  communication, February 4,1992.

Bar90
  Bartholomew, J. C., Quality Conservation, Spokane, WA,
  personal communication, May 23,1990.

Be84
  Becker, A.  P., and T. M. Lachajczyk, "Evaluation of
  Waterborne Radon Impact on  Indoor Air Quality and
  Assessment of Control Options," EPA-600/7-84-093
  (NTIS PB84-246404), September 1984.
Bo91
  Bohac, D. L., L. S. Shen, T. S. Dunsworth, and C. J.
  Damm, "Radon Mitigation Energy Cost Penalty Research
  Project," report prepared by the Minnesota Building Re-
  search Center, University of Minnesota, September 15,
  1991.

Br89
  Brennan,  T. M., M. R. Watson, C. E. Kneeland, J. P.
  Reese,  and W. Evans,  "Pressurizing Beneath Slabs in
  New York State," presented at The 1989 Annual AARST
  National Conference, sponsored by the American Asso-
  ciation of Radon Scientists and Technologists, Bethesda,
  MD, September 1989.

Br91a
  Brennan, T. M., M. E. Clarkin, M. C. Osborne, and W. P.
  Brodhead, "Evaluation  of Radon Resistant New Con-
  struction Techniques," in Proceedings: The 1990 Interna-
  tional Symposium on Radon and Radon Reduction Tech-
  nology.  Volume 2,  EPA-600/9-91-026b  (NTIS
  PB91-234450), pp. 8-1 through 8-13, July 1991.

Br91b
  Brennan, T. M., "Interpreting the Vacuum Suction Test,"
  in Proceedings:  The  1990 International Symposium on
  Radon  and Radon Reduction Technology, Volume 3,
  EPA-600/9-91-026c  (NTIS PB91-234468), pp. 5-15
  through 5-20, July 1991.

Br92
  Brennan,  T. M., Camroden Associates, Inc., Oriskany,
  NY, personal communication, March 26,1992.

Bro90
  Brodhead, W. P., WPB Enterprises, Inc.,  Riegelsville,
  PA, personal communication, April 26,1990.

Bro92
  Brodhead, W. P., WPB Enterprises, Inc.,  Riegelsville,
  PA, personal communication, March 11,1992.

Bru83
  Bruno, R. C., "Sources of Indoor Radon in Houses: A
  Review,"  Journal of  the Air Poll. Control Assoc.,
  33(2): 105-109,1983.

Ca60
  Carrier Air Conditioning Company, "Carrier System De-
  sign Manual: Part 2 - Air Distribution," Syracuse, NY,
  1960.
                                                    291

-------
C191
  Clarkin, M.  E., T. M. Brennan, and D.  Fazikas, "A
  Laboratory Test of the Effects of Various Rain Caps on
  Sub-Slab Depressurization Systems," inProceedings: The
  1991 International Symposium on Radon and Radon Re-
  duction Technology. Volume 3, EPA-600/9-91-037c (NTIS
  PB92-115377),pp.P4-31 through P4-4 I.November 1991.

CMHC88
  Canadian Mortgage and Housing Corp., "Procedure for
  Determining  the Safety of Residential Chimneys," in
  Chimney Safety Tests User's Manual (Second Edition),
  Ottawa, Ontario, January 1988.

Cr91
  Craig, A. B.,  K. W. Leovic, D. B. Harris, and B. E. Pyle,
  "Radon Diagnostics and Mitigation in Two Public Schools
  in Nashville,  Tennessee,"  in Proceedings: The 1990 In-
  ternational Symposium on Radon and Radon Reduction
  Technology.  Volume 2,  EPA-600/9-91-026b  (NTIS
  PB91-234450), pp. 9-15 through 9-33, July  1991.

Cr92a
  Craig, A. B., U. S. Environmental Protection Agency,
  Research Triangle Park, NC,  personal communication,
  January 1992.

Cr92b
  Craig, A. B., D. B. Harris, and K. W. Leovic, "Radon
  Prevention in Construction of Schools and Other Large
  Buildings—Status of EPA's Program," in Proceedings:
  The 1992 International Symposium on Radon and Radon
  Reduction Technology.  Volume 2, EPA-600/R-93-083b
  (NTIS PB93-196202), pp. 10-151 through  10-171, May
  1993.

Cra91
  Crawshaw, D. A., and G.  K. Crawshaw, "Mitigation by
  Sub-Slab Depressurization Under Structures Founded on
  Relatively Impermeable Sand," in Proceedings: The 1991
  International Symposium on Radon and Radon Reduction
  Technology.  Volume 3,  EPA-600/9-91-037c  (NTTS
  PB92-115377), pp. P4-15 through P4-30, November 1991.

Cu92
  Cummings, J. B., J. J. Tooley, Jr., and N. Moyer, "Radon
  Pressure Differential Project, Phase I. Florida Radon Re-
  search   Program,"   EPA-600-R-92-008   (NTIS
  PB92-148519), January 1992.

DcP91
  DePierro, N., T. Key, and J. Moon, "The Effectiveness of
  Radon Reduction in New Jersey," in Proceedings: The
  1990 International Symposium on Radon and Radon Re-
  duction Technology. Volume 2, EPA-60Q/9-9 l-026b (NTIS
  PB91-234450), pp. 7-83 through 7-99, July  1991.

Di86
  Dietz, R. N., R. W. Goodrich,  E. A. Cote, and R. F.
  Wieser, "Detailed Description and Performance of a Pas-
  sive Perfluorocarbon Tracer System for Building Ventila-
  tion and Air Exchange Measurements," in Measured Air
  Leakage of Buildings, ASTM Special Technical Publica-
  tion 904, H. R. Trechsel and P. L. Lagus, Eds., American
  Society for Testing and Materials, pp. 203-264, August
  1986.

DOE87
  U. S. Department of Energy, Energy Information Admin-
  istration, "Household Energy Consumption and Expendi-
  tures 1987. Part I: National Data," p. 81, Washington, D.
  C., 1987.

