EPA-600/R-95-013
January 1995
AIR INFILTRATION MEASUREMENTS
USING TRACER GASES:
A Literature Review
By
Max M. Sam field. Consultant
915 W. Knox Street
Durham, NC 27701
EPA Purchase Order 3D 1279 NATA
EPA Project Officer: David C. Sanchez
U.S.Environmental Protection Agency
Air and Energy Engineering Research Laboratory
Research Triangle Park, NC 27711
Prepared for:
U.S. Environmental Protection Agency
Office of Research and Development
Washington, DC 20460
-------
EPA REVIEW NOTICE
This report has been reviewed by the U.S. Environmental Protection Agency, and
approved for publication. Approval does not signify that the contents necessarily
reflect the views and policy of the Agency, nor does mention of trade names or
commercial products constitute endorsement or recommendation for use.
This document is available to the public through the National Technical Informa-
tion Service, Springfield, Virginia 22161.
Abstract
The use of tracer gases for the measurement of air infiltration into structures and interzonal
flows within a structure is not new. This technique has been investigated over the past 15
years. Numerous tracer gases have been used, among which are: sulfur hexafluoride,
hydrogen, carbon monoxide, carbon dioxide, nitrous oxide, and radioactive argon and
krypton. Sulfur hexafluoride is the most common tracer gas of choice — primarily because its
presence may be accurately measured in the ppb range using electron capture/gas
chromatography techniques. Most of the other gases used may be accurately measured in the
ppm range using infrared technology.
There are generally three types of methods used: tracer gas decay, constant concentration,
and constant injection.
Investigations comparing tracer gases have led to the following conclusions: (a) Even though
sulfur hexafluoride is appreciably heavier than air, mixing is not a problem; and (b) The
inherent uncontrollable variables present in tracer gas work limit the accuracy of
determinations to +1-5% - 10%. There is thus no reason why one tracer gas should be
selected over another provided other criteria are met. In the case of hydrogen, diffusion of
the gas through the surrounding walls can pose a problem.
Tracer gases may be used in air flow measurements in large buildings where the building may
be treated as several coupled zones. In such a case, the decay technique can still be used by
having the system repeat the injection at regular intervals. A computer-controlled injection
system is described in the text.
ii
-------
Table of Contents
Page No.
Abstract ii
List of Tables and Figures iv
Introduction 1
The Characteristics of a Good Tracer Gas 2
A Comparison Between Tracer Gases Which Have
Been Used 4
Error Analysis of Tracer Gas Methods 7
Tracer Gas Techniques for Use in Measuring Air
Movement in Multicell Buildings 13
Instrumentation Used in Tracer Gas Work 15
Summary and Conclusions 20
References 21
Bibliography 23
iii
-------
List of
Tables and Figures
Table Page
1
Published Comparisons Between
Tracer Gases
5
2
Instruments Used by Various Researchers
to Detect Tracer Gases
5
3
Air Change Rates
11
Page
1
Sampling and Calibration Valves
18
2
Block Diagram of SF6 Computer-
Controlled System in Load Mode
18
3
Block Diagram of SF6 Computer-
Controlled System in Inject Mode
19
4
Electron Capture Detector Block Diagram
19
iv
-------
Introduction
During the past 15 years a multitude of diagnostic procedures associated with the evaluation of
air infiltration and air leakage sites have been developed. The spirit of international cooperation
and exchange of ideas have greatly facilitated the adoption and use of these measurement
techniques. Wide application of such diagnostic methods is not limited to air infiltration alone.
For example, they have been applied to the evaluation and improvement of radon reduction in
buildings. (1) Radon problems are not unique to the United States, and the methods have to a
degree been applied by researchers of other countries faced with similar problems. The radon
problem involves more than harmful pollution of the living spaces of our buildings - it also
involves energy to operate radon removal equipment and the loss of interior conditioned air as
a direct result. The techniques used for air infiltration evaluation have been shown to be very
useful in dealing with the radon mitigation challenge. (1)
Various researchers (2,3,4) have reported that a strong relationship exists between ventilation and
indoor air quality. Unfortunately, little work has been done to determine the interaction between
mechanical ventilation, infiltration, and indoor air quality. One reason that such data are not
available- is the complexity involved in measuring ventilation rates of large buildings (4,5,6).
The American Society for Testing and Materials (ASTM) defines a tracer gas as "A gas that can
be mixed with air and measured in very small concentrations in order to study air movement"(7).
This definition does not go far enough to characterize a suitable tracer gas, as brought out in the
next section.
1
-------
The Characteristics of a Good Tracer Gas
Tracer gas methods have been used for many years to measure the air exchange rates of a variety
of different types of buildings. Three basic methods are currently used for the measurement:
tracer gas decay, constant concentration, and constant injection. The tracer gas decay method is
the simplest and most inexpensive of the three methods to implement A standard practice,
ASTM E741-93, describes the technique (7). The constant concentration method employs an
automated system to simultaneously measure the tracer gas concentration and inject the
appropriate quantity of tracer gas required to maintain a constant concentration (B); the air
exchange rate is related to the tracer gas release rate. With the constant injection method, tracer
gas is released over extended time periods at a constant rate. The method requires a system for
release at well-controlled rates because the air exchange rate is related to the tracer release rate.
