United States
Environmental Protection
Agency
Office of Monitoring Systems and
Quality Assurance
Washington DC 20460
EPA-600/4-84-019
February 1984
Research and Development
&EPA
Field Surveys of
Carbon Monoxide in
Commercial Settings
Using Personal
Exposure Monitors
-------
Tech. Report No. EPA-600/4-84-019
February 1984
FIELD SURVEYS OF CARBON MONOXIDE
IN COMMERCIAL SETTINGS
USING PERSONAL EXPOSURE MONITORS
by
Peter G. Flachsbart"*"
Department of Urban and Regional Planning
University of Hawaii at Manoa
Honolulu, Hawaii 96822
and
Wayne R. Ott
U.S. Environmental Protection Agency
Office of Research and Development (RD-680)
401 M St. S.W.
Washington, D.C. 20460
+ Supported by Contract No. A-1598-NNLX, Office of Mobile Source Air
Pollution Control, U.S. Environmental Protection Agency, Washington, D.C.
20460.
Work completed while on assignment to the Department of Statistics,
Stanford University, under the EPA Innovative Research Program.
-------
DISCLAIMER
The Office of Research and Development of the U.S. Environmental
Protection Agency has reviewed this report and approved it for publication.
Approval does not signify that the contents necessarily reflect the views
and policies of the U.S. Environmental Protection Agency, nor does mention
of trade names or commercial products constitute endorsement or
recommendation for use.
11
-------
FOREWORD
Polluted air, foul water, and spoiled land offer tragic testimony to
the deterioration of environmental quality. The United States has responded
to the threat of environmental pollutants to health and welfare by creating
regulatory programs designed to control pollutants at their sources.
Physical, chemical, and biological measurements of environmental quality are
the most important yardsticks available for assessing the progress of these
programs in improving and protecting the quality of the environment.
Accurate data on the duration, magnitude, and physical distribution of
pollutants in the environment also are essential for determining the degree
of source control that is required and the manner in which control should be
implemented.
Past air pollution monitoring programs usually have relied on
measurements at fixed stations located outdoors. Unfortunately, outdoor
measurements do not always reflect the levels to which members of the public
actually are exposed in their daily lives. With the introduction of small
personal exposure monitoring instruments, it should become increasingly
possible to obtain accurate measurements of the pollutant levels actually
impacting the population with convenience and at low cost. The present
investigation is one of the first studies to utilize these new instruments
in a large-scale field investigation. We hope that the findings of such
field studies will have important implications for future policies affecting
the control of environmental pollution and the protection of public health
and welfare.
Courtney Riordan
Director, Office of Monitoring
Systems and Quality Assurance
111
-------
EXECUTIVE SUMMARY
Commercial areas attract and generate relatively large volumes of urban
traffic that typically circulates at low speeds with frequent stops and
starts. Such traffic patterns are usually associated with high emissions of
carbon monoxide (CO). This air pollutant stems from incomplete fuel
combustion, which most frequently occurs while motor vehicles are
accelerating and decelerating. Since commercial areas attract large numbers
of people, one would expect to find high levels of personal exposure to CO.
Yet, relatively few field studies have been undertaken to measure CO
concentrations in commerical settings, and particularly inside these
settings.
This study employed recently developed miniaturized personal exposure
monitors (PEM's) to measure CO in commercial settings in five California
cities and suburbs. Since high CO levels have been recorded in colder
weather not typical of California's coastal cities, the CO levels reported
here do not completely reflect CO problems elsewhere. A total of 588
commercial facilities were visited, including retail stores, office
buildings, hotels, restaurants, department stores, and adjacent sidewalk and
street intersections. Altogether, 5,000 CO observations were recorded
instantaneously at 1-minute intervals as the investigators walked along
sidewalks and into buildings.
The study had five principal objectives:
1. To determine the CO concentrations typically found in commerical
settings;
2. To determine the variability of CO concentrations in commercial
settings and the time and spatial factors that may be associated
with that variability;
3. To define and classify microenvironments which are applicable to
commerical settings;
4. To determine how accurately fixed station monitors measure the CO
concentrations to which the public is exposed in commercial
settings; and
5. To develop research methodology for measuring CO concentrations in
field surveys using PEM's.
Typical CO Concentrations
* Of 210 indoor commercial settings, excluding 10 parking garages, 204
(97.1%) had an average CO concentration of less than 9 parts per
million (ppm), the National Ambient Air Qualilty Standard for an 8-hour
exposure.
IV
-------
* Of 368 outdoor settings, 356 (96.7%) had an average CO concentration of
less than 9 ppm.
* CO concentrations measured at various heights in six garages of eight
high-rise office buildings and hotels were fairly widespread and
sometimes quite high (7-31 ppm); levels in their hallways on several
floors were fairly homogeneous and often low (less than 8 ppm).
* In Palo Alto, average CO concentrations in a 15-story office building
with an attached parking garage were equal to or greater than 9 ppm on
most of the first 10 floors on four of seven visits, each visit on a
different date.
* For this same 15-story office building in Palo Alto, CO concentrations
were relatively homogeneous for lobbies on different floors; for
stairwells, however, concentrations tended to decline linearly with
greater elevation.
Time Variability of CO Concentrations
* Of 74 settings visited for relatively long periods of time (6 to 111
minutes), the standard deviations of CO concentrations of 55 (74.3%)
did not exceed 2 ppm, indicating that for a given setting and date, CO
concentrations appear relatively stable over time.
* Of 17 (12 indoor and 5 street intersections) visited twice on the same
date, the average difference in mean CO concentration between visits
was small (1.3 ppm) for 15 of the cases, further indicating that CO
concentrations are stable over time.
* CO concentrations for both indoor and outdoor settings tended to vary
with date and were affected by wind speed, with lower CO values
encountered on windier days.
* If two dates had dissimilar wind speeds, CO concentrations varied more
from one date to another than from one setting to another on the same
date; this was true for settings of all types, except parking garages.
Spatial Variability of CO Concentrations
* CO concentrations indoors were statistically, but not substantially,
less than outdoors, and then only when the entrance door of the setting
was closed to the street and both measurements had been made within a
short time (3 minutes).
* On a given date, CO concentrations inside varied as much from, for
example, one retail store to another, as they did from a retail store
to a bank, hotel, office building, .restaurant, or other commercial
setting. A notable exception was the parking garage, for which CO
concentrations were substantially greater than for all other types of
indoor settings.
-------
* For geographic locations with heavy traffic, such as the Union Square
district of San Francisco, CO concentrations inside varied more from
one city block to another than from one setting to another on the same
block. This conclusion does not imply, however, that adjacent indoor
settings in commercial areas have identical concentrations.
* In the Union Square district of San Francisco, there was no
statistically significant difference in CO concentrations between the
corners and faces of city blocks for sidewalk settings.
* For the same geographic area and date, the CO concentrations varied
more from one intersection to another than from one corner to another
of the same intersection.
* For the same geographic area and date, some intersections had similar
CO concentrations and some differed, although the differences usually
were small.
* No statistically significant difference was found for CO concentrations
measured outdoors on sidewalks on opposite sides of the same street,
suggesting that the windward-leeward effect was quite weak for these
locations.
* In downtown Palo Alto, CO concentrations were dissimilar for indoor
settings located on opposite sides of the same street, giving support
to a windward-leeward effect for these settings.
* In general, outdoor settings were difficult to distinguish with respect
to variation in CO concentrations, indicating that some outdoor
settings (sidewalks, plazas, arcades, etc.) by themselves may
constitute one microenvironment, at least for some commercial areas.
* CO concentrations inside settings varied as much within a given
geographic location as they did between geographic locations. This
implies that data for different locations could be consolidated, if
precautions were taken to use dates with similar wind speeds and
atmospheric stability.
Microenvironments
* Based upon time and spatial variability, CO concentrations were defined
for four types of commercial microenvironments.
Enclosed parking garages: 10 enclosed parking garages had a
mean CO concentration of 21.7 ppm (excluding the ambient
concentration as measured at the fixed monitoring station) and
a standard deviation of 12.5 ppm.
Settings attached to enclosed parking garages: 7 settings
attached to enclosed parking garages had a mean CO
concentration of 6.1 ppm (excluding the ambient concentration)
and a standard deviation of 2.9 ppm.
VI
-------
All other indoor settings: 202 other indoor settings had a
mean CO of 2.1 ppm(excluding the ambient concentration) and a
standard deviation of 1.6 ppm.
All outdoor settings: 368 outdoor settings had a mean CO
concentration of 3.0 ppm (excluding the ambient concentration)
and a standard deviation of 2.6 ppm.
Quality Assurance of Personal Exposure Monitors
* For 7 of 11 surveys, during which two PEM's were used simultaneously
"side-by-side," the average of the absolute difference between
instruments was less than 1 ppm.
* For 9 of 11 surveys, during which two PEM's were used simultaneously
"side-by-side," the correlation coefficients between instruments
equalled or exceeded 0.97.
Comparison of Personal and Fixed Station Monitors
* For indoor settings, excluding parking garages and one "hot" building
in Palo Alto, the average CO concentration as determined by PEM's was
3.02 ppm. This value was statistically, but not substantially, greater
than the average CO concentration of 2.00 ppm as determined
simultaneously by fixed monitoring stations.
* For outdoor settings, the average CO concentration as determined by
PEM's was 4.00 ppm. This value was statistically, but not
substantially greater than the average CO concentration of 1.98 ppm as
determined simultaneously by fixed monitoring stations.
In general, this study demonstrated that personal exposure monitors are
an efficient and inexpensive means for trained technicians to collect a
large quantity of data on the distribution of air pollutant concentrations
in time and space in an urban area.
VII
-------
CONTENTS
Foreword
Executive Summary lv
Figures .x
Tables xii
Acknowledgments ' xv
1. Introduction 1
A. Approaches for obtaining exposure data 1
B. Research questions 4
2. Literature Review - 7
A. Enclosed mobile settings 7
1. Defective vehicle exhaust systems 7
2. Traffic densities 8
3. Passenger smoking behavior 8
B. Enclosed stationary settings 9
1. Indoor/outdoor relationships 9
2. Indoor sources 10
3. A comprehensive model 13
C. Outdoor settings 13
1. Pedestrian settings 13
2. Occupational exposures 15
3. Explanatory studies 17
D. Discussion 19
3. Methodology 21
A. Personal exposure monitors 21
B. Monitoring procedures and quality assurance tests. ... 22
C. Sampling procedures 23
D. Geographic locations and commercial settings 26
E. Procedures for data reduction 27
4. Time Variation of CO Concentrations 38
A. Settings monitored for extended periods
without interruption 38
B. Settings visited twice on the same date 43
C. Same settings visited on different dates 45
5. Spatial Variation of CO Concentrations 50
A. Comparison of indoor and outdoor settings 50
B. Horizontal variation for indoor settings 55
1. Purpose of the setting 55
Vlll
-------
2. Location of setting within a city block grid . . 57
3. Settings on opposite sides of the same street. . 57
C. Horizontal variation for outdoor settings 60
1. Purpose of the setting 60
2. Physical coordinates of a setting 62
3. Settings on opposite sides of the same street. . 66
D. Vertical variation for indoor settings 67
6. A "Hot" Office Building 70
A. Time variation 71
B. Spatial variation 77
C. Discussion 77
7. Geographic Variation of CO Concentrations 80
8. Comparison of Personal and Fixed Station Monitors 85
9. Microenvironments of Commercial Areas 91
10. Conclusions and Recommendations 98
References 101
Appendices 107
A. Photographs of PEM Instruments, Sampling Approach,
and Several Sampling Locations 107
B. Sampling Procedure for Union Square District of
San Francisco for 13 June 1980 116
C. Typical CO Concentrations for Commercial Settings. . . . 120
D. Cumulative Frequency Distributions of CO Concentrations. 142
IX
-------
FIGURES
Number Page
1 Four geographic locations in the San Francisco Bay Area .... 28
2 One geographic location surveyed in the Los Angeles
metropolitan area 29
3 Financial-Chinatown district of San Francisco, California ... 30
4 Union Square district of San Francisco, California 31
5 University Avenue central business district of
Palo Alto, California 32
6 Castro Street central business district of
Mountain View, California 33
7 Westwood Village district of Los Angeles, California 34
8 Average concentrations of CO for selected levels of
eight high-rise buildings 69
9 Vertical profile of CO concentrations measured in a
15-story building in Palo Alto, California, during
the daytime hours on 4 April 1980 78
10 Cumulative frequencies of CO concentrations for four
microenvironments 96
B-l Field map of Union Square district of San Francisco,
California 117
D-l Cumulative frequencies of CO concentrations for all
commercial settings, except parking garages,
in downtown Palo Alto on five dates 143
D-2 Cumulative frequencies of CO concentrations for
indoor and outdoor settings near Union Square
in San Francisco on 13 June 1980 144
-------
D-3 Cumulative frequencies of CO concentrations for eight
types of indoor commercial settings near Union Square
in San Francisco on 13 June 1980 145
D-4 Cumulative frequencies of CO concentrations for eight
city blocks in the Union Square district of
San Francisco on 13 June 1980 146
D-5 Cumulative frequencies of CO concentrations for
selected indoor commercial settings at five
geographic locations 147
XI
-------
TABLES
Number
1 Statistical Summary of Quality Assurance Tests,
2 Field Survey Dates, Hours, Locations, and Numbers
of CO Samples 35
3 Number of Commercial Settings by Type of Setting and
Geographic Location 36
4 Statistical Summary of CO Concentrations for Commercial
Settings Visited for Lengthy Periods of Time on the
Same Date 39
5 Statistical Summary of Mean CO Concentrations for Commercial
Settings Visited Twice on the Same Date 44
6 Meteorological Summary for the Redwood City Station of
the Bay Area Air Quality Management District for Dates
and Hours Corresponding to Five Surveys Conducted in
the Central Business District of Palo Alto, California ... 47
7 Statistical Summary of CO Concentrations for all
Commercial Settings, Except Parking Garages,
Visited on Five Dates in the Central Business
District of Palo Alto, California 48
8 Statistical Summary of CO Concentrations Measured on
3 June 1980 at Indoor and Adjacent Outdoor
Commercial Settings Located in the Union Square
District of San Francisco, California 51
9 Statistical Summary of CO Concentrations Measured on
13 June 1980 for Indoor Commercial Settings Broken
Down by Purpose and Located in the Union Square
District of San Francisco, California 56
10 Statistical Summary of CO Concentrations Measured on
13 June 1980 for Indoor Commercial Settings Broken
Down by City Blocks Located in the Union Square
District of San Francisco, California 58
XII
-------
11 Statistical Summary of CO Concentrations Measured on
13 June 1980 at Corners and Faces of City Blocks
Located in the Union Square District of
San Francisco, California 61
12 Statistical Summary of CO Concentrations Measured on
13 June 1980 for Corners of Intersections Located in
the Union Square District of San Francisco, California ... 63
13 Average CO Concentrations by Date and Level for a
15-Story General Purpose Office Building in Palo Alto,
California (1980) 72
14 Statistical Summary of CO Concentrations for Visits to
Five Selected Floor Lobbies of a 15-Story General
Purpose Office Building in Palo Alto, California,
Broken Down by Date and Hour of Visit 73
15 Meteorological Summary for the Redwood City Station of the
Bay Area Air Quality Management District for Dates and
Hours Corresponding to Seven Surveys Conducted in the
Central Business District of Palo Alto, California 76
16 Percentages of Indoor Commercial Settings with Entrance
Doors in Designated Positions for Five Selected
Field Surveys 82
17 Statistical Summary of CO Concentrations for Selected
Indoor Commercial Settings Situated at Five
Geographic Locations 84
18 Straight-Line Distance and Direction of Geographic
Locations with Respect to Their Nearest Fixed
Monitoring Stations 86
19 Summary of CO Concentrations Collected Simultaneously
From Fixed Monitoring Stations and Personal Exposure
Monitors 87
20 Summary of Ambient CO Concentrations for Each Field
Survey 94
21 Statistical Summary of CO Concentrations for Four
Commercial Microenvironments 95
B-l City Block Assignments for Surveyors 118
B-2 Sample Data Form 119
C-l Mean Concentrations of CO for Indoor Commercial Settings
Broken Down by Geographic Location 121
Xlll
-------
C-2 Statistical Summary of CO Concentrations for Indoor
Commercial Settings Located near Union Square in
San Francisco, California, and Spanning Selected
Dates from 9 November 1979 through 13 June 1980 , 128
C-3 Statistical Summary of CO Concentrations for Indoor
Commercial Settings Located along University Avenue in
Downtown Palo Alto, California, and Spanning Selected
Dates from 24 January 1980 through 11 July 1980 131
C-4 Mean Concentrations of CO for Outdoor Commercial Settings
Broken Down by Geographic Location , 133
xiv
-------
ACKNOWLEDGMENTS
Several individuals assisted with collections of CO samples in the
Union Square district of San Francisco. J. Botsford assisted Wayne Ott on 9
November 1979 and David Fairley and Robin DeMandel assisted both Wayne Ott
and Peter Flachsbart on 13 June 1980. Their assistance is gratefully
acknowledged and appreciated.
Stanley Blacker of EPA's Office of Mobile Source Air Pollution Control
served as project officer on the contract which supported a significant part
of this research project. A portion of this project, including the cost of
the personal monitoring instruments and one author's salary, was supported
under EPA's Innovative Research Program and was facilitated by the SIAM
Institute for Mathematics and Society (SIMS) program at Stanford University.
We especially wish to thank the technical reviewers of this report:
Robin DeMandel of the Bay Area Air Quality Management District; Richard
Ziskind of Science Applications Inc.; Nihua Duan of the Rand Corporation;
and Charles Brunot, David Mage, James Repace, Joe Sommers, and Lance
Wallace, all of EPA. Finally, we wish to thank Irene Kiefer for her
invaluable editorial assistance.
xv
-------
CHAPTER 1
INTRODUCTION
The goal of air pollution control programs in the United States, as
mandated by Federal law and implemented by the States, is to attain National
Ambient Air Quality Standards (NAAQS). The NAAQS for carbon monoxide (CO),
for example, specify two different concentrations and averaging times,
neither of which is to be exceeded more than once per year:
35 parts per million (ppm) for 1 hour
9 ppm for 8 hours
Both standards are intended to protect against the accumulation of more than
2% carboxyhemoglobin in the blood.
Nondispersive infrared (NDIR) monitoring at fixed stations is the usual
way for determining a given city's compliance with the NAAQS for CO. During
the past decade, a number of studies have revealed that concentrations
observed at fixed air monitoring stations have not been representative of
concentrations sampled throughout an urban area. Some field studies have
shown, for example, that commuters in traffic and pedestrians on downtown
streets encountered CO levels above the NAAQS on a given date, while
official air monitoring stations reported CO values below the NAAQS at the
same time. Furthermore, studies of human activities suggest that most
people spend the greatest proportion of any given 24-hour period indoorsin
residences, stores, offices, factories, etc. These settings are not
necessarily identical to sites selected for fixed air monitoring stations.
These studies have raised questions about the usefulness of data
generated by today's monitoring stations for protecting public health. An
unanswered question is the degree to which conventional fixed stations
either underestimate or overestimate the actual exposure of people as they
go about their daily activities. These studies have stimulated interest in
"exposure monitoring," which treats the person as a receptor and measures
the pollutant levels actually contacting the person's body.
A. APPROACHES FOR OBTAINING EXPOSURE DATA
A recent review of the literature by Ott(l) indicates that a great
number of scientific papers and governmental research reports have sought to
develop means for estimating human exposures to air pollution from existing
monitoring data. Because such estimates usually rely on extrapolation of
data from fixed stations, these authors have found it necessary to introduce
-------
many assumptions, and their calculations may be subject to error. There is
general agreement that the most accurate exposure information comes from
field measurements at the physical boundary of the person, but, prior to the
last few years, instruments for making such measurements generally have not
been available.
As indicated by Duan(2), two basic approaches exist for studying human
exposure: (1) the direct approach, in which we simply measure exposure
directly; and (2) the indirect approach, in which we measure the time
allocations and concentrations separately and then combine them to recon-
struct exposures. The direct approach requires the availability of
accurate, affordable personal monitoring instruments, while the indirect
approach requires the existance of some model by which activity patterns
(human time allocations) can be combined with concentration data from field
measurements at actual locations. Some authors(3) refer to the first
approach as an "exposure profile" field study technique and the second
approach as an exposure modeling technique. The second approach requires
field surveys for collecting the concentration data that will be combined
with the activity patterns, as well as a conceptual basis for designing such
field surveys. The senior author of this report currently is developing
methodologies for conducting such field studies.
Duan(2,4) has proposed a conceptual framework for the various ways that
one might go about estimating or measuring human exposure to air pollution
using the indirect approach. During daily activities, people are likely to
pass through many different "locations" with different pollutant concen-
trations. Duan(2,4) refers to such locations as "microenvironments." A
microenvironment is a "chunk of air space with homogeneous pollutant
concentration." His examples include "Room 106," "Room 107," "Sidewalk
along 1700 Main Street," etc. Because of the large number of possible
microenvironments, he suggests that microenvironments of a similar nature
should be grouped to avoid too many different categories. He refers to such
groupings as "microenvironment types." Presumably, such groupings could be
made by examining the concentrations observed in a large number of
microenvironments and then identifying those that are statistically similar.
In the present investigation, a "commercial setting" consisted of
indoor and outdoor locations in downtown areas selected on the basis of a
sampling plan. These settings may be considered rudimentary building blocks
of microenvironments and, ultimately, microenvironment types. Thus a
commercial setting, as used in this study, is a less formal concept than a
microenvironment. Except for one date in which a very structured sampling
plan was used in downtown San Francisco's Union Square area, a commercial
setting corresponds to a downtown sampling location where members of the
general public might be found as they go about their daily activities.
Despite the lack of a formal effort ,to identify and construct microenviron-
ments and our limited data for this purpose, we have carried out statistical
analyses of our data in a manner that is intended to identify those settings
(i.e., locations) which are appropriate to treat as CO microenvironments
(Chapter 9).
-------
Ultimately, such microenvironmental field data, when represented by
proper statistical parameters, can be incorporated into
"microenvironment-activity pattern" models designed to accommodate such
data. One such model, the Simulation of Human Air Pollution Exposure
(SHAPE) model,(5) has been designed specifically to utilize information on
CO concentrations and activity patterns. It uses computer simulation
techniques to expose one person after another to the CO concentrations
associated with selected microenvironments, according to the times spent by
each person in these microenvironments. These times are treated
stochastically but are based on real observations, such as population
surveys and diary information. The SHAPE model requires statistical
parameters, such as the geometric mean and standard geometric deviation of
one-minute average CO concentrations, for each microenvironment in the
program. One purpose of the present research is to determine if appropriate
microenvironments can be found among commercial settings. It is important,
if possible, to identify the statistical parameters appropriate for each
microenvironment, so it can be represented in the SHAPE model and other
models of this kind.
Although some previous field studies have measured CO in a few
important microenvironments, such as inside motor vehicles moving in
traffic, a great number of microenvironments have not been studied
systematically. For example, few field studies have measured CO
concentrations in commercial settings, particularly inside these settings.
Yet, theory suggests that CO exposures in such settings may be significant
at times.
CO in the air is strongly associated with traffic from motor vehicles.
CO is emitted from incomplete fuel combustion, which most frequently occurs
while engines are accelerating or decelerating. Commercial areas attract
and generate relatively large volumes of urban traffic that move at low
speeds with frequent stops and starts. Office buildings and retail stores
close to commercial shopping streets in downtown and suburban areas often
form small, semiclosed spaces. These canyon-like spaces afford relatively
limited opportunity for traffic-generated pollutants to disperse.
Furthermore, since commercial areas attract large numbers of people, one
would expect them to be exposed to high CO concentrations.
Previous exposure monitoring field studies for CO generally have been
handicapped by the complexity and size of the instruments used to monitor
compliance with the NAAQS. These instruments continuously sample the
outdoor air at one or more locations in U.S. cities. These instruments
employ the NDIR measurement principle that has been evaluated and approved
by the U.S. Environmental Protection Agency (EPA) as the Federal reference
method. Monitoring instruments incorporating the NDIR reference method are
large, complex, and expensive and they require an air-conditioned structure
to produce accurate and reliable data. Such instruments cannot easily be
used to follow persons as they go about their daily activities in stores,
restaurants, residences, offices, and other locations.
-------
Prior to the late 1970"s there was no low cost, accurate means
available for measuring CO concentrations to which people ordinarily were
exposed in their daily lives. The advent of microelectronics has brought
considerable progress in developing reliable, compact air quality monitoring
instruments that can operate on batteries. The most dramatic of these are
the new miniaturized personal exposure monitors (PEM's) sometimes called
"dosimeters." The instruments can go nearly anywhere, attached to a person
or carried on a shoulder strap. Furthermore, like NDIR instruments, the CO
PEM's can operate continuously. Of course, a PEM's operation is limited by
its battery life.
The present investigation is the first large-scale microenvironmental
field study to make use of the new CO PEM instruments. We were impressed by
the ease with which it was possible to collect large quanities of field data
of relatively high quality. Data on the reliability of the PEM instruments
when carried side-by-side (Chapter 3) and photographs of the sampling
methodology (Appendix A) are included in this report.
B. RESEARCH QUESTIONS
Several research questions can be addressed in a field study of
commercial settings. One set of questions concerns the CO concentrations
typically found in commercial settings:
* What levels of CO ordinarily are present in typical commercial
settings?
* Are CO levels in typical commercial settings usually zero,
negligible, or above the NAAQS?
Another set of questions concerns the varibility of CO concentrations
and the factors that may be associated with that variability:
* How do CO concentrations vary over time for a given commercial
setting?
* How do concentrations of CO vary within and between different
cities for a given commercial setting?
* How do indoor and outdoor CO concentrations compare in a given
commercial setting?
* Can CO concentrations indoors be reduced substantially by
keeping entrance doors of commercial buildings closed to the
street?
* How do CO concentrations vary over the height of high-rise
buildings?
If CO is a street-level pollutant associated with vehicular
traffic, do workers have greater protection in offices on the
upper floors of a high-rise building?
-------
Other research questions concern how accurately the fixed-station
monitors operated by air quality management districts measure the air
pollution to which the public actually is exposed. Specifically:
* Do CO concentrations measured in commercial settings using
PEM's correlate with ambient concentrations measured at fixed
stations using NDIR instruments?
* Can commercial settings be separated into those having
concentrations that closely match ambient levels, as measured
by fixed-station monitors, and those having concentrations
that can be related to specific CO sources?
Microenvironments pose another set of research questions. As indicated
by Duan(2,4) microenvironments must be defined "fine" enough so that the
variance of an air pollutant concentration within each environment is
smaller than that between environments. On the other hand,
microenvironments must be defined "coarse" enough to collect data to
determine how much time people spend in each microenvironment. (In such
studies, humans can detect neither levels nor changes in CO levels, but they
can distinguish ordinary changes in the settings of their activities. Thus,
the definition and classification of microenvironments should capture real
variation in CO concentrations.) Research should answer several questions:
* Do several commercial settings form a microenvironment with
similar CO concentrations?
* Allowing for ambient CO concentrations, can the level in
commercial settings be represented approximately by a single
frequency distribution or a set of frequency distributions?
* Does a statistical definition of microenvironments produce
important and useful distinctions?
Finally there are questions related to research methodology:
* Is the CO PEM an effective tool for sampling air quality at a
variety of urban locations?
* Do PEM's provide valid and reliable data?
* Can a systematic sampling method be devised that will allow
collecting data in other cities comparable to those in this
study?
* What are the implications of the present study for future
research on exposures of the population to CO?
This report presents results of a project that used PEM's to measure
instantaneously 5,000 CO concentrations at 1-minute intervals in a variety
of commercial settings. Among indoor settings were banks, barber shops,
book stores, coffee shops, department stores, drug stores, and travel
-------
agencies. Both indoor and outdoor settings were sampled, often more than
once, in several California cities on a variety of dates in the 9-month
period from November 1979 to July 1980.
-------
CHAPTER 2
LITERATURE REVIEW
This chapter reviews the literature concerning measurements of CO
concentrations in environmental settings commonly inhabited by people.
Since the literature is quite large, studies are emphasized that focus on
settings similar to those we have investigated, thus providing a basis for
comparison.
The literature is not far advanced in terms of capabilities to predict
CO concentrations for given settings. Most studies have described the CO
concentrations of a particular setting to determine the severity of the
potential health risk. Fewer studies have attempted to explain CO
concentrations in terms of traffic volume and meteorological variables.
Part A of this chapter reviews studies that focused on enclosed mobile
settings, that is, the passenger compartments of motor vehicles. Part B
describes studies dealing with the large variety of enclosed stationary
settings, such as dwelling units and commercial facilities. Part C presents
the numerous studies of outdoor settings, usually near traffic, which are
open and stationaryfor example, sidewalks and intersections. Part D
closes the chapter with a discussion of these studies.
A. ENCLOSED MOBILE SETTINGS
Several studies have focused on the relatively high CO concentrations
associated with the passenger compartments of different modes of
transportation. These high concentrations are related to defective vehicle
exhaust systems, traffic densities, and passenger smoking behavior.
1. Defective Vehicle Exhaust Systems
The diffusion of CO from the exhaust system into the passenger
compartment of the vehicle has been labeled the "CO intrusion problem."
Ziskind ejt a^. ,(6) discuss this problem, which is especially critical to the
operators of sustained-use vehicles. Even short-term users of public
vehicles may incur some risk. A study supported by the U.S. Department of
Transportation(7) estimated that CO concentrations of about 20 ppm may exist
in school buses occupied by up to 2.1 million school children nationwide.
Of 64 buses tested, 20% exceeded the 20 ppm level, and 5.4% exceeded 50 ppm.
-------
2. Traffic Densities
A study made by Brice and Roesler(8) in 1966 found that occupants of
motor vehicles in six large cities were exposed to CO levels ranging from
1.3 to 6.8 times the corresponding CO values measured at fixed air
monitoring stations. Another early study by Lynn et_ aL.,(9) found similar
results for CO in 15 cities by using a mobile van equipped with a NDIR
detector. Neither study, however, attempted to measure traffic volumes
associated with the CO measurements.