Du90
  Dudney, C. S., L. M. Hubbard, T. G. Matthews, R. H.
  Socolowl A. R. Hawthorne, K. J. Gadsby, D. T. Harrje, D.
  L. Bohac,  and  D. L. Wilson, "Investigation of Radon
  Entry and Effectiveness of Mitigation Measures in Seven
  Houses  in New Jersey," EPA-600/7-90-016 (NTIS
  DE89016676), August 1990.

Du91
  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 Val-
  ley Houses," in Proceedings: The 1990 International
  Symposium on Radon and Radon Reduction Technology.
  Volume 2, EPA-6QO/9-91-026b (NTIS PB91-234450), pp.
  7-59 through 7-71, July 1991.

EI88
  Ellison, H., Safe-Aire, Inc., Canton, IL, personal commu-
  nication, September 1988.

EPA87a
  U. S. Environmental Protection Agency, "Radon Reduc-
  tion in New  Construction:  An Interim  Guide,"
  OPA-87-009, August 1987.

EPA87b
  U. S. Environmental Protection Agency,  "Radon Refer-
  ence Manual," EPA-520/1-87/20 (NTIS PB88-196654),
  September 1987.

EPA87c
  U. S. Environmental Protection Agency, "National Pri-
  mary and Secondary Ambient Air Quality Standards,"
  Code of Federal Regulations, Title 40, Part 50, as amended,
  July 1,1987.

EPA88a
  Henschel, D. B., "Radon Reduction Techniques for De-
  tached Houses: Technical Guidance (Second Edition),"
  EPA/625/5-87/019 (NTIS PB88-184908), January 1988.

EPA88b
  U. S. Environmental Protection Agency, "Reducing Ra-
  don in  Structures: Manual (Second Edition)," student
  manual utilized in EPA's radon mitigation training course,
  September 1988.
                                                    292

-------
EPA89a
  Mosley, R. B., and D. B. Henschel, "Application of
  Radon  Reduction Methods (Revised)," EPA/625/5-88/
  024 (NTIS PB89-205975), April 1989.

EPA895
  U. S. Environmental Protection Agency, "Radon Reduc-
  tion Techniques in Schools: Interim Technical Guidance,"
  EPA-520/1-89-020 (NTIS PB90-160086), October 1989.

EPA89c
  U. S. Environmental Protection Agency, "Radon Tech-
  nology for Mitigators: Exercises and Supplemental Mate-
  rial," supplement to the student manual utilized in EPA's
  radon mitigation training course, 1989.

EPA91a
  Clarkin, M. E., and T. M. Brennan, "Radon-Resistant
  Construction Techniques for New Residential Construc-
  tion: Technical Guidance," EPA/625/2-91/032, February
  1991.

EPA91b
  U. S. Environmental Protection Agency, "Radon Contrac-
  tor Proficiency Program Interim Radon Mitigation Stan-
  dards," December 15,1991.

EPA92a
  U. S. Environmental Protection Agency, "A Citizen's
  Guide to Radon (Second Edition)," EPA 402-K92-001,
  May 1992.

EPA92b
  U. S. Environmental Protection Agency, 'Technical Sup-
  port Document for the 1992 Citizen's Guide to Radon,"
  EPA400-R-92-011 (NTIS PB92-218395), May 1992.

EPA92C
  U. S. Environmental Protection Agency, "Consumer's
  Guide to Radon Reduction," EPA 402-K92-003, August
  1992.

EPA92d
  U. S. Environmental Protection Agency, "Indoor Radon
  and Radon Decay Product Measurement Device Proto-
  cols," EPA 402-R-92-004 (NTIS  PB92-206176), July
  1992.

EPA93
  U. S. Environmental Protection Agency, "Protocols for
  Radon and Radon Decay Product  Measurements in
  Homes," EPA 402-R-92-003, June 1993.

Er84
  Ericson, S.-O., H. Schmied, andB. Clavensjo, "Modified
  Technology in New Construction, and Cost Effective
  Remedial Action in Existing Structures, to Prevent Infil-
  tration of Soil Gas Carrying Radon," Radiation Protec-
  tion Dosimetry, 7:223-226,1984.
Fe92
  Femto-Tech, Inc., "Instruction Manual: Model RS410F
  Radon Survey Instrument," Carlisle, OH, 1992.

Fi89
  Findlay, W. O., A. Robertson, and A. G. Scott, "Testing
  of Indoor Radon Reduction Techniques in Central Ohio
  Houses: Phase 1 (Winter 1987-88)," EPA-600/8-89-071
  (NTIS PB89-219984), July 1989.

Fi90
  Findlay, W. O., A. Robertson, and A. G. Scott, "Testing
  of Indoor Radon Reduction Techniques in Central Ohio
  Houses: Phase 2 (Winter 1988-1989)," EPA-600/8-90-050
  (NTIS PB90-222704), May 1990.

Fi91
  Findlay, W. O., A. Robertson, and A. G. Scott, "Follow-Up
  Durability Measurements and Mitigation Performance
  Improvement Tests in 38 Eastern Pennsylvania Houses
  Having Indoor Radon Reduction Systems," EPA-600/
  8-91-010 (NTIS PB91-171389), March 1991.

Fis92
  Fisher, E. J., Office of Radiation Programs, U. S. Envi-
  ronmental Protection Agency, Washington, D. C., per-
  sonal communication, January 9, 1992.

Fit92
  Fitzgerald, J., Jim Fitzgerald Contracting, Minneapolis,
  MN, personal communication, July 22,1992.

Fo89
  Fowler, C. S., A. D. Williamson, B. E. Pyle, F. E. Belzer
  m, D. C. Sanchez, andT. Brennan, "Sub-Slab Depressur-
  ization Demonstration in Polk County,  Florida,
  Slab-on-Grade Houses," in Proceedings: The 1988 Sym-
  posium on Radon and Radon Reduction Technology. Vol-
  ume 1, EPA-600/9-89-006a (NTIS  PB89-167480), pp.
  7-65 through 7-78, March 1989.

Fo90
  Fowler, C. S., A. D. Williamson, B. E. Pyle, F. E. Belzer,
  and R. N. Coker, "Engineering Design  Criteria for
  Sub-Slab Depressurization Systems in Low-Permeability
  Soils," EPA-600/8-90-063 (NTIS PB90-257767), August
  1990.

