f> <<
?,-
"i
7 .HgjjSSBgEr.AI. THE -73rd. AHHUAL MZEE3D OF TEE AIH 50LHJTION COHTSOL.
22r27,' -JcfiO.AMOMSZAL, CAHADA., Paper 30^1.6
TOTAL HUMAN EXPOSURE TO AIR POULUTION
James L. Repace
Wayne R. Ott"
Lance A. Wallace
80-61.6
U.S. EPA Headquarters Library
Mail code 3201
1200 Pennsylvania Avenue NW
Washington DC 20460
U.. S. En v i ro nment a 1 P rot act 1 q n A gen cy
Washington, D. C-
U.S. Environmental Protection Agency-
Library, Room 2404 PM-211-A
401 M Street, S.W.
Washington, DC 20460
S >
-------
80.61.6
The Clean Air Act provides for the establishment of National Ambient
Air Quality Standards (NAAQS) in order to protect the public health from
the effects of air pollution. The U.S. Environmental Protection Agency (EPA), '
in its regulations, pursuant to the Act, has defined "ambient air" as "that
portion of fthe,,atmosphere, external to buildings, to which the general public
has access."];; .'.Consistent with this definition, state and local governments,
who have the primary.,responsibility for the prevention and control of air
pollution; have..established a network of outdoor air quality monitoring
stations that are located -at fixed sites. These monitoring stations sample
outdoor concentrations of various pollutants, which are analyzed over fixed
averaging periods specified by the^NAAQS, to determine compliance with these
standards within an Air Quality Control Region. Thus, .implicit in the process of
protecting public health from air pollution is the assumption that "exposure"
to air pollutants occurs only outdoors, and that measurements at fixed sites
are accurate ways of monitoring this exposure.
However, a number of studies have shown that fixed monitoring stations"
do not accurately reflect the exposure of the population to outdoor pollutants
in the outdoor environment.^-? Furthermore, various studies have indicated
that indoor exposures to air pollutants can be very important contributors to
air pollution burdens of the population.8-11 in this paper we will explore tJ
.'.factors affecting" an individual's "exposure" by modeling and by measurement,
'and discuss research needs and implications for regulatory action.
t
TOTAL EXPOSURE: A CONCEPT FOR PROTECTING PUBLIC HEALTH
Any true assessment of the impact of air pollution on public health must
take into account all the combined events making up an individual's exposure:
namely, his daily and hourly activities. The individual is exposed to
different levels of carbon monoxide and particulate matter when driving a car
to work, walking to the office cr factory, spending a day in a congested
office with smokers, going to a crowded restaurant at lunch, returning in
heavy traffic during the evening commute, shopping in crowded 'stores'after
work, and spending the evening in a home with gas appliances and- other
members cf the family who are smokers. An individual's exposure on a
given day is the weighted sum of all of these exposure events.
Although the ten'n "exposure" is coming into frequent use in the
environmental community, it is often defined^-14 imprecisely.
To avoid the ambiguity occurring when .outdoor fixed-station measurements
are used as a surrogate for actual exposures to air pollution, we shall.
adept the statistical definition of "exposure" suggested by Ott.^ Here,
an "exposure" to a pollutant is defined as.an event in which "a person comes
into contact with the pollutant." Thus, if we consider pollutant concentration
at some location j to be a random variable C(j) = c, and if person i is
present at this same location, we say that "person i is exposed to concentrate
-------
80-61.6
-2-
c" at that instant of time. Thus, "an exposure" can be viewed probabilistically
as the intersection (i.e., joint occurrence) of two events: .
Person i is present at location, j > n{c(j)' = c 1
Here, the brackets denote the occurrence of a particular event- Similarly,
if N is a random variable denoting the number of persons present at location
j, we say that N ป n persons are exposed to the concentration c when the
following events occur jointly:
/N ป n persons are present at location jj-H^Ctj) = c>
- For the joint occurrence to take place, the pollutant must come into actual
contact with the physical boundary of the person. Note that nothing is
said, however, about whether the pollutant actually crosses this boundary.
If the pollutant does cross the boundary, then a "dose" is said to occur.
Thus, one can have an exposure without a dose, but not a dose without an
exposure. If person i encounters a^ series of exposures over time, we speak
about his "integrated exposure," C-j, which is the sum of the individual
exposures, weighted by their durations t;
With the above definition as a guide, we see that there are only two
possible methods for determining the total exposure of a person (or of the
population) to air pollution: (1) modeling, which considers both the
activities of persons as a function of time and the concentrations to which
they are exposed, and (2) field studies, which utilize personal air pollution
monitors to.measure an individual's exposure (or the exposures of a sufficiently
large population sample).
' . . MODELING EXPOSURE
Total ExDOSure .
' Fugas -^ made one of the first attempts to compute total exposures-
from real data obtained from official ai-r .monitoring stations in the city,
and from measurements taken at the breathing zone in several streets during
business hours, indoors close to the streets, and in the countryside. By
estimating .the time spent by city residents in five different locations
home, work, street No. 1, street No. 2, and the countryside Fugas
calculated the "weighted weekly exposure" (WWE) for sulfur dioxide,
lead, and manganese (see Table 1). An intermediate computation is the
"integrated exposure," which is the product of the average concentration
and the time over which it occurs.
-------
80.61.6
-3-
has extended the approach suggested by Fugas and has substituted
the term "microenvironment types" for the "locations" used to compute
weighted weekly exposure. In this model, an individual's integrated
exposure over some time period (for example, a week) is computed as a weighted
average of the exposures from various microenvironment types, weighted by the
proportion of time spent in each microenvironment type:
where
k-l
integrated exposure of -the
during the j^ time period
ith individual
cijk
average concentration in the kth microenvironment type
during the jth time period
activity pattern coefficient denoting the time the ith
individual spent* in the kth microenvironment type during
the jth time period
Moschandreas and Morse^ have suggested an analogous approach for com-
puting air pollution exposures and have applied this approach to real data.
They introduce the application of "mobility patterns," which are designed
to capture "...the daily movements of population as they move to and from .
work, from home to points of amusement, adventure, business, and so on."
