indoor-outdoor air leakage of apartments and commercial buildings

L/ s m 2 at 50 Pa, and a geometric standard deviation (GSD) of about 2.3. Figure 1 shows a
histogram of the distribution of the logarithms (base 10) of the data.
0
r-
0
(D
0
l{)
(/J
0>
C
:2 0
·s '¢
0)
(3
Q; 0
.0 (')
E
::l
Z
0
0J
0
0.3 1 3 10 30
Air Flow at 50 Pa [Liters per second per square meter]
Figure 1 Histogram of air flow (liters per second per square meter of building shell) at 50 Pa indoor­
outdoor pressure difference, for the 267 buildings in the commercial buildings database. These data do
not constitute a representative sample of all commercial buildings. The distribution is approximately
lognormal, with a geometric mean (GM) of 4 LIs m 2 and a geometric standard deviation (GSD) of 2.3.
By contrast with the commercial building distribution, our recent analysis of the air
leakage of US single-family houses (Chan et al. 2005) found that the leakage follows a
lognormal distribution with a GM of 2.6 L/ (s·m 2 ), and a GSD of 1.6, at a 50 Pa indoor-outdoor
pressure difference. Thus, based on this cursory summary of the data, commercial buildings
seem to be somewhat leakier than single-family houses, and also to have leakiness that is more
variable than single-family homes.
Based on the published information about the buildings that were measured, we
classified each building according to usage (e.g. school, retail, etc.) and construction type
(masonry, steel frame, etc.). "Manufactured building" refers to trailers or portable structures.
Inevitably, there is some ambiguity in the classification of building usage and construction
types. Our classifications are based on those used in the original studies, but we had to interpret
some entries that did not perfectly match any of our categories. Table 2 summarizes the number
of buildings in each classification and construction type, and the fraction of the total database
18