Fo92
  Fowler, C. S., A. D. Williamson, B. E. Pyle, F. E. Belzer,
  and R. N. Coker, "Radon Mitigation Studies: South Cen-
  tral Florida Demonstration,"  EPA-600/R-92-207 (NTIS
  PB93-122299), October 1992.

Fu91
  Furman, R. A., and D. E. Hintenlang, "Sub-Slab Pressure
  Field Extension Studies on Four Test Slabs Typical of
  Florida Construction," in Proceedings: The 1990 Interna-
  tional Symposium on  Radon and Radon Reduction Tech-
  nology.  Volume 2,  EPA-600/9-91-026b  (NTIS
  PB91-234450), pp. 8-29 through 8-43, July  1991.
                                                    293

-------
Ga92
  Gadgil, A. J., Y. C. Bonnefous, and W. J. Fisk, "Relative
  Effectiveness of Sub-Slab Pressurization and Depressur-
  ization Systems for Indoor Radon Mitigation: Studies
  with an Experimentally Verified Numerical Model," in
  Proceedings: The 1992 International Symposium on Ra-
  don  and Radon Reduction  Technology.  Volume  1,
  EPA-600/R-93-083a (NTIS PB93-196194),  pp.  6-23
  through 6-39, May 1993.

Gad89
  Gadsby, K. J., L. M. Hubbard, D. T. Harrje, and D.  C.
  Sanchez, "Rapid Diagnostics: Subslab and Wall Depres-
  surization Systems for Control of Indoor Radon," in Pro-
  ceedings: The 1988 Symposium on Radon and Radon
  Reduction Technology. Volume 2, EPA-600/9-89-Q06b
  (NTIS PB89-167498),pp. 3-69 through 3-85, March 1989.

Gad91
  Gadsby, K. J., and D. T. Harrje, "Assessment Protocols:
  Durability of Performance of a Home Radon Reduction
  System - Sub-Slab Depressurization Systems," EPA/625/
  6-91/032, April 1991.

Gad92
  Gadsby, K. J., Princeton University, Princeton, NJ, per-
  sonal communication, March 20,1992.

Gi90
  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.

Ha89
  Harrje, D. T., L. M. Hubbard, K.  J. Gadsby, B. Bolker,
  and D. L. Bohac, "The Effect of Radon Mitigation Sys-
  tems on Ventilation in Buildings," ASHRAE Transac-
  tions, 95 (Part 1):107-113,1989.

Ha91
  Harrje, D. T., K. J. Gadsby, and D. C. Sanchez, "Long
  Term Durability and Performance of Radon Mitigation
  Subslab Depressurization Systems," in Proceedings: The
  1990 International Symposium on Radon and Radon Re-
  ductionTechnology. Volume 3,EPA-&X)/9-9l-Q26c(pmS
  PB91-234468),pp. 7-15 through 7-32, July 1991.

He87
  Henschel, D. B., and A. G. Scott, "Testing of Indoor
  Radon Reduction Techniques in Eastern Pennsylvania:
  An  Update," in Indoor Radon II: Proceedings of the
  Second APCA International Specialty Conference on In-
  door Radon, APCA Publication  SP-60, pp. 146-159,
  Cherry Hill, NJ, April 1987.

Hc89
  Henschel, D. B., and A. G. Scott, "Some Results from the
  Demonstration of Indoor Radon Reduction Measures in
  Block Basement  Houses,"  Environment International,
  15(l-6):265-270,1989.
He91a
  Henschel, D. B., A. G. Scott, A. Robertson, and W. O.
  Findlay, "Evaluation of Sub-Slab Ventilation for Indoor
  Radon Reduction in Slab-on-Grade Houses," in Proceed-
  ings: The 1990 International Symposium on Radon and
  Radon  Reduction Technology. Volume  2, EPA-600/
  9-91-026b (NTIS  PB91-234450), pp. 7-1  through 7-18,
  July 1991.

He91b
  Henschel, D. B., "Cost Analysis of Soil Depressurization
  Techniques for Indoor Radon Reduction," Indoor Air,
  1(3):337-351,1991.

He91c
  Henschel, D. B., "Parametric Analysis of the Installation
  and Operating Costs of Active Soil Depressurization Sys-
  tems for Residential Radon Mitigation," EPA-600/
  8-91-200 (NTIS PB92-116037), October 1991.

He91d
  Henschel, D. B., and A. G. Scott, "Causes of Elevated
  Post-Mitigation Radon Concentrations in Basement Houses
  Having Extremely High Pre-Mitigation Levels," in Pro-
  ceedings: The 1991 International  Symposium on Radon
  and Radon Reduction Technology. Volume 1, EPA-600/
  9-91-037a (NTIS  PB92-115351), pp. 4-3  through 4-19,
  November 1991.

He92
  Henschel, D. B., "Indoor Radon Reduction in Crawl-Space
  Houses: A Review of Alternative Approaches," Indoor
  Air, 2(4):272-287,1992.

Hi92
  Hintenlang, D. E., University of Florida, Gainesville, EL,
  personal communication, January 23,1992.

Ho91
  Hoornbeek, J., and J. Lago, "Private Sector Radon Miti-
  gation Survey," in Proceedings: The  1990 International
  Symposium on Radon and Radon Reduction Technology.
  Volume 3, EPA-600/9-91-026c (NTIS PB91-234468), pp.
  4-17 through 4-30, July 1991.

How92
  Howell, T., Radon Reduction and Testing, Atlanta, GA,
  personal communication, February 5,1992.

Hu89
  Hubbard, L. M., B. Bolker, R. H. Socolow, D. Dickerhoff,
  and R. B. Mosley, "Radon  Dynamics in a  House Heated
  Alternately by Forced Air and by Electric Resistance," in
  Proceedings: The 1988 Symposium on Radon and Radon
  Reduction  Technology. Volume 1, EPA-600/9-89-006a
  (NTIS PB89-167480), pp. 6-1 through 6-14, March 1989.