Both active and passive systems of tracer release and sampling have been used (9). Fundamentals
and applications of tracer gas methods for the measurement of air exchange have been presented
in a number of published reviews (10,11,12).
The energy cost of excessive air infiltration in buildings has spurred a sharp increase in the
number of research projects investigating the magnitude of this phenomenon. The most common
technique used to monitor air infiltration rates requires measurements of the concentration of a
tracer gas. The tracer gas, a material easily monitored which normally is not present in the
atmosphere, is injected into the space to be tested. When the injection ends, the concentration
of tracer is measured as a function of time. Outdoor air, leaking into the test space, replaces the
tracer/indoor air mixture which leaks out at the same rate. The rate of change of the
concentration of tracer in the indoor air is therefore proportional to the concentration of tracer
in the test space; i.e., the concentration decreases exponentially (13).
2
-------
This verbal description of the measurement process contains several assumptions about the nature
of air infiltration. For example, it assumes that the rate of air infiltration remains constant during
the measurement period. In addition, it assumes that the outdoor infiltrating air mixes uniformly
with the indoor air during measurement. Most investigations also assume that measured air
change rates are independent of the type of tracer gas used if adequate mixing of the tracer gas
occurs in the test space (13).
An ideal tracer gas should meet the following criteria (10):
a. Be inexpensive
b. Be easily measurable at low concentrations
c. Be non-toxic and non-allergenic
d. Be non-inflammable
e. Have approximately the same molecular weight as air
f. Not be absorbed on any surface within the space under test
g. Not be a normal constituent of air in the test space
No tracer gas meets all these requirements. Within recent years improvements in instrumentation
have permitted measurements of concentration of sulfur hexafluoride, SF6, at levels of parts per
billion. This is three orders of magnitude smaller than previous techniques which yield
measurements in the range of parts per million. The ability to observe such low concentrations
makes SF6 an attractive possibility for use as a tracer gas. However, several researchers have
expressed the concern that the large molecular weight of SF6 [146] will cause stratification of the
tracer gas after injection. Hunt and Burch (14) have discussed the errors which will result if a
tracer gas is poorly mixed within the test space. Stratification, which will yield poor mixing, will
3
-------
lead to erroneous infiltration results unless multiple sampling and properly weighted averaging
of the tracer gas from several points occur (13).
A Comparison Between Tracer Gases Which Have Been Used
The paper of Hitchin and Wilson (15) is an excellent review of the experimental techniques used
in measuring air infiltration. This work has been extended to include work through 1975 by
Hunt's review of current techniques. These authors cite previous direct intercomparison results.
These are shown in Tables 1 and 2 (13).
Warner (16) reported comparisons between coal gas (a mixture of H2 and CO obtained by passing
steam over hot carbon) detected using a katharometer and C02 whose concentration was
measured by Haldane gas analysis (13).
Collins and Smith (17) used the radioactive argon isotope41 Ar as a tracer; its concentration was
measured with a geiger counter and a ratemeter. A direct comparison was made of the infiltration
rate detected using H2 with a katharometer and 41Ar; agreement within 8 % was seen in two trials.
Howland et al. (18) reported comparisons between air changes measured with a radioactive
isotope, 85Kr, using a geiger counter and a ratemeter. The decay rates were compared with
measurements which used C02 as the tracer. Its concentration was determined by drawing
samples of air periodically and using chemical analysis (the Haldane apparatus) to find the
amount of tracer remaining in the test space. Results of three tests varied by about 9%.
4
-------
TABLE 1
Published Comparisons Between Tracer Gases
Reference
Tracer Gases
No. of Tests
Results *
Warner (16)
Coal Gas, CO,
3
1.05+0.18
Collins and Smith (17)
Ha, 41 Ar
2
0,93 ±0.01
Rowland, et al. (18)
C02, *Kr
3
1.00±0.09
Lidwell (19)
NjO, CjHjO
1
0.97
Howard (20)
H„ NjO
many
agreement
Howard (20)
02, NjO
many
agreement
Hunt and Burch (14)
He, SFe
6
1.17 ± 0.14
*Tbe results quoted are the mean values of the ratios of the
measured air change rates.
The ratio is formed by dividing the air change rate of the heavier
gas by the air change rate of the lighter gas.
