United States       Air and Energy Environmental    EPA/600/9-90/005e
           Environmental Protection    Research Laboratory       January 1990
           Agency         Research Triangle Park NC 27711

           Research and Development	_^_^^___________
c/EPA      The 1990 International
           Symposium on Radon
           and Radon Reduction
           Technology:
           Volume V. Preprints

           Session VIII: Radon
            Prevention in New
            Construction
           Session C-VIII: Radon
            Preventionin New
            Construction—POSTERS
           Session \X:  Radon in
            Schools and  Large
            Buildings
           February 19-23.1990
           Stouffer Waverly Hotel
           Atlanta, Georgia

-------
            Session VIII:
Radon Prevention in New Construction

-------
                                                       VIII -1
EVALUATION OF RADON RESISTANT NEW CONSTRUCTION TECHNIQUES

                    by :  Terry Brennan
                         Mike Clarkin
                         Camroden Associates, Inc.
                         Oriskany, NY 13424

                         Michael C. Osborne
                         USEPA/AEERL
                         Research Triangle Park, NC 27711

                         Bill Brodhead
                         WPB Enterprises
                         Riegelsville, PA 18077
ABSTRACT

     In  1989 a  project to  evaluate three  approaches  to  radon
resistant new construction was undertaken by the Radon Mitigation
Branch of the EPA Air and Energy Engineering Research Laboratory.
Test houses  were selected.  Indoor  radon was measured to evaluate
the effectiveness of foundation sealing and passive and active soil
depressurization at preventing radon entry into the houses. Tracer
gas methods were developed to estimate the fraction of air that was
being drawn  through the cracks and holes in the foundation by the
soil  depressurization  system.    Below  grade  leakage   area  was
estimated using tracer gas data.  It was  found  for the small number
of houses  in the study that  a  very small amount  of below grade
leakage can  still result in elevated indoor rtdon levels. Passive
soil depressurization  systems were found  to  perform better than
mechanical barriers alone, but did not keep radon  levels  as low as
active systems. Active soil depressurization systems were found to
perform very well.
KEY WORDS

     Radon, Active Sub-slab Depressurization, Passive Sub-slab
Depressurization, Barrier.
     This paper has been reviewed in accordance with the u. s,
     Environmental Protection Agency peer and administrative
     review policies and approved for presentation and
     publication.

-------
INTRODUCTION

     Growing concern about the health risks associated with indoor
radon, a radioactive  gas  found  in varying amounts  in  nearly all
houses, has  underscored  the need for  dependable  radon-resistant
residential construction  techniques.  In response to this public
health exposure the United States Environmental Protection Agency
(EPA) has developed and demonstrated a variety of methods that have
been used to reduce indoor radon  levels in  existing houses.  Many
of these methods are being included in the design and construction
of new houses.  In an effort to determine whether a house built with
radon resistant techniques would have had elevated radon levels if
the radon resistant techniques  had not  been  used  and to evaluate
the effectiveness of those techniques,  EPA sponsored the Evaluation
of Radon Resistant Construction Techniques  project.  This project
addressed active versus passive  sub-slab depressurization systems,
investigated  the  effectiveness of  using  the foundation  as  a
barrier, and assessed the  energy penalties associated with the use
of a sub-slab depressurization system. Because of the limited scope
of this paper, the energy penalty issues will not be covered.

     Two houses in Northern Virginia and two houses in Allentown,
Pennsylvania,  were selected  for this  project.  All houses  had
sub-slab depressurization systems  installed or  provided for during
construction.  Careful  attention was  paid  to  foundation detail.
Measurements  made  in these  houses  included continuous  radon
measurements with  the sub-slab  depressurization  system operating
under  various  conditions  and,  to  some extent,  monitoring  of
pressure differentials and airflows.

INVESTIGATION OF HOUSES ~ RADON RESISTANT TECHNIQUES USED

     The two Virginia houses, designated VA1 and  VA2,  are nearly
identical 2-story  frame  houses built  in a  subdivision.  The two
houses are approximately 300 ft apart. Each house was constructed
with radon resistant techniques  including a layer of DOT #2 or ASTM
#57  stone  pebbles placed  beneath the  slab  prior to  pouring,  a
single-point sub-slab depressurization system,  and perforated pipe
placed within  the stone  layer. Additional  techniques  used were
poured concrete foundation walls, sump holes capped and vented to
the  outdoors,  and caulked floor/wall  joints,   form   ties,  and
expansion joints.

       The two  Pennsylvania houses, designated PA1 and PA2, each
has a layer of DOT #2 or ASTM 157 stone pebbles beneath the slab,
a single-point  sub-slab  depressurization system,  poured concrete
foundation walls, capped sump holes, and caulked floor/wall joints
and form ties. Both houses have perforated sub-slab interior

-------
footing drains  connected to  the  sump holes.  In House  PA1,  the
sub-slab depressurization penetration is located approximately 18
in.*  away  from  the  footing  drain.  The  House  PA2  sub-slab
depressurization system  is connected  directly  to the sump hole.

RADON SOURCES

     Sub-slab radon grab samples were  taken at all houses. The two
Virginia houses averaged less than  100 pCi/L beneath the slab. The
two Pennsylvania houses averaged concentrations in excess of 1000
pCi/L beneath the slab.

ACTIVE VERSUS PASSIVE SUB-SLAB DEPRESSURIZATION SYSTEMS

     Radon measurements were taken  with the sub-slab system in the
active and passive modes at Houses VA1,  PA1,  and PA2. The results
of the radon measurements are illustrated  in  Figures 1,2, and 3,
respectively.  As can  be seen,  the radon concentrations in House
VA1 are quite  low.  This was to be  expected because  the sub-slab
radon grab  samples  indicated radon concentrations less than 100
pCi/L. For this reason  it was decided that this house and House VA2
were not appropriate for this portion of the  study.  In order to
determine  the  effectiveness  of  radon  resistant  construction
techniques, it was  felt that radon levels of  at least 700 pCi/L
(Br88) beneath the slab should be present.
(*) Readers more familiar with metric units may use
    the factors listed at the end of this paper.

-------
      C Sub-slab passive 1/20/89 • 1/31/89
      B Sub-slab active 1/31/89 - 2/3/89
      m Barrier only 2/24/89 -3/25/89
Figure  1.  House  VA1  basement radon  concentrations with  sub-slab
system  in  the passive and  active  nodes and with the  exhaust pipe
capped.  (Error bars show  1  standard Deviation)

     House  PA1  averaged  1.1  ± 0.2 pCi/L  in the active node fron
March 31 to April 5, 1989.  During this sane period,  the  sub-slab
depressurization   system   was  maintaining  an  average   pressure
difference between the basement and the sub-slab  air of -210 Pa. The
system was placed in the passive node on April 7, 1989, and allowed
to run  passively  until  April 13. During this period,  the  basement
radon  concentration  averaged  8.2  ±  2.4  pCi/L.   Unfortunately,
equipment  malfunctions  resulted  in  the  loss  of  the   pressure
differentials during  this period;  however, for  the  period between
April 5  and  7, the sub-slab system maintained an average  pressure
differential of -1 Pa in the passive mode.  While this should not be
considered the pressure differential that the system was developing
during  the  April  7 to  13 period,   it  does represent the  pressure
differential that the passive  system can develop. A  pressure

-------
differential of 1 Pa is very  small  and can be easily overcome.  In
fact the pressure differential  switched from negative to positive
during the April 5 to  7 monitoring  period.  On April 13,  1989, all
radon  monitoring  equipment  was   removed  from  the  field  for
calibration checks, and not placed  back in the field until Nay 5.
The  system  was again  placed  in the  passive mode  on May  5,  and
allowed to run  passively until May 11.  Basement radon concentrations
during this period averaged 9.9 ± 3.6 pCi/L. In order to determine
how  the active system  would perform under stress, the basement was
depressurized to an average of  -10  Pa during the period from June
6 to 27,  1989. During this  period basement  radon concentrations
averaged 1.5 i 0.6 pCi/L. This is not significantly different from
the non-stressed test.

     While it  is  obvious that active  soil  depressurization has a
dramatic impact on indoor radon concentrations the case is much less
clear for passive soil depressurization. During the two periods in
April and May that passive soil depressurization was monitored the
radon concentrations were 8 to 10 pCi/L. However in the barrier-only
testing in  May immediately following the  passive  test  the radon
concentration was 9.4 ± 2.8  pCi/L (see Figure 4).  This  implies that
the passive soil depressurization had  little effect  on  the basement
radon concentrations, but is not conclusive.  Later the barrier-only
method resulted in indoor radon levels of 20 ± 8  pCi/L.

-------
   20-

—  IB"

£  16-

S  14-

 i  12.

   10-

    e-

    6-

    4-

    2-

    0-
E2 Sub-slab active 3/31/89 • 4/5/89
ED Sub-slab passive 4/7/89 -4/13/89
D Sub-slab passive 5/5/89 -5/11/89
E2 Sub-slab active basement depressurized to 10 Pa 6/7/89 - 6/27/89
                    1 1, I  ,! | t
    Figure 2. House PA1 basement radon  concentrations with  sub-slab
    system in the active and passive nodes and active node with basement
    depressurized to 10 Pa.

         House  PA2  was  also subjected  to cycling  between active  and
    passive sub-slab depressurization systens. From February 6  to 8,
    1989   (admittedly  a  very   short  period),  the   basement   radon
    concentration averaged  0.4  +  0.3  pCi/L with  the system in  the
    active node.  From February 8 to 11,  the basement averaged 3.5 ± 3.4
    pCi/L with  the system in the passive  node. On March  3,  the  system
    was placed in the active node and run  until April 1. Basement radon
    concentrations over this period averaged 0.6 ± 0.7 pCi/L. The system
    was  again placed in  the passive node on  May  5 and run  passively
    until May 18. Basement radon concentrations averaged 1.5 ± 0.5 pCi/L
    over  this tine period.  After a series of  experiments  that involved
    the testing  of the  barrier in this house,  the system  was again run
    passively from July 24  to August  18. Basement radon concentrations
    over  this period averaged 2.23 ±  0.7 pCi/L.

-------
      Note that, in this house, passive soil  depressurization worked
 better  than barrier-only  in consecutive  tests in  May [1.5  ± 0.5
 pCi/L  passive and  5.2  ± 2  pCi/L  barrier-only  (see Figure'5)].
 Another   interesting   bit   of   data   for   the   passive   soil
 depressurization  in this house was that  it kept radon levels at 5.2
 ±5.1 pCi/L when a 16 sq in. hole  in the slab was  resulting in 36.7
 ±8.9 pci/L with barrier-only technique. This is rather encouraging
 for passive soil  depressurization in some cases.
15
n-
13-
12-
11 -
10-
 9-
 6-
 7-
 6-
 5-
 4-
 3
 2
 1
 0
n
   m Sub-slab active 2/6789 - 2/B/B9
   U Sub-slab passive 2/8/89 -2/11/89
   E Sub-slab active 3/3/89 -4/1/89
   ^ Sub-slab passive 5/5/89 -5/18/89
   D Sub-slab passive 7/24/89 - 8/18/89
                            \\x\\
Figure  3.  House PA2  basement radon  concentrations  with sub-slab
system  in the active and passive nodes.

-------
EFFECTIVENESS OF BARRIERS

     Although the radon source at House VA1 was considered too low
to include the house in this portion of the  study, measurements were
made to  see  how  effective the  barrier was.  The basement  radon
concentration was continuously monitored from February 24 to March
25, 1989. Basement radon  concentrations during this period averaged
1.3 ± 0.3 pCi/L.

     House PA1 was subjected  to  a  series of experiments  designed
to determine  both the effectiveness  of  a barrier and  the effect
various  size holes   in  the  barrier  have  on the  indoor  radon
concentration. As can be  seen  in  Figure 4,  even the best laid plans
of  the  most  conscientious  researchers  often  go  awry.  Radon
concentrations in  the basement varied greatly with  no  regard for
hole size even when radon source strengths were similar. One would
have expected that  increasing the  size of the  hole  would have
resulted in higher radon concentrations.

     It  is  hypothesized  that  the  leakage  area  already  present is
large enough  so that  increasing  the area and allowing more air to
be drawn in through the soil  is diluting the soil air and  lowering
the soil air concentration faster than the  production and transport
of radon can  replenish it.

-------
   60

   65-

   60-

 '  45-

   40-

   35-

   30-

8  25 H

O  20 H

I  ,,-

   10 -

    5-

    0
m Barrier only 5/12/89 -6/18/89
B Barrier with 144 sq in. hole 6/29/89 - 7/10/69
D Barrier with 10 sq In. hole 7/10/89 - 7/16/89
K Barrier with 1 sq in. hole 7/18/89 - 7/31/89
D Barrier only 7/31/89 - 8/17/89
   Figure  4.  House PA1 basement  radon concentrations during barrier-
   only tests.
        The effectiveness of the barrier in House PA2 was also tested.
   This house followed the idea that increasing the below-grade leakage
   area   should  result   in  an  increase   in  basement  air  radon
   concentrations.  As seen on Figure 5, a 16 sq in. hole in the barrier
   resulted  in increased  concentrations,  compared  to  the  intact
   barrier.  Radon  concentrations with the 16 sq in. hole  increased
   rapidly to  an average  of 36.7 ± 9 pCi/L.  From July 13 to  24,  the
   sub-slab  system was operated  in the  passive node with  the  same 16
   sq  in. hole in the barrier  as in  the previous period.  The  passive
   sub-slab  system maintained the basement radon  concentration at 5.2
   ± 5 pCi/L.  Because of this behavior it is believed that the amount
   of  below-grade leakage area is smaller than  the critical  size at
   which  soil  air  dilution   puts  a  cap  on  the  basement  air
   concentration.  This hypothesized phenomenon would be very  dependent
   on  the strength  of radon  sources and their geometry with respect to
   low airflow resistance pathways beneath the house  and  through the
   foundation.

-------
70
60-
50-
40-
30 -
10-
m Barrier only 6/18/89-5/26/89
E3 Barrier only 6/21/89-7/10/89
D Barrier with 16 sq In. hole 7/10/89 - 7/13/89
& Barrier with 16 sq in. hole.Sub-slab passive 7/13/89 - 7/24/89
Figure 5. House PA2 basement radon concentration during barrier-only
tests.

BELOW GRADE LEAKAGE AREA  (BGLA)

     The  BGLA was estimated  in three of the houses (VA1, VA2,  and
PA1).  The measurement was not made  in House PA2  because it  was
occupied  and access was  limited.  This  is  unfortunate  because  it
seems to  have had the best results  for the mechanical-barrier-only
technique.  The technique  used to estimate  the BGLA is as  follows.
The total airflow out of the  active sub-slab stack and the  fraction
10

-------
of that exhaust air coining from the basement was estimated using a
tracer gas technique (described in detail in a poster paper  (Cl 90).
The  air pressure  differential between the  sub-slab air  and  the
basement  air was  measured using a micromanometer  at several test
holes.  Because  of  the  tightness  of  the  foundation  and  the
surrounding  low-permeability  soil  it was found  that the layer of
stone pebbles beneath  the  slab acted like an air plenum. That is,
there  was very  little difference  in pressure  differences  (DPs)
measured  at  different locations (all  DPs were within  3%  of each
other  from  one  end of  the basement  to  the  other in  all  three
houses).  Given  this special case it is possible to calculate the
leakage area of a sharp-edged orifice (SaB9)  that would give the
same airflows at the same DPs as was measured in each house.  Figure
6 summarizes the measurements  and the calculated BGLAs.
House
ID
VA1
VA2
PA1
Average DP
(in. WC)
0.44
0.22
0.78
Total stack
Airflow (cfra)
78
24
16
Basement
Airflow(cfm)
34
11
3
BGLA
(sq in
3
1.4
0.2
     Notice that the BGLAs are a few square inches or less. In fact,
for one of the houses it is only 0.2  sq in.  As a check on the BGLA
soil gas concentrations, basement infiltration rates and the airflow
characteristics of  the BGLAs were  used to estimate basement radon
concentrations if the basements were depressurized by 1 to 2 Pa. The
results  for  Houses VA2  and PA1 compared well with  actual  radon
measurements made  in them  (an estimated  range of  20 to 30  pCi/L
compared to 9 to 20 pCi/L measured for PA1 and an estimated 1 to 2
pCi/L compared to <1 pCi/L measured for VA2). However, the estimated
concentration for VA1 was 20 to 30 pCi/L and the measured indoor
radon concentration in VA1 was 1.3  ±  0.3 pCi/L.  It  is hypothesized
that the BGLA in VA1 actually is close to the 3 sg  in. estimate: the
explanation for the much  lower measured indoor radon concentrations
than would be predicted is the result of outdoor air leaking under
the slab  at  the corner of  the  basement with a walkout  door.  Air
leaking in at this site would not pass through much soil and could
be diluting the air beneath the slab. The  exterior door was tightly
sealed.

     The  remarkable thing about these BGLA measurements  is  that
under some circumstances  it takes only a tiny leakage area to result
in elevated indoor radon. The implications are not encouraging for
depending on a mechanical barrier technique to control indoor radon.
11

-------
CONCLUSIONS

     The  conclusions   of   this  study  must  be   considered  as
representing  only a  few houses,  located in  specific soils  and
climates, and (except for  PA2)  unoccupied.  Because all  of  these
houses had tight foundations and relatively tight soils, they also
represent only one  of  four possibilities  (tight  soil  -  tight
foundation,   loose soil  -   loose  foundation,  tight  soil -  loose
foundation,  and loose soil - tight foundation).

     The  limited  sample  studied shows  that  making  mechanical
barriers that can prevent soil air entry nay be impractical with the
ordinary amount of quality control found in the construction of most
houses. This study includes an example of a house with a very tight
foundation (10 to 100 times tighter  than  the  tightest of building
shells) that  still has  elevated indoor radon  levels. Additionally
it was found that enlarging the  leakage area in this foundation had
no dramatic  impact on the indoor radon concentrations, implying that
tightening efforts had  little impact on the radon levels observed
in the barrier-only mode in this particular house.

     Passive  soil depressurization seemed to work  fairly well in
House PA2, but did not  seem to  work  at all  in PA1.  This result is
not unexpected  considering the wide  variety of  environmental and
site specific variables that impact on indoor radon concentrations.
The question that remains to be  answered is how often or under what
circumstances is passive soil depressurization a viable option? Both
houses meet a goal of radon levels as low as possible using active
soil depressurization.

     Active soil depressurization proved to be extremely effective
in all houses tested.  In PA1 indoor radon levels averaged 0.57 pCi/L
and never got above 3 pCi/L in  the active node,  and in PA2 indoor
radon averaged 1.4 pCi/L and was never above 2 pCi/L in the active
mode. Both houses spent a good deal of time below 1 pCi/L of radon.
At these low levels, uncertainty due  to measurement accuracy begins
to become an  important  source of error in hourly measurements.

METRIC CONVERSION FACTORS

     Readers  more familiar with metric units may use the following
to convert to that system:

     Nonmetrlc           Multiply by              Yield Metric

       ft'/min                0.00047                  m'/sec
       ft                     0.305                    m
       ft1                    0.028                    m'
       in.                    2.54                     cm
       in.'                    6.45                     cm'
       in. WC                0.249                    kPa
12

-------
                            REFERENCES

Br(88) Brerman,  T.  (Camroden Associates), Informal poll of EPA field
investigators at EPA contractors  meeting,  Research Triangle Park,
NC, 1988.

Cl(90) Clarklin, M. "Energy Penalties Associated with the Use of a
Sub-slab Depressurization system,11 Presented at 1990 International
Symposium on  Radon and  Radon  Reduction Technology,  Atlanta,  GA,
February 19-23,  1990.

Sa(89) Saum, D.  (Infiltec, Inc.), Personal communication, August 30,
1989.
13

-------
                                                        VI! 1-2
Radon Mitigation Performance of Passive Stacks In
Residential Hew Construction
                   By:    David  W.  Saum
                          Infiltec
                          Falls  Church,  VA 22041

                   and    Michael C.  Osborne
                          U.S. EPA, AEERL
                          Research  Triangle Park,  NC 27711
                             ABSTRACT

       Passive stacks have been proposed and installed as a radon
   resistant measure in new houses, but little quantitative data
   on their performance has been collected.  This study involved
   continuously monitoring several  houses  that were  recently
   built with radon resistant features including crack sealing,
   porous subslab aggregate, and a stubbed off pipe penetrating
   the slab for use in installing a radon  mitigation system.  For
   this project, the pipe  systems were  completed  so that they
   exited  from the  roof,   and  half  of  the  houses  had  radon
   mitigation  fans  installed  on  the  pipes.    Houses  were
   continuously monitored with the pipes sealed, then with the
   pipes open but no fans  operating,  and finally with the fans
   (if installed) operating. The results show significant radon
   mitigation performance by the passive stack systems in most
   cases,  and  excellent  mitigation by the  active  systems.
   Failures by the  passive stack systems  appear  to be  due  to
   basement depressurization by heating,  ventilation,  and air-
   conditioning (HVAC) duct leakage, poor  installation of subslab
   piping, and poor communication between multilevel  slabs.

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

-------
INTRODUCTION

    A passive stack (PS) radon mitigation system is a type of
subslab depressurization  (SSD)  system  where the  source  of
exhaust power  in the  stack  is buoyant  air rather  than  an
electric fan.   The buoyant force is generated  whenever the
air in the pipe  is  at  a higher  temperature than the outdoor
air.   Since  the major cause of radon entry into  houses  is
thought to be  pressure driven  flow of radon into  the house
that is primarily caused by temperature differentials, the PS
has been suggested as an inexpensive, low energy solution to
radon problems that automatically compensates for changes in
temperature.

    Several  problems  with  PS  radon mitigation  have  been
suggested: reverse pressures from wind, pressure losses under
the slab,  reverse pressures  during the  summer,  and reverse
pressures due to mechanical and  heating appliances.  Although
some mitigators report installing PS systems, there is little
quantitative data on their performance.   This  paper reports
on a study of  PS in 16 new houses  constructed  by Ryan Homes
in Maryland  and  Virginia.   These houses  had basement radon
levels  between  4  and 20  pCi/L  before  the  stacks  were
installed.  During construction, a pipe stub was imbedded in
the  slabs,   4  in.  (10  cm) of  clean  coarse aggregate  was
installed beneath the  slabs, and the floor-wall cracks were
sealed with polyurethane caulk.   In half of these houses the
passive stack  performance  was compared  to conventional fan
powered SSD performance.  This paper contains a discussion of
the  passive   stack theory,  a summary  of  results,  and
conclusions.
PASSIVE STACK THEORY

    The term stack  effect is commonly used  to describe the
pressures and flows that are generated when buoyant warm air
is  enclosed  in  a   building.     The  ASHRAE  Handbook  of
Fundamentals1  contains an extensive discussion  of building
stack effect.  Figure 1  is a schematic diagram of the stack
pressures that would be expected in a house when there is no
wind and the inside temperature is higher than the outdoors.
In order to understand the stack effect, two terms — column
pressure and neutral pressure level —• must be defined.

Column Pressure

    Column pressure  is the maximum differential pressure that
can be induced across any point  on the building shell by an
inside-outdoor temperature  difference.   The column pressure
is proportional  to  the building height  and the temperature
difference between inside  and outdoors.   For example,  a  45°F
(25°C)  temperature difference and  an 8  ft  (2.4  m) building
height will induce a  column pressure of 0.01 in.  (0.025 cm)

-------
we pressure  across  the building envelope.  In  Figure  1 the
column pressure is  the  sum  of  the top and bottom pressures.
Note  that  the  column  pressure   is  not  affected  by  the
airtightness of the building.

Neutral Pressure Level

    The neutral pressure  level (NPL)  is  the  imaginary line
around the enclosure where the differential pressure inside-
outdoors is  zero.   In  Figure 1 the top pressure is shown to
be equal in  magnitude  to  the bottom pressure.   This results
in a NPL that is half way up the side of the enclosure.  The
NPL  location  and   stack  pressure  distribution across  the
building envelope   are  determined  by  the building  airflow
resistance,  including  the  resistance  of both  the building
shell  and  any internal partitions.   In houses,  there are
generally large openings (doors) or leaks between floors, and
the  internal  airflow   resistance   is  assumed  to  be  small
relative to  the shell  airflow resistance.   Therefore, the
shell  openings determine the  location of the NPL  and the
pressures across the top and bottom of the shell.  Note that
the NPL and the pressures on the enclosure are determined by
the distribution of leaks on the  enclosure surface,  and not
by the  overall airtightness or leakiness  of  the enclosure.
Leakier enclosures  will require  more  heat to  maintain the
inside-outdoor temperature difference, but the column pressure
will be independent of  the airtightness.  If the majority of
leaks are near -the  top of the enclosure, then the NPL will be
near the top, but the maximum pressures will be  at the bottom.
When top and bottom openings are  equal in size,  the NPL is
midway up the side  of the enclosure.

Effects of Sealing

    Sealing leaks in the enclosure will not change the column
pressure, but  the NPL  will  be  shifted if leaks near the top
or bottom of the enclosure are  sealed.  Since radon entry may
be proportional to  the  depressurization of the slab, sealing
the upper part of the shell should be beneficial since it will
reduce this depressurization, but  sealing leaks in the lower
part of the shell should be detrimental since it will increase
the depressurization of the lower  part of the enclosure.  If
a house could  be sealed as  completely as a hot air balloon,
with all  remaining leaks concentrated at the  bottom, then
there would be no stack induced depressurization on the slab
that would pull radon  into  the house.    Appendix A contains
a calculation which relates shell leakage distribution to
pressure on the lower surface of the enclosure.

    Figure  2 shows  the  changes in pressure on the upper and
lower  surfaces of   an  enclosure under a  variety  of leakage
configurations.    Figures  2. A and 2.B  have   already been
discussed,  Figure 2.C shows the effect of sealing half the

-------
leaks in  the bottom,  while Figure  2.D  shows the effect  of
sealing half  the  leaks in the top.   As a general  rule  for
radon control,  air sealing should  be  limited to the  upper
surfaces of the house  in  order to  minimize depressurization
of the slab.   Since a PS system does not  have  much suction
power, the reduction of stack effect in the  building may  be
necessary to maximize its performance.

Passive Stack System

    A PS  radon  mitigation system is shown  schematically  in
Figure 3,  and its performance is determined by the interaction
of  two  stacks.   The house  acts as a  stack which creates
depressurization above the slab,  and the PS pipe generates a
counteracting depressurization below the slab.  For the PS to
operate  successfully,  the  PS  must  induce  slightly  more
depressurization  beneath  the slab than  the house induces
above.  For best  performance of  the PS  five conditions must
be met.

Raise the column pressure of the PS

    The  PS  depressurization beneath  the  slab  should  be
maximized by keeping the temperature of the stack as high as
possible  relative  to  the  outdoor air  temperature.   Most
important, the PS  should  not be  run though unheated spaces,
such as garages.

Raise the NPL of the stack

    The flow though the stack should be minimized by sealing
the slab to keep the NPL of the PS as high as possible which
will increase the depressurization  below the  slab.   A PS that
is open at both ends  will have a NPL  at its midpoint,  and a
PS that is connected to an airtight subslab cavity will have
a NPL at its top.

Lower the NPL of the house

    The house stack depressurization above  the slab  should be
minimized by  air-sealing  of the  house  envelope  to lower the
NPL of  the house  as  much as possible.  Air-sealing of the
upper surfaces  of the house will  lower the house  NPL,  but
sealing the  lower part of the house will  raise the NPL and
increase the stack depressurization above the slab.  Lowering
the  temperature  of the  house  would  decrease  house  stack
depressurization,  but it is not  generally practical,  and
would also result in  a lower stack temperature.  Note that
if the house NPL is below the stack NPL, then the PS will not
provide radon mitigation.

-------
Maximize air-flow communication under the slab

    If there is any airflow  into  the  stack,  there will be a
pressure  drop  in  the   aggregate   that  will  reduce  the
depressurization of some areas of the subslab.  Good airflow
communication  by  using  clean coarse  aggregate  and  laying
perforated pipe under the slab may minimize this problem.

Minimize anv mechanical source of house deoressurization

    Any negative  pressure in  the house  will add  to house
depressurization above the  slab.   Examples  of this problem
include  imbalances  in  forced  air  distribution  systems,
fireplaces, exhaust fans  such  as  dryers, and combustion air
exhausted from the house by  fossil fuel heating systems.

SUMMARY OF RESULTS

An Example of Passive System Performance

    Figures 4 and 5 show the radon mitigation  performance of
a typical  house (#MIL) in which  a PS was  installed.   This
house  was built  in 1987  and has  1600  sq ft  (149  m2)  of
finished space with 800 sq ft  (74 m2) of unfinished basement
area.  The basement has poured concrete walls, is about 4 ft
(1.2 m) below ground,  and has a sump  connected to  an external
footing drain.  The passive stack goes up about 12 ft  (3.7 m)
through a chase inside the house  and  exits about 2 ft  (0.6 m)
below the roof line.  During construction, a perforated pipe
was laid in the 4  in. (10 cm) deep gravel bed beneath the slab
and  was  run  diagonally  across  the  basement  to  provide
communication.  One end  of  the perforated pipe is connected
to the sealed sump, and a T fitting was used to bring  a 4 in.
(10 cm) diameter  pipe  stub  up through the slab for possible
connection to a future  radon  control system.  All visible slab
cracks were sealed with polyurethane caulk.   When the passive
stack was installed in late 1988, a long twisted run of 4 in.
(10 cm)  pipe was necessary  to connect the pipe stub at one
side of the basement to the chase located in the center.  The
heating system is a heat pump with a fan system located in the
basement,  and  the distribution ductwork is contained within
the shell  of the  building.   Blower door measurements showed
about 5 air changes per hour (ACH) at 20 in.  we (50 Pa) which
is relatively  airtight for the Maryland  climate.

Winter performance

    Figure  4 shows  the  winter  performance  of  the  passive
stack.     Although  the  radon  mitigation   performance   is
significant  (85%),  the  radon spikes  suggested  that some
unknown depressurization source in the house was occasionally
overcoming the  PS  system.  When  the  basement was  reexamined,
it was found that the homeowner had sealed off all the supply
ducts  in  the unfinished  basement to save energy, but he had

-------
neglected to seal a large return vent. Therefore any tine the
basement door was closed,  the  basement  was depressurized by
about 0.030 in.  (0.076 cm)  vc.  This counterpressure overcame
the passive stack system and brought in radon but only when
the basement door was closed.

Summer performance

    Figure 5 shows the summer performance in house #MIL after
the air  return  vent in the basement was sealed  in order to
prevent  basement  depressurization  by  the heat  pump  fan.
Although the  summer radon levels were  much  lower  than the
winter levels in this house, the PS radon reduction was still
substantial (70%).  There is no sign of the spikes that were
seen  in  the winter,  and the  pressures measured  under the
basement door were less than 0.004 in. (0.01 cm)  we.  During
the period shown  in Figure 5,  the house was air conditioned,
and  there  were   several  periods   when  the  outdoor  air
temperature  was  higher  than   indoors.   These  data  do not
indicate that passive  system performance decreases when the
temperature differentials are reversed.

PS performance conclusions

    In several  of the PS  houses,  the stack  provided radon
mitigation well below 1 pCi/L during  all  seasons of the year.
Performance did not seem to be  affected by pipe straightness,
position of the vent on the roof, or wind conditions.  Houses
with poorer mitigation performance seemed  to  have four main
problems: 1) depressurization of the basement because of large
leaks in the return ducts in the basement, 2) multilevel slabs
that were not connected to  the  PS pipe system, 3) stack pipes
run through unheated spaces, and 4) improperly  connected stack
pipes that were blocked by  debris during  construction.   Even
in houses where there  was  a combination of problems,  the PS
generally gave mitigation performance of at least 50%.