BL90
  ULMASTI  Electronics,  Ltd., "IlmaRadon meter is the
  instrument for  measurements of  radon gas concentra-
  tions," Vantaa, Finland, 1990.
                                                    294

-------
Jo91
  Jones, D. L., Radon Reduction and Testing, Atlanta, GA,
  personal communication, March 1991.

Ka89
  Kaschak, W. M., D. G. Gilroy, R. H. Tracey, and D. B.
  Henschel, "Assessment of the Effectiveness  of
  Radon-Resistant Features in New House Construction,"
  in Proceedings: The 1988 Symposium on Radon and
  Radon Reduction  Technology.  Volume 2, EPA-600/
  9-89-006b (NTIS PB89-167498), pp. 4-91 through 4-103,
  March 1989.

KI89
  Kladder, D. L., Colorado Vintage Companies, Inc., Colo-
  rado Springs, CO, personal communication, June 1,1989.

K192
  Kladder, D. L., and S. R. Jelinek, Colorado Vintage
  Companies, Inc., Colorado Springs, CO, personal com-
  munication, March 30, 1992.

Kn90
  Kneeland, C. E., J. P. Reese, and M. R. Watson, "Diag-
  nostic and Mitigation Techniques Used in Radon Field
  Workshops," Paper No. 90-89.8,83rd Annual Meeting of
  the Air & Waste Management Association, Pittsburgh,
  PA, June 1990.

MaSS
  Matthews, T. G., C. S. Dudney, D. L. Wilson, R. J. Saultz,
  P. K. TerKonda, andR. B. Gammage, personal communi-
  cation, September 1988.

MaS9a
  Matthews, T. G., D. L. Wilson, P. K. TerKonda, R. J.
  Saultz, G. Goolsby, S. E. Burns, and J. W. Haas, "Radon
  Diagnostics: Subslab Communication and Permeability
  Measurements," inProceedings: The 1988 Symposium on
  Radon and  Radon  Reduction Technology. Volume 1,
  EPA-600/9,89-006a (NTIS PB89-167480),  pp.  6-45
  through 6-66, March 1989.

Ma89b
  Matthews, T. G., D. L. Wilson,  R. J. Saultz,  and C. S.
  Dudney, personal communication, May 1989.

Me92
  Menetrez, M. Y., U. S. Environmental Protection Agency,
  Research Triangle Park, NC, personal  communication,
  March 1992.

Mes90a
  Messing, M., "Testing of Indoor Radon Reduction Tech-
  niques in Basement Houses Having Adjoining Wings,"
  EPA-600/8-90-076  (NTIS  PB91-125831), November
  1990.

Mes90b
  Messing, M., Infiltec Radon Control, Inc., Falls Church,
  VA, personal communication, April 25-26,1990.
Mes90e
  Messing, M., Infiltec Radon Control, Inc., Falls Church,
  VA, personal communication regarding  follow-up
  alpha-track measurements in Maryland study houses, No-
  vember 21,1990.

Mes91
  Messing, M., Infiltec Radon Control, Inc., Falls Church,
  VA, personal communication, February 14,1991.

Mes92
  Messing, M., Infiltec Radon Control, Inc., Falls Church,
  VA, personal communication, March 31,1992.

Mi87
  Michaels, L. D., T. Brennan, A. Viner, A. Mattes,
  and W. Turner, "Development and Demonstration
  of Indoor Radon Reduction Measures for 10 Homes
  in Clinton, New Jersey," EPA-600/8-87-027 (NTIS
  PB87-215356), July 1987.

Na85
  Nazaroff, W. W., and S. M. Doyle, "Radon Entry into
  Houses Having a Crawl Space,"  Health Physics,
  48(3):265-281,1985.

Ne92
  Nelson, G., The Energy Conservatory, Minneapolis, MN,
  personal communication, July 30,1992.

NFGC88
  National Fuel Gas Code, "Recommended Procedures for
  Safety Inspection of an Existing Appliance Installation,"
  Appendix H, p. 2223.1-98,1988.

M85  .
  Nitschke, I. A., G. W. Traynor, J. B. Wadach,  M. E.
  Clarion, and W. A. Clarke, "Indoor Air Quality, Infiltra-
  tion and Ventilation in Residential Buildings," Report
  85-10, New York State Energy Research and Develop-
  ment Authority, Albany, NY, March 1985.

Ni89
  Nitschke, I. A., "Radon Reduction and Radon Re-
  sistant Construction Demonstrations in New York,"
  EPA-600/8-89-001 (NTIS  PB89-151476),  January
  1989.

NYSEO91
  New York State Energy Office, "Radon: A Diagnostic
  Field Guide for Professionals," Albany, NY, 1991.

Os89a
  Osborne, M C., "Resolving the Radon Problem in Clinton,
  New Jersey, Houses," Environment International,
  15:281-287, 1989.

Os89b
  Osborne, M. C., D. G. Moore, Jr., R. E. Southerlan, T. M.
  Brennan, and B. E. Pyle, "Radon Reduction in Crawl
  Space House," /. Environ. Engineering, 115(3):574-589,
  1989.
                                                    295

-------
Pe90
  Pelican Environmental Coip., "Pelican S-3 Blower and
  Pelican HPLF System," Framingham, MA, 1990.

Pr87
  Prill, R. J., B. H. Turk, W. J. Fisk, D. T. Grimsrud, B. A.
  Moed, and R. G. Sextro, "Radon and Remedial Action in
  Spokane River Valley Homes. Volume 2: Appendices to
  LBL-23430," LBL-24638, Lawrence  Berkeley Labora-
  tory, Berkeley, CA, December 1987.

Pr89
  Prill, R. J., W. J. Fisk, and B. H. Turk, "Monitoring and
  Evaluation of Radon Mitigation Systems Over aTwo-Year
  Period," in Proceedings: The 1988 Symposium on Radon
  and Radon Reduction Technology. Volume 1, EPA-600/
  9-89-006a (NTIS PB89-167480), pp. 7-93 through 7-109,
  March 1989.

Py90
  Pyle, B. E., and A.  D. Williamson, "Radon  Mitigation
  Studies: Nashville Demonstration," EPA-600/8-90-061
  (NTIS PB90-257791), July 1990.