TABLE 2
Instruments Used by Various Researchers
to Detect Tracer Gases
Research Group
Honeywell
Princeton
Lawrence Berkeley Laboratory
Lawrence Berkeley Laboratory
Lawrence Berkeley Laboratory
Tracer Gas
CH4
sf4
NjO
Cft
SFe
Interference between CH4 and N20 prevented simultaneous
measurements of tracer gas concentrations using these gases;
therefore, the tests were organized using the schedule shown in Table 3,
Instrument
IR Analyzer
Electron Capture
IR Analyzer
IR Analyzer
Electron Capture
5
-------
Lidwell (19) compared results when nitrous oxide, N20, is compared with acetone, C3H60, as
tracer gas. Infrared absorption was used to measure the concentration of N20; acetone
concentration was determined by measuring the change in pH which occurred when air
containing acetone is absorbed into solutions of hydroxylamine hydrochloride. A single
measurement (judged to be accurate within 10%) produced 3% agreement.
Howard (20) compared N20 with both H2 and 02. N20 concentrations were measured using an
infrared analyzer, H2 concentrations with a katharometer, and 02 by absorption in aqueous
chromous chloride. Specific results are not quoted in the paper. The author states that close
agreement between decay rates using N20 and 02 were seen over wide ranges of wind speeds.
On the other hand, H2 decay rates were substantially higher than N2G. The evidence suggested
that diffusion of H2 through the walls of the unpainted gypsum of the test space was the source
of the discrepancy. This hypothesis was tested by repeating the tests after the walls were sealed
with two coats of latex paint and also repeating the tests in a laboratory with masonry walls. The
discrepancy was not present in the latex paint and masonry wall tests.
Hunt and Burch (14) compared air change rates using He and SF6 as tracer gases to examine the
influence of molecular diffusion on the infiltration process. Their test space was a four-bedroom
townhouse constructed within an environmental test chamber. If molecular diffusion were
important in the infiltration process, the air change rate measured with He would be significantly
larger than that measured with SF6. In fact, slightly larger air change rates were seen when SF6
was used as a tracer rather than He. Six trials were made. The ratio of the air change rate
measured with SF6 to that measured with He was 1.17 with a standard deviation of 0.14.
6
-------
As described by Knoepke (21), SF6 has been successfully used to determine the volumetric flow
rate of flue gases. The tracer gas was injected at a constant flow rate, and a flue gas sample was
taken downstream of the injection point. A mass flowmeter (Matheson Mass Flowmeter LF10Q
with F-100 Transducer) was used to measure injection flow rate. SF6 was selected as the tracer
gas because it:
(1) Can be accurately measured from a few ppb to many ppm by using a gas
chromatograph equipped with an electron capture detector.
(2) Is chemically stable at red heat.
(3) Is not present in the atmosphere.
(4) Is non-toxic and odorless
The temperature at the point of injection was 870°C. When compared with pitot-tube flow
measurements, the largest deviation was 7%, and the average deviation was 4%. The advantages
of the tracer gas method over the pitot tube method include: (a) the pitot tube method requires
that a traverse of the flue be made, hence, instantaneous flow measurements cannot be made, and
(b) a flue survey can be completed in much less time using the tracer gas method.
Error Analysis of Tracer Gas Methods
Grimsrud et al. (13) did their intercomparisons of three tracer gases using the decay technique.
After injection, the rate of change of tracer gas concentration is the product of the air change
rate, A, and the concentration in the test space:
7
-------
dC/dt = -AC
(1)
If the air change rate is constant, the solution of Eq.(l) is amply;
C(t) — Cat'M (2)
where C0 is the concentration of the tracer at time t = 0.
The data may be analyzed by plotting concentration as a function of time using semilog graph
paper.
Grimsrud et al. (13) compared SF6 with CH, and NaO. Tests were made in a one-story
unoccupied residence in California. The volume of the living space of the house was 230 m3, its
floor area was 100 m2, and the area of the six surfaces bounding the living apace was 300 m2.
Table 3 shows the results of the measurements. The mean value of the ratio for all tests was
1.10 ± 0.10. The mean value for the comparisons of SF6 with N20 was 1.09 ± 0.09, while the
value for the comparison with CH< was 1.16 ± 0.09.
The uncertainties listed with each of the ratios is the standard deviation of a single measurement.
The "t" distribution with nine degrees of freedom was used to calculate the expected range of the
ratio. This result predicts that the actual range of the ratio, r , Ees within the range;
8
-------
1.01 l z i 1.20
at the 99% level of confidence. The "t" distribution assumes from an infinite sample in which
the scatter of results is due only to random effects. Their results, therefore suggest that;
(a) Systematic errors exist in the measurement procedure which results in SF4
concentration decays that are too large, or N20 and CH* decays which are too small; or
(b) SF6 overestimates the "true* air filtration rate when used as a tracer gas.
The above results refer to two independent sets of measurements of the concentration decay of
SF6 that were made using equipment at Princeton and Lawrence Berkeley Laboratory (LBL).
Hunt and Burch (14) compared tracer gas measurements described by Grimsrud et al. (13) and
a ratio of ASF6/AiightoJ„ of 1.13 + 0.12. If the t- distribution with 15 degrees of freedom is used,
it was predicted that the actual ratio, r , will lie within the range
1. 04 * r * 1.22
with 99% confidence.