Passive Versus Active System Performance

    In half the  houses, SSD systems with fans were installed:
these houses were studied  to compare the performance of the
PSs with the active systems.  The stack was  sealed for several
days to approximate premitigation conditions, the stack was
opened without fan operation to approximate a PS system, and
finally the fan was turned  on to demonstrate fan assisted SSD
performance.   The  resistance  of the  fan  in the  pipe was
assumed to  be  small enough to ignore because of  the small
airflows in the pipes  under passive  conditions.   House #TIN
had some of the  highest radon levels in this study,  but it was
otherwise quite typical of  the  active houses.  This house had
2000 sq ft (186 m )  of  finished floor area on  two floors, and
an unfinished 800 sq ft (74 m2) walk-out basement below.  The
basement has poured concrete walls with the  same crack sealing

-------
and  stubbed off pipe described in the passive house.  There
was  no  sump and  the pipe  stub  was connected  only to  a
perforated pipe in the aggregate beneath the slab.  The blower
door measurement showed that this house was about  4 ACH at 20
in.  we  (50  Pa) which is quite tight.

Winter performance

    When the  fan was on,  the radon levels were well below l
pCi/L, but  they quickly rose to about 30 pCi/L when the fan
was  turned  off and  the pipe was sealed.   When the passive
stack was simulated  by opening the stack without turning on
the  fan, the radon levels were significantly depressed,  but
there  were   occasional  large  spikes  that  could  not  be
explained.   Even so, the radon reduction  due  to  the PS  was
about 75% (Figure 6).

Summer performance

    The house was retested during warm weather  in September:
the  radon levels during the  PS test were lower than for the
comparable  winter  levels.   Since the measurements  did  not
include  a test  period when the stack  was  sealed,  it is  not
possible to determine the absolute mitigation performance.
There were some spikes in the radon levels  but  not as many as
the winter data showed. After this monitoring was concluded,
a  reexamination of the heat pump fan system  showed that a
construction  defect had left a large hole in the return duct
in the  basement.   The hole  generated a  depressurization of
0.050 in. (0.13  cm) we in the basement every time the  fan came
on and the  basement  door  was closed.   The September PS data
may  be  low  for  this house because  the mild weather did  not
require much  heat pump operation  (Figure 7).

Active versus passive

    The  conventional active SSD  system  is a  very reliable
solution to radon problems in new  construction houses because
it will overcome most of the inadequacies of PSs highlighted
in this  study.   Even the severe  depressurization  that  was
found in this house was negligible compared to the pressures
of about 0.75 in.  (1.9 cm)  we that  are commonly generated
under the slab  by  most  SSD fans.  The  only problem in  the
test houses that this excess fan power could not overcome is
the  lack of communication between multilevel slabs that was
seen in several houses.

CONCLUSIONS

    This  study  suggests  that  PS  systems  in  new  house
construction  can provide radon mitigation that is comparable
to  the  performance  of   active  systems  if  there  are  no
interfering sources of house depressurization.  However,  the
study also suggests that forced air heating  systems are major

-------
 sources of house depressurization due to duct leakage.   Since
 PS  systems  are very sensitive to pressure imbalances  in  the
 house,  they  provide  a  sensitive  tool  for  studying  the
 interaction of  the  other  systems  in  the  house.

 Duct Leakage

    The pressure imbalances in houses due to  duct leakage  and
 flow  imbalance were  recently studied in  Florida houses by
 Cummings'. That study indicated that Florida  houses typically
 have heat pumps in  the attic  outside the house  envelope,  and
 that duct leakage can cause depressurization or pressurization
 of  the entire house.   Since the houses  in this  study  had no
 ductwork outside the conditioned space, the leakage could only
 cause room to room variations in pressure rather than changes
 in  whole  house  pressure.  However, when  the  HVAC fans  are in
 the basement, as they  were  for  all of the houses  in this
 study,  the  probability  of  significant  pressures  in  the
 basement  is quite high.   All  of the  houses in this study  had
 pressurization  of the upstairs bedrooms when the  doors were
 shut  because  of the  lack  of returns  in  the  rooms.   This
 problem   does   not  seem  to   produce  significant basement
 depressurization.    It   seems to   be   more of  an   energy
 conservation problem than a contributor to the radon problem.
 Cummings2 reports a  similar problem in Florida.

 Adequate Ventilation

    The houses in this study were found to be  almost airtight.
 If  they did not have forced air heating  and  cooling systems,
 and the  pressure  or  leakage  problems previously  discussed,
 then they might be  under-ventilated  when compared to  ASHRAE
 recommendations1.    If  the duct leakage  and  imbalances were
 corrected  for  reasons   of   energy  conservation  or   radon
 mitigation, then the ventilation impact should be considered.

 Limitations and Suggestions for Further  Work

    This study was  very limited in the housing stock studied
 since it  only dealt with heat pump  houses within a limited
 area.   Future  studies  might  look  at Florida  or Minnesota
 houses with their different climates and HVAC systems.   All
 of the stacks  in this study were 4 in. (10 cm) schedule  40  PVC
 pipe.  It would be useful to  study 3 in. (7.6 cm)  pipe  since
 it  would be simpler and cheaper to run smaller pipe through
 the house.

Recommendations to Builders

    Passive  stacks  appear to  be the  most  effective passive
radon mitigation technique for new construction.  This  study
suggests that the stack should be run through the warm part

-------
of the house,  excellent subslab communication can be provided
with 4 in.  (10 cm)  of  clean coarse aggregate,  and the stack
pipe should be run up to the roof line.  Additional guidance
should include avoidance of duct leakage that depressurizes
the  basement  and connecting  stack pipes  to  each  level  of
multilevel  slabs.   Things  that can be  ignored include wind
caps for the stack,  multiple bends in the pipe, and failures
due to cooling situations.
REFERENCES

1.  ASHRAE  Handbook of Fundamentals, Chapter 23 Infiltration
    and Ventilation,  Section 23.2 Driving Mechanisms, American
    Society of Heating,  Refrigeration,  and Air-Conditioning
    Engineers, Inc., Atlanta,  Georgia,  1989.

2.  Personal communication with  J. B. Cummings, Florida Solar
    Energy  Center,   300   State   Road  401,  Cape  Canaveral,
    Florida, Sept.  13,  1989.

3.  ASHRAE Std. 62-1989, Ventilation for Acceptable Indoor Air
    Quality,  American Society of Heating, Refrigeration, and
    Air-Conditioning Engineers,  Inc., Atlanta, Georgia, 1989.

-------
                          APPENDIX A

                         CALCULATIONS

    Column  pressure  Peol  can be computed  from barometric
pressure  Pb, enclosure  height H, inside  temperature Tin,  and
outdoor temperature T^:
          Peol = 0.52PbH(l/Tout - l/Tln)                   (1)

where            H  = height of building in  ft
          Pb = ambient (barometric) pressure in psia (14.7 at
sea level)
          T     =  Rankine  temperature  (459  +   Fahrenheit
temperature)

    The  stack  pressures  can  be computed  in the  case of an
enclosure with  sharp edged holes  in  the top and bottom  with
areas Ahi and A10.   The equation for air flow Q  under standard
conditions through a sharp edged hole is given by:

          Q  = 16.9AP0'5                                  (2)

where     Q  = flow in cfm
          A  = Area in sq in.
          P  = Pressure in in. we

If air  flows through  an enclosure because  of  stack effect,
then flow rate  in Q10 must equal flow rate out  QM:

          Qio =  Qhi                                      (3)

or        16.9A10P10°-5 = 16.9AhiPhl°-5                     (4)

Rearranging, Plo = Phl (Ahl/A10) 2                           (5)

But column pressure Pcol  is the sum of the high Phi  and low P10
pressures:

          Peol - Phi + Plo                                (6)

Combining 5  with  6:

          PIO =  P.oi/[(Alo/Ahi)2 + 1]                      (7)

Note that  lower  pressure P10  is determined by the  ratio of
upper Abi and lower A10 leaks and not  by total leakage A.
                              10

-------
                  Neutral Pressure Level
                           t
      Assume no wind and  warmer  indoors  than outdoors
          Arrow  length represents pressure  magnitude


FIGURE 1 Schematic Diagram of Stack Effect Pressures in a House

          A ^^        B
1
NuM frm*
i
n M (IH)
•^ X

t
t
4
m.
A
1
•^ x-


t
1
m.
4



      Arrow  length  represents pressure  magnitude
   Assume  houses  are warmer indoors  than outdoors
  FIGURE 2 Stack Pressures for Several Leakage Configurations
                         11

-------
        Outdoor
        colder air
                           Stack neutral pressure level
                           House neutral pressure level
                     Indoor
                     warmer air
            Lower house depresssurization
                                                   I Stack
         Underground
                         Stack induced depressur'aaVion
FIGURE 3 Schematic Diagram of Passive Stack Radon Mitigation System
Stack inside  house
Sump,  deep basement
                                       Passive stack =  2.9  pCi/L
                                       stack  sealed «  19-9  P?i/L
            - Stack open
                                       Stack  sealed
          350           360
     Thursday 12/15/88


 — Basement Radon Levels
                          370           380

                   1988 Julian  Day
          Figure 4 Winter Performance of Passive Stack System
                                 12

-------
^^x




O
 CL
 O
•O
 (0
20






 15






 10






  5
            Stack inside  house

            Sump,  deep  basement
Passive stack = 0.7  pCi/L

 Stack sealed = 2.4 pCi/L
            - Stack open
                             Stack sealed  I    Stack open
          240        250


     Monday 8/28/89





  — Basement Radon Levels
                        260       270



                      1989 Julian Day
               280
290
          FIGURE 5 Summer Performance of Passive Stack System
b
o.
o
•o
(0
ex.
70



60



50



40



30



20



10
            Stack inside house

            Walkout basement, no sui
 Passive stack = 7.5 pCi/L

 StaclJ sealed = 29.0 pCiA

       Fan on =  0.2 pCi/L



       Fan off, stack open
          80           90


     Tuesday 3/21/89





 — Basement  Radon Levels
                             100           110



                      1989  Julian  Day
                          120
     FIGURE 6 Winter Performance of Active and Passive Stack Systems
                                  13

-------
       30
.i     20
O
Q.
I
        10
           Stack inside  house

           Walkout  basement, nb sump
                                Fan off, passive = 3.7 pCi/L

                                    SSO fan on = 0.3 pCi/L
                                   SSO fan turned on
         250       260

    Thursday 9/7/89




 — Basement  Radon Levels
                             270        280


                            1989  Julian  Day
290       300
     FIGURE 7 Summer Performance of Active and Passive Stack Systems

-------
                                                                        VIII-3
                   SUB-SLAB  PRESSURE FIELD EXTENSION STUDIES
              ON FOUR TEST SLABS TYPICAL OF FLORIDA CONSTRUCTION

                  by:   Richard A. Furman
                        School of Building Construction
                        University of Florida
                        Gainesville,  FL  32611

                        David E. Hintenlang, Ph.D.
                        Department of Nuclear Engineering Sciences
                        University of Florida
                        Gainesville,  FL  32611
                                   ABSTRACT
      The State  of Florida is currently  in the process of  developing,  under
legislative mandate, a radon resistant building  code for new construction.  One
of the research projects funded by the State is to examine the influences that
various  construction   practices   have  on  the  effectiveness   of  sub-slab
depressurization systems.   Four test  slabs  have been  constructed and pressure
field extension studies have commenced. The influences of several construction
practices and techniques including stemwall curtains,  types  of sub-slab fill,
fill depth and sub-slab plumbing influences will be investigated.  Climatic and
other environmental conditions will be monitored to determine their influence
on the sub-slab depressurization systems.

      This paper will discuss which typical Florida construction practices have
the  most  significant  impact on  the effectiveness  of  different  sub-slab
depressurization systems.


                                 INTRODUCTION
      Indoor radon has been identified as a problem in Florida for more than a
decade.  Early research in the phosphate mining areas of central Florida reported
that substantially elevated levels of radium and radon were present in reclaimed
mining  lands.    Not until  the mid  to  late  1980's  did  indoor radon  become
acknowledged  as  a legitimate  health  hazard,   thereby  prompting  nationwide
political action.  In 1988 the Florida Legislature passed legislation mandating
the development of a radon resistant building code for new construction.  As a
result of that legislation the State University System was  tasked with developing

-------
a draft building code for delivery to the Florida Department of Community Affairs
by  February 1, 1990.    The  Florida  Board of  Regents  established  the  State
University System Radon Advisory Board (SUSRAB) to coordinate the research and
development activities relating to the production of the draft code.

      Early  on,  the SUSRAB  recognized the  proven effectiveness  of sub-slab
depressurization  systems  and  funded  several  studies  to  investigate  the
operational characteristics of these  systems.  Regional construction techniques
and procedures play a critical role in the  success or failure of these systems.
Differences between the construction features of northern homes with basements
and  the  Sunbelt-type   construction   have  resulted   in  varying  levels  of
effectiveness of the sub-slab depressurization technique.  In order to study the
effects  of  various construction  techniques  on the sub-slab  depressurization
system, four test slabs were constructed and tested.

      The study, from which this paper was  developed, was  funded by the Florida
Board  of Regents  to determine  what  construction processes have  the greatest
impact on the sub-slab pressure field.  From  this work guidance was provided to
the SUSRAB in the development of the draft radon resistant building code.


                         FLORIDA TYPICAL CONSTRUCTION

      An informal survey of homebuilders conducted by the authors during the 1988
Summer Board Meeting of the  Florida  Home  Builders  Association found that over
95X of all residential construction built in Florida is  constructed  with concrete
slabs-on-grade.  In the colder climates of the more northern states where there
is less  rainfall, sub-grade  construction  of  basements  is common.   In Florida,
however, sub-grade construction is a rarity.  This result was not unexpected but
confirmed that the emphasis of radon  control  efforts should be focused on slab-
on- grade  construction.   Further analysis  of the survey  data  indicated  that a
geographical distribution of the type of slab-on-grade system used was evident.
Monolithic concrete slabs were  found  to be the  slab system of choice in those
areas  south  of Orlando; slabs constructed on masonry  stemwalls or foundation
walls were  more prevalent in those geographical areas north  of Orlando.   The
preference for slab systems with stemwalls north of Orlando is due to increased
topographic relief.

      Florida's sub-tropical climate and high precipitation levels  significantly
influence  the  construction  techniques used.   The most  significant regional
construction difference is the type of media used under the slab.   Gravels, used
primarily in northern U.S. sub-grade construction to facilitate water removal,
have substantially  higher air permeabilities than the sand fills used  in the
Sunbelt  region.   The  permeability of these base  materials coupled with the
effective leakage of the confining construction are of major importance in the
effectiveness  of  the  sub-slab depressurization system.   However, in slab-on-
grade construction  sub-grade water control is  not normally required resulting
in the extremely infrequent use of gravels as a slab base material.  Gravel or
stone  aggregate  is a  non-standard  building  component  in  typical  Florida
construction.

-------
      To  facilitate the  construction of  concrete  slabs on  grade,  Florida's
abundant supplies of clean construction-grade sands are used as compactable fill
material.  Fill-sands are frequently relocated from one area of the project site
to another or  may  be hauled in from distant borrow pits.  Some of these sands
may  possess  moderate  to  high radium  content  thereby  raising  a  concern  of
transporting radon contamination from one  locale to another.

SLAB CONSTRUCTION  CATEGORIES

      Concrete floor slabs  constructed  on  grade in Florida  typically represent
one  of the following  three general  slab  construction categories:  monolithic
slabs,  floating slabs  and  slab-in-stemwall  slabs (Figure 1).
       A.   Monolithic
                                                    C.   Slab-in-stemwall
      B.  Floating

Figure 1. Typical Florida slab construction techniques and soil gas entry routes.
Monolithic Slabs

      Slabs constructed on near-level  ground in areas not subject to flooding
are constructed most  cost effectively  as  monolithic slabs (Figure 1-A).  When
site modifications to create a level building platform are more cost effective

-------
in areas of irregular terrain.  Monolithic slabs are characterized as having the
footing and the slab cast as one integral unit.  Foundation depth is minimized
and radon entry routes  through the  slab are usually confined to cracks (planned
or unplanned) and mechanical system penetrations for plumbing, electrical, etc.
In  general,  fewer foundation  entry  conditions  are present  in  monolithic
construction than in the two other categories.

Floating Slabs

      Where terrain features, water considerations and/or other conditions demand
the use of foundation walls, which  elevate the  slab some distance above natural
grade, floating slabs are one of two construction techniques commonly employed.
Floating slabs  are cast against,  not into,  the foundation wall  (Figure 1-B).
Expansion joint materials are normally used to separate the slab from the inside
face of  the foundation wall forming a continuous radon entry  route  along the
perimeter of the slab.   In addition to the typical entry conditions associated
with slab cracking and penetrations,  foundation wall  conduction  of radon from
below  the   floor  slab  into  the superstructure walls is also  common.    Most
superstructure walls  erected on  floating slabs are  constructed of masonry block.
The masonry block wall's  thickness  is sufficient  for  the baseboard to conceal
the perimeter  crack; the crack is virtually  inaccessible  after  construction.
The continuous perimeter crack and the foundation/superstructure wall conduction
are the most significant  radon  entry  routes  associated with  this  slab system.

Slab-in-Stemwall Slabs

      Many  contractors  prefer frame  superstructure  walls to masonry  and have
eliminated  the  perimeter crack associated with floating slabs by  adopting a
system of casting the slab into the masonry foundation wall (Figure 1-C).  Two
types of masonry block may  be used to form the edge  of the  slab.   The  lintel
block can be used  so only the outer  face  shell remains as  the  slab form.  The
header block, however,  when used retains part of the web partitions as well as
the outer face shell.  This condition is important because,  depending upon the
method utilized to prevent concrete from being lost down the block cores, entry
routes from the foundation wall into the superstructure  may result.   If the
contractor is careful,  the foundation wall superstructure conduction problem can
be eliminated.  However, contractors have been observed draping the vapor barrier
over the header block.   After the slab was cast, large  holes were found to exist
where the concrete had  been held away from the  outer face shell of the block by
the vapor barrier. These holes result in foundation/superstructure conduction.

                              TEST SLAB PROGRAM
      The U.S. Environmental Protection Agency and others have conducted numerous
studies in the eastern U.S. over the past several years and have found that sub-
slab depressurization systems  appear to be  more  consistently successful than
other experimental radon mitigation  systems.  Most  of the early  tests of sub-
slab depressurization systems were conducted on basement structures using stone
aggregate as the slab-bed material. Permeabilities in these stone materials were
sufficient  to  provide adequate  pressure field  development  from  usually  one

-------
suction  point.    Permeability of  the sub-slab  media was  recognized as  the
principle characteristic affecting system performance  and effectiveness.  It is
expected  that  this  mitigation  technique  will  maintain  a  high  level  of
effectiveness in the  event  that post-construction penetrations  occur (2).   It
was believed that if major entry conditions could be prevented near the suction
point, then  the negative sub-slab  pressure  field could  successfully protect
against radon transport through new pathways caused by aging of the building.

      The following testing program was developed to investigate the performance
characteristics of  sub-slab depressurization  systems  under  conditions typical
of Florida.  The primary factors that may significantly impact the effectiveness
of sub-slab depressurization systems  installed in typical Florida homes and that
the researchers felt necessary to investigate were:  1) the effects of the type
and permeability of fill material;  2)  the vertical depth or volume of the fill
material;  3) the size and  configuration of the "suction pit";   4) the effects
of air infiltration through the  stemwall; and   5)  the influence of sub-slab
plumbing systems (3).

      The  objective  of  this testing  program was  to determine  the  area  of
effective depressurization  under  various conditions created by these factors.
Four test slabs were  constructed using the slab-in-stemwall technique ensuring
that  the concrete  completely sealed  the  slab/stemwall junction  against  any
leakage  (Figure 2).

      Each slab was built with  the  outside dimensions of  24' x 48' by a local
contractor using standard construction practices.  Following  construction of the
footings  and masonry  stemwalls  each  slab  was  provided  with  a  polyethylene
"stemwall curtain" placed along the inside  face  of the stemwall for half of the
slab's perimeter (Figures 2-A & 2-B).  The curtain extended  from the  footing to
the  top  of  the  fill material where it folded over  the  vapor barrier  for a
distance of  24 inches.  The purpose of this curtain was to effectively seal the
masonry  stemwall against  air  infiltration for that portion of the slab.

      Two  simulated waste  plumbing systems were  installed to determine  the
significance of the pipe or its  trench of disturbed soil on  the pressure field.
Where the plumbing penetrated the slab and stemwall, great care was  taken to seal
against  air  infiltration.

      The  clean sand fill  used in  each test slab was provided from the same
borrow pit by one supplier.  Uniformity of the fill material was maintained as
much as  could be reasonably achieved.

      All four slabs were constructed such that the foundations  were at the same
elevation, and they penetrated into the native soil to the same distance.  This
procedure provided uniform  conditions so the movement of atmospheric air under
the foundation could  be examined.

      Specific construction details  for each test slab follow.

-------
                           V
                . ••• • .' : * •  •».:>
                                   VAPO* |A«*IC«
                                   STCUWALL CURTAIN
                                   (sec  FLAN)
        A. Typical  Section
       c.o.
TWCAL
PLUUVIMO  J—
LAYOUT
(2 CACM SLA*)
                •CVCKSC
                ^LuweiNC
                LAYOUT
         4..
            B.   Typical Plan
                                                                                    •" STONC
                                                           C.  Test Slab #1
                                                                  «" COMOCTC WITH
                                                                          (•>P)
                     E. Test  Slab //3   /


Figure  2.  Test  slab construction details
                                                                                              D.   Test  Slab //2

-------
Test Slab #\

      Test Slab #1 was constructed to simulate the sub-slab conditions found in
basement structures previously studied  and to provide a means of evaluating the
permeability of various sub-slab materials  (Figure 2-C).   A six-inch layer of
3/4"  to  1 1/2"  crushed drain- field stone  similar to  that used  in basement
construction was placed as the slab bed over several inches of clean sand fill.
It was assumed that during placement and grading one to two inches of stone would
be compressed  into  the underlying sand fill resulting in  a nominal  four inch
permeable stone layer.

Test Slab #2 & #3

      These slabs were constructed in order to study the effects of fill depth
and stemwall leakage.  Both Test Slab #2 and #3 were constructed with sand fill
to different depths  (Figures 2-D & 2-E).   Test Slab #2 had a 24" deep layer of
clean sand fill while  Test  Slab  #3 had only an  8"  deep layer of fill over the
native soil.   Test Slab #2,  after final grading,  had 288 square feet of stemwall
with half of it protected by the stemwall curtain.   Test Slab #3 had 96 square
feet of exposed stemwall with half of it protected by the stemwall curtain.

Test Slab #4
      While the other three test slabs were  designed to study various media and
leakage conditions, Test Slab #4 was constructed to evaluate an extended suction
pressure  distribution system.   This  slab  was constructed with  four separate
strips of plastic  drainage matting installed over 8 inches of  fill with each
strip having a separate suction  point  (Figure 2-F). This  test platform provided
the opportunity to study the effectiveness of a continuous,  linear suction pit
versus the point source suction pit used on the  other three  test slabs.   The
stemwall  curtain on this slab was  not only  installed around half the perimeter
of the slab but extended across  the midsection of the slab to effectively divide
the fill into two separate  regions.  This was done to minimize the communication
of pressure fields developed in different suction strips.

Test Procedure

      Following construction of the  test  slabs, multiple suction  points and
monitoring points were installed.  The suction points were installed  at the time
the testing was to commence at  that location.  A3" hole was cored through the
slab and  the pit excavated by hand  to  the desired configuration.  Following the
pit excavation  a 4"  PVC clean-out  adapter with plug was  installed and sealed.
The suction device was then attached  to this adapter  when  testing commenced.
The monitor points  were installed by drilling a hole through the slab  and sealing
a 3/4" PVC pipe 6" long into place.   The pipes  were  then  plugged with rubber
stoppers.   Figure 3  illustrates  the  arrangement of  the suction  and monitor
points.

-------















* 1 "l 'l
o
ir.i
*< 'i "«
n >
" rj

" j * i *
O
' 10 " n * n
^V n
'» '•« "'»
O
*i» 'IT *"





Plur
Roo
• — S
(
^-






                                StcrawiU
                                Curuin
                                 Sicmwall
                                                                          Rjilon
                                                                          Contfol
       Test Slabs #1,  2,  3
Test Slab //4
            Figure 3.   Test slab suction & monitor point locations.
      Pressure  testing was  conducted  using an industrial  vacuum cleaner as the
suction source  regulated through a bleed-valve assembly.  Pressure measurements
were made  at the suction  point and  the  monitoring points using  a  Neotronics
Model MP20SR micromanometer.  Air flow measurements  were taken with a Kurz model
440  air velocity  meter at the  suction point  and periodically  at a  remote
monitoring point.

      Testing routinely started with a measure of air flow at the suction point
at suction pressures ranging from 500 Pa to as high  as  6000 Pa.   After  this
procedure was finished the suction pressure was reduced to approximately 500 Pa
and  a  pressure measurement at  each  monitor point was  taken.   A  commercial
software package (Surfer,  Golden Software) was used to  interpolate between data
points,  using  a Kriging algorithm, and to develop contour lines of constant
pressure.

-------
                            RESULTS AND DISCUSSION
      Figure  4  illustrates  the  difference  In  air  flow  rates  for a  high
permeability condition (Test Slab f 1) and a low permeability condition (Test Slab
#2).   A comparison of the four curves from the sand fill  (low permeability) with
the stone fill (high permeability) demonstrates the dramatic increase In air flow
that is generated at any given pressure  for high  permeability fills.  The line
representing  Test Slab  #1  (stone  media)  indicates  that high permeability
materials require fans capable of handling high flow rates in order  to achieve
the desired pressure conditions under the entire slab.   The large amount of air
Infiltration  into the system on Test Slab #1 must be entering both through the
unsealed stemwall and/or under  the  foundation from the atmosphere. The pressure
field for this test slab is nearly uniform in all directions.  The half of the
slab protected by  the  stemwall curtain has slightly high pressure produced by
limiting infiltration through  the stem wall.
               3000
                      FIST MAK II  (stone fill)
                                   2         3
                                Suction Pressure (KPa)
  Figure 4.  Comparison of airflow  rates  of high and low permeability fills.
      Figure  5  and Figure  6  illustrate the effects  of different suction  pit
configurations.  Four different  suction pit  configurations  were constructed at
the same  suction location on Test Slab #2.  Each configuration was  tested  for'
air flow and pressure field development.  Figure 5 illustrates the relationship
of pit contact area to air flow.  The greater area of sub-slab media exposed to
the highest suction pressure  allows  a  larger pressure field to develop and  the
induces a larger air flow.

      Figure  6  shows  the  pressure  field  contours  for  each  suction  pit
configuration.   Note must be made of the contour values to properly compare  the
effectiveness of the suction  systems.   All plots on Figure  6 indicate a better
field development  toward  that half of  the  slab with the stemwall curtain.

-------
                      600
                    .£• *oo
                      200
                                2        «
                              Suction Pressure 
-------
1200 •
 000
                                                       000
   000  400  600  1200 1600 2000 2400 2800 3200 36 OO 4000  4400 4800  000  400  800  12 00 16 00 20 00  24 DO 2800 1200 3600 4000  4*00 4800
         1" diameter 12"  deep cylinder
4" diameter 12"  deep  cylinder  (no stone)
   000  400  BOO  1200 1600 2000 2400 2600 MOO  3600 4000 4400 48 00  000  400  8 OO 12 OO 16 OO 20 OO 2< OO 2800 35 OO 36 CO  4000 44 CO 4800
                                                                                                              2*00
 000
   000 <00 SOO 1200 1600 2000 2400  2800  3200 3600 «0 00 44 CO 4800  000  400  800  1200  16 OO 2000 2400 2800 3200  3600 «0 00 4400 4600
      A" diameter  12" deep cylinder  with  stone                       12"  hemisphere
     Figure  6.  Pressure  contours produced on Test Slab //2  by various  suction pit configurations.

-------
   c.oo  «.oo  aoo  12.00 it.oo  20.00  2«.oo za.oo 12.00  js.oo  40.00 «4.oo «6.op   200   «.oo  a.oo  12.00  ieoo 10.00 2«.oo  js.oo 32.00 J&.oo »o.oo 44.00  *a.oo
20.00
1J.OO
 4.00
                                                                                                                           - 20.00
                                                                                                                             12.00
   cso  4.00  B.CB  :j.x  i«cc  2; :c  z< os 2sc: 32 co  3500  «co3 <«:: 4»c:   ;:D  4 co  sea  -.2.00  '.ec-c  ZD.OO 2«.oo zsoo  J2.DG  ie.c1: 'coo 400 's.oo
     A. Test Slab  #2   -   Suction  point  //2
                                                                 B.   Test  Slab  #2   -   Suction point it3
   0.00  4.00  800  12.00 1600  20.00  24 00 2800 32.00  3600  4000 44.CO 4S.30   0 -,O  » CC  100  12.00  '.i.OO  2000  24.00 38 CO 32.00  36 OD  4000  4400 4800
                             T—i	1—I	!—I—I—1—:—1—i—I—I—| 2< ~  ,—i	:—:—:	;—I—i—!	!—i—i—I	1—1	i	:—i	;	:—>	:	!	1—i 24.00

                                                             i                                                    .	.



                                                            -i 'e::  -


                                                            -i -2."  -
                                                                                                                             3.00
                                                                   L
000  4.00  8.00  '7.30 1600 JO 00  ?« 00 2800 JJ.OO  36 CO  
-------
  Che stemwall  induces significant infiltration through both the stemwall and soil
  under the footing.  Since Suction Point  #2  is  located immediately adjacent to
  a curtained stemwall, the air flow at this suction point must come primarily from
  under the footing.  The curve for Suction Point  #3  results from both stemwall
  infiltration and  atmospheric  transport under  the  footing since  the stemwall
  curtain is absent.  The difference between  the plots for Suction Point //I and
  //4  is very small and results from the  fact that  the pressure differential at the
  stemwall and ground surface  is  greatly reduced by having located the suction
  strip several feet  in  from the  edge  of the slab.   The  pressure  field plots
  contained in Figure 9 illustrate the  degree of pressure field  development for
  each suction strip.  Figures  9-A &  9-D show that, where  the strip is several
  feet inside  the stemwall, the  field development is virtually identical for both
  suction strips.   Marginal stemwall leakage  is expected.   However,  Figure 9-B
  and 9-C show marked differences in pressure field development.   These results
  demonstrate how eliminating stemwall leakage allows higher suction pressures to
  be  developed and  that  stemwall  infiltration limits  pressure field extension.
  These tests  demonstrate a very  promising approach to  cost effective pressure
  field development.   Locating the suction  strip 8  ft.  to  15  ft.  inside  the
  stemwall would likely  have  resulted,  on this  slab,  in only one  strip being
  necessary to  develop adequate pressure field coverage.
                3000
                                             Suction Point 12
                2000
                1000
Figure 8.  Airflow rate comparison for different  suction points  on Test Slab

-------
    0.00   4.00  8.00  12.00  16.00 20.00 24.OO 28.00 32.00 36.00 40.00 44.00 48.00  0.00  4.00  8.00  12.00  16.00  20.00  2».00  38.00  32.00  36.00  40.00  44.00  48.00
24-00 r-7-1	1	   III—|	1	1	1	1	1	1 --+-- i	1	1	1	1	1	1	1	1	1 24.00



                                                               - 20.00  -
                                                               - U.OO  -
                                                                -  4.00
    0.00  4.00   8.00  12.00  16.00  20.00  24.00  28.00 32.00 36.00  40.00 44.00 *8.00   0.00   4.00   8.00  12.00 16.00 20.00 24.00  28.00  32.00  36.00 40.00 44.00 43.00

         Suction  point  #1   (stemwall  curtain)                 Suction  point  #2  (stemwall  curtain)
24.00



20.00



te.oo



12.00



 8.00



 4.00
   0.00  4.00  B.OO  12.00  16.00 20.00 24.00 28.00 32.00 36.00 40.00 «4.00 48.00   0:00  4.00  8.00  12.00  1600  20.00  24.00  28.00  32.00  3600  4000  4400  «S 00
                                                                 24.00
 0.00
20.00  -
 8.00



 4.00



 0.00
    0.00   4.00  8.00  12.00  16.00  20.00 24.00 28.00 32.00 36.00 40.00 44.00 48.00   0.00   4.00   8.00  12.00 16.00 20.00 24.00  28.00  32.00  36.00 4C.OO 44.00 "8.00

         Suction  point #3 (no  stemwall  curtain)             Suction point  #4 (no  stemwall  curtain)
             Figure 9.    Pressure contours  produced at  each  suction  point on Test  Slab  M.