Py91
  Pyle, B. E., and K. W. Leovic, "A Comparison of Radon
  Mitigation Options for Crawl-Space School Buildings,"
  in Proceedings: The 1991 International Symposium on
  Radon  and Radon Reduction  Technology. Volume 2,
  EPA-60Q/9-91-037b (NTIS PB92-115369),  pp. 10-73
  through 10-84, November 1991.

I»y92
  Pyle, B. E., Southern Research Institute, Birmingham,
  AL, personal communication, April 14,1992.

Ra92
  RadonAway, Inc., "Catalog: Fans and More  for Radon
  Professionals," Andover, MA, 1992.

Ro90
  Robertson, A., American ATCON, Inc., Toronto, Ontario,
  personal communication regarding April 1989-90 annual
  alpha-track detector measurements in radon  mitigation
  study houses in Ohio, September 1990.

Roc91
  Roessler, C. E., R. Morato, D. E. Hintenlang, and R. A.
  Furman, personal communication, 1991.

Ru91
  Ruppersberger,  J.  S., "The Use of Coatings  and Block
  Specification to Reduce Radon Inflow Through Block
  Basement Walls," in Proceedings: The 1990 Interna-
  tional Symposium  on Radon and Radon Reduction Tech-
  nology,  Volume  2,  EPA-600/9-91-026b (NTIS
  PB91-234450), pp. 8-51 through 8-59, July 1991.

Sa84
  Sachs, H. M., and T. L. Hernandez, "Residential Radon
  Control by Subslab  Ventilation," presented at the  77th
  Annual Meeting of the Air Pollution Control Association,
  San Francisco, CA, June 24-29,1984.

Sau89
  Saum, D. W., and  M. Messing, "Guaranteed  Radon
  Remediation Through Simplified Diagnostics," in Pro-
  ceedings of the Radon Diagnostics Workshop, April 13-14,
  1987,  EPA-600/9-89-057 (NTIS PB89-207898), June
  1989.

Sau91a
  Saum, D. W, and M. C. Osborne, "Radon Mitiga-
  tion Performance of Passive Stacks in Residential
  New Construction," in Proceedings: The 1990 Inter-
  national Symposium on Radon and Radon Reduction
  Technology.  Volume 2, EPA-600/9-91-026b  (NTIS
  PB91-234450), pp. 8-15 through 8-28, July 1991.

Sau91b
  Saum, D. W., "Mini Fan for SSD Radon Mitigation in
  New Construction,"  in Proceedings: The 1991 Interna-
  tional Symposium on Radon and Radon Reduction Tech-
  nology.  Volume  2, EPA-600/9-91-037b   (NTIS
  PB92-115369), pp. 8-45 through 8-55, November 1991.

Sau92
  Saum, D. W., Infiltec, Falls Church, VA, personal com-
  munication, May 20,1992.

Sc88
  Scott, A. G., A. Robertson, and W. O. Findlay, "Installa-
  tion and Testing of Indoor Radon Reduction Techniques
  in 40 Eastern Pennsylvania Houses," EPA-600/8-88-002
  (NTIS PB88-156617), January 1988.

Sc89
  Scott, A. G., and A. Robertson, "Follow-Up Alpha-Track
  Monitoring in 40 Eastern Pennsylvania Houses with In-
  door Radon  Reduction  Systems (Winter  1988-89),"
  EPA-600/8-89-083 (NTIS PB90-134172), October 1989.

Sc90a
  Scott, A. G., American ATCON, Inc., Wilmington, DE,
  personal communication, August 1990.

Sc90b
  Scott,  A. G.,  and A. Robertson, "Follow-Up Annual
  Alpha-Track  Monitoring in 40 Eastern  Pennsylvania
  Houses with Indoor Radon Reduction Systems (Decem-
  ber 1988-December 1989)," EPA-600/8-90-081  (NTIS
  PB91-127779), November 1990.

Sc92
  Scott, A. G., American ATCON, Inc., Wilmington, DE,
  personal communication, November 24, 1992.

Sh90
  Shearer,  D. J., Professional House Doctors, Inc.,  Des
  Moines, IA, personal communication, April 25,1990.
                                                   296

-------
Sh91
  Shearer, D. J., Professional House Doctors, Inc., Des
  Moines, IA, personal communication, February 15,1991.

Sh92
  Shearer, D. J., Professional House Doctors, Inc., Des
  Moines, IA, personal communication, February 6,1992.

SheSO
  Sherman, M. H., and D. T. Grimsrud, "Measurement of
  Infiltration Using Fan Pressurization and Weather Data,"
  in Proceedings of First Air Infiltration Centre Conference
  on Air Infiltration Instrumentation and Measuring Tech-
  niques, pp. 277-322, Air Infiltration Centre, Berkshire,
  UK, 1980.

She86
  Sherman, M. H., "Infiltration Degree Days: A Statistic for
  Quantifying Inffltation-Related Climate," AXffiJAE Trans-
  actions, 92 (Part 2A): 161-181, 1986.

Si91
  Simon, R. P., R. F. Simon Company, Inc., Barto, PA,
  personal communication, April 4,1991.

St90
  Staley, T. L., Radon Screening Service, Inc., Englewood,
  CO, personal communication, April 25,1990.

Str91
  Strom, D. J., W. D. Ulicny, J. B. Mallon, Jr., and R. W.
  Benchoff,  "A  Cost-Effectiveness  Comparison  of
  Private-Sector Radon Remediation with Traditional Ra-
  diation Protection Activities," in Proceedings: The 1990
  International Symposium on Radon and Radon Reduction
  Technology.  Volume 2,  EPA-600/9-91-026b  (NTIS
  PB91-234450), pp. 7-73  through 7-82, July 1991.

Ta85
  Tappan, J. T.? "Radon Mitigation Remedial Action Dem-
  onstration at the Watras Residence," report to Philadel-
  phia Electric Co. by Arix Corp., June 1985.

TEC87
  The Energy Conservatory, "Minneapolis Blower Door
  Operation Manual, Model 3," Minneapolis, MN, 1987.

TEC92
  The Energy Conservatory, "Combustion Safety Test Pro-
  cedure," Minneapolis, MN, 1992.