A later study by Godin, Wright, and Shephard(lO) used 100 milliliter
(ml) glass syringes to measure CO concentrations inside passenger vehicles.
Traffic volume was measured subjectively"heavy," "moderate-to-heavy,"
"moderate," "light-to-moderate," and "light." The investigators concluded
that vehicle compartment concentrations were much higher than measurements
taken at conventional air monitoring stations and that traffic volume was
the dominant variable affecting CO exposures for vehicle occupants.
Cortese(ll) equipped 65 nonsmoking volunteers in Boston, Massachusetts,
with CO monitors (ESI Ecolyzers). Each volunteer wore the monitor for 3 to
5 days for a 6-month period. Some volunteers were commuters who lived in
Boston suburbs but worked in offices and commercial settings in downtown
Boston. Other volunteers did not commute. Cortese found that the CO
exposures averaged 11.9 ppm for commuters during trips lasting 45 to 60
minutes. Simultaneous measurements taken at six fixed stations averaged
only 6 ppm. Two of the six stations, located in the urban portion of the
metropolitan area, recorded concentrations which matched the commuter
exposures better than the four suburban stations. The 8-hour average CO
concentrations measured at the six stations were a better match to the
exposures of those volunteers who did not commute.
Willits and Ott(12) recently conducted a study of CO exposure of two
motorists using PEM's. Altogether, 85 drives were made over the same
segment of an arterial in Palo Alto and Menlo Park, California, either once
or twice a week throughout 1980. Two test vehicles were used, both known to
be free of CO intrusion from the exhaust system. Measured exposure was
related somewhat to traffic volume at intersections, but more so to
occasional vehicles with highly visible emissions. Average exposure was
related strongly to average traffic volume at intersections.
3. Passenger Smoking Behavior
Sebben, Pimm, and Shepherd(13) found elevated CO levels on public
vehicles that permitted passengers to smoke. On an airport bus in which the
windows were closed and the air conditioning system was operating, the CO
level was 6.5 ppm when the bus departed from the terminal. Ten passengers
immediately began to smoke, and the CO level climbed to 12 ppm after 5
minutes. The level returned to 6.5 ppm 25 minutes later when the bus
arrived at its destination and the passengers disembarked.
On a commuter railway car departing from the terminal, CO levels
averaged 7.3 3.3 (standard deviation) ppm in the smoking compartment and
-------
6.2 ± 1.8 ppm in the nonsmoking section. Several passengers in the smoking
compartment immediately began to smoke. After 10 minutes, levels reached 12
ppm, compared to 7 ppm in the nonsmoking section. After passengers
disembarked at the destination, levels reached 8 ppm in both compartments.
Sebben, Pimm, and Shepherd also investigated the Toronto Island ferry during
Winter when windows were closed. They found that CO levels in the crowded
smoking compartment were 22 ppm, compared to 2 ppm in the nonsmoking
compartment.
B. ENCLOSED STATIONARY SETTINGS
A report prepared for EPA(14) summarized a number of studies of CO
levels in enclosed stationary settings. The studies revealed that, in the
absence of indoor sources, CO concentrations in enclosed settings tend to be
similar to outdoor concentrations, but tend to lag behind the peak
concentrations observed outdoors. These studies of indoor/outdoor
relationships are reviewed below in the first section. The second section
considers the contribution to indoor CO levels made by specific enclosed
sources of CO, such as motor vehicles in enclosed garages and tunnels, gas
appliances, and smoking. The third section briefly descibes a comprehensive
model which attempts to explain indoor concentrations in terms of all
relevant factors.
1. Indoor/Outdoor Relationships
Yocum, Clink, and Cote(15) studied buildings in Hartford, Connecticut,
by monitoring CO concentrations at four locations with respect to the
building's exterior walls: "far outdoors", "near outdoors", "near indoors",
and "far indoors". The ratio of far indoor levels to far outdoor levels was
1.0 for all those buildings for which long averaging times were used. In
contrast, short-term CO concentrations near indoors lagged behind
concentrations near outdoors. These outdoor CO concentrations peaked from
6:00 to 9:00 a.m. During these hours, the indoor CO concentrations were
lower than the outdoor levels, but in time tended to catch up with outdoor
levels. However, as the outdoor levels subsequently were dissipated, indoor
concentrations rose, but to less than the highs found earlier outdoors. The
authors determined an exponential decay constant for CO by closing the
ventilation system and then "spiking" the indoor air with CO.
General Electric Company(16) investigated CO concentrations in two
high-rise buildings, one constructed over a highway. As in the previous
study, indoor CO concentrations normally were lower than outdoor
concentrations at all heights above the roadway when outdoor concentrations
were high. Conversely- when outdoor concentrations were low, indoor
concentrations were not as low. Hence, in the absence of indoor CO sources,
indoor concentrations tend to follow outdoor levels with some degree of time
lag and with a tendency not to reach either the extreme high or low values
that may be found outdoors.
The General Electric study also reported that at heights greater than
100 feet above the roadway, CO concentrations were larger indoors than
-------
outdoors. This result can be attributed to the entrapment of CO within the
building. No such entrapment existed outside the building at the same
height.
Godin, Wright, and Shepherd(lO) reported findings similar to those
reported so far. They sampled a variety of indoor settings in and around
Toronto, Ontario, using 100-ml glass syringes. For downtown offices, they
found that indoor CO concentrations tended to match outdoor levels, but with
a lag of 1 or 2 hours. They also observed an inverse relationship between
CO concentrations and distance from the central city. For a building near
the city's center, outdoor values were, 2.7 ± 1.8 ppm, while corresponding
indoor values on the first floor were 2.2 ± 1.3 ppm. At a suburban home 10
miles north of the city's center, CO values had diminished to 2.0 ± 1.4 ppm
(outdoors) and 1.9 ± 1.3 ppm (indoors). Finally, at a semirural farm house,
outdoor concentrations averaged 0.8 ± 0.6 ppm and indoor values were almost
the same at 1.0 ± 0.8 ppm; the slightly higher indoor values could be
explained by a log fire in the house.
Godin and coworkers also investigated vertical variation in CO
concentrations for the Toronto Dominion Center building. While the sidewalk
concentration was 6.4 ppm, the concentrations on the 1st, 3rd, and 54th
floors were 4.6 ppm, 4.0 ppm and 2.4 ppm, respectively.
Ziskind(17) has described similar results for buildings in Los Angeles,
California. Although he has not yet published his results, he found a
general decline in CO concentrations with greater height in several of the
tall buildings that he studied. ^
Wilson and Schweiss(18,19) monitored five indoor settings in downtown
Seattle, Washington, using a bag sampling methodology. Again, the CO
concentrations measured indoors were usually lower than at selected outdoor
settings, but did not show a constant relationship to outdoor values.
Furthermore, the indoor 8-hour average CO concentration exceeded the 9 ppm
NAAQS only once at one of the five indoor sites.
Finally, Sebben, Pimm, and Shepherd(13) measured CO concentrations in
indoor commercial settings in Toronto using a portable CO analyzer. A total
of 33 retail stores were sampled, mainly from 1:00 to 4:00 pm. Each store
employed fewer than 25 persons, and the City of Toronto's bylaws permitted
smoking by both patrons and staff. Indoor CO levels averaged 10.0 ±4.2
ppm, slightly less than the outdoor readings of 11.5 ± 6.9 ppm. Lobbies of
three downtown hospitals were also sampled. CO concentrations ranged from 4
to 8 ppm, close to outdoor levels nearby.
2. Indoor Sources
Among primary sources of CO in enclosed settings are gas appliances
such as kitchen stove burners and ovens, gas-fired space heaters that are
unventilated, smokers, and motor vehicles operating in parking garages and
tunnels.
10
-------
In conjunction with Battelle Columbus, Elkins e_t £l_. ,(20) surveyed 57
dwelling units equipped with kitchen gas appliances. In one home, the
hourly average CO concentration was 39.4 ppm, due to a poorly adjusted
oven. After the oven was adjusted the level dropped to 7.1 ppm. The other
homes had CO values much lower than the NAAQS of 35 ppm for 1 hour. In 71%
of these homes, CO concentrations were less than the 8-hour NAAQS of 9 ppm.
Cote, Wade, and Yocum(21) monitored CO concentrations in the kitchen
and in three other locations in four private residences. Data collected for
2-week periods during four different seasons showed that CO concentrations
were higher inside than outside. With one exception, concentrations always
were higher in the kitchen than elsewhere in the house. In one of the four
houses, 40% of the indoor CO concentrations were above the 8-hour NAAQS.
Radford(22) reported that, of 302 old houses surveyed in East
Baltimore, 20% had levels above 10 ppm; 4% had levels above 20 ppm; and 1%
exceeded 50 ppm. In houses with high concentrations, the problem was traced
to unvented gas space heaters or stoves used for heating; to oil or
gas-fired furnaces used for central heating; and to various combinations of
these sources.
A number of studies have focused on the effect of smoking on CO levels
indoors. Bridge and Corn,(23) for example, monitored CO concentrations
before and after in two rooms, each the site of a 90-minute party in which
roughly half the participants were smokers. Prior to the parties the CO
levels in the two rooms (volumes of 5,120 cubic feet and 3,570 cubic feet)
were 1 and 2 ppm, respectively. During the parties the average CO
concentrations were 7 ppm and 9 ppm, respectively. These levels could be
attributed to 50 cigarettes and 17 cigars smoked at the party in the first
room and to 63 cigarettes and 10 cigars smoked at the party in the second
room. Because the levels were below the 35 ppm NAAQS for 1 hour, the
authors concluded that CO concentrations from cigarette and cigar smoking
did not represent a CO inhalation hazard to nonsmokers.
Sebben, Pimm, and Shepherd(13) used a portable analyzer to measure CO
concentrations in nine night clubs (12 rooms) where smoking was permitted.
A repeat visit was made to the most heavily contaminated club. At these
clubs, the primary increase in CO concentrations occurred from 6:00 to 7:00
p.m., with relatively constant readings thereafter. The peak occupancy of
the clubs was reached between 8:30 and 9:00 p.m. Patrons tended to smoke
most heavily at first, then somewhat less after dancing and formal
entertainment began. The investigators reported a significant correlation
between cigarette smoke density, as measured on a five-point scale, and
measured CO concentrations.
For all clubs and times, the average indoor CO concentration of 13.4
pprn exceeded the outdoor figure of 9.2 ppm by a significant margin
(4.2 ± 1.5 ppm; p < 0.01). For the most heavily contaminated club, the
concentrations for the first visit ranged from 29.8 to 41.0 ppm and averaged
35.3 ppm. The concentration on the street outside was 9.9 ppm. There was
some improvement on the second visit when the CO concentrations averaged
11
-------
28.8 ± 2.2 ppm (5.0 ppm outside). There were near-capacity crowds at this
club on the evenings surveyed, and clouds of smoke were visible near ceiling
light fixtures.
These researchers also briefly visited 14 licensed restaurants during
the noon lunch hour. The average lunchtime reading of 9.9 ± 5.5 ppm was
substantially lower than that found for the night clubs. One heavily
contaminated setting yielded a reading of 28 ppm, a value associated with
"very smoky" conditions. Family restaurants also were surveyed. Except for
one high reading, the mean indoor CO concentration for the 45 restaurants
was 8.2 ± 2.2 ppm, versus 7.1 ± 1.7 ppm outside. One high reading, 37 ppm,
was observed in a pizza restaurant for which the outdoor level was only 9
ppm. Instead of tobacco smoke, the high reading in this restaurant was
attributed to either a malfunction of the ventilating system or to a greater
use of ironwork in the structure of the cooking grill. The physical
chemistry of CO formation and ironwork was not explained.
Sebben, Pimm, and Shepherd(13) concluded that it is relatively uncommon
to find a CO level over 15 ppm in public places that permit smoking.
However, concentrations two to three times higher can be encountered in
buildings that are poorly ventilated, particularly if they are tightly
constructed and therefore depend heavily on recirculated air. The authors
also concluded that the higher CO levels observed indoors than outdoors can
be attributed to cigarette smoke, when a substantial number of people are
smoking. However, it appears that a large part of the average indoor
concentration could be attributed to outdoor levels and not necessarily to
indoor sources only.
Although all levels were below Ontario's 8-hour occupational health
standard of 50 ppm, a substantial proportion of the observations exceeded
the "desirable" 8-hour exposure limit of 13 ppm. The authors noted that the
less restrictive occupational health standard is designed to apply to
persons who recognize that a special hazard exists in their work. Still,
most restaurant and bar employees are recruited on a casual basis and
probably do not expect to be exposed to high levels of CO.
For certain types of indoor settings, whether partially or totally
enclosed, motor vehicles may be the primary source of CO. Waller, Commins,
and Lawther(24) measured average CO concentrations in the Blackwell and
Rotherhithe Tunnel in London. The average CO concentrations during the
morning and evening rush hours were slightly above 100 ppm. Peak CO
concentrations of 500 ppm, 450 ppm, and 340 ppm were recorded on days for
which fans in the tunnel were shut off.
Other studies have investigated enclosed parking garages. Trompeo,
Turletti, and Giarrusso(25) measured CO levels in 12 underground garages in
Rome, Italy. The average CO level based on 132 observations was 98 ppm and
ranged between 10 ppm and 300 ppm. Of the 132 observations, 42% were above
100 ppm. Chovin(26) has recorded average levels of 150 ppm between 7:30 and
8:00 a.m. in a police garage in Paris, France. Finally, Wright et al..(27)
12
-------
reported that concentrations in poorly ventilated underpasses and
underground garages were higher than concentrations observed at other places
in Toronto.
3. A Comprehensive Model
Moschandreas e_t a_l.,(28) described a numerical model, the GEOMET
Indoor-Outdoor Air Pollution (GIOAP) model, which predicts indoor air
pollutant concentrations as a function of time. The model requires
information on outdoor concentrations at the boundaries of a residence as a
function of time, indoor source emissions, air exchange rates, indoor air
cleaning devices, and chemical sinks and recirculation mechanisms.
Moschandreas(29) has done extensive field studies to validate this
model and "...to produce new data and insights concerning the occurrence,
behavior, and significance of air pollution in nonworkplace, indoor
environments through observation, evaluation, and analysis of 46 months of
monitoring data." His study included experiments of episodic releases of
air pollutants near residences.
C. OUTDOOR SETTINGS
Outdoor settings near traffic have been the focus of a large number of
studies of CO concentrations. Some studies describe the typical
concentrations to which sidewalk pedestrians could be exposed. These
studies are presented below in the first section. The second section
considers those studies of people who may be at risk to high CO
concentrations due to their occupations, such as traffic police. The third
section reviews those studies which have attempted to explain CO
concentrations for outdoor settings as a function of traffic volume and
other variables.
1. Pedestrian Settings
Pedestrian areas of city streets have been a logical focus of several
studies of CO concentrations. By walking along the sidewalks of a busy
street, Ott(30) collected CO concentrations in downtown San Jose,
California, using a bag sampling methodology. As reported by Ott and
Mage,(31) 425 samples were collected on 21 days during midwinter. On seven
days, samples were collected over at least 8 hours, allowing the pedestrian
exposures of the investigator to be compared with the Federal NAAQS of 9 ppm
for 8 hours. For three of these seven days, the investigator's exposure
exceeded the NAAQS, while the official air monitoring station reported
values below the NAAQS.
Overall, the 8-hour average CO concentrations for the seven days ranged
between 1.4 and 3 times the values observed simultaneously at the official
air monitoring station. The highest values were in late December when
streets were heavily congested with traffic due to Christmas shopping. Ott
and Mage concluded that the bag sampling methodology illustrates "...the
feasibility of obtaining true measures of the exposure of critical popula-
tion groups with relative ease and economy...."(31)
13
-------
Wright et_ aK ,(27) used Ecolyzers to measure 4-to-6 minute average CO
concentrations" encountered by pedestrians in Toronto. They focused on the
sidewalks of a street that had been closed to traffic to become a pedestrian
mall. Before the street was closed, the average CO concentrations at two
cross streets were 9.4 ± 4.0 ppm and 7.9 ± 1.9 ppm. After the street was
closed, the averages dropped to 3.7 ± 0.5 ppm and 4.0 ±1.0 ppm, roughly
equivalent to the urban background level.
Wilson and Schweiss,(32,33) working with EPA's Regional Office in
Seattle, measured CO concentrations at 40 outdoor sites and 4 sidewalks in
Boise, Idaho, during 20 days in November-December 1977. On 19 of the 20
days, the 8-hour CO NAAQS was exceeded at one or more sites, while the same
standard was exceeded at the fixed air monitoring station on only 9 of 19
dates for which data were available. Altogether, about 70% of the 40
outdoor sites experienced one or more days in which the 8-hour CO average
exceeded 9 ppm. The highest 8-hour CO average was 17.2 ppm.
Pedestrian exposures also exceeded the 9 ppm NAAQS at times, ranging
from 2.9 ppm to 14 ppm for 2- to 4-hour averages. When CO concentrations
exceeded the 9 ppm NAAQS at the fixed air monitoring station, they also
exceeded it at all the outdoor sites surveyed. Although the fixed station
data generally were representative of higher concentrations in the area,
they did not reflect the very highest concentrations observed or the
frequency of such concentrations in the study area.
Wilson and Schweiss(18,19) conducted a similar study in downtown
Seattle in October-November 1977. Using a bag sampling methodology, they
measured CO concentrations over a 20-day period at 36 outdoor sites
(including small parks and isolated areas) and two sidewalks. For 80% of
these 20 days, the 8-hour average CO concentrations exceeded the 9 ppm NAAQS
at these sites, while the same standard was exceeded on only 45% of these
days at the fixed station. On the other hand, Wilson and Schweiss found
that the highest 8-hour value (15.8 ppm) observed at the fixed station did
not differ significantly from the highest value (16.1 ppm) observed for the
survey sites.
At 39% of the 36 outdoor sites, the 8-hour CO concentrations did not
exceed the 9 ppm NAAQS during the 20 days. Nor did the highest 8-hour
average always occur at the same site. Rather, it occurred at each of eight
sites, including the fixed station, over the 20 days. There were no
violations of the 9 ppm NAAQS on 4 days at the fixed station. Yet, on these
same days, at least one other survey site violated the standard and, during
these times, was more than 1.5 times greater than the fixed-station value.
CO levels along sidewalks ranged from 1.1 ppm to 11.9 ppm (4-hour
average). Wilson and Schweiss concluded that pedestrians probably were
exposed to levels above 9 ppm for more than 4 hours over the period. Of
course, few people remain pedestrians for that length of time.
A variety of other field studies of CO levels in downtown areas have
been stimulated by EPA's recommendations that data be collected at CO "hot
14
-------
spots" in those cities where fixed stations have reported high CO levels.
EPA has issued guideline documents, authored by Midurski,(34,35) that
provide criteria for selecting urban microenvironments in which CO levels
are expected to be high. The work of Wilson and Schweiss suggests that no
single location in a downtown area can give consistently high values over a
finite period. Hence, the concept of "hot spots" seems to be somewhat vague
and difficult to defend scientifically. However, in the search for hot
spots, a significant quantity of data has been collected on CO levels in
downtown areas, but most have not been published.
Similar efforts also have been conducted by transportation agencies to
assess the effects of traffic on CO levels near highways. For example,
Baker(36) collected CO data next to highways at Lake Tahoe for the State of
California's Department of Transportation.
2. Occupational Exposures
The early studies of occupational exposure to CO are compiled in a
bibliography by Cooper,(37) and some are discussed in the various air
quality criteria documents.(38,39) A recent study by Jabara £££^.,(40) is
reviewed here primarily because they used the new, miniaturized PEM's.
(Their instrument is similar to the General Electric instrument used in our
investigation, with some exceptions. The instrument used by Jabara et al.,
is smaller and does not show the instantaneous reading on a digital readout
display. Their instrument integrates concentration by its internal
electronics. The integrated CO value (ppm - hours) can be read from a
"console" at the end of the sampling period, which can be 8-hours or more.)
Using these instruments, Jabara ejt al., collected 8-hour exposures of
65 employees of the Denver, Colorado, Police Department who worked close to
heavy traffic for extended periods of time. These employees enforced
parking regulations, directed traffic at special events, and responded to
all traffic problems. Office employees served as a control group. The 33
control employees, plus the 65 experimental employees, gave a total of 98
subjects, each monitored for many 8-hour periods from September 1978 to
January 1979. Breath samples also were collected to allow measurements of
blood carboxyhemoglobin (COHb).
The CO concentrations are summarized below:
Nonsmokers Smokers
Median
(ppm)
7.5
20.4
Range
(ppm)
4.2-22.1
7.8-44.3
Median
(ppm)
7.1
22.1
Range
(ppm)
4.1- 9.6
8.0-55.3
Controls
Experimental
Some members of the experimental group exceeded the 8-hour CO standard
of 35 ppm, which is recommended by the National Institute of Occupational
Safety and Health (NIOSH). In some cases, smokers even exceeded the
time-weighted standard of the Occupational Safety and Health Administration
15
-------
(OSHA), which is 50 ppm for 8 hours. Among controls, the median
concentration was greater for nonsmokers than for smokers, contrary to
expectations. This result was attributed to five high readings among
employees who worked in offices above a parking garage in the police
station. The investigators concluded that CO leaked upstairs from the
garage on very cold mornings when its doors were closed and vehicles were
warmed up before leaving.
In the experimental group, CO exposures depended on job type. The six
members of the "special enforcement" section spent little time working close
to traffic. In contrast, the 29 "meter people" and 6 "highway" section
officers, who spent considerable time working in heavy traffic, were exposed
to higher CO concentrations:
CO 8-hour
Section Exposure (ppm)
Special Enforcement 11.0
Meter 21.8
Highway 27.3
When the investigators compared exposure data with ambient measurements
at fixed stations, they found that fixed stations gave only a very rough
estimate of personal CO exposures and were considerably lower. They
concluded that "...exposure to ambient CO levels, as monitored at fixed
sites, comprise only a small portion of the total occupational exposure."(40)
The breath samples showed that the percentages of COHb in the blood of
smokers were approximately four times greater than those of nonsmokers:
Nonsmokers Smokers
Controls 2.14% 8.7%
Experimentals 3.04% 11.2%
Among all experimental subjects, the COHb levels were higher than for
all control subjects. However, these differences were not as striking as
the differences between smokers and nonsmokers. The investigators found
that COHb levels of smokers before work were higher for the experimental
group than for the control group, perhaps the result of "...some carryover
from the preceding day."(40) Because the experimental group had high COHb
levels after work, the authors suggest that this group may not have
completely eliminated its COHb burden before beginning work the next day.
Even among nonsmokers, there were three instances in which COHb levels after
work exceeded the NIOSH recommended limit of 5.0%.
Bellin and Spengler(41) measured the CO exposures of baggage handlers
at Boston's Logan International Airport using the Series 2000 Ecolyzer
monitors. Both the upper (open) and lower (semienclosed) levels of the
entrance area of one terminal were monitored. Monitoring was performed
intermittently between 12:00 noon and 10:00 p.m. for more than 100 hours
during 14 days during the last 3 months of 1978.
16
-------
Based upon 15-minute averages, the indoor (ticket counter)
concentrations were significantly lower than the outdoor (curb-side)
concentrations. The 95% confidence interval for all indoor locations ranged
from 5.7 to 6.8 ppm, while the same interval for all outdoor locations was
9.7 to 12.5 ppm. The 1-hour maximum concentrations were 15 ppm (indoors)
and 23 ppm (outdoors), both occurring on the same day. The authors observed
no violations of OSHA's occupational standard of 50 ppm for 8 hours, nor
violations of the 1-hour NAAQS for CO of 35 ppm. Violations of the 8-hour
NAAQS for CO of 9 ppm could not be determined.
3. Explanatory Studies
Various researchers have studied CO concentrations on sidewalks near
roads and attempted to relate "curb-side" CO concentrations to traffic
volume and meteorological factors. A number of additional papers discuss
various studies of CO concentrations as a function of distance from streets,
many of them summarized in the literature review by Ludwig and Kealoha.(42)
Numerous models also have been developed for calculating concentrations as a
function of distance from the roadway and of various meteorological
factors. Many are discussed in a book by Noll and Miller.(43) Rao et
£^.,(44) also have undertaken comparative studies to evaluate the
performance and effectiveness of various microscale highway air pollution
models. Finally, a mathematical model of CO exposures in traffic has been
developed by Simmon and Patterson.(45)
McCormick and Xintaras(46) carried out one of the early "curb-side"
studies. They measured CO concentrations and traffic volumes in Nashville,
Tennessee, and Cincinnati, Ohio. While they found good qualitative
agreement between CO concentrations and traffic volumes, they stated that
"...the statistical correlation between these quantities was not impressive,
since it was on the order of 0.6 in both cities." They also reported very
low correlations between, on the one hand, CO concentrations and, on the
other, hourly wind speed, wind direction, and stability. They concluded:
"This investigation has shown that in urban areas some order can be found
but that a complete interpretation of such 'curb-side1 data in terms of
source and meteorological factors is very complex."
Brief, Jones, and Yoder(47) counted all cars and trucks passing a
selected location for 5 to 10 minutes each half hour and simultaneously
measured CO concentrations at this location. They found that CO
concentrations were highly correlated with traffic volume and obtained the
following regression equation:
CO = 0.281 + 0.136T (1)
where:
CO = the concentration of carbon monoxide (ppm); and
T = the traffic volume passing a sampled location (vehicles/minute).
17
-------
They speculated that other factors besides the number of vehicles per minute
also affected CO concentrations. For example, if traffic is congested and
vehicles pile up, they suggested that the number of cars per 1,000 square
feet near the sampled location might be a better measure of CO emissions
than traffic volume. They also developed relationships for lead
concentrations, CO -concentrations, and traffic volume.
Ramsey(48) attempted to collect a more representative sample than the
one collected in the previous study. Instead of one location, Ramsey
surveyed 50 intersections and other urban sites during a 6-month period in
Dayton, Ohio. He also related concentrations of CO to time of the day, day
of the week, and season of the year, in addition to traffic volume and
meteorological conditions. He reported a strong relationship between
traffic volume and CO concentrations at intersections, ranging from 56.1 ±
18.4 ppm for heavy traffic (94 readings), to 31.4 ± 31.5 ppm for moderate
traffic (108 readings), to 15.3 ± 10.2 ppm for light traffic (73 readings).
Ramsey reported that concentrations were greater at intersections along
major arteries somewhat removed from downtown, perhaps because they have
more traffic lanes than arteries closer in. The mean concentration at these
intersections was 3.4 times the mean of intersections a block away and
perpendicular to the axis of the arterial. Finally, concentrations were
highest during periods of high atmospheric stability.
Besner and Atkins(49) further investigated the relationship between
motor vehicle pollutants and perpendicular distance to the axis of an urban
expressway in Austin, Texas. CO and lead were measured using NDIR at two
sites, one 16 feet and the other 95 feet from the road. At both sites, they
found a strong positive association between CO and lead concentrations, but
they did not find a significant linear correlation between lead and traffic
volume. However, they reported a decline of CO concentrations with greater
distance from the road, verifying Ramsey's finding. At the site nearest the
road, CO concentrations ranged from 3.4 to 6.0 ppm, while at the other site
they ranged from 2.4 to 3.9 ppm.
\
Colucci and Begeman(50) departed substantially from the approach of the
previous two studies. They measured CO concentrations in Detroit, New York,
and Los Angeles at 12 locations which varied not only in distance from a
roadway but in land use. Roadway distance ranged from 2 to 150 meters,
although most ranged from 2 to 6 meters. Data were collected on weekdays
and weekends, as well as during Autumn and Spring.
In general, Colucci and Begeman found that CO concentrations wer.e about
50% higher in Los Angeles than in New York pr Detroit. The highest average
CO concentrations were observed at Pico Boulevard in Los Angeles, where they
measured 38 ppm for 1 hour, 27 ppm for 8 hours, and 17 ppm for 24 hours. By
comparison, 10.5 ppm was the highest weekday concentration of CO observed in
New York at Herald Square. In Detroit and New York, CO concentrations
correlated well with traffic volume, but less so in Los Angeles.
The authors also reported that concentrations generally were highest in
the commercial areas of these cities, intermediate in the freeway areas, and*
18
-------
20% lower than on weekdays. Finally, concentrations generally were highest
in Autumn and lowest in Spring, reflecting an inverse relationship with wind
speed and stability which varied with season of the year.
A number of investigators have looked at the vertical distribution of
CO concentrations for outdoor settings near traffic. Georgii, Bush, and
Weber(51) made many measurements at various heights along four streets in
Frankfurt, West Germany. They used these data to develop equations
expressing CO concentrations as a function of traffic volume and height
above the roadway. Schnelle, Ziegler, and Kreukel(52) developed similar
empirical relationships for the vertical distribution of CO as a function of
meteorological factors at an urban intersection. Johnson et al.,(53)
developed an urban scale diffusion model that contained a microscale
submodel for calculating CO concentrations at various heights above the
sidewalk in urban street canyons (downtown streets surrounded on both sides
by tall buildings). They applied their diffusion model to streets in San
Jose, California, and St. Louis, Missouri. Finally, DeMarrais(54,55)
examined the relationship of hourly CO concentrations measured by several
state air quality networks to traffic counts and meteorological variables.
He concluded that the daily variations of CO concentrations correlated
poorly with traffic patterns, which he ascribed to daily variations in
vertical mixing and wind speeds.
D. DISCUSSION
Many of the studies described concluded that fixed air monitoring
stations do not accurately reflect the true exposures of the population-,
especially the exposure of occupational groups such as traffic police. Such
personal exposures to CO can be affected by highly localized phenomena,
including traffic, exhaust leakage into passenger compartments, gas
appliances, and smoking. Furthermore, concentrations are modified by
horizontal and vertical distance from the source, as well as by
meteorological factors, such as wind speed and stability; which in turn vary
in time and place. Fixed stations cannot be sensitive to localized
^phenomena because they are too remote from them. However, the public is not
remote from these phenomena and does risk personal exposure. Consequently;
personal exposures should be monitored. However, these monitors are new,
and the methodology for their use is not well developed.
