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
------- |