The results show that a difference exists between air exchange rates measured using SF6 and air
exchange rates measured using lighter tracer gases. Grimsrud et al. (13) conclude that the
difference is small, representing the range of uncertainty which they estimate is present in any
tracer gas measurement (5 to 10%).
9
-------
An example of this is test 12 shown in Table 3. The air exchange rates measured using C2H6,
SF6, and N20 were 0.68, 0,66, and 0.61 hr"These values represent the range of values seen
whenever air exchange rates are measured; consequently, Grimsrud et al, (13) conclude that the
differences seen in this intercomparison are scarcely large enough to be significant. They
examined the measurement process for physical effects which would bias the data in the direction
observed. Two effects, molecular diffusion and absorption of N20 by water vapor, were
eliminated because:
(a) If molecular diffusion were important in air infiltration, air change rates measured
with light gases which have higher thermal speeds would be higher than those measured with
heavy gases. This is the opposite of what was observed.
(b) If a significant amount of N20 were absorbed by water vapor in the test space, air
exchange rates measured with N20 would tend to be larger than those using SF6. Again, the
opposite result was actually seen.
Another possibility was considered; i.e., settling of the tracer gas in the test space. Since the
tracer was sampled in the return duct of the furnace, located in the ceiling in the test space,
settling of the heavy gas during the course of the measurement would appear to increase the air
exchange rate measured using a heavy tracer gas such as SF6.
Calculations showed that this was quite unlikely. The tracer gases were injected into the return
duct of a forced air heating system and are well mixed after a short time. From other results, a
mixing time of the order of 5 minutes was determined. Therefore, after 5 minutes the tracer gas
is well mixed throughout the test space - and this mixing continues throughout the concentration
decay measurement.
-------
TABLE 3
Air Change Rates
Test No.
N,0 (hfl\
cF?,fvn
SF/hr'l
Wind SnfWsl
At_£?Cl
Ratio*
1
0.49
-
-
4
1
-
2
-
0.75
-
5
-3
-
3
0.64
-
0.71(P)
4
-2
1.11
4
-
0.63
0.76(P)
6
-4
1.21
5
0.69
-
0.76(B)
7
7
1.10
6
0.91
-
-
8
6
-
7
-
0,89
0.94(B)
8
5
1.06
8
-
1.27
1.59(B)
9
4
1.25
9
1.25
-
1.19(B)
9
2
0.95
10
-
0.72
0.80(B)
7
3
1.11
11
0.51
-
-
4
6
-
12**
0.61
-
0.66(B)
6
-1
1.08
13
0.58
-
0.70(B)
3
-3
1.21
14
0.47
4
-6
* The ratio quoted is the air change rate of SF6 divided by the air change rate of the lighter tracer gas.
** During test 12, ethane (C^H^ was also used as a tracer gas. Hie air exchange rate measured was 0.68
hr"1, yielding a ratio of 0.97.
U
-------
Buoyancy effects were also considered: The fractional difference in density between a
macroscopic volume of gas containing air and one containing 1 ppb of SF6 is 4 x 10"9. Since the
acceleration due to buoyant forces is (Vp/p)g, the effective acceleration of the volume element
containing SFfi is 4 x 10"10 g. It would take about 3 hours for such an element to settle 2 m in still
air under an acceleration of that magnitude. However, since the furnace blower moves 6
volumes of house air through it each hour, forced mixing and convective mixing certainly
dominate buoyancy effects. The conclusion was that stratification due to the heavy SF6 molecule
is unlikely after the gas is initially mixed with room air.
Perhaps more space was consumed than is justified in describing the calculations involved, but
they do illustrate the many factors which must be considered in comparing the use of different
tracer gases.
Other possibilities exist: One may, in fact, be comparing instrumentation systems. The SF6 is
detected with electron capture gas chromatographic techniques, while the concentration of the
lighter gases is measured with infrared absorption techniques or the change in thermal
conductivity of helium/air mixtures.
Another possibility is that the absorption rate could have had a significant effect in the ppb range
while it would not be noticed in measurements in the ppm range.
The authors conclude that, while differences may be real, they should not preclude the use of one
gas in preference for another when choosing a tracer gas.
12
-------
By way of summary, variables which may affect the results when using a tracer gas to measure
ventilation rates are:
(1) Stratification of gases using a tracer gas of substantially higher density of air unless
adequate mixing exists.
(2) Changes in barometric pressure and wind velocity during the time of measurement.
(3) Absorption of the tracer gas on the surfaces to which the tracer is exposed.
(4) Error differences between type of detectors used; e.g., infrared absorption vs electron
capture gas chromatograph.
(5) Changes in relative humidity during measurements; e.g., N20 will react with water
vapor.
(6) At least in the case of H2 use as a tracer gas, diffusion may become important in
measuring ventilation rates. Grimsrud et al. (13) concluded that, in any tracer gas
measurements, one can expect a 5 - 10% uncertainty to exist.