Tu87
  Turk, B. H., R. J. Prill, W. J. Fisk, D. T. Grimsrud, B. A.
  Moed, andRi  G. Sextro, "Radon and Remedial Action in
  Spokane River Valley Homes.  Volume 1: Experimental
  Design and Data Analysis," LBL-23430, Lawrence Ber-
  keley Laboratory, Berkeley, CA, December 1987.
TufiSa
  Turk, B. H., J. Harrison, R. G. Sextro, L. M. Hubbard, K.
  J. Gadsby, T. G. Matthews, C. S. Dudney, and D. C..
  Sanchez, "Evaluation of Radon Reduction Techniques in
  Fourteen Basement Houses: Preliminary Results," Paper
  No. 88-107.2, 81st Annual Meeting of the Air Pollution
  Control Association, Dallas* TX, 1988.

Tu88b
  Turk, B. H., J. Harrison, R. J. Prill, and R. G. Sextro,
  "Preliminary Diagnostic Procedures for Radon Control,"
  EPA-600/8-88-084 (NTIS PB88-225115), June 1988.

Tu89
  Turk, B. H., J. Harrison, and R. G. Sextro, "Performance
  of Radon Control Systems," LBL-27520, Lawrence Ber-
  keley Laboratory, Berkeley, CA, June 1989.

Tu90
  Turk, B. H., J. Harrison, R. J. Prill, and R. G. Sextro,
  "Developing Soil Gas and 222Rn Entry Potentials for Sub-
  structure Surfaces and Assessing ^Rn Control Diagnos-
  tic Techniques," Health Physics, 59:405-419,1990.

Tu91a
  Turk, B. H., J. Harrison, R. J. Prill, andR. G. Sextro, "Soil
  Gas and Radon Entry Potentials for  Substructure Sur-
  faces," in Proceedings: The 1990 International Sympo-
  sium on Radon and Radon Reduction Technology. Vol-
  ume  2, EPA-600/9-91-026b (NTIS PB91-234450), pp.
  5-33 through 5-51, July 1991.

Tu91b
  Turk, B. H., D. Grumm, Y. Li, S. D. Schery, and  D. B.
  Henschel, "Soil Gas  and Radon Entry  Potentials  for
  Slab-on-Grade Houses," in. Proceedings: The 1991 Inter-
  national Symposium on Radon and  Radon Reduction
  Technology.  Volume  1,  EPA-600/9-91-037a (NTIS
  PB92-115351), pp. 5-53 through 5-67, November 1991.

Tu91c
  Turk, B. H., Mountain West Technical Associates,  Santa
  Fe, NM, personal communication, April 3,1991.

Tu92
  Turk, B. H., Mountain West Technical Associates,  Santa
  Fe, NM, personal communication, March 31,1992.

Vi79
  Vivyurka, A., "Assessment of Subfloor Ventilation Sys-
  tems," presented at the Workshop on Radon and Radon
  Daughters in Urban Communities Associated with Ura-
  nium Mining and Processing, Bancroft, Ontario, March
  12-14,  1979.

We90
  West, D.,  Insul-Tech, Inc.,  Westerville, OH,  personal
  communication, April 25,1990.
                                                     297

-------
Wi90                                                 Wi92
  Wiggers, K. D., American Radon Services, Ltd., Ames,        Wiggers, K. D., American Radon Services, Ltd., Ames,
  IA, personal communication, October 1990.                  IA, personal communication, December 1992.

Wi91                                                 Zu92
  Wiggers, K. D., American Radon Services, Ltd., Ames,        Zucchino, A. P., RadonAway, Inc., Andover, MA, per-
  IA, personal communication, September 1991.                sonal communication, December 11,1992.
                                                  298

-------
                                         Section  15

                              Sources of Information

The first point of contact for information concerning indoor  If further information is desired, additional assistance and
radon and radon reduction measures should be the appropriate  contacts can be provided by the EPA Regional Office for the
state agency. Table 7 lists the appropriate agency to contact  region that includes your state. Table 8 lists the address and
for each of the states.                                  telephone number of the radiation staff for each of EPA's 10
                                                   regional offices. The table also includes the appropriate re-
                                                   gional office to contact for each state.
                                                 299

-------
Tablo 7.  Radon Contacts for Individual States
Alabama
  Radiological Health Branch
  Alabama Department of Public Health
  434 Monroe St, Room 510
  Montgomery, AL 36130-1701
  (205) 261-5315
  1 •800-582-1866 (In state)

Alaska
  Alaska Department of Health and Social Services
  Division of Public Health
  P, O. Box H
  Juneau.AK 99811-0610
  (907) 465-3019

Arizona
  Arizona Radiation Regulatory Agency
  4814 South 40th Street
  PhoonJx, AZ 85040                           	
  (602) 255-4845

Arkansas
  Division of Radiation Control and Emergency Management
  Arkansas Department of Health
  4815 Markham Street
  Uttte Rock, AR 72205-3867
  (501)661-2301

California
  California Department of Health Services
  601 North 7th Street
  Sacramento, CA 94234-7320
  (916) 324-2208

Colorado
  Radiation Control Division
  Colorado Department of Health
  4210 East 11th Avenue
  Denver. CO 80220
  (303)331-8480

Connecticut
  Radon Program
  Connecticut Department of Health Services
  150 Washington Street
  Hartford, CT 06106-4474
  (203)566-3122

Delaware
  Division of Public Health
  Delaware Bureau of Environmental Health
  P. O. Box 637
  Dover, DE 19901
  (302) 739-3787 or-3839
  1-800-554-4636 (in state)

District of Columbia
  DC Department of Consumer and Regulatory Affairs
  614 H Street, NW, Room 1014
  Washington, DC 20001
  (202) 727-7218

Florida
  Office of Radiation Control
  Florida Department of Health and Rehabilitative Services
  1317 Wine wood Boulevard
  Tallahassee, FL 32399-0700
  (904) 488-1525
  1-800 543-3279 (consumer inquiries only)
Georgia
  Georgia Department of Human Resources
  Environmental Protection Division
  878 Peachtree Street, Room 100
  Atlanta, GA 30309
  (404) 894-6644
  1-800-745-0037 (in state)