-------
                                 CONCLUSIONS


      The studies conducted thus far in this research effort seem to provide the
following guidelines.

      The amount of suction contact area is vital to pressure field development.
Larger pits shaped  to maximize surface  area  function  much better  than pits of
smaller surface  area.   The drainage  matting material used as  suction strips
appear to be very promising in its  effectiveness  if placed  8-15 ft.  inside of
a perimeter stemwall.  This suction system  took several minutes  to install versus
several hours for the excavated pits.

      Stemwall curtains in slabs with deep  layers of fill and corresponding large
stemwall areas will perform better with the protection of the stemwall curtain.
Slabs with short stemwalls probably  will not  realize a positive cost benefit of
the stemwall curtain.

      Finally, highly permeable fill,  such as  the stone used in Test Slab #1,
provides  for  excellent pressure  field development if it  is placed  over low
permeability soil and could be used effectively as an active sub-slab vent?. Ration
system.   On the  other  hand, in Florida, where stone is not a locally produced
material  it appears to be  the most expensive option.

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


                                  REFERENCES


1.    Roessler,  C.E.,  Kautz, R.,  Bolch, W.E., Jr. and Wethington, J.A., Jr.
      The  effects  of  mining  and  land  reclamation  on   the  radiological
      characteristics  of  the  terrestrial environment  of  Florida's  phosphate
      region.   The Natural Radiation Environment  III  (Proceedings of  symposium
      April 1978)  U.S  Department of Energy Publication CONF-780422.  pp.  1476-
      1493, 1980.

2.    Hintenlang,  D.E.  and Furman, R.A.   Sub-slab suction system design for low
      permeability soils.   In:  Proceedings of the 1990 International  Symposium
      on Radon and Radon Reduction Technology, Atlanta, Georgia, to be published.


 3.    Fowler,  C.E., Williamson, A.D.,  Pyle,  B.E.,  Belzer,  F.E., Coker,  R.N.,
       Sanchez,  D.C., and Brennan, T.   Engineering design criteria for sub-slab
      depressurization systems in low-permeability soils.   1m   Proceedings of
       the 1990  International Symposium on Radon and Radon Reduction Technology.
      Atlanta,  Georgia, to be  published.

-------
                                                                          VIII-4
        EVALUATING RADON RESISTANCE  OF FILMS AND SEALANTS

                 USING  PERFLUOROCARBON TRACER GASES

                          Mark Nowak and Ban-Huat Song
                          NAHB National Research Center
                             Upper Marlboro, Maryland


                                     Abstract
Movement of radon into a home is controlled by two mechanisms:  diffusion and convective
flow.  Although many different materials have been suggested as mechanical barriers to reduce
radon flow, there  is very little  data to substantiate a  material's effectiveness.   This paper
discusses a relatively low-cost laboratory method to evaluate materials using perfluorocarbon
tracer (PFT)  gases to  simulate diffusion of radon.

The test method uses a modified version of the Air Infiltration Measurement Service (AIMS),
originally developed for measuring infiltration rates  into homes.  The procedure is conducted
in an enclosed  glass desiccator that is divided into two zones by the test material. The PFT
source is placed in the lower chamber  and samplers are located in the upper chamber.  Two
results  are reported:  the reduction in diffusion achieved by the  material as compared to
diffusion in an open chamber, and the  rate of diffusion through the barrier.
INTRODUCTION
Movement of radon from soils into  a  home is controlled by  two mechanisms: Diffusion,
which is a random scattering of a gas across a concentration gradient; and convection, which

results from driving forces that transport radon with other gases.


The driving force behind convective flow can occur in homes as  a result of a reduced pressure
that develops in the lower level of a home relative to surrounding soil.  The thermal stack
effect, wind,  and  operation  of  HVAC  equipment  are principal factors  contributing to  this

pressure differential.  Diffusion  is generally  thought to be less significant than convective

flow, and consequently most attention  has been focused on developing barriers  to reduce

convective flow of radon into  homes.   However, in some areas diffusion  could result in

-------
elevated indoor radon levels.  Barriers that effectively retard  both  mechanisms of transport
will increase the radon-resistance of a building.

Barriers are recommended as the first line of defense  in  new construction  by  most radon
mitigation experts.  Barriers are frequently used under floor slabs and on below-grade walls
to reduce radon entry through cracks, pores, and other openings.   Another important use of
barriers - as interior sealants applied  over building materials -   will  become increasingly
important if the national goal to reduce radon levels to outdoor levels is to be  taken seriously.

Testing the  effectiveness of barriers is  difficult  One method  is to locate a home with high
radon  levels and measure indoor levels  or radon flux through  the wall before and  after
application of the barrier.  Another is  to  test the  barrier's resistance to radon movement in
a laboratory setting using a controlled radon source.  Each of these methods have practical
limitations  that, in  effect, prohibit most  manufacturers from  testing their products.   Costs
associated with construction  and uncertainties in comparing before and after test results in
the first method, and the specialized equipment and  use of a radioactive source in the
laboratory method are the primary factors that discourage testing.

THE AIMS METHOD
An alternative method discussed here employs a tracer  gas to simulate movement of radon.
The method incorporates the perfluorocarbon tracer (PFT) gas technology used in the Air
Infiltration Measurement Service (AIMS)  operated by the NAHB  National Research  Center.

AIMS was  developed in the mid-1980s at Brookhaven National Laboratory as an alternative
to higher cost tracer gas systems in use at that time.  The system can be used effectively to
simultaneously measure infiltration rates in up to four different zones.  PFT sources used in
the AIMS program emit the gas at a constant rate.  Sampler tubes, which passively adsorb
the tracer gas,  are placed in the measurement area.  At the end of the test period, samplers
are returned to the  AIMS Laboratory and analyzed using a gas chromatograph equipped with
an electron capture detector to determine the concentration of the tracer  gas in the sampled
airspace.

-------
Dietz et  aL,(1985) discussed  the  analytical procedure and  assumptions  for determining
infiltration rates using PFT tracers,  termed the steady-state tracer gas method.  As the term
implies, the critical assumption  is that steady-state infiltration conditions occur within a well
mixed chamber.  Calculations (by Dietz et. al.) shown below are based on conservation of
mass (mass-balance) within a single  zone. The mass-balance equation for a single-zone room
or chamber yields:
      dC/dt = (S/V) - nC                                                (equation 1)
             where,
             C = tracer concentration in chamber (pL/L),
             S = tracer source strength (nL/hour),
             V = volume of area to be sampled (m3),
             n = air changes per hour (ACH).
Under steady-state conditions, dC/dt = 0, and the equation becomes:
             n = S/VC                                                  (equation 2)

MEASURING RADON RESISTANCE
Perfluorocarbons are gases whose physical behavior is similar to radon.  PFTs are also very
similar to radon in size: radon has a molecular weight of 222, the PFTs used in the procedure
have a molecular weight of  3SO.  This  similarity permits application of AIMS technology to
evaluate the radon resistance of a barrier using a low-cost laboratory procedure.

TESTING PROCEDURE AND  EQUIPMENT
The modified AIMS test of a radon barrier employs a small scale dual-zone chamber.  The
two chambers of the glass testing device are separated by an aluminum disk with a six-inch
diameter opening.  The test  sample is  sealed to the opening  to form a barrier between the
upper and lower chambers.  A PFT  source with an emission rate of 24.1 nL/min at 2S°C was
placed in the lower chamber and multiple sampler tubes were deployed in the upper chamber.
The samplers  were removed and  analyzed with a gas chromatograph after  a  30 minute
sampling period.

Two tests  of the procedure were conducted on four different samples.  Sample No. 1 consisted
of 4-mil polyethylene film  typically used as a vapor barrier in construction and generally
believed to be a good radon-resistant material.  No. 2 consisted of a single coat of a common
water-based acrylic paint applied over a porous paper backing.  The third sample consisted of

-------
a second coat of paint applied over sample No. 2.  Sample No. 4 consisted of the paper used
as a backing for the paint samples.

RESULTS AND CONCLUSIONS
Results of the testing are shown in Table  1.   Reduction percentages shown in Column 3 are
based on a  comparison with a value calculated for diffusion in an unobstructed chamber.
The diffusion  rates in  Column 4  are of primary importance in evaluating  the  relative
effectiveness of barriers.  This value was calculated from the measured concentration  per unit
time and the internal sampling area.

Under the conditions of this test, the one-coat paint sample was partially effective at  slowing
the diffusion rate of the PFT.  The  polyethylene and the two-coat paint sample were much
more effective at slowing diffusion.  Diffusion  through the polyethylene occurred at a rate
approximately 2.8 times that of the two-coat paint sample.  Due to different properties of the
backing material, it is  expected that the painted sample would not perform as well if applied
to a more porous surface like concrete.

In summary, this procedure offers a relatively inexpensive method to measure and compare
the effectiveness of materials in resisting movement of gases.   It is expected that future
modifications will be made to more closely reflect  conditions of use.  For example, the test
can be run with a pressure differential between  the chambers induced with a small  vacuum
pump.  The pressure differential could range from  three to five pascals, which is similar to
pressure differentials in some homes during winter.  Modification to the procedure could also
be made to more closely  simulate the block or concrete substrates to which paints and other
sealants are  applied  in construction.

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

-------
REFERENCES
Dietz, D.N., Goodrich, R.W. and Cote, E.A. 1984. "Detailed Description and Performance of
a  Passive  Perfluorocarbon  Tracer System  for Building Ventilation  and Air  Exchange
Measurements."  Symposium on Measured Air Leakage Performance of Buildings, American
Society for Testing and Materials, Philadelphia, PA.

Mortimer, C., Chemistry. A  Conceptual Approach. 1975, Van Nostrand Co., New York, NY
                                           Table  1 - Test Results
                                Average
                             Concentration*                         Diffusion Rate
       Test Sample              (pL/L)        % Reduction**        (mole/L-hr.-sq.in.)

       Polyethylene              19.524           99.977               7.063 x lfru
       Paint-1 coat              619.048          99.268               2.281 x lO"
       Paint-2 coats               7.024          99.992               2.542 x I0ra
       Paper                  51.310.242         39.322               1.856 x 104

      *   1 picoliter (pL) of PFT gas is equivalent to 1.79 x 10* grams.
      **  Based on a  comparison of the average concentration to a value of 84,561.40 pL/L in
          the open chamber (calculated from an  emission rate of 24.10 nL/min and a volume of
          8.55 L).

-------
                                                                          VIII-5
               The Use of Coatings and Block Specification to Reduce
                   Radon Inflow Through Block Basement Walls
                              John S. Ruppersberger
             U. S. EPA, Air and Energy Engineering Research Laboratory
                        Research Triangle Park, NC   27711
ABSTRACT - Samples of six different coatings  were evaluated in specially designed
chambers built around 1.5 m2 concrete block wall sections.  Data were collected over a
pressure range of 1 to 12 Pa with flows ranging from less than 0.01 standard liters per
minute  (SLPM)  to  50 SLPM.   A major preliminary finding is  that all these coatings
proved  to be highly  effective (98+%) when enough material was carefully applied.
Baseline (uncoated) flows varied by  a factor of 2 (12.1  to 23.5 SLPM/m2 at 3  Pa)
between the two batches of lightweight block used for coatings testing that came from
a North Carolina manufacturer; these differed by an order of magnitude from normal
weight blocks received later from a Minnesota manufacturer (1.8 SLPM/m2 at 3 Pa).  This
large difference found in a small sampling of blocks is significant not only in the potential
impact on coating performance, but more significantly that specification of blocks with
low air permeability for new construction of substructures could greatly reduce soil gas
entry, even if left uncoated.

INTRODUCTION - Pressure driven transport of soil gas carrying radon  is  believed to be
the major entry mechanism for indoor radon. For houses with basements that have been
mitigated as part of the EPA's research program, the perimeter crack where the floor slab
meets the basement walls has often been considered the primary entry route.  Generally,
sound poured concrete has not been found to offer a significant entry  route, even with
typical hairline cracks. Therefore, coating poured concrete as part of a  radon mitigation
effort is not considered worthwhile  when working toward a guideline of 4 pCi/L.
Basement walls built of hollow concrete masonry units (CMUs) have always been suspect,
and mitigation has included some work with this type of wall. The surfaces of these
walls have been known to  be an  entry route,  but given less treatment since other
mitigation  techniques such  as active  (fan  driven) suction on the soil  side  of  the
substructure/soil interface have been shown to often be effective in achieving a 4 pCi/L
      This paper has been reviewed in accordance with the U. S.
      Environmental  Protection Agency's peer and administrative
      review policies and approved for presentation and publication.

-------
guideline.  Active systems  are often powerful enough to extend their suction into the
core of the CMU wall, effectively reversing the premitigation condition of the basement
air, typically being the suction side of a  slight pressure on the order of -3 to -5 Pa.
Efforts  to quantify the CMU wall source term is not commonly included in radon
mitigation  work  to  date.   When quantified,  it was found  to contribute  up  to
approximately 20% (1). Reports of the effectiveness of painting the surface of CMU walls
have  been mixed with success credited to the treatment by some mitigators and no
benefit  even after  thorough treatment with  expensive  paints by  others.   The recent
development of the long term goal of ambient radon concentration supports more effort
toward  the more common houses, those with  a premitigation level of less than 4 pCi/L.
The relative significance of radon entry routes may be different for this large group of
houses. Another major concern of active soil  ventilation systems is the large volume of
conditioned  indoor air they draw  through a typical uncoated CMU  basement wall,
producing inefficiency in the system and resulting in  an  energy penalty of conditioning
the outside replacement air that is brought to the temperature of the house. Trace gas
experiments indicated that 50% of the air exhausted from one active sub-slab mitigation
system was from the basement. (1)  Extension of the suction to all essential areas of the
soil/substructure interface is made more difficult if CMU walls allow large flows of indoor
air into the system.

      In a Canadian study, external coatings were evaluated for their ability to form an
airtight membrane  that would  remain intact  even if cracks occurred in the substrate
subsequent to their application. Coatings were applied to two adjacent concrete blocks,
which were then moved apart to simulate opening of a crack. (2)

      Recent  work performed  at  Princeton University  found that  block  wall air
permeability was reduced by 99.5+% with two  coats of a special rubberized paint,  a
polysulfide copolymer. One coat was 91% effective.  Two coats of either ordinary latex
or  oil base  paints reduced wall permeabilities by 95%.   Air  permeabilities were
determined from flow versus pressure data. (3)

      A "standard test method for rate of air leakage  through exterior windows, curtain
walls, and doors" has been established by ASTM. (4)   This test method covers  the
determination of resistance of curtain walls to pressure driven air infiltration.  The EPA
test method complies with the ASTM test method in every major aspect, and differs only
in data collections at lower pressure differentials, in the range of 1 to 12 Pa.  Its
instrumentation exceeds accuracy requirements to permit precise measurements at these
very low pressures and flows.

MATERIALS AND PROCEDURES - Concrete masonry  units of the type used in this test
are covered by an ASTM "standard specification for hollow load-bearing masonry units".
(5) This specification covers hollow load-bearing concrete masonry units made from

-------
Portland cement, water; and mineral aggregates with or without the inclusion of other
materials.  The three weight classifications are normal, medium, and lightweight.

      Coatings selected for evaluation include a two part catalyzed water based epoxy
paint, an elastomeric paint, a cementaceous block filler, a fiber reinforced surface bonding
cement, a polysulfide  vinyl acrylic paint, and a latex paint.  Selection criteria included
an attempt to sample various  types of coatings that might be used on CMU basement
walls under various conditions.  Some coatings are not well suited, or even recommended
by the producers for negative side (inside) basement walls, but were evaluated because
they may already exist on some walls of basements needing mitigation, or might be better
suited for application to walls under certain conditions than other coatings.  Coatings
were applied  separately to bare wall  specimens  even though the desired result of a
continuous gas flow resistant film might better be achieved by a combination, such as a
block filler and a coat of paint.  This was done to collect data on each coating separately
- - performance of various combinations may be estimated from the data, but to be more
precise,  and  perhaps  more  realistic  in some  instances, reasonable  recommended
combinations  should be subjected to further evaluations.  The effectiveness of a single
coat of the two part water based epoxy  argues against a  strong need for further tests.
Freshness of samples and adherence to application directions were emphasized.  No two
coatings were produced by the same manufacturer.

      The test stand was designed for a 16 ft2 (1.5 m2) CMU wall.  The wall assembly
is made by pouring a  concrete footing (48 x 16 x 6  in., or 122 x 41 x 15 cm) on which
a block wall of 15  standard blocks  and  6 half  blocks is  carefully built.  Mortar
construction techniques vary;  these walls were built with two fairly generous strips of
mortar on which the base  course of blocks are laid.  Mortar is applied to all horizontal
surfaces of the previous course and to the end of the next block that will butt up against
the last block on a course in progress.  After the wall has set up  for over a week it is
caulked generously and, while wet, the side and top panels are assembled then fitted with
covers to encapsulate the wall with a plenum on either side.  Closed cell rubber gasket
material is sandwiched between all mating metal surfaces, and between the metal and
acrylic plastic covers.

       The completed  assembly is leak tested by pressurizing to between 2 and 3 in.
H20 (500 to 750 Pa) using helium gas,  and tested for leaks  using a halogen gas leak
detector.   Leak testing is also  conducted  on the pressure  side  of  the air delivery
equipment. The control panel is composed  primarily of computer controlled mass flow
controllers, a pressure transducer a pump, and a bypass valve that provide precise control
of flows from less than 0.01  to 50 SLPM over a pressure range of 1  to 12 Pa.  The
acrylic plastic cover over  the side of  the wall to be  painted is then removed.  After
baseline data are collected, the coating is carefully applied.   Care is taken to quantify
material used. Additional coats were applied a day  after the previous coat except when
data collection needs  dictated longer periods or when it might be reasonable to stop

-------
application of that paint with that coat. It is important to note that the coatings were
applied by brush, carefully working material into the block surface and leaving as much
material on the wall as possible without runs.  Primary consideration was sealing the
porous block surface, not the amount of material used. Exceptions are the elastomeric
paint which was  applied at the maximum recommended rate and the surface bonding
cement which was applied using a steel trowel.

      The initial wall was constructed  as part of prototype development and was used
for the polysulfide vinyl acrylic paint evaluation. Its baseline flow was as shown in
Figure  1, Line a, 35 SLPM at 3 Pa, or about 2 SLPM per full block. The remaining five
coatings were evaluated from walls built later; from a different batch^of blocks. Although
from the same local North Carolina manufacture]; meeting the same ASTM specifications
for CMUs, and  with no noticeable  difference  in appearance, the baseline flows were
approximately half of that of the prototype wall, as shown  in Figure 1, Line b, 18 SLPM
at 3  Pa, or about 1 SLPM per full block.  Then an opportunity developed to evaluate
blocks received from a Minnesota manufacturer.  The baseline flow for these blocks is on
average approximately an order of magnitude lower than the  local blocks, as shown  in
Figure  1, Line c,  2.7 SLPM at 3 Pa or about 0.15 SLPM per full block.  Blocks from both
manufacturers are typical of those  used in their geographical  regions for residential
construction. The more air permeable North Carolina block is lighter, 12.1 kg, and has
become common  in the southeast.  It  contains expanded  lightweight  aggregate, filled
with numerous  discrete voids  that do  not appear to be interconnecting.  The less air
permeable Minnesota block weighed  16.9 kg, uses natural aggregate, and has a smoother,
less porous surface appearance.

WATER BASED EPOXY PAINT - This is  a water based catalyzed  (two part) epoxy resin
paint.  Its  analysis by weight, as supplied by the manufacture];  is 16.5% titanium dioxide
pigment and 83.5% vehicle (7.7% epoxy resin, 6.6% ethylene glycol and alcohol, 20.7%
acrylic  resin, 46.5% water, and 2.0% additives). The data for this paint  are summarized
with other coatings in Table 1. Even with a slightly higher baseline flow of 19.8 SLPM,
a single coat of this epoxy paint resulted in the lowest airflow for one coat of any paint
evaluated, 0.75 SLPM (96.2% reduction) at 1 day drying time, and was also lowest for
any two coats of paint evaluated, at 0.01  SLPM (99.9% reduction).  The paint film is
very smooth, and dried specimens exhibited unexpected elongation and strength upon
being pulled  apart, although  these  observations were not quantified by  any standard
testing techniques.   Application was  considered  easier  than  average to provide a
continuous film for both the first and second coats.

ELASTOMERIC PAINT - The analysis of this elastomeric acrylic emulsion paint was not
given.  Application rate for concrete block was specified as 50 to 125 ftVgal. (1.23  to
3.07 mVL). The first coat was applied at 50 ftVgal.  Based on the performance of the
first coat,  a second coat was also evaluated.  About a third  of the quantity of paint used
for the first coat was used for the second coat.  The data for this paint  are summarized

-------
with other coatings in Table 1.  The measured effectiveness of the elastomeric paint was
second only to the epoxy: 1.36 SLPM (92.1% reduction) for one coat, and 0.025 SLPM
(99.9% reduction) for two coats. Application by brush was considered the easiest of any
paint evaluated.

CEMENTACEOUS BLOCK FILLER - This is a portland cement plaster that can be troweled,
sprayed, or brushed. There was no information concerning composition on the container
(a bag holding 50  lb,  or 23 kg) or sales literature, other than references to portland
cement. In keeping with the general trend of brush application, this product was applied
with a "masonry brush" purchased from the dealer It required a different technique than
typical paints, but application progressed satisfactorily after a brief time.  Application was
considered more difficult than average to provide a continuous coating with reasonable
surface  appearance.  Experience would probably improve both application rate and
appearance.   The other cementaceous  product  evaluated was  troweled on,  but  that
product specified a 1/8-in. thick coating; this cementaceous block  filler (also called a
"finish coat" by the manufacturer) was applied at a thinner consistency and thinner than
1/8-in. by brush. The data for this coating are summarized with other coatings in Table
1. The single thick coating resulted in an air flow of only 0.06 SLPM (99.7% reduction);
only about half of the flow through the fiber reinforced surface bonding cement, and over
an order of magnitude less than one coat of the most effective paint.

SURFACE  BONDING CEMENT  - This  is  a  mixture of  portland cement,  fiberglass
reinforcement fibers, and unspecified (proprietary) ingredients. Application is specified
as a minimum of 1/8-in. thick with coverage per 50 lb bag of approximately 50 ft2.
Trowel or spray application options are in the product literature supplied by the producer.
Application was with a steel trowel by an experienced mason to a thickness of slightly
over 1/8-in (0.32-cm). The data for this coating are summarized with other coatings in
Table 1.  The single application resulted in an air flow of only 0.10 SLPM.

      This is more than the other portland cement coating evaluated in this study, but
still is highly  effective at 99.5% flow reduction in one application. This single application
allowed less than 14% of the flow of one coat of the most effective paint.

POLYSULFIDE VINYL ACRYLIC PAINT -   The  supplier described it as polysulfide/vinyl
acrylic dispersion without giving any further specifics on composition.  It was offered at
the time the program was started and was used for the original prototype test stand and
equipment testing.   Since  it  was accepted for those first developmental tests, it was
decided to evaluate it as the first coating  using the equipment after it was fully calibrated
to QA/QC specifications.   It is  currently available commercially to radon mitigators.
Application was considered average  to provide a continuous film for both the first and
second coats, although pinholes were observed soon after application and their apparent
number and size increased with drying  time.  The data  for this paint  are summarized
with other coatings  in Table 1.  Specifically, the baseline flow was much higher than for

-------
the walls built later from another batch of blocks.  The measured effectiveness of one
coat of this paint was 78.6%; many pinholes were apparent in one coat.  The second coat
reduced air flow at 3 Pa substantially, with an effectiveness of 99.4%: 0.20 SLPM of the
baseline  flow of 35 SLPM.   The observed time dependence on measured effectiveness
resulted  in reconsideration of a standard condition of 2 weeks drying time since data
collected at different drying  times exhibited  different slopes and a consistent trend  of
increasing flows with increasing drying time.  Performance deteriorated further between
2 weeks  and 2 months with  visual pinholes becoming more numerous  and larger.

LATEX PAINT - This is a latex semi-gloss paint.  It is commonly sale priced at retail and
might be considered the type of latex paint homeowners would  often buy; there are both
less  and  more expensive latex paints available.  Its analysis did  not list  ingredients by
weight, merely as pigment (titanium dioxide, and hydrous aluminum silicate) and vehicle
(polystyrene  resin, acrylic latex, vinyl  acrylic  resin, 1,  2 propanediol,  additives, and
water). Application information on the  label  stated 400 ftYgal.(9.82 mz/L) or less, and
that textured surfaces may require more paint.  Actual  application rate on the test wall
was approximately 100  frYgal (2.46 m2/L).   That first coat took as much paint as the
next two coats combined, resulting in a coverage after  the three coats of approximately
50 ftVgal (1.23 mz/L).  Application of the paint by brush was considered slightly more
difficult and time consuming  than the average of the paints evaluated.  The data for this
paint are summarized with other coatings in Table 1. The measured effectiveness of this
latex paint was less than any of the other three paints evaluated for either one or two
coats.  One coat was not very effective,  about an order of magnitude less effective than
the epoxy or  elastomeric paints, allowing 11 of the baseline 19 SLPM (42.1% reduction)
to pass at 3 Pa after 1 day drying time.   It was the only paint  applied with three coats,
but that third coat greatly increased the effectiveness, from 84.2% with two coats up to
98.1% (0.37  SLPM) at three coats.

DISCUSSION - Reduction of  air entry is the primary concern that motivated  this work
and all aspects of the test program.  Several observations are especially noteworthy. The
first is that, of these coatings, everything works well in reducing air flow across these test
walls (initially at least, under these ideal conditions) if sufficient  material and care are
used in their application.  A major finding, and a surprising one  after  discussions with
several paint  chemists working to formulate especially, effective radon  flow resistant
paints, and seeing fairly expensive specialty paints being advertised specifically as radon
resistant,  is that all these  coatings, when  carefully  applied  with sufficient quantity,
demonstrated  that they can be highly effective in reducing gas flow across the face  of
concrete  masonry units.   Another is the interesting observation that the data  for any
particular set of  flow  vs  pressure  plots are a straight  line  on arithmetic paper.
Apparently, flow through these small openings  at low pressure differentials is laminar.

-------
       A comparison of percent effectiveness in reducing the baseline flow and estimated
cost associated with do-it-yourself (approximated by the material cost) and professional
application total cost is summarized in Table 1. Among the paints, the polysulfide  vinyl
acrylic offered as specifically formulated for radon control performed much less effectively
(78.6%) than either the  epoxy (96.2%)  or elastomeric (92.1%) with only one  coat.
Cost of the polysulfide vinyl acrylic paint sold by the manufacturer to radon mitigators
is nearly equal to that  of the water based expoxy.  Of commercially available paints, the
latex is the least expensive material for three coats.  However, a do-it-yourselfer would
have to be strongly motivated to save a few pennies per square foot  and get about 2%
more effectiveness,  to go to the trouble of applying three coats of a slightly more difficult
to apply paint, than one coat of the epoxy that provided almost equivalent effectiveness.
The better appearance and other desirable properties of this  epoxy over the latex  paint
might  also influence a final decision.   Considering just  the paints, one coat of an
equivalent water based epoxy might be a reasonable choice since the trouble and expense
of a second coat produce  only a little over 3% increase in effectiveness.

       From these data, the  clear performance value leader  is the cementaceous block
filler. This product contains portland cement, but no fiberglass. The  product evaluated
was white; if a natural  grey color similar to the original block wall is acceptable, it would
be approximately^  penny less per square foot material cost.  This type of coating, highly
effective at the lowest  cost, and with the effect of significantly changing the  block wall
appearance  to a much smoother plastered look,  is an apparent first choice unless
conditions in the basement do not favor its application. Such adverse conditions would
also produce  problems with paints  in general.  The surface bonding cement  is only
slightly more  expensive.   It has added advantages  of its fibers providing added tensile
strength that may be helpful for walls experiencing problems with minor cracking, and
also has  a higher portland cement content that  imparts improved waterproofing
performance ~ although this and most others are recommended for exterior application.

       In summary,  considering both cost and effectiveness, for coating an existing wall,
a cementaceous product such as the cementaceous block filler is apparently the first
choice.  If a paint is desired, the choice is more complicated based on these results, but
one coat of  a similar performing water based  epoxy would be good if about  96%
effectiveness is acceptable.  If top effectiveness is the only criterion, two coats of the
water based epoxy is found to be the most effective of the coatings evaluated. Of course,
building the wall with low air permeable blocks in the first place could decrease the need
for any coating at all. Dry stacking with fiber reinforced surface bonding cement would
also provide soil gas entry resistance.

-------
PRELIMINARY CONCLUSIONS AND RECOMMENDATIONS:

      1)    All of the six coatings selected for evaluation can provide highly effective
reductions in air  flow under the conditions of the tests when sufficient quantity is
carefully applied.

      2)    Considering both cost and effectiveness, a cementaceous product such as the
cementaceous block filler offers highly effective flow reductions in a single application at
low cost. Among paints, two coats of the water based catalyzed epoxy was found to be
most effective in these tests, although one coat may be adequate for some situations.

      3)    The  variation in air flow characteristics between similar looking blocks is
large; in this limited sampling of only two batches of lightweight blocks from a North
Carolina manufacture]; it was found to vary by  about  a  factor of 2 (12.1 to 23.5
SLPM/m2 at 3 Pa). The difference between the North Carolina blocks (lightweight, 17.8
SLPM/m2 average flow at 3 Pa) and Minnesota blocks (normal weight, 1.8 SLPM/m2 at
3 Pa) is an order of magnitude. This large difference found in a small sampling of blocks
is  significant not  only  in  the potential impact  on  coating performance,  but  more
significantly that specification of blocks with low air permeability for new construction
of substructures could greatly reduce soil gas entry, even  if left uncoated.

      4)    Flow vs pressure data collected for these carefully constructed test walls and
at this very low pressure range were found to be linear.

REFERENCES:

1.    Hubbard, L.M., et al. Research on Radon Movement in  Buildings in  Pursuit of
            Optimal Mitigation. Proceedings of American Council for an Energy Efficient
            Economy 1988 Summer Study. Vol. 2. Asilomar. California.  August 1988.
2.    DSMA ATCON LTD.   Atomic Energy Control Board Development  Program for
            Radon Reduction,  Report 9, "Laboratory Tests External  Coatings," August
            1979.
3.    Maryonowski, J.M. "Measurement and Reduction Methods of Cinder Block Wall
            Permeabilities," Center for Energy and Environmental Studies,  Princeton
            University, Working Paper No. 99, 1988.
4.    ASTM E 283-84, "Standard Test Method for Rate of Air Leakage Through Exterior
            Windows, Curtain Walls, and  Doors," September 1984.
5.    ASTM C 90-85, "Standard Specification for Hollow Load-Bearing Concrete Masonry
            Units," July 1985.
                                        8

-------
                                             First botch of local blocks
                                             Second botch of local blocks
                                             Blocks from Minnesota
§  20  -
    10  -
                          456789

                              PRESSURE (Pa)
             FIGURE 1. Air flow  through block walls.
10   11  12  13
   TABLE 1. EFFECTIVENESS OF COATINGS IN REDUCING AIR FLOW THROUGH
            CONCRETE BLOCKS AT 3 PASCALS
Effectiveness f%)
Estimated Cost (dollars) 1989
Per Sauare Foot
Greater Than 99% Effective
Epoxy, 2 coats water based
Elastomeric, 2 coats
Cementaceous Block Filler
(brushed thick)
Surface Bonding Cement
(1/8 • in. troweled)
Polysulfide Vinyl Acrylic,
2 coats
Greater Than 98% Effective
Latex, 3 coats
Greater Than 90% Effective
Epoxy, 1 coat, water based
Elastomeric, 1 coat
Greater Than 75% Effective
Latex, 2 coats
Polysulfide Vinyl Acrylic,
1 coat
Less Than 75% Effective
Latex, 1 coat
99.9
99.9
99.7
99.5
99.4

98.1

96.2
92.1

84.2
78.6

42.1
Material
0.37
0.50
0.14
0.20
0.36

0.22

0.25
0.37

0.17
0.23

0.11
Labor
0.34
0.30
0.20
0.22
0.30

0.42

0.20
0.18

0.30
0.18

0.18
Total
0.71
0.80
0.34
0.42
0.66

0.64

0.45
0.55

0.47
0.41

0.29
Typical Basement'
D.I.Y Professional
440 850
600 960
170
240
430

260

300
440

200
280

130
410
500
790

770

540
660

560
490

350
    1 Typical" is assumed  to be approximately 1200 ft* of wall surface,  in good
    condition, ready to be painted. Costs for brushes, dropcloths, preparing the base-
    ment area and walls, etc. are not included.  D.I.Y. Means Do It Yourself.