Tracer Gas Techniques for Use in Measuring
Air Movement in Multicell Buildings
"The past decade has seen the development of many techniques utilizing tracer gases for
measuring air infiltration, ventilation, and airflow within buildings. These range from
measurements of air infiltration rate where the building is treated as a single uniformly mixed
zone to more complex measurement techniques that treat the building as several coupled zones
exchanging air with one another and with the outside.
13
-------
"Furthermore, some techniques give air flow rates averaged over a short time interval, while
others directly yield values averaged over an extensive periods of time -- from days to many
weeks.
"The various methods may be grouped into the following categories— tracer gas dilution, constant
injection of tracer gas, and maintaining a constant concentration of tracer gas. One or more
tracer gases may be used,"
"The most common approach for measuring air infiltration in buildings uses the dilution or decay
of a single tracer gas. This method assumes that the building interior can be treated as a single
uniformly mixed space within which the tracer gas concentration is everywhere the same. A
small quantity of a single tracer gas is injected into the building. As outside air leaks into the
building, the tracer gas concentration in the building air falls. The building's air infiltration rate
is determined from periodic measurements of the tracer gas concentration within the building.
"The tracer gas decay technique does not tell one the airflows in a building that cannot be treated
as a single uniformly mixed zone. These include large buildings lacking a single air circulation
system. Even for houses, one cannot reliably determine the air infiltration rate if there is no
forced air heating or central air conditioning system and where the interior doors are kept closed.
"The single tracer decay method can, however, be adapted to a building made up of two zones,
each of which may be treated as uniformly mixed. In this case, the tracer gas is injected into one
of the two zones and the concentration of the tracer gas is monitored in both zones. The airflow
-------
rates among the two zones and the outside can be calculated from the concentration
measurements" (22).
An important limitation of the tracer gas decay technique is that it yields a reliable measurement
only for a certain time interval following injection while the tracer gas concentration is high
enough to be reliably measured. Two ways of getting around this problem have been developed.
In one approach, an automated system repeats the injection at regular intervals. Air samples are
also automatically collected, analyzed, and recorded at regular intervals to determine the air
infiltration rate variation over extended periods of time (6).
Another approach involves injecting tracer gas continuously at a constant rate (the constant
injection method). Periodic sampling, analysis, and automatic recording of tracer gas
concentration permits direct measurement of long-term variations in air infiltration rates (23).
Instrumentation Used In Tracer Gas Work
Tracer gas dilution is a technique which involves releasing a certain amount of gas (usually in
the heating, ventilation, and air-conditioning (HVAC) return with the outdoor dampers shut while
the tracer is mixing with indoor air) into the space under investigation, allowing at least 30 min.
for mixing so that the tracer concentration is uniform throughout the space, and collecting air
samples as the gas concentration decays in accordance with the ASTM test for determining air
leakage rate by tracer dilution (E741-93). Usually the air samples are collected in sampling bags
or bottles (6,7) and analyzed later in the laboratory. There are several problems with this
-------
method of sampling, including the possible alteration of tracer gas concentration between
sampling and analysis, and the fact that, for each time period, samples must be collected
throughout the building. Alevantis (24) has described a system capable of sequentially monitoring
tracer gas decays at up to 16 different locations. The system allows on-site determination of the
overall ventilation rate under steady-state conditions (i.e., when the slopes of the decay curves
are equal).
At this time, (SF5) appears to be the most desirable tracer gas for near real-time, single-tracer
ventilation measurements (25,13). Some advantages of this non-toxic, relatively low-cost gas are
that it can be detected at very low concentrations (ppb and below) and that ambient SF6 levels
are low [< lppt (26)] with respect to concentrations used in this method. Concerns include the
ability of SF6 to mix well with the room air (SF6 has a higher molecular weight than normal
constituents of air) and the possibility that SF6 could decompose into harmful compounds if it
comes in contact with a high temperature surface (ASTM E741) such as a burner or the burning
end of a cigarette. At the present time it is not known what is a "safe" concentration of SF6 in
a building or what compounds result from its decomposition (10).
SF6 can be detected either with infrared methods or with gas chromatograph(GC)-electron capture
methods. Infrared instruments are very easy to set up, but are unfortunately capable of
monitoring SF6 concentrations only at ppm levels. The measurement results using the latter
method are in the form of chromatographic peaks, which must be individually integrated and
16
-------
converted into concentration numbers with the use of calibration data. Calibration of the electron
capture detector (ECD) is required before each test and after sufficient time (30 minutes to 2
hours) has been allowed for the gas flow rates and the detector and column temperatures to
stabilize. Recalibration will be needed if the GC column or the ECD becomes contaminated
during the test or if the carrier gas flows or temperatures are changed, since the areas under the
chromatographic peaks depend on the temperature, flow rates, and cleanliness of the GC column
and ECD. Commercially available SF6 detection instruments are designed basically for laboratory
application, not for field use. Such systems are not capable of :
(a) Sampling from more than one location.
(b) Performing auto-calibration using more than one span gas.