Unfortunately, localized phenomena can reveal dramatic results, which
may explain why so many investigators have been preoccupied with explaining
these phenomena. Their studies have shown that people can be exposed to
high levels of CO in many isolated locations and at isolated times. The
studies failed to report, however, just how many people were exposed, and if
any"of these exposures were so long that they violated Federal standards.
Ideally, future studies should monitor the public, not specific settings,
although such settings are easier and cheaper to study.
Finally, several studies revealed high CO concentrations in commercial
settings. However, studies of specific commercial sites have been limited
to nightclubs, restaurants, and a few office buildings. The present study
attempts to expand that range. Office buildings especially need more
19
-------
attempts to expand that range, because many people spend nearly 8 hours or
more there. Past studies of garages have given very high CO levels, often
above 100 ppm. Until recently, however, few studies had been made of these
settings. Some studies have indicated that smoking can be associated with
levels of 7 to 9 ppm in some commercial settings. However, the contribution
from smoking may be exaggerated because the studies failed to account for
ambient CO levels. For cities covered in the present study, ambient CO
levels measured outdoors and more than 100 yards from streets with traffic
generally were negligible, simplifying the analysis of the components of
concentrations measured in different settings.
20
-------
CHAPTER 3
METHODOLOGY
The methodology of this study is presented in this chapter. Part A
describes the CO PEM's employed in the field surveys. Part B discusses
monitoring procedures and quality assurance tests. The sampling procedures
are presented in Part C, and the geographic locations and commercial
settings that were visited are described in Part D. Finally, Part E
discusses the procedures for data reduction employed in the study.
A. PERSONAL EXPOSURE MONITORS
The PEM's used in this project consisted of the Energetics Science,
Inc. (ESI) "dosimeter" (Model 9000) and the General Electric portable
"detector" (Model 15ECS3C01). (See Appendix A for photographs.) In both
instruments, CO is oxidized to carbon dioxide (C02) in an electrolyte in
the following reaction:
CO + H20 -> C02 + 2H+ + 2e~ (2)
The resulting electrical signal is proportional to the partial pressure of
CO contained in the gas sample passing through the cell. As a result, a
current flows between the sensor and counter electrode, and this reaction
occurs at the counter electrode:
02 + 4H+ + 4e~ -v 2H20 (3)
Thus the overall reaction in the cell is:
2CO + 02 + 2C02 (4)
Although the measurement principles are the same, the GE and ESI
instruments differ in how the principle is implemented. The sensing cell in
the GE instrument contains a hydrated solid polymer electrolyte membrane and
is filled with deionized (distilled and filtered) water. Gas sensing and
reference electrodes are bonded to one side of the membrane, and a
collecting counter electrode is bonded to the other side. The circuit is
compensated to correct for shifts in the sensor span signal due to
temperature changes. No zero compensation is required. The only periodic
maintenance required is to fill the cell with deionized water. At low
temperatures the water freezes and the GE monitor becomes inoperable, a
problem that we did not encounter in our study. The cell is relatively
insensitive to changes in humidity. The GE instrument does not react to
21
-------
mild vibration encountered in driving or normal walking, although it does
react to severe vibration. The instrument weighs 0.5 kilogram and is fully
self-contained with its own batteries, pump, pump flow regulator, flow rate
meter, chemical filter (potassium permanganate on alumina), span and zero
controls, electronics, and liquid crystal display digital readout device.
It can operate a minimum of 10 hours.
Unlike the GE instrument, the ESI Ecolyzer personal monitor uses a
sensor that contains a liquid aqueous electrolyte. The electrolyte is
maintained by a "humidification process" in which moist air is pumped
through the cell. The Ecolyzer does not contain its own pump or readout
device, but a.companion digital readout device available from ESI can be
connected to the monitor to give both continuous and integrated readings.
Attached to the monitor is a portable pump manufactured by Dupont (Model
P-200). The time range is a minimum of 8 hours.
B. MONITORING PROCEDURES AND QUALITY ASSURANCE TESTS
One ESI instrument and four GE instruments were available for this
project. Not all were employed for each field survey. Before and after
each survey, any instruments used were zeroed and spanned with either 48 ppm
or 60 ppm CO (NBS traceable). Very little, if any, drift was observed in
the instruments for each survey.
The GE detector was attached to a shoulder strap, with the intake probe
located approximately 1 meter above the ground. The ESI instrument and its
attached readout device and pump were placed on a clipboard, which was held
by hand with the intake probes located at a height similar to the probes of
the GE detector. The instruments were separated in horizontal distance
about 1 meter (range 15 centimeters to 6 meters). The display device of
each instrument permitted the reading of the CO concentration in ppm at any
instant of time. Because CO levels near traffic could change rapidly over
time, instantaneous readings were taken at 1-minute intervals.
Because this measurement approach differs from the Federal reference
method, NDIR, we had to perform quality assurance tests whenever two
instruments were operated simultaneously side-by-side. This occurred for 11
of the 15 surveys (73%) conducted in the field.
We performed two types of quality assurance tests. The first tested
the precision of the instrument readings. To perform this test, the
absolute difference in recorded values was computed as follows:
di = lxl ~ X2l (5)
where:
d£ = absolute difference in recorded values for pair i of
simulataneous observations;
xi = CO concentration in ppm recorded from instrument #1; and
X£ = CO concentration in ppm recorded from instrument #2.
22
-------
The value of d (the mean difference) was generally lowless than 1 ppm
for 7 of the 11 surveys (Table 1). Furthermore, the highest d occurred on
the first survey, the lowest on the llth, and the second lowest on the
15th. This pattern suggests that the value of d£ was not totally random
over time, but dropped as we improved the calibration procedure for zero and
span.
The second type of quality assurance test determined if the variation
in CO concentrations recorded by the two instruments was in the same
direction. As CO concentrations change over time, one instrument should not
increase in value as the other decreases. The correlations between
simultaneous measurements were evaluated using
Pearson's product moment coefficient.(56)
For 9 of 11 surveys, these coefficients equalled or exceeded 0.97
(Table 1). The z statistic(56) for each coefficient indicated that one may
reject the hypothesis that the correlation was zero with a probability of p
of rejecting a true null hypothesis. In every case, the value of p was
extremely low.
In general, these quality assurance tests indicated very good agreement
between concentrations recorded by two instruments. These results were
encouraging, with one caveat. The discrepancy in recorded values between
instruments increased in significance when observed CO concentrations were
very low. For example, if typical CO concentrations in a given setting were
on the order of 1 or 2 ppm, a 1-ppm difference in recorded values
represented a relatively large error. For typical CO concentrations on the
order of 50 ppm, a 1-ppm difference represented a small relative error. In
this study there were many more settings with very low readings than with
very high readings.
Error can usually be reduced by increasing the number of samples
obtained in a given setting, in accordance with the central limit theorem.
If a large number of CO measurements are taken, CO concentrations which are
either higher or lower than the "true" value will tend to offset each
other. In this study we obtained a large number of measurements for generic
settings (e.g., retail stores of all kinds) and for some specific settings
(e.g., a particular shoe store). In other words, the mean CO concentrations
reported for retail settings in general have greater precision than the
concentrations reported for a specific store, unless that store was sampled
intensively. Furthermore, measurement error can be minimized only if there
are no systematic biases, such as improper calibration to reference gases.
Consequently, we made an extensive effort to calibrate the instruments with
both zero and span gases before and after each survey, as well as during
some surveys. The calibration gases themselves were assumed to be accurate.
C. SAMPLING PROCEDURES
With so many different kinds of cities, plus different commercial
establishments and locations within each city, it was not possible to obtain
a representative sample of all possible locations visited by members of the
general public. Nor was it possible to obtain a sample representative of
23
-------
TABLE 1. STATISTICAL SUMMARY OF QUALITY ASSURANCE TESTS
Descriptive statistics
Survey
1
2
3
4
5
7
10
11
12
13
15
Mean
(d)
(ppm)
2.32
0.73
0.88
0.66
0.65
1.37
1.02
0.11
0.62
1.11
0.26
Std.
dev.
(s)
(ppm)
3.52
1.80
1.14
0.84
0.88
1.68
0.86
0.85
1.14
1.01
0.89
Number of
paired CO
samples
n
88
354
171
116
185
95
117
168
217
47
148
Correlation statistics
Pearson's
coef f .
r
0.98
0.87
0.97
0.99
0.99
0.99
0.99
0.98
0.99
0.83
0.98
z
statistic
9.14
16.35
11.96
10.62
13.43
9.60
10.66
12.59
14.55
5.63
11.88
Probability
P
1.2 x
3.3 x
1.5 x
1.3 x
1.5 x
1.5 x
1.6 x
1.7 x
5.9 x
9.1 x
1.8 x
10-10
10-10
10-10
10-10
10-10
10-10
10-10
10-10
10-10
10-9
lO-10
1706
24
-------
all possible meteorological and traffic conditions during an entire year.
Such representative samples would require listing of every possible location
and time of visit by the public prior to field surveys. Such a sampling
procedure was too costly and time consuming, considering the financial
support available for this project.
Given these constraints, plus the absence in previous studies of
commercial settings and of a rigorous sampling procedure, our procedure
evolved over time. Initially, we followed an exploratory approach to gain
an appreciation of the factors which might affect the spatial and time
variation of CO concentrations in commercial settings. We decided on the
frequency, duration, and location of field surveys according to six
guidelines:
* Samples should be taken in settings with high CO emissions, using
average daily traffic (ADT) counts as surrogate indicators
whenever available, to identify the more heavily traveled streets
in these zones;
* Samples should be taken in settings which are intensively used by
workers, shoppers, or pedestrians, and land use maps should be
used to locate such settings, which normally are found in the
commercial zones of a city;
* Samples should be taken in settings in which CO concentrations
vary due to differences in distance and height to sources of CO,
indoor/outdoor location, and shielding, such as doors;
* Samples should be taken in different geographic locations to
capture differences in areawide emissions, meteorological
conditions, and types of commercial facilities;
* For some geographic locations, samples should be taken in a large
variety of indoor commercial establishments which vary in total
volume, number of smokers, and ventilation conditions;
* For some geographic locations, samples should be taken at various
hours of the day, days of the week, and seasons of the year.
With the limited resources available, we sought to attain these
guidelines to the fullest extent. However, with observations on only 15
dates, our research should be regarded as just a pilot study rather than an
exhaustive and comprehensive field survey. These guidelines achieved a data
base with varied CO concentrations. Such variation could then be associated
statistically with observable physical, spatial, and time factors to define
and classify appropriate microenvironments. Hence, although 15 dates of
data were not sufficient to adhere to these guidelines completely, they did
permit us to make comparisons and to explore the factors which affect
variation in CO concentrations.
While these guidelines were helpful, they did not pinpoint specific
locations (addresses) in the field at which we should monitor. Neither did
25
-------
these guidelines specify the duration of monitoring at specific sites, nor
the frequency of return visits. Such decisions followed some less formal
guidelines plus our discretion. First, we generally walked a path similar
to that of other shoppers, but with random choices of city blocks, streets,
and stores within a defined geographic location. If a specific street
section was the sole focus of study, both sides of the street were sampled.
If a specific tall building was the sole focus of study, lobbies on
different floors were chosen at random. Second, specific sites along a path
were selected randomly. Very seldom did proprietors refuse us admittance or
harass us in any way. Furthermore, despite some embarrassment, we entered
some sites that generally catered to women shoppers only. Third, once at a
site, either indoors or outdoors, we generally recorded 3 to 10 CO
concentrations per instrument. More than 10 recordings were generally taken
if the CO concentrations were unusually elevated. Such "hot spots"
typically warranted several return visits to determine the persistence of
and causes of the higher concentrations.
We followed these guidelines for each survey except for the one
conducted on 13 June 1980 in the Union Square district of downtown
San Francisco. On this occasion, the study site was defined by two
criteria: (1) average daily traffic exceeded 10,000 vehicles on a given
street; and (2) that area designated C-3-R ("downtown retail") as shown on
the San Francisco land use map. (See Figure B-l in Appendix B for a
simplified map.)
Four investigators, all males, conducted the survey. Each block was
numbered, randomly selected, and then assigned to one investigator. Each
investigator was assigned two or three blocks, depending upon the size of
the block.
Investigators were instructed to begin with the first block on their
list starting at the northwest corner. Next, they took three 1-minute
instantaneous observations of CO concentrations at each corner of that
intersection, proceeding clockwise around the intersection. Then they
proceeded clockwise around the block, taking three 1-minute instantaneous
observations of CO concentrations at establishments with an entrance at
street level. Observations were made for every multistory establishment,
such as department stores, and for every third establishment, such as small
shoe stores, occupying only the first floor.
The investigator measured both outdoor and then indoor CO
concentrations, usually noting whether the entrance door was opened or
closed. For indoor observations, only the street-level floor was sampled.
This entire procedure was then repeated for the next block on the list of
assigned blocks. Most surveyors completed at least two blocks in a 5-hour
period. The information was recorded on a form (Table B-2 in Appendix B),
which was attached to a clipboard.
D. GEOGRAPHIC LOCATIONS AND COMMERCIAL SETTINGS
We conducted 15 field surveys of CO concentrations in four geographic
locations of the metropolitan San Francisco Bay area and in one location of
26
-------
the Los Angeles metropolitan area (Figures 1 and 2). The four Bay area
sites were selected such that they provided variation in the volume of
traffic, in addition to variation in micrometeorological conditions. These
four sites included the Financial-Chinatown district (Figure 3) and the
Union Square district (Figure 4) of San Francisco's central business
district, plus the commercial shopping streets of University Avenue of
downtown Palo Alto (Figure 5) and Castro Street of downtown Mountain View
(Figure 6). Both University Avenue and Castro Street intersect with
stations serving the commuter railroad operated by the Southern Pacific
Railroad Co. Consequently, both streets are susceptible to both heavy
automobile and pedestrian traffic. Finally, Westwood Village in West Los
Angeles (Figure 7) was selected primarily to include a meteorological
setting different from that of the San Francisco Bay area.
Exactly 5,000 instantaneous samples of CO concentrations were collected
at 1-minute intervals between November 1979 and July 1980, at various hours,
dates, and locations (Table 2). To catch the heaviest pedestrian activity
along streets, we sampled those hours when stores and offices were open for
business. Furthermore, nearly every minute of the working day was sampled
in the 15 surveys. Samples were collected over a number of months to give
some indication of the effect of seasonal variation on CO concentrations.
Specifically, for the Union Square district of San Francisco, samples were
collected at four different times over an 8-month period. For the
University Avenue shopping street of Palo Alto, samples were collected at
five different times over a 7-month period.
Our study was limited to daytime and weekday hours of operation for
commercial settings. If we assume that our commercial settings were open
all 5 days a week and 8 hours each day for a 9-month period from November
through July, total business hours of operation for just one setting equal
1,560 hours. Our sample of 5,000 minutes is equivalent to 83.3 hours. If
we further assume that most commercial settings have similar CO
concentrations, then our observations represent about a 5.3% sample of total
business hours for this 9-month period.
Altogether, 588 different commercial settings were visited, both
indoors (including department stores, hotels, office buildings, parking
garages, retail stores, restaurants, banks, travel agents, and theaters) and
outdoors (including street intersections, sidewalk locations in the middle
of blocks, arcades, parks, plazas, and parking lots) (Table 3). Typical CO
concentrations for both indoor and outdoor commercial settings broken down
by geographic location are summarized in Appendix C.
E. PROCEDURES FOR DATA REDUCTION
The raw data consisted of instantaneous CO concentrations collected at
1-minute intervals from a given PEM. Simultaneous observations taken from
two monitors were highly correlated. Thus, in our quality assurance tests,
the average of two instantaneous, 1-minute observations taken simultaneously
was computed first, whenever two separate instruments were employed
side-by-side.
27
-------
o Oakland
Financial-Chinatown
PACIFIC OCEAN
Figure 1. Four geographic locations in the San Francisco Bay area.
28
-------
kWestwood Village
PACIFIC OCEAN
Los Angeles
Miles
0 10
Figure 2. One geographic location surveyed in the Los Angeles metropolitan area.
29
-------
* +-^
*f
co
c
o
CO
o
CO
-3
Battery St.
Custom House PI.
Sansome St.
Hotaling St.
^^
o
c
CO
CO
Wentworth St.
Merchant St.
.*-*
co
CO
O
Portsmouth
Square
Brenham St.
Front St.
5)
'o
Comme
Sacramento St.
55
o
I
Leidesc
California St.
Jorff St.
Montgomery St.
Spring St.
Kearney St.
Grant St.
Figure 3. Financial - Chinatown district of San Francisco, California.
30
-------
Bush St.
N
Figure 4. Union Square district of San Francisco, California.
31
-------
Stanford University Campus
" -.
1 1 1 1
1 1 1 1
I
05
3.
o
<
CD
1 1 1 1 1 1
1 1 1 1 1 1
>
^X
City Hall
C
CD
-^
w
l-t-
CD
' ,
I I I 1 1 1 1 1
1 1 1 1 1 1 1 1
v
x^
Florence St.
15-story office bldg.
Tasso St.
mmm^^m^*m
I I I I
I I 1 I
0
CD
EICaminoReal
Alma St.
High St.
Emerson St.
Ramona St.
Bryant St.
Waverly St.
Kipling St.
Cowper St.
Figures. University Avenue central business district of Palo Alto, California.
32
-------
City Hall.
School Way
Central Expressway
I I I I1_|I I Southern Pacific RR
Evelyn Ave.
O
W
21
o
CO
I
o
"O
CD
CO
Villa St.
Dana St.
California St.
Mercy St.
Church St.
Yosemite Ave.
Fairmont Ave.
El Camino Real
Figure 6. Castro Street central business district of Mountain View, California.
33
-------
UCLA Campus
Le Conte Ave.
Figure 7. Westwood Village district of Los Angeles, California.
34
-------
TABLE 2. FIELD SURVEY DATES, HOURS, LOCATIONS, AND NUMBERS OF CO SAMPLES
CO
Ul
Survey
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
9
13
24
31
7
6
7
13
27
4
11
11
9
13
11
Date
Nov 1979a
Dec 1979a
Jan 1980a
Jan 1980a
Feb 1980a
March 1980
March 1980a
March 1980
March 1980
April 1980a
April 1980a
April 1980a
May 1980a
June 1980
July 1980a
Hour
10
8
2
9
11
11
4
10
2
1
10
2
1
10
2
:13
:42
:00
:20
:14
:00
:22
:00
:26
:42
:55
:06
:20
:44
: 18
a
a
P
a
a
a
P
a
P
P
a
P
P
a
P
»TD .
*m .
.m .
«m .
.m .
*m .
.m.
.m.
.m.
.m.
.m.
.m.
.m.
.m.
.m .
- 11
- 4
- 4
- 11
- 2
- 12
- 6
- 1
- 7
- 3
- 1
- 5
- 2
- 3
- 6
:41
:45
:37
:18
:30
:30
:00
:15
:27
:46
:47
:58
:07
:09
:32
a.m.
p.m.
p .m.
a.m.
p.m.
p.m.
p.m.
p .m.
p.m.
p.m.
p.m.
p .in.
p.m
p.m.
p .m.
Geographic location
Union Square, San Francisco
Union Square, San Francisco
University Avenue, Palo Alto
University Avenue, Palo Alto
University Avenue, Palo Alto
Financial-Chinatown district,
San Francisco
University Avenue, El Camino,
Palo Alto
Castro Street, Mountain View
Westwood Village, Los Angeles
15-Story building, Palo Alto
Financial-Chinatown district,
San Francisco
Union Square, San Francisco
15-Story building, Palo Alto
Union Square, San Francisco
University Avenue, Palo Alto
No. of CO
samples
181
766
342
232
370
89
190
181
300
234
336
434
94
787
464
5000
Note:
a. Instruments were deployed side-by-side on these dates during some portion of
the sampling.
-------
TABLE 3. NUMBER OF COMMERCIAL SETTINGS BY TYPE OF SETTING AND GEOGRAPHIC LOCATION
1
Commercial
setting
Indoor
Department stores
Hotel
Office buildings
Parking garages
Retail goods
Restaurants, etc.3
Services
Theaters
Subtotal
Outdoor
Arcades, etc.b
Intersections
Midblocks
Subtotal
Grand Total
Union
Square
District,
San Francisco
6
13
11
3
51
14
11
2
111
1
30
155
186
297
Geographic
Financial-
Chinatown
District ,
San Francisco
0
1
7
2
5
5
4
0
24
10
15
24
49
73
location
University
Avenue ,
Palo Alto
0
1
3
3
33
5
7
1
53
3
11
66
80
133
Castro
Street,
Mountain
View
0
0
0
0
8
2
1
0
11
2
9
16
27
38
Westwood
Village,
Los Angeles
0
0
2
2
14
2
1
0
21
2
8
16
26
47
Total
6
15
23
10
111
28
24
3
220
18
73
277
368
588
Notes:
a.
b.
Includes cafeterias, coffee shops, delicatessens, bars, lounges,
restaurants, and sandwich shops.
Includes arcades, parks, plazas, and parking lots.
-------
Most of our analyses focused on the commercial setting. However, the
length of a visit to each commercial setting varied. Often, more CO
readings were taken at settings with higher concentrations than settings
with lower concentrations. Consequently, the average concentration per
visit to a given setting was computed second. Finally, if several visits
were made to the same setting, the average concentration was computed based
upon the total number of visits; we gave equal weight to each visit
regardless of the number of observations per visit.
Procedures for further data reduction depended upon the purpose of the
analysis. Furthermore, they depended upon an understanding of how CO
concentrations vary in time and space. The next two chapters discuss the
time and spatial variations of CO concentrations measured in commercial
settings.
37
-------
CHAPTER 4
TIME VARIATION OF CO CONCENTRATIONS
This chapter describes the time variation of CO concentrations measured
in commercial settings using personal exposure monitors. Part A considers
settings monitored for extended periods of time without interruption. Part
B looks at settings visited twice on the same date. Finally, Part C
describes settings visited on five dates that differed in recorded wind
speed.
A. SETTINGS MONITORED FOR EXTENDED PERIODS WITHOUT INTERRUPTION
In studying the time variations of CO concentrations, we had to decide
how long to stay at a given setting. Casual observations during the early
surveys led us to suspect that indoor CO concentrations did not vary
appreciably over long periods of time, especially when readings were taken
in small physical spaces. Consequently, most of the commercial settings
were monitored for 2 to 5 minutes each, enabling us to sample a greater
variety of commercial settings.
Some settings were visited for longer times to evaluate the homogeneity
of CO concentrations. If concentrations were homogeneous for extended time
periods (longer than 5 minutes), samples collected over a shorter time span
could be assumed to be fairly representative of a longer time period.
We visited 74 commercial settings for times varying from 6 to 111
minutes (Table 4). The average of two simultaneously taken 1-minute samples
is the basic variable of interest. The mean CO concentration ranges from a
low value of 0 ppm, observed at a real estate office and an outdoor plaza,
to a high of 51.3 ppm, recorded for an enclosed parking garage. Despite
this variation, the standard deviation in CO concentrations for 74.3% of
these settings is less than 2 ppm, relatively small compared to the mean
concentration, regardless of its size. Furthermore, for six settings
(8.1%), the standard deviation is greater than the mean concentration. For
three, the mean concentration is near 0 ppm, and because the PEM is accurate
to only 1 ppm, the standard deviation exceeds the mean concentration.
Consequently, allowing for instrumental insensitivity, we see that for
nearly 96% of the settings visited for extended periods, the standard
deviation is quite small and less than the mean concentration.
The small standard deviations suggest that no matter when or where a
specific commercial setting, indoors or outdoors, was visited, its CO
concentrations were likely to be fairly homogeneous for the duration of the
38
-------
TABLE 4. STATISTICAL SUMMARY OF CO CONCENTRATIONS FOR COMMERCIAL SETTINGS VISITED
FOR LENGTHY PERIODS OF TIME ON THE SAME DATE
Survey
1
1
1
1
2
2
2
2
2
2
2
2
2
2
2
3
3
3
4
4
4
4
4
Date
9 Nov 79
9 Nov 79
9 Nov 79
9 Nov 79
13 Dec 79
13 Dec 79
13 Dec 79
13 Dec 79
13 Dec 79
13 Dec 79
13 Dec 79
13 Dec 79
13 Dec 79
13 Dec 79
13 Dec 79
24 Jan 80
24 Jan 80
24 Jan 80
31 Jan 80
31 Jan 80
31 Jan 80
31 Jan 80
31 Jan 80
Geographic
location
San Francisco
San Francisco
San Francisco
San Francisco
San Francisco
San Francisco
San Francisco
San Francisco
San Francisco
San Francisco
San Francisco
San Francisco
San Francisco
San Francisco
San Francisco
Palo Alto
Palo Alto
Palo Alto
Palo Alto
Palo Alto
Palo Alto
Palo Alto
Palo Alto
Commercial
setting
Garage
Florist
Sandwich shop
Garage
Garage
Coffee shop
Book store
Cutlery store
Dept. store #1
Restaurant
Dept. store #2
Dept. store #1
Dept. store #3
Florist
Sandwich shop
Bank
Office bldg.a
Intersection^
Office bldg.#la
Garage
Coffee shop
Office bldg. #2C
Office bldg. #la
Mean
CO
cone.
(ppm)
35.4
6.7
7.9
34.2
16.8
2.6
7.0
3.8
6.8
4.6
4.6
2.7
2.7
4.3
11.7
5.9
13.3
2.9
5.0
28.3
3.3
4.1
6.9
Std. dev.
of CO
cone .
(ppm)
7.8
0.7
4.9
8.3
10.2
0.5
1.0
0.8
4.7
0.7
0.9
0.3
0.4
0.8
7.7
0.5
1.9
1.3
0.9
4.8
1.2
1.8
0.5
Range
of CO
cone.
(ppm)
21-44
6-8
0-17
20-48
7-37
1-4
5-10
3-5
3-17
3-8
4-7
2-4
2-4
3-5
3-32
5-7
8-15
0-7
3-7
23-37
3-4
2-8
6-8
Length
of
visit
(min. )
7
20
8
19
10
34
33
7
27
98
10
13
35
32
56
6
27
9
11
8
8
11
7
(continued)
-------
TABLE 4 (continued)
Survey
5
5
5
5
5
5
5
5
6
6
6
7
7
8
8
8
8
8
8
9
9
9
9
9
9
9
9
9
Date
7 Feb 80
7 Feb 80
7 Feb 80
7 Feb 80
7 Feb 80
7 Feb 80
7 Feb 80
7 Feb 80
6 March 80
6 March 80
6 March 80
7 March 80
7 March 80
13 March 80
13 March 80
13 March 80
13 March 80
13 March 80
13 March 80
27 March 80
27 March 80
27 March 80
27 March 80
27 March 80
27 March 80
27 March 80
27 March 80
27 March 80
Geographic
location
Palo Alto
Palo Alto
Palo Alto
Palo Alto
Palo Alto
Palo Alto
Palo Alto
Palo Alto
San Francisco
San Francisco
San Francisco
Palo Alto
Palo Alto
Mountain View
Mountain View
Mountain View
Mountain View
Mountain View
Mountain View
Los Angeles
Los Angeles
Los Angeles
Los Angeles
Los Angeles
Los Angeles
Los Angeles
Los Angeles
Los Angeles
Commercial
setting
Office bldg.#la
Sidewalk cafe
Real estate office
Card & gift shop
Electronics store
Office bldg. #la
Garage
Restaurant
Plaza #1
Plaza #2
Restaurant
Office bldg. #la
Office bldg. #3d
Intersection #1^
Midblock #1
Intersection #2
Midblock #2
Restaurant
Midblock #3
Midblock
Bus stop #1
Office bldg. #le
Intersection
Office bldg. #2f
Book store #1
Drugstore
Book store #2
restaurant
Mean
CO
cone.
(ppm)
3.5
0.3
0.0
0.1
0.3
4.0
27.9
1.5
1.3
0.2
2.0
8.5
2.6
13.7
1.7
1.6
6.3
2.7
2.4
4.2
30.2
1.7
4.7
4.6
3.5
2.8
2.0
4.9
Std. dev.
of CO
cone.
(ppm)
1.8
0.5
0.0
0.2
0.3
1.2
1.2
0.0
0.5
0.4
0.0
2.7
2.6
15.5
1.2
1.1
3.3
0.5
1.4
1.5
43.4
0.5
3.2
1.5
0.5
0.4
0.0
1.1
Range
of CO
cone .
(ppm)
1-6
0-2
-
0-1
0-1
2-6
27-31
-
1-2
0-1
-
3-14
0-7
2-45
0-4
0-4
3-11
2-3
0-4
2-6
9-159
1-2
1-10
3-7
3-4
2-3
0
3-7
Length
of
visit
(min. )
8
9
8
7
6
10
7
46
6
6
36
15
7
7
11
10
7
37
7
6
11
11
15
16
6
9
33
69
(continued)
-------
TABLE 4 (continued)
Survey
10
11
11
11
11
12
12
12
12
12
12
12
12
12
13
13
14
14
15
15
15
15
15
Date
4 April 80
11 April 80
11 April 80
11 April 80
11 April 80
11 April 80
11 April 80
11 April 80
11 April 80
11 April 80
11 April 80
11 April 80
11 April 80
11 April 80
9 May 80
9 May 80
13 June 80
13 June 80
11 July 80
11 July 80
11 July 80
11 July 80
11 July 80
Geographic
location
Palo Alto
San Francisco
San Francisco
San Francisco
San Francisco
San Francisco
San Francisco
San Francisco
San Francisco
San Francisco
San Francisco
San Francisco
San Francisco
San Francisco
Palo Alto
Palo Alto
San Francisco
San Francisco
Palo Alto
Palo Alto
Palo Alto
Palo Alto
Palo Alto
Commercial
setting
Office bldg. #18
Hotel lobby
Plaza
Restaurant
Garage
Garage
Book store
Coffee shop
Office bldg.h
Sandwich shop
Art gallery #1
Hotel #1£
Art gallery #2
Hotel #2J
Office bldg. #lk
Garage
Park
Restaurant
Real estate office
Res taurant
Office bldg. tfl1
Garage
Cocktail lounge
Mean
CO
cone.