                                   9

-------
                Session C-VIII:
Radon Prevention in New Construction—POSTERS

-------
                                                              C-VIII-1
                                   N WOOD FOUNDATION SYSTEM
                  by:    Roscoe J. Clark
                        Permanent Wood Foundation, Inc.
                        Flint, Michigan  48501
                                   ABSTRACT

      Radon, ...an issue of growing concern to the building industry.
Silently, invisibly, it invades existing structure ...as it will future
foundation structures.

      This paper will address the nature and causes of radon, and cost-
effective prevention and retrofit techniques used for wood foundation systems.

      Radon also can enter homes with foundations that use the under-floor as
an air distribution system.  These building practices will be shown; even
materials used in construction may release radon, for example, this may be a
problem in a house that has a solar heating system in which its heat is stored
in large beds of stone.  Stone is most often used in wood foundation
construction.

      The common radon entry points will be looked at, and the latest
prevention techniques will be illustrated, such as natural and forced
ventilation, sealing major radon sources and entry routes, and sub-slab and
sump crock ventilations.

P.W.F. STUDY

      The U.S., Canada and overseas are studying the effectiveness of various
ways to reduce high concentrations of radon in houses.
Each year the study gives us a better understanding about radon and its
effects.  This paper describes methods that have been designed for successful
reduction in permanent wood foundations.  The information presented here is
concerned primarily with radon which enters a house from the underlying soil.

FACTS

      The first fact about radon reduction is "no two houses are alike."
Houses that are built exactly the same have small  differences in them that
affect radon entry.  These differences can effect the design and effectiveness

-------
of radon reduction techniques.  P.W.F. homes can have major differences in
radon sealing techniques compared to conventional foundations.

      This paper is intended primarily for permanent wood foundations and
those who have decided that they need to incorporate radon-reduction methods.
The two basic goals are to minimize radon entry and the removal of radon gas.
In fact, it is potentially more cost-effective to build a P.W.F. Home that
resists radon entry than to remedy a radon problem after construction.  This
paper should only be used as an attempt to remedy a radon problem, and to give
an understanding so not to cause other performance problems to be discussed
below.

THE RISKS

      The known health effect associated with exposure to elevated levels of
radon is an increased risk of developing lung cancer.  There is no doubt that
sufficient doses of radon and its progeny can produce lung cancer in humans.
Radon is believed to be the greatest risk of getting lung cancer for non-
smokers .

      As with other pollutants, when studying the exposure to radon, there is
some uncertainty about the amount of health risk.  The risk estimates are
based on scientific studies of miners exposed to varying levels of radon in
underground work.  A more certain estimate of risk (than on studies) rely
solely with animals.  Despite that same uncertainty in the risk estimates for
radon, the greater your exposure to radon, the greater your risk of developing
lung cancer.

RADON ENTRY

      Radon is a gas which can move through small spaces in the soil and
gravel on which a house is built.  Radon can seep into a home through the sump
crock, joints, floor drains, cracks in concrete floors, etc.  Radon has been
found in well water and can release radon into the home when the water is
used.

      Some building materials have released radon; for example, stone
fireplaces or solar systems which used stone beds for heat storage.  However,
building materials are not a major source of indoor radon.

      The physical relationship between the major sources of radon and the
indoor  structure of a house is illustrated in figure  1.  Common entry routes
for radon gas into the house are shown.  The major entry points for radon  into
the house include:

      A.  Cracks in concrete floors
      B.  Joints in building materials
      C.  Sumps - exposed soil and tile
      D.  Stone materials  (i.e., granite)
      E.  Water  (i.e., shower)

-------
      The soil is believed to be the greatest contributor of indoor radon in
typical homes.

NEW CONSTRUCTION

      1.  Homes should be designed and built to maintain a
          neutral pressure differential between indoors and
          outdoors.

      2.  Homes should be designed and constructed to
          minimize pathways for soil gas to enter.

      3.  Features can be incorporated during construction that
          will facilitate radon removal after completion of the
          home if prevention techniques prove to be inadequate.

      4.  The first step in building new radon-resistant
          permanent wood foundations is to determine the
          potential for radon problems at the building site.
          Check with neighbors or state and local health
          agencies for soil test results.

CONSTRUCTION TECHNIQUES

      Some of the radon prevention techniques are common building practices in
permanent wood foundations.  Others are less costly if accomplished during
construction.  The cost to retrofit an existing home with the same features
would be significantly higher.  These construction techniques do not require
any fundamental changes in building design.  Supervision over quality control
is needed for certain construction details.  Radon entry techniques can be
grouped into two basic categories:

      1.  Methods to reduce pathways for radon entry.
      2.  Methods to reduce the vacuum effect of a home
          on surrounding and underlying soil.

These techniques are used in conjunction with each other  (see figure 2).


UNDERSTANDING THE PERMANENT MOOD FOUNDATION BASIC REQUIREMENTS

      The Permanent Wood Foundation is a load-bearing wood frame wall system
designed for below-grade use as a foundation for light-frame construction.

      The stress-graded lumber framing and plywood sheeting in  the system are
carefully engineered to support lateral soil pressure as well as live, dead
and climatic loadings.  Vertical loads are distributed to the supporting  soil
by a composite footing consisting of a wood footing plate and a structural
gravel layer.

-------
      All lumber and plywood in contact with or close to the gravel is
protected against decay and insects by pressure treating with time proven wood
preservatives.

      Moisture control measures based on the latest development in foundation
engineering, construction practice and building materials technology are
employed to achieve dry and comfortable living space below grade.

      The most important of these measures is a porous gravel envelope
surrounding the lower part of the basement.  This porous layer directs ground
water to a positive drain or sump, thus preventing hydrostatic pressure on the
basement walls or floor.  Similarly, moisture reaching the upper part of the
basement foundation wall is deflected downward to the gravel drainage system
by polyethylene sheeting or by the treated plywood itself.  The result is a
dry basement space that is readily insulated for maximum comfort and
conservation of energy.

WORD OF CAUTION

      What works for conventional foundations may have significant negative
effects when applied to Permanent Wood Foundations.  It is important to
understand the basic requirements of the Permanent Wood Foundation system
before applying mitigation measures which could cause problems to the system's
performance.  This Radon Check List may be helpful when applying mitigation
measures.
                               RADON CHECK LIST

SEALING WALL POLY FOR RADON CONTROL

      Six-mill thick polyethylene sheet should be applied over the below grade
portion of exterior basement wall prior to backfilling.  Joints in the
polyethylene sheet shall be lapped 6 inches and bonded.

      The top edges of the polyethylene sheeting should be bonded to the
plywood sheeting.  A treated lumber strip should be attached  to the wall to
cover the top edges of the polyethylene sheeting.  The wood strip shall extend
8 inches above and 4 inches below finish grade level as required to protect
the polyethylene from exposure to light and from mechanical damage at or near
grade.  The joint between the strip and the wall shall be caulked full length
prior to fastening strip to the wall.  The polyethylene sheet shall extend
down to the bottom of the wood footing plate but shall not overlap or extend
into the grave footing.  Do not seal the bottom of the polyethylene sheet to
the plywood wall for radon resistance.  Moisture is deflected downward to the
gravel drainage system by this polyethylene sheeting and by the treated
plywood wall itself.  Sealing this joint would result in preventing positive
drainage to the gravel.  Simply put, the basement may leak.

SEALING FLOOR POLY FOR RADON CONTROL

-------
      A six mill thick polyethylene moisture barrier shall be applied over the
porous layer of gravel.  Overlap the seams in the barrier 12 inches and seal.
Penetrations of the barrier by plumbing should be sealed and care should be
taken to avoid puncturing the barrier when pouring the floor.  The floor
polyethylene should not lap under the footer plate or seal to the wall poly
for radon control.  This will prevent positive drainage of the wall poly or
moisture from inside the basement wall plates draining through the footer
plate to the gravel footer.  Moisture would be drained on top of the floor
poly causing hydrostatic pressure and dampness.  The poly can be brought up on
the inside wall surface and sealed to the screed board.

INSTALLING POLY FOR WOOD FLOOR SYSTEM

      Where wood basement floors are used, the polyethylene sheeting six mill
shall be placed over the wood sleeper supporting the floor joist and,
provisions are made for drainage to the porous layer below at the end of each
bay.  The sheeting should not extend beneath the wood footer plate.  The poly
is overlapped 12 inches and is not sealed for radon control.

      Water that accumulates on top of the moisture barrier during
construction can be dried out by venting the wood floor.  Radon gas can also
be vented out through these vents by convection.  The studies show that three
vents work better than two and should be placed 4 feet or farther away from
any corners for best performance, (see figure 3)

USING LARGE GRAVEL TO VENT RADON

      Gravel should be washed and well graded.  The maximum size stone should
not exceed 3/4 inch.  Gravel shall be free from organic, clay or salty soils.

      Sand shall be coarse, not smaller than 1/16 inch grains and shall be
free from organic, clay or salty soils.

      Crushed stone shall have a maximum size of 1/2 inch.  Crushed stone must
also be compacted before installing footer plate.

      The Permanent Wood Foundation system incorporates a composite footing
consisting of a wood footing plate and a layer of gravel, coarse sand or
crushed stone.  The wood footing plate distributes the axial design loads from
the framed wall to the gravel layer which in turn distributes it to the
supporting soil:

      *the use of larger than 3/4 inch stone for radon
       mitigation can cause inadequate bearing transfer.
       This means less wood to stone surfaces that can cause
       the stone to crush into the wood footer plate causing
       settling of the foundation.

      *the use of larger than 1/2 inch crushed rock will have
       jagged edges that will break off under load causing
       settling.

-------
      *larger stone makes it harder to level foundation walls.

CAULKING THE P.W.F.

      The sealant used for caulking joint in plywood wall sheeting should be
capable of producing an effective moisture seal under the conditions of
temperature and moisture content at which it will be applied and used.  Buytl
caulking is commonly used for this purpose.  To correctly apply caulking into
joint, apply caulking on edges of plywood joint, push next sheet into place.
Be careful not to slide off caulking.  This has several purposes; one is to be
an expansion joint for the plywood, and another is to give resistance to
moisture flowing down the wall and can have sealing effects in radon control.
Caulk end and center seam of wall plywood.  Do not caulk the bottom edge of
plywood where it buts up next to bottom plate.  This will stop the pathway for
any water that accumulates from condensation on the inside plywood surface
during winter construction.  This damming of water can be picked up by the
wall insulation and lead to mildew and drywall staining.

TOOLING BASEMENT FLOOR/WALL FLOOR JOINT

      The more common radon entry pathway are inside perimeter floor/wall
joints.  To reduce radon entry through these joints, install a wide plywood
screed board along the perimeter of the foundation.  This wider screed board
can be sealed to the drywall finish installed later on the foundation wall
above.  Mud and tape this joint when finishing the drywall above.

      The concrete floor is poured against the screed board and the concrete
edge is tooled round for a caulk sealer to be applied later.

      Oversize tooling of this joint may cause lateral wall deflection at the
floor/wall assembly.  Proper floor height against studs will allow adequate
bearing area.  This bearing area is figured from the bottom of the tool joint.

      Remove all grade stakes and fill the holes as the slab is being
finished.  This will prevent future radon pathways through the slab which
might otherwise be created, as embedded untreated wood eventually
deteriorates.

      Carefully seal around all pipes and wires penetrating the slab.  Pay
careful attention to the bathtub, shower, and toilet openings around traps.

      Floor drains, if installed, should drain to a sealed sump crock, and
used traps in all floor drains.

      Sump should be sealed at the top with a plywood and gasket lid.  Use
only submersive-type sump pumps in the crock to prevent high humidity
corrosion.  For sump crock sub-slab ventilation, drill holes near the top of
the sump wall to let soil gases enter crock, (see figure 4)

METHODS TO REDUCE PATHWAY FOR RADON ENTRY

-------
      The most effective way to seal a P.W.F. basement wall is by drywalling
the wall.  First, the bottom seal must be applied.  This is done by applying a
gasket to the bottom edge of a added plywood screed board.  This is pushed
down into place and the wood strip is nailed into place (use 1/2 inch plywood
for 1/2 inch drywall).  Next, install a gasket at the top top-plate.  Now
install new drywall, compressing the gasket between the plate and the drywall.
(see figure 5)

SUB-SLAB VENTILATION USING WOOD FLOOR SYSTEM

      To seal a wood basement floor system, first install a wider plywood
screed board.  A gasket is installed to the plywood screed where it will be
sandwiched between the wood floor ban joist.  A gasket is installed on the top
edge of the ban joist so that gasket will be sandwiched between the plywood
and ban.  The gasket makes a more permanent seal over caulk or glue.

      Caulk plywood edges as the floor is installed with buytl caulking.  If
mechanical ventilation is needed, put suction pipe under poly film for better
performance.  Keep above the water line.  A tee fitting will add to the
performance,  (see figure 6)

SUB-SLAB VENTILATION USING SUMP SUCTION APPROACH

      Where better performance is needed, the P.W.F. sump can be vented by
sealing the top with a plywood lid and gasket.  The water enters the sump from
the bottomless crock.  This system will cause a water trap effect.  Therefore,
drilling holes at the top of the crock will help let in the soil gases so it
can be vented, (see figure 7)

SUB-SLAB VENTILATION OF WOOD FLOOR SYSTEM

      To vent a wood basement floor, cut out a section of the ban joist.
Install 2x2 and sheeting to make a pathway to the outside up the foundation
wall.  Install a screen vent cover on the outside,  (see figure 3)

SUB-SOIL VENTILATION USING HORIZONTAL RUN UNDER VAPOR RETARDER

      When using the crawlspace as a heat plenum, always have the crawlspace
in a positive air flow.  Running the blower constantly will minimize the
negative pressure differences.  A polyethylene film is installed over a 4 inch
layer of gravel and the polyethylene is covered with 2 inches of sand.
Overlap the polyethylene 12  inches at all seams and seal.  Seal polyethylene
to film on the wall.  For better performance, a tee or cross network of 4 inch
p.v.c. can be connected to a fan-driven system.
 (see figures  8 and 9)

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

-------
MAJOR  RADON ENTRY  ROUTES
                            CRACKS
                            JOINTS
                            SUMP
                            MATERALS
                            WATER
 FIGURE 1

-------
      METHODS TO REDUCE THE VACUUM EFFECT
                    SEAL AROUND
                    FLUES AND
                    CHIMNEYS
                                              AVOID RECESSED
                                              CEILING LIGHTS
                                              IN UPPER CEILING
                     v-SEAL AROUND
                     \ALL  OPENINGS
                                           SEAL AROUND
                                           PLUMBING
                                           PENETRATIONS
                            SEAL
                            ACCESS
MAKEUP
AIR
CPAWL SPACE

                           KNGHT
                            SEALING
                            WINDOWS
       FIGURE z
                                VAPOR RETARDER-

-------
   SUB-SLAB VENTILATION OF WOOD FLOOR SYSTEM
    VENT—o-?0
GRADE
SEAL DRYWALL
TO PLATE
                    SHEATHING
      &
    DRYWALL TO
SCREED BOARD
                                           SEAL JOINTS
                       ff      *~       W
                     —J  _ o_	 _ .. _P xrxav	
                     0f°&~SG* ° <2,°1&° ^O*0 Q° O " °/5*O
                     o/^«	e °_To .—i <>*?__l»^ a ^o o^i er v%» r~l ^o^f1
                     •QPop «CSc7 ^ o o^P.0 "oC? o o 0-0-o° . °^

                     BSBiilfisH    '
          SOIL GAS
           VAPOR  RETARDER
    FIGURE

-------
     Sub-slab ventilation using  individual suction point  approach.
Exhaust
                                   Seal  drywall to plate
Grade level
                                  Tape drywall to screed board

                                        Seal  joints

                                                                     4
Connect
to other
suction  points
                                                                   House air through
                                                                      sealed cracks
                                                                   and joints
                                          Soil  gas
      FIGURE 4

-------
       METHODS TO REDUCE  PATHWAYS  FOR  RADON  ENTRY


               rAm
GRADE LEVEL
                         SEAL DRYWALL
                         TO PLATE
                                   SEAL ALL PLUMBING
                                      PENETRATIONS
                            TAPE DRYWALL TO
                            ADDED SCREED BOARD
                            SEAL JOINTS-7
               CAULK CRACKS
 SOIL GAS
         *-VAPOR RETARDER
        FIGURE 5
TRAPPED DRAIN
TO SUMP, SEWER
OR DAYLIGHT

-------
    Sub-slab ventilation using wood  floor system
     Exhaust
Grade  level
                                I—Seal  drywall to
                                   plate
       Soil gas
                                                   Soil  gas'
    FIGURE 6

-------
 Sub- slab ventilation using sump suction approach.
Exhaust'
                                Seal drywall
                                to plate
                                                             House air through
                                                             unsealed  cracks
                                                             and joints
                                                              Utility pipe
   FIGURE 7

-------
    Perimeter drain tile ventilation where tile drains to sewer or  daylight.
                 Protective cover
          Exhaust

Riser connecting drain  tile —
to fan

Condensate
                       |(35)--Fan
                                                  Seal drywall
                                                  to  plate
                                                       House air  through
                                                       unsealed cracks
                                                       and drains
Capped  riser toadd
water to trap
                                     j^-»-/*Va
                                     *S»o0oo<'<'Ws<» «5U
          •Water trap to prevent air from being
           drawn up from sewer or daylight
       FIGURE 8

-------
   Sub- soil  ventilation using horizontal   run  under  vapor retarder.
 Exhaust

'V
Ife
I I y-Floor register
U1
Grade  level
                                                              HVAC
                                                               Unit


                                                              3///////I
1
                                                        Positive  air  crawl
                                                        space plenum
     FIGURE  9

-------
            Session IX:
Radon in Schools and Large Buildings

-------
                                                        IX-1
           Radon Measurements in 130 Schools Across the US
                         R.  Thomas Peake
                          Anita Schmidt
               US  Environmental  Protection Agency
                             ABSTRACT
     During  the winter  of 1989,  screening Rn measurements were
made in 130 schools geographically dispersed across the U.S.  The
purpose  of  the study   (Phase  I)   was  to  identify  a subset  of
schools  suitable for year long follow-up  study  (Phase  II),  the
results  of which will  be used to udpdate  EPA's  guidance  for Rn
testing  in schools.   The 130  schools  were  selected nonrandomly
using  school  characteristics  and accessibility  in  areas  where
there were known or suspected  radon problems in  homes.   Because
of  this  selection,  it   is postulated that  levels found  in this
study  may represent   an upper  boundary  for  screening  radon
measurements in US schools.

     The  findings  from  this  study confirm previous  findings (1)
that Rn  concentrations  can vary significantly from  room to room
within a school.  The average Rn level  for the 130 schools is 3.7
pCi/L  (geometric mean  of 1.4  pCi/L) ;  over  half of  the schools
tested had at  least one screening Rn measurement above  4 pCi/L.
The study  also indicates  that  schools  in the same  area  can have
similar or significantly different radon concentrations.

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

-------
                                                                IX-2
            RADON DIAGNOSTICS  AND MITIGATION IN TWO PUBLIC SCHOOLS
                            IN NASHVILLE, TENNESSEE

               by:  Alfred B. Craig, Kelly V.  Leovic,
                    and D. Bruce Harris
                    U.S. Environmental Protection Agency
                    Air and Energy Engineering Research Laboratory
                    Research Triangle Park, North Carolina  27711

                    Bobby E. Pyle
                    Southern Research Institute
                    Birmingham, Alabama  35255-5305
                                   ABSTRACT

     Diagnostic measurements  and mitigation studies  were  carried out  in two
schools in Nashville, Tennessee, as part of the Environmental Protection Agency's
(EPA's) School Radon Mitigation Development/Demonstration Program.  Diagnostic
studies included architectural plans  and building examination,  sub-slab radon
concentrations, sub-slab  communications  measurements, and detailed  classroom
radon measurements using  2-day charcoal  canisters,  electret  ion chambers, and
continuous monitors.   Although  sub-slab communications varied significantly
between the two schools, both were amenable to mitigation using sub-slab suction.
Average premitigation levels of 39.5 and  29.7 picocuries per liter (pCi/L)  were
reduced to 0.78 and 1.7 pCi/L in the two schools.

     This paper has  been reviewed in accordance with the U.S. Environmental
Protection Agency's peer  and administrative review policies and approved for
presentation and publication.
 See  conversion factors listed at end of paper.

-------
                                  INTRODUCTION

      In February 1989, the Environmental Protection Agency  (EPA),  in coopera-
 tion with  state and  local officials,  measured  radon  levels  in over  3,000
 classrooms  in 130 schools located in 16 states spread across the United States
 using 2-day charcoal canisters.  The  purpose of these tests was  to  select  20
 schools  for detailed  testing  over  a  1-year period  to  improve  interim  radon
 measurement protocols  for schools.  Two of the schools in Nashville, Tennessee,
 with very high  levels of radon  in  most of  the  classrooms  were  selected  for
 inclusion in  EPA's School Radon Mitigation  Development/Demonstration Program.
 These were  the Two Rivers Middle  School and Glenview Elementary  School, both
 located  in  a  high radon-risk area of Nashville where numerous  outcroppings  of
 Mississippian-Devonian black shale,  commonly referred to as  Chattanooga shale,
 occur.  This shale deposit runs through the middle  of the state of Tennessee  and
 is  the source of high radon  levels in houses  in the central  Tennessee  area,
 including parts of Nashville.   EPA has also been studying  the mitigation of radon
 in existing houses  in  Nashville over the  past 3 years.

      Results  of  diagnostic and mitigation techniques  used in  the  two  schools
 are  discussed  separately  and compared.

                           TWO RIVERS  MIDDLE SCHOOL

      Figure 1  is a floor plan of the  first  floor of Two Rivers  Middle School
with dates of construction of each area and classroom numbers.  This school was
           TWO RIVERS MIDDLE SCHOOL
  \9M
uDoniw
                          Figure 1.  Floor plan and eonccrueclon ditci

-------
built as a high school between 1960 and 1963 in four phases.  The school contains
19 slab-on-grade  classrooms  and 2 classrooms located over a crawlspace on the
first floor, and 2 basement classrooms.  In addition to these classrooms,  there
are two gymnasiums with boys' and girls' locker rooms,  cafeteria, kitchen, and
three rooms for music and shop.   Part of the locker rooms and gymnasium are over
cravlspace  and the rest is slab-on-grade.   The  two classroom wings  contain a
second story with 23  classrooms and a library. Mitigation studies at this school
were limited  to the  two,  two-story classroom wings and the two basement  rooms
under the end of  one classroom wing.

     Thirty classrooms  were initially tested over  the weekend of  February 4,
1989.   The average  of  these tests was 39.5  picocuries per liter  (pCi/L) and
ranged from 1.5 to 136.2 pCi/L.  Only three rooms  were below 4 pCi/L.   These
measurements  are  shown on the floor plan in Figure  2.   The  three highest  rooms
were tested 2 weeks  later to verify these high readings.   One  room,  initially
measured at 136.2 pCi/L, increased to 148.8 pCi/L, and radon levels in the  other
two rooms decreased  to about 50 pCi/L.  The average decrease of the three  rooms
was 18.5  percent.   Retest data are given  in  parentheses  in Figure  2.  Testing
was carried out in accordance with Report  EPA-520/1-89-010, "Radon Measurement
in Schools, An Interim Report."(1)
                                                   TWO RIVERS MIDDLE SCHOOL
             DOTE-  IKMIM. TCS1S MIX IN {»Rl»
                      1989  MOUSES IK
                 PARENTHESES U( IEPEMS WK
                 III UUC FEBRUABT
                                Baden •••mr«B«nct. February 1969. pCl/L
     Because  of the  very high  readings,  school maintenance  personnel sealed
floor  cracks  and openings in the 10 classrooms  with the highest readings, all
but  one of which  were above 50 pCi/L.   These  10  rooms were  reduced from an
average of 81.7 pCi/L  in  the  initial  test to  29.0 pCi/L,  a  64.5 percent
reduction.   However,  the  readings  in all  10 rooms were still  above 20 pCi/L
after  sealing,  the highest level being 41.7 pCi/L.

     Prior  to initiating diagnostic studies in early June, the building was re-
tested over a weekend under the same  test conditions as the first set.  The 26
rooms  had an  average  level of 10.1 pCi/L and ranged from 1.9  to  32.4 pCi/L, with
6  of  the  26  rooms  being below 4 pCi/L.    This  decrease  could  be partially
attributed  to the  sealing carried out by the school maintenance personnel, but
it was also probably  the result of the difference in summer and winter readings.
However,  the  levels were  sufficiently  high   to cause  a  health  concern;
consequently, diagnostic measurements were made and mitigation carried out.

-------
DETAILED DIAGNOSTIC STUDIES

Inspection of Architectural Drawings and Building

     Examination of the foundation and wall section plans disclosed that the two
hall walls in the slab-on-grade areas are load bearing, resting on a thickened
slab footing,  with aggregate  shown extending under  the footing.   The  walls
between classrooms are not load-bearing and do not extend through the  slab.  The
addition of the four classrooms at the north end of the building resulted in what
appeared to be a solid wall under the slab between the last four rooms and the
initial part of the building.

     Eight classrooms in the south wing are slab-on-grade and the southernmost
part is a four classroom addition built partly over crawlspace and partly over
basement.  This is shown in Figure 1.  The drawings show that aggregate had been
used under the slab throughout the building, including the basement.

     The building  has hot water  radiant  heat  along the outside  walls  in all
areas except  the  basement.   (These  two  basement rooms  had  two separate HVAC
systems.)   The portion of the building heated  with  hot water  has  no active
ventilation except infiltration.  The original windows have been replaced with
a  combination of  solid  panels  and weather stripped  windows,  reducing  the
potential for ventilation through infiltration.   A blower test  run on a segment
of four rooms  indicated a very  tight building.   The entire  building is air-
conditioned by window units installed in  one  of the solid panels  in each of the
classrooms.  All of the fresh air supplies for the air-conditioning units were
closed for energy conservation.

Measurement of Sub-slab Radon Levels

     Sub-slab radon levels were measured  by  "sniffing"  through a 1/4 in. hole
in the center of each room using a Pylon AB-5 continuous  monitor.  Levels under
the original building ranged  from 2600 to 5700 pCi/L.   The four-room addition
on the north end of the building  ranged  from 1300 to 1700  pCi/L.  The portion
of the  south wing of the building which was slab-on-grade ranged from 200 to 2200
pCi/L.   The levels under  the  cafeteria were  4100 pCi/L.   Levels under the two
basement rooms at the south end of  the building  were  500 and 800 pCi/L.   Sub-
slab radon levels for individual rooms are shown in Figure 3.
  •»««*»!
            TWO RIVERS MIDDLE SCHOOL
                         Figure 3.  Sub-*Ub radon
                                                   net. pCIA

-------
Sub-slab Communication Testing

     Sub-slab  communication was measured using  an industrial vacuum cleaner.
The hose of the vacuum was  inserted through a 1-1/2 in.  hole drilled through the
slab in the closet in Classroom 104  (north wing).  The speed of the industrial
vacuum cleaner was varied  to draw vacuum in the hose  of 2, 4, and 6 in. V.C.
Pressure difference between the sub-slab and the room was  then measured at the
center of each room (as a minimum) using a micromanometer sensitive to less than
0.001 in. V.C.  The measurement was made in the hole (in the middle of the room)
which had been drilled for radon sub-slab profile measurements.  The suction hose
of the vacuum cleaner and the hose from the micromanometer were  carefully sealed
at the floor surface using rope  caulking.   Negative sub-slab pressures shown in
Figure 4 were  measured at 4 in. V.C. on the vacuum cleaner line.

     Excellent sub-slab depressurization was measured from Classrooms 101 to 106.
Depressurization under the  rooms on the east side of  the  hall appeared to be as
good as  those on the west side of  the  hall,  showing that pulling under the
thickened slab sections  under the  two hall walls did not  cause any significant
loss of pressure confirming that  there was aggregate under the thickened slab
as indicated on the plans.   The  negative pressure found in the  slab under Rooms
107 through 110 was surprising:  it was not anticipated  that these rooms would
communicate with the rest  of the  rooms in the  wing because they were added at
a later date,  leaving the  north wall of the building  in place.    (These brick
walls were left exposed  in Rooms  107 and 108.)  However, when the wall at the
end of the initial hall  was broken out to extend  the hall, the foundation was
apparently broken below  the aggregate level;  the aggregate continues down the
hall.  Consequently, it  was possible for the suction to  reach  these other four
rooms.  The center of Room 109 was approximately 120  ft from the suction point.

     Sub-slab  communication was measured in the south wing by placing a suction
point  in the  office in the south  west  corner of  Classroom 121.   Negative
pressures under the slab in these  rooms are also shown on Figure  4.  Note that
the suction field extension is not nearly as great in this wing as in the north
             TWO RIVERS MIDDLE SCHOOL
                           Figure 4.  N«g*clv« cub-slab pr«nur««, In. V.C.

-------
wing;  apparently sub-slab communication  is  not as  good.   Although  Room 126
(farthest from the suction hole) did not show a negative pressure under the slab
with  the  vacuum cleaner test, it  was anticipated that the  installation  of a
suction point where the suction hole was would  reach Room 126.   The  larger
suction hole  and fan of the permanent system  normally give  a greater suction
field extension  than is typically seen with the vacuum cleaner test.

     Sub-slab communication was measured in the basement rooms (containing the
art and TV  rooms)  at the south end of the building.   A suction point was made
in one corner of the art room, and it was possible  to measure negative pressures
in the middle of the  room.  However, there was no  indication of any negative
pressure under the slab of the TV room.

MITIGATION DESIGN AND INSTALLATION

     The degree  of sub-slab communication in  the north wing was much greater
than had been found in any other school  tested  in EPA's School Radon Mitigation
Development/Demonstration Program.   Sub-slab  depressurization in the original
building was measured at 0.05 in.  V.C. as far as 60 ft from the suction point.
Even  in  the  four  rooms  of  the  addition  at the  north  end  of the  wing,
depressurization was readily achieved from the one suction point in  the original
building.  The center of the farthest classroom (109)  in the north addition was
110 ft from the suction point and still  showed  a negative pressure of 0.003 in.
V.C. at a vacuum suction of 6 in.  V.C.,  in spite of the intervening sub-slab
wall.  Consequently, it was  anticipated  that  one  suction point in the storage
room of Classroom  104  would be satisfactory for the  entire  north wing (about
15,000 sq ft).

     Although the  communication  in the south  wing  was not as good  as  in the
north wing, it was  adequate  to expect mitigation  in all  the rooms (except the
basement rooms and possibly the rooms over the crawlspace) with a single suction
point located in the office of Classroom 121.