(c) Easily interfacing with a computer so that acquired
data can be reduced.
(d) Venting the oxygen before it enters the ECD,
(e) Being easily transported and set up.
Alevantis (24) describes a unit capable of meeting all these requirements with the additional
advantage of being able to control all the parameters as frequently as the operator wishes.
Sampling and calibration valves are shown in Figure 1. Figure 2 shows a block diagram of the
SF6 computer-controlled system in the load mode. Figure 3 is a block diagram of the SF6
computer-controlled system in the inject mode. Figure 4 shows the ECD in block diagram
form.
17
-------
COlWOti OCTIXt TO MMfUHC VOCATION
to VMXVH nuT
10 mans
Man ic lacATtsm
TO &W tt.HH.tnc
LOOT <««• tig, J)
TO $2.4 j»pb
irt uki
to SAnmie uscaiiok
smu» rain» v*m
co#ur nut roi. ajie snow)
FIG. 1-
TO 1*4 ffh
SF4 tAKX
TW ItlllllO*
-------
AHJVSUJU
—<*£IXX£»C£
CUU£K7
VCLTACi-SOKTHCUtt
wjjuutu-mou&tc?
ms£i
-^.ru^ctwcr artm
mqUtROf-
*A#G£-$VlTCHHi
VOttACE
ATTEHJATO*
vqhagc ocrrur
coNvom
orrscr
COHTfcCL
FlG 4— Electron capture detector block diagram.
19
-------
Summary and Conclusions
(1) The use of tracer gases for the measurement of air infiltration into structures and
interzonal flows within a structure is not new. This technique has been investigated over the past
15 years.
(2) Numerous tracer gases have been used, among which are: sulfur hexafluoride,
hydrogen, carbon monoxide, carbon dioxide, nitrous oxide, and radioactive argon and krypton.
(3) Sulfur hexafluoride is the most common tracer gas of choice — primarily because its
presence may be accurately measured in the ppb range using electron capture/gas chromatography
techniques.
(4) Most of the other gases used may be accurately measured in the ppm range using
infrared technology.
(5 There are generally three types of methods used: tracer gas decay, constant
concentration, and constant injection.
(6) Investigations comparing tracer gases have led to the following conclusions:
(a) Even though sulfur hexafluoride is appreciably heavier than air, mixing is not a problem; and
(b) The inherent uncontrollable variables present in tracer gas work limit the accuracy of
determinations to +1-5% - 10%. There is thus no reason why one tracer gas should be selected
over another provided other criteria are met. In the case of hydrogen, diffusion of the gas through
the surrounding walls can pose a problem.
(7) Tracer gases may be used in air flow measurements in large buildings where the
building may be treated as several coupled zones. In such a case, the decay technique can still be
used by having the system repeat the injection at regular intervals.
20
-------
References
(1) Harrje, D. T, and Gadsby, K. J. (1989). "Airflow Measurement Techniques Applied to
Radon Mitigation Problems." Progress and Trends in Air Infiltration and Ventilation Research.
10th AIVC Conference, Dipoli, Finland.
(2) Turiel, I., et al. (1981). "The Effects of Reduced Ventilation on Indoor Air Quality in an
Office Building," Lawrence Berkeley Laboratory Report LBL-10479, University of California,
Berkeley.
(3) Hollowell, C.D., et al. (1980). "Building Ventilation and Indoor Air Quality," Lawrence
Berkeley Laboratory Report LBL-10391, University of California, Berkeley.
(4) Sterling, T.D., et al. (1983). "Air Quality in Public Buildings with Health Related
Complaints," ASHRAE Transactions, Vol. 89, Part 2A, pp 198-212.
(5) Silberstein, S., et al. (1983). "Air Exchange Rate Measurements in the National Archives
Building," National Bureau of Standards Report NBSIR 83-2770, Washington, D.C.
(6) Grot, R. A., et al. (1980). "Tracer Gas Automated Equipment Designed for Complex Building
Studies," First Air Infiltration Center (AIC) Conference: Air Infiltration and Measuring
Techniques, Berkshire, U.K., pp 103-128.
(7) ASTM (1993). "Standard Test Method for Determining Air Change in a Single Zone by
Means of Tracer Gas Dilution." ASTM E741-93.
(8) Harrje, D.T., et al. (1985). "Documenting Air Movements and Air Infiltration in Multicells
Using Various Tracer Techniques," ASHRAE Transactions, Vol. 91, Part 2.
(9) Dietz, R.N., et al. (1986). "Detailed Description and Performance of a Periluorocarbon
Tracer System for Building Ventilation and Air Exchange Measurements." Measured Air
Leakage of Buildings, ASTM STP 904, H.R. Trechsel and P.L. Lagus Eds., American Society
for Testing and Materials, Philadelphia, PA, pp 203-264.
(10) Nagda, N.L., et al. (1987). Guidelines for Monitoring Indoor Air Quality. Hemisphere
Publishing, New York.