Guam
  Guam Environmental Protection Agency
  Hand E. Harmon Plaza D-107
  130 Rojas Street
  Harmon, Guam 96911
  (671)646-8863

Hawaii
  Environmental Protection and Health Services Division
  Hawaii Department of Health
  591 Ala Moana Boulevard
  Honolulu, HI 96813-2498
  (808) 586-4700

Idaho
  Bureau of Preventative Medicine
  Idaho Department of Health and Welfare
  450 West State Street
  Boise, ID 83720
  (208) 334-6584

Illinois
  Illinois Department of Nuclear Safety
  1301 Knotts Street
  Springfield,  IL 62703
  (217)786-7126
  1-800-325-1245 (in state)

Indiana
  Radiological Health Section
  Indiana State Board of Health
  1330 West Michigan Street
  P. O. Box 1964
  Indianapolis, JN 46206-1964
  (317)633-0150
  1-800-272-9723 (in state)

Iowa
  Bureau of Radiological Health
  Iowa Department of Public Health
  Lucas State Office Building
  Des Moines, IA 50319-0075
  (515)281-7781
  1-800-383-5992 (in state)

Kansas
  Radiation Control Program
  Environmental Health Services
  Kansas Department of Health and Environment
  6th floor, Mills Building
  109 SW 9th Street
  Topeka, KS66612
  (913) 296-1560

Kentucky
  Radiation Control Branch
  Division  of Community Safety
  Department of Health Services
  275 East Main Street
  Frankfort, KY 40621
  (502) 564-3700
                                                                                                                (continued)
                                                            300

-------
Table 7.  (Continued)
Louisiana
  Radiation Protection Division
  Louisiana Department of Environmental Quality
  P.O. Box 14690
  Baton Rouge, LA 70898-4690
  (504) 925-4518

Maine
  Division of Health Engineering
  Maine Department of Human Services
  State House, Station 10
  Augusta, ME 04333
  (207) 289-5692

Maryland
  Center for Radiological Health
  Maryland Department of the Environment
  2500 Broening Highway
  Baltimore, MD21224
  (301)631-3300
  1-800-872-3666 (in state)

Massachusetts
  Radiation Control Program
  Massachusetts Department of Public Health
  150 Tremont Street, 11 th floor
  Boston, MA 02111
  (617) 727-6214

Michigan
  Division of Radiological Health
  Bureau of Environmental and Occupational Health
  Michigan Department of Public Health
  3423 Logan/Martin Luther King, Jr. Blvd.
  P. O. Box 30195
  Lansing, Ml 48909
  (517) 335-8190

Minnesota
  Indoor Air Quality Unit
  Minnesota Department of Health
  925 Delaware Street, SE
  Minneapolis, MN 55459-0040
  (612)627-5012
  1-800-798-9050 (in state)

Mississippi
  Division of Radiological Health
  Mississippi Department of Health
  P.O. Box 1700
  Jackson, MS 39215-1700
  (601)354-6657
  1-800-626-7739 (in state)

Missouri
  Bureau of Radiological Health
  Missouri Department of Health
  1730E. Elm
  P. O. Box 570
  Jefferson City, MO 65102
  (314)751-6083
  1-800-669-7236 (in state)

Montana
  Occupational Health Bureau
  Montana Department of Health and Environmental Sciences
  Cogswell Building A113
  Helena, MT 59620
  (406) 444-3671
Nebraska
  Division of Radiological Health
  Nebraska Department of Health
  301 Centennial Mall, South
  P. O. Box 95007
  Lincoln, NE 68509
  (402)471-2168
  1-800-334-9491 (in state)

Nevada
  Radiological Health Section
  Nevada Department of Human Resources
  505 East King Street, Room 203
  Carson City, NV 89710
  (702) 687-5394

New Hampshire
  Bureau of Radiological Health
  New Hampshire Division of Public Health Services
  Health and Welfare Building
  Six Hazen Drive
  Concord, NH 03301-6527
  (603)271-4674

New Jersey
  Bureau of Environmental Radiation
  New Jersey Department of Environmental Protection and Energy
  CN-415
  Trenton, NJ 08625-0145
  (609) 987-6389
  1-800-648-0394 (in state)

New Mexico
  Radiation Licensing and Registration Section
  New Mexico Environmental Improvement Division
  1190 St. Francis Drive
  Santa Fe, NM 87503
  (505) 827-2948

New York
  Bureau of Environmental Radiation Protection
  New York State Health Department
  Two University Place
  Albany, NY 12202
  (518) 458-6461
  1-800-458-1158 (in state)

North Carolina
  Radiation Protection Division
  North Carolina Department of Environment, Health and Natural
  Resources
  P. O. Box 27687
  Raleigh, NC 27611-7687
  (919)571-4141

North Dakota
  Division of Environmental Engineering
  North Dakota Department of Health
  1200 Missouri Avenue, Room 304
  P. O. Box 5520
  Bismarck, ND 58502-5520
  (701)224-2348

Ohio
  Radiological Health Program
  Ohio Department of Health
  246 North High Street
  P.O. Box 118
  Columbus, OH 43266-0118
  (614) 644-2727
  1-800-523-4439 (in state)
                                                                                                                 (continued)
                                                            301

-------
Table 7.  (Continued)
Oklahoma
  Radiation Protection Division
  Oklahoma Stata Department of Health
  P.O. Box53551
  Oklahoma City, OK 73152
  {405)271-5221

Oregon
  Hsafth Division
  Oregon Department of Human Resources
  1400 S.W. 5th Avenue
  Portland, OR 97201
  (503) 229-5797

Pennsylvania
  Bureau of Radiation Protection
  Pennsylvania Department of Environmental Resources
  200 North Third Street
  P. O. Box 2063
  Ham'sburg, PA 17120
  (7f 7) 787-2163 or-2480
  1-800-237-2366 (in state)

Puerto Rico
  Puerto Rico Radiological Health Division
  Q.P.O. Cal!Box70184
  Rio Ptofdras, PR 00936
  (809) 767-3563

Rhode Island
  Division of Occupational Health and Radiation
  Rhode Island Department of Health
  205 Cannon Building
  Davis Street
  Providence, R1 02908
  (401) 277-2438