(ppm)
17.9
1.9
0.0
1.1
15.8
45.9
6.1
3.2
3.9
9.1
1.3
4.5
1.1
5.3
8.7
19.1
2.2
8.6
0.0
0.7
7.1
51.3
2.3
Std. dev.
of CO
cone.
(ppm)
25.9
0.2
0.0
0.8
3.8
8.8
0.8
0.8
1.1
1.1
0.4
1.4
0.5
1.7
2.9
11.6
1.3
1.9
0.0
0.5
2.6
1.5
0.4
Range
of CO
cone .
(ppm)
1-38
1-2
-
0-4
12-24
30-62
5-8
2-5
3-6
5-11
0-2
2-7
0-2
2-9
4-22
1-34
1-5
6-11
-
0-1
2-11
48-53
1-3
Length
of
vis it
(min. )
111
7
7
30
6
8
12
7
8
11
6
16
11
24
39
8
12
7
7
6
48
9
77
(continued)
-------
TABLE 4 (continued)
Notes:
a.
b.
c.
d.
e.
f.
g-
h.
i.
j-
k.
Five levels (1, 3, 6, 11, 15) of a 15-story building.
One corner of intersection.
Four levels (I, 3, 6, 8) of an 8-story city hall office building.
Three levels (1, 3, 5) of a 5-story general office building.
Three levels (1, 6, 12) of a 12-story medical office building.
Five levels (1, 4, 8, 12, 16) of a 16-story general office building.
All levels (lobbies and stairwells) of a general office building.
All levels of a general office building.
Three levels (1, 6, 12) of a 12-story hotel.
21) of a 21-story hotel.
4, 6, 8, 10, 12, 14, 15) of a 15-story general
10, 15,
A, 1, 2,
20,
3,
Six levels (1, 5
Twelve levels (B
office building.
Eight levels (B, A, 1, 3, 6,
building.
9, 11, 15) of a 15-story general office
-------
visit. Compared to the momentary fluctuations in CO concentrations at a
given setting, the concentrations differ more from one setting to another
and even from one date to another for the same setting. This implies
that only a few samples are necessary to characterize the air quality of
a given commercial setting on a given date. Staying in a setting for
more than 6 to 10 minutes gives a great deal of redundant information.
There are exceptions to this generalization. In three casesan
intersection in Mountain View, a bus stop in Los Angeles, and an office
building in Palo Altothe standard deviation exceeded a rather large
mean concentration. In the intersection and the bus stop (itself also
near an intersection), this result can be explained by a few spikes in
concentration, which skewed the distribution. Without these spikes,
these settings also would be fairly homogeneous, at least for short
visits. The case of the general office building, sampled on 4 April 1980
in Palo Alto, is discussed in Chapter 6 in greater detail.
B. SETTINGS VISITED TWICE ON THE SAME DATE
Five intersections and 12 indoor settings were visited twice on the
same date, permitting us to evaluate the homogeneity of CO concentrations
in periods separated by from 43 to 470 minutes (Table 5). The number of
observations taken per visit was less than 10, too small to permit a
rigorous statistical analysis of whether or not the mean CO concentration
of the first visit to a given setting was equal to the mean concentration
of the second visit. Rather, the analysis must be qualitative.
The 17 settings cannot claim to represent the variety of all settings
visited as part of this study. These settings appear here for several
reasons. One is that the surveyors had particular reasons to visit some
settings twiceparking garages for example. Normally one parks a car
and returns to it later. Hence, garages compose nearly a third of all
settings visited twice in one day. Another reason for repeat visits is
that one's path in a particular geographic location may naturally cross
the same point on two occasionsintersections near garages, for
example. Third, the researchers consciously revisited some settings
because previous visits had revealed very high CO concentrations. Hence,
the stability of those high concentrations over time was in question.
The general office building located in Palo Alto is a good example.
Finally, some settings, such as the real estate office and variety store
located in Palo Alto, were revisited for personal reasons.
Our results show that the mean concentration on the second visit
differed greatly from that on the first for only 2 of the 17 settings
(Table 5). Values for the two visits to a parking garage differed by a
factor of about two, and in the case of a print shop by exactly four.
For the remaining 15 settings, the average difference in mean
concentration between visits was only 1.3 ppm, supporting the results for
settings monitored for extended periods without interruption.
Since concentrations at one of the six garages did not remain stable
over time, it may be useful for future research to make distinctions
43
-------
TABLE 5. STATISTICAL SUMMARY OF MEAN CO CONCENTRATIONS FOR COMMERICAL SETTINGS
VISITED TWICE ON THE SAME DATE
Survey
1
2
3
3
4
4
4
4
5
5
5
8
10
10
11
11
14
Commercial
setting
Garage
Garage
Intersection
Real estate office
Office building lobby
Garage
Variety store
Intersection
Office building lobby
Garage
Intersection
Intersection
Office building lobby
Garage
Garage
Intersection
Print shop
Mean CO
concentration
(ppm)
1st visit 2nd visit
35.4
16.8
1.5
3.5
5.3
28.3
2.1
1.8
2.0
26.8
0.0
1.0
11.6
40.0
16.5
1.5
32.3
34.5
37.1
2.9
3.4
7.0
31.6
2.6
1.5
3.8
27.9
2.0
0.0
8.6
41.6
15.8
1.5
8.0
Difference in mean
concentrations
between visits
(ppm)
0.9
-20.3
-1.4
0.1
-1.7
-3.3
-0.5
0.3
-1.8
-1.1
-2.0
1.0
3.0
-1.6
0.7
0.0
24.3
Time
between
visits
(min. )
122
470
142
155
130
103
44
84
115
98
84
179
57
43
168
70
69
-------
between different types of garages. Garages which predominantly provide
parking for shoppers may have more frequent arrivals and departures than
garages used by office workers. For example, people and cars come and go
frequently at the Union Square garage, which is in the midst of a
shopping area. The 121% increase in mean CO concentration, observed in
Survey 2, may be due to the almost 8 hours between visits. Virtually no
change in mean CO concentration was observed in Survey 1 for this garage,
perhaps because of the relatively shorter span (about 2 hours) between
visits and/or the time of day (midmorning period as stores were opening).
Both the garage located in Palo Alto (Surveys 4, 5, and 10) and the
one in the Financial-Chinatown district of San Francisco (Survey 11)
serve nearby offices. The peak vehicular movement periods in these
garages are the morning and evening rush-hour periods. Hence, if the
ventilation systems in these garages were working properly, CO
concentrations should drop during the midday. They did for the San
Francisco garage (Survey 11), dropping by 0.7 ppm after 3 hours. The
concentrations for the Palo Alto garage increased slightly on each of
three dates. The garage's maintenance engineer reported that the
ventilation system was shut down periodically during the day to conserve
energy, which perhaps explains the increases.
Ostensibly, the print shop would appear to belong in the same
category as the variety store and the real estate office, since all three
appear to have no indoor or nearby sources of CO. Yet, the mean
concentration was 32.3 ppm (average of three observations) on the first
visit. This extremely high concentration may have beem due to the van
that stood with its engine idling in the alley outside the shop's open
door. On a return visit 69 minutes later when the van was gone, a
concentration of 8.0 ppm was recorded.
For the print shop, the arrival and departure of delivery vans cannot
be predicted without more information about their schedules. Neither can
one predict whether or not the driver will leave the engine idling on a
given visit and whether or not the door to the shop will be open. Such
events are essentially random, and fortunately they are rare in most
settings.
Finally, there are some indoor settings for which the CO
concentrations varied only slightly over several hours. Such settings
can be protected by closed doors and properly functioning ventilation
systems. The effect of door position is discussed in the next chapter.
C. SAME SETTINGS VISITED ON DIFFERENT DATES
Commercial settings in Palo Alto's central business district had been
visited on five separate dates. For these dates, records of the wind
speed and direction were obtained for the nearest station of the Bay Area
Air Quality Management District, which is in Redwood City approximately
5.6 kilometers from the survey site. While the wind was generally from
the north and/or west on each date, the wind speed was three to five
times greater, depending upon the hour, on some dates compared to others
45
-------
(Table 6). Specifically, winds were light on 24 January, 31 January, and
7 March, and higher on 7 February and 11 July. To examine the effect of
wind speed on mean CO concentrations of commercial settings, we used a
one-way analysis of variance. Our statistical summary of CO
concentrations for each date (Table 7) included all indoor and outdoor
settings, except for garages, which, because of their high
concentrations, tended to seriously elevate the distribution. (See
Figure D-l in Appendix D for cumulative frequency distributions. Observe
that the line plots as approximately straight on logarithmic probability
paper. Such straightness implies a tendency toward a lognormally
distributed random variable.)
The one-way analysis of variance requires several assumptions.
First, the distribution of the mean CO concentration of commercial
settings visited on a given date must be normally distributed. The
central limit theorem states that as the sample size increases without
limit, whether the distribution of the population is normal or otherwise,
the sampling distribution of the mean approaches a normal
distribution.(56) Samples of 100 or more are usually sufficient to cause
the distribution of the mean to approximate the normal distribution, even
if the parent population from which the sample was drawn is radically
skewed. When samples of this size are unavailable, samples of 30 or more
are considered adequate to ensure close approximation of the normal.(56)
Since fewer than 30 settings were visited on 7 March, this date was
eliminated from further consideration here. However, in accepting
smaller samples we set a more stringent level of statistical significance
( a ) 0.01 instead of 0.05.
A second assumption is that the standard deviations of concentrations
are approximately equal. This assumption applies only to 31 January,
7 February, and 11 July. Hence, 24 January was eliminated from further
consideration.
A third assumption (made for ease of manual computation) is that the
number of commercial settings visited is approximately equal on all
dates. To meet this assumption, an equal number of settings was selected
randomly for each date. The mean and standard deviation were
recalculated for these three samples as follows:
Date x B n
1
2
3
The hypotheses to be tested were that:
IT . c= *f c= V
RI : at least one mean does not equal the others
46
31 Jan 80
7 Feb 80
11 July 80
3.56 ppm
1.43 "
1.26 "
1.54 ppm
1.65 "
1.78 "
33
33
33
-------
TABLE 6. METEOROLOGICAL SUMMARY FOR THE REDWOOD CITY STATION OF
THE BAY AREA AIR QUALITY MANAGEMENT DISTRICT FOR
DATES AND HOURS CORRESPONDING TO FIVE SURVEYS CONDUCTED IN
THE CENTRAL BUSINESS DISTRICT OF PALO ALTO, CALIFORNIA
Date
24 Jan 80
31 Jan 80
7 Feb 80
7 March 80
11 July 80
Hour
2 p.m.
3 p.m.
4 p.m.
10 a.m.
11 a.m.
12 noon
1 p.m.
2 p.m.
5 p.m.
6 p.m.
3 p.m.
4 p.m.
5 p.m.
6 p >m.
Wind
speed
(mph)
3
4
4
3
5
15
15
12
4
3
12
11
11
10
Wind
direction
NE
N
NW
NW
NW
N
N
NW
NW
W
W
W
W
W
47
-------
TABLE 7. STATISTICAL SUMMARY OF CO CONCENTRATIONS FOR
ALL COMMERCIAL SETTINGS, EXCEPT PARKING GARAGES,
VISITED ON FIVE DATES IN THE CENTRAL BUSINESS DISTRICT OF
PALO ALTO, CALIFORNIA
Date
24 Jan 80
31 Jan 80
7 Feb 80
7 March 80
11 July 80
Mean
CO
concentration
(ppm)
3.61
3.56
1.52
3.46
1.24
Std. dev.
of CO
concentration
(ppm)
2.50
1.54
1.65
3.06
1.56
Number
of
commercial
settings
58
33
48
13
81
48
-------
In a one-way analysis of variance, F = 19.68. With 2 degrees of freedom for
the variance between groups and 96 degrees of freedom for variance within a
group, p = 6.9 x 10~8. Thus, the null hypothesis was rejected as
p < a = 0.01, the assigned level of significance.
The means for 31 January and 11 July were sufficiently dissimilar to
cause the null hypothesis to be rejected. Also, the mean concentration for
7 February may be sufficiently different from the mean for
31 January. In testing this hypothesis using a t-statistic for independent
samples and pooled variance, t = 5.10. With 64 degrees of freedom, p = 3.3
x 10 , which is less than a = 0.01. Hence, we concluded that wind speed
has an effect on CO concentrations for both indoor and outdoor commercial
settings. Greater wind speeds appear to lower CO concentrations in
commercial settings. In this case, windy daysthose with minimum winds of
10 miles per hour (mph)reduced average concentrations to less than half
the average concentration computed for calm days (5 mph maximum).
A t-test comparing the mean concentrations for 11 July and 7 February
(dates with similar high winds) revealed a value of t = 0.40. With 64
degrees of freedom, p = 0.69. Since p > a = 0.01, the null hypothesis was
accepted. Thus, the mean CO concentration for settings surveyed on
7 February is equal to the mean concentration for settings surveyed on
11 July.
Thus, wind speeds appear to play a role in determining the CO
concentrations found in the commercial settings of a particular geographic
location. If two dates have dissimilar wind speeds, individual exposure to
CO varies more from one date to another than from one setting to another on
the same date, assuming the same geographic location applies for all dates
and settings. This result has at leas£ one implication for efforts to
estimate human exposure to air pollution through computer simulation. These
estimates must first cluster similar settings by dates with similar wind
speeds and perhaps other meteorological variables beyond the scope of this
study.
49
-------
CHAPTER 5
SPATIAL VARIATION OF CO CONCENTRATIONS
This chapter describes the spatial variation in CO concentrations for
commercial settings, which permits development of a classification scheme of
commercial microenvironments with respect to air pollution. Ideally: in
this scheme the variance of CO concentrations within each microeavironment
is relatively small and less than the variance between microenvironments.
This chapter is limited to small-scale spatial separations (both horizontal
and vertical), as opposed to variation in CO concentrations between
geographic locations, which is explored in Chapter 7.
Horizontal separations raise several important questions. The names
used to classify commercial settings, such as bank, restaurant, hotel, or
retail store, may or may not have relevance to air pollution monitoring. Is
one bank similar to or different from another bank in terms of CO
concentrations? How does an open entrance door affect the indoor CO
concentrations of a commercial setting which fronts a street with heavy
vehicular traffic? Do concentrations of CO peak at intersections and taper
off between intersections? Do CO concentrations vary greatly for different
intersections or even different corners of the same intersection? Are
opposite sides of the same street similar or different with respect to
concentrations of CO? Vertical separations raise questions such as whether
or not CO concentrations vary appreciably over the height of tall structures.
Part A of this chapter compares indoor concentrations, with doors and
opened and closed, with outdoor concentrations. Parts B and C examine the
horizontal spatial variation in CO for indoor and outdoor commercial
settings. Finally, Part D considers vertical variation in CO in tall
buildings.
A. COMPARISON OF INDOOR AND OUTDOOR SETTINGS
Indoor and outdoor concentrations of CO were compared using data
collected on 13 June 1980 for 88 commercial settings in the Union Square
district of San Francisco (Table 8). Blocks and settings had been selected
randomly through a process described in Chapter 3. At each setting the
surveyor took two sets of samples. The first set of 3 readings was taken
outdoors at points within 3 meters of the entrance. The second set of 3
readings was taken indoors within 6 meter,s of the entrance. The entrance
door (or doors) was recorded as opened or closed for 81 of the settings.
(Figure D-2 in Appendix D shows the cumulative frequency distributions.)
50
-------
TABLE 8. STATISTICAL SUMMARY OF CO CONCENTRATIONS MEASURED ON
13 JUNE 1980 AT INDOOR AND ADJACENT OUTDOOR COMMERCIAL SETTINGS
LOCATED IN THE UNION SQUARE DISTRICT OF SAN FRANCISCO, CALIFORNIA
Mean
indoor
Commerical concentration
Case setting (ppm)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
Clothing store #1
Clothing store #2
Bank #1
Office bldg. lobby #1
Clothing store #3
Department store #1
Clothing store #4
Jewelry store #1
Fabric store #1
Shoe store #1
Sandwich shop
Currency exchange
Drug store #1
Airline ticket office #1
Office building lobby #2
Shoe store #2
Office building lobby #3
Burlesque theater entrance
Novelty shop #1
Restaurant lobby #1
Clothing store #5
Delicatessen #1
Novelty shop #2
China & glass shop
Office building lobby #4
Clothing store #6
Office building lobby #5
Linen shop
Hotel lobby #1
Church
Restaurant lobby #2
Hotel lobby #2a
Airline ticket office #2
Bar
Hotel lobby #2b
Restaurant lobby #3
Hotel lobby #3
Movie theater lobby
0.7
1.7
1.0
2.0
1.7
1.0
3.0
1.0
2.0
3.7
3.0
2.7
1.3
1.3
2.3
1.0
5.0
6.7
3.3
2.3
3.7
4.7
4.0
3.7
2.3
1.3
0.3
0.0
3.3
1.3
1.3
1.3
4.3
5.7
6.0
3.0
4.3
0.0
Mean Position
outdoor of
concentration entrance
(ppm) door
2.3
0.7
2.3
4.3
7.7
1.7
1.3
2.0
0.7
4.3
3.0
8.0
1.7
1.3
4.3
1.7
5.7
4.3
3.7
6.0
4.3
10.0
3.0
2.0
2.7
2.7
0.7
0.0
3.7
1.7
6.7
2.3
5.0
6.3
16.3
4.0
2.7
0.0
Open
Open
Closed
Closed
Open
Closed
Open
Closed
Closed/open3
Open
Open
Closed
Open
Open
Open
Not recorded
Closed
Open
Open
Open
Open
Open
Open
Open
Open
Closed
Open
Open
Open
Closed
Closed
Closed
Closed
Closed
Closed
Closed
Open
Open
(continued)
51
-------
TABLE 8 (continued)
Mean
indoor
Commerical concentration
Case setting (ppm)
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
Office building lobby #6
Art gallery #1
Art gallery #2
Antique store
Stamp store
Hotel lobby #4
Bank #2
Furniture store #1
Private club
Hotel lobby #5
Restaurant lobby #4
Gift store #1
Gift store #2
Drug store #2
Delicatessen #2
Gift store #3
Hotel lobby #6
Hotel lobby #7
Camera repair shop
Hotel lobby #8
Cafeteria
Clothing store #7
Fabric store #2
Clothing store #8
Department store #2
Clothing store #9
Jewlery store #2
Bank #3
Hotel lobby #9
Book store
Stereo store
Airline ticket office #3
Department store #3a
Department store #3b
Department store #4a
Department store #4b
Department store #3c
Department store #3d
Restaurant lobby #5b
Art gallery #3
Hotel lobby #10
5.0
2.0
4.7
1.7
1.0
1.0
2.0
0.0
0.0
6.0
8.6
6.0
2.0
2.7
3.2
5.4
3.7
2.7
1.0
3.0
1.0
1.0
1.0
1.0
0.7
1.0
0.3
2.0
1.3
3.0
1.0
5.3
5.0
3.3
2.7
4.0
3.3
2.6
2.3
1.7
2.7
Mean Position
outdoor of
concentration entrance
(ppm) door
4.0
5.7
6.3
2.3
0.0
2.0
4.0
0.0
0.0
4.3
3.7
6.3
3.7
3.0
2.0
9.0
2.7
1.3
1.7
2.0
0.3
1.0
1.0
1.0
0.3
1.0
0.0
1.7
2.3
3.3
2.7
9.3
3.6
5.3
2.7
6.7
2.6
2.0
1.7
2.3
3.3
Open
Closed
Open
Closed
Not recorded
Closed
Closed
Not recorded
Closed
Closed
Open
Open
Open
Open
Open
Open
Closed
Closed
Closed
Closed
Closed
Not recorded
Open
Not recorded
Not recorded
Closed
Open
Closed
Open
Open
Open
Closed
Closed
Closed
Closed
Closed
Open
Closed
Open
Closed
Closed
(continued)
52
-------
TABLE 8 (continued)
Mean Mean Position
indoor outdoor of
Commerical concentration concentration entrance
Case setting (ppm) (ppm) door
80
81
82
83
84
85
86
87
88
Frame shop
Coffee shop
Clothing store #10
Real estate office
Office building lobby #7
Office building lobby #8
Furniture store #2
Fitness center
Clothing store #11
1.3
1.3
2.3
2.0
2.7
9.0
2.7
2.3
1.3
1.3
2.3
2.3
2.0
1.3
4.3
2.7
3.3
1.0
Closed
Open
Open
Closed
Open
Open
Closed
Open
Open
Notes:
a. The door was closed while outdoor samples were taken, but then
opened while the second indoor sample was taken.
b. This setting is one flight below street level.
53
-------
For this survey, the following variables were defined:
x(c)i = mean CO concentration of indoor setting i with entrance door
closed;
y(c)- = mean CO concentration of outdoor setting i with entrance
door closed;
k(o)- = mean CO concentration of indoor setting j with entrance door
open; and
y(o)- = mean CO concentration of outdoor setting j with entrance
door open.
We expected that for a given setting the indoor and outdoor CO
concentrations would be correlated, and that the indoor concentrations would
be appreciably less than the outdoor concentrations when the entrance door
was closed. The Pearson's product moment correlation coefficient (r = 0.60)
for all settings suggested that correlation did indeed exist. This
coefficient had a z-statistic value of 5.596 and a probability of p = 1.1 x
10~8 of rejecting a true null hypothesis that the correlation is zero.
The difference in mean concentrations between indoor and outdoor
concentrations for two different door positions was tested next. The null
hypotheses were that:
HQ (1): X(0) = y-(o)
H0 (2): 5?(c) = y(c)
In effect, these hypotheses state that indoor and outdoor concentrations are
equal and the position of the door has no effect. The alternative
hypotheses were stated as follows:
H! (1): x(o)
Hj_ (2): 5(c) * y(c)
In effect, these hypotheses state that indoor and outdoor concentrations are
not equal and that the door's position does have an effect, although the
direction of the effect is not specified in advance. The two-tailed test of
these hypotheses was performed using the t-statistic for dependent
samples. (56) Using a two-tailed test imposed a more stringent criterion of
statistical significance ( a /2 = 0.025) for rejecting a true null
hypothesis. This precaution was taken to offset the fact that indoor
measurements had been taken shortly after the outdoor measurements. This
delay was not considered too important, because, as shown in Chapter 4,
indoor concentrations do not vary a great deal over short periods of time.
For the 44 settings with an opened entrance door:
54
-------
x(o) = 2.91 ppm y(o) = 3.18 ppm
sx(o) = 2'04 PPm sy(o) = 2'22 PPm
For these settings, t = -0.872 with df = 43 and p = 0.338. For the 37
settings with a closed entrance door:
x(c) = 2.58 ppm y(c) = 3.67 ppm
Sx(c) = 1'62 PPm sy(c) = 3-09 PPm
For these settings, t = -2.846 with df = 36 and p = 0.007- Hence, we
concluded that, when indoor/outdoor measurements are taken within a short
time span of each other, indoor CO concentrations were statistically less
than outdoor concentrations only when the entrance door was closed. This
conclusion does not counter the findings of other studies (Chapter 2) which
suggest that over a long term, indoor levels tend to lag behind outdoor
levels and may even be higher when outdoor levels are decreasing.
B. HORIZONTAL VARIATION FOR INDOOR SETTINGS
The results of the previous chapter indicate that concentrations in
commercial settings are relatively homogeneous over short time periods.
Thus, the CO exposure for a person moving within a given setting for short
periods of time appears to be fairly uniform. Yet, what happens as a person
moves from one indoor setting to another? Are the CO concentrations of
adjacent and nearby indoor settings similar to each other? If the
concentrations are substantially different, can that difference be explained
by the purpose of the setting, its location within a grid of city blocks, or
its orientation to the street?
The previous chapter also has shown that exposure to CO varies more
from one date to another than from one setting to another, assuming the same
geographic location applies for all dates and settings. Hence, to discover
variation in CO between settings requires focusing on a particular
geographic location and a specific date. Furthermore, the settings should
have been selected randomly.
1. Purpose of the Setting
The question of purpose was studied by comparing samples collected on
13 June 1980 in the Union Square district of San Francisco (Table 9).
(Figure D-3 in Appendix D shows the cumulative frequency distributions.)
The hypotheses to be tested were that:
HQ: Xl = x2 = ... = x8
H^: at least one mean does not equal the others.
Here X£ equals the mean CO concentration of indoor commercial settings
with purpose i. In a one-way analysis of variance, F = 1.86. With 7
55
-------
TABLE 9. STATISTICAL SUMMARY OF CO CONCENTRATIONS MEASURED ON
13 JUNE 1980 FOR INDOOR COMMERCIAL SETTINGS
BROKEN DOWN BY PURPOSE AND LOCATED IN THE
UNION SQUARE DISTRICT OF SAN FRANCISCO, CALIFORNIA
Commercial setting3
Clothing, fabric, & shoe store
Department stores
Home furnishings
Hotel lobbies
Office building lobbies
Restaurants
d
Service centers
Q
Miscellaneous
Mean
CO
cone.
(ppm)
1.6
2.8
1.9
3.2
3.6
3.3
2.4
2.8
Std. dev.
of CO
cone.
(ppm)
1.1
1.4
1.3
1.7
2.7
2.3
1.5
1.8
Number
of
settings
17
8
8
11
8
11
9
12
Notes:
a. Two theaters were omitted.
b. Includes appliances, antiques, furniture, and paintings.
c. Includes burger stands, cafeterias, coffee shops, delicatessens,
and restaurants.
d. Includes airline ticket agencies, banks, currency exchanges, real
estate offices, and camera repair shops.
e. Includes book stores, china and crystal shops, drug stores, gift
shops, jewelry stores, novelty stores, and stamp stores.
56
-------
degrees of freedom for the variance between groups and 76 degrees of freedom
for variance within a group, p = 0.088. Thus, the null hypothesis was
accepted as p > a = 0.01.
The results indicate that exposure to CO inside commercial settings
varies as much from, for example, one retail store to another as it does
from a retail store to a bank, hotel, office building, or restaurant. A
notable exception is the parking garage. The one garage monitored in the
Union Square district was not included because the analysis of variance
requires that cell sizes should be approximately equal for ease of manual
computation. Nonetheless, the range of mean CO concentrations in this
garage vary between 30.8 ppm and 60.5 ppm, depending upon the time of
measurement and the monitoring instrument. Clearly, these concentrations
are several orders of magnitude greater than all other types of indoor
commercial settings. Enclosed parking garages stand out as a clear and
distinct class with respect to CO concentrations.
2. Location of Setting Within a City Block Grid
The locations of indoor commercial settings within a city block grid
could be related to variation in traffic volumes and micrometeorological
factors. The question of location within a block was studied by comparing
samples collected on 13 June 1980 in the Union Square district of San
Francisco (Table 10). (Figure D-4 in Appendix D shows the cumulative
frequency distributions.) The hypotheses to be tested were that:
H0: *! = 52 = . . . = xg
H^: at least one mean does not equal the others.
Here 1?^ equals the mean CO concentration of indoor commercial settings
located on city block i. In a one-way analysis of variance, F = 3.48. With
7 degrees of freedom for the variance between groups and 80 degrees of
freedom for variance within a group, p = 0.0026. Thus, the null hypothesis
was rejected as p < a= 0.01.
The results suggest that for geographic locations with heavy traffic
patterns, such as the Union Square district of San Francisco, CO exposure
inside commercial settings varies more from one city block to another than
from one setting to another on the same block. Differences in traffic
volume and in micrometeorology for different blocks probably account for
these differences in exposure, even though they are relatively small. This
conclusion does not imply, however, that adjacent indoor settings in
commercial areas have identical concentrations.
3. Settings on Opposite Sides of the Same Street
During the last survey of University Avenue in Palo Alto (made on
11 July 1980), we collected data simultaneously on opposite sides of the
street from 2:20 p.m. to 3:30 p.m. The purpose was to determine if
small-scale wind currents make CO concentrations different on opposite sides
of the street. For this date and time period, the wind was coming out of
57
-------
TABLE 10. STATISTICAL SUMMARY OF CO CONCENTRATIONS MEASURED ON
13 JUNE 1980 FOR INDOOR COMMERCIAL SETTINGS
BROKEN DOWN BY CITY BLOCKS LOCATED IN THE
UNION SQUARE DISTRICT OF SAN FRANCISCO, CALIFORNIA
City
block
number3
1
2
4
8
11
12
15
19
Mean
CO
cone.
(pptn)
1.9
3.0
4.0
3.3
3.1
1.8
2.8
1.1
Std. dev.
of CO
cone.
(ppm)
1.8
2.2
2.2
1.1
1.9
0.9
2.4
0.8
Number
of
settings
9
9
11
10
13
16
9
11
Note:
Refer to Figure B-l in Appendix B for the location of each city
block.
58
-------
the west at a speed of 10 to 12 mph. Hence, the wind was blowing strongly
at roughly a 60 degree angle to the axis of the street. This condition
enabled a test of the windward-leeward hypothesis. This hypothesis suggests
that the windward or east side of University Avenue would have lower
concentrations than the leeward or west side of the street.
Because we knew that CO concentrations vary for indoor settings (with
entrance door open or closed) and outdoors, we had to control these factors
to isolate the effect of small-scale winds. Thus, the analysis of
University Avenue was segmented into indoor and outdoor settings. (Outdoor
settings are discussed in Section C of this chapter.) For indoor settings,
equal percentages of settings with doors opened to the street were visited
on both sides of the street without violating the random selection process.
The following variables were defined for indoor settings:
x(i)£ = mean CO concentration of indoor setting i
on the west side of University Avenue; and
y(i); = mean CO concentration of indoor setting j
on the east side of University Avenue.
The equivalence in mean CO concentrations for indoor settings on opposite
sides of University Avenue was tested next. The hypotheses were as follows:
H0: x(i) = y(i)
HI: x(i) > y(i)
The alternate hypothesis made a prejudgment regarding which side of the
street would have a higher average CO concentration for indoor settings.