(11) Lagus, P. and Persily, A.K. (1985). "A Review of Tracer Gas Techniques for Measuring
Airflows in Buildings," ASHRAE Transactions, Vol. 91, Part 2.
(12) Sinden, F.W. (1978). "Multi-Chamber Theory of Infiltration," Building and Environment,
Vol. 13, pp 21-28.
21
-------
(13) Grimsrud, D. T., et al. (1980). "An Intercomparison of Tracer Gases Used for Air
Infiltation Measurements. M ASHRAE Transactions. Vol. 86, No.l, pp 258-67.
(14) Hunt, C.M. and Burch, D.M. (1975). "Air Infiltration Measurements in a Four-Bedroom
Townhouse Using Sulfur Hexafluoride as a Tracer Gas." ASHRAE Transactions. Vol. 81,
No. 1, pp 186-201.
(15) Hitchin, E.R. and Wilson, C.B. (1967). "Review of Experimental Techniques for the
Investigation of Natural Ventilation in Buildings," Build. Science, Vol. 2, pp 59-82.
(16) Warner, C.G. (1940). "Measurements of the Ventilation of Dwellings," Journal of
Hygiene, Vol. 40, pp 125 - 153.
(17) Collins, B.G. and Smith, D.B. (1955). "Measurements of Ventilation Rates Using a
Radioactive Tracer," Journal of Inst. Heat.Vent.Eng., Vol. 23, pp 270-274.
(18) Howland, A.H., et al. (1960). "Measurements of Air Movements in a House Using a
Radioactive Tracer Gas," Journal of Inst. Heat. Vent. Eng., Vol. 28, pp 57-71.
(19) Lidwell, C.M. (1960). "The Evaluation of Ventilation," Journal of Hygiene, Vol. 58, pp
297-305.
(20) Howard, J.S. (1966). "Ventilation Measurements in Houses and the Influence of Wall
Ventilators," Building Science, Vol. 1, pp 251-257.
(21) Knoepke, John (1977). "Tracer- gas System Determines Flow Volume of Flue Gases,"
Chem. Eng. (N.Y.), Vol. 84, No.3, pp 91-94.
(22) Harrje, D.T., et al. (1985). "Documenting Air Movements and Air Infiltration in Multicell
Buildings Using Various Tracer Techniques," ASHRAE Transactions, Vol. 91, Part 2.
(23) Dietz, R.N. and Cote, E.A.(1982). "Air Infiltration Measurements in a Home Using a
Convenient Perfluorocarbon Tracer Technique," Environment International, Special Issue on
Indoor Air Pollution, Vol 8, No.l, pp 419-433.
(24) Alevantis, L.E. (1989). "A Computer-Controlled System for Measuring Ventilation Rates
of Buildings Using Sulfur Hexafluoride as a Tracer Gas." ASTM Spec. Tech. Publ. Vol. 1002.
(25) Bassett, M.R., et al. (1981). "An Appraisal of the Sulfur Hexafluoride Decay Technique
for Measuring Air Infiltration Rates in Buildings," ASHRAE Transactions, Vol.87, Part 2, pp
361-371.
(26) Grot, R.A., et al. (1980). "Tracer Gas Automated Equipment Designed for Complex
Building Studies," First Air Infiltration Center (AIC) Conference: Air Infiltration and Measuring
Techniques, Berkshire, United Kingdom, pp 103 - 128.
22
-------
Bibliography
(The Following Are Not Used As References
But Contain Pertinent Data)
ASTM Standard El554-94 (1994 ). "Standard Test Methods for Determining External Air
Leakage of Air Distribution Systems by Fan Pressurization."
Bohae , D.L. (1986). "The Use of Constant Concentration Tracer Gas System to Measure
Ventilation in Buildings, " Report No. 205, Princeton University Center for Energy and
Environmental Studies, Princeton, NJ.
Bohac, D.L., et al. (1987). "Field Study Comparisons of Constant Concentration and PFT
Infiltration Measurements." Proceedings of the 8th AIVC Conference - Ventilation Technology
Research and Application, Supplement, AIVC Bracknell, Berkshire, Great Britain , Document
AIC-PROC-8-5-87, pp 47-62.
Clark, T.L., et al. (1989). "Comparison of Modeled and Measured Tracer Gas Concentrations
During the 'Across North America Tracer Experiment (ANATEX)1." NATO Challenges Mod.
Soc. Vol. 13. Air Pollut. Model. Its Appl. p 307.
Dabberdt, W.F. and Dietz, R.N., "Gaseous Tracer Technology and Applications." Probing the
Atmospheric Boundary Layer, D.H. Lenschow, Ed. American Meteorological Society, Boston,
MA, pp 103-128.
Flanagan, B.S., et al. (1976). "A Microprocessor Controlled Infiltrometer Using Sulfur
Hexafluoride Tracer Gas." In strum. Aerosp. Ind., Vol. 22, pp 705-712.