By examining the existing literature on activity patterns and "time
budgets" (discussed in the following section), they arrive at the estimated
time that all persons all races, ages, socioeconomic groups, workers,
students, etc. spend in various "environmental modes" (see Table 2). An
"environmental mode" is analogous to the /'microenvironment type" suggested
by Duan^' and the "locations" used by Fugas.16 Overall, the population
spends 72.3% of the time inside their homes, but these figures will differ
greatly for different population subgroups (workers, children, the elderly,
etc.)- By reviewing 35 oublications on human activities, for example,
Moschandreas and Morse conclude that housewives spend 3.0% of the 24-hour
period shopping, while fully employed persons spend only 1.8% of the day
shopping.
3y considering data on typical diurnal ozone concentrations for three
environmental modes (residential, office, and indoors) from a GEGMET study
of the Boston area, and by combining these data with estimates of the per-
centage of the population within each environmental mode as a function of
-------
80-61.5
-4- '
time, Moschandreas and Morse^ estimate the percentage of the population
exposed to ozone concentrations of 80 parts per billion or more.
The approaches of Fugas, Duan, and Moschandreas and Morse yield the-
populatton's average exposure over some specified time period. However,
a problem arises from emphasis upon the arithmetic mean. Within any
given time period, some people will exhibit combinations of activities
that result in exposures much higher than the mean value, while for
others, exposures will be much lower than the mean values. In addition,
the time spent in each activity will vary from day to day and from person
to person. Thus, the mean value is not adequate to characterize the
highest exposures of members of the population, and the variance of
exposures also must be considered. Ideally, what is needed is an effective .
technique for determining the entire frequency distribution of exposures
of the population to air pollution-
Ott^ is developing a computer simulation model of human exposures to
air pollution that considers the movements of each person in a metropolitan
area as a series of transitions from one "microenvironment" to another. A
microenvironment is analogous to the "environmental mode" of Moschandreas
and Morse,'ฐ the "microenvironment type11 cf Duan,'? and the "locations" of
Fugas.16 The resulting computer program, called Simulation of Human Air
Pollution'Exposures (SHAPE), uses probability distributions of the time,-that
people spend in each microenvironment; these probability distributions will be
derived from studies of human activity patterns. Within each microenvironment,
the concentration to" which an' individual is exposed is treated stochastically,
using distributional models that will be based on field studies of air pollutant
concentrations .reported in the research literature. In the simulation, the
computer keeps track of the exposure received as each person moves forward in
time and occupies successive microenvironments.
These studies illustrate the need for improved data on the time spent by
members of the public in various locations and activities throughout the day.
A. considerable body of literature now exists on human activity patterns and
"time budgets" that is.derived from large population surveys using
questionnaires and home interviews. .For example, Szalai^ has surveyed
the activity patterns of 29,392 persons in 12 countries. Ott^i currently
is preparing a 'literature review of these studies, with particular emphasis
on their applicability, to air pollution exposure estimation. 3y reviewing
data in 44 U. S. cities on the time spent by the population in various
locations, 20-22 ancj by introducing various assumptions, Ott^'
estimates that, on the average, employed men spend 90% of the day (21.7
hours out of 24 hours) indoors, while housewives spend 95% of the day
(22.8 hours) indoors. Overall, employed men in the United States are
estimated to spend 2.9% of the day (0.7 hours) in transit, while housewives
are estimated to spend 1.7% (0.4 hours) in transit. Unfortunately, more
-------
80-61.5
-5-
detailed estimates of exposure are very difficult to extract from these
studies, because the original investigators were not seeking information
relevant to air pollution. Thus, future research should concentrate on
filling the gaps in earlier studies of activity patterns. Once these
data are collected, they can be used to construct increasingly accurate
models for estimating the frequency distribution of human exposures to
air pollution. Table 3 lists some of the questions on Which future studies
should collect information. The time budget, coupled with information about
the microenvironment, allows the modeling of exposures of the population.
If there are known sources of pollutants in the microenvironment which have a
known frequency of use, and if information about the volume of the space and
air exchange rate is available, the concentration of the pollutant can be
calculated. For each of the indoor spaces given in Table 3, design occupancy
and ventilation rate can be obtained from ASHRAE Tables22a. Thus, exposures
in the indoor and in transit environment, as well as the outdoor environment,
can be modeled, given the appropriate diffusion equation for the pollutant.
Indoor Exposure
In modeling exposure in certain indoor microenvircnments, account must
be taken of not only-the frequency of the exposed population, but the
frequency distribution of certain human activities which affect the rate of
generation or removal of certain pollutants. -
Repace and Lowrey^ have demonstrated a model for estimating the exposure
of a nonsmoker to respirable participate (RSP) from cigarette smoking as a
function of design building occupancy and effective ventilation rates. Repace
and lowrey3 estimate that, relative to nonsmokers exposed only to the background
ambient of an airshed in compliance with the annual secondary TSP NAAQS of
60 ug/rn^, some nonsmokers may experience exposures more than 15 times greater
due to indoor air pollution from cigarette smoke. The model estimated
concentrations are in good agreement with RS? levels from smoking obtained
in a field survey of 52 indoor, outdoor, and in transit microenvironments.
Moschandreas and Stark^ have demonstrated an indoor-outdoor air pollution
model that predicts exposure to indoor air pollution by simulating interactions
involving outdoor pollutant levels, structural characteristics of dwellings,
and behavioural patterns of the inhabitants. The model appears to yield
reasonable predictions (within 25%) of indoor levels of COj, CO, NO, N02,.
and nonrhethane hydrocarbons.
'Kusuda, Silberstein, and McNall^ have developed a time-dependent model
for estimating indoor concentrations of radon and daughters in buildings as
a function of radon emission rates, air infiltration rates and the frequency
of opening of windows. Anderson, Lundqvist, and Molhave^ have developed a
-------
-6- 80-61.6
model which permits the calculation of the.concentration of formaldehyde
emitted from particle board in residences, given the quantity of particle
board and the ventilation rate. Agreement with measurements is very good.
They conclude that the use of particle board in its present form, containing
urea-formaldehyde glue, may result in higher indoor formaldehyde concentrations
than permitted for continuous outdoor exposure. They argue that air quality
standards and control procedures should be established for indoor.air in the home.
Such models, when coupled with the frequency distribution of human
activity patterns, may serve to give good estimates of the human exposure . '
to certain pollutants, particularly those, (such as radon, tobacco smoke, .
formaldehyde, and gas stove emissions) that are virtually encountered only
indoors.