Consequently; the one-tailed test of these hypotheses was performed using
the t-statistic for independent samples.(56)
Of the 17 indoor settings sampled on the west side of the street, 11
(64.7%) had the entrance door opened to the street, as did 13 (61.9%) of the
21 indoor settings sampled on the east side. A statistical summary of CO
concentrations for these indoor settings shows that:
x(i) = 1.79 ppm y(i) = 0.28 ppm
sx(i) = 1'07 PPm Sy(i) = °'58 Ppm
nx(i) = 17 ny(i) = 21
Before comparing these settings, the assumption that the variances are
equal must be tested. In an analysis of variance, F = 3.357. With 16
degrees of freedom for the larger variance and 20 for the smaller
variance, p = 5.91 x 10" . Since p < a = 0.01, a t-test could be
performed using the formula for unequal variances. Then t = 5.23. With
approximately 24 degrees of freedom, p = 2.3 x 10~^. Since p < a = 0.01,
the null hypothesis was rejected. Hence, statistical analysis indicates
59
-------
that the CO concentrations were dissimilar for indoor settings located on
opposite sides of the same street. In effect, there appeared to be windward
and leeward sides of this street.
C. HORIZONTAL VARIATION FOR OUTDOOR SETTINGS
The relevant question and requirements for analyzing outdoor settings
are similar to those just discussed for indoor settings. The samples for
Sections 1 and 2 were collected near Union Square in San Francisco on 13 June
1980, while those for Section 3 were collected along University Avenue in
Palo Alto on 11 July 1980.
1. Purpose of the Setting
The purpose of an outdoor setting had been classified as either: (1)
an intersection; (2) a midblock location between two intersections; or (3)
an arcade, park, plaza or parking lot. Such descriptors were readily
discernible to the public and could be related to the nearness and/or
strength of CO sources. Arcades, parks, plazas, and parking lots were not
considered in the statistical analysis which follows, however, because a
sufficient number could not be found.
Because the Union Square district data had been collected in randomly
selected city blocks, the following variables were defined:
x^ = mean CO concentration at the corners of city block i; and
y^ = mean CO concentration of all monitored midblock locations along
the faces of city block i.
CO was measured for the corners and faces of each city block surveyed
(Table 11). For a given city block, we expected that the CO concentration
measured at the corner of the block would be greater than that at midblock
locations along the block face. At an intersection, vehicles are stopped in
one direction. When they are stopped, the vehicle queue may extend along
the block face. Hence, the individual waiting to cross a street at an
intersection may be exposed to concentrations greater than those found along
the sidewalk. Consequently, the hypotheses tested were as follows:
V x = y
HI: x > 7
The one-tailed test of these hypotheses was performed using the
t-statistic for dependent samples.(56) The t-statistic for dependent
samples was appropriate if the concentrations of the corner sites and of the
block faces were correlated. Pearson's coefficient (r = 0.59) for all city
blocks surveyed indicated that correlation did exist in the expected
direction. This coefficient had a z-statistic value of 1.56 and a
probability p = 0.059 of rejecting a true null hypothesis that the
correlation is zero.
60
-------
TABLE 11. STATISTICAL SUMMARY OF CO CONCENTRATIONS MEASURED ON
13 JUNE 1980 AT CORNERS AND FACES OF CITY BLOCKS LOCATED
IN THE UNION SQUARE DISTRICT OF SAN FRANCISCO, CALIFORNIA
Mean CO
concentration
(ppm)
City block
number3
1
2
4
7
8
11
12
15
19
Corners
4.1
3.6
4.2
5.8
4.7
4.2
1.9
2.5
4.0
Faces
2.7
4.3
3.7
4.8
3.9
3.5
3.0
2.0
1.9
Note:
Refer to Figure B-l in Appendix B for the location of each city
block.
61
-------
For the 9 city blocks:
5 = 3.89 ppm J = 3.31 ppm
Sx = 1.15 ppm Sy = 0.99 ppm
For these blocks, t = 1.77. With df = 8, p = 0.115. Since p was greater
than the significance value a = 0.05, we concluded that for this small
sample no difference in CO concentrations existed between corners and faces
of city blocks in commercial areas. This conclusion echoes the conclusion
made for indoor commercial settings classified by purpose. However, further
investigation, perhaps with a larger sample, is recommended, since it
counters the expectation that intersections should have higher CO levels due
to more idling and acceleration.
2. Physical Coordinates of a Setting
Outdoor settings with different physical coordinates were compared next
using nine corners of intersections located in the Union Square district of
San Francisco and sampled on 13 June 1980 (Table 12). The results raise the
question of whether variation in CO concentrations is greater between
intersections or between different corners of the same intersection. Here
the issue is whether macro- or micro-scale variation in CO concentrations is
greater. The fact that observations were not taken simultaneously at all
intersections will be ignored for the present and considered later.
A one-way analysis of variance is appropriate for resolving this
question if certain assumptions can be met. First, this test assumes that
the mean CO concentrations of intersections are normally distributed. Since
fewer than 30 observations were taken at each intersection, this assumption
could not be made categorically. Rather than scratching this analysis due
to this assumption, we decided to pursue the analysis with this statistical
test and impose a more rigorous standard, 0.005, for the level of
significance.
A second assumption is that the standard deviations are approximately
equal. The results revealed clearly that such was not the case for all
intersections. However, the standard deviations for four intersections were
approximately equal: Mason and O'Farrell, Powell and Geary, Stockton and
Sutter, and Stockton and Post. Hence, only these four intersections were
considered in subsequent analysis.
The third assumption is that the number of 1-minute samples taken at
one intersection is approximately equal to the number of samples taken at
each of the other intersections. This assumption was met because 12
samplesthree at each of four cornershad been taken at each intersection.
The following variables were then defined:
xi = mean CO concentration for Mason and O'Farrell;
X£ = mean CO concentration for Powell and Geary;
62
-------
TABLE 12. STATISTICAL SUMMARY OF CO CONCENTRATIONS MEASURED ON
13 JUNE 1980 FOR CORNERS OF INTERSECTIONS LOCATED IN THE
UNION SQUARE DISTRICT OF SAN FRANCISCO, CALIFORNIA
Intersection
& corner
Mason & Sutter
southeast corner
southwest corner
northwest corner
northeast corner
all corners
Mason & Post
southeast corner
southwest corner
northwest corner
northeast corner
all corners
Mason & O'Farrell
southeast corner
southwest corner
northwest corner
northeast corner
all corners
Powell & Post
southeast corner
southwest corner
northwest corner
northeast corner
all corners
Powell & Geary
southeast corner
southwest corner
northwest corner
norhteast corner
all corners
Mean
CO
cone.
(ppm)
10.3
2.3
0.3
0.0
3.3
0.8
0.3
1.7
0.0
0.7
7.3
2.7
3.7
2.0
3.9
5.0
8.7
2.3
2.8
4.7
8.3
5.3
6.7
8.3
7.2
Std. dev.
of CO
cone .
(ppm)
9.1
0.6
0.6
0.0
5.8
0.3
0.6
1.2
0.0
0.9
1.5
0.6
1.5
0.0
2.4
3.0
8.1
0.6
2.8
4.7
2.1
3.5
2.1
1.2
2.4
Number Time
of of
observa- observa-
tions tions
1:46 p.m-l:57 p.m.
3
3
3
3
12
11:37 a.m. -11:48 a.m.
3
3
3
3
12
12:28 p.m. -12:39 p.m.
3
3
3
3
12
10:44 a.m. -10:55 a.m.
3
3
3
3
12
10:47 a.m. -10:58 a.m.
3
3
3
3
12
(continued)
63
-------
TABLE 12 (continued)
Intersections
& corner
Stockton & Sutter
southeast corner
southwest corner
northwest corner
northeast corner
all corners
Stockton & Post
southeast corner
southwest corner
northwest corner
northeast corner
all corners
Grant & Bush
southeast corner
southwest corner
northwest corner
northeast corner
all corners
Fifth & Market
southeast corner
southwest corner
northwest corner
northeast corner
all corners
Mean
CO
cone.
(ppm)
5.0
4.7
2.3
5.7
4.4
0.3
3.7
2.7
3.3
2.5
2.0
16.3
7.0
3.7
7.3
1.3
0.0
1.7
2.7
1.4
Std. dev.
of CO
cone.
(ppm)
3.0
2.5
0.6
2.1
2.3
0.6
2.1
1.5
2.3
2.0
1.0
14.0
1.0
0.6
8.3
1.5
0.0
0.6
1.2
1.3
Number Time
of of
observa- observa-
tions tions
1:06 p.m. -1:17 p.m.
3
3
3
3
12
10:45 a.m. -10:56 a.m.
3
3
3
3
12
1:19 p.m. -1:30 p.m.
3
3
3
3
12
1:06 p.m. -1:17 p.m.
3
3
3
3
12
64
-------
x.j - mean CO concentration for Stockton and Sutter;
5?4 = mean CO concentration for Stockton and Post.
The hypotheses to be tested were as follows:
H^: at least one mean does not equal the others.
In a one-way analysis of variance, F = 8.97. With 3 degrees of freedom for
the variance between groups and 44 degrees of freedom for the
variance within a group, p = 9.45 x 10"^. Thus, the null hypothesis was
rejected as p< a= 0.005, the assigned level of significance for this case.
Thus, exposure to CO may vary more from one intersection to another
than from one corner to another at the same intersection, assuming the same
geographic location and date of measurement apply for all intersections.
This conclusion still leaves open the hypothesis that variation in CO
concentrations between corners of an intersection may or may not be greater
than the variation in concentrations at a given corner. Sufficient data had
not been collected at each corner to offer a valid test of this latter
hypothesis. The significance of this hypothesis may be inconsequential,
however, since the corners of intersections receive only transitory visits
by most pedestrians.
Of the four intersections, only two had been monitored at roughly the
same time. Hence, one could argue that variation in CO concentrations for
these intersections was due to time as well as spatial factors. To isolate
the effect of spatially related factors required comparisons of
intersections monitored at approximatedly the same time. Three
intersections were monitored simultaneously during a brief period in the
morning and two in the afternoon.
The following variables were defined for the three intersections
sampled simultaneously at roughly 10:44 a.m. to 10:58 a.m.:
*2 = mean CO concentration for Powell and Geary;
£3 = mean CO concentration for Stockton and Post;
x^ = mean CO concentration for Powell and Post.
The relationships among the means were as follows:
x2 = 7.22 ppm > £3 = 4.7 ppm > x^ = 2.5 ppm
The analysis of variance had shown that at least one of the four
intersections discussed earlier had a mean CO concentration signficantly
different from the others. In fact, Powell and Geary (x2 =7.2 ppm) had
the highest average concentration and Powell and Post (S/ = 2.5 ppm) the
65
-------
lowest. Since both intersections had been included in the analysis of
variance, one could conclude that they had significantly different mean CO
concentrations .
Consequently, the intersection of Stockton and Post (33 = 2.5 ppm)
and the intersection of Powell and Post ($4 = 4.7 ppm) were compared
because their means were closer to each other than either was to the mean
concentration for Powell and Geary (x2 = 7.2 ppm). Before comparing these
two intersections, the assumption that the variances are equal must be
tested. An analysis of variance showed that F = 5.52. With 11 degrees of
freedom each for the larger and smaller variances, p = 4.29 x 10~J. Since
p < a = 0.005, a t-test was performed using the formula for unequal
variances.
The hypotheses were stated as follows:
V X4 = X5
Analysis revealed that t = 1.02. With approximately 16 degrees of freedom,
p = 0.32. Since p > a/2 = 0.0025, the null hypothesis was accepted. Hence,
statistical analysis had indicated that one could not reject the hypothesis
that the mean CO concentrations for these two intersections were equal.
Next, the following definitions were made for two intersections
monitored simultaneously from 1:06 p.m. to 1:17 p.m.:
X3 = mean CO concentration for Stockton and Sutter;
Xg = mean CO concentration for 5th and Market.
Again one must first test the assumption of equal variances. With 11
degrees of freedom each for the larger and smaller variances, an analysis of
variance revealed that F = 3.13 and p = 0.036. Hence, the variances were
assumed equal since p > a = 0.005. Thus, a t-test was performed using the
formula for pooled variances. The hypotheses were stated as follows:
V
x3
With approximately 22 degrees of freedom, t = 3.93 and p = 7.2 x 10~\
Since p
-------
x(o)£ = mean CO concentration of outdoor setting i located on the
west side of University Avenue; and
y(o)j = mean CO concentration of outdoor setting j located on the
east side of University Avenue.
Again the hypotheses were stated as follows:
H0: x(o) = 5Ko)
HI: 5(o) > 5Ko)
Since we had prejudged which side of the street had a higher average CO
concentration, a one-tailed test was performed using the t-statistic for
independent samples.
A statistical summary of CO concentrations for these outdoor settings
indicated the following results:
x(o) = 2.07 ppm y(o) = 0.77 ppm
sx(o) = !'63 PPm sy(o) = 1'30 PPm
nx(o) " U ny(o) = 13
Again, the assumption that the variances are equal must first be tested. An
analysis of variance shows that F = 1.575. With 10 degrees of freedom for
the larger variance and 12 for the smaller variance, p = 0.23. Thus, a
t-test was performed using the formula for pooled variances, since
p> a= 0.01. Then t = 2.18. With 22 degrees of freedom, p = 0.04. Since
p > a = 0.01, the null hypothesis was accepted. Hence, statistical
analysis had indicated that one could not reject the hypothesis that CO
concentrations were similar for outdoor settings located on opposite sides
of the same street. Nevertheless, because of the small sample size, this
result warrants more study.
This result reinforces the general conclusion that has been building in
previous sections: Without more observations outdoor settings are difficult
to distinguish with respect to variation in CO concentrations. Furthermore,
it is difficult to distinguish between intersections and midblock locations
and between opposite sides of the same street for outdoor settings only.
D. VERTICAL VARIATION FOR INDOOR SETTINGS
The indoor CO concentrations for a 15-story general office building
located in downtown Palo Alto were discovered to be unexpectedly high on the
first visit to this building on 24 January 1980. That discovery stimulated
repeated visits to this structure and to other tall buildings located in
different geographic locations, eight of which were surveyed in sufficient
detail to permit an analysis of the vertical variation of CO. Because the
office building in Palo Alto was monitored more intensively, it has been
singled out for separate discussion in the next chapter.
67
-------
The monitoring instruments were positioned at various points of the
lobby of each floor near the elevators. With some exceptions, CO
concentrations are fairly uniform over the height of each building (Figure
8). The exceptions generally occur at garage floor levels where CO
concentrations are substantially higher than those recorded on other
floors. Another exception occurred at the 16th floor of the general office
building located in Westwood Village in Los Angeles. This concentration of
7 ppm of CO (versus 4-5 at lower floors) probably can be attributed to a
cocktail lounge and restaurant at this level.
68
-------
21
19-
17
15 -
13-
o
o
7 -
5-
1 -
G-
X 8-story city hall office building in Palo Alto visited 31 January 1980;
it 13-story general office building in the financial district of San Francisco
visited 6 March 1980;
6-story general office building in Palo Alto visited 7 March 1980;
O 12-story medical office building in Westwood Village of Los Angeles
visited 27 March 1980,
A 16-story general office building in Westwood Village of Los Angeles
visited 27 March 1980,
O 6-story general office building in the Union Square district of San Francisco
visited 1 1 April 1980;
12-story hotel in the Union Square district of San Francisco
visited 1 1 April 1980;
A 21-story hotel in the Union Square district of San Francisco
visited 1 1 April 1980.
8
24
28
32
12 16 20
CO Concentration (ppm)
Figure 8. Average concentration of CO for selected levels of eight high-rise buildings.
69
-------
CHAPTER 6
A "HOT" OFFICE BUILDING
On the afternoon of 24 January 1980, we surveyed University Avenue in
Palo Alto for the first time, focusing on the retail stores which
comprise the central business district. The average CO concentration at
a bank at one end of this district was 5.9 ppm, two to three times higher
than concentrations being recorded that day for other settings. Next, we
sampled a 15-story building adjacent though not connected to the bank.
The average concentration for the main lobby was 13.9 ppm, which prompted
an investigation of the lobbies on floors 3, 6, 11, and 15, and a law
office on the llth floor. On these floors the average concentrations
were high, ranging from 12.0 ppm for the 15th floor lobby to 15.0 ppm for
the 6th floor lobby. The CO concentrations were fairly uniform over the
height of this building, as was the case with eight other high-rise
buildings visited later.
How long these elevated concentrations lasted was not known yet.
Consequently, we surveyed the building on seven different dates between
24 January and 11 July 1980, twice on two dates. (The time variation of
CO concentrations, described in Chapter 4, included data for the main
lobby only of this office building, as previous data had indicated that
the concentrations recorded for the main lobby were fairly typical of the
upper floors.)
On six dates, the same four floor lobbies mentioned earlier were
monitered on each visit. After the first survey, the garage and two
lower levels near the garage were also monitored on each visit. These
seven levels constituted the regular procedure for visits, with one
exception. On 4 April 1980, the surveyor first went up the building by
elevator, taking three observations from each of two instruments in the
lobbies of each floor. From the top floor, the surveyor came down by the
stairwell, taking two observations from each of two instruments at each
floor.
No samples were taken inside particular offices for two reasons.
During the first survey, the average CO concentration measured in a law
office on the llth floor was 14.5 ppm, not much different from the
average concentration of 14.0 ppm recorded for the adjacent lobby.
Second, unsympathetic occupants of the office usually requested lengthy
explanations. Consequently, only lobbies and other public areas in the
building were monitored.
70
-------
In effect, the surveys conducted in this building constituted a
small-scale study within the overall study. Also, the data were
collected such that both the time and spatial variation of CO
concentrations within the building could be analyzed.
A. TIME VARIATION
Detection of time variation in CO concentrations in this building
requires controlling for the effects of spatial variation over its
height. Consequently, only locations visited repeatedly the lobbies on
floors 1, 3, 6, 11 and 15 are considered here. While the parking garage
was visited on nearly every occasion, it has been excluded here since its
mean CO concentration is substantially greater than the rest of the
building (Table 13). Although the garage is attached to the building, it
must be treated as a separate microenvironment .
Two PEM's were used on every visit. Observations taken
simultaneously at 1-minute intervals from the two instruments were
averaged. For some dates, three pairs of observations per floor were
taken, while on other dates only one pair per floor was taken (Table 14).
On 31 January and 7 February 1980, the five floors were visited
twice. The time between visits was approximately 1 hour and 40 minutes
for 31 January, and 1 hour and 55 minutes for 7 February. The mean CO
concentration for the second visit on both dates was greater than on the
first. Since a third of the above ground-level of the building had been
sampled, a rigorous statistical comparison of the mean CO concentration
for the first and second visits was judged appropriate, even though there
were only five sample pairs for each date and visit.
The following variables were defined:
x(l) = mean CO concentration for the first visit on 31 January;
x(2) = mean CO concentration for the second visit on 31 January;
y(l) = mean CO concentration for the first visit on 7 February; and
y(2) = mean CO concentration for the second visit on 7 February.
The hypotheses for the visits on 31 January were as follows:
HQ: x(l) = 3c(2)
HI: x-(l) t 5t(2)
The alternate hypothesis was so stated because there was no reason to
believe that the CO concentrations observed during the second visit should
be consistently different than those on the first visit. Furthermore, there
was no reason to believe that the CO concentrations on these floors were
related from visit to visit. Consequently; the t-test for independent
samples was selected as the test statistic. (56)
71
-------
TABLE 13. AVERAGE CO CONCENTRATIONS BY DATE AND LEVEL
FOR A 15-STORY GENERAL PURPOSE OFFICE BUILDING IN
PALO ALTO, CALIFORNIA (1980)
Mean CO concentration (ppm)
Level 24 Jan
Garage
Lobby B
Lobby A
Lobby 1 13.9
Lobby 2
Lobby 3 13.8
Lobby 4
Lobby 5
Lobby 6 15.0
Lobby 7
Lobby 8
Lobby 9
Lobby 10
Lobby 11 14.0
Lobby 12
Lobby 13
Lobby 14
Lobby 15 12.0
31 Jan
27. 9a
31.6
8.1
5.3
7.0
5.5
^n
5.8
7.5
5.5
6.5
3.8
6.3
7 Feb
23.0
7779"
24.0
2.0
178-
3.0
3.5
5.0
5.5
5.5
5.0
4.5
3.3
7 March 4 April
51.0 40.6
35.0
12.5
9.9 10.2
11.5
10.3 13.5
11.3
11.5
9,8 12.4
11.8
12.3
12.3
18.5
9.0 12.0
2.1
2.0
2.2
4.0 4.0
9 May
19.2
21.0
8.8
4.8
6.7
7.0
8.5
8.4
4.5
11 July
51.3
7.0
8.5
9.2
9.2
9.0
5.3
Note:
a.
Entry above line represents first visit and entry below line
represents second visit.
72
-------
TABLE 14. STATISTICAL SUMMARY OF CO CONCENTRATIONS
FOR VISITS TO FIVE SELECTED FLOOR LOBBIES OF A
15-STORY GENERAL PURPOSE OFFICE BUILDING IN PALO ALTO, CALIFORNIA,
BROKEN DOWN BY DATE AND HOUR OF VISIT
Date
24
31
31
7
7
7
4
9
11
Jan
Jan
Jan
Feb
Feb
1980
1980
1980
1980
1980
March 1980
April 1980
May
July
1980
1980
3:00
9:21
11:10
11:14
1:07
4:24
1:42
1:20
3:41
Hour
of
visit
- 3:26
- 9:43
- 11:15
- 11:21
- 1:16
- 4:45
- 3:00
- 1:51
- 4:02
Std. dev. Number
Mean of CO of
CO cone. cone. sample
(ppm) (ppm) pairs
p.m. 13
a.m. 5
a.m. 6
a.m. 4
p.m. 4
p.m. 9
p.m. 11
p.m. 6
p.m. 7
.5 1.5
.0 1.0
.9 0.5
.1 1.3
.3 1.0
.2 3.0
.4 3.0
.3 1.6
.8 1.7
15a
5b
5b
5b
5b
15a
15a
15a
15a
Notes
a.
b.
Three
One pt
pairs of 1-minute observations taken on each
lir of 1-minute observations taken on each of
of five floors.
five floors .
73
-------
To compare these two visits requires first testing the assumption that
the variances are equal. In an analysis of variance, F = 4.0. With 4
degrees of freedom each for the larger and smaller variances, p = 0.10.
Since p> a= 0.01, a t-test was performed using the formula for pooled
variances. Then t = 3.8. With 8 degrees of freedom p = 0.0052, which is
not significant since a/2 = 0.005. Hence, the null hypothesis was
accepted. Statistical analysis indicated that the mean CO concentration for
the second visit on 31 January was equal to the mean concentration of the
first visit.
The hypotheses for the visits on 7 February were stated as follows:
HO: yd) = 7(2)
f y(2)
In an analysis of variance, performed first to test the assumption of equal
variances, F = 1.69. With 4 degrees of freedom each for the larger and
smaller variances, p = 0.30, which is greater than a = 0.01. Consequently,
in a t-test, performed on the means using the formula for pooled variances,
t =0.27. With 8 degrees of freedom, p = 0.79, which is greater than a/2 =
0.005. Hence, the null hypothesis was accepted. For 7 February;
statistical analysis indicates that the mean CO concentration for the first
visit was equal to the mean concentration for the second visit.
Analysis in Chapter 3 had indicated that CO concentrations were
relatively homogeneous over time for a given date, geographic location, and
commercial setting. The results just presented support that analysis. For
both 31 January and 7 February, the CO concentrations were homogeneous over
a period of nearly 2 hours.
Variation in CO concentrations among the seven dates that the office
building had been visited was considered next. That variation did indeed
exist: The mean CO concentration ranged from a low 4.1 ppm recorded for the
first visit on 7 February to a high of 13.5 ppm on 24 January. A one-way
analysis of variance was employed to test if mean CO concentrations on at
least one date were significantly different from all other dates.
The one-way analysis of variance imposes three major assumptions.
First, it assumes that the mean CO concentrations for the building are
normally distributed. This assumption usually can be made if the number of
sample pairs exceeds 30. Although this many sample pairs had not been
collected on any date, the normality was still assumed here because the
standard deviation was small relative to the mean in all cases. As an added
precaution, a more stringent level of significance was adopted, a= 0.01.
A second assumption is that the standard deviations are approximately
equal. To meet this assumption, only three dates (24 January, 9 May, and 11
July) were retained for further analysis: Two had a standard deviation of
1.6 ± 0.1 ppm, and the third had a standard deviation of exactly 1.6 ppm.
74
-------
The third assumption is that the number of sample pairs taken on one
date should be approximately equal to the number of sample pairs taken on
each of the other dates. This assumption was met for the three remaining
dates, because 15 sample pairs had been collected on each date.
The following variables were defined:
xj = mean CO concentration for samples collected on 24 January;
x"2 = mean CO concentration for samples collected on 9 May; and
£3 = mean CO concentration for samples collected on 11 July.
The hypotheses to be tested were as follows:
H0: xx = x2 = *3
H^: at least one mean does not equal the others.
In a one-way analysis of variance, F = 84.22. With 2 degrees of freedom for
the variance between groups and 42 degrees of freedom for the variance
within a group, p = 4 x 10 . Thus, the null hypothesis was rejected as
p< a= 0.01, the assigned level of significance.
The mean concentrations for 24 January and 9 May were sufficiently
dissimilar to force rejection of the null hypothesis. Also, the mean
concentration for 24 January may be sufficiently different from the mean for
11 July. This hypothesis was tested using a t-statistic for independent
samples and pooled variance. The value of t = 9.73. With 28 degrees of
freedom, p = 2.1 x 10~10, which is much less than a = 0.01. Hence, the
mean concentration of samples collected on 24 January was significantly
different from the mean concentration of samples taken on 11 July. In a
t-test for 9 May and 11 July, t = 2.49. With 28 degrees of freedom, p =
0.02. Since p> ct = 0.01, the null hypothesis of equal means was accepted.
Hence, the mean concentrations for these two dates were similar.
A partial explanation of this result may be found in the meteorological
conditions on each date. Wind speed and direction were obtained for the
nearest station to downtown Palo Alto for all seven dates and the hours
nearest to the time that the office building was visited (Table 15). The
correlation between mean CO concentration and wind speed was calculated
using Pearson's correlation coefficient. For the nine cases, r = 0.49. The
significance of this coefficient can be tested with the t-statistic since
the number of cases is less than 30. The value of t = 2.306. With 7
degrees of freedom, p = 0.054, which is not strictly significant if a =
0.05. Nevertheless, there may be grounds to conduct further research to
determine if greater wind speeds are associated with lower mean
concentrations of CO for this office building. Exactly why is not certain.
However, if true, this result would reinforce a similar result for both
indoor and outdoor commercial settings sampled in Palo Alto.
75
-------
TABLE 15. METEOROLOGICAL SUMMARY FOR THE REDWOOD CITY STATION OF THE
BAY AREA AIR QUALITY MANAGEMENT DISTRICT FOR DATES AND HOURS
CORRESPONDING TO SEVEN SURVEYS CONDUCTED IN THE
CENTRAL BUSINESS DISTRICT OF PALO ALTO, CALIFORNIA
Date
24 Jan 1980
31 Jan 1980
31 Jan 1980
31 Jan 1980
7 Feb 1980
7 Feb 1980
7 March 1980
4 April 1980
4 April 1980
9 May 1980
9 May 1980
11 July 1980
Hour
3 p.m.
9 a.m.
10 a.m.
11 a.m.
11 a.m.
1 p.m.
5 p.m.
2 p.m.
3 p.m.
1 p.m.
2 p.m.
4 p.m.
Wind
speed
(mph)
4
1
3
5
15
15
4
7
8
7
9
11
Wind
direction
N
S
NW
NW
N
N
NW
S
S
S
S
w
76
-------
B. SPATIAL VARIATION
The spatial variation of CO concentrations over the height of this
general office building was investigated extensively. On 4 April 1980, the
lobbies and stairwell were surveyed on every floor, including two floors
below street level, as well as the parking garage. During the sampling,
done by the investigator going up the elevator and down the stairwell, the
door between the lobby and the stairwell was closed at every level when
observations were taken, except for the garage level and the 10th floor.
The results show that the CO in the building apparently originated, not
surprisingly, at the garage level, where concentrations were between 30 and
35 ppm, depending upon the location (Figure 9). Concentrations appeared to
drop off going up the building, the rate depending upon the location. For
the lobbies, the decline was already dramatic for the level immediately
above the garage. From that point, concentrations stayed relatively uniform
(except for two high readings) until the llth floor. At the 12th floor, the
concentration in the lobby fell off again and stayed that way for the
remaining upper floors. For the stairwell, CO concentrations appeared to
decline with height more gradually until the 10th floor, where they fell off
considerably until the 12th floor. There they remained relatively uniform
through the 15th floor.
The difference between the upper and lower level concentrations may be
explained by the two ventilation systems serving the building and the
location of their respective vents. One system served the building from the
garage level to the llth floor, with a vent at street level. The other
system served the upper four floors, with a vent on the roof.
Thus, a drop at the 12th floor would be expected. An additional explanation
is that the ventilation system serving the upper portion of the building may
be more efficient in removing CO from the building. The difference between
the lobbies and the stairwell may be due to the fact that the door to the
stairwell at the garage level was always kept open. Furthermore, the
stairwell has a relatively smaller volume compared to the rest of the
building and may be less well ventilated. How much CO from the stairwell
migrated into the lobbies, which were protected by doors closed to the
stairwell, was not known.
The two aberrations in the relatively uniform concentrations for the
lobbies occurred at the 3rd and 10th floors. The bulge at the 3rd level may
be due to CO infiltration into the lobby from gas stoves in a restaurant at
that level. The bulge at the 10th floor can be explained by a freshly
painted door, which was left open between the lobby and the stairwell. Yet,
while the concentrations for the lobby and stairwell were similar at the
10th floor, they were not identical, most likely because the measurements
had not been taken simultaneously. The same statement applies to the lobby
and stairwell at the garage level.