     Two  temporary mitigation  systems  were   installed  using 6-in.  diameter
flexible PVC drain pipes exiting through the classroom windows and attached to
individual fans.  The turbo fan had a rating of 410 cfm at 1  in. V.C.  Vith the
temporary system operating in the north wing, all rooms measured below 4 pCi/L
during weekend charcoal canister measurements except for the classroom closest
to the  exhaust  line  through the  exterior of the  building.  This classroom
measured 5.3  pCi/L; it  was suspected that the higher  level  was  the result of
reentrainment.  All the rooms in the  south wing were below 4 pCi/L except those
over the crawlspace and basement and the basement rooms.  (These tests were made
before the  basement sub-slab system was  installed  in the TV  and art rooms.)
Sub-slab communication was again measured in  all  the rooms  with the temporary
systems in place  and later with the permanent systems in operation.  As expected,
the negative  sub-slab  pressure,  achieved with the  temporary  and permanent
systems, was greater than had been  achieved with the vacuum cleaner as shown by
comparing the negative  sub-slab pressures in  Figures  5  (permanent systems in
operation) and 4 (vacuum cleaner tests).  The center of the classroom at the far
end of the north wing was 112 ft from the  suction point and had a negative sub-
slab pressure of 0.007 in. V.C. with the permanent system in operation.

-------
            TWO RIVERS MIDDLE SCHOOL

              SUCTION rams
                                                 Kf
                                                 il^M     ^SUCTION nilT -1.91
                                            M-.l? ,
                                    :TIM POINT • 1.75
                 Figure 5.  Post-Bltlgetlon fugitive tub-slab preuures. In  U C.
     In view of these findings,  a decision was made  to convert the temporary
systems into permanent systems.  In both  cases  the pipes were run overhead  to
the rear of the building and up the outside of the building with the fan located
under the overhang and the exhaust extended over the top of the roof.  There are
no air intakes within 50 ft of these exhaust points.  (Note that Tennessee State
codes require  any exhaust to be a minimum of 10 ft from any fresh-air  intake.)

     The two basement  rooms had much poorer communication  than the first floor
slab-on-grade  wings.   It appeared that there was a subslab barrier between the
TV room and  the  art room; consequently, a 4-in.  suction point  was installed  in
each of the  two rooms  manifolded to a  common exhaust line run out the  back  of
the building  and  to  the  roof using  the  same  configuration  as  in  the  other
systems.    With this  third system  in  operation,  pressure  field  extension
measurements under the slab  of  the  two  basement  rooms  showed  that adequate sub-
slab depressurization  had been achieved as shown in Figure 5.

MITIGATION RESULTS

     Installation of the sub-slab system in the two  basement rooms was completed
during the last  week of  July.  All rooms in the two classroom  wings, including
the basement, were  tested  over  the   first  weekend  of August  with  charcoal
canisters under  closed building conditions.   Results of these tests are shown
in Figure 6.  The  average level of all measurements in August, with the three
mitigation systems operating, was 0.78 pCi/L, ranging from 0.5 to 1.3  pCi/L.
            TWO RIVERS MIDDLE SCHOOL
                    Figure 6.  Posc-Blclgeclon redan levels. Auguet 1989, pCl/L

-------
     Pressure-actuated switches  served  the three permanent systems:   a green
light indicated when negative  sub-slab  pressures  were being maintained, and a
red  light  shoved when the  system was  inoperable.   These warning  lights  are
checked daily by the custodial staff.

     All the rooms were  retested during cold weather in December.   Results of
mid-winter tests will be reported at the Symposium.

                          GLENVIEV ELEMENTARY SCHOOL

     Glenview is a  21-classroom elementary school.   The original part of the
building   was   constructed  in   1954   and  included   13  classrooms,   an
auditorium/cafeteria, a kitchen,  and administrative offices. The first addition
was built  in 1957 and consisted of four  classrooms  at the south  end of the
original building.   Both the original structure and the  1957 addition are slab-
on- grade; the addition is about  4 ft lower than  the  original building because
of the topography of the site.   The  second addition, built in 1964, consisted
of a four-classroom separate structure  on  the northwest side of the building.
This addition is a raised slab over a crawlspace.   Hitigation studies during the
summer of 1989 were limited to the slab-on-grade portion of the building.  The
crawlspace wing will be mitigated during the 1989-90  school session.  Figure 7
is a floor plan showing dates of construction and classroom numbers.
                                   GLENVIEW ELEMENTARY SCHOOL
                     Figure 7  Floor plan and ceiutrucclon dat»

     All classrooms  and the two administrative offices  were initially tested
over the weekend  of  February 4, 1989.  The average  level  of the 22 locations
tested was 29.7 pCi/L,  ranging from 8.9 to 52.5  pCi/L.   Individual classroom
levels are shown in Figure  8.  The four rooms with  highest radon levels (average
of 44.5 pCi/L,  ranging  from 40.1 to 52.5 pCi/L)  were  retested the weekend of
February 18.   Retests of these four rooms gave an average level of 37.5 pCi/L,
ranging from 30.0 to 42.6 pCi/L.  This was  less of a  decrease than found in the
Two Rivers School during the sane period, indicating  that seasonal variation is
school-specific.  All the rooms were retested in late June and again in July.

-------
NOfE:  INITIAL TESTS MADE IN EARlf
     FEBRUARY 1M9  FIGURES IN
     PARENTHESES ARE REPEATS MADE
     IN LATE FEBRUART.
                                         GLENVIEW ELEMENTARY SCHOOL
                                           28.4   12.21

                   Flgur* 8. Radon ocaiureaent. February 1989. pCl/L
Levels  in both  tests were  approximately 25 percent  lower than  they were  in
February.  The room that measured highest in February decreased from 52.5 to 13.4
pCi/L in June and to A.A pCi/L in July.  The next highest room, which was 44.8
pCi/L in February,  increased to 93.4 pCi/L in June and to 108.9 pCi/L in  July.

DETAILED DIAGNOSTIC TESTING

Inspection of Architectural Drawings and Building

      Examination of the foundation and wall section plans disclosed  that  block
walls on all  four sides  of  all  rooms and closets extended to footings under the
slab with no indication of breaks in these sub-slab walls,  reducing the potential
for sub-slab  communication  between rooms.  The wall section drawings  showed the
presence of 4 in. of aggregate of unspecified size under the slab.

      Each  room is heated by a fan coil unit mounted above the dropped ceiling.
The heated air from the  fan coil is ducted to the ceiling registers  near  the
outside wall.  Cold air  return  is through an unducted opening very close to the
air intake of the  fan coil.   Consequently,  it  is unlikely that the  fan coil
causes much of a negative pressure in the plenum above the ceiling.   The  rooms
are cooled in the wanner months  by  individual window-mounted air-conditioners
in each room.  The  fresh air intakes for these have all been closed  off.   Five
wind-driven turbine roof ventilators exhaust air from the building.   These  are
located in  the hallways  in  the  plenum above the suspended ceiling.  Host of the
classrooms  in the original  structure also have small ventilator fans  mounted in
the hall wall and exhausting into the hall.  It  is not certain how  these fans
are operated,  if at all.  There are  two large kitchen exhaust fans,  an exhaust
hood over the cooking unit, and an exhaust fan for the dishwasher.

     The foundation drawings of the 1957 addition, containing four slab-on-grade
rooms, indicated that not  all  the walls go  through  to footings;  there was a
possibility of good sub-slab communication between these four rooms.   This was
confirmed by  communication  testing reported later in this paper.

-------
Measurement of Sub-slab Radon Levels

     Sub-slab radon levels were measured  through  a 1/4-in.  hole in the center
of each room using a Pylon AB-5 continuous monitor.  Sub-slab radon levels were
between 300 and 1900 pCi/L as shown in Figure  9, much lower than those found at
the Two Rivers School.  There appears to  be no correlation between the levels
in the room and the sub-slab levels.
                                        OLENVIEW ELEMENTARY SCHOOL

                                                           To!
                                                                   800
j8ft f IT   IO'J
|   fl.LJLJ 400  |
                                                               '"
                                                             400    l?00
                           Figure 9.  Sub-sUb radon Icvcli. pCl/L

Sub-slab Communication Testing

     Sub-slab communication  test measurements were carried out using the test
method described previously for the Two Rivers School.  The initial suction hole
was placed Just inside Room 110,  near the door.   It was  possible  to pull a
negative pressure  in Classrooms 109 and 111,  and  in the  hall just outside of
Classroom  110.  However, no communication could be  detected with Classroom 120,
the room across the hall.   Similar results were obtained at the other end of the
building.  From these tests,  it was concluded  that  sub-slab communication could
be extended about 30 ft in the original building and could pull through one sub-
slab wall, but not two.  Even within the same room,  pressure-field  extension was
much poorer than  found at  Two  Rivers.   Consequently,  it was felt that the
aggregate  was  not as open as  it was at Two  Rivers.   This was confirmed when
aggregate  from  the  two  slabs was removed during the  installation  of mitigation
systems.    At  Two Rivers,  the stone was  very large, from 3/4   to  1 in.  in
diameter.  At Glenview, it was much finer,  averaging about 1/4  in. in diameter
and contained some  fines  and dirt.  Both were screened river  gravel.

     A suction  point  in Classroom 114 of the  four-room addition indicated good
sub-slab  depressurization  in  Classrooms   112  to  115 as  expected  from the
examination of  the  foundation drawing.

MITIGATION DESIGN AND INSTALLATION

     In view of the relatively poor sub-slab communication, it was decided that
a minimum  of one  suction point in every other room would be  necessary on both
sides of  the  hall.   Three systems were laid out  as  shown in Figure 10.  Two
multi-point systems had 6 in. trunk  lines in  the hall plenums with 4  in.

-------
                                        GLENVIEW ELEMENTARY SCHOOL
        fVCTIO* ft
                         Figur. 10. Mitigation lyttem layout
drops to  suction points in the rooms.  Trunk  lines  were run over the dropped
ceiling of the hall and 4-in. drops were placed as indicated in Figure 10.  In
the trunk lines,  T's  were  installed and capped off to facilitate the addition
of suction points to  all of the remaining rooms if necessary.  One mitigation
system had five  suction points and the other  had  six,  as shown in Figure 10.
Each used a turbo fan rated at 410 cfn at 1 in. W.C.

     Communication testing in the 1957 four-room slab-on-grade addition at the
south end of the  building indicated that one suction point would likely mitigate
all four rooms.  A 4-in. suction point was put  in the corner of ROOD 114 behind
the door and the  pipe was  run overhead through the outside wall at the rear of
the building.  Suction was achieved using a  turbo fan rated at  210 cfm at  1 in.
W.C.
     Pressure  activated visual alarms,  as described  for  Two Rivers,  were
installed for each mitigation  system.   These were  put on the mitigation  trunk
lines and are checked daily by the custodial force.

MITIGATION RESULTS

     Mitigation  installation was completed  the first week of August,  and all
systems were put  in continuous operation.  The  following weekend all rooms were
tested with charcoal canisters under closed conditions with all air handlers off.
Average radon levels in all rooms were 1.7 pCi/L, ranging from 1.2 to 2.9 pCi/L.
Post-mitigation radon levels in  each room are  shown  in Figure  11.

     Glenview Elementary School was also retested in December, and the average
levels during cold weather will be reported at the Symposium.

     As stated above,  the  1964  four-classroom  addition is built over a crawl-
space and will be mitigated in early 1990.   Suction under polyethylene sheeting
in the crawlspace will be compared  to  pressurization and depressurization of
the crawlspace.  This will be EPA's  first  detailed  study of mitigating a school
building over a crawlspace.  This addition is  ideal  for  these  studies since it
is small with adequate headroom  and contains no asbestos.

-------
1.
                                    121
                                   1.1
                                   GLENVIEW ELEMENTARY SCHOOL
                                             121
                                             MDITUUIM'
                                             CAftHBIA
                                               1.«
        uri    112'
       1.6  I   2.1
                             nT"~" n«i   i«!»-i;Tf?T"   toil   iu|    l
                          ,.."|,.. 1  ..,  j*j.uj;?f TryTi
                                          "                  ^^^^^^^^
             Figur* 11. Po«t.nltig«tlon radon level*. August 1989. pCl/L

                            CONCLUSIONS

The following conclusions are drawn from these studies.

1.   The two schools studied had  widely  different sub-slab communication
     resulting  from different  sub-slab  construction details  and  from
     differences  in aggregate.    The  poorer  sub-slab  communication  in
     Glenview was  overcome by increasing  the number of suction  points.
     The number used was based on the communication determined using vacuum
     cleaner communication testing.

2.   Two Rivers Middle  School  was built  using  a  thickened slab  under
     interior load-bearing walls but contained no interior  sub-slab block
     walls on separate footings.   This resulted in a continuous  aggregate
     layer under each classroom wing.   As a result,  one suction  point was
     capable of depressurizing  the sub-slab area of  an  entire  classroom
     wing (IS rooms) of over 15,000 sq ft.  Negative sub-slab pressure of
     0.004 in. V.C. was measured as far as 120 ft from the  suction point.
     Average radon  levels  in  the mitigated portion of the building were
     reduced from 39.5 to 0.78 pCi/L.

3.   All interior walls  in Glenview Elementary School extended through the
     slab to footings.   This resulted in a compartmentalized sub-slab area
     equivalent to the room configuration of the school.  In addition, the
     aggregate had a smaller average particle size  and  contained more fines
     and dirt.  As  a result of these  two factors, sub-slab communication
     was found  to  be much poorer  than  at Two Rivers.    Pressure  field
     extension  was  limited  to  a  maximum of  30  ft.    Mitigation  was
     accomplished with three suction systems containing 12 sub-slab suction
     points or  an average of  about 1 suction  point  for  every 2 rooms.
     Average radon levels were reduced from 29.7 to 1.7 pCi/L.

                             REFERENCES

United States Environmental Protection Agency,  Office of Radon  Programs,
•Radon Measurements in  Schools -  An  Interim Report.*   Washington.  D.C.
20460, EPA-520/1-89-010, NTIS PB89-189-419AS, March 1989.

-------
                              CONVERSION FACTORS



1 pieoeurie/liter (pCi/L) - 37 becquerels/cubic meter



1 inch (in.) - 2.4 centimeters



1 inch (in.) water column (U.C.) - 249 pascals



1 foot (ft) - 300 centimeters



1 square foot (sq ft) - 929 square centimeters



1 cubic foot/minute (cfm) - 472 cubic centimeters/second

-------
                                                                        IX-3
                THE EFFECTS OF HVAC SYSTEM DESIGN  AND  OPERATION
                     ON RADON ENTRY INTO SCHOOL BUILDINGS

                    by:  William A. Turner
                         Harriman Associates
                         Auburn, ME  04210

                         Kelly W. Leovic and Alfred B. Craig
                         U.S. EPA, AEERL
                         Research Triangle Park, NC  27711
                                   ABSTRACT

     Heating, ventilating, and air conditioning (HVAC) systems in schools vary
considerably and tend to have a greater  impact on pressure differentials - - and
consequently radon  levels  -- than do heating and  aiu-conditioning  systems in
houses.  If the HVAC system induces  a negative pressure relative to the subslab
area, radon can be pulled into  the building.  If  the HVAC system pressurizes the
building, it can prevent radon entry as long as the  fan  is  running.   However,
school HVAC systems  are  normally set back or turned off on evenings and weekends
and, even  if the  HVAC system pressurizes  the school  during operation,  indoor
radon levels may build up during setback periods.

     Many of the historical methods utilized to deliver ventilation air (outdoor
air) over the past 40 years are summarized.  In addition,  for  each type of system
presented, the possible  impact the ventilation system  might be expected to have
(positive or negative) on the pressure of the building envelope (and subsequent
radon levels in the building) is discussed.

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

-------
                                 INTRODUCTION

THE EFFECTS OF VENTILATION ON BUILDING PRESSURE DIFFERENTIALS

     The primary mode of  radon entry into a school is normally  from soil gas
that is  drawn in by pressure  differentials  between the soil  surrounding the
substructure and the building interior.   If the building interior is at a lower
pressure than  the  soil  surrounding  the  substructure,  radon can be  pulled in
through cracks and openings  that are  in  contact with  the  soil.   The  amount of
radon in a given classroom will depend on the level of radon in the underlying
material, the  ease with which  the  radon moves as a component  of  the soil gas
through the soil, the magnitude and direction of the pressure differentials, the
number and size of the radon entry routes, and dilution and mixing of the room
air.

     Pressure  differentials  that contribute  to  radon  entry can  result  from
operation of a heating, ventilating,  and air conditioning  (HVAC)  system under
conditions that cause negative pressures (in the building relative  to the subslab
area),  indoor/outdoor temperature differences  (including  the "stack  effect"),
use of appliances or other mechanical devices  that depressurize  the  building,
and wind.

     HVAC systems in  schools and other  large buildings vary considerably and
tend to have a greater impact on radon levels than do heating and
air-conditioning systems  in  houses^ ?  The design,  installation,  and  operation
of  the  ventilation  equipment  will  cause the building  envelope  to be  at  a
positive, negative,  or neutral  pressure with  respect to the outdoors,  depending
on the system  design, how it was initially  installed  and  balanced, and how it
has been historically maintained and operated.   Sometimes  schools and similar
buildings were not designed with adequate ventilation, and in other instances,
ventilation systems are not operated properly due to factors such as  increased
energy costs  or uncomfortable  conditions caused by  a design or  maintenance
problem.   Such circumstances may enhance radon entry into the building.

     If the HVAC system  induces a negative pressure in the building relative to
the subslab area, radon can be pulled into the building through floor and wall
cracks or other  openings  in  contact with the soil.   (Even  if  the HVAC system
does not  contribute  to pressure differentials in the building, the natural stack
effect in a leaky building can cause  the building to be under  a negative pressure
so that radon-containing  soil gas is pulled into the school.)

     If the HVAC system  pressurizes  the building  -- which is  a common design in
many systems  --  it  can  prevent radon entry as  long  as  the  fan  is running.
However,  school HVAC systems  are normally set back or  turned  off during evenings
and  weekends,  and  even  if  the  HVAC system  pressurizes  the school  during
operation,  indoor radon levels may  build up  during setback  periods.   Once the
HVAC system is turned back on,  it may take several hours for  radon  levels to be
reduced.   Consequently, how the ventilation air (outdoor air) is supplied Co the
building (i.e. , whether the ventilation system is pressurizing or depressurizing
the  building)  can be expected to  drastically affect the  radon  levels  in  a
building when  the  ventilation system  is  operating.   Even when  a  building is
operated  overall slightly positive with respect  to the outdoors,  localized

-------
negative pressures may exist.   If openings to the earth are present where these
localized negative pressures exist, soil gases will be drawn in.

BUILDING VENTILATION HISTORY

     Buildings designed  for human occupancy (in particular,  public  buildings
such as  schools) have historically  been required  to  be designed to  provide
ventilation air  (outdoor air) to  the occupants.    This  outdoor air has  been
provided historically by non-powered  design features  such as operable  windows
and gravity ventilators (which allow air movement created by wind and temperature
differences),  more recently by powered ventilating equipment,  or some combination
of both.   In many  areas  of the country where air conditioning has not  been
utilized,  it  has  been popular  to  provide  a  base level  of ventilation  by
mechanical systems  and to allow  supplemental ventilation to occur  through the
use of large operable windows as weather conditions allow.

     Historically, the introduction of fresh outdoor air into buildings has been
relied upon  to  dilute the  contaminants  which  are generated  by  the  occupants
within  the building  and to  provide  free  (economizer)  cooling when  weather
conditions permit.   The  American Society of  Heating, Refrigerating  and Air
Conditioning Engineers (ASHRAE) standards documents have provided, and continue
to provide, engineering professionals with guidelines to the suggested minimum
quantities of outdoor air which should be provided  to the occupants  .  Building
Codes (laws) also govern the amounts  of outdoor air  to be provided; often these
codes refer to the  ASHRAE guidelines.   The  most up-to-date ASHRAE ventilation
standard, ASHRAE 62-1989, prescribes  that 15 cfm (1 cfm - 0.47 L/s) of outdoor
air be supplied  to  a classroom for acceptable indoor air quality ^  I

MULTIPLE ZONING OF  VENTILATION AIR

     In  large buildings  with multiple exhaust fans,  supply  air  systems,  and
system  balancing dampers,  some  section of  the building will  frequently  be
designed to be  operated  negative and other sections  positive with respect to
adjoining  areas  in order to  minimize the  spread  of  an  internally  generated
irritant or odor.  Thus,  with regard to radon entry  from  the soil, the expected
and measured  overall  pressure balance of the  total building envelope  and the
expected and measured pressure relationship  in  individual  areas of the building
must be considered.

     Typical areas which might be expected to be designed and operated negative
with respect to  adjoining areas  include any area where identifiable sources of
pollutants may be  generated;  e.g.,  toilets,  locker rooms,  shops,  print rooms,
art areas,  laboratories,  kitchens,  gymnasiums,  hallways,  lounges, and janitor's
closets. Areas which might be expected to be designed and  operated positive with
respect  to adjoining  areas  include classrooms,  computer  rooms,  and libraries
Thus, it is important  to  know the expected  and measured pressure relationship
of individual zones within the envelope as well as the overall building envelope.

     In  addition to affecting the  pressure  relationships,  the ventilation air
(outdoor air) will also  be  available to dilute radon  gas  once  it  has  entered
the building.  The dilution  effect of outdoor air (ventilation air) is primarily
a control  strategy for other  pollutants  (bioeffluents) generated primarily by

-------
the occupants; however,  dilution alone is seldom adequate  to reduce elevated
levels of radon without the proper pressure relationship.

     The  following  section presents  an  overview  of  the  various  types  of
ventilation systems which might be found in school buildings and the potential
(positive or negative) impact each type of system would be expected to have with
regard  to radon  entry  (i.e.,  the  expected  overall impact on the  pressure
relationship between  the building envelope and the soil).

                            TYPES OF  HVAC  SYSTEMS

     Many of the HVAC systems discussed below have the option of being designed
to supply a  fixed or variable  amount of outdoor air.   In addition,  the total
supply air moved  by an air handling  system (i.e.,  the  combination  of outdoor
and recirculated  air) may  be  fixed  (constant volume)  or modulated  (variable
volume).  A variable  air volume  (VAV) system  is  typically designed  to deliver
more total supply air as additional  cooling is called for.  With a VAV system,
the amount  of  outdoor air delivered may  also  be  designed  to  be fixed  or
modulated.

EXHAUST-ONLY SYSTEMS

     One  of  the  basic  systems  is  an exhaust-only  system;  i.e.,  the  system
consists of  exhaust fans (often installed in hallways,  bathrooms,  and locker
areas).  Building leakage or the opening of windows is typically the source of
outdoor makeup air.   Even more basic systems include gravity ventilators (non-
powered exhaust shafts dependent upon the  building  stack effect,  and operable
windows).  Such systems  that do  not supply tempered makeup air typically lead
to stuffy conditions  in  the winter  time, when occupants  are hesitant  to open
windows due to cold drafts.

     Exhaust-only systems would be expected to cause  the  overall pressure within
the building to be  negative with respect to  the  outdoors, thus increasing the
flow of soil  gas into  the building envelope.  Depending on the  degree of building
depressurization, and the location and size of the envelope leakage,  the radon
levels in a building should  increase during operation of  an exhaust-only system.

RADIANT HEAT SYSTEMS

     Radiant heat systems in schools  tend  to be  of  three types:   hot water or
steam radiators,  baseboard heaters,  or warm water radiant  heat  within the slab.
Schools heated with radiant systems should  have a ventilation system to achieve
the fresh air requirements recommended by ASHRAE;  however, many of these schools
provide no ventilation other than natural infiltration.   In other schools, there
are exhaust  ventilators on the  roof.   These can  be  passive, allowing some
ventilation  through the  stack  effect, or  they can  be powered.   Powered roof
ventilators (PRVs) can cause significant building depressurization, particularly
if a fresh air  supply is  not provided.  This can cause considerable radon entry
into the building while such exhaust systems are  operating.

-------
UNIT VENTILATORS

     The use of unit ventilators  in  schools has been and continues  to be very
popular.  They are available in a number of different  arrangements:  horizontal,
vertical, draw-through,  and  blow-through; and are made  by a wide  variety of
manufacturers.

     In a typical unit ventilator system,  by  design, there is a connection to
the outdoors,  providing makeup air  for ventilation  and free cooling.   In a
typical unit  ventilator  configuration,  the outdoor  air  mixes with  return air
from the classroom in the plenum portion of the unit  ventilator and is supplied
to the space typically through the top.

     The advantages of this type of system are  often economics and architectural
flexibility: generally no  ductwork is  required.   Some of the disadvantages of
this system are  the noise levels generated  by  the  unit  ventilators  and the
numerous  wall  penetrations  that at  some points downgrade the  architects'
elevation aesthetics.  Also,  a serious  concern is  the draftiness of these types
of units especially in northern climates.  Drafts are of concern because, with
20 to  25  students in a  typical modern  classroom,  coupled with well insulated
walls  and  ceilings  and  1.5-2 W/ft2  (335-446  kW/m2)  of  lighting,  the internal
heat gains often outweigh any envelope losses of the classroom. This can require
fresh air to be introduced for cooling during major portions of the school year.

     Unit ventilator  systems  would be  expected to cause the overall pressure
within the  building to be  positive with respect to the outdoors, thus reducing
the flow of soil  gas into  the building  envelope when the unit is operating and
outdoor air is being  drawn into the  unit,  whether or not the space served by
the unit ventilator is actually  pressurized with respect to the soil will depend
on the degree  of overall building pressurization or depressurization.   If other
areas of the building have exhaust-only systems which exhaust more air than is
made up by  the unit  ventilators,  then  soil gases  will still be drawn into the
building  in areas where the net  pressure  is  negative.   The radon levels in  a
building should decrease when a unit ventilator system  is operating properly,
if adequate overall makeup air  is provided.

TERMINAL AIR  BLENDERS

     Terminal  air blenders have also been used. Initially, this  type  of system
was a good alternative to unit ventilators. There are a number of ways  terminal
air blenders have been  used.   They  were  installed  to  help combat the energy
crunch, while still  delivering outdoor air for cooling  and ventilation.  With
these  systems in the classroom, even in northern climates, over 90 percent of
the time  approximately 5 cfm (0.00236  ft3/s)  of outdoor air per student would
typically  be  introduced.   Because  of  the  high   internal heat gains  in  the
classroom,  the intent was  to thermostatically control the outdoor air  quantity,
bringing  in  the  appropriate amount of outdoor  air required  to  satisfy  the
internally  generated heat load.  These  systems are generally connected to an air
duct system to distribute  the ventilation air evenly, reducing drafts, and  are
less  noisy than  unit  ventilators.    A consistent  outdoor  air  supply is  not
provided; however, typically, 90 percent of the time  more than 5 cfm per student

-------
outdoor air is provided with a  thermostatically controlled terminal air blender
ventilation system that is functioning properly.

     Terminal air blender systems would be expected to cause the overall pressure
within the building to be positive with respect to the outdoors, thus reducing
the flow of soil gas into the building envelope when the unit is operating and
outdoor air  is being  drawn  into the  unit  and distributed to  the  occupants.
whether or not  the space  served by the  terminal  air blender is  actually
pressurized with respect to  the soil will  depend on  the  degree  of  overall
building pressurization or  depressurization.   That is,  if other  areas  of the
building have exhaust-only systems which exhaust more air than is made up by the
terminal air blender, then soil gases will still be drawn into  the building in
areas where the resultant pressure  Ls negative.  The radon levels in a building
should decrease  when a terminal air blender system is  operating  properly,  if
adequate overall makeup air is provided.

UNITARY HEAT PUMPS OR FAN-COIL UNITS

     Heat pump units have been  utilized to  a limited  degree  in  schools.   They
appear similar to a fan-coil unit and may or may not have outdoor air ducted to
the unit.   Fan-coil units consist of  a fan and heating and/or cooling coils and
may or may not have outdoor  air ducted to the unit.  (They may just recirculate
air.)

     Unitary heat pumps or fan-coil units would be expected to cause no overall
pressure change  within  the  building  even  when outdoor air has  been ducted to
the unit unless  additional dampers and controls have  been added to  convert it
to function as a  unit ventilator.  If outdoor air has been provided and the unit
converted  to  a unit ventilator, then the  unit would be  expected  to cause  a
positive pressure inside the building with respect to the  outdoors, thus reducing
the flow of soil gas into the building envelope when the unit is operating and
outdoor air  is being  drawn  into the  unit  and distributed to  the  occupants.
Whether or not the  space served by the unit is actually pressurized with respect
to the  soil  will depend on  the degree of overall building  pressurization or
depressurization.  That  is,  if other areas of the building  have  exhaust-only
systems which  exhaust  more  air than  is made up  by the heat pump or fan-coil
units, then soil  gases  will  still be  drawn into the building in areas where the
net pressure is negative.  The radon levels  in a building should decrease only
when  a  fan-coil  or heat  pump  unit  has  been equipped  with  outdoor air,  and
converted  to  a unit  ventilator (if adequate overall makeup  air is  provided).
If the units are not supplied with outdoor  air then  the only impact should be
from the normal natural stack effect of a  leaky building.

HEAT RECOVERY VENTILATORS (HRV)

     In general, HRVs  are  either ducted systems with supply and  return ducts
servicing different parts of the building or room,  or wall-mounted units, similar
to wall-mounted  air-conditioning units.  In both  types  of units,  fresh  air is
brought in through a heat recovery device,  then distributed,  or  passed through
a preheat  coil and  then out to the system's zones.   The  exhausts from these zones
pass through a separate section of the  heat recovery device,  and then discharge
far  enough  from  the  fresh  air  intake  to  minimize  re-entrainment.    One

-------
disadvantage of  these  types  of systems is that condensation on the surface of
the  heat exchanger  may frost  up  and block  the  heat exchanger  when outdoor
temperature drops  below 20°F  [°C=5/9(°F-'32) ] .  To  avoid that problem, bypass
sections  and  defrost  controls  are often  available with  units  as  standard
features.   By temporarily  bypassing the outdoor  air and,   thus,  raising the
exchange  surface temperature  to above the  dewpoint,  frost  on  the exchanger
surface  is avoided.

     If balanced correctly, HRV  systems would be expected to  cause  the overall
pressure within the building to be neutral  or very slightly positive with respect
to the outdoors, thus reducing  the  flow of soil gas  into the building envelope
when the  unit  is operating and  outdoor air is being drawn  into  the  unit and
distributed to the occupants.  Whether or  not the space served by  the device is
actually  pressurized with respect  to  the soil will depend on the degree of
overall building pressurization or  depressurization.   That  is,  if other areas
of the building have exhaust-only  systems which exhaust  more air than is made
up by the HRVs, then soil gases  will be drawn into the building in areas where
the  resultant  pressure is negative.   The  radon  levels  in  a building should
decrease when a HRV system is operating properly, if adequate  overall makeup air
is provided.   One  exception  is exhaust-only heat recovery devices which would
be expected to raise radon  levels  similar to exhaust-only ventilation systems
discussed earlier.

CENTRAL STATION AIR  HANDLERS

There are many types of central  station systems, many with features similar to
those discussed above.   The  common features  of all central units include an air
handling  unit  supply fan and/or return  fan and associated  tempering coils,
dampers and controls, distribution  ductwork,  exhaust (or relief), mixing box,
and outdoor air intake.   In the  past, constant volume systems, which consisted
of central station air handling systems that generally had fixed minimum outdoor
air dampers, were used in schools. Typically the outdoor air would  be controlled
by  a  two-position   damper  closed  and opened  to  whatever   percentages  were
predetermined, to be mixed with  return air,  passed through the supply fan, and
introduced to the occupied space.

     If designed correctly, central station air handler systems would be expected
to cause the overall pressure within the building to be slightly positive with
respect to the outdoors,  thus  reducing  the  flow of  soil  gas  into the  building
envelope when the unit  is  operating  and outdoor air  is being drawn  into the unit
and distributed to the occupants.  However,  in a building with multiple zones,
some spaces served by  the central  system may be adjusted to be  positive with
respect to the  soil,  and other areas may be negative.  Many areas of the building
tray have  exhaust  fans  which exhaust more air than  is made  up by the central
system by  design.    Thus, even  if  the  overall pressure  relationship  for the
building is positive, soil gases will still be drawn  into the building in areas
where the local resultant pressure  is negative.  The  radon levels  in a building
should decrease when  a central station system is balanced to be  slightly positive
and operated properly, if adequate overall  makeup air is provided.  The following
sections present a few of these central station air handler  systems in detail.