MEASURING EXPOSURE
Personal monitoring of air pollutants can be used as an alternative
to modeling, or to validate models. Coupled with monitoring at fixed
stations including homes, offices, and other areas personal monitors
may provide much more precise estimates of total exposure of whole"populations
than modeling. Such estimates would give greater support to epidemiologic
studies and regulatory actions. Because personal monitors were not
available in the past, and because of the mistaken assumption that outdoor
exposures characterized total exposures, personal monitors have not been
deployed in populations large enough statistically to-'draw general. inferences
about the total exposure of the entire population. Lee and Mage?ฐ
recently surveyed available physiological, chemical, and physical personal
monitors. However, we are not aware of a compact summary of commercially
available personal monitors and their cost, capabilities, and limitations.
This is presented in Table 4. In addition, Table 5 lists experimental
personal" monitors currently being developed or deployed in research
studies.
As new monitors are developed, whether by private companies or government-
sponsored research, it is important that they be evaluated promptly by an
independent testing organization such as The National Bureau of Standards (NBS),
the National Institute of Occupational Safety and Health {NIQSH}, or EPA.
Many commercial instruments have not undergone this testing. As a result,
problems with precision, accuracy, interferences,- shelf life, fragility, and
span instability have'been allowed to persist undiscovered except in piecemeal
fashion "by individual users. Recent government efforts to rectify this
situation have included (1) EPA-sponsored evaluations at NBS of the Palmes
diffusion tube, the West permeable membrane system (for N02), the TSI Piezobalance,
the Harvard cyclone system (for respirable particulates) and (2) NIOSH evaluations
of Dupont, 3M, and Abcor badges (for organics) and the Reiszner Mini-Monitor
for vinyl chloride."
In considering total exposure to-pollutants, pathways into the body other
than air may be significant. Therefore, analytical -methods are needed
to detarmine'the concentrations of certain pollutants in air and drinking
water, as well as a much longer list of pollutants in food. The Food and
Drug Admin-i strati on can measure- only about half of the 129 pollutants in . .
food about which there is great concern. Some pollutants can be measured
-------
-7-
in certain food groups,' such as beverages or fatty foods, but not in others.
Methods for measuring many pollutants in body fluids {breath, blood,
urine) also need to be developed.
Sampling techniques and equipment capable of accompanying individuals on
their daily rounds need to be developed. Active monitors require rugged,
quiet, light pumps with dependable flow rates and long pumping periods
(24 hrs. minimum) on a single battery charge. Such monitors also require
impactors or cyclones designed to make dependable size cuts for a variety of
aerosols. Sampling heads need to be designed to be independent of air
flow velocity. Microcomputer technology must be further developed to
provide readout and storage capabilities for real-time measurements.
Development of passive badge-type monitors requires solutions to the
problems of stagnant boundary layers and interferences. Finally, automation
rs needed to permit the wearer of a personal monitor to record, with a minimum
of effort, the time at which a given exposure occurred, and information
about his activity during the exppsure.
Deployment of Personal Monitors
Although there have been no large scale (greater than 100 persons) studies
of total exposure using personal monitors, several small scale studies are
now underway. To illustrate of the kind of data that such studies will
generate and the conclusions that can be drawn from the data, we deployed
personal monitors on a very limited scale to measure actual exposures to RSP
and carbon monoxide.
Figure 1 shows a 24-hr, total RSP exposure measured by one of the authors,
James Repace, using a TSI model 3500 Piezobalance (each data point represents
several 2-minute averages). On October 16, 1979, he commutes from his
residence in suburban Maryland to EPA Headquarters in downtown Washington, D. C,
His office is smoke-free. After that, he journeys to the Goddard Space
Flight Center in suburban Maryland, a trip made behind a smoky diesel
truck. Lunch in a cafeteria is followed by a short tour of various
buildings. The RSP level in the smoking section of the cafeteria is 55%
greater than in the non-smoking section. On the return trip to Washington,
RS? levels are much lower in the absence of diesel exhaust. Another
period in a smoke-free office is followed by second encounter with a
puff of diesel exhaust on the city sidewalk. The commute back to his
suburban residence is followed by jogging in the outdoors and then dinner.
The highest RS? levels'of the day are encountered during preparation of
dinner, despite open windows in the kitchen and living room, and despite-
a ceil ing-mounted exhaust fan operating in the kitchen during the cooking
period. Elevated levels of RSP from cooking persist in the residence'
for several hours and permeate other rooms.
Several tentative generalizations may be drawn from this experiment.
The indoor RS? concentrations appear to be higher than those in transit,
which in turn ara higher than those in the .outdoors. Diesel-powered vehicles
appear to produce higher RSP levels than do gasoline powered vehicles.
-------
-8-
The time budget reveals that Repace spent 84% of his time indoors, 9% in
transit, and 7% outdoors. His total (24-hour integrated) RSP exposure
for this day is 1428 microgram-hours per cubic meter (ug-hrs/m3). Contri-
butions to the total exposure break down into about 82.3% from exposure in indoor I
microenvironments, 9.8% from in transit microenvironments, and 7.9% from I
outdoor microenvironments. .
Figure 2 shows another 24-hr exposure determined using a portable RSP
monitor on June 13-14, 1979 for the same subject. With more windows open
in the house, the RSP levels from cooking are considerably lower. Levels
of several hundred micrograms per cubic meter are encountered in transit
in the subway (Metro) station; investigation of the chemical nature of this
aerosol is planned. Elevated levels encountered in the office early in
the morning are from infiltration of smoke from an adjacent office;
levels are not elevated in the afternoon due to the absence of smoking.
James Repace1s time budget for this day in June shows that he spent 71.8%
time indoors 10.5% in transit, and 17.7% .of his time outdoors. The
calculation of total.RSP exposure for the 24-hour period is 1043 ug-hrs/rn2,
73% of his total exposure for the day in October. For June 13-14, 71.2%
of his total exposure comes from indoor microenvironments, 15.8% from in
transit microenvironments, and 13% from outdoor microenvironments.
Figure 3, from a paper by Ott^, shows the daytime CO exposure
measured-on Cheryl Boyter, an. EPA employee who consented to be a research
subject. She was provided with an Energetics Science (ESI) Series 9000
personal "dosimeter," a DuPont miniaturized portable pump,, and an ESI portable'
readout device. She carried these instruments along with her nn her daily '
activities and wrote down her activities, the time they occurred, and the
concentrations.