C. DISCUSSION
The important findings about this building were the consistently high
levels of CO observed on most visits. Average CO concentrations were equal
77
-------
CO
15
14
13
12
11
10
9
1 6
3
CO
5
4
3
2
1
A
B
o Lobby
D Stairwell
Door Open to Stairwell
I I I I I I I I
Main Floor
I I I I
I I I I
10 12 14
16 18 20 22 24
CO Concentration (ppm)
26 28 30 32 34 36 38
40
Figure 9. Vertical profile of CO concentration measured in a 15-story building
in Palo Alto, California during the daytime hours on 4 April, 1982.
-------
to or greater than 9 ppm on most of the first 10 floors on four of seven
visits, with each visit on a different date. On the two dates during which
the building was visited twice in a 2-hour period, the concentrations of CO
were relatively stable. If these high CO levels persisted for as long as 8
hours on the four dates when they were recorded, it is very probable that
the 8-hour CO NAAQS was violated on most floors. Exposures could be reduced
by improving the building's ventilation system, making structural changes,
and closing doors from the building to the garage.
Under normal occupational exposures, employees are aware that high
exposures to air pollutants are possible. But it is unlikely that any of
these office workers were aware that their place of work involved a health
risk from high CO exposure.
Since the present study was completed, the owners of this building became
aware of the elevated CO concentrations present inside, and the managers
the building undertook steps to reduce interior CO concentrations by changing
the operation of the ventilation system.* It would be highly desirable to con-
duct a follow-up survey of CO concentrations inside this building to evaluate
the effectiveness of their efforts to mitigate this indoor air quality problem.
*Based on a telephone conversation with R. Brainard of Hare, Brewer, and Kelly,
Inc., 305 Lytton Ave., Palo Alto, CA 94301, January 1984.
79
-------
CHAPTER 7
GEOGRAPHIC VARIATION OF CO CONCENTRATIONS
"Geographic location" is a complex variable representing differing
traffic volumes and hence differing levels of CO emissions from motor
vehicles. It also represents various spatial configurations, mixing
volumes, and possibly micrometeorological patterns. Up to now in our
analyses of variation in CO concentrations, we looked at data while holding
geographic location constant.
In this chapter, geographic location is treated as the variable for
analysis. Ideally, in comparing the CO concentrations of commercial
settings among different locations, factors should be held constant that
have been shown to be associated with variation in CO concentrations.
Specifically, to hold constant the effects of weather, consideration should
be limited to just those geographic locations surveyed on the same date.
For this study, this limitation would be too restrictive, narrowing
consideration to just those geographic locations with low wind speeds and
similar atmospheric stability.
Previous analysis has shown significant differences in concentrations
between indoor and outdoor settings. Consequently, we will consider indoor
settings only, since the time variation in concentrations is less subject to
peaks. Intersections, for example, are affected by spikes in CO
concentrations, which increase the standard deviation. Furthermore,
previous analysis has shown that the position of the entrance door affects
the CO concentration indoors. However, considering just those settings with
door opened or door closed would severely reduce the number of suitable
settings at each location. Thus, both types of settings are now included in
our analysis, with particular attention to the percentage of settings of
each type for each location. Finally, previous analysis has shown that
certain types of indoor commercial settings, such as parking garages and
some office buildings attached to garages, have elevated concentrations. If
such settings are sampled at one particular location and not at another,
differences might result in mean concentrations between such locations.
Consequently, these types of settings were eliminated from consideration.
Two of the five geographic locations, Castro Street in Mountain View
and Westwood Village in Los Angeles, had been surveyed only once during
March 1980 about 2 weeks apart. Since it was desirable to analyze all five
locations, these two sites dictated the wind speeds which the other three
sites would have to match. The average wind speed for the Castro Street
site was 5 mph, with a standard deviation of 0.6 mph for the time period
80
-------
during which it was visited. For Westwood Village, the average wind speed
was 2.6 mph with a standard deviation of 1.1 mph. These wind speeds
eliminated the surveys conducted in Palo Alto on 7 February and 11 July 1980
because the average wind speeds were 13.8 mph and 10.8 mph, respectively.
The surveys conducted in some of the other geographic locations on
specific dates were eliminated from consideration for other reasons. The
survey of the Chinatown-Financial district of San Francisco on 6 March 1980
was eliminated because it was raining that day. Consequently, for this
location only the 11 April 1980 survey remained. The meteorological station
recorded an average wind speed of 6.3 mph, with a standard deviation of 0.5
mph for the time period of the survey. Since a survey had been conducted in
Union Square on the same date (11 April 1980), this site also was included
in the analysis. The Union Square site had an average wind speed of 5.0 mph
and a standard deviation of 1.6 mph.
The two surveys conducted in Palo Alto on 4 April and 9 May of 1980
were excluded since only one setting (a 15-story office building) had been
visited on each date. Another survey in Palo Alto on 7 March 1980 was
eliminated as only six indoor settings had been visited. Of the two surveys
remaining for Palo Alto, the survey conducted on 24 January 1980 was
selected because more indoor settings had been visited.
This screening process resulted in a set of five surveys, one for each
geographic location. Since the position of the entrance door could not be
controlled, the percentages of settings with entrance doors to streets in
the opened and closed position were then determined. This information
indicates no uniformity with respect to door position among the five surveys
(Table 16). Hence, the variable of "door position" may explain apparent
differences in CO concentrations for indoor commercial settings among the
five geographic locations. For example, one can expect that locations with
a greater percentage of opened doors probably also had a greater average CO
concentration for indoor settings.
The following variables were then defined for the five geographic
locations.
v^ = mean CO concentration of indoor commercial setting i sampled
on 24 January 1980 along University Avenue in Palo Alto;
W4 = mean CO concentration of indoor commercial setting j sampled
on 13 March 1980 along Castro Street in Mountain View;
Xfc = mean CO concentration of indoor commercial setting k sampled
on 27 March 1980 in Westwood Village in Los Angeles;
yi = mean CO concentration of indoor commercial setting 1 sampled
on 11 April 1980 in the Chinatown-Financial district of San
Francisco; and
zm = mean CO concentration of indoor commercial setting m sampled
on 11 April 1980 in the Union Square district of San Francisco.
81
-------
TABLE 16. PERCENTAGES OF INDOOR COMMERCIAL SETTINGS WITH ENTRANCE
DOORS IN DESIGNATED POSITIONS FOR FIVE SELECTED FIELD SURVEYS
Date
Geographic location
No. of Percentage
indoor Opened ClosedNot
settings door door recorded
24 Jan 80 University Avenue 24
Palo Alto
13 March 80 Castro Street 11
Mountain View
27 March 80 Westwood Village 19
Los Angeles
11 April 80 Chinatown-Financial 18
San Francisco
11 April 80 Union Square 24
San Francisco
25.0 50.0 25.0
9.1 81.8 9.1
31.6 63.2 5.2
27.8 72.2
60.9 26.1 13.0
82
-------
The hypotheses to be tested were as follows:
H0: v = w = x = y-=2
H-^: at least one mean does not equal the others.
A one-way analysis of variance is appropriate for resolving this
question if the assumptions of this statistical test can be met. First,
this test assumes that the mean CO concentrations at geographic locations
are normally distributed. This assumption can be made if the number of
indoor commercial settings visited at each location is 30 or more. Second,
this test assumes that the standard deviations of concentrations for each
location are approximately equal. The third assumption is that the number
of indoor settings sampled at one location is approximately equal to the
number of settings sampled at each of the other locations.
Our results show that the assumptions of the analysis of variance test
could not be met (Table 17). (See Figure D-5 in Appendix D for the
cumulative frequency distributions.) Rather than scratching this analysis,
we decided to pursue the analysis with this statistical test and impose a
more rigorous standard (0.001) for the level of significance.
In a one-way analysis of variance, F = 8.96. With 4 degrees of freedom
for the variance between groups and 90 degrees of freedom for the variance
within a group, p = 3.8 x 10~ . Thus, the null hypothesis was rejected as
p< ct= 0.001, the assigned level of significance for this case. The mean
concentration for the Chinatown-Financial district is the dissimilar value.
Its low mean concentration may be due to micrometeorological factors related
to that district's closeness to the coastline. The meteorological station
for this location is further inland and is probably not sensitive to these
factors. Consequently, it's possible that meteorological factors actually
have not been held constant for this location.
Suppose meteorological evidence supported removal of the
Chinatown-Financial district from the set of five locations. In a one-way
analysis of variance for the remaining four locations,
F = 2.02. With 3 degrees of freedom for the variance between groups and 73
degrees of freedom for the variance within a group, p = 0.12. Thus, the
null hypothesis could not be rejected as p> a= 0.001.
Since the null hypothesis could not be rejected, the variable of door
position may be disregarded. No uniformity in the variable of door position
could have explained significant differences in mean CO concentrations for
indoor commercial settings. However, since these differences were not
significant in this case, one may assume that door position had no effect
here. Tentatively, we can conclude that individual exposure to CO inside
commercial settings varies as much within a given geographic location as it
does between geographic locations. This is important if it is true. It
implies that data for different geographic locations can be consolidated if
only dates with similar wind speeds and atmospheric stability are used.
Even so, this conclusion is tentative and limited since all of the
geographic locations are in California.
83
-------
TABLE 17. STATISTICAL SUMMARY OF CO CONCENTRATIONS FOR
SELECTED INDOOR COMMERCIAL SETTINGS SITUATED AT
FIVE GEOGRAPHIC LOCATIONS
Date Geographic location
Mean
CO cone .
(ppm)
Std. dev.
of CO
cone.
(ppm)
Number
of
indoor
settings
24 Jan 80
University Avenue
Palo Alto
13 March 80 Castro Street
Mountain View
27 March 80 Westwood Village
Los Angeles
11 April 80 Chinatown-Financial
San Francisco
11 April 80 Union Square
San Francisco
2.7
2.2
2.6
0.9
3.4
1.2
1.2
0.9
0.7
2.1
24
11
19
18
23
84
-------
CHAPTER 8
COMPARISON OF PERSONAL AND FIXED STATION MONITORS
The research and development of PEM's have been predicated partly on
the belief that air quality measurements recorded at fixed station monitors
(FSM) do not indicate population exposure. One reason is that FSM's are too
sparsely distributed to capture the rich variation in concentrations which
presumably exists. Indeed, we have documented both horizontal and vertical
spatial variation with respect to CO concentrations. Another reason is
that, unlike FSM's, populations are highly mobile. They engage in
activities in a variety of settings, which predispose them to a range of air
pollutants and concentrations. Consequently, we compared between PEM's and
FSM's as part of this study.
The five geographic locations of this study were served by three FSM's
(Table 18). These stations monitor CO continuously and the results are
available on strip charts. However, these data were not very accessible
because of the difficulty of interpreting instantaneous ink trace readings
on the charts. On the other hand, the hourly average concentrations of CO
are tabulated and are available upon request to the districts.
Since the ambient concentrations measured by different monitors in
different locations were being compared, it was necessary to exclude certain
indoor settings with CO concentrations attributable to specific
sourcesprimarily parking garages and the 15-story general office building
in Palo Alto attached to a garage. Furthermore, visits to commercial
settings were of different lengths, and previous analysis had shown that
concentrations were fairly homogeneous over time for a given setting.
Therefore, it was appropriate not to compute a time-weighted average for
personal monitors for any given 60-minute period. Instead, an average CO
concentration of setting averages was computed for a given 60-minute period.
The data from PEM's were broken down according to whether the
measurements were taken indoors or outdoors, since earlier analysis had
shown that concentrations differed between these two locations when the
entrance door to the street was closed (Table 19). For the present
analysis, we did not have enough data to make further distinctions between
indoor settings with open doors versus those with closed doors. The
following variables were defined for the data from PEM's and FSM's:
X£ = average CO concentration reported by the FSM over time period
i:
85
-------
TABLE 18. STRAIGHT-LINE DISTANCE AND DIRECTION OF
GEOGRAPHIC LOCATIONS WITH RESPECT TO THEIR
NEAREST FIXED MONITORING STATIONS
Geographic location
Address of
nearest fixed
monitoring station
Intervening Direction
distance to
(kilometers) station
Union Square
San Francisco
939 Ellis St.
San Francisco a
1.4
east
Ch ina town-Financ ia1
San Francisco
939 Ellis St.
San Francisco a
2.3
northeast
University Avenue
Palo Alto
Castro Street
Mountain View
Westwood Village
Los Angeles
897 Barren Ave. 5.6 southeast
Redwood City a
897 Barren Ave. 14.8 southeast
Redwood City a
1535 S. Robertson 5.5 west
Los Angeles "
Notes:
a. Operated by Bay Area Air Quality Management District
b. Operated by South Coast Air Quality Management District
86
-------
TABLE 19. SUMMARY OF CO CONCENTRATIONS COLLECTED SIMULTANEOUSLY FROM FIXED
MONITORING STATIONS AND PERSONAL EXPOSURE MONITORS
Fixed stations
Date
9 Nov 79
9 Nov 79
13 Dec 79
13 Dec 79
13 Dec 79
13 Dec 79
13 Dec 79
13 Dec 79
13 Dec 79
13 Dec 79
24 Jan 80
24 Jan 80
31 Jan 80
31 Jan 80
7 Feb 80
7 Feb 80
7 Feb 80
6 March 80
7 March 80
7 March 80
13 March 80
13 March 80
13 March 80
Hour
11 a.m.
12 noon
9 a.m.
10 a.m.
11 a.m.
12 noon
1 p.m.
2 p.m.
3 p.m.
4 p.m.
3 p.m.
4 p.m.
10 a.m.
11 a.m.
12 noon
1 p.m.
2 p.m.
12 noon
5 p.m.
6 p .m .
1 1 a.m.
12 noon
1 p.m.
Geographic
location
Union Square
Union Square
Union Square
Union Square
Union Square
Union Square
Union Square
Union Square
Union Square
Union Square
Palo Alto
Palo Alto
Palo Alto
Palo Alto
Palo Alto
Palo Alto
Palo Alto
Chinatown-
Financial
Palo Alto
Palo Alto
Mountain View
Mountain View
Mountain View
Ave .
CO
(ppm)
2
2
4
4
4
3
3
3
3
4
3
3
3
3
2
2
2
3
2
3
1
1
1
Personal exposure monitors
Indoors
Ave . CO
(ppm)
5.2
7.2
2.8
5.0
5.4
6.4
4.6
4.5
2.7
10.8
2.8
2.4
3.2
3.2
0.4
1.1
2.4
1.6
2.6
2.3
2.2
1.0
2.9
No. of
Settings
2
4
1
2
5
3
1
5
2
2
14
6
2
8
4
10
1
7
4
2
5
2
4
Outdoors
Ave . CO
(ppm)
6.7
6.9
3.6
4.5
5.3
6.2
a
4.2
7.0
7.3
3.9
5.3
3.9
3.6
0.8
1.5
3.6
1.4
7.9
3.1
2.3
2.6
4.7
No. of
Settings
8
8
2
10
9
8
0
15
6
2
12
8
4
13
11
14
7
13
4
7
15
15
4
(continued)
-------
TABLE 19 (continued)
00
OO
Fixed stations
Date
27 March 80
27 March 80
27 March 80
27 March 80
27 March 80
11 April 80
11 April 80
11 April 80
11 April 80
11 April 80
11 April 80
13 June 80
13 June 80
13 June 80
13 June 80
13 June 80
11 July 80
11 July 80
11 July 80
11 July 80
Hour
3 p.m.
4 p.m
5 p.m
6 p .m.
7 p.m
12 noon
1 p.m.
3 p.m.
4 p.m.
5 p.m.
6 p.m.
11 a.m.
12 noon
1 p.m.
2 p.m.
3 p.m
3 p.m.
4 p.m.
5 p.m.
6 p.m.
Geographic
location
Los Angeles
Los Angeles
Los Angeles
Los Angeles
Los Angeles
Chinatown-
Financial
Chinatown-
Financial
Union Square
Union Square
Union Square
Union Square
Union Square
Union Square
Union Square
Union Square
Union Square
Palo Alto
Palo Alto
Palo Alto
Palo Alto
Ave.
CO
(ppm)
1
1
1
1
2
1
1
2
2
1
1
1
2
1
1
1
1
1
1
1
Personal exposure monitors
Indoors
Ave . CO
(ppm)
2.2
1.8
2.8
2.4
4.4
0.8
1.2
0.9
5.0
4.1
3.0
b
1.6
1.9
4.2
1.8
2.6
1.4
0
2.2
No. of
Settings
6
2
8
2
1
10
6
3
6
10
3
0
9
6
7
6
10
11
2
1
Outdoors
Ave. CO
(ppm)
4.3
6.9
3.1
3.0
6.9
1.8
1.4
3.7
5.9
4.9
5.2
2.4
2.7
3.2
5.2
1.7
2.2
2.5
0.9
b
No. of
Settings
4
9
10
6
10
14
26
9
8
14
12
2
11
7
8
8
9
6
8
0
Notes:
a. Averages for
b. Insufficient
settings visited during
data.
60 minutes
preceding shown.
-------
Y£ - average CO concentration obtained from the PEM for indoor
settings over time period i; and
z-[ = average CO concentration obtained from the PEM for outdoor
settings over time period i.
The difference in recorded concentrations between these two types of
monitoring systems was tested next. Two sets of hypotheses were
constructed. For indoor settings, the hypotheses were as follows:
H0: x = y
HI: x jt y
For outdoor settings, the hypotheses were as follows:
H0: x = z
H]_: x 7* 2
Before testing these hypotheses, it was appropriate to determine if the
observations should be matched as pairs, using Pearson's correlation
coefficient.(56) Matching is appropriate if this coefficient is both
nonzero and positive. Hence, the hypotheses constructed for the FSM and PEM
(indoor) data were as follows:
V rxy = 0
HI: rxy > 0
Similarly, these hypotheses were constructed for the FSM and PEM (outdoor)
data:
H
o
rxz = 0
xz
A one-tailed test of these hypotheses was performed for each case using
a t-statistic. For FSM and PEM (indoor) data, r = 0.50, and t = 3.67; for
FSM and PEM (outdoor) data, r = 0.30 and t = 1.993. Each value of t is
significant at a = 0.05, with 40 degrees of freedom for the indoor case and
39 for the outdoor case. Hence, the null hypotheses were rejected for each
case, and we concluded that it was appropriate to pair observations and to
proceed with the t-test for dependent samples. A two-tailed test was
selected because there was no reason to expect that the concentrations
recorded by one type of monitoring system would differ systematically from
those recorded by another. Using a two-tailed test imposed a more stringent
criterion of statistical significance (a/2 = 0.025) for rejecting a true
null hypothesis.
89
-------
A statistical summary of the 42 pairs of observations for the indoor
case indicates the following means and standard deviations:
X = 2.00 ppm J = 3.02 ppm
Sx = 1.04 ppm Sy = 2.03 ppm
For this case, the t-dependent statistic is -3.782, which is significant at
40 degrees of freedom for a/2 = 0.025. A statistical summary of the 41
pairs of observations for the outdoor case indicates:
x = 1.98 ppm z = 4.00 ppm
Sx = 1.04 ppm Sz = 1.95 ppm
In this case, the t-dependent statistic is -6.362, which is significant at
39 degrees of freedom for ot/2 = 0.025. Therefore, one must reject the null
hyptheses for both indoor and outdoor settings. Hence the mean
concentrations of CO estimated from FSM and PEM observations are equal. In
fact, the average concentration of CO for both indoor and outdoor settings
as measured by the PEM's was slightly higher than that reported by the
FSM's. This conclusion is consistent, in general, with the findings of past
studies.
90
-------
CHAPTER 9
MICROENVIRONMENTS OF COMMERCIAL AREAS
The definition of microenvironments used in this study possesses both
statistical and practical conflicting criteria that must be resolved. The
definition of microenvironments using statistical criteria has focused on
variation in CO concentrations due to time and spatial factors. Analysis of
the time factor suggests that for a given setting and date, CO
concentrations appear relatively stable over time, but that they may vary
between windy and calm days. Analysis of the spatial factor indicates that
statistical distinctions can be made between indoor and outdoor commercial
settings, but only when the entrance door is closed to the street.
Furthermore, enclosed parking garages, and some buildings which are
physically attached to these garages, have elevated CO concentrations.
Also, statistical distinctions can be made for indoor commercial settings on
opposite sides of the same street. Other statistical distinctions for
either indoor or outdoor settings either could not made (e.g., among indoor
commercial settings with different purposes) or could not be made
convincingly (e.g., among intersections).
From a practical standpoint, however, some of these distinctions based
on statistical criteria do not appear to be important. First, most people
can distinguish between calm and windy days, but not borderline cases.
Furthermore, the difference in mean CO concentrations between calm and windy
days for settings in the same geographic location is on the order of 1.94
ppm to 2.37 ppm. This difference in CO concentration is too small to
justify classifying separate microenvironments for calm and windy days.
Basically, the same conclusion can be reached for indoor commercial
settings with entrance doors closed to the street. On the average, the mean
indoor concentration of settings with closed doors was only 1.09 ppm smaller
than the mean concentration measured outdoors near the front entrance. The
GEOMET Indoor Outdoor Air Pollution (GIOAP) model,(28) as well as other
studies, has shown that indoor concentrations tend to lag behind outdoor
levels. This lag occurs because it takes some time for outdoor air
containing the pollutant to diffuse indoors. Closing the door should reduce
the flow rate and thereby increase the diffusion time constant. Thus,
indoor concentrations should be lower than outdoor concentrations initially,
but should eventually rise with time. Although indoor levels may never
reach outdoor levels exactly, as indicated by the GIOAP model, the indoor
levels may be elevated for relatively long periods. Consequently, the
average concentrations of the indoor and outdoor levels will be approximately
equal. The lower levels observed indoors in this investigation can be
91
-------
attributed partly to the "filtering" effect of the structure and are
consistent with the findings of other indoor air pollution studies. Thus,
indoor settings with opened entrance doors, together with settings with
closed doors, should be treated as one microenvironment, other things being
equal. This is fortuitous, since it doesn't seem realistic to ask people to
keep track of the position of entrance doors whenever they are asked to keep
records of their personal exposure to CO.
Likewise, it does not appear practical to ask subjects to keep detailed
records as to which side of a street they happen to be on. In this study,
indoor settings on the windward side of University Avenue in Palo Alto had
an average CO concentration that was only 1.51 ppm smaller than those
settings on the leeward side; both sides had been monitored simultaneously
on 11 July 1980. This finding is consistent with results reported by
Georgii, Bush, and Weber(51) and by Johnson e_t a_t^. (53) They postulate a
"leeward-windward" effectthat is, CO concentrations on the windward side
of the street always are lower than those on the leeward side, because the
wind forms a helical pattern in canyon-like streets. The leeward side of
the street receives air containing freshly mixed emissions from traffic,
while the windward side receives air relatively free of emissions from the
street. In our opinion, this difference in CO concentration between
opposite sides of the same street is too small to justify separate
microenvironments for each side.
What classification of commercial microenvironments is justified based
upon statistical and practical criteria? Our initial answer is an
hypothesis based upon the results of foregoing chapters. We will use the
distinction between indoor and outdoor settings as a starting point, even
though statistical differences could not always be found for every
situation. Despite these similarities, the indoor-outdoor distinction is a
very useful basis upon which to hinge certain modifiers. One modifier is
the purpose of the setting, the prime example being the CO concentrations in
the enclosed parking garage. A second modifier is physical design, the best
example being the 15-story general purpose office building located in Palo
Alto. In this case, the physical attachment of the building to an enclosed
parking garage appears to be responsible for elevated CO levels in the
building.
A third modifier may be internal sources of CO, such as commercial
settings with gas appliances (e.g., some restaurants). In this study,
settings with gas appliances were not systematically identified. Hence, our
initial assumption was that all indoor settings preparing hot foods
possessed internal sources of CO, and that these sources were in operation
when the setting was visited. Then these settings and all other indoor
settings, excluding enclosed parking garages and buildings attached to such
garages, were compared.
For this comparison, the following variables were defined:
X£ = mean CO concentration of setting i which prepares hot foods
(e.g., restaurants, cafeterias, bakeries, sandwich shops); and
92
-------
Yj - mean CO concentration of all other settings j, excluding
parking garages and buildings attached to such garages.
We expected that settings which prepare hot foods would possess higher mean
concentrations. Consequently, we considered the following hypotheses:
H0: x = y
HI: x > y
A one-tailed test of these hypotheses was performed using the t-statistic
for independent samples. A significance level of ct = 0.05 was specified for
rejecting a true null hypothesis. The following descriptive statistics were
prepared from Table C-l of Appendix C:
5 = 2.8 ppm y = 2.0 ppm
Sx = 2.2 ppm S = 1.5 ppm
nx =23 ny = 179
For these settings, t = 1.696, which is not significant for 23 degrees of
freedom at a = 0.05. Hence, the null hypothesis was accepted, and the two
types of settings were lumped together for subsequent analysis, since no
difference could be shown in mean CO concentrations.
Consequently, we propose four different types of commercial
microenvironments: (1) enclosed parking garages; (2) settings physically
attached to enclosed parking garages; (3) all other indoor commercial
settings; and (4) all outdoor settings. {(See Tables C-l and C-4 of
Appendix C for descriptive statistics.) In preparing these statistics, we
assumed that the measured CO concentration of a given setting could be
partitioned into two components: (1) the ambient concentration and (2) the
concentration associated with nearby sources of CO, most of which were motor
vehicles. The ambient concentration was defined as the minimum of all CO
concentrations measured instantaneously at 1-minute intervals for outdoor
settings only during the course of a particular field survey. The ambient
CO concentration (Table 20) was then subtracted from the mean concentration
for each setting to determine the concentration which could be associated
with nearby sources of CO.} A statistical summary of CO concentrations for
these four microenvironments was prepared (Table 21) and their cumulative
frequency distributions determined (Figure 10).
For all other indoor commercial settings, CO concentrations usually
were low and were well below the NAAQS of 9 ppm for 8 hours. Approximately
72% of these settings averaged less than 3 ppm, and about 84% were under 4
ppm. Yet within this class, those few commercial settings that exhibited
levels above 5 ppm often had special features which helped explain the
higher levels. For example, a bookstore in the Union Square district of San
Francisco had a mean CO concentration of 5.8 ppm (excluding the ambient
concentration), with a standard deviation of 1.2 ppm (based on 70
observations during two visits). This bookstore was located on the corner
93
-------
TABLE 20. SUMMARY OF AMBIENT CO CONCENTRATIONS FOR EACH FIELD SURVEY
Minimum CO concentration
Field
survey
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
Date
9 Nov 79
13 Dec 79
24 Jan 80
31 Jan 80
7 Feb 80
6 March 80
7 March 80
13 March 80
27 March 80
4 April 80
11 April 80
11 April 80
9 May 80
13 June 80
11 July 80
Geographic
location
Union Square
Union Square
Palo Alto
Palo Alto
Palo Alto
Chinatown-
Financial
Palo Alto
Mountain View
Los Angeles
Palo Alto
Chinatown-
Financial
Union Square
Palo Alto
Union Square
Palo Alto
Instr. #1
0
1
0
2
0
0
0
0
1
a
0
1
a
0
0
(ppm)
Instr. #2
2
1
0
1
0
b
0
b
b
a
0
0
a
b
0
Average
1
1
0
1
0
0
0
0
1
a
0
0.5
a
0
0
Notes:
a.
b.
No outdoor setting was monitored
Only one instrument was used for
for this survey
this survey.
94
-------
TABLE 21. STATISTICAL SUMMARY OF CO CONCENTRATIONS FOR
FOUR COMMERICAL MICROENVIRONMENTS
Microenvironment
Mean
CO
cone.
(ppm)a
Std. dev.
of CO No.
cone. of
(ppm) settings
Enclosed parking garages
Settings attached to enclosed
parking garages
All other indoor commercial
Note:
21.7
6.1
12.5
2.9
10
settings
All outdoor settings
2.1
3.0
1.6
2.6
202
368
a. The value reported excludes ambient concentrations of CO.
95
-------
0.01 0.05 0.1 0.2 0.5 1
CUMULATIVE FREQUENCY
10 20 30 M 50 60 70 80
CONSTRUCTED BY THE AUTHORS FROM PAPERS OF-
KC PROBABILITY X 2 LOG CYCLES
C KEUFFEL ft E5SER CO "A« null
90 95 98 99 99.5 99.8 99.9 99.99
100
Z
O
oc
Z
O
O
<
DC
I-
z
UJ
u
z
O
O
0.1
0.01 0.050.1 0.2 0.5 1 2 5 10 20 30 M 50 60 70 80 90 95 98 99 99.5 99.899.9 99.99'
CUMULATIVE FREQUENCY (%)
FIGURE 10. CUMULATIVE FREQUENCIES OF CO CONCENTRATIONS FOR FOUR MICROENVIRONMENTS.
96
-------
of an uphill street with a traffic signal and heavy traffic. It appears
that the relatively high CO emissions generated by vehicles accelerating
uphill accounted for the elevated CO levels inside this store.
97
-------
CHAPTER 10
CONCLUSIONS AND RECOMMENDATIONS
Because of their small size and portability, the miniaturized personal
exposure monitors used in this study generated a large volume of data on CO
concentrations to which California people actually are exposed as they
transact their daily business. If two of these instruments are carried
side-by-side, and if they are calibrated to known concentrations of
reference gases, they generally agree very well. Therefore, we concluded
that:
* PEM's are sufficiently well developed to be deployed by
trained technicians in large-scale field studies.
s
As this research project evolved, it became apparent that standard
survey methodology was necessary to enable other investigators to compare
their findings with ours. The methodology developed for the last survey of
the Union Square district of San Francisco represents an initial step. We
concluded that:
* A study area should be defined by several rigorous criteria
pertaining to both pedestrians and motor vehicle traffic.
Within that area, city blocks and commercial settings 'on each
block should be chosen at random. How often measurements are
taken also should be rigorously specified.
The present field study concentrated on both indoor and outdoor
commercial settings where people transact their daily business, such as
banks, offices, retail stores, hotels, restaurants, and department stores.