South Carolina
  Bureau of Radiological Health
  South Carolina Department of Health and Environmental Control
  2600 Bull Street
  Columbia, SC 29201
  (803) 734-4700 or-4631
  1-800-768-0362 (in state)

South Dakota
  Division of Environmental Regulation
  South Dakota Department of Water and Natural Resources
  Joe Foss Building, Room 217
  523 E. Capitol
  Pierre, SD 57501-3181
  (605)773-3153

Tennessea
  Division of Air Pollution Control
  Bureau of Environmental Health
  Tennessee Department of Environment and Conservation
  Customs House, 4th floor
  701 Broadway
  Nashville, TN 37243-1531
  (615)741-4634
  1-800-232-1139 (in state)
Texas
  Bureau of Radiation Control
  Texas Department of Health
  1100 West 49th Street
  Austin, TX 78756-3189
  (512)835-7000

Utah
  Bureau of Radiation Control
  Utah State Department of Health
  288 North. 1460 West
  P. O. Box 16690
  Salt Lake City, UT 84116-0690
  (801)538-6734

Vermont
  Occupational and Radiological Health Division
  Vermont Department of Health
  10 Baldwin Street
  Montpelier, VT 05602
  (802) 828-2886

Virginia
  Bureau of Radiological Health
  Department of Health
  1500 E. Main Street, Room 104A
  P. O. Box 2448, Main Street Station
  Richmond, VA 23218
  (804) 786-5932
  1-800-468-0138 (in state)

Washington
  Office of Radiation Protection
  Washington Department of Health
  AirDustrial Building 5, LE-13
  Olympia, WA 98504
  (206) 586-3303
  1-800-323-9727 (in state)

West Virginia
  Radiological Health Program
  Industrial Hygiene Division
  West Virginia Department of Health
  151 11th Avenue
  South Charleston, WV 25303
  (304) 348-3426 or -3427
  1-800-922-1255 (in state)

Wisconsin
  Division of Health
  Radiation Protection Unit
  Wisconsin Department of Health and Social Services
  P. O. Box 309
  Madison, Wl 53701-0309
  (608) 267-4795

Wyoming
  Environmental Health
  Wyoming Department of Health and Social Services
  Hathway Building, 4th Floor
  Cheyenne, WY 82002-0710
  (307) 777-6015
                                                            302

-------
Table 8.  Radiation Contacts for EPA Regional Offices

         Address and Telephone
     States in EPA Region
Region 1
  U. S. Environmental Protection Agency
  John F. Kennedy Federal Building
  Boston, MA 02203
  (617) 565-4502

Region 2
  2AWM:RAD
  U. S. Environmental Protection Agency
  Jacob K. Javits'Federal Building
  26 Federal Plaza
  New York, NY 10278
  (212)264-4110

Region 3
  3AT12
  U. S. Environmental Protection Agency
  841 Chestnut Building
  Philadelphia, PA 19107
  (215) 597-8320

Region 4
  U. S. Environmental Protection Agency
  345 Courtland Street, N.E.
  Atlanta, GA 30365
  (404) 347-3907

Region 5
  AT-18J
  U. S. Environmental Protection Agency
  77 West Jackson Blvd.
  Chicago, IL 60604-3590
  (312)886-6175

  Region 6
  6T-ET
  U. S. Environmental Protection Agency
  1445 Ross Avenue, Suite 1200
  Dallas, TX 75202-2733
  (214) 655-7223

Region 7
  U. S. Environmental Protection Agency
  726 Minnesota Avenue
  Kansas City, KS 66101
  (913)551-7020

Region 8
  8AT-RP
  U. S. Environmental Protection Agency
  999  18th Street, Suite 500
  Denver, CO 80202-2405
  (303) 293-1709

Region 9
  A-1-1
  U. S. Environmental Protection Agency
  75 Hawthorne Street
  San Francisco, CA 94105
  (415) 744-1045

Region 10
  AT-082
  U. S. Environmental Protection Agency
  1200 Sixth Avenue
  Seattle, WA 98101
  (206) 553-7299
Connecticut, Maine,
Massachusetts, New Hampshire,
Rhode Island, Vermont
New Jersey, New York, Puerto
Rico, Virgin Islands
Delaware, District of Columbia,
Maryland, Pennsylvania,
Virginia, West Virginia
Alabama, Florida, Georgia,
Kentucky, Mississippi, North
Carolina, South Carolina,
Tennessee
Illinois, Indiana, Michigan,
Minnesota, Ohio, Wisconsin
Arkansas, Louisiana, New
Mexico, Oklahoma, Texas
Iowa, Kansas, Missouri,
Nebraska
Colorado, Montana, North
Dakota, South Dakota, Utah,
Wyoming
American Samoa, Arizona,
California, Guam, Hawaii,
Nevada
Alaska, Idaho, Oregon,
Washington
                 Correspondence should be addressed to the EPA Radiation Program Manager at each address indicated.

                                                                                                                (continued)
                                                            303

-------
Table 8. (Continued)
Alabama
Alaska
Arizona
Arkansas
California
Colorado
Connecticut
Delaware
District of Columbia
Florida
Georgia
Hawaii
Idaho
Illinois
Indiana
Iowa
Kansas
Kentucky
Louisiana
Maine
Maryland
Massachusetts
Michigan
Minnesota
Mississippi

EPA Region
4
10
9
6
9
8
1
3
3
4
4
9
10
5
5
7
7
4
6
1
3
1
5
5
4


Missouri
Montana
Nebraska
Nevada
New Hampshire
New Jersey
New Mexico
New York
i
North Carolina
North Dakota
Ohio
Oklahoma
Oregon
Pennsylvania
Rhode Island
South Carolina
South Dakota
Tennessee
Texas
Utah
Vermont
Virginia
Washington
West Virginia
Wisconsin
Wyoming
EPA Region
7
8
7
9
1
2
6
2
4
8
5
6
10
3
1
, ^
8
4
6
8
1
3
10
3
5
8
U.S. GOVERNMENT PRINTING OFFICE: 1994 — 550  -001   /80385




                        304

-------