-------
Conventional Constant Volume (CV)

     Central stations,  predominantly with constant volume air systems with fixed
minimum outdoor air entry and reheat  coils, have  been  utilized for many years
in a  limited number of schools.   They are somewhat energy efficient  and are
easily balanced.

Variable Air Volume (VAV)

     With the energy crunch, several  manufacturers  of  air handlers introduced
VAV control.  The most immediate savings are fan energy and elimination of reheat
coils (if individual room VAV diffusers are used); however, outdoor air control
is difficult.   In  many VAV  systems,  there is  no way to control outdoor air to
bring the  room  above bare minimum fresh-air  quantities.   Central station VAV
systems, with static pressure devices in the outdoor air stream and addition of
reheat  coils,  were an  answer  to outdoor  air control.   With  static  pressure
control of the outdoor  air stream, it is possible to maintain an  overall positive
pressure within the building under various operating conditions.

VAV with Economizer

     As just noted,  the VAV helps cut down on fan energy  when  not dealing with
peak  cooling loads.   If only minimum air movement  is  needed,  the reduced air
flow will  save  energy  dollars.   Whenever  outdoor  air is critical  (e.g.,  if all
areas must  be kept under positive pressure to  keep radon out),  shutting off the
VAV distribution boxes in individual  spaces is  a  concern.  One disadvantage of
most  VAV  systems   is that they  have no sensing  in the  outdoor air stream  that
would guarantee the correct amount of outdoor air  during part-load operation.
One way to  avoid this situation is to  use  an outdoor air flow sensor.  With this
type  of metering  system,  a drop  in  velocity in the  outdoor  air stream  will
control  the air dampers,  bringing In more air from the  outdoors.   (This  would
be typically called an outdoor  air reset.)

VAV with Outdoor Air Control and Heat Recovery

      This  type  of  package combines efficient operation with temperature  control.
Most  importantly,  the  ability to  deliver   outdoor  air capacity  is  greatly
 increased,  and the facility is not penalized in  terms  of energy  costs,  nearly
 as much as without the heat recovery feature.

      Central station  heat  exchangers are currently being considered  in  many
 schools being designed for northern climates.   In this  type of  design,  a central
 air handling system incorporates a heat exchanger.  Some reheat may be  required
 in such system.

                        HVAC  SYSTEMS AND RADON MITIGATION

      A potential  mitigation approach  for schools is  adjustment of  the  air-
 handling system to maintain a  positive pressure  in the school relative to the
 subslab area, discouraging the  inflow of  radon.   This technique, referred to as
 pressurization, has been shown to be an  effective  temporary  means of  reducing
 radon levels in some  schools,  depending  on the  design of the  HVAC system.  If

-------
pressurization through the  HVAC  system is under consideration as  a  long-term
radon mitigation solution in a given  school,  proper  operation and  maintenance
of the system are critical.  Responses  to changes  in environmental conditions
and any additional maintenance costs  and  energy penalties  associated with the
changes in operation of the HVAC system must also  be carefully considered1 i

     Important factors that  need to be considered when utilizing the HVAC system
for radon control include:  (1) How much outdoor air was the system originally
designed to supply under what design conditions? (2) How leaky is the shell of
the building and  can pressurization be utilized? and (3)  Is  the system currently
operating as designed, or has  it been modified purposely  or  through neglect?
Once  the  limitations  of  the  HVAC system  and building shell  are  determined,
decisions can be  made  on  the best  or most  reasonable course of action which can
be taken.  Some  approaches  to  radon  control through the HVAC  system that have
been used temporarily and permanently are generalized below.

EXHAUST-ONLY VENTILATION AND RADIATION HEAT SYSTEMS

     For schools with either exhaust-only ventilation systems or radiant heat,
positive pressurization will probably require major modifications  if the HVAC
system is considered as part of the mitigation strategy.

UNIT VENTILATORS AND EXHAUST-ONLY SYSTEMS

     Radon mitigation strategies in schools with unit ventilators might include
(1) opening the fresh-air vents (if they have been closed) to improve ventilation
and running  the  unit  ventilator fans continuously  (or  prior  to  occupancy) to
pressurize the room;  (2)  replacing an exhaust-only ventilation  system with a
system that  operates  under  a  slight  positive pressure;  or (3) installation of
a  subslab depressurization (SSD) system that could  overpower all negative
pressures in the building.   If the current HVAC  system is providing adequate
ventilation  to the building or if options  (1) and (2) are not feasible, option
(3), installation of a SSD system,  would be the most practical near-term strategy
if there is  good subslab communication.

CENTRAL AIR  HANDLING  SYSTEMS

     Although most  central HVAC  systems  are commonly  designed  to operate at
positive or  neutral pressures, pressure measurements in schools have indicated
that  such  systems can cause  significant negative pressures  in  the building.
HVAC  system  modifications  (such as reducing  the  amount of fresh-air intake),
lack of maintenance (such as dirty filters), unrepaired damage, or other factors
can result in substantial negative pressures in some rooms,  thus increasing soil
gas  entry.    In addition,  operation  of  localized  exhaust  fans  can   cause
significant  negative  pressures in areas of operation.

      If positive pressures are not being achieved in a central single-fan system,
the system should be  checked  to ensure that the fresh-air intake meets design
specifications and that the  intake has not been closed or restricted.  Increasing
the fresh-air  intake  if  it has  been restricted,  and operating  the  fan  for  a
sufficient time prior to  occupancy and continuously while the school is occupied
will help to reduce radon levels  that have built up during  setback periods and

-------
will maintain low radon levels during occupied hours by preventing radon entry
by maintaining a positive pressure and by providing fresh (dilution) air.

     In  a  central  dual-fan  system,  the  return-air fan  can be  set back  or
restricted so that all of the rooms are under a positive pressure.  The fresh-
air intake to the  supply  fan can also be increased up  to  the  design limit  of
the system,  if it  has  been reduced.   If radon  control  through  HVAC  system
operation  is  under consideration as  a permanent mitigation strategy,  proper
system operation and maintenance are critical.

     Many schools with highly elevated radon levels have installed SSD systems
in order to control radon levels even when the HVAC system is not operating.

                                  REFERENCES

1.   Leovic,  K.  W. ,  Craig, A. B. , and Saum, D. W.   The influences of HVAC design
     and operation  on  radon mitigation  of  existing school  buildings.   In:
     Proceedings of ASHRAE IAQ '89, San Diego, November 1989, NTIS PB89-218-762.

2.   ASHRAE Std.  62-1989.    "Ventilation  for acceptable indoor  air quality,"
     American Society of Heating, Refrigerating, and Air-Conditioning Engineers,
     Inc.,  Atlanta, Georgia,  1989.

3.   Radon reduction techniques in schools - interim technical guidance.  EPA-
     520/1-89-020, U.S.  Environmental Protection Agency, Office of Research and
     Development and  Office  of Radiation Programs, Washington,  D.C.,  October
     1989.

                               ACKNOWLEDGEMENTS

     The authors would  like  to  thank  the  support  staffs of AEERL and Harriman
Associates for their assistance in preparing this paper.

-------
                                                                      IX-4
         RADON MITIGATION EXPERIENCE IN DIFFICULT-TO-MITIGATE SCHOOLS
               by:
                    Kelly W. Leovlc and A. B.  Craig
                    Air and Energy Engineering Research Laboratory
                    U.S. Environmental Protection Agency
                    Research Triangle Park, NC  27711

                               and

                    David W. Saura
                    Infiltec
                    Falls Church, VA  22041
                                   ABSTRACT

     Initial  radon  mitigation  experience   in   schools   has  shown  subslab
depressurization (SSD) to be generally effective in reducing elevated levels of
radon  in schools  that have  a continuous  layer  of  clean,   coarse  aggregate
underneath  the  slab.    However,  mitigation experience  is limited  in  schools
without subslab aggregate and in schools with characteristics such as return-air
ductwork underneath the slab or unducted return-air plenums in the drop ceiling
that are open to the  subslab area  (via open tops  of block walls).   Mitigation
of schools with utility tunnels and of schools constructed over crawl spaces is
also limited.

     Three Maryland schools exhibiting some of the  above  characteristics are
being researched to help understand the mechanisms that control radon entry and
mitigation in schools  where standard SSD systems are not effective.  This paper
discusses specific characteristics  of potentially difficult-to-mitigate schools
and, where applicable, details examples from the three Maryland schools.

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

-------
                                 INTRODUCTION

     Subslab depressurizaCion (SSD) has typically been a very effective radon
mitigation approach in schools and in houses that are underlain with a continuous
layer of clean, coarse aggregate.  Recent experience has shown that in schools
with 4 in. (1 in. = 2.54 cm) of subslab aggregate -- approximately 0.75 to 1.5
in. in diameter with no  fine material  - -  one  SSD point can sometimes mitigate
an entire school wing of 10 classrooms  or about  15,000 sq (1 ft = 0.09 sq m).
Costs of mitigation under  these  conditions  of excellent subslab communication
may be as  little as $3,500  for materials,  in  addition  to  approximately 120
person hours for diagnostics and 160 person hours  for installation,  depending
on the number of suction points needed (1,2).

     However, mitigation experience has also identified certain types of schools
that are difficult  --  and consequently expensive --  to mitigate with current
technology.     Characteristics   of  potentially  difficult-to-mitigate  schools
include (but are not limited to):  1) schools  with poor subslab communication,
2) schools with return-air ductwork underneath  the slab, 3) schools with unducted
return-air plenums in the  drop ceiling  (that  are open to the subslab area via
open tops  of block walls), 4)  schools with  utility  tunnels, and  5)  schools
constructed  over crawl  spaces.    Mitigation  of such  schools  could be  very
expensive, especially  if the  mitigation  involves  a complete  retrofit  of the
heating,  ventilating, and air-conditioning (HVAC) system.

     These characteristics that may cause a school to be difficult-to-mitigate
are discussed  below,  and  examples  from  three difficult-to-mitigate schools,
currently being researched by  the U.S.  EPA in Maryland,  are discussed.   School
A  is  located  in Prince  Georges  County,  and  Schools  B and  C are  located in
Washington County.   Note  that since  School A exhibits more than one  of the
difficult-to-mitigate characteristics,  it is discussed  in two sections.

                    SCHOOLS WITH  POOR SUBSLAB  COMMUNICATION

     Although poor subslab communication  is a relative term, for the purposes
of this paper it is roughly  defined as  the  inability to measure  a negative
pressure (in the subslab relative  to the  building interior)  of at least 0.001
in. WC (1 in. WC - 250 Pa) in a 0.25 in. test hole located approximately 10 ft
from a 1.5  in.  suction point when  maximum suction  is applied to the suction point
with a variable speed Industrial vacuum cleaner.  As a  comparison, in the school
with excellent communication mentioned  in  the  Introduction, subslab depressuri-
zation was measurable in a test hole  100 ft from the suction point.

     Schools with poor subslab communication  typically have slabs that are poured
directly onto tightly packed soil such as sand  or  clay,  with little or no subslab
aggregate.    The tightly  packed fine  material  greatly restricts  the subslab
airflow and  limits  the  practicality of installing  a  SSD system.    Even  if a
school does show relatively good subslab communication  within  a given classroom,
communication  between  classrooms  may  be limited  by  below-grade walls  and
footings.   As a result,  it may be necessary  to  install a suction point in every
or every other classroom in order  to control  radon  entry.  An example of such
a school is discussed in Reference 1.

-------
     If poor communication reduces the subslab area depressurized by a suction
point, it may  be possible to achieve effective mitigation  by installing more
suction points.  Typically these  suction points are  installed around the edge
of each room in order  to depressurize the major cracks between the floor and the
walls.  A  related technique  involves the installation of an enclosed channel
between the floor and  the wall (channel drain) which can be depressurized.  Both
of these  approaches have the disadvantage  that they represent  a significant
investment in time and labor, and they be unsightly or take up excessive space.
The  area  depressurized by  each  SSD  suction point  may  also be  increased by
measures  that   increase  the   suction,  reduce   air  leakage,  and  increase
communication.   Sealing of floor  cracks, excavation  of  subslab cavities under
suction points,  and increased fan suction are often  used to increase pressure
field  extension.   Previous  school  research  has shown  that pressure  field
extension is often doubled by these measures (2).  Research on houses  in Florida
has shown the effectiveness  of very high pressure  fans (greater than 4 in. WC)
in cases of very poor communication due to sandy soils.

SCHOOL A

     As shown  in Figure 1, initial charcoal canister measurements made in this
school in March 1988  averaged 4.2 pCi/L  (1 pCi/L  = 37 Bq/m3), with the highest
room  measuring 12.3 pCi/L.    A  second  set  of charcoal measurements made in
November 1989  averaged 5.0 pCi/L, with two rooms measuring 10.0 pCi/L.

     The  school is slab-on-grade construction.   Inspection  showed that the
material under  the  slab is a mixture of sand and clay.   A utility tunnel runs
parallel to the corridor in part of the building, as indicated on  the  floor plan
in Figure 1.   The original HVAC system consists of  a perimeter hot water system,
with  air movement  by  convection.   There are  also overhead exhaust fans in the
classrooms; however, school personnel stated that these exhausts are rarely used.

     As a result of other  indoor air quality problems, overhead air-conditioning
units  in  the classrooms --  that  were designed with  the  capability  to provide
outdoor air --  have been  modified to provide heating so that  they can be used
year round.  Measurements  are in progress to determine the ability of these units
to pressurize  the building to prevent radon entry; however,  operation of each
unit is controlled in the  classroom by the teacher and, consequently,  continuous
operation cannot currently be ensured.

     Room 13,  the classroom with  the highest initial  radon level  (12.3 pCi/L in
Figure 1) was selected to  evaluate the applicability of subslab depressurization
in  this  school  with  poor subslab communication.  (Previous  efforts by  school
personnel   to   seal  the   floor/wall  cracks  did  not  reduce  radon   levels
sufficiently.)     As   seen  in   Figure  2,   three  3  in.   diameter  subslab
depressurization points were installed in the corners of  Room 13  and manifolded
to  a 4 in. overhead  line,  exiting through the window at point  A.  To improve
communication,  pits  approximately  2 ft in  diameter and  1 ft  in depth were
excavated  under the slab  at  the three suction points.   A fan  rated  at 270 cfm
(1 cfm = 0.47  L/s)  at 0 in.  WC  was installed.

-------
     In schools vith good subslab communication, 4  to 6 in.  diameter vertical
drops are typically installed because of  the high air flow;   however, because
of the low flow rates anticipated in the mixture of subslab  sand and clay, 3 in.
diameter vertical drops were used in this school.

     Communication measurements made in Room 13  with  suction being applied to
all  three  points are summarized  in Table 1.   With all three suction points
operating, a  slight  depressurizatLon was  measurable in the  far  corner of the
room.

         TABLE 1.  Subslab Communication Measurements, December 1989
                   (School A - Room 13)
       Suction      Suction          Distance  from        Pressure  in
        Point    in Pipe,  in.  WC   Suction Point,  in.     Test Hole,  in.

          A         - 1.48             1                 - 0.860
          A         - 1.48            60                 - 0.171
          B         - 1.47             1                 - 0.045
          B         - 1.47            60                 - 0.011
          B         - 1.47           120  (toward center) - 0.086
          C         - 1.45             1                 - 1.300
          C         - 1.45            60                 - 0.331
          C         - 1.45           180  (near door)     - 0.001
     Radon levels measured in this school with the SSD system in operation  are
shown in Figure 2.   It should be noted that a utility tunnel depressurization
system  -- that will be discussed later --  was  also in  operation in another part
of the building during these measurements.   The radon levels in Room 13 and the
adjacent room were reduced from premitigation radon levels  of 7.1 and 7.4 pCi/L
to 1.3  and  1.6 pCi/L,  respectively.   It  is  obvious  from  Figure 2  that radon
levels  are  lower  throughout  the building  --  not  only  in the  areas  where
mitigation was being applied.  However, radon reductions in parts of  the building
where  no  mitigation was  being applied were  about 45 percent  (attributed to
natural variations resulting from weather,  for example), and reductions in these
two classrooms were about 80 percent,  indicating the positive effects of the SSD
system.

FUTURE RESEARCH PLANS

     Further work in School A  will assess the possibility of installing SSD
systems in the other rooms with elevated radon  levels (other than the rooms being
treated by  the  tunnel depressurization system).    In  spite  of the poor subslab
communication,  it is reassuring to learn that the  radon levels could be reduced
in this classroom by installing enough suction points.  However, questions that
still remain include: 1) Would this approach work  in a similar school with much
higher  radon levels?  2)   What are the mechanisms controlling radon reduction
in the adjacent classroom?  and  3)   If every room  in a school with poor subslab
communication has elevated levels of radon,  would one suction point (or more)

-------
in each classroom be a practical mitigation approach or should alternatives be
sought?

             SCHOOLS WITH RETURN-AIR DUCTWORK UNDERNEATH THE SLAB

     Schools with return-air ductwork under the slab are a concern if elevated
levels of radon are present in the  surrounding soil.  Since the ducts are under
a negative pressure when the return-air fan is in operation, radon can be pulled
into the ducts through unsealed openings in the ductwork.  Any radon that enters
the ducts  from the  soil  under  the  slab  can then be distributed throughout the
school  by the  HVAC  system.     In  fact,  the American  Society of  Heating,
Refrigerating  and Air-Conditioning Engineers  (ASHRAE) has  recommended that,
where  soils  contain high concentrations of  radon,  ventilation practices that
place crawl spaces, basements, or underground ductwork below atmospheric pressure
be avoided since such practices tend to increase indoor  radon concentration (3).
Unsealed supply-air ductwork located underneath the slab may also be of concern
since radon can enter the ductwork  if it is subjected to negative  pressures when
the HVAC system  is off and the ductwork is below the neutral pressure plane.

SCHOOL  B

     In April 1988, initial charcoal canister measurements in this school showed
that  radon levels  in 28 of  39  rooms exceeded  4 pCi/L.   Follow-up canister
measurements made In 24 of these rooms in December 1988 showed all 24 rooms above
4 pCi/L.   All of these measurements were made  with  the  HVAC system off and the
building unoccupied.

     The original school was constructed  in  1954, and  four four-room additions
were added in 1964.   The entire school is slab-on-grade  construction, and the
foundation plans and specifications call for 4 in.  of aggregate under  the slab.
For the purposes of this paper,  only the  four  additions (which are referred  to
as pods) are discussed.  Each  pod  has four classrooms and a central media area.

     There is a  two-fan  HVAC system with the return-air ducts  located  under the
slab in each pod. Room air enters the return-air ductwork from registers  located
on the exterior wall of each classroom and is  pulled through the subslab ductwork
to a central  cold-air return  located in the  media  area.   Supply  air  (including
recirculated air from the subslab ducts) is distributed  to each classroom through
overhead ducts mounted near the interior wall.  Radon grab samples collected  in
the supply air when the system was operating were about  20 pCi/L, indicating that
the return-air  ducts  under the slab were  drawing in soil  gas  and recirculating
it throughout the school.  Subslab differential pressure measurements  (relative
to the building interior)  made in  one of  the pods  showed  a pressure  field  that
ranged from  - 0.001 to  - 0.1   in. WC with the HVAC system in operation.  The
greatest  negative pressures were near the central cold-air return in  the center
of the pod.   These measurements  suggested that it  may  be  possible to  install a
SSD system to exert a greater negative  pressure under  the slab than  the return
air fan.

      To evaluate the  ability of  a  SSD system to overcome the negative  pressures
generated by the return-air  system in one of the pods,  an exhaust fan rated  at
500  cfm (at  0   in. WC)  was  mounted on the  roof,  with a 6  in.  manifold  pipe

-------
connected to two 4 in.  diameter vertical drops.   These  two suction points were
placed near the return-air exhaust  stack in order to maximize  the  suction  in the
area where the negative pressures of the return-air ducts were the  strongest.
The  pod with  the  highest initial  screening measurements  was  selected  for
installation of the SSD system;  however,  it should be noted that the most recent
measurements in this school show radon  levels in this pod to be slightly lower
than in the other pods -- even when the mitigation system is not in operation.

     Post-mitigation  diagnostics  indicated  that  this  SSD  system could  not
overcome the  negative  pressures generated by the  return-air ducts and,  as a
result, radon-containing soil gas was still being pulled into the ductwork with
both the HVAC and SSD systems in operation.   Pressure measurements made through
a test hole drilled in one of  the  return-air ducts showed a negative pressure
of  0.80 in.  WC  in  the  duct;  whereas,  the  initial  differential  pressure
measurements made in the subslab aggregate measured only the negative pressure
caused by the duct leakage, rather than the actual negative pressure in the duct.

     Based on the above  results,  the school  recently installed a new overhead
return-air system in all four pods.   The new return-air registers are located
overhead near the interior walls of the classrooms, and  the supply-air ductwork
has been extended  to  supply air closer to the exterior walls.    However,  the
abandoned return-air registers have not yet been sealed off and, consequently,
there are still openings from the classrooms to the subslab.

     Continuous radon monitors (CRMs) were placed in  Classrooms 111 and 115 for
a 10-day period as  shown  in Figure 3.  Rooms  111  and 115 are  in different pods
but are at the same orientation within the pods. The  two SSD points are located
in Room Ill's pod, with one point adjacent to Room 111.    As seen in Figure 3,
radon levels in Room 115  are consistently higher than  those  in Room 111, even
with the SSD system off.   These results are consistent with those from electret
ion chamber (E1C) measurements in the other three classrooms in each pod.   As
shown in Figure 3,  the SSD system was on in Room Ill's pod from 0 to 120 hours
and off from 120 to  240 hours.   The  averages indicate that the  SSD  system clearly
has an effect on the radon levels, although the effect is much  less dramatic than
one might expect since the return-air registers have not yet been sealed.

     The next plan  for this school is  to seal the  return-air registers in all
16 classrooms.  Radon  measurements  will be made with the  registers sealed and
the new  HVAC system  in operation.   If necessary,  SSD systems will  then be
installed in the other three pods.

     An effort will also be made to tap into the  sealed-off  return-air system
with the SSD system in Room Ill's pod to compare its effectiveness in reducing
radon levels with that of the current system.

FUTURE RESEARCH PLANS

     School B showed that it was not reasonably possible to  overcome the negative
pressures of leaky  return-air  ductwork with  a typical SSD system,  even if the
slab is constructed on  aggregate.  These results will  be  confirmed with research
in  additional  schools;    however,   under   such  circumstances,  the  current
recommendation is to abandon (and seal off) the subslab ductwork and replace the

-------
system with an overhead  return-air  system.   Although this may be an expensive
retrofit  (about  $40,000  for the 16  classrooms  in School B),  the  guidance is
directly  applicable  in  the  design  of new buildings  or in  the  remodeling of
existing ones.

         SCHOOLS WITH UNDUCTED RETURN-AIR PLENUMS IN THE DROP CEILING
       (THAT ARE OPEN TO THE SUBSLAB AREA VIA OPEN TOPS OF BLOCK WALLS)

     Schools with  unducted return-air plenums  in the drop  ceiling may  be of
concern if elevated levels of radon are present in the surrounding soil and there
are openings between the  soil and the plenums through the unsealed tops of block
walls.  The potential problem for radon entry exists  if  the load-bearing hollow
block walls penetrate the slab as shown in Figure  4.  The radon-containing soil
gas can enter the  block  walls  and be pulled into  the unducted plenum which is
under negative pressure when the fan is in operation.  The soil gas that enters
from  the  block  walls could then be  mixed with the  recirculated room  air and
redistributed to  the building, presenting a potentially difficult mitigation
problem.

     This specific issue is not currently being  addressed in  the  field but will
be included in future studies.

                         SCHOOLS WITH UTILITY TUNNELS

     In some  slab-on-grade schools,  utility lines  are located in  a  subslab
utility tunnel  that  typically runs  parallel  to  the corridor  with  sections
sometimes branching off to individual rooms.  Sizes of utility tunnels may vary
from about 5 ft wide and 5 ft deep  (to  allow  maintenance workers  to enter them)
to 1  ft wide and  0.5  ft deep.   Utility tunnels may or may  not  have poured
concrete floors.  Even tunnels with poured concrete floors may have many openings
to the soil beneath the slab-on-grade, facilitating radon entry.   Risers to unit
ventilators or fan-coil units frequently pass through unsealed penetrations in
the  floor so that soil  gas  in the tunnel has  an  easy  entry route  to the
classrooms.

     Although utility tunnels can be a major radon entry route if the surrounding
soil contains elevated levels of radon, depressurization of  the utility tunnel
has been  considered  as a potential  mitigation  approach.    If utility tunnel
depressurization is attempted, any asbestos  in the tunnel should be removed or
encapsulated according to   the Asbestos Hazard  Emergency Response  Act (AHERA)
before attempting any radon reduction activities (4).

SCHOOL A

     A large utility tunnel (about 5 ft wide and 5 ft deep)  runs under or near
eight of the classrooms in this school as shown in Figures 1 and 2. Radon levels
in these eight classrooms  averaged  4.4 and  4.3  pCi/L without depressurization
being applied to  the tunnel.   (It  should be noted that  --as  with  the other
classroom in school A discussed earlier -- sealing of the floor/wall cracks was
attempted in two of these classrooms with negligible  results.)  Radon levels in
the utility tunnel typically range from about 10 to 30 pCi/L.

-------
     An exhaust  fan rated at  270 cfm  (at 0 in. WC) was Installed  in  the  tunnel
at the end of the corridor.   A 4 in.  diameter pipe penetrates the tunnel  from
the  outdoors, and  the fan  is  mounted on  a vertical riser at  roof level.
Differential  pressure  measurements  were made with the depressurization  fan  in
operation  at  the two  tunnel access doors; one is  located about 3 ft from the
suction pipe penetration, and the other is located near the tunnel T  by the two
farthest classrooms.   These measurements showed  depressurization  of  the  tunnel
of  about   0.03  in.  WC  at  both points  in the  tunnel.    A subslab pressure
measurement was also made in a classroom test hole about  10 ft  from the tunnel,
and depressurization from the tunnel was negligible.   (Remember from  above  that
this school is constructed on a mixture of sand and clay  and, consequently, has
poor subslab  communication.)

     Charcoal  canister  measurements shown  in Figure  2 indicate a reduction  in
radon levels  in  the eight classroom area  with the tunnel  depressurization fan
in operation.   The classroom levels averaged  1.8 pCi/L with the fan in operation
compared to 4.4 and 4.3 pCi/L (Figure  1),  a reduction of  about  60  percent.  (It
should again  be  noted  that the unmitigated areas averaged  about  45 percent
reduction  in  these measurements.)

     Figure 5 shows continuous radon levels for 22 days in one of the  classrooms
with the tunnel depressurization fan cycled on and off. To include measurements
during both occupied and unoccupied periods,  the  fan  was  on from Wednesday noon
to Saturday noon and off from Saturday noon to Wednesday  noon.  Radon levels  in
the classroom  averaged  1.2  pCi/L for  the  8 days  that the fan on and  5.1 pCi/L
for the 14 days that the fan was off.   The overall average for the 22 days was
3.6 pCi/L.

     Recent continuous  measurements collected in the tunnel showed  that radon
levels are similar when the  fan is  on and when  the fan is off.   This implies
that the radon levels in the classrooms are being reduced because of the reversed
pressure differentials between the tunnel and the rooms, rather than by dilution
of the  radon  in  the  utility tunnel.   This  will  be  investigated  further by
collecting simultaneous radon measurements in the classrooms and  tunnel.

FUTURE RESEARCH PLANS

     Research on this school will continue at least through the winter months.
In addition, utility depressurization will also be studied in other  schools to
determine  its overall applicability.

     In slab-on-grade schools with  utility tunnels it may also be possible to
reduce radon levels with a SSD system if subslab communication is good.  However,
two potential problems could be  1) too much SSD system air may be lost  to the
tunnel,  and 2)  the  depressurization may not be able  to  reach  the radon entry
routes in the tunnel.   This will be addressed in future research.

                    SCHOOLS CONSTRUCTED OVER CRAWL SPACES

     Mitigation techniques applied  in  crawl  space  houses include: submembrane
depressurization  (SMD)  in the crawl space, depressurization or pressurization
of the crawl  space,  and natural ventilation  of  the  crawl space.      To  date,

-------
research of crawl space houses has shown SHD to be the most successful technique
of the four in reducing radon levels in the living area.   However, since schools
constructed over crawl spaces are typically much larger than houses constructed
over crawl spaces, the practicality of installing a SMD  system in a school crawl
space may be very limited.    In addition to the  larger  size of the crawl space,
school crawl spaces often have structural support walls and  piers throughout,
which could quickly increase the  cost of installing a SMD system.   If the crawl
space contains asbestos, any techniques that may increase air movement or require
entering the crawl space should be avoided.

SCHOOL C

     Crawl  space depressurization is  currently being  tested  in a  school in
Washington County, Maryland.   Half of this single-story school is constructed
over a crawl space with two slab-on-grade additions.  The  crawl  space part of
the building was built in 1936 and is wooden floor joist  construction with a dirt
floor.  The HVAC system  is a single-fan system with overhead ductwork. The crawl
space is approximately  14,000  sq ft  and has numerous support piers that would
make installation of  a SMD system very difficult.

     A 500 cfm  (at 0 in. WC) fan was  installed to depressurize the crawl space.
Figure 6  shows  continuous  radon levels in the  school  office and in the crawl
space for  a  16-day period  cycling the crawl space depressurization fan on and
off for 4-day periods.   Operation of the depressurization fan tends to smooth
out the peaks in the indoor radon levels,  although spikes above 4 pCi/L do occur.
Spikes exceeding 10 pCi/L  occur  during periods  when the fan is not operating.

     School vacation began on day 357 and continued until day 369.  Radon levels
in the office drop considerably  on day 369 when school  was back in session even
though the crawl space depressurization fan was off and radon levels in the crawl
space were  still quite  high.    The lower radon  levels  on days 369 to 371 seem
to  indicate  that normal HVAC operation creates a  positive pressure, but that
during setback  periods  (nights,  weekends, and holidays), the building is under
negative pressure.

FUTURE RESEARCH PLANS

     Research of mitigation approaches in a relatively  small  crawl space (4,800
sq  ft) will be  initiated in early 1990.  The school is located in Nashville,
Tennessee, and  is  a four-classroom addition to one  of the  slab-on-grade schools
discussed  in Reference  1.   SMD, crawl  space  depressurization,  crawl  space
pressurization,  and natural ventilation of the  crawl space will be tested and
compared for effectiveness.   Available data will be presented at  the  Symposium.

                                  CONCLUSIONS

Based on the authors'  experience in these schools, the following conclusions can
be  made:

      1.    In schools  with poor  subslab communication, it may  be possible to
           adequately  depressurize the  subslab  area by  adding enough suction
           points (three  in one classroom  in  this case), excavating a  pit under

-------
           the  suction  point,  and  using a  high suction  fan.    However,  the
           applicability and practicality of this  approach in schools with highly
           elevated radon  levels throughout the entire building, must still be
           addressed.

      2.    Since  subslab return-air ductwork is under a negative pressure when
           the HVAC system is in operation,  radon  can be pulled into the ductwork
           through  unsealed openings, and then  distributed  in  the building with
           the  recirculated  supply air.   In some cases, it may be necessary to
           relocate the  subslab  ductwork  overhead since a SSD  system may not be
           able  to overcome  the negative  pressures  generated  in  the  subslab
           ductwork.  If this is done, the return-air  registers should be sealed
           since  they are  a  radon entry route.   Future research will determine
           the possibility of depressurizing  the  sealed subslab duct system for
           radon  reduction.

      3.    Schools  with unducted return-air plenums in the drop ceiling that are
           open to  the subslab via  uncapped block walls should be researched to
           determine their impact on  radon entry.

      4.    In slab-on-grade schools with utility tunnels, it is sometimes possible
           to reduce radon  levels in the classrooms by depressurizing the  utility
           tunnels.  This approach  needs  to be  studied in additional schools.

      5.    Thus far, crawl space depressurization has shown some potential for
           reducing radon levels  in schools constructed over crawl spaces.   Future
           research will  look  at   the  applicability  of  SMD,  crawl  space
           depressurization, crawl space pressurization, and natural ventilation
           of the crawl space in more detail  to determine their performances in
           large  crawl spaces typical of  schools.

                                  REFERENCES

1.    Craig, A.B.,  Leovic,  K.W., Harris,  D.B. and R.E. Pyle, "Radon Diagnostics
      and Mitigation in  Two  Public Schools  in Nashville,  Tennessee," presented
      at the  1990 Radon Symposium  on Radon and  Radon  Reduction Technology in
     Atlanta, GA,  February 19-23,  1990.

2.    Saum, D.W., Craig, A.B.,  and Leovic,  K. W.   "Radon Reduction Systems in
      Schools," in  Proceedings: The 1988  Symposium on Radon and Radon Reduction
     Technology.   Volume  1.  Symposium  Oral  Papers.  EPA-600/9-89-006a,  NTIS
      PB89-167480, March 1989.

3.   ASHRAE Draft  Standard  62-1981R.   Ventilation for Acceptable  Indoor Air
     Quality.  ASHRAE,  Atlanta, GA, 1986.

4.   Asbestos Hazard Emergency Response Act (AHERA), Public Law 99-519, October
     1986.

-------
                                                   Prealtlgatlon ridnn lenels (ptl/L)

                                                   Set 1  March 11-13. 1988 (upper)
                                                   Set 2- Nmeafter 17-20. 1989 (Iner)
            Figure  1.   Premitigation  radon  levels in  School  A.
                                                  Radon leveli IpCI/L) «'th 3-polnt SSD
                                                  syitn and tunnel depressuriiation lysten
                                                  in operation
                                                  Oecenber 1-4. 1989
                       —vpn^ui       up,  _,-,.,
                       .-tJ"... I     ,.,       l.rn
                       -TTTT"!           I  J

Figure 2.   Radon levels  in School A  with  SSD and tunnel depressurization

systems  in operation.

-------
^-»
o
                                              AVERAGE RADON


                                          ROOM 115       ROOM 111
                      40
200
                                                                                240
                                         HOUR
     Figure  3.   Continuous radon measurements in Rooms  111  and 115;SSD fan

     in Room 111  on and off  (12/1-11/89).

-------
          POSSIBLE
       FRESH-AIR INTAKE
                       ROOF
              I
  AIR
HANDLER
                              AIR SUPPLY DUCT
                        RETURN AIR
                              RETURN
                                AIR
                 HALL
                                    UNDUCTED PLENUM
                                     T  -
                                  DROP
                                 CEILING
                                  CLASS  ROOM
SUPPLY
  AIR
Figure 4.  Example of school with unducted return-air plenum in drop ceiling.

-------
    30
    20
 C_5
 O.
 o
 O
    10
     0
           Total period = 22   days,  radon * 3.6 pCi/L
           Fan on 8   days, radon - 1.2  pCi/L
           Fan off  14   days,  radon  *  5.1 pCi/L
                                     HOURS

    Depressurization fan.on (Wednesday noon  to  Saturday noon)
    Depressorization fan off (Saturday noon  to  Wednesday noon)
Figure 5.  Continuous radon measurements  in  School  A with utility tunnel
depressurization  fan on and off (9/27 -  10/22/89).

-------
           Ion on	",<.	Fort off
--- X	fon ofj  	> - On
  0
                         120
                                       HOURS
Figure 6.   Crawl  space depressurization in School C with  fan  on and off
(12/22/88  -  1/7/89).

-------
                                                                       Paper IX - 5
       AIR PRESSURE  DISTRIBUTION AND RADON ENTRY PROCESSES
                            IN EAST TENNESSEE SCHOOLS


                   L. D. Sinclair', C. S. Dudney", D. L. Wilson, and R. J. Saultz

                              Health and Safety Research Division
                                Oak Ridge National Laboratory
                                  Oak Ridge, TN 37831-6113
                                         ABSTRACT

       Many building characteristics have been found to influence radon entry, including building size and
configuration, substructure, location of utility supply  lines, and design  and operation of the heating,
ventilation, and  air conditioning  (HVAC)  system.   One of the  most significant factors is room
depressurization resulting from the HVAC system exhausting more than it supplies. This paper represents
a preliminary assessment of HVAC characteristics and  how they may relate to radon entry.  During the
summer of 1989, a limited survey was made of air pressure and radon levels in  four schools in  eastern
Tennessee.  Short-term samples of radon and pressure were made in all rooms in contact with the soil using
alpha scintillation cells and an electronic micromanometer, respectively. The pressure differences and radon
concentration changes induced by operation of the building ventilation system varied  among sites within
individual schools.
" Permanent address: Lexington High School, Lexington, SC 29072.

" Author to whom correspondence should be addressed.

-------
                                        INTRODUCTION

        Radon-222 is a radioactive, colorless, odorless, chemically inert gas with a half-life of 3.8 days.  It
is the parent of several short-lived, alpha-emitting radionuclides that occur naturally in the environment and
have caused lung cancer in some exposed human populations (1). Recent attention has focused on radon
in schools because of legislation passed by the U.S. Congress (2).

        Radon studies have been conducted in schools in 13 states. Elevated radon levels have been found
in many of these schools.  Although there are no studies to determine whether children are more sensitive
to radon  than adults,  some studies of other radiation  exposures  indicate that children may be  more
sensitive (3).  Consequently, children exposed to radon could at greater risk than adults from exposure to
radon.  Indoor radon in large structures such as schools is a topic of concern to the U.S. Environmental
Protection Agency (EPA),  the U.S. Department of Energy, and various public health officials in all  levels
of government.  It now appears that the benefits of making schools more energy efficient by sealing up
cracks and cutting back on  HVAC  operation must be evaluated against the substantial risks associated with
decreased indoor air quality.  In addition, Saum et al. (3) have shown that air pressure varies significantly
between rooms in a single elementary school  in Maryland.  We suspect that at the time of initial installation,
the HVAC system is balanced for each room (i.e., the system is adjusted so that supply exceeds exhaust by
a small  amount), and there is positive pressure in every room. As the system ages, some rooms come  under
negative pressure.  In rooms with negative pressure, some of the air that is drawn into the room comes from
cracks in or near the floor. The source of most radon is uranium in the soil(4). So if a schoolroom is both
under negative pressure and connected to a portion of the soil with elevated radon concentrations, the room
may have elevated radon levels.  Therefore, the purpose of this study was to monitor HVAC operation,
pressure, and radon concentrations at selected sites in four school buildings to gain greater insight into the
variations and relations among these variables in large structures such as school buildings.

                          EXPERIMENTAL MATERIALS  AND METHODS

        The primary criteria used  to select  the four schools were:

        1.      Some rooms with  radon levels above 4 pCi/L, were expected based on earlier surveys (6,7),

        2.      Schools  were under normal or near normal operation due to summertime  use of the
                building,

        3.      Schools were similar in size, construction, and design, and

        4.      The full range of public school  grade levels were represented.

        School A is a high school, constructed in 1969.  It  is a block and brick veneer structure built on
 a concrete slab with  underlying  aggregate.  It consists of a  two-story classroom wing containing 40
 compartmentalized  rooms, a one-story gymnasium  wing with  perimeter  classrooms, and a  one-story
 auditorium wing which also houses  the  band and chorus facilities.  The  three wings are connected to a
 center  commons area which contains the cafeteria, library, and administrative offices.  The HVAC system
 consists of individual unit  exchangers in the classroom wing.  These exchangers  allow air  to pass through
 heating and cooling coils.  The heat is  furnished  by hot water from a central boiler and cooling is furnished
 by chilled water from a  central chilled water unit.  Air returns are located in  the exchanger.   In the
 auditorium wing, air is split into two streams after the air supply fan:  one stream is  heated by hot water
 coils, and the other is chilled by cold water coils.  The  two streams are carried in parallel  ducts to each
 room.  A mixing box  in each  room controls the percentage of heated and cooled air entering  the room
 depending on the room thermostat.   Returns were located  in the wall or in the  slab.   There  are three

-------
 ventilation systems.  One (roof mounted) for the first and second story classroom wing, one for the office
 and cafeteria area, and a third for the auditorium areas. There exists a sub-slab ventilation system (return)
 unique to the band practice room area located between the band and chorus rooms.  Other returns in this
 area (assembly) were located in the wall.

        School C is an elementary school, constructed in 1968.  It is a two-story block and brick veneer
 structure of modified open space design. The lower level includes three clusters (pods), each consisting of
 five classrooms and a teacher planning center. A maintenance corridor is located along the back side. Since
 this school was built into the side of a  hill, the  lower level classrooms are walk-out basements and are in
 contact with the soil on three sides. The upper  level consists of three pods identical  to and directly above
 the lower level.  There is also a one-story gym, cafeteria, media center, and office area which fronts the
 school. The HVAC system is of the dual air supply design with returns located in special light fixtures that
 have a chamber around them for exhausting air  to the ceiling plenum.  It  is turned on locally at 6:30 a.m.
 and off at 2:30 p.m.  The system is shut down at 2:30 on Friday for the weekend.

        School D is a junior high school, constructed in 1954. Also of block and brick veneer construction,
 it consists of a separate two-story classroom building with 35 self-contained classrooms, library, and offices.
 Another separate building contains the gym and still another two-story  building contains  the cafeteria,
 auditorium, and basement level shop classrooms  and storage.  The classroom building is  built into the side
 of a hill and has one wall in total  contact with  the soil.  The basement level has no air conditioning.  A
 dual air supply system provides both heated and cooled air (added  in 1988) in both the  two-story  building
 and in the auditorium-cafeteria level of the other building.  The gym area has forced air heat exchangers
 but no air conditioning. The HVAC system is turned on at 6:30 a.m. and off at 3:30 p.m. and is shut down
 for the weekend.

        Constructed in 1976, School E  is a high school and  is constructed of a slab on grade with block
 walls and a veneer of bricks. The main building consists of two stories  of classrooms with a hexagonal
 commons area, cafeteria, and administrative offices at the center.  This building was built into the side of
 a hill  and portions of it have all areas in contact  with  the soil.   The lower west  wing was  situated
 considerably below  grade.  In fact,  the hallway leading to it is inclined.  A separate building contains the
 gymnasium-auditorium complex.  It has a basement (girls and boys locker and shower rooms) and a first
 and second floor (balcony level). The HVAC system consists of roof mounted ventilators and individual
 room air exchangers that  differ from those at school A in that they are on the outside wall and vented to
 the outside.  They bring in fresh air as  well as circulate room air.  The operation of the HVAC  is under
 computer control from the district  office and operates from 6:30 a.m. until 3:30 p.m. on weekdays during
 the summer with some provision for manual override at the school. The vocational complex was not studied
 because of time limitations.  However, it has construction and location characteristics that would make it
 of interest.

        From June 22 until August  14, 1989,  four schools were surveyed for  radon  using  several
 measurement methods.  Results reported here were all obtained with  a Pylon AB-5 monitor operating with
 room air flowing through a  163 cm3 Lucas cell at about 1 L min'1. After steady state conditions were
 achieved within the Lucas cell, at least ten consecutive 1-minute readings were recorded and averaged.  The
 efficiency (CPM per pCi/L) of each Lucas cell was determined in a chamber at ORNL  using instruments
 that had been compared with instruments at EPA's Eastern Environmental Radiation Facility or instruments
 at DOE's Environmental  Measurements Laboratory.   During any  day in  which data were collected,  the
 operation of the instrument was checked with a ^Ra source and the background count rate was  determined
 by sampling outdoor air for at least ten minutes.  Background count rates were subtracted from observed
 count rates before calculating radon concentrations. The sampling was performed according to established
 EPA guidelines, in that, sampling was done in the center of the room, away from windows, doors,  corners,
and HVAC ducts.

-------
       Pressure readings (referenced to the air mass in a central hallway) were recorded for each location
with a digital micromanometer. The micromanometer was turned on and, after a warm-up period, the zero
point was established by connecting the two pressure connections together with tubing.  For differential
pressure measurements, the micromanometer was connected to a length of tubing inside the room  to be
tested,  and to  an identical length of tubing that was placed under the closed door and opened into the
adjacent hallway.  The instrument was electronically zeroed immediately before each measurement.  Data
were recorded  after allowing a few seconds for the instrument readings to stabilize.  The position of the
HVAC return  and supply in the room was noted and whether or not the HVAC system was  on or off in
that particular location.

                                  RESULTS AND DISCUSSION

       Figure 1 and Table 1 summarize the results from  the survey.  The highest ^Rn concentrations
were seen in School A in the auditorium wing during periods of HVAC operation.  Moderately elevated
concentrations were also seen in Schools  A,  C,  and E when  the HVAC was not operating.  Lower
concentrations were observed in School D. In Schools A and D, the cases of elevated ^Rn were generally
limited to those instances when the pressure was low (see Figure 1), similar to what has been reported in
a school in Maryland (4). A similar trend may be present in the data from School E, but because the ^Rn
concentrations are lower, the signal to noise ratio is lower and no conclusive statements are possible. There
is no apparent correlation between radon and pressure data  from School C, where the ^Rn concentrations
are lowest on average among  the  four schools.

       In  School A, we  observed considerable variation among zones  of  the building.   For further
evaluation of this phenomenon, we divided the building into zones.  Zone U consisted of all rooms not in
contact with the soil. Zone A consisted of the auditorium and music rooms. Adjoining Zone A was  Zone
B, consisting of the cafeteria and nearby rooms. Adjoining Zone B was Zone C, consisting of the central
office and nearby rooms.  Zone D adjoined both Zones  C and E and consisted of the ground floor of the
classroom wing.  Zone E was  the gymnasium and shop areas.  Figure 2 summarizes  the means of all mRn
and pressure measurements made in these zones during periods of HVAC operation and non-operation.
There is much variation among zones with regard to the mean levels of ^Rn and pressure.  There is also
considerable variation among  zones as to  both (he magnitude and algebraic sign of the change induced by
operation of the HVAC system.

       Results in this paper  confirm and extend the observations of Craig and his  coworkers (4). These
results strongly suggest that the impact of HVAC operation on radon entry processes in schools and  other
large buildings is very important  and can  vary considerably  within a building.

       These results  have important implications  for the  design of surveys of mRn in  large buildings.
This study shows that care must be taken in the analysis of results from surveys of large buildings to reflect
the status of the HVAC system during the survey in comparison to its status during normal occupancy.  To
reduce costs, some survey designs may include only sparse sampling of rooms within a large building using
passive monitors.  Such survey designs have limited ability to reduce  the occurrence of falsely negative
findings (i.e., cases where buildings are falsely declared "radon-free"). An alternative approach may  be to
screen  a building with a manometer and to place radon monitors in known low-pressure rooms.

                                    ACKNOWLEDGEMENTS

       The authors are very grateful to the staff of the schools in which these data were collected. Without
their patience, this study would not have been possible. One of us (L.  D. S.) was a Teacher Research
Associate at Oak Ridge National Laboratory under the auspices of a program sponsored by the Office of

-------
Energy Research of the U.S. Department of Energy.  This research was sponsored by the Office of Health
and Environmental Research of the U.S. Department of Energy under Martin Marietta Energy Systems, Inc.,
contract DE-AC05-84OR21400 with the U.S. Department of Energy.

        The work described in this paper was  not funded by the U.S. Environmental Protection Agency
and therefore the contents do not necessarily reflect the views of the Agency and no official endorsement
should be inferred.
                                        "The  suborned  manuscript  has  bMn
                                        authored by  a  contractor of  the U S
                                        Government  under  connect   No  0€-
                                        ACOS-B40R21400 Accorehnoiy. the US
                                        Government  retorts  e  normdusrva.
                                        royalty-free license to pubhsh or reproduce
                                        tn6 pubbstwd  form of this coflintxrtioo, or
                                        aDow others to do so. for U S Government
                                        purposes'

-------
                                         REFERENCES

1.      National Research Council.  Health risks of radon and other internally deposited alpha-emitters.
       Washington, DC:  National Academy Press; 1988.

2.      Indoor Radon Abatement Act.  U.S. Congress.  1988.

3.      Cook-Mozaffari, P.; Darby,  S.; Doll, R.; Forman, D.; Hermon,  C;  Pike,  M.;  Vincent, T.
       Geographical variation in mortality from leukaemia  and  other cancers in England and Wales in
       relation to proximity to nuclear installations 1969-78. British Journal of Cancer. 1989.

4.      Saum, D.; Craig, A. B.; Leovic, K.  W.; Radon reduction systems in schools.  Paper  X-4 in Proc.
       of  1988  EPA Symposium on Radon and  Radon Reduction  Technology (Available  from U.S.
       Environmental Protection Agency,  Washington, DC 20460). 1988.

5.      Nero, A. V.; Nazaroff, W. W.  Characterising  the source of radon indoors.  Radiation  Protection
       Dosimetry 7(l/4):23-39; 1984.

6.      Hawthorne, A. R., Gammage, R. B., and Dudney, C. S. An indoor air quality study in forty  east
       Tennessee homes.  Environment International 12:221-239. 1986.

7.      Dudney,  C S.;  Hawthorne, A. R.;  Wallace, R. G.;  Reed, R. P.  Radon-222, mRn progeny,  and
             progeny levels in 70 houses. (In  press) 1989.

-------
n *° "
3E -
1*
£. 30 -
K
:? I »•
O £
3 | 2.-
J 5
2?
B9
O. 10 -
•o
{3 E
K
3
'

•





O
O
•
•

*
1







School A
HVAC
• OH
O OFF

u

•


2

I
o
« « 0
!iai^. 	 ?. .
E -


4 -

5
^
•— 3 -
e
K
2 -


1 -






1

« •
'


i


•

„ m
•
V
•

o
o m


School C
HVAC
• ON
0 OFF

m
*
B •
• B
a


1 •

0 "
1 •
DOE -0003 -0001 0001 0003 0 OOE 0007 0009 0011 0013 001E _0006 ' _„ OM 0JOX T 0 O'os ' „*, * „,',. ' J















18
§j "•'"""• "n 	 w""> P,....,. (inch.. .1 ».!.,)
e
3 2*-
n
1
3 "-
t>
n* 2C ~
09 18 -
3 I '•-
3 f "-
- f 12-
JJ i
S
»— I fl
a
g,
e
•i « -i
(5 OH

0




e
0



•

S
o °





School D
HVAC
• ON
O OFF






m
> ^
o •
•i .-'.
m -
9 -
e -
7 -

^ 8 -
— 8 -
g
1 4-

3 -

2 -

1 -
0 -

0


c
C

o
• *


0

o






School E
HVAC
• ON
O OFF



"
"





B "















GT -0012 -0008 -0004 0 OOO4 0008 0012 0018 -001S -0 OOE 0 DOS 0018 0 02S
Pr.nur. (Inoh.l ol w.l.r) Pr.llur. (Inch.! of w.l.r)

-------
                                                                        IX-6
                    RADON  IN  SCHOOLS  OF  MASSACHUSETTS
                  by:  Lee Grodzins
                      NITON Corporation, Bedford,  MA 01730
                      Massachusetts Institute of Technology
                                  ABSTRACT

      NITON  Corporation  has  completed  the  testing  of  more  than  7,000
classrooms in about  500  buildings of more than 80 public  and private school
systems in Massachusetts.   This paper presents our protocols and summarizes
our  results  based on  about 5,000 of  the tests.  Only  6% of the  rooms  had
radon levels, measured in weekend  screen  tests, above 4  pCi/L,  a  factor of 3
smaller than found in a recent  EPA survey.  Our protocols are similar to those
recommended  by  the  EPA,  with  three exceptions:   1.  Long-term tests  that
include nights,  weekends and school vacations are never  carried out; they are
manifestly unrealistic measures of radon levels during occupancy.   2. Follow-
up tests  of  elevated measurements are made during hours of occupancy;  i.e.,
during the daytime when schools are in use.  3. Successful  screen tests, with
the buildings closed and unoccupied  and with  ventilation systems  dampened or
off,  are carried out  in warm weather.

-------
                                INTRODUCTION

      We  treat  our children differently  from ourselves.   We  set  for them a
higher  standard.   Their environment must be  ultra-clean,  whether  ours is or
not.  So  we demand that their schools be free  from asbestos,  free from lead
paint, free from radon.
       A  few  states have passed  laws requiring  that  schools be  tested and
mitigated, if necessary.  Most states will probably  follow. But schools cannot
and are not waiting.   Massachusetts, for example,  has  no radon  testing law,
nor is  one  imminent.   Nevertheless,  thousands of classrooms of Massachusetts
have already been tested, for school administrators  are  all too aware that the
wide  knowledge  of the  risks of radon  have,  de  facto,  made  the  schools
accountable.
      We report here on  the  testing  of more  than  5,000  classrooms  in some 70
school systems.  NITON's basic protocol is a three-stage procedure similar to
that  advocated  by the  EPA1.  We  begin  with a  broad-coverage  screening test
using charcoal,  liquid scintillatoc,  passive  detectors.  The tests are carried
out under "worst case conditions," over a week-end or during a vacation.  The
initial test  is followed, where  necessary,  with tests carried  out  when the
school  is in  session,   again  using liquid  scintillation, passive  charcoal
detectors.  Finally, in  those cases  where  there are  elevated  radon  levels
during occupancy,  we make diagnostic tests  with electronic radon monitors to
recommend mitigation procedures.
      Schools  are different  from  houses,  but two generalizations  are much the
same:   There  is yet no way of telling, a priori, which school buildings,  or
which  rooms  in a  given building,  will have a  high  radon  level;  every
frequently occupied room on or below grade must be tested. And the mitigation
of elevated levels  is  idiosyncratic;  there  are  few  commonalties  and no magic
bullets that will solve the  radon  problem  in   all  schools.
      This  report  begins by reviewing those differences  in  construction and
usage that  bear on the protocols  one should use.   Section III describes the
NITON detectors and our early results for a  few school systems that led us to
adopt the protocols described in Section IV.   Section v presents  the full
results of our  school  tests.  Case studies give  examples of different classes
of school buildings,  demonstrate  the differences between  radon  levels found
over  weekends  and during  week  days,  and between  daytime  and  nighttime
concentrations,  and give examples of successful mitigation. The final section
presents our conclusions relevant  to  this  conference.
                 II.  SCHOOLS  ARE  DIFFERENT  FROM  HOUSES

      The upper  section  of  Table  1  summarizes  the  main  differences  in
construction,  ventilation  and usage  that  bear on  the  radon  problem.  In
particular,  the almost universal use  of on- or below-grade rooms in sprawling,
decentralized schools make a strong case for  testing every such occupied room.
      The lower section of Table 1 summarizes the obvious  differences  in the
profiles of  occupancy between a  home  and a school,  differences  that  argue
compellingly that long-term, uninterrupted radon  testing is inappropriate for
schools.

-------
                                  Table  I.

              Houses                            Schools

Small Footprint.                    Very Large Footprint.

Basements often not  used.           Ground Floors almost always  used.

Basements  usually  isolated  from Ground  floor often open  to upper
the first floor.                    floors through wide stairwells.

The  rooms of  a  given floor  are Classrooms,   especially  in  the
generally open to each other.       lowest grades, are often isolated,
                                   with  rooms  closed  for  extended
                                   periods.

The stack effect of the furnace is The  boiler  room   has   little
a paramount concern  in the  winter,  relevance to the radon problem.

N.E. homes are rarely  on slab.      Schools are often on slab.

N.  E  homes usually have  2 floors Schools  are  often single  or two
plus a basement and  attic.          story; on or below grade.

No ventilation code.               Mass,  code   requires  10   cfm  of
                                   fresh outdoor air per person.

One type of heating  system.         Usually a complex HVAC system.
Occupied day and night, especially Occupied during the day. Rarely at
at night.                          night.

Occupied on weekends.               Used  sparingly  on  weekends  and
                                   then only during daytime.

Occupied during the summer.         Used sparingly during the summer.


      School construction is remarkably varied.  Some buildings are built  on
slab,  some  over  crawl space.  Old  buildings  may  still have  fieldstone
foundations. A school may  have  one to several  additions, each  constructed
years apart under different  codes and designed by different architects.
      Schools are supposed to  have  ventilation  systems operating whenever the
building  is occupied.   Massachusetts, for  example,  requires  that schools
supply  at  least  10  cfm of fresh  outdoor air  for each  occupant;  for most
classrooms  that  is about 2  air exchanges per  hour.  To  meet this code, some
schools  have  central  HVAC  but many have  complex,  hybrid systems.   Modern
schools in  our area  often  have central ventilation plus  independent univent
heating systems  in every classroom to heat fresh make-up  air. Nevertheless,
the ventilation system is often the offender  when a  radon problem  is found  in
a school.  One  reason is that  older school buildings generally have antiquated
ventilation systems,  or none at all.  Another can be traced to the  1970's when
the sharp increase in the cost of energy prompted many custodians  to minimize
the amount of fresh air brought in  during the colder months.

-------
       Finally,  we note  that the  stack  effect can  be important  in every
building,  but  the causes and consequences  are different in  a  school and a
home. In particular,  unrelated parts of a  large,  one-story school may  exhibit
widely   different  stack  effects  and,   hence,  widely   different  radon
concentrations.
                    III.    Early  Results  and  Protocols

     All screening measurements have been carried out  with  our  patented NITON
Liquid Scintillation  detectors.   These small vials,  weighing  about half an
ounce, contain a perforated plastic container filled  with  activated charcoal
and  desiccant.    The  adsorption  time constant  makes them  appropriate  for
sensitive testing from 8 to 72 hours.  Removing  the cap  exposes the charcoal
to the ambient air; screwing  the  cap  back on seals the, gasket and completes
the test.  The detectors  are highly resistant to problems  of high humidity and
air movements.
     We carried out  radon tests of several school systems  in the fall of 1988.
Fig.  1 shows the  extreme  variations that one can encounter in different school
buildings of the  same  town.  School A had no radon level above 1 pCi/L; School
B had a  few isolated  high radon values, difficult to  find  without a thorough
test; the majority  of the rooms in  School C had radon levels above 2 pCi/L,
though no room had  a  serious problem; the majority of tests in school D were
greater than 10 pCi/L.
  V
  .a
  S
                          Radon Concentration, pCI/L
  Figure  1.  Radon  Screen  Teat Results  in  4  School  Buildings

-------
     In one town, we worked closely with a superintendent who had an excellent
custodial staff  and was himself conversant  with  the construction  and  use  of
every building.  He had also done  his  radon  homework.   When we walked through
the system, he knew where he wanted the  detectors  to be placed.   One  modern
school in particular was especially suspect, being  built into a granite ledge
so as to be surrounded on  two  sides by stone up to  the  top floor.  Another,  a
modern, one-story,  well-ventilated school,  situated on  top of a knoll  would,
he  felt,   be   no problem.   However,   former  school building had a  radon
distribution similar to  School  A of Figure 1, no radon  levels above 2  pCi/L.
The latter school building had the radon distribution of School D.
      When these results were confirmed with a second weekend test,  in which
the   radon  concentrations  reached   100  pCi/L   in   one  classroom,   the
superintendent considered shutting the school down until the radon levels were
mitigated.   William Bell  of  the  Mass.  Department of Public Health  and I
persuaded  him  to allow us  to first carry out  some  long-term, time-dependent
radon tests.   These results, which are commented on more thoroughly below, are
shown  in  Fig.  2.   The radon levels in that school were varying  from day to
night  by  more than a  factor of 25; on one  day the variation was  close to a
factor of 50.
     u
     a
      •s
     O
     5
     at
     u
     i
     u
                       Central Air Exchange
                       On Continuously
                                              (• 914 4M 431 43< 4i» 3«4 311 SSI
                                                  U
                                                  p.
                                                  a
                                                  5
                                                  o
            24 41  71 »« 11*
144  161 1»2 11« !'• «« »•• "2 >}< '

 TIME, Each Unit Is One-Half Hour
   Figure 2.    12 Days  of  Radon  Tests  in One  Classroom  in  School  D.

       A careful examination of our  results convinced  us  that  we could neither
 predict  which buildings nor which  rooms  in  a given  building  might  show high
 concentrations  of radon.  School  B,  with 5% of its  rooms  above 4 pCi/L, showed
 that  every  occupied room on grade  should be tested,  or one  can  miss the odd
 room  or  the special wing that might be  seriously polluted.
        The results  of  School D  also  made evident  that  follow-up  tests must
 measure  the radon concentrations during occupancy; we care minimally about the

-------
concentrations when no one is  in  the  building to be at risk.
     Our early  conclusions  have  been confirmed and strengthened by a year of
testing  school  systems.   The protocols outlined  in the  next  section are
designed to uncover problems and  obtain  insight into a mitigation  strategy.


                            XV.   NITON  PROTOCOLS

1. THE INITIAL SCREEN TEST:   A screen test is a short-term test carried out in
such a manner as  to create "worst case" conditions.  The radon concentration
found in the screen test should be higher than  the radon  concentration in that
location averaged over an entire  year of occupancy.

      The key word is occupancy.  Schools are typically occupied by a given
staff member  for  no more  than 25% of the hours  of  the year;  a  child is in
school for no more than  15% of the year.  It is the radon levels  during these
fractions of a year that we seek  to screen.

      The data  in section V  show that  a "closed school"  test almost always
yields a higher elevated concentration than that found during  a  school day.

2. WHEN  TO  TEST:   Start  the  test on a Friday  afternoon,  harvest them Monday
morning.   By  Saturday  morning,  the school  building will have  had closed
conditions  for  about  12  hours;  by Monday  morning, a charcoal  diffusion
detector, which gives greater weight to  the radon  levels  in  the  later times of
the tests, will give a good measure of the  radon concentrations.
      A Friday to  Monday test  is not written  in  stone.   Two-day  tests during
vacation periods  or  tests pulled early  because of impending  school functions
or stormy weather are also successful.   But a weekend  test is generally the
most economical, requiring the minimum overtime pay  for the  custodians.
      Winter or Summer?  The EPA  states that "radon screening measurements in
schools  should be made in  the  colder   months  (October through  March)  when
windows and doors  as  well  as  interior room doors are  more likely  to be closed
and the heating system is  operating."  We know of no  evidence  to  support this
recommendation.   Indeed, one can generally get better "closed conditions" in
the  summer since  schools are often completely closed  during  parts  of the
summer.  The  remaining justification for cold weather screen  testing presumes
that  the  heating  system exacerbates the radon problem.   Our view is that the
emphasis on the heating  system is misplaced.   The  proper  emphasis should  be on
the fresh  air input systems,  required by Massachusetts  and most states.  If
such  systems  are  shut off,  and the school is closed and  unused,  then a screen
test  is  likely to be as  "worst  case" in the  summer  as  in the winter.   NITON
tested several  large school systems this past summer.  We expect to retest  a
number of the buildings this winter with the hope that  the data  will  provide
convincing evidence  for  or against year  round screen testing of schools.

3. WHERE  TO TEST:  Every frequently occupied room on  or  below  grade should be
tested.  Closed rooms that are occupied  for only short periods  — file rooms,
storage  rooms,  boiler rooms and  the like  — should  be  tested  only with  full
knowledge that  acceptable levels  can exceed 4 pCi/L without undue risk.
      Cafeterias,  gymnasiums,libraries and other large open  rooms  seem  to have
higher radon  levels  than do classrooms.   Perhaps it is because they are seldom
well  sealed against soil gas.  Auditorium stages, for example,  are  sometimes
built  directly over  the ground.  We recommend one  detector for every  2,000

-------
square feet of floor space;  that  is,  a  60 by 60 room would have 2 detectors.
     Washrooms and showers.   These tend to have elevated concentrations due to
gaps around pipes  that  go through  the floor.   But a  given  individual spends
little time in these rooms,  so that  testing should be done with insight.
     Upper floors: The open structure of schools expose the upper floors to a
radon problem if it exists  in the ground floor.   A  thorough screening of the
on-grade rooms  should uncover the  problem.   We  recommend that a  few upper
floor rooms be tested in every wing.