Typically, these settings had a front entrance opening onto a sidewalk along
a two- or four-lane street with moderately heavy traffic. These sidewalks
were usually filled with pedestrians and shoppers. We expected that CO
concentrations measured inside these commercial settings would be low or
negligible, especially when ambient background levels were zero. This
expectation was based on the results of previous studies which had shown
that CO concentrations rapidly decline with distance from streets.
As expected, we found that CO concentrations inside commercial settings
on busy streets usually were low, but seldom zero. We attributed this to
the diffusion of CO into buildings from vehicular emissions on the street or
from attached parking structures. It is likely that a given indoor
commercial setting can be represented by a "box model," such as the GIAOP
model. None of our results were inconsistent with such a representation.
98
-------
Several distinctions can be made among commercial settings in business
districts with respect to the definition of microenvironments. First,
parking garages, and buildings attached to these garages, represent two
distinct indoor microenvironments. All other indoor commercial settings,
including those with gas appliances, either could not be distinguished
statistically, or if statistically significant differences existed, they
were deemed inconsequential. Therefore, all other indoor commercial
settings can be treated as a single microenvironment. For both practical
and statistical reasons, all outdoor commercial settings were judged to be a
fourth microenvironment. The cumulative frequency plots for these four
microenvironments appear as reasonably straight lines on logarithmic
probability paper (Figure 10). We concluded that:
* Four distinct microenvironments can be defined in commercial
settings: enclosed parking garages, buildings attached to
them, all other indoor settings, and all outdoor settings.
Underground parking garages structurally attached to buildings proved
to be unexpected and important in explaining elevated indoor CO levels in
these buildings. Our data for one such "hot" building in Palo Alto show
that air pollutants from its attached garage can diffuse into the building
and become well mixed on many floors. Average CO concentrations in this
15-story office building were equal to or greater than 9 ppm on most of the
first 10 floors during four of seven visits to the building, all on a
different date. On two dates, when the building was visited twice during a
2-hour period, CO concentrations were relatively stable. This suggested
that high CO levels could persist in this building for as long as 8 hours.
If such high levels did persist that long, the 8-hour CO NAAQS was probably
violated on most floors.
Other studies have shown that motor vehicle operators, especially taxi
and bus drivers, are at risk to high CO exposures. It is possible that some
receptionists, secretaries, clerks, bookkeepers, lawyers, salesmen, etc.,
now must be added to this list of occupations. It is unlikely that any of
the employees of the 15-story office building in Palo Alto were aware that
their place of work involved a health risk from high CO exposure.
In this study, we visited nine high-rise buildings with attached
garages that were substantially enclosed. Only the office building in Palo
Alto was "hot," showing more than 9 ppm of CO on the upper floors on several
visits. This exposure could be reduced by improvements to the building s
ventilation system, by some changes in its structural design, and by simply
keeping doors from the building to the garage closed.
Unfortunately; our sample of buildings with attached garages was too
small and not rigorously selected to indicate how widespread and serious the
problem is. We concluded that:
* Buildings with attached garages have the potential for
violating the NAAQS for CO indoors, and some may do so almost
on a daily basis throughout the year.
99
-------
This study focused on commercial areas, consciously excluding shopping
centers with pedestrian malls surrounded by large parking lots. The
physical separation between the pedestrian mall and the parking lot is
usually sufficient to diminish CO exposure. However; some of these shopping
centers are attached to semienclosed parking garages. Such physical
enclosure, coupled with high volumes of pedestrian and vehicular traffic,
raises the potential for high levels of personal exposure to CO.
Finally, several other contributions of this study should be
highlighted here:
* For a given date and geographic location, the CO
concentrations of a particular commercial setting appear
relatively stable over time. However, they do vary slightly
for dates with different wind speeds.
* CO concentrations are dissimilar for indoor settings located
on opposite sides of the same street, giving support to a
windward-leeward effect for these settings. Although past
studies have shown a windward-leeward effect for outdoor
sidewalks, the present investigation is the first to show that
such an effect also applies to indoor settings.
* Average CO concentrations, as determined by personal monitors
for both indoor and outdoor settings, are statistically, but
not substantially, greater than the average CO concentrations
as determined simultaneously by fixed monitoring stations.
On the basis of our study, we recommend that:
* A great many technical improvements should be made in the data
storage and retrieval capacities of PEM's to make them more
convenient for both technicians and the general public to
use. Without these improvements, keeping records can be quite
tedious.
* A standardized methodology should be applied to several other
urban areas, but with a larger number of dates and times of
day than in our study. Specifically, both seasonal variations
and daily patterns in CO concentrations for commercial
settings need further research.
* A well-designed study should be undertaken of a representative
sample of buildings with attached garages in a large city.
* Future research projects should investigate shopping centers
with attached parking garages.
100
-------
REFERENCES
1. Ott, Wayne R. , "Concepts of Human Exposure to Air Pollution,"
Environmental International, Vol. 7, 1982, pp. 179-196.
2. Duan, Naihua, "Models for Human Exposure to Air Pollution,"
Environmental International, Vol 8, Nos. 1-6, 1982, pp. 305-309.
3. Repace, James L., Wayne R. Ott, and Lance A. Wallace, "Total Human
Exposure to Air Pollution," Paper No. 80-61.6 presented at 73rd
Annual Meeting of the Air Pollution Control Association, Montreal,
Canada, June 22-27, 1980.
4. Duan, Naihua, "Micro-Environment Types: A Model for Human Exposure
to Air Pollution," Technical Report No. 47, Department of Statistics,
Stanford University, Stanford, Calif., May 1981.
5. Ott, Wayne R. , "Exposure Estimates Based on Computer Generated
Activity Patterns," Paper No. 81-57.6 presented at the 74th Annual
Meeting of the Air Pollution Control Association, Philadelphia, Pa.,
June 21-26, 1981.
6. Ziskind, R.A., M. B. Rogozen, I. Rosner, and T. Carlin, "Carbon
Monoxide Intrusion in Sustained-Use Vehicles," Report No.
SAI-068-535, submitted to the U.S. Environmental Protection Agency -
Washington, D.C. , by Science Applications, Inc., November 15, 1979.
7. U.S. Department of Transportation, "Carbon Monoxide Intrusion into
School Buses," National Highway Traffic Safety Administration,
Washington, D.C., 1978.
8. Brice, Robert M., and Joseph F. Roesler, "The Exposure to Carbon
Monoxide of Occupants of Vehicles Moving in Heavy Traffic," Journal
of the Air Pollution Control Association, Vol. 16, No. 11, November
1966, pp. 597-600.
9. Lynn, David, Elbert Tabor, Wayne Ott, and Raymond Smith, "Present and
Future Commuter Exposures to Carbon Monoxide," paper presented at the
60th Annual Meeting of the Air Pollution Control Association,
Cleveland, Ohio, June 1967.
10. Godin, G., G. R. Wright, and R. J. Shephard, "Urban Exposure to
Carbon Monoxide," Archives of Environmental Health, Vol. 25, No. 5,
November 1972, pp. 305-313.
101
-------
11. Cortese, Anthony D., "Ability of Fixed Monitoring Stations to
Represent Personal Carbon Monoxide Exposure," Ph.D. Dissertation,
Harvard University, School of Public Health, Boston, Mass., 1976.
12. Willits, Neil H., and Wayne Ott, "The Effect of Intersections on the
CO Exposures of Motorists," Environmetrics-81 Conference, Washington,
D.C., April 1981.
13. Sebben, John, Peter Pimm, and Roy Shepherd, "Cigarette Smoke in
Enclosed Public Facilities," Archives of Environmental Health, Vol.
32, No. 2, March/April 1977, pp. 53-58.
14. GEOMET, Inc., "The Status of Indoor Air Pollution Research 1976,"
report prepared for the U.S. Environmental Protection Agency under
Contract No. 68-02-2294, Research Triangle Park, N.C.,
EPA-600/4-77-029, May 1977.
15. Yocum, J. E., W. L. Clink, and W. A. Cote, "Indoor-Outdoor Air
Quality Relationships," paper presented at the 63rd Annual Meeting of
the Air Pollution Control Association, St. Louis, Mo., June 1970.
16. General Electric Co., "Indoor-Outdoor Carbon Monoxide Pollution
Study," National Technical Information Service, Springfield, Va.,
NTIS No. PB220428, 1972.
17- Ziskind, Richard A., personal communication to Wayne Ott, Stanford,
Calif., 1980.
18. Wilson, C. B., and J. W. Schweiss, "Carbon Monoxide StudySeattle,
Washington, October 6 - November 2, 1977, Part 1, Executive Summary,"
U.S. Environmental Protection Agency, Region X, Seattle, Wash., EPA
910/9-78-054a, December 1978.
19. Wilson, C. B., and J. W. Schweiss, "Carbon Monoxide StudySeattle,
Washington, October 6 - November 2, 1977, Part 2," U.S. Environmental
Protection Agency, Region X, Seattle, Wash., EPA 910/9-78-0546,
December 1978.
20. Elkins, R. H., D.Y.C. Ng, J. Zimmer; and R. A. Marcus, "A Study of CO
and NOX levels in the Indoor Environment," paper presented at the
67th Annual Meeting of the Air Pollution Control Association, Denver;
Colo., June 1974,
21. Cote, W. A., W. A. Wade, III, and J. E. Yocum, "A Study of Indoor Air
Quality," U.S. Environmental Protection Agency, Research Triangle
Park, N.C., EPA-650/4-74-042, 1974.
22. Radford, E. P., "Survey of Carbon Monoxide in East Baltimore Housing,
Preliminary Report," Johns Hopkins School of Hygiene and Public
Health, Baltimore, Md., 1973.
102
-------
23 Bridge, D.P., and M. Corn, "Contribution to the Assessment of
Exposure of Nonsmokers to Air Pollution from Cigarette and Cigar
Smoke in Occupied Spaces," Environmental Research, Vol. 5, 1972, pp.
192-209.
24. Waller, R.E., B. T. Commins, and P.J. Lawther, "Air Pollution in Road
Tunnels," British Journal of Industrial Medicine, Vol. 18, No. 4,
October 1961, pp. 250-259.
25. Trompeo, G., G. Turletti, and 0. T. Giarrusso, "Concentration of
Carbon Monoxide in Underground Garages," Rass. Med. Ind., Vol. 33,
Nos. 3-4, May-August 1964, Rome, Italy, pp. 392-393.
26. Chovin, P., "Carbon Monoxide: Analysis of Exhaust Gas Investigations
in Paris," Environmental Research, Vol. 1, No. 2, October 1967, pp.
198-216.
27. Wright, G. R., S. Jewczyk, J. Onrot, P. Tomlinson, and R. Shepherd,
"Carbon Monoxide in the Urban Atmosphere: Hazards to the Pedestrian
and Street-Worker," Archives of Environmental Health, Vol. 30, No. 3,
March 1975, pp. 123-T21H
V.
28. Moschandreas, Demetrios J., John W. C. Stark, James E. McFadden, and
Sallie S. Morse, "Indoor Air Pollution in the Residential
Environment: Volume I. Data Collection, Analysis, and
Interpretation," U.S. Environmental Protection Agency, Research
Triangle Park, N.C., EPA-600/7-78-229a, December 1978.
29. Moschandreas, Demetrios J., "Indoor Air Pollution in the Residential
Environment: Volume II. Field Monitoring Protocol, Indoor Episodic
Pollutant Release Experiments and Numerical Analyses," U.S.
Environmental Protection Agency, Research Triangle Park, N.C.,
EPA-600/7-78-229b, December 1978.
30. Ott, Wayne R., "An Urban Survey Technique for Measuring the Spatial
Variation of Carbon Monoxide Concentrations in Cities," Ph.D.
Dissertation, Stanford University, Department of Civil Engineering,
Stanford, Calif., 1971.
31. Ott, Wayne R. and David T. Mage, "A Method for Simulating the True
Human Exposure of Critical Population Groups to Air Pollutants,"
Proceedings of the International Symposium on Recent Advances in the
Assessment of the Health Effects of Environmental Pollution, June
24-28, 1974, Report No. EUR 5360, Commission of the European
Communities, Luxembourg, 1975, Volume IV, pp. 2097-2107.
32. Wilson, C. B., and J. W. Schweiss, "Carbon Monoxide StudyBoise,
Idaho, November 25-December 22, 1977, Part 1, Executive Summary,"
U.S. Environmental Protection Agency, Region X, Seattle, Wash., EPA
9lO/9-78-055a, December 1978.
103
-------
33. Wilson, C. B., and J. W. Schweiss, "Carbon Monoxide StudyBoise,
Idaho, November 25-December 22, 1977, Part 2," U.S. Environmental
Protection Agency, Region X, Seattle, Wash., EPA 910/9-78-055b,
December 1978.
34. Midurski, Theodore, "Carbon Monoxide Hot Spot Guidelines, Volume I:
Techniques," U.S. Environmental Protection Agency, Research Triangle
Park, N.C., EPA-450/3-78-033, August 1978.
35. Midurski, Theodore, "Carbon Monoxide Hot Spot Guidelines, Volume III:
Workbook," U.S. Environmental Protection Agency, Research Triangle
Park, N.C., EPA-450/2-78-035, August 1978.
36. Baker, Norman L., "Air Pollution Concentrations in the Tahoe Basin,"
CALTRANS, State of California, Department of Transportation, District
3, May 1978.
37. Cooper, A. G., "Carbon Monoxide. A Bibliography with Abstracts,"
U.S. Department of Health, Education, and Welfare, Public Health
Service, Division of Air Pollution, Washington, D.C., PHS Publication
No. 1503, Bibliography No. 68, 1966.
38. "Air Quality Criteria for Carbon Monoxide," U.S. Department of
Health, Education, and Welfare, Public Health Service, National Air
Pollution Control Administration Publication No. AP-62, Washington,
D.C., March 1970.
39. "Air Quality Criteria for Carbon Monoxide," Committee on Challenge of
Modern Society, North Atlantic Treaty Organization, NATO/CCMS Report
No. 10, Brussels, Belgium, June 1972.
40. Jabara, James W., Thomas J. Keefe, Harry J. Beaulieu, and Roy M.
Buchan, "Carbon Monoxide: Dosimetry in Occupational Exposures in
Denver, Colorado," Archives of Environmental Health, Vol. 35, No. 4,
July/August 1980, pp. 198-204.
41. Bellin, Peter, and John D. Spengler, "Indoor and Outdoor Carbon
Monoxide Measurements at an Airport," Journal of the Air Pollution
Control Association, Vol. 30, No. 4, April 1980, pp. 392-394.
42. Ludwig, F. L., and J. H. S. Kealoha, "Selecting Sites for Carbon
Monoxide Monitoring," prepared for the U.S. Environmental Protection
Agency, Washington, D.C., under Contract No. 68-02-1471, September
1975.
43. Noll, Kenneth E., and Terry L. Miller, "Air Monitoring Survey
Design", Ann Arbor Science Publishers, Ann Arbor, Mich., 1977.
44. Rao, S. Trivikrama, Michael Keenan, Gopal Sistala, and Perry Samson,
"Dispersion of Pollutants Near Highways. Data Analysis and Model
Evaluation," U.S. Environmental Protection Agency, Research Triangle
Park, N.C., EPA-600/4-79-011, February 1979.
104
-------
45. Simmon, Patricia B., and Robert M. Patterson, "Commuter Exposure
Modeling Methodologies," U.S. Environmental Protection Agency,
Research Triangle Park, N.C., EPA-600/4-79-010, February 1979.
46. McCormick, R. A., and C. Xintaras, "Variation of Carbon Monoxide
Concentrations as Related to Sampling Interval, Traffic, and
Meteorological Factors," Journal of Applied Meteorology, Vol. 1, June
1962, pp. 237-243.
47- Brief, Richard S., Allen R. Jones, and John D. Yoder, "Lead, Carbon
Monoxide, and Traffic," Journal of the Air Pollution Control
Association, Vol. 10, No. 5, October 1960, pp. 384-388.
48. Ramsey, James M., "Concentrations of Carbon Monoxide at Traffic
Intersections in Dayton, Ohio," Archives of Environmental Health,
Vol. 13, July 1966, pp. 44-46.
49. Besner, David, and Patrick Atkins, "The Dispersion of Lead and Carbon
Monoxide from a Heavily-Travelled Expressway," Paper No. 70-12
presented at the 63rd Annual Meeting of the Air Pollution Control
Association, St. Louis, Mo., June 1970.
50. Colucci, Joseph M., and Charles R. Begeman, "Carbon Monoxide in
Detroit, New York, and Los Angeles Air," Environmental Science and
Technology, Vol. 3, No. 1, January 1969, pp. 41-47-
51. Georgii, H. W., E. Bush, and E. Weber, "Investigation of the Temporal
and Spatial Distribution of Emission Concentration of Carbon Monoxide
in Frankfurt/Main," Report No. 11 of the Institute for Meteorology
and Geophysics of the University of Frankfurt/Main, Translation No.
0477, National Air Pollution Control Administration, Washington,
D.C., May 1967-
52. Schnelle, K. B., F. G. Ziegler, and P. A. Kreukel, "A Study of the
Vertical Distribution of Carbon Monoxide and Temperature Above an
Urban Intersection," Paper No. 69-152 presented at the 62nd Annual
Meeting of the Air Pollution Control Association, New York, N.Y.,
June 1969.
53. Johnson, W.B. , et_ a_l_. , "Field Study for Initial Evaluation of an
Urban Diffusion Model for Carbon Monoxide," Comprehensive Report for
Contract CAPA-3-68 (1-69), prepared by Stanford Research Institute
for the Coordinating Research Council and the U.S. Environmental
Protection Agency, June 1971.
54. DeMarrais, Gerald A., "Diurnal Variations in Carbon Monoxide
Concentrations, Traffic Counts and Meteorology," U.S. Environmental
Protection Agency, Research Triangle Park, N.C., EPA 600/4-77-009,
February 1977.
105
-------
55. DeMarrais, Gerald A., "Diurnal Variations in Carbon Monoxide
Concentrations, Traffic Counts and Meteorology," U.S. Environmental
Protection Agency, Research Triangle Park, N.C., EPA 600/4-77-016,
April 1977-
56. Roscoe, John T. Fundamental Research Statistics for the Behavioral
Sciences, Holt, Rinehart and Winston, Inc., New York, 1975.
106
-------
APPENDIX A
PHOTOGRAPHS OF PEM INSTRUMENTS, SAMPLING APPROACH,
AND SEVERAL SAMPLING LOCATIONS
The 14 photographs that follow illustrate the measurement methodology
used in the present research investigation:
1. Two photographs of the PEM instruments and associated equipment.
2. Two photographs of samples being collected outdoors in downtown
San Francisco near Union Square.
3. Four photographs of samples being collected indoors in various
stores in San Francisco.
4. Two photographs of the downtown main street of Palo Alto.
5. A photograph of the "hot building" in Palo Alto.
6. Three photographs of samples being collected outdoors in Palo
Alto.
7. Two photographs showing how PEM instruments were carried
side-by-side.
107
-------
o
CD
Personal exposure monitoring instruments used in the present research investigation: General
Electric CO "Detector" Model 15ECS3C01 with case and calibration gases (left); and Energetics
Science, Incorporated (ESI) CO "Dosimeter" Model 9000 with Dupont pump, digital readout system,
and calibration gases (right).
-------
Measurement of CO outdoors in San Francisco using the GE
personal exposure monitor on a sidewalk near Union Square
(top) and while crossing a street at a downtown intersection
(bottom).
109
-------
Measurement of CO indoors in San Francisco using the GE
personal exposure monitors in two commercial settings:
inside a cutlery store (top); inside a coffee and gift store
(hot torn).
110
-------
Measurement of CO in a San Francisco department store using
the GE personal exposure monitor in an aisle (top) and near
a jewelry counter (bottom).
Ill
-------
University Ave. in Palo Alto showing close-up view of street
(top) and wide-angle view of street and downtown area
(bottom).
112
-------
II
iii!I«!!!H
The 15-story office building in downtown Palo Alto that was
"hot" with respect to interior CO levels (top), and
investigator taking a reading on University Ave. with GE
personal exposure monitor (bottom). (Since completion of
this studv, the managers took steps to reititre levels in this
building; see p. 79.)
-------
Measurement of CO in downtown Palo Alto using ESI personal
exposure monitor on a corner (top) and in front of a store
(bottom). Digital readout device, monitor, and pump were
carried on the clip board.
114
-------
Measurement of CO with two investigators walking
side-by-side to compare readings from two >",K personal
monitors (top) and to record instantaneous readings on
clipboards at one-minute intervals (bottom).
115
-------
APPENDIX B
SAMPLING PROCEDURE FOR UNION SQUARE DISTRICT OF
SAN FRANCISCO FOR 13 JUNE 1980
The boundaries of the Union Square district (Figure B-l)
were defined on the basis of:
1. CO emissions are expected to be high as average daily traffic
(ADT) exceeds 10,000 vehicles on a given street.
2. Pedestrian exposure is expected to be high as this area
encompasses the C-3-R District ("downtown retail") as shown on
the San Francisco land use map.
Surveyors followed this sampling procedure:
1. Begin with the first block on your list (Table B-l);
2. Start at the corner where the block number has been circled
and take 3 1-minute observations of CO concentrations at each
corner of the intersection;
3. Proceed clockwise around the block taking 3 1-minute
observations of CO concentrations at each kth establishment
with an entrance at the street level. Let k=l for multistory
establishments and k=3 for first-floor only establishments.
4. At each establishment selected above, take both indoor and
outdoor CO concentrations and note whether the entrance door
is opened or closed. For indoor observations, measure only
the street-level floor;
5. Repeat the above procedure for the next block on your list;
6. Record all information on the form provided (Table B-2).
116
-------
Bush St.
N
Figure B-l. Field map of Union Square district of San Francisco, California
117
-------
TABLE B-l. CITY BLOCK ASSIGNMENTS FOR SURVEYORS
City block Surveyor
number DeMandel Fairley Flachsbart
1 X
2 X
4
7 X
8 X
11 X
12 X
15 X
19
Ott
X
X
118
-------
TABLE B-2. SAMPLE DATA FORM
Block No.
Date
Instrument
Type of Outdoor Indoor Door
Time Intersection Corner establishment CO CO position Comments
119
-------
APPENDIX C
TYPICAL CO CONCENTRATIONS FOR COMMERCIAL SETTINGS
This appendix contains four tables summarizing typical CO concentrations
for commercial settings:
TABLE C-l. Mean concentrations of CO for indoor commercial settings
broken down by geographic location.
TABLE C-2. Statistical summary of CO concentrations for indoor
commercial settings located near Union Square in San
Francisco, California, and spanning selected dates
from 9 November 1979 through 13 June 1980.
TABLE C-3. Statistical summary of CO concentrations for indoor
commercial settings located along University Avenue
in downtown Palo Alto, California, and spanning
selected dates from 24 January 1980 through 11 July 1980.
TABLE C-4. Mean concentrations of CO for outdoor commercial settings
broken down by geographic location.
120
-------
TABLE C-l. MEAN CONCENTRATIONS OF CO FOR INDOOR COMMERCIAL SETTINGS
BROKEN DOWN BY GEOGRAPHIC LOCATION
Antique shop
Art gallery #1
Art gallery #2
Art gallery #3
Art gallery #4
Bakery shop
Bank //I
Bank #2
Bank #3
Bank #4
Bar
Barber shop
Bicycle shop
Book store #1
Book store #2
Book store #3
Cafeteria
Camera repair shop
Camera store #1
Camera store #2
Candy store
Mtn. View
Mean
CO No.
cone. of
(ppm) obs.
1.0 4
3.5 2
1.0 4
2.8 3
Palo Alto
Mean
CO No.
cone. of
(ppm) obs.
0.5 2
2.9 6
4.6 33
1.4 6
0.8 8
1.7 16
0.0 3
0.5 2
1.3 4
2.6 8
1.4 7
Chinatown-
Financial ,
San Francisco
Mean
CO No.
cone. of
(ppm) obs.
0.5 4
0.5 4
0.3 4
2.6 10
2.0 2
Union Square^
San Francisco
Mean
CO No.
cone. of
(ppm) obs.
1.7 3
1.3 12
1.2 22
2.0 3
4.7 3
1.0 3
2.0 3
2.0 3
5.7 3
6.7 70
1.5 4
3.0 3
1.0 3
1.0 3
3.8 10
Westwood,
Los Angeles
Mean
CO No.
cone. of
(ppm) obs.
1.8 4
3.5 6
2.0 33
(continued)
-------
TABLE G-l. (continued)
Type of
commercial
setting
Card shop
China & glassware
Clean air store
Clothing store #1
Clothing store #2
Clothing store #3
Clothing store #4
Clothing store #5
Clothing store #6
Clothing store #7
Clothing store #8
Clothing store #9
Clothing store #10
Clothing store #11
Clothing store #12
Clothing store #13
Clothing store #14
Clothing store #15
Cocktail lounge
Coffee shop #1
Coffee shop #2
Copy center
Currency exchange
Cutlery shop
Mtn. View
Mean
CO No.
cone. of
. (ppm) , obs.
Palo Alto
Mean
CO No.
cone. of
. (ppm) , obs.
0.1 16
3.0 2
2.6 4
0.3 4
0.0 3
2.3 154
3.1 18
1.7 6
Chinatown-
Financial,
San Francisco
Mean
CO No.
cone. of
(ppm) obs.
1.8 4
Union Square,
San Francisco
Mean
CO No.
cone. of
(ppm) obs.
3.7 3
4.3 4
2.0 2
3.7 10
6.3 6
2.0 4
0.7 3
1.7 3
1.7 3
3.0 3
3.7 3
1.3 3
0.3 3
1.0 3
1.0 3
1.3 3
2.7 82
2.3 3
2.7 3
4.7 18
Westwood ,
Los Angeles
Mean
CO No.
cone. of
(ppm) obs.
2.6 3
N>
to
(continued)
-------
TABLE C-l. (continued)
Type of
commercial
setting
Delicatessen #1
Delicatessen #2
Dept. store #1
Dept. store #2
Dept. store #3
Dept. store #4
Dept. store //5
Dept. store #6
Drug store #1
Drug store #2
Dry cleaning
Electronics
Fabric store #1
Fabric store #2
Fast food arcade
Fitness store
Florist
Furniture store #1
Furniture store #2
Game store
Gift store #1
Gift store #2
Gift store #3
Gift store #4
Mtn . View
Mean
CO No.
cone. of
(ppm) . obs.
1.0 4
4.0 2
Palo Alto
Mean
CO No.
cone. of
(ppm) . obs.
2.6 8
1.7 11
1.0 3
0.0 2
0.5 15
2.2 6
2.7 21
2.2 6
2.0 7
0.0 2
Chinatown-
Financial,
San Francisco
Mean
CO No.
cone. of
(ppm) obs.
0 2
1.0 1
0.7 10
Union Square,
San Francisco
Mean
CO No.
cone. of
.(ppm) obs.
4.7 3
3.2 4
4.6 79
4.6 20
2.7 80
1.2 5
1.8 6
0.8 4
1.3 3
2.7 3
2.0 3
1.0 3
2.3 3
5.2 39
2.7 3
0.0 3
6.0 3
2.0 3
5.4 5
1.7 3
Westwood,
Los Angeles
Mean
CO No.
cone. of
(ppm) obs.
1.8 4
2.8 9
2.2 4
OJ
(continued)
-------
TABLE C-l. (continued)
Type of
commercial
setting
Health store
Hotel //I
Hotel #2
Hotel #3
Hotel #4
Hotel #5
Hotel #6
Hotel #7
Hotel #8
Hotel #9
Hotel #10
Hotel #11
Hotel #12
Hotel #13
Ice cream store #1
Ice cream store #2
Import store
Jewelry store #1
Jewelry store #2
Linen shop
Music store
Novelty store #1
Novelty store #2
Mtn. View
Mean
CO No.
cone. of
(pptn) obs.
2.3 3
Palo Alto
Mean
CO No.
cone. of
Cppm") obs.
3.0 4
0.3 3
0.0 2
2.0 16
Chinatown-
Financial,
San Francisco
Mean
CO No.
cone. of
fpiW) . obs.
1.8 14
Union Square,
San Francisco
Mean
CO No.
cone. of
-------
TABLE C-l. (continued)
Type of
commercial
setting
Office building #1
Office building #2
Office building #3
Office building #4
Office building #5
Office building #6
Office building #7
Office building #8
Office building #9
Office building #10
Office building #11
Office machines
Paint store
Parking garage #1
Parking garage #2
Parking garage #3
Mtn. View
Mean
CO No.
cone. of
(ppm) obs.
Palo Alto
Mean
CO No.
cone. of
(ppm) obs.
9.7 384
4.1 22
2.6 14
3.6 14
1.7 9
25.0 110
42.8 14
30.0 8
Chinatown-
Financial,
San Francisco
Mean
CO No.
cone. of
(ppm) obs.
0.9 16
2.0 1
1.0 4
0.5 2
0.5 4
1.0 2
0.3 4
15.5 30
15.0 2
Union Square,
San Francisco
Mean
CO No.
cone. of
(ppm) obs.
2.0 2
3.9 16
3.3 4
2.0 3
2.3 3
5.0 3
2.3 3
0.3 3
5.0 3
2.7 3
9.0 3
35.0 108
12.8 4
6.8 4
Westwood,
Los Angeles
Mean
CO No.
cone. of
(ppm) obs.
1.7 11
4.6 16
15.0 4
13.0 4
(continued)
-------
TABLE C-l (continued)
Type of
commercial
setting
Stereo store
Theater #1
Theater #2
Tile store
Tobacco shop
Toy store
Travel agency #1
Travel agency #2
Travel agency #3
Travel agency #4
Variety store
Mtn. -View
Mean
CO No.
cone. of
(ppm) obs.
1.3 3
Palo Alto
Mean
CO No.
cone. of
(ppm) obs.
0.0 3
2.1 6
2.1 22
Chinatown-
Financial,
San Francisco
Mean
CO No.
cone. of
(ppm) obs.
Union Square,
San Francisco
Mean
CO No.
cone. of
(ppm) obs.