On June 28, 1979, she 1 eaves-Manassas at 6:06.a.m. on her way to EPA
headquarters. Prior to leaving her house, her CO exposure is negligible.
It begins to increase rapidly after.she leaves the "Main Gate" in Manassas,
exceeding 9 part-per-mi 11 ion (ppm) (the NAAQS 8-hr level) as soon as she
reaches Route 66, a major artery into Washington. Thereafter, the levels
vary between 10 and 20 ppm, with a peak of 22 ppm when she passes automobiles
waiting in line at a gasoline station near EPA headquarters. Her highest
exposure, 27 ppm, occurs not while she is driving, but when she leaves her
car and walks, in the EPA garage under the EPA building.
Once inside her office, the levels drop to about 2 ppm, remaining fairly
constant with time, and she turns the instrument off. Her office is well-
ventilated, and it is probable that, in a less-ventilatad office with many
smokers, exposure during the work hours would be higher. Her return commute
shows a pattern similar to her morning commute, except in reverse order.
Her highest exposure, 50 ppm, occurs while she waits briefly at the collection
booth inside the EPA garage, the levels in commuting traffic once again are
in the range of 10-20 ppm, and as she departs Route 56 and enters the neighbor-
hood streets of Manassas, CO concentrations decline below 9 ppm.
-------
-9-
Her highest exposures occurred in the EPA garage. However, they were of
relatively short duration, and hence contributed a relatively small portion
of the total exposure. Her overall running average exposure for the day was
less than the CO NAAQS of 9 ppm for 8 hours. This can be seen by noting
that the area under the jagged CO exposure graph is always less than the
dashed rectangular area represented by 9 ppm for 8 hours (72 ppm-hours),
regardless of the time during which the average is taken. Nevertheless,
once again, for another pollutant, we see the-pattern repeated: the indoor
(garage) concentrations and in transit levels dominate the outdoor levels.
Based on integrated exposure the largest contribution to the total exposure
conies from the in transit microenvironment. Thus, the personal monitor
shows where controls are needed most to limit total exposure to CO.
Recently, several short-term studies of RSP and CO have been made on
a limited scale using personal monitors. Wallace28 has measured RSP
concentrations in a number of environments in the Washington, D. C.
area using the TSI Piezobalance. Figure 4 shows 5-day composite concentrations
in several common microenvironments. Outdoor concentrations were the
lowest of all environments tested. Figure 4 shows that a typical 24-
hour exposure would average close to 75 ug/me, the NAAQS for total
suspended particulates, even though the outdoor mean is less than half
that value. (No standard exists for RSP). This finding raemphasizes
the importance of measuring total exposure, particularly in the indoor
and in transit environments, to avoid drawing erroneous conclusions
about human exposure from the outdoor concentrations only.
Figure 5 compares the ranges (vertical bars) and average values (dots)
of RSP concentrations observed in two limited studies, one by Gage29 in China,
and the other by Wallace2** in Washington, D.C. Data from 40 measurements in
China are labeled on the figure, while data from several hundred measurements in
Washington, 0. C. are unlabeled. The highest levels of RSP are found
indoors in China, where smoking occurred, and indoors in several Washington
subway stations where smoking is prohibited. These high RS? levels in
the subway are similar to those obtained by Repace (Figure 2); the cause
is unknown. In general, RSP levels in China seem higher than in Washington,
D. C. for all environments tested. The sample size, however, is too
limited to permit general conclusions to be drawn about relative levels
of RSP in China and in Washington, D. C. It is clear that more personal
monitoring, studies of this kind are needed.
Wallace^, in a bus trip from Reston, Virginia to Washington, D. C., recorded
his CO exposure with the ESI 9000 personal CO monitor. Like the earlier exposure
data on CO (see Figure 3), Figure 6 shows that high CO exposures (5-35 ppm) occur
-------
-10-
while an Individual is in transit. The' integrated exposure shows a steady rise
during the early morning trip.
EPA recently completed a study30,31 of CO intrusion into more than
1000 school buses; police cars, and taxis. The study employed stain tubes
to screen all the vehicles for CO levels. A subset of 100 vehicles showing
higher concentrations was monitored using the ESI series 9000 personal.
"dosimeter." (Figure 7 shows the frequency distribution of concentrations
measured by the personal "dosimeters"). If the data from this study are
extrapolated to 8 hours, we would find that 5% of the vehicles would exceed
the Occupational Safety and Health Administration (OSHA) standard of 400
ppm-hrs. More than.15% of the vehicles exceeded the EPA 1-hour standard
of 35 'ppm, and over 50* exceed the 9 ppm 8-hr standard. These important
findings illustrate the discoveries that may be expected with increasing
use of personal monitors.
A major EPA field study just getting underway, "The Public Health
Initiative" is based on the concept of determining individual exposure
to selected pollutants.through air, food,'and drinking water. Individuals
in the study wi.ll be equipped with personal monitors consisting of commercially
available components (pump, filter, tubing, and polymer absorbent) slightly
modified to form a rugged individual air sampling system. The pump
collects 8-hour samples of organic vapors on a cartridge of Tenax absorbent;
which can then be analyzed by gas chromatography/mass spectrometry to
determine air exposures to about a dozen halocarbons, most of which are
carcinogenic. The same pollutants in those individuals' food and drinking
water will be simultaneously measured to determine the relative importance
of .the three routes of exposure. Pollutant levels in breath, blood and
urine will be determined to correlate exposure to dose.32
Contribution of Smoking to Exposure
Repace andLowrey^, using a TSI model 3500 portable monitor for RSP,
surveyed 52 indoor, outdoor, and in transit microenvironments in a systematic
study of the effects of cigarette smoking on RS? levels (see Figure 8)
The respirable particulate concentrations measured are plotted versus
the observed cigarette density. With no cigarette smoking, RS? concentrations
during five rush-hour auto trips ranged from 25 to 54 ug/m^. With no
cigarette smoking, RS? concentrations in five residences, five restaurants,
two libraries, a church, and an office ranged from 20 to 50 ug/irP.