Fortmann, R.C., et al. (1989). "A Multiple Tracer System for Real-Time Measurement of
Interzonal Airflows in Residences." ASTM Spec. Publ., Vol. 1002, pp 287-297.
Grimsrud, D.T., et al. (1989). "An Intercomparison of Tracer Gases Used for Air Infiltration
Measurements." ASHRAE Transactions, Vol. 86, Part 1.
Harrje, D.T., et al. (1982). "Sampling for Air Exchange: Rates In a Variety of Buildings."
ASHRAE Transactions, Vol. 88, pp 1373-1384.
Harrje, D.T., et al. (1990). "Tracer Gas Measurement Systems Compared in a Multifamily
Building." Air Change Rate and Air Tightness in Buildings, ASTM STP 1067, M. H.Sherman,
ed. American Society for Testing and Materials, Philadelphia, PA.
23
-------
Jenkins, A. and Cornell, R.C. U.S. Reissue Patent Re 28, 417 (1975). (Original Filed Aug. 13,
1970.) Reissue of Patent U.S. 3,699,3342. "Apparatus and Method for Detecting Tracer Gas."
Liddament, M. and Thompson, C. (1983). "Techniques and Instrumentation for the
Measurement of Air Infiltration in Buildings - A Brief Review and Annotated Bibliography," Air
Infiltration Centre Document AIC-TN-10-83.
Niemela, R., et al. (1991). "Comparison of Three Tracer Gases for Determining Ventilation
Effectiveness and Capture Efficiency," Ann. Occup. Hyg., Vol. 35, No.4, pp 405-417.
Pitts, W.M. and McCaffrey, B.J. (1986). "Response Behavior of Hot Wires and Films to Flows
of Different Gases." J. Fluid Mech, Vol. 169, pp 465-512.
Sherman, M. and Dickerhoff, D. (1989). "Description of the LBL Multitracer Measurement
System." Proceedings, Fourth Building Thermal Envelope Conference, ASHRAE, Atlanta, GA.
Vinson, R.P., et al. (1981). "Sulfur Hexafluoride Tracer Gas Tests of Bagging-Machine Hood
Enclosures." Rep. Invest. - U.S.Bureau of Mines, Number RI 8527.
24
-------
TECHNICAL REPORT DATA
(Phase read Instructions on the reverse before comp
1. REPORT No.
E PA ~ 600 / R- 9 5- 013
4. TITLE AND SUBTITLE
Air Infiltration Measurements Usin^
A Literature Review
Tracer Gases:
PB95-173225
5. REPORT DATE
January 1995
6. PERFORMING ORGANIZATION CODE
7. AUTBOBESi
Max M. Sainfield
8. PERFORMING ORGANIZATION REPORT NO.
9. PERFORMING ORGANIZATION NAME AND ADDRESS
Max M. Samfield, Consultant
915 W. Knox Street
Durham, North Carolina 27701
10. PROGRAM ELEMENT NO.
11. CONTRACT/GRANT NO.
EPA Purchase Order
3D 1279 NATA
12. SPONSORING AGENCY NAME AND ADDRESS
EPA, Office of Research and Development
Air and Energy Engineering Research Laboratory
Research Triangle Park. NC 27711
13. TYPE OF REPORT AND PEFilOD COVERED
Final; 5-7/93
14. SPONSORING AGENCY CODE
EPA/600/13
15. SUPPLEMENTARY NOTES AERRL pr0ject officer iS David C.
541-2979.
Sanchez, Mail Drop 54, 919/
16. ABSTRACT
The report gives results of a literature review of air infiltration measure-
ments using tracer gases, including sulfur hexafluoride, hydrogen, carbon monoxide,
carbon dioxide, nitrous oxide, and radioactive argon and krypton. Sulfur hexafluoride
is the commonest' tracer gas of choice, primarily because its presence may be accu-
rately measured in the parts per billion range using electron capture/gas.chromato-
graphy techniques. Most of the other gases used may be accurately measured in the
parts per million range using infrared technology.- There are three basic types of
measurements: tracer gas decay, constant concentration, and constant injection. In-
vestigators comparing tracer gases conclude that: (a) even though sulfur hexafluoride
is appreciably heavier than air, mixing is not a problem; and (b) the inherent uncon-
trollable variables present in tracer gas work limit the accuracy of determinations
to +/- 5 to 10%. Thus, if all other criteria are met, there is no reason why one tra-
cer gas should be selected over another. The report describes a computer-control-
led injection system.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b. IDENTIFIERS/OPEN ENDED TERMS
c. COSATI Field/Gioup
Pollution
Measurement
Fluid Infiltration
Sulfur Hexafluoride
Gases
Pollution Control
Stationary Sources
Tracer Gases
13 B
14B
20D
07B
07D
18. DISTRIBUTION STATEMENT
Release to Public
19. SECURITY CLASS (This Report)
Unclassified
21. NO. OF PAGES
20. SECURITY CLASS (This page)
Unclassified
22, PRICE
EPA Form 2220-1 (9-73}
------- |