4. CONDITIONS FOR  TESTING:  We all  agree that the school  should be  closed as
much as is practical during the weekend test.   And no one disagrees that high
winds and  stormy  weather should be  avoided,  if  at  all possible.   There is,
however, disagreement on the matter  of  the operation of the HVAC system and on
summer time versus winter time testing.  I will take up the latter point again
in the  last  section.   On the question of  the operation  of  the HVAC system,
NITON prefers to make the tests with the ventilation system throttled so as to
bring in the  least amount of fresh outside make-up air and provide a better
"worst  case"  screening  test.   The EPA, however,  recommends  that the heating
system  be  kept on during the entire  weekend,  a procedure that produces a
"worst case" only when the ventilating  system  is not bringing in outside air.

5. WHO  SHOULD DO  THE  TESTING? A school can  save a  great  deal of money by
putting out and harvesting the tests themselves.  The testing procedure itself
is simple.  So too are the directions for placement. If one can handle taking a
pill from a child-proof bottle, one can conduct a passive radon test. But our
experience  shows  that  there  are  no free  lunches  here.    The  school
superintendent  must decide whether  there  are competent  custodians who will
responsibly put  out and harvest the tests  and,  most important,  keep proper
records  of what they  have  done.   If  not,  we believe  that the  school will
benefit from having  a  professional radon expert  who  will  assume  all
responsibilities;  in any  case,  the expert  will be  needed if  problems are
uncovered.   On the other hand, the budgets of many schools preclude hiring an
outside professional for the  first screen test.  In that case, NITON provides
a number of aids:
a. Clear written instructions as  to  how to make the test.
b. Record-keeping sheets so that  each custodian will provide  us with the basic
information of location and testing  periods.
c. Where practical, we hold a meeting of the custodial staff  to explain radon,
hand out material,  demonstrate the  test, stress  the  need for record keeping,
accuracy, promptness and diligence,  and answer questions.
d. A  successful strategy is  to  give a free test to  custodians  so  that they
could test their own home to become  familiar with  the  entire  procedure.

FOLLOW-UP TESTS:

1. Criteria;  what should be  the criteria  for retesting?   At what screening
level  do  we  pronounce  a  room  cleared?   There  are  no one-line answers.
Prudence dictates  that  we recognize that  radon levels can fluctuate markedly
from  day  to  day,  from week  to week,  from season  to  season,  and  that a
screening test is not invariably  a  "worst case."
NITON asks for follow-up tests whenever:

     a.  A room has  a radon concentration above 3 pCi/L.
     b.  A cluster  of rooms has radon  concentrations above 2 pCi/L.
     c.  An  entire  school  has  a mean  radon level above 1 pCi/L.

-------
      The isolated  high-radon-level  room may be a  fault  of the test, but we
have found that the problem is almost always real and often  caused by  a faulty
ventilation  unit  or a unique radon entry point. Isolated rooms with  elevated
radon levels should be tested with two detectors placed side-by-side.
       Areas  of  radon concentrations exceeding  3  pCi/L should  be  retested
thoroughly,  with additional  tests in rooms  above  or adjacent to  the area.
Duplicate tests are satisfying but not necessary since the tests validate each
other.
     Clusters of rooms that have weekend values  between 2  and 4 pCi/L  point to
a problem that could  be  serious in another  season or  year.   We highlight this
region and ask for early follow-up tests on most of  these  rooms.
      Some  school buildings show no radon concentration greater than  3 pCi/L,
but yet  are  far  from being radon  free.   These buildings must be watched.  We
begin with follow-up tests of about 20%  of the rooms.
      Having  determined which rooms to retest, we now must decide how to make
the tests in the most direct, economical manner  possible.  Follow-up tests can
simply repeat the initial screening test to confirm  their  findings.  NITON has
long  since discarded this approach  since we found that  we rarely  failed to
confirm the earlier findings.  Our preferred follow-up tests address the radon
problems directly.
      The NITON  follow-up tests are carried  out  during school  days when the
rooms  are  occupied.  The  tests  are  done  with  our  LS   vials   so as  to
differentiate between radon levels in the daytime and those  in the nighttime.

2.  The  Day   versus  Night  Follow-up   Test:  We  seek  to  know   the  radon
concentration that students and staff are exposed to. We  thus need  to measure
the radon concentrations  during a  weekday,  from roughly 7 A.M.  to  7 P.M; the
duration depends on  the  school use.   Continuous monitors,  in place for at
least one day, would give the most complete information, but such monitors are
expensive and the first  follow-up tests, like  the  initial screening,  often
involve a number of rooms.
     We prefer to make the tests using our inexpensive NITON LS detectors.  We
have carefully calibrated the sensitivity and accuracy of these detectors for
periods ranging  from  8 to 72  hours.   The sensitivity at 8 hours is more than
adequate; a  10  minute liquid scintillation count is  sensitive to  0.4 pCi/L
with  an  uncertainty  of  25%.   The  accuracy of an  8  hour  measurement,  as
determined from  tests at a National Radon  Facility,  is well within 15%; the
typical precision of  multiple tests  is  about 10%.   Both the accuracy and the
precision are considerably worse  than we generally  obtain with  a  2 day test,
but they are  quite acceptable for  screening tests of  elevated radon  values.
     Consistency checks  of this procedure are constantly made by exposing, in
classrooms,  three sets of detectors:  one during the day;  a second  set during
the night;  a  third set for the full 24 hours.
       The  custodial  staff  (now expert in handling our  detectors)  or  a
professional exposes  one set of detectors  from, typically,  8 A.M.  to 6 P.M.
and another set  from  6 P.M.  to  8  A.M. An equivalent combination is a daytime
exposure together with an overlapping 24  hour exposure.

DIAGNOSTIC TESTS:
      Daytime  concentrations  are generally 30% to 50% smaller than  the "worst
case" screen  tests.   Thus, many  of the rooms found to  have screen-tested radon
levels in the 2 to  6 pCi/L range,  are  found,  with  daytime  testing  to have
acceptable values during occupancy.  These  rooms must be checked periodically
but need not  be mitigated.
      The rooms  that have elevated radon concentrations during occupancy must
be dealt with expeditiously.   Section V  gives a  few examples.

-------
                                V.
Results
SCREENING TESTS:

     Fig.  3  presents the results of 5,200 screening tests  carried out in 350
school buildings in  68  school  systems  of  Massachusetts,  mainly public, during
the past 12  months.   Most of the tests were conducted during the period from
November,  1988 to May,  1989  but  every  season  of the  year  is  represented.
Most of  the  screening tests  were  carried out by  the  school  custodians.   An
additional 2,000 screening tests (not shown)  were carried out by professionals
in private schools  in this area. The  results  are shown on  a  semilog plot to
show the  overall trend  to  high radon values.   The abscissa  values  above 10
pCi/L are grouped in 5  pCi/L bins,  shown as 5 identical vertical bars.   Some
overall results:

       61% of the rooms  had radon levels below 1 pCi/L.
       6% of the rooms had radon levels above 4 pCi/L.
       1.1%  of the rooms had radon concentrations above 10 pCi/L.
       0.7%  of the rooms  (37)  had radon concentrations exceeding 20 pCi/L.

    The  elevated radon  concentrations  are far less frequent than found in the
recent EPA study of  3,000  classrooms.2  In that study,  20%  of the school rooms
had radon  levels exceeding 4  pCi/L,  and  3%  had radon levels  above 20 pCi/L.
We  have  no  explanation  for  the factor  of  3  to  4 between the  EPA results
obtained in  130  schools  and the Massachusetts school results.
                 Radon Distribution in 5,200 School Rooms In Massachusetts.
         100 -t
         .001
                           Radon Concentration, pCl/L
 Figure   3.  Radon  Distribution  of  5200  School  Rooms   in Massachusetts

-------
     The gross averages tell little  about  the individual schools, which have
widely  different  distributions,  as  Figure  1  illustrates.  The point  is
emphasized by noting that 25% of the buildings accounted for all of the radon
levels exceeding the EPA  "action  level"  of 4 pCi/L.   Said the other way,  260
buildings, that is, 75% of  the  total,  had no "elevated" radon concentration.
And  25  school buildings,  7% of  the total, accounted  for all of  the radon
concentrations above 10 pCi/L. Some of these "high radon" buildings were old,
most were not.   Some  were built on slab,  some were  not.   Some were in towns
that have higher than  average radon  levels,  others were not.   No pattern has
emerged so far.
     The elevated  levels  found in the weekend tests are  only indicators of
possible problems.  Before  one  rushes  to mitigate, one must confirm that the
results apply to people when they  occupy  the building.

DAY-NIGHT FOLLOW-UP TESTS:

      Radon  concentrations  observed during  a regular school  day are almost
always  lower than  the values found  either on the weekend or at  night.   An
example  of  our observations is given in Table  2,  which shows  the  results
obtained for two school buildings in one system.  The second column gives the
values  found over a weekend in October.  The  third column gives  the values
found a  few  weeks  later during the  daytime (12 hours) of a  school day;  the
fourth  column  gives  the concentration found during the  following night  (12
hours.)

            Table  2.    Weekend,   Daytime  and  Nighttime  Testa
                           in  One School  System

Room              pCi/L, Weekend   pCi/L,  Daytime    pCi/L Nighttime

5: 13                   6.8               3.0               2.8
E: Gym                  5.5               2.3               2.9
E: 10                   6.2               1.9               3.0
E: Library              6.4               2.9               3.8
E: 16                   5.2               2.5               2.9

F: Teachers Room        7.0               1.8               7.9
F: Cafeteria            5.1               2.5               6.4
F: Spec. Ed             3.6               1.5               7.1
F: 11                   3.6               1.2               3.6
F: 10                   5.4               2.1               6.0
F: Janitor             10.0               2.4               6.1
F: Girls Room           5.4               2.1               6.0

     In building E,  the daytime  and nighttime values  are similar, and both are
about a  factor of  2 less than  the weekend  results.   In  this  building,  the
ventilation system  was  never turned off on  weekdays.
     In building F,  the nighttime  concentrations are  similar to those found on
the  weekend, and  both are about a factor  of  2  higher  than the  daytime
concentrations.   In this building,  the ventilation  system is damped at night.
      The results of 40 such tests are summarized in Figure  4;  four  sets  in
which a ratio exceeded 1.5  have been excluded to  emphasize the  main  body  of
data.  On the Y axis,  we have plotted the ratio of the daytime to the weekend
concentration; on  the  X  axis is  the ratio  of  the nighttime to  the  weekend
concentration.   (The division of the  two  values gives the ratio of the daytime
to nighttime concentrations.)

-------






c"
55

_
e
8
C
o *;
US
•O u
e «•
v X
ft) C
.-
2
V
g
£
n
Q




140-
130-


120-
110-


100-

00-

80-

70-
60-
so-

40-

30-
20-
10-


:
j
j
i
| "
j
i
i
i
j
j
i
i
j •

; •
• j
! ,
• * » ' '
M ' ^

|"
ft '
' j* '
1:
r
i 	
              20     40     60     80     100    120    140

                    Nighttime to Weekend Concentration, In Percent
                                                         in
                                                               180
                                                                     200
     Figure  4.  Day  to  Night  to  Weekend  Concentrations of  Radon.

     The scatter of both  ratios  is wide.  But  on  average,  using all data from
tests  of some  60  rooms,       the  daytime concentrations were  42% of  the
nighttime concentrations,  with a standard  deviation  of 32%; on average,  the
nighttime concentrations were 97% of the weekend results; 
-------
systems being turned  off.   When the air exchange system was left running all
the  time,  the maximum values dropped  to  just below  4 pCi/L.   The univent
system lowered the radon levels to below 2  pCi/L.
       The  one-week  average for  the radon  concentration  in School  B  was
approximately 11  pCi/L,  far above  the  EPA guideline,  and  a demand  for
mitigation.   But  the  one-week average of the  radon concentration from 8 A.M.
to 6 P.M was only 1.9 pCi/L, well within the EPA  guideline.

LONG-TERN TESTS

      In  every  school building  we have investigated,   a  long-term  test that
includes  the nighttime  and weekend concentrations  would  give a  totally
erroneous representation of the radon exposure to  the occupants.

EXAMPLES OF MITIGATION

1.  The Odd High Radon Level.  We frequently find that  a building has only one
room or a limited area with an elevated concentration.  Such problems are often
easy  to  diagnose  and correct.  An example was  the multi-winged,  one-story
school built on slab.    An examination uncovered  two problems.
     The ventilation  system was defective.   For one  thing,  the  air returns
inside the classroom  closets were  so covered  with books and paraphernalia as
to be ineffective.   For another, the make-up  air brought  in by the univents
and central ventilation  system  had been deliberately  restricted years ago to
reduce the cost of heating.  When  the returns were cleared, the radon problem
in the rooms of one wing disappeared.
    But the radon problem in two end  rooms of another wing not only persisted
but remained well above  10  pCi/L during the daytime.   Further tests with the
LS detectors, carried out by the custodians,  showed that  the back  of one of
the classrooms was  about 20 pCi/L -  almost twice the  concentration found in
the front of the room.
    When William Bell  and I  investigated, we found that the radon was gushing
in from  the  clearance space around  the sink  drain pipe.  Foaming this space
solved most of the remaining radon  problem.

2.  School D.  School  D, the modern grade school building of Figs. 1 and 2 is
a one-story building built over a crawlspace whose height ranges from about 3
to more  than  6  feet.   Service  ducts  and pipes form a  tangle above  the dirt
floor.  The radon concentration  in the  crawl  space was very high and clearly
the cause of the radon problem in the  school.
      Sealing  the  dirt floor,  with the possibility of  pumping under  the seal,
was  and  remains  an  expensive  option.   Instead, the  decision was made  to
mitigate  by  controlling the cycles  of heating  and  ventilation.   The  air
exchange  system  is  kept  on during  the week  to supress excessively  high
concentration.  The  univent systems  are turned  on by  6 A.M.  and turned off
around 6  P.M., later if there is an evening activity.  We periodically monitor
the building and have  advised installing a permanent radon monitor.

                             VI.     Conclusions:

      NITON has tested some  7,000  classrooms  in  about  500  buildings  in more
than 80 school systems.  The first measurements were screen tests carried out
over  a  weekend under "worst  case"  conditions.   The  follow-up tests  were
carried out during school  session;  daytime and nighttime concentrations were
obtained  separately.    Finally,  where  necessary,  diagnostic  tests  were
conducted with electronic sniffers  and monitors to determine'the origin of the

-------
radon infiltration.  He draw the following conclusions  from these data:
1. A week-end test, beginning Friday afternoon and ending Monday morning, when
carried  out  under closed building  conditions,  provides  a  reliable screening
measurement.   The radon concentrations are,  with few exceptions,  higher than
concentrations found during occupancy.

2.  The ventilation  system  in  the  school should be  restricted  during the
screening test if one is to simulate "worst case" conditions.

3.  Follow-up  tests  of elevated  readings  should  distinguish daytime and
nighttime  radon  concentrations.  Such tests  can be  carried out economically
using passive LS charcoal detectors.

4. Long-term testing of radon that includes nighttime, week-end and vacation
concentrations produces manifestly improper measurements  of the  radon exposure
to humans.   EPA'a insistance in suggesting alpha track and long-term EPERM
measurements as  a preferred alternative  for  screening  measurements weakens
their entire program for schools.
     The results  of  a long-term test can  be dangerously wrong if radon levels
at night  and weekends are  much lower than those during the daytime.  In that
•case, a serious radon problem may be missed.   The more  likely scenario is that
the  higher radon concentrations generally found on weekends  and nighttimes
will  result in  a falsely high radon concentration.   Schools will  then be
compelled to spend considerable funds to "fix what  ain't  broke."

5.   We are  now  testing whether warm-weather testing gives erroneously low
radon results, as is  implied by the EPA  recommendation to test only in cooler
months.  We  hope to present the data at the conference.

6.   Our data  reinforces the EPA's conclusion  that  one  cannot predict which
buildings,  or which  rooms  in a given building, will have  a  radon problem.
Every occupied room on or below grade should be tested.

7.   Finally, we  emphasize  that school buildings differ from  one another in
construction and usage.   Testing  them for  radon  is now  a straightforward
protocol.    Mitigation  of  serious radon problems is,  however, not  well
understood.  We do not yet know how to systematically approach  a radon problem
in buildings with complex,  hybrid heating and ventilating systems.

                              Acknowledgements

     The data  described here was  obtained by  the  technical  staff of NITON
Corporation,  whose watchful and meticulous  work  is deeply appreciated.   I
would  also like  to  thank  William  Bell,  with  whom  I  did most of the early
diagnostic measurements; his teaching and advice are  of continuing  benefit.

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

REFERENCES:

1. Radon Measurements in Schools. An Interim Report.  United States  Environ-
   mental Protection Agency, March, 1969.  EPA 520/1-89-0010.

2. Phase I School Protocol Development Study.   EPA, Private Communication.

-------
                                                                  IX-7
                    RADON GAS TESTING IN KENTUCKY SCHOOLS:
     SUMMER TESTING PRAGMATIC CONCERNS AND PRESSURE / HVAC  CONSIDERATIONS

                  by:    Patrick Holmes
                        Alpha Spectra of KY.,  Inc.
                        Louisville,  KY  40243
                                   ABSTRACT

      This study examines the radon gas levels and related mechanisms in
elementary, middle, high and special schools.  The Jefferson Co.  Project
consists of 4,000 to 5,000 screening data points on 158 sites in Louisville,
Kentucky.  The primary instruments consist of short-term electret perms and
open-face design AC canisters with CRM's as cross-checking devices.

      Due to political, operational and financial considerations, this project
was mandated to be initiated and completed in summer 1989.  Modifications and
specific concerns were addressed, in order to adhere as closely as possible to
EPA's Interim Guidelines for school testing (see summary).

-------
                                                                  IX-8
                     RADON SURVEYS IN LARGE BUILDINGS;
                           THE UCF RADON PROJECT

                             Ralph A. Llewellyn
                           Department of Physics
                       University of Central Florida
                             Orlando, FL.32816
                                  ABSTRACT

     Documented protocols for surveying the distribution of radon in large
buildings do not currently exist.  Those developed for one/two family  •
residential structures and draft versions of those being developed for
schools provide inadequate guidance for investigators and diagnosticians
charged  with determining radon levels throughout large office buildings or
interconnected multi-building complexes.  To support the development of
protocols for large buildings, the UCF Radon Project has completed the
initial phase of a detailed data collection and analysis program.
Measurements of radon levels were made under known, controlled conditions
in twelve large buildings ranging in size from 25,000 sq.ft. to
225,000 sq.ft. and up to five floors above ground.  An average of more
than 100 radon measurements were made per building.  Results from the
study indicate that (i) radon levels often do not decrease as expected in
the upper floors of multistory buildings and (ii) sampling rates currently
being proposed for the above ground-level floors of large buildings may be
too low.
                                INTRODUCTION

     As is the case in several states, Florida is currently in the throes
of developing a building construction code intended to make all structures,
from single family residences through schools to high-rise office towers,
resistant to the intrusion of radon gas.  A legislatively mandated team
drawn from the Florida State University System has been working with
pertinent state agencies since September 1988 on the drafting of the
comprehensive code.  At the outset of their task the group recognized
that existing experimental data that could provide the basis for a
radon-resistant construction code which could apply to large buildings

-------
was exceedingly meager.  Even the most fundamental knowledge, such as
characteristic concentrations, radon movement pathways, and the extent of
exposures to occupants, was (and is) not yet available in the literature.
Indeed, standard protocols to govern the conduct of radon surveys and
follow-up measurements for large and/or high structures do not exist.  As
a consequence of the dearth of information, the drafting of codes and
regulations pertaining to the prevention of radon hazards in large
structures must await additional measurements and research.

                             UCF RADON PROJECT

     In mid-1989 the small environmental physics group at the University of
Central Florida initiated ~a research program, the UCF Radon Project,
directed toward filling a small part of the information gap.  The project
is organized in three phases:

Phase 1:  Baseline determination.  Collection and analysis of detailed radon
          concentration data in a number of large buildings.  Development
          of a computer program and database to aid the analysis and to
          facilitate diagnostic measurements and interpretation.

Phase 2:  Study the effects on radon concentrations and transport of
          changing the environmental control settings in selected
          structures.  Extend computer programs to include the results of
          phase 2 work.

Phase 3:  Utilize the results to develop a draft protocol for radon
          analysis of large buildings.

     The University's complex of large buildings of varying numbers of
floors, up to 5,  and areas,  up to 225,000 sq.ft. (20,900 mZ), provided a
convenient and in many ways ideal "laboratory" for the project.   All of
the buildings are of masonary construction and relatively new, the oldest
just over 20 years and the newest less than a year.  Types of building
utilizations include exclusively offices, mixtures of offices and
laboratories, a library,  and exclusively residential suites.  The HVAC of
all but the latter are controlled by a central computer system on a floor-
by-floor basis.   Table 1 identifies 12 of the buildings used in Phase 1
of the study as to their size and type of utilization.

DATA COLLECTION

     The data collection for the project was to  serve a dual purpose.
First, it was to provide research data pertaining to large buildings that
would assist in the development of future measurement protocols and
construction codes.  Secondly, it was to provide the University with timely
information regarding the levels of exposure to radon gas received by its
employees and student body.   To this end, all rooms were assigned to one
of four categories based on occupancy time, ranked in order of

-------
                      TABLE 1.  BUILDING DESCRIPTIONS
Building
No . Name
1
2
5
12
14
18
20
21
29
32
40
45
Administration
Library
Chemistry
Physics
Phillips Hall
HFA
Biology
Education
ecu
Seminole Hall
CEBA I
CEBA II
Utilization
type (a)
A
B,A
D,A
D,C,A
A,C
A
D,A
A,C
A,D
E
D,C,A
A,C
Number of
floors
4
5
3
4
4
5
4
3
2
4
4
4
Gross area
(sq.ft.)
87,700
226,500
49,100
106,500
64,600
84,000
62,800
110,300
23,400
42,100
130,900
119,700
 a   Key to utilization type:    A = offices;  B = library;  C = classrooms;
                                 D = laboratories;  E = residence suites
measurement priority.  They were:  1 - offices, 2 - residential suites,
3 - laboratories and classrooms, and 4 - mechanical and service areas.
Elevator shafts were initially of high priority interest and were
included in category 1.

     Data collection and measurements were carried out on the basis of an
a priori draft "UCF Radon Measurement Protocol"  (1) for large
structures, the detector deployment and measurement procedures of which were
based closely on pertinent USEPA documents (2) (3) (4).  The resources
available to the project have thus far enabled screening measurements to be
made in all first priority rooms (offices) and some second priority rooms
(residential suites).  A total of approximately 1500 measurements have
been completed to date.

     Radon samples were collected using charcoal canisters  exposed from 48
to 72 hours.  The radioactivity of each exposed canister was measured using
one of two 3 in x 3 in (7.6 cm x 7.6 cm) NaI(T£) scintillation detectors,
ORTEC electronics, and an IBM PC/XT configured as a 2048-channel analyzer
operating in four 512-channel segments.  The individual analyzer segments
were programmed to record gamma rays from radon daughters in the energy
range from 0.25 MeV to 0.61 MeV.  Typical operational characteristics of
the two detector systems are recorded in Table 2.
   F&J Specialty Products, Inc.  model RA40V

-------
                     TABLE 2.  DETECTOR CHARACTERISTICS
Channel Resolution
%
A 6.0
B 6.0
Efficiency

0.263
0.272
Error
% b
7.7
7.3
Minimum Detectable
Activity(pCi/liter) a
0.27
0.25
  a   At the level of 3 standard deviations.
  b   At the level of 2 standard deviations for a 4.0 pCi/liter result.
DATABASE

     The net gamma ray counts measured for each canister were entered into
a specialized database (UCFRADON), together with information that enabled
the program to calculate the radon concentration in pCi/liter according to
the procedures outlined in reference 4.  Its design is especially 'user
friendly!    Information is also entered that permits relating the radon
results to the HVAC status of the building maintained by the University's
indoor environmental control computer.  File structures in UCFRADON have
been arranged to facilitate analysis on a building-by-building, floor-by-
floor, and room-by-room basis.

     There are currently approximately 1500 records in the UCFRADON data
files.  The data retrieval routines currently provide a variety of options,
including, e.g., printouts of rooms with radon concentrations exceeding
user selected levels.  Additional options under study include floor-by-
floor contour plots and 3-dimensional building-wide concentration
histograms to aid in radon transport analysis.  The existing data and any
added subsequent to this writing as available to interested individuals
on request.


                                  RESULTS

ELEVATION DEPENDENCE

     Initial analysis of the data collected in the first phase of the
UCF Radon Project has yielded some interesting results.  That fickle
predictor, conventional wisdom, tells us that the concentration of radon
should decrease as we move upward in a multistory structure.  Assuming
that any radon in the building emanates from the ground rather than from
the materials used in the construction, it is easy to understand physically
why the concentration should decrease with elevation.

     Treating radon diffusion in air as a random walk problem, let's
consider some concentration of radon, n(z,t) atoms/in-*, which is introduced

-------
 at  the ground  level slab where z = 0  at  time t  =  0.   Straightforward
 application of statistical mechanics leads  to the  conclusion that
                           z2 = (l/3)v2 t t
                                                                         (1)
      —
where v  = mean square velocity of the atoms and f
collisions (5) .
                                                      mean  free  time  between
Equation 1 implies that the standard deviation of the z component of
                                               )^
the displacement vectors of the radon atoms Az =
                                                        is  proportional
 to  t5, where  t  is  the  time  after  the  radon was  introduced.   Figure  1  shows
 curves for  the  concentration n(z,t) vs  z  for  three different times
 0
-------
that the air in buildings circulates, to a certain extent due to natural
forces, but mainly because of very effective HVAC systems.  Still, the
expectation has always been that the radon concentration will decrease
as we move upward through the building.  The few sampling protocols that
have been drafted for large structures seem based on that assumption  (7).
However, results of experimental measurements on the large buildings
studied in the UCF Radon Project suggest strongly that the expectation is
not well founded.  Results for the five largest buildings studied, graphed
in Figure 2, show that in 80% of the cases the average concentration  on
the 2nd and 3rd floors substantially exceeds that on the 1st floor, as do
more than half of the cases at the 4th floor level.  For the two buildings
with 5 floors, the average radon concentration on the top floor was still
40% of that on the ground floor.

     These results, if substantiated by further work, have serious
implications for the development of protocols intended to guide the
accurate assessment of the health risks to occupants of large buildings
arising from long-term exposure to radon gas.  They also have something to
say about the design of big buildings.  Namely, the way to reduce radon
intrusion into floors above the ground level would be to isolate the  ground
floor HVAC system from that of the rest of the building.  Our interest
here, however, is in the first of the implications, which will be discussed
briefly in the next section.


SAMPLING FREQUENCY

     Draft protocols currently being discussed for measuring radon
concentrations in large buildings specify sampling rates on floors above
the ground level that are much lower than that for the ground floor.  For
example, reference 7 specifies a 20% sampling rate for second floors  and
10% for the third floors.  However, the results shown in Figure 1 suggest
that such low sampling rates will not yield a reliable profile of the
radon concentration distributions on the upper floors.  The large number of
measurements made during this study enabled a test of that suggestion.

     Using a number of floors in the larger buildings for which 100%  of
the first priority rooms had been measured, a pt? 'goodness of fit1 test was
done for random samples at several sampling rates from 20% to 80%.  This
test answers the question, "What is the probability that the sample
distribution agrees with the parent, or actual distribution?" (8).
Table 4 records the results of the test for three floors in two of the
large buildings studied.  Figure 3 shows the curves of probability of
goodness of fit used in evaluating the reliability of the random samples.
Clearly, a 20% sampling rate on the 2nd floor of CEBA I provided a poor
representation of the actual distribution of radon.  Indeed, sampling rates
in excess of 60% were necessary in order to achieve probabilities of  'good
fit* in the 0.7-0.8 range.  While these numbers are in part a function of
the statistical uncertainties of the relatively low radon concentrations in
the University's buildings, they suggest that sampling rates defined by
protocols be considered with great care.

-------
                        TABLE 4.   RELIABILITY  OF  SAMPLE
Building
and
Floor
CEBA I, 2nd
CEBA I, 3rd
HFA, 4th
Sampling Rate %

20 33 50
0.02 0.05 0.18
0.01 0.01 0.11
0.01

67
0.75
0.75
0.25

80
-
-
0.85
     Tabulated values  are probabilities of agreement between sample and
     parent distributions.
CONCLUSION

     Based on the results discussed above, it would be prudent for those
developing protocols for guiding radon measurements in large structures
to assess very carefully the sampling rates proposed for floors above the
ground level in large buildings.  Assumptions typically made regarding
radon concentrations and adequate sampling rates that are implicit in
current draft protocols may be seriously in error and could well lead to
substantial underestimation of the radon exposures received by occupants
of large buildings.

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


ACKNOWLEDGEMENTS

     The author wishes to express his appreciation to Hugh Ivie, Director
of the UCF Office of Environmental Health and Safety for his steadfast
encouragement and support of the UCF Radon Project, to Mark Llewellyn for
his design of a truly first rate database for the project,  and certainly
not least to Bruce Dean for the countless hours he worked on data
collection and measurement.

-------
                                 REFERENCES


1.  Llewellyn, R.A.   UCF Radon Measurement Protocol,   unpublished draft,
    University of Central Florida, Orlando, Florida,  1989.

2.  Ronca-Battista, M.,  Magno, P., Windham, S., and Sensintaffar,  E.
    Interim indoor radon and decay product measurement protocols.
    EPA 520/2-86-04, U.S. Environmental Protection Agency,   Washington,
    D.C., 1986.  50 pp.

3.  Ronca-Battista, M.,  Magno, P., and Nyberg,  P.    Interim protocols for
    screening and followup radon and radon decay product measurements.
    EPA 520/1-86-014, U.S. Environmental Protection Agency, Washington,
    D.C., 1987.  22 pp.

4.  Gray, D.  and Windham, S.   EERF standard operating procedures  for
    radon-222 measurement using charcoal canisters.  EPA 520/5-87-005,
    U.S. Environmental Protection Agency, Montgomery,  Alabama,  1987.   30 pp.

5.  Reif, F.    Elementary Theory of Transport Processes.  In;   Statistical
    Physics,  McGraw-Hill, New York, NY, 1967.

6.  Feynman,  R., Leighton, R., and Sands, M.   Molecular Diffusion.
    In;  The  Feynman Lectures on Physics, Addison-Wesley, Reading, MA,  1963.

7.  'Summary  of Proposed Radon Standard' developed by Occupational Health
    Conservation, Inc. for the Florida Department  of  Health and
    Rehabilitative Services for use by schools  and 24 hour  care
    facilities, 1989.

8.  Young, H.   Probability Distributions.  In:  Statistical Treatment
    of Physical Data, McGraw-Hill, Reading, MA, 1962.

9.  Morse, P. and Kimball, G.   Probability.  In:   Methods  of  Operations
    Research, The Technology Press/Wiley, New York, NY,  1958.

-------
                             H|(z.f)
                                    0                     z

Figure 1.  Concentration n(z,t)  vs z for various  tjnn.-s  aftei  t  = 0.
                             1.0
                                                               *   +
                                                              21    0

   2.0    Normalized
Radon Concentration
  Fipuri- 2.   R.idon  concentration vs elevation  in  5  l.irgc  buildings.

-------
       TOTAL  OF n
       mi OF THEM
       WHERE THERE
       RESULTS {; «1
                                                EXPECTED VALUE OF m, IS npif
                                                      WHERE  I  •«        /
                                                                           60 100
Figure 3.   Contours of  probability  for goodness  of fit.   (9)
                                            * U.S. GOVERWCNT FRINT1NG OFFICE- 1990  748-010/25005

-------