With cigarette smoking, however, concentrations in" restaurants (E,H,K,L,M,N,P ,R,S)
cocktail lounges (C, F), bingo games (0, G), a dinner dance (8), a bowling
alley (I), a sports arena (0), a hospital waiting room (J), and a residence
(A), ranged from 90 to 700 ug/irr. In-an office building conference
room (datum not shown) a level of 2000 ug/u? was observed. The RS?
levels were directly proportional to the smoker density, and inversely
proportional to the effective ventilation rat6,5 RS?" readings outside
the smoking premises ranged from 22 to 60 ug/m-. This study raises serious
-------
-ii-
doubt about the ability of outdoor TSP monitoring stations to reflect
indoor levels, and it also impugns the results of epidemiological studies
based on outdoor TSP levels, since RSP levels outdoors may frequently be
negligible in comparison with indoor levels.
Any measurement of the total exposure of the population to air pollution
should account for the self-inflicted exposure of the cigarette-smoking
population. The range of this exposure may be determined from the mainstream
tar and nicotine levels published by the U. S. Federal Trade Commission33.
Tar and nicotine levels range from about 1 milligram(mg) to more than 30
ing. Figure 9 shows the enormous 24-hour average levels of respirable
particulate to Which cigarette smokers are exposed, as high as 100,000
ug/m3 (assuming a respiration rate of 20 mP/day). Clearly the NAAQS
for particulates does not protect smokers from more than a small fraction
of total exposure. Carboxyhemoglobin levels in smokers also appear to
be considerably in excess of the levels that the NAAQS for CO is designed
to prevent.34
IMPLICATIONS FOR CONTROL STRATEGIES
How can the total human exposure to air pollutants be minimized? In the
outdoor air environment, limits on emissions from stationary sources primarily
serve to control the outdoor exposure levels. In the in transit environment,
emission controls on mobile sources primarily serve to limit the exposure of
people in vehicles. However, at present, control of indoor air pollution is
haphazard. There has been no scientific effort, based on public health
grounds, to establish air pollution standards specifically for the indoor non-
industrial environment.35 Berk, Hollowell, and Lin 36 and the World
Health Organization37 have tabulated a larae number of indoor air pollution
sources. Currently radon38, formaldehyde,-9 gas-stove emissions40,
and tobacco smoked'37 have been mentioned as serious public health concerns.
Berk-et al ป3ฐ have suggested three approaches to control of indoor air
pollution: air-to-air heat exchangers which permit venting of contaminants
while minimizing thermal losses to the outdoor air, air circulation through
filtration devices, and sealing or eliminating the sources of pollutants. The
latter measure might also include restrictions on certain human activities, such
as smaking3ป42,43
The American Society of Heating, Refrigerating, and Air-Conditioning
Engineers (ASHRAE) sets design ventilation requirements for spaces intended
for human occupancy in the residential, commercial, and industrial areas.
These standards are codified under ASHRAE Standard 62-73, Standards for
Natural and Mechanical Vantilation.22a ASHRAE has reduced its minimum (health-
related) ventilation rates from 30 cubic feet per minute (CFM) per occupant in
1900 to 10 CFM in 1936 to 5 CFM in 1973. ASHRAE 62-73 also specified recommended
(we!fare-related) rates, but in 1975, ASKRAE Standard 90-75R reduced
these to the minimum, in the interest of energy conservation in new
buildings.44 Since ventilation systems may require more than half of
~he total energy consumed in buildings, 3ฐ>~5 they have become
popular targets for energy conservation measures. Two recent Department of.
-------
-12-
Energy (DOE) programs have proposed energy conservation measures which
may cause reductions in ventilation rates in both existing and new buildings^.
DOE has recognized the potential for indoor air quality problems and has
called for the establishment of indoor air quality standards46. Such
standards have been promulgated in-Japan.*' ASHRAE has also recognized
the indoor air quality problem, and has proposed a revision to 62-73.
However, neither DOE nor ASHRAE have a public health mission, posing an important
.leadership challenge for agencies concerned with public health.
CONCLUSIONS
Our study indicates a great need to consider the contributions to total
human exposure to air pollution from the indoor and in transit environments as
well as the outdoor environment. Modeling and field studies are valid alterna-
tive approaches to assessing the relative contributions. Models for concen-
trations of pollutants in indoor, in transit, and outdoor environments,
together with time budget studies, will allow the frequency distribution of
exposures of the population to be determined. Knowledge of this distribution
will permit rational exploration of alternative control strategies for the
limitation of total exposure. Measurements of total human exposures to air
pollution in field studies statistically large enough to draw inferences about.
the exposures of large populations in different regions will have several
advantages over the current limited system based on outdoor fixed monitoring
stations. They will-permit: Identification for the first time, of those
microenvironments associated with the highest exposures; Identification,
for the first time, of subgroups of the population receiving the highest
exposures; Determination by the public health decision-maker where controls
are necessary to. limit total exposures; Epidemiological studies concerning
health effects of pollutants to be based on precise determinations of total
exposure.
DISCLAIMER: The opinions given in this paper represent the views of the
authors and should not be construed as views of the U.S. Environmental
Protection Agency. The mention of trade names does not constitute endorse-
ment by the Agency or recommendations for use.
-------
-13-*
80-61.6
TABLE 1
Example by Fugas^ Illustrating Computation of
Weighted Weekly Exposure (WWE)
Type of
Exposure
Home
Work
Street 1
Street 2
Count rv-side
Total
WWE
Hours
Per Week
no
42
10
4
2
168
S02
c
89
a
.600
180
25
0
C'x-t
9790
336
6000
720
50
16896
101-
Pb
C
2.5
0.3
6.0
3.5
0.1
C x t
275
12.6
60
14
0.2
361.8
2.2
Mn
C
0.04
0.02
0.80
0.12
0.01
C x t
4.4
0.84
8.0
0.48 -
0.02
13.7^
0.08
Note: All values for C (concentration) are expressed in units of ug/ni3, and all
values for C x t (integrated exposure) are expressed in units of ug-hr/m3.
Table 2
Time Spent by the Overall Population in Each Environmental
Mode, As Estimated by Moschandreas and Morse'8
Environmental
Mode
Inside One1 s Home
In- transit
At Work
At other locations
Percent of
the Day
72.8
5.S
13 .9
7.4
-------
-14-
Table 3
QUESTIONS OF IMPORTANCE FOR AIR POLLUTION EXPOSURE
ESTIMATION, ADAPTED ROM
1. How long, and during what hours of the day, Is the person physically present
in each of the following microenvironments (given as examples)?