1.0 3
6.7 3
0.0 3
3.8 6
9.5 2
4.3 4
1.3 3
4.3 3
5.3 3
Westwood,
Los Angeles
Mean
CO No.
cone. of
(ppm) obs.
-------
TABLE C-l.
(continued)
Type of
commercial
setting
Pet store
Picture frame shop
Print shop
Realty office
Record store #1
Record store #2
Restaurant #1
Restaurant #2
Restaurant #3
Restaurant #4
Restaurant #5
Restaurant #6
Restaurant //7
Sandwich shop
Shoe store //I
Shoe store #2
Shoe store #3
Sporting goods #1
Sporting goods #2
Stamp store
Stationery store //I
Stationery store #2
Mtn. View
Mean
CO No.
cone. of
(ppm) obs.
4.0 3
2.7 37
1.0 3
Palo Alto
Mean
CO No.
cone. of
(ppm) obs.
0.3 6
0.3 25
1.4 72
2.4 10
3.5 6
1.6 16
2.3 3
1.2 15
2.4 13
Chinatown-
Financial,
San Francisco
Mean
CO No.
cone. of
(ppm) obs.
1.5 2
2.0 36
2.3 4
1.1 60
1.0 4
Union Square ,
San Francisco
Mean
CO No.
cone. of
(ppm) obs.
1.3 3
26.3 4
2.0 3
2.3 3
4.6 196
2.0 2
2.3 3
1.3 3
3.0 3
8.6 7
10.2 74
2.0 2
3.7 3
1.0 3
1.0 3
Westwood ,
Los Angeles
Mean
CO No.
cone. of
(ppm) obs.
3.0 5
2.2 5
2.0 3
4.9 69
2.0 3
2.3 3
(continued)
-------
TABLE C-2.
STATISTICAL SUMMARY OF CO CONCENTRATIONS FOR INDOOR
COMMERCIAL SETTINGS LOCATED NEAR UNION SQUARE IN
SAN FRANCISCO, CALIFORNIA, AND SPANNING SELECTED
DATES FROM 9 NOVEMBER 1979 THROUGH 13 JUNE 1980
Type of
commercial
setting
Antique shop
Art gallery //I
Art gallery #2
Art gallery #3
Art gallery #4
Bank //I
Bank #2
Bank #3
Bar
Bookstore #1
Bookstore #2
Bookstore #3
Cafeteria
Camera repair shop
Candy store
China & glassware
Clothing store #1
Clothing store //2
Clothing store //3
Clothing store #4
Clothing store #5
Clothing store #6
Clothing store #7
Clothing store #8
Clothing store #9
Clothing store #10
Clothing store #11
Clothing store #12
Clothing store #13
Clothing store #14
Clothing store #15
Coffee shop #1
Coffee shop #2
Currency exchange
Cutlery shop
Delicatessen #1
Delicatessen #2
Mean
CO
concentration
(ppm)
1.7
1.3
1.2
2.0
4.7
1.0
2.0
2.0
5.7
6.7
1.5
3.0
1.0
1.0
3.8
3.7
4.3
2.0
3.7
6.3
2.0
0.7
1.7
1.7
3.0
3.7
1.3
0.3
1.0
1.0
1.3
2.7
2.3
2.7
4.7
4.7
3.2
Std. dev.
of CO
concentration
(ppm)
0.6
0.6
0.7
0.0
2.5
0.0
0.0
1.0
0.6
1.2
0.6
0.0
0.0
0,0
0.4
1.2
0.5
0,0
1.2
1.2
0.0
0.6
0.6
0.6
0.0
0.6
0.6
0.6
0.0
o.o
0.6
0.8
0.6
0.6
1.9
1.5
1.3
Number
of
visits
1
1
1
1
1
1
1
1
1
2
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
2
1
1
2
1
1
Total
number
of
observations
3
12
22
3
3
3
3
3
3
70
4
3
3
3
10
3
4
2
10
6
4
3
3
3
3
3
3
3
3
3
3
82
3
3
18
3
4
(continued)
128
-------
TABLE C-2 (continued)
Type of
commercial
setting
Department store #1
Department store #2
Department store //3
Department store #4
Department store #5
Department store #6
Drug store #1
Drug store #2
Fabric store #1
Fabric store #2
Fitness store
Florist
Furniture store #1
Furniture store //2
Gift store #1
Gift store #2
Gift store #3
Gift store #4
Hotel #1
Hotel //2
Hotel #3
Hotel #4
Hotel #5
Hotel #6
Hotel #7
Hotel #8
Hotel #9
Hotel #10
Hotel #11
Hotel #12
Hotel #13
Jewelry store #1
Jewelry store #2
Linen shop
Novelty store #1
Novelty store #2
Office building //I
Office building #2
Office building #3
Office building #4
Office building #5
Office building #6
Office building #7
Mean
CO
concentration
(ppm)
4.6
4.6
2.7
1.2
1.8
0.8
1.3
2.7
2.0
1.0
2.3
5.2
2.7
0.0
6.0
2.0
5.4
1.7
9.9
4.3
2.5
5.3
3.3
4.3
6.0
3.7
2.7
3.0
1.3
2.7
1.0
1.0
0.3
0.0
3.3
4.0
2.0
3.9
3.3
2.0
2.3
5.0
2.3
Std. dev.
of CO
concentration
(ppm)
3.6
1.2
0.7
0.4
0.4
0.5
0.6
0.6
0.0
0.0
0.6
1-.5
0.6
0.0
1.0
1.0
1.1
0.6
1.2
1.7
0.6
1.7
2.5
0.6
0.0
0.6
0.6
0.0
0.6
0.6
0.0
0.0
0.6
0. 0
0.6
1.0
0.0
1.1
1.0
0.0
0.6
0.0
2.1
Number
of
visits
4
1
3
2
1
1
1
1
1
1
1
3
1
1
1
1
1
1
1
3
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
Total
number
of
observations
79
20
80
5
6
4
3
3
3
3
3
39
3
3
3
3
5
3
10
38
4
48
3
3
3
3
3
3
3
3
3
3
3
3
3
3
2
16
4
3
3
3
3
(continued)
129
-------
TABLE C-2 (continued)
Type of
commercial
setting
Office building #8
Office building #9
Office building #10
Office building #11
Parking garage #1
Parking garage #2
Parking garage #3
Picture frame shop
Print shop
Realty office
Restaurant #1
Restaurant #2
Restaurant #3
Restaurant #4
Restaurant #5
Restaurant #6
Restaurant #7
Sandwich shop
Shoe store #1
Shoe store #2
Shoe store #3
Stamp store
Stereo store
Theater #1
Theater #2
Tobacco shop
Toy store
Travel agency #1
Travel agency #2
Travel agency #3
Travel agency #4
Mean
CO
concentration
(ppm)
0.3
5.0
2.7
9.0
35.0
12.8
6.8
1.3
26.3
2.0
2.3
4.6
2.0
2.3
1.3
3.0
8.6
10.2
2.0
3.7
1.0
1.0
1.0
6.7
0.0
3.8
9.5
4.3
1.3
4.3
5.3
Std. dev.
of CO
concentration
(ppm)
0.6
1.0
0.6
2.0
13.6
1.3
0.5
0.6
12.3
0.0
1.2
1.1
0.0
0.6
0.6
0.0
1.9
6.3
0.0
0.6
0.0
0.0
0.0
1.5
0.0
1.2
0.7
1.0
0.6
0.6
0.6
Number
of
visits
1
1
1
1
7
1
1
1
2
1
1
1
1
1
1
1
1
4
1
1
1
1
1
1
1
2
1
1
1
1
1
Total
number
of
observations
3
3
3
3
108
4
4
3
4
3
3
196
2
3
3
3
7
74
2
3
3
3
3
3
3
6
2
4
3
3
3
130
-------
TABLE C-3.
STATISTICAL SUMMARY OF CO CONCENTRATIONS FOR INDOOR
COMMERCIAL SETTINGS LOCATED ALONG UNIVERSITY AVENUE
IN DOWNTOWN PALO ALTO, CALIFORNIA, AND SPANNING
SELECTED DATES FROM 24 JANUARY 1980 THROUGH 11 JULY 1980
Type of
commercial
setting
Bakery shop
Bank //I
Bank //2
Barber shop
Bicycle shop
Book store //I
Book store #2
Book store //3
Camera store //I
Camera store //2
Candy store
Card shop
Clean air store
Clothing store //I
Clothing store //2
Clothing store #3
Cocktail lounge
Coffee shop
Copy center
Delicatessen
Drug store //I
Drug store //2
Dry cleaning
Electronics
Fabric store
Fast food arcade
Florist
Furniture store //I
Furniture store //2
Ice cream store #1
Ice cream store //2
Jewelry store
Mexican restaurant
Music store
Office building //I
Office building #2
Office building #3
Office machines
Mean
CO
concentration
(ppm)
0.5
2.9
4.6
1.4
0.8
1.7
0.0
0.5
1.3
2.6
1.4
0.1
3.0
2.6
0.3
0.0
2.3
3.1
1.7
2.6
1.7
1.0
0.0
0.5
2.2
2.7
2.2
2.0
o.o
3.0
0.3
0.0
1.4
2.0
9.7
4.1
2.6
3.6
Std. dev.
of CO
concentration
(ppm)
0.7
0.6
1.3
2.2
0.9
1.1
0.0
0.7
0.5
1.7
1.4
0.3
0.0
0.9
0.5
0.0
0.5
0.8
1.5
1.4
0.7
0.0
0.0
0.5
1.2
1.9
1.0
1.9
0.0
0.8
0.6
0.0
0.6
1.0
5.4
1.8
2.6
1.9
Number
of
visits
2
1
5
2
2
4
1
1
1
3
2
2
1
1
1
1
1
2
2
1
3
1
1
2
2
4
2
2
1
2
1
1
3
4
9
1
1
3
Total
number
of
observations
4
6
33
6
8
16
3
2
4
8
7
16
2
4
4
3
154
18
6
8
11
3
2
15
6
23
6
7
2
4
3
2
72
16
384
22
14
14
Paint store 1.7 0.5 2 9
(continued)
131
-------
TABLE C-3 (continued)
Type of
commercial
setting
Parking garage #1
Parking garage #2
Parking garage #3
Pet store
Realty office
Shoe store #L
Shoe store #2
Sporting goods //I
Sporting goods #2
Stationery store #1
Stationery store #2
Theater
Travel agency
Variety store
Mean
CO
concentration
(ppm)
35.0
43.8
30.0
0.3
0.3
2.4
3.5
1.6
2.3
1.2
2.4
0.0
2.1
2.1
Std. dev.
of CO
concentration
(ppm)
12.1
12.0
4.0
0.5
1.0
1.7
1.8
1.3
0.6
0.8
1.4
0.0
0.8
1.2
Numb er
of
visits
8
1
1
2
4
3
2
2
1
4
2
1
2
4
Total
number
of
observations
110
14
8
6
25
10
6
16
3
15
13
3
6
22
132
-------
TABLE C-4.
MEAN CONCENTRATIONS OF CO FOR OUTDOOR COMMERCIAL SETTINGS BROKEN DOWN BY GEOGRAPHIC LOCATION
Type of
commercial
setting
Arcade #1
Arcade #2
Intersection #1
Intersection #2
Intersection #3
Intersection #4
Intersection #5
Intersection #6
Intersection #7
Intersection #8
Intersection #9
Intersection #10
Intersection #11
Intersection //12
Intersection #13
Intersection //14
Intersection #15
Intersection #16
Intersection #17
Intersection #18
Intersection #19
Intersection #20
Mt. View
Mean
CO No.
cone. of
(ppm) obs.
0.5 2
3.9 8
0.5 2
1.5 4
10.4 11
3.8 4
1.5 8
3.0 1
1.6 10
Palo Alto
Mean
CO No.
cone. of
(ppm) obs.
5.7 3
1.1 6
2.2 31
2.0 26
3.7 30
2.9 23
1.4 24
3.4 44
3.9 8
6.5 2
2.0 2
Chinatown-
Financial,
San Francisco
Mean
CO No.
cone. of
(ppm) obs.
1.5 10
2.0 4
1.0 2
0.3 3
2.0 2
0.5 2
6.7 6
8.0 2
5.0 4
2.0 2
0.0 2
0.5 2
7.3 4
0.0 2
9.0 2
Union Square,
San Francisco
Mean
CO No.
cone. of
(ppm) obs .
3.5 16
1.0 15
0.3 3
6.1 17
3.7 3
12.2 9
3.9 23
8.9 26
1.8 6
1.7 3
5.0 4
3.2 3
1.9 21
6.3 25
4.1 15
4.1 7
2.0 3
8.8 12
7.3 36
0.5 2
Westwood,
Los Angeles
Mean
CO No.
cone. of
(ppm) obs .
6.0 2
2.0 3
3.4 7
2.7 7
9.8 25
4.7 9
7.5 8
2.0 1
2.0 1
5.0 3
iiTinf-T niiorl ^
OJ
-------
TABLE C-4 (continued)
OJ
-P-
Type of
commercial
setting
Intersection #21
Intersection #22
Intersection #23
Intersection #24
Intersection #25
Intersection #26
Intersection #27
Intersection #28
Intersection #29
Intersection #30
Midblock #1
Midblock #2
Midblock #3
Midblock #4
Midblock #5
Midblock #6
Midblock #7
Midblock #8
Midblock #9
Midblock #10
Midblock #11
Midblock #12
Mt. View
Mean
CO No.
cone. of
(ppm) obs .
0.0 3
0.0 2
3.0 1
1.0 3
2.0 1
2.0 4
1.7 11
1.2 5
1.7 3
8.0 1
2.4 7
3.5 2
Palo Alto
Mean
CO No.
cone. of
(ppm) obs.
5.5 4
10.1 6
4.3 2
3.7 6
1.5 8
4.5 2
5.0 2
4.3 4
4.0 2
9.0 4
7.5 2
3.5 6
Chinatown-
Financial,
San Francisco
Mean
CO No.
cone. of
(ppm) obs .
1.0 2
3.0 4
1.8 6
0.0 2
0.0 2
0.0 2
0.0 4
0.8 4
0.0 2
1.5 2
0.0 2
0.0 2
Union Square,
San Francisco
Mean
CO No.
cone. of
(ppm) obs .
4.1 18
2.9 8
1.0 3
3.6 15
6.3 6
5.0 10
1.4 12
2.7 3
5.8 6
6.7 3
3.8 4
3.8 4
3.1 8
5.8 10
3.3 6
3.8 4
2.0 2
2.8 4
3.8 6
2.3 4
4.0 2
2.5 2
Westwood,
Los Angeles
Mean
CO No.
cone. of
(ppm) obs.
4.2 6
3.0 2
7.3 3
8.8 4
5.5 2
2.0 1
4.0 1
2.0 1
2.0 1
2.0 1
2.0 1
2.0 2
(continued)
-------
TABLE C-4 (continued)
Type of
commercial
setting
Midblock #13
Midblock #14
Midblock #15
Midblock #16
Midblock #17
Mibblock #18
Midblock #19
Midblock #20
Midblock #21
Midblock #22
Midblock #23
Midblock #24
Midblock #25
Midblock #26
Midblock #27
Midblock #28
Midblock #29
Midblock #30
Midblock #31
Midblock #32
Midblock #33
Midblock #34
Midblock #35
Midblock #36
Mtn. View
Mean
CO No.
cone. of
(ppm) obs.
5.0 1
6.3 7
7.0 4
2.5 2
Palo Alto
Mean
CO No.
cone. of
(ppm) obs.
2.0 2
2.0 4
4.3 6
3.3 4
6.5 2
1.5 2
2.3 6
2.3 2
3.1 4
1.8 2
2.8 2
6.3 2
4.4 4
6.8 2
3.8 4
3.0 6
3.5 4
0.5 2
0.0 2
1.0 2
0.5 4
0.5 2
0.0 2
1.0 2
Chinatown-
Financial,
San Francisco
Mean
CO No.
cone. of
(ppm) obs .
0.0 2
1.3 4
1.0 2
0.5 4
1.0 4
0.0 2
1.0 2
5.0 4
1.3 6
1.7 6
1.7 6
1.5 2
Union Square^
San Francisco
Mean
CO No.
cone. of
(ppm) obs.
5.3 4
6.5 4
2.5 4
2.8 4
5.5 2
20.5 2
2.5 2
4.3 8
3.3 4
3.5 2
3.8 6
5.7 6
2.0 2
2.2 8
2.5 2
3.3 4
2.5 2
3.3 4
6.8 4
4.5 4
4.5 2
8.0 1
1.5 2
1.5 2
Westwood,
Los Angeles
Mean
CO No.
cone. of
(ppm) obs.
5.0 1
5.0 1
2.0 3
6.3 3
(continued)
-------
TABLE C-4 (continued)
Type of
commercial
setting
Midblock #37
Midblock #38
Midblock #39
Midblock #40
Midblock #41
Midblock #42
Midblock #43
Midblock #44
Midblock #45
Midblock #46
Midblock #47
Midblock #48
Midblock #49
Midblock #50
Midblock #51
Midblock #52
Midblock #53
Midblock #54
Midblock #55
Midblock #56
Midblock #57
Midblock #58
Midblock #59
Midblock #60
Mtn. View
Mean
CO No.
cone. of
(ppm) obs .
Palo Alto
Mean
CO No.
cone. of
(ppm) obs.
1.0 2
4.0 2
0.5 4
0.5 2
2.5 2
0.0 2
3.0 2
6.5 2
1.5 12
2.0 2
2.3 4
0.0 2
3.0 1
2.0 1
5.0 1
3.0 1
4.0 1
3.5 2
0.0 1
0.0 1
0.0 4
1.0 1
0.0 1
1.0 1
Chinatown-
Financial*
San Francisco
Mean
CO No.
cone. of
(ppm) obs.
Union Square,
San Francisco
Mean
CO No.
cone. of
(ppm) obs.
2.0 4
1.5 2
9.0 2
3.3 4
2.0 2
1.0 2
4.5 2
5.5 2
3.0 2
9.0 2
5.5 2
2.0 2
4.0 4
8.5 2
3.0 4
5.0 2
0.5 2
4.8 8
3.2 6
2.5 2
4.0 2
4.0 4
4.5 2
7.0 2
Westwood,
Los Angeles
Mean
CO No.
cone. of
(ppm) obs.
(continued)
-------
TABLE C-4 (continued)
Type of
commercial
setting
Midblock #61
Midblock #62
Midblock #63
Midblock #64
Midblock #65
Midblock #66
Midblock #67
Midblock #68
Midblock #69
Midblock #70
Midblock #71
Midblock #72
Midblock #73
Midblock #74
Midblock #75
Midblock #76
Midblock #77
Midblock #78
Midblock #79
Midblock #80
Midblock #81
Midblock #82
Midblock #83
Midblock #84
Midblock #85
Midblock #86
Mtn. View
Mean
CO No.
cone. of
(ppm) obs.
Palo Alto
Mean
CO No,
cone. of
(ppm) obs.
3.0 1
0.0 1
0.0 4
0.0 4
0.5 4
7.8 4
Chinatown-
Financial,
San Francisco
Mean
CO No.
cone. of
(ppm) obs .
Union Square,
San Francisco
Mean
CO No.
cone. of
(ppm) obs.
2.3 3
0.7 3
2.3 3
4.3 3
7.7 3
1.7 3
1.3 3
2.0 3
0.7 3
4.3 3
3.0 3
8.0 3
1.7 3
1.3 3
4.3 3
5.7 3
4.3 3
3.7 3
6.0 3
4.3 3
3.0 3
2.0 3
2.7 3
2.7 3
0.7 3
0.0 3
Westwood,
Los Angeles
Mean
Co No.
cone. of
(ppm) obs.
(continued)
-------
TABLE C-4 (continued)
Type of
commercial
setting
Midblock #87
Midblock #88
Midblock #89
Midblock #90
Midblock #91
Midblock #92
Midblock #93
Midblock #94
Midblock #95
Midblock #96
Midblock #97
Midblock #98
Midblock #99
Midblock #100
Midblock #101
Midblock #102
Midblock #103
Midblock #104
Midblock #105
Midblock #106
Midblock #107
Midblock #108
Midblock #109
Midblock #110
Midblock #111
Midblock #112
Mtn. View
Mean
CO No.
cone. of
(ppm) obs.
Palo Alto
Mean
CO No.
cone. of
(ppm) obs.
Chinatown-
Financial,-
San Francisco
Mean
CO No.
cone. of
(ppm) obs.
Union Square,
San Francisco
Mean
CO No.
cone. of
(ppm) obs .
3.7 3
2.0 3
7.0 3
3.3 3
6.8 3
1.7 3
6.7 3
2.3 3
5.0 3
6.3 3
16.7 3
1.7 3
4.0 3
2.7 3
0.0 3
0.3 3
4.0 3
5.7 3
6.3 3
2.3 3
0.0 3
2.0 3
4.0 3
0.0 3
0.0 3
Westwood,
Los Angeles
Mean
CO No.
cone. of
(ppm) obs.
to
CD
(continued)
-------
TABLE C-4 (continued)
Type of
commercial
setting
Midblock #113
Midblock #114
Midblock #115
Midblock #116
Midblock #117
Midblock #118
Midblock #119
Midblock #120
Midblock #121
Midblock #122
Midblock #123
Midblock #124
Midblock #125
Midblock #126
Midblock #127
Midblock #128
Midblock #129
Midblock #130
Midblock #131
Midblock #132
Midblock #133
Midblock #134
Midblock #135
Midblock #136
Midblock #137
Midblock #138
Mtn. View
Mean
CO No.
cone. of
(ppm) obs.
Palo Alto
Mean
CO No.
cone. of
(ppm) obs.
Chinatown-
Financial,
San Francisco
Mean
CO No.
cone. of
(ppm) obs.
Union Square.,
San Francisco
Mean
CO No.
cone. of
(ppm) obs.
3.7 3
6.3 3
3.7 3
3.0 3
2.0 3
9.0 3
2.7 3
1.3 3
1.7 2
2.0 3
3.0 1
0.3 3
1.0 3
1.0 3
1.0 3
0.3 3
1.0 3
0.0 3
1.7 3
2.3 3
3.3 6
7.0 3
2.7 3
9.3 3
3.6 3
5.3 3
Westwood,
Los Angeles
Mean
Co No.
cone. of
(ppm) obs.
(continued)
-------
TABLE C-4 (continued)
Type of
commercial
setting
Midblock #139
Midblock #140
Midblock #141
Midblock #142
Midblock #143
Midblock #144
Midblock #145
Midblock #146
Midblock #147
Midblock #148
Midblock #149
Midblock #150
Midblock #151
Midblock #152
Midblock #153
Midblock #154
Midblock #155
Park
Parking lot #1
Parking lot #2
Parking lot #3
Mtn. View
Mean
CO No.
cone. of
(ppm) obs.
0.0 2
0.0 1
Palo Alto
Mean
CO No.
cone. of
(ppm) obs.
1.0 2
1.5 2
6.7 6
Chinatown-
Financial,
San Francisco
Mean
CO No.
cone. of
(ppm) obs.
Union Square,
San Francisco
Mean
CO No.
cone. of
(ppm) obs.
2.7 3
6.7 3
2.6 3
2.0 3
1.7 3
2.3 3
3.3 3
1.3 3
0.3 3
2.3 3
2.3 3
2.0 3
1.3 3
4.3 3
2.7 3
3.3 3
1.0 3
Westwood,
Los Angeles
Mean
CO No.
cone. of
(ppm) obs.
-o
o
(continued)
-------
TABLE C-4 (continued)
Type of
commercial
setting
Plaza //I
Plaza #2
Plaza #3
Plaza #4
Plaza #5
Plaza #6
Plaza #7
Plaza #8
Plaza #9
Plaza #10
Mtn. View
Mean
CO No.
cone. of
(ppm) obs.
Palo Alto
Mean
CO No.
cone. of
(ppm) obs.
Chinatown-
Financial ,
San Francisco
Mean
CO No.
cone. of
(ppm) obs.
1.0 3
1.0 8
0.1 12
0.2 8
1.0 1
2.0 2
2.0 2
0.0 8
0.0 14
0.0 4
Union Square,
San Francisco
Mean
CO No.
cone. of
(ppm) obs.
4.5 24
Westwood,
Los Angeles
Mean
CO No.
cone. of
(ppm) obs.
-------
APPENDIX D
CUMULATIVE FREQUENCY DISTRIBUTIONS OF CO CONCENTRATIONS
This appendix contains five figures showing cumulative frequency
distributions as plotted on log-probability paper:
Figure D-l. Cumulative frequencies of CO concentrations for all
commercial settings, except parking garages, in
downtown Palo Alto on five dates.
Figure D-2. Cumulative frequencies of CO concentrations for
indoor and outdoor settings near Union Square in San
Francisco on 13 June 1980.
Figure D-3 Cumulative frequencies of CO concentrations for eight
types of indoor commercial settings near Union Square
in San Francisco on 13 June 1980.
Figure D-4 Cumulative frequencies of CO concentrations for eight
city blocks in the Union Square district of San
Francisco on 13 June 1980.
Figure D-5. Cumulative frequencies of CO concentrations for
selected indoor commercial settings at five
geographic locations.
142
-------
CUMULATIVE FREQUENCY
K-E
PROBABILITY X 2 LOG CYCLES
100
001 0.05 O.I 0.2 0.5 1 2 5 10 20 30 40 50 60 70 80 90 95 98 99 99.5 99.8 99.9
99.99
100
z
g
u
z
o
u
1-44-4-
24 JANUARY 1980
(n = 58)
< 31 JANUARY 1980
(n = 33)
0.8 ~
0.6
0.5 ^44
0.4
0.3
0.2
0.1
0.01 0.05 0.1 0.2 0.5 1 2 5 10 20 30 M 50 60 70 80 90 95 98 99 99.5 99.8 99.9
CUMULATIVE FREQUENCY (%)
FIGURE D-1. CUMULATIVE FREQUENCIES OF CO CONCENTRATIONS FOR ALL COMMERCIAL SETTINGS, EXCEPT
PARKING GARAGES, IN DOWNTOWN PALO ALTO ON FIVE DATES.
0.1
143
-------
CONSTRUCTED BY THE AUTHORS FROM PAPERS OF:
\j r PROBABILITY X 2 LOG CYCLES
CUMULATIVE FREQUENCY "»"^ os"
001 005 01 02 05 1 2 5 10 20 30 40 50 60 70 80 90 95 98 99 99.5 99.899.9
100
70
60
50
40
20
10
9
8
7
E 6
£ 5
O 4
H
<
£ 3
z
bU
U
Z
O
o 2
1
0.9
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
- - -
-- - -.
-: -.-- '
--- _
a^-.^i i
- -f
-
i i- '
i
:
1 \ '
t '
I
j \
\
\ :
' ' t '
> t
* i
!
; i
1 4-
j
;
- J
..__
:
: -- - I
- = j=^':
- T
:
M i^
','. \
; i
II
: i : 1 i
|
: 1 i
;
[
i * ^
;: i , i
' ' t
.- . i | .
- - t <
,. i i .
4i - 4f
. j
'i i i
-U 1
nl T T
i: -:
U: -
1
- }
- - 4
-
-
,
' VH^
i '
i: : i
>'
i ,, i
M i i
1 : I i :
1 :l; i i
i : : ! : :
! :: :
i , . > i
i ; :
I ': i
| j i
{
: i
! ' ; i
t i i -
I '
' i .
! 14
1 -
- i\
. 11.
_:_ ..,;,: ,_,.:: ...I
jt . - - T t
I J:
: - T
[
1 i ' 1 i
: 111
': ::';:'; : : :::
"^"r|i::ii:]i §
: ::: ;!:: : :: :::
|j|- | !
; . ! ' .i'M1 i : i : . ill!
I i ! :! :!' ' : ! i : ! i i I i I ! !
i j ': III III II ] II' III1
' 1 > ' ,*J
i ; :i ii! iiii :; ;i:^
I : ; :!. ; I:::::,:. '!#'ii i
^£j;%|il ^!^
^ i ::i i it : '
" \ " \\ \
:::: :::::] : . __ : :
^^: El Jill;--, -
| j_
t ' -1 '
1 '
1 1 ;
1 if -: - - - : - --_;:
- - i i i - - . -- - .:-_- . . --
1 ! 1 . .:
1 : 1 ! IJ! ] -:- ±^r '
; .:.-- -^=. : .
.'.'... .-- ;=' "--
i - : - '
|i-:;. -if. II. .-_-|_ - -"[-,.
rffl \m\ ililfli IS- H iil
-C:EE-rE=f::;;:::::EE --^ ----
----
:E~- ^ ;.=
:-- : -r --
- '- ' .' -
- _
---3
..__ -- - -
'- -- -- -
F -:-- :£-
] f ~ jj
)UTDOOR SET
n = 88)
i
)OR SETTING
M DOORS n =
i 1 H!1
! 1 I'1
3R SETTINGS,
ED DOORS n
;_; ;;;
. : : .' '..':'- -J-
:C.- ~~-----~
:"_-:.:^-^
;;|: ^--- ;
;.:E;;;: :
: = =- = = -^= :
- -..- = == ::
. _ _ _ . _ . .
T- ~ r -
-. "= IT ZZI : :
. ; -~.EZE I
1^: --- .- ;
..-. :---"- -
99.99
..... . . 100
-- _
. :
_ . _ __
-^ =
-: : - ^~
::::::= ~
'-"'- >r
TINGS
5 _l£j
44) _^
111
= 37) JZ
- 1 T-"
; i _- :
:^^
~EE
: ; i._
.____
: -;^: -
' ~ 7-
- :~r=-
- = --
. ._ E -.;
-.- - =
::---.
90
80
70
60
50
40
20
10
9
8
7
6
5
t
3
2
1
0.9
0.8
0.7
0.6
0.5
0.4
0.3
0.2
n i
0.01 0.05 0.1 0.2 0.5 1 2
20 30 40 50 60 70 80 90 95 98 99 99.5 99.8 !
CUMULATIVE FREQUENCY (%)
FIGURE D-2. CUMULATIVE FREQUENCIES OF CO CONCENTRