Environmental Health Criteria 214

HUMAN EXPOSURE ASSESSMENT
estimated 150 000 per filter) in subsequent scanning electron
microscope (SEM) analyses, but their mass does not appear to account
for more than 10% of the excess personal exposure (Mamane, 1992).
Mean PM 2.5 daytime concentrations were similar indoors
(48 µg/m 3 ) and outdoors (49 µg/m 3 ), but indoor concentrations fell
off during the sleeping period (36 µg/m 3 ) compared to 50 µg/m 3
outdoors. Thus the fine particle contribution of PM 10 concentrations
averaged about 51% during the day and 58% at night both indoors and
outdoors. Unweighted distributions are displayed in Fig. 37 for 24-h
average PM 10 personal, indoor and outdoor concentrations. About 25%
of the population of Riverside was estimated to have 24-h personal
PM 10 exposures exceeding the 150 µg/m 3 24-h US National Ambient Air
Quality Standard (NAAQS) for ambient air. Central-site PM 2.5 and PM 10
concentrations agreed well with backyard concentrations. Overall, the
data strongly suggest that a single central-site monitor can represent
well the PM 2.5 and PM 10 concentrations throughout a wider area such
as a town or small city, at least in the Los Angeles basin.
Stepwise regressions resulted in smoking, cooking, and either air
exchange rates or house volumes being added to outdoor concentrations
as significant predictors of personal exposure. Smoking added about 30
µg/m 3 to the total PM 2.5 concentrations. Cooking added 13 µg/m 3 to
the daytime PM 2.5 concentration, but was not significant during the
overnight period. At night, an increase in air exchange of one air
change per hour resulted in a small increase of about 4.5 µg/m 3 to
the PM 2.5 concentration, but was not significant during the day. The
house volume was not significant at night, but was significant during
the day, with larger homes resulting in smaller PM 2.5 concentrations.
Since air exchange and house volume were weakly correlated
(negatively), they were not included together in the same regression.
Following Koutrakis et al. (1992), a non-linear least squares
regression equation was used to estimate penetration factors, decay
rates and source strengths for particles and elements from both size
fractions (Ozkaynak et al.1996). Penetration factors were very close
to unity for nearly all particles and elements. The calculated decay
rate for fine particles (< PM 2.5 ) was 0.39 ± 0.16 h-1, and for PM 10
was 0.65 ± 0.28 h -1 . Since PM 10 contains the PM 2.5 fraction, a
separate calculation was made for the coarse particles (PM 10 - PM 2.5 )
with a resulting decay rate of 1.01 ± 0.43 h-1. Decay rates for
elements associated with the fine fraction were generally lower than
for elements associated with the coarse fraction, as would be
expected, due to their lower settling velocities. For example, sulfur,
which is associated with the fine fraction of aerosols in the form of
sulfate, had calculated decay rates of 0.16 ± 0.04 and 0.21 ± 0.04 h-1
for PM 2.5 and PM 10 fractions, respectively. The crustal elements
http://www.inchem.org/documents/ehc/ehc/ehc214.htm
Page 206 of 284
6/1/2007