A. Indoors .
Residential
Windows Open or Closed
Commercial
Restaurant
Bar
Recreation center
Office
Waiting Room
Conference Room
Auditorium
Haircutting Salon
Public garage
Organizational
Church or temple
Legislative hall
Lodge hall
Industrial
Light manufacturing
Chemical
Heavy industrial
Agricultural
Food, processing
Institutional
School
Hospital or nursing home
Research institute
Military installation
B. Outdoors
Occupational
(Light, moderat e ,
or heavy work)
NonoccuDational
Resting, walking, or running
C. In transit "
Automobile
Bus
Truck (gas or diesel)
Train(subway or elevated)
Plane
Motorcycle
Boat
1. If the person is present on or near a roadway, is the traffic density low,
medium, or high?
3. Within each indoor or in transit microenvironment, are others smoking, and if.
so, how many?
4. Does the person's microenvironment include other known sources of air pollutants'
If so, what kind and during what hours are they operated?
5. What is the physical volume, air exchange rate, and occupancy of the
microenvironmerit (pertains mainly to indoor and enclosed in transit spaces)
if different from the design occupancy and ventilation?
-------
IV C
*J O
=3 O
A
J_>
e
1 VI
01 i-
* ซrt O
Zซ- = g
5M CSJ -" c
B3 (/> CO VI
*> 3 -"3
ซi o 2 o
ecu c o;
c
= i
S_ O
o m
O
O
in
ง
C3
in
o E
o 2
in a-
o
o
CO
o
CM
01
1
ai
ซ/)
0)
O
a.
. O OJ
ซ/>
feฃ go.'
* ฃ ฃ ง a
ซ c c w rr
-3
*
a
(t3 i-
O!
21
u
0)
o
Nl
^5 4
rs 0)
C. U
o "
u u
4-1 -5?
u ฃ
0) 0)
=
CJ
4^
I- U
+J -^
u =
a> ai
aj L. S*
111
a>
i^ซ.
u. o o
in t-
0) OJ
O
J-J
u
ซ
ซซ
O 4_ '-ซ
^- O >
CO
c
o
U
OJ
Genera
0)
u
u
in
I/I
u
u
wป
u
O)
4->
c
a.
s_
o
OJ
c
in
"oJ
o
D
u
us
o
o
"s
X
0)
o
i.
a
-------
en 01
<9 55
1_ '
Ol'i
o
Oi
in
U
o ^
(O
O
I/I
u
-o s-
*_. 01 O
ฐ S"1"*
,1 ฃV
r^ 01 O
> ฃ
0)
ป-s
tn
01
r p
ซJ5
a>
= "&
01
01
o
0) -^
p- U
en ฃ
= S
19 *>
i-> <^
u 2
^01 *
JS JJ O
5 if
= 2ฎ
o
V)
01
t/t l/>
OJ O)
U 3
01
iJ
(O
id
c.
.
"*w ST.
a] ^_. 01
01
4*1
19
*O
01
a
i
U1
U
.1
OJ U
N O
c ป
at Q
as. ~
(Q
O
o
01
ซ- ซ V,
ฐr ^
0 ซ
u
H3
i.
0
^J
O
s.
01
4^
'
c u
f3 (13
en
s_
u
"c
ซ
sn
^
c
1
4-J
ฃ
L.
a
^j
u
(*1
bf
"o
u
Jฃ
o
ui 0)
w >
E 4->
rc U
31 (C
U '
O
c
i
i/i
u
*e
Q
en
^
c
4>
&
>i
^^
0)
O
re
~
-------
-17-
REFERENCES
U C.F.R. 40, 1978, Part 50,
2. W. R. Ott, An Urban Survey Technique for Measuring the Spatial Variation
of Carbon Monoxide Concentrations in Cities, Ph.D. Dissertation, Department
of Civil Engineering, Stanford University, 1971.
3. A. Cortesi and J. D. Spengler, J. Air Pollut. Control Ass. 26.: 1144 (1976).
4. G. Godin, G. R. Wright, and R. J. Shepherd, "Urban Exposure to Carbon
Monoxide," Arch. Environ. Health .25.: 305 (1972).
5. G. R. Wright et. al, Arch. Environ. Health, 30.: 123 (.1975).
6. L. A. Wallace, "Use of Personal Monitor to Measure Commuter Exposure to
Carbon Monoxide in Vehicle Passenger Compartments," Paper No. 79-59.2
presented at the 72nd Annual Meeting of the Air Pollution Control
Association, Cincinnati, Ohio, June, 1979.
7. M. .Waldman, S, Weiss, and W. Articola, "A Study of the Health Effects of
Bicycling in an Urban Atmosphere," U. S. Department of Transportation,
'Office of Environmental Affairs, Report No. DOT-TES-78-001, October 1977.
8. J. L. Repace and A. H. Lowrey, "Indoor Air Pollution, Tobacco Smoke,
and Public Health," SCIENCE, to be published, 1980.
9. J. D. Spengler, B. G. Ferris, and D. W. Dockery, Environ. Sci. Techno!.
J3j 1236 (1979),
10. R. E. Binder et al., Arch. Environ. Health 31.: 277 (1976).
11. T. D. Sterling and D. M. Kobayashi, Environ. Res. 13:1(1977); T. D. Sterling
and E. Sterling, J. Air Poll. Contr. Ass. 29_: 238 (1979).
12. Y. Horie and A. C. Stern, "Analysis of Population Exposure to Air
Pollution in the New York-New Jersey-Connecticut Tri-State Region,"
U.S. Environmental Protection Agency, Research Triangle Park, N. C.,
EPA-450/3-76-027, March 1976.
13. N. H. Frank, F. Hunt, Jr., and M. Cox, "Population Exposure: An Indicator
of Air Quality Improvement," Paper No. 77-44.2 presented at the 70th
Annual Meeting of the Air Pollution Control Association, Toronto, Ontario,
Canada, June 1977.
14. Y. Horie, S. Chaplin, and 0. Helfenbein, "Population Exposure to
Oxidants and Nitrogen Dioxide in Los Angeles, Volume II: Weekday/Weekend
and Population Mobility Effects." U. S. Environmental Protection Agency,
Research Triangle Park", N. C., E?A-450/3-77-004b. January 1977
-------
" . -18-
15. W. R. Ott, "Concepts of Human Exposure,to Environmental Pollution," SIMS
Technical Report No. 32, Stanford University, Department of Statistics,
1980. .
16. M. Fugas, "Assessment of Total Exposure to an Air Pollutant," Proceedings
of the International Conference on Environmental Sensing and Assessment,
L,as Vegas, Nev. IEEE .#75-CH 1004-1 ICESA, September 1975.
17. N. Duan, "Microenvironment Types: A Model for Human Exposures to Air
Pollution," SIMS Technical Report, Stanford University, Department of
Statistics, to be published* .
18. D. J. Moschandreas and S. Morse, "Exposure Estimation and Mobility
Patterns," Paper No. 79-14.4 presented at the 72jtd Annual Meeting of the
Air Pollution Control Association, Cincinnati, Ohio, June 1979.
19. W. R. Ott, "Development of Activity Pattern Models for Human Exposure
Monitoring," U. S. Environmental Protection Agency, Office of Research
and Development, Washington, D.. C., Innovative Research Program Proposal,
approved July 1979.
20. A. Szalai, editor, The Use of Time: Daily Activities of Urban and
Suburban Populations in Twelve Countries, Mouton, -the-Hague, Paris, 1972.
21. W. R. Ott, "Human Activity Patterns: A Review of the Literature for Air
Pollution Exposure Estimation," U. S. Environmental Protection Agency,-
technical report,.to be published, 1980..
22. J. P.. Robinson, How Americans Use Time: A Social-Psychological Analysis
of Everyday Behavior, Praeger Publishers, Praeger Special Studies, New Yark,
^jj~ . ,
22a. ASHRAE Standard 62-73, American Society of Heating, Refrigerating, and
Air Conditioning Engineers, N. Y.., N. Y. 10017.,
23. D. J. Mcschandreas and J. W. C. Stark, 'The GEQMET Indoor-Outdoor Air
" Pollution Model" U. S. Environmental Protection Agency, Office of
Research and Development, Washington, D. C., EPA/600/7-78-106, 1978.
24. T. Kusuda, S. Silberstein, and P. E. McMall, Jr., "Modeling of Radon and
Daughter Concentrations in Ventilated Spaces," National Bureau of Standards,
to be published. - . -
25. I. Anderson, G. R. Lundqvist, and'L. Molhave, Atmos. Environ^ 9_: 1121 (1975),
26. R. E. Lee, Jr. and D. T. Mage, "Personal Exposure Monitors - A Survey" . .
paper No. 79-14.1, presen^d at the 72nd Annual Meeting of Air Pollution
Control Association, Cincinnati, Ohio, 1979.
27., National Bureau of Standards, yearly progress report, EPA-NBS IAG,
December 1979; Personal, communication, Mary Lynn Woebkenberg, National
Institute of Occupational Safety and Health, February 1980.
-------
-19-
28. L. Wallace, "Exposure to Respirable Particulates", unpublished.
29. Personal Communication, S. Sage, Assistant Administrator, Office of
Research and Development, U.S. Environmental Protection Agency,
Washington, D. C.
30. R. Ziskind, "Carbon Monoxide Intrusion into the Passenger Compartments of
Sustained-Use Vehicles," Science Applications Incorporated, Interim
Report to Congress, U. S. Environmental Protection Agency, Washington, D. C.
July 1978.
31. R. Ziskind, "Carbon Monoxide Intrusion into the Passenger Compartments of
Sustained-Use Vehicles," Final Report to Congress, U. S. Environmental
Protection Agency, May 1980.
32. Office of Research and Development, U. S. Environmental Protection
Agency, Washington, D. C.'
33. Federal Register, 45_ #7,'Thursday, Jan. 10, 1980, p. 2102, Federal Trade
Commission Report on Tar and Nicotine Content of Cigarettes.
34. A. Kahn et aT., "A Study of Carbon Monoxide Sources in the St. Louis Metroregion,"
Cuers Report #4, September 1975, Southern Illinois University, Edwardsville,
111.
35. J. E. McFadden, J. H. Beard, and D. J. Moschandreas, "Survey of Indoor
' Air Quality Health Criteria and Standards," EPA 600/7-7-027, 1978.
> ป '
36. J. V. Berk, C. D. Hollowell, and C.I. Lis, "Indoor Air Qual ity Measure-
ments in Energy Efficient Houses" paper 79-14.2 presented at the 72nd
Annual Meeting of the Air Pollution Control Association, Cincinnati, Ohio,
June, 1979.
37. "Health Aspects Related to Indoor Air Quality," Euro Report #21, 1979
, World Health Organization, Copenhagen, Denmark, 1979.
38. U. S. Environmental Protection Agency, Washington, D. C,, Office of
Radiation Programs, Guidance issued to Governor of Florida, "Maximum
Recommended Radon Concentrations in Homes," 1978.
39. P. A. Breysse, Environ. Health Safety News 2j[, (1977).
40. R. J. Melia and C.. Flory, Brit. Med. J. 2.: 149 (1977).
41. I. Tager, Amer. J. Epidemic!. 110; 15 (1979).
42. R. N. Kalika et a!., ASHRAE J. JJ.:44 (1970).
43. C.3. Barad, Occup. Health and Safety, 48_, No. 1: 21 (1979).
-------
-20-
44. J. E. Woods, "Energy Conservation Ventilation, and Acceptable Indoor
Air Quality," paper presented at Human Factors Society meeting Detroit,
Mich., Oct. 1978.
45. U. S. Department of Energy Washington, D.C., Reports: DOE/EIS-0050,
November 1979, DOE/EIS - 0061D, November 1979.^
46. Letter from M. Savitz, Acting Assistant Secretary for Conservation and
Solar Energy, U.S. Department of Energy, to D. Costle, Administrator, U. S,
Environmental Protection Agency, Washington, D. C., November 27, 1979.
47. Japanese Law #20',. April 14, 1970, "Sanitary Environment for Buildings."
Japanese Cabinet Order #304, October 1970.
-------
30-61.6
POLLUTANT CONCENTRATION, (ug/m3
m
O
^i
ง
m
m
O
3
3i
Tfl
'a
-------
80-61.6
POLLUTANT CONCENTRATION
c
o
D O O O O o O
i i 11.
a
i
ป
*
g ง
i i
M
f1^
I T
M)
o
O5
O
f
o>
o
T
CJ
0
o
5
z
9 a
m
O
O
5
O
c
o
o
o
"0
m
X
i"0
o
CO
c
a
m
s
m
O
(O
a
m
m
O
C
2
m
ro
(O
I'M
(0
(O
I
5 2
111
M 3
-------
80-61.6
CO CONCENTRATION (ppm)
01
10
O
NJ
01
<0 Ml
o o
ง0
g m
m x
a T
co o
o co
S O
o""
z o
o E
Qw
m
co
a
m
a
m
-1 f
o
o
8
i
s>
o
o
o
s
ง
a
o
o
s
>
i
(O
1
ง s 5
JS 5
1 1
CO CONCENTRATION (ppm)
22
-------
POLLUTANT CONCENTRATION
80-61.6
W
CO
0
A
ซ1
co
ui
IB
e
o
M
a
o
o
o
tn
II-
q > i
m
O
O rn
2. ฐ
O
Z
rn
m
m
2
33
m
2 ง
a
a
09
O
o
o
o
o
\>
b
o
If)
b
ง
o
C3
b
e
o
o
o
So
_ o
o
Z ^
tป
CJ
e
O
m
O
C
H
O
o
o
33
Crt
1 1
v<>
-sen
>CD
ง<
Jv
22
O
m
1 . V 1
o
c
31
m
-s >
I
P m
Po
r m
> x
si
i- r-
m
ss
5
m
01
ซ
o
a
8
S 'S. s
1
U)
-------
FIGURE 5.
LEVELS Of RESPIRABLE P ARTICULATES IN CHINA AND WASH. D.C.
SPRING AND SUMMER, 1379
440
400
360
320
280
240
>
2
Ml
U
I
ฃ 200
s
ฃ 160
_t
a
c
120
80
40
INDOORS
SMOKtNG
(CHINA)
1 AUTOS
(CHINA)
INDOORS T-
OUTDOORS NONSMOKING
(CHINA) (CHINA)
I
SUBWAY
PLATFORM
INDOORS
SMOKING
SUBWAY
CARS
AUTOS
DIESEL
INDOORS BUSES
NONSMOKING)
OUTDOORS
(SUBURBS)
1
INDOORS
(SUBURBAN
HOME)
i
-------
80-61.6
INSTANTANEOUS CO CONCENTRATION (ppm)
m
*
m
x
I
o
30
m
3
g
9
00
O
If
gl
O _
o
o
I i
3
| I
O
ffl
ON
f
o
ro
O
(O
CO
> o
ro
O
CO
o
.
INTEQRATED'GO.EXPOSURE (ppm-hrs)
24
-------
:.i :-..:V:".-v~~."T'.:.ii--1
.;.:'.: t^Js-" Jf i'^fijl' ^ v^'.
100
OSHA 8-HRSTWASTD
10
20 30 40 50
CARBON MONOXIDE CONCENTRATION, PPM
60
FIGURE 7. CUMULATIVE DISTRIBUTION OF PERSONAL SAMPLER READING
CO INTRUSION STUDY 30,31
-------
!
.1,*; ;r-" *'~**r-****?*:
"-* ~"r-v .
>ซ..-? ซซ?-r _ ,>yjฃ-
1000
ฃ900
IU
S
o .
I 800
o .
oc
Ul
1700
O "
ce
i 50ฐ
CO
i 500
Q
OL .
g 400
oc
Ul
CO
CD
O 300
200
100
R6URE 8, RESULTS OF A FIELD SURVEY OF RESPIRABUE
PART1CUIATE IN 52 MICROENVIRONMENTS,
FROW REPACE AND LOWREY*:
NAAQS 24-HR AV. SIGNIFICANT HARM LEVEL FOR TSP
NAAQS 24-HR AV. AIR POLLUTION EMERGENCY LEVEL FOR TSP
B Ds ESTIMATED
MEASURED DATA
ORT CALCULATED
AIR CHANGES PER HR
F
NAAQS 24-HR PRIMARY LEVEL FOR TSP
H*
NAAQS 24-HR SECONDARY LEVEL FOR TSP
M N
ปQ
SQ> 9.2 AIR CHANGES PER HR
NAAQS ANNUAL PRIMARY LEVEL FOR TSP
31 DATA POINTS
.'5 -
1.0
1.5
2.0
2.5
3.0
^.t..
3 *
w.v-
DISSERVED ACTIVE SMOKER DENSITY,
(BURNING CIGARETTES PER 100 CUBIC METERS)
-------
.....-.--." -.-: - "--.:'" -?' '.:":*"I
' ' *:!^ป.ซS^^^.*^^-^^J^!^.^:' '-r'*Z'' -ป
' 4^']M: ฃฃงฃ? i^f^l
'^:^B*:ai^3^.^fet^
;' T" 'ซ , :. /.*. . ,* -j^ .ป.'.'^!^^iป--- . ! ' .^i**'* iv'^'^
^^^^p^"^^:^:'7"S%?i'
- - JrJ;._1-4'.
.--.-'. ^- ''"~ Y-^-^-^^'^x:!^ :*:;':
:;ซฃ*-.. ..'.-"'> " 4Ci - ""MS*'.'?'
..-' /.****. SWP_.~ -".'ป"_ .. -v7r ""-..*.' I
-yr-:-: .;ป^..,iaw-fv:;"j
.- .v^^r---.'^-..-.*^---^-^!^-1- :r ?.V:v.-?:::
""". " f -*.-_' r/T.-.-T; I- .;-;?;. .-.i'.-ป-- _.'*-/_."^J
" .XZ-ill-ti 1 ' X*-*',"*> J,\**^".""''*"V'".=**
;..^:^ป^?^::^:ci::
. ,
.. ..
.. ... . . ..... __ ._-*v- - ; V
'.:^i;.- -./-" ;-^;T^ '**^7 *. "
**v"
.
-401 M Street.
Washington, D
200000
150000
100000
50000
n 20000
a
%
UJ
(C
3 10000
X
Ul
5000
LU
(9
-cc
UJ
^. 2000
CM
1000
500
200
100
Figure9. RANGE OF EXPOSURE OF SMOKERS
TO RESPIRABLE PARTICULATE
FROM CIGARETTES
24-HR SIGNIFICANT HARM LEVEL
24-HR PRIMARY NAAQS
AVERAGE ^^
TAR CIGARETTE
_L I t 1
34 5 6 7 8910
20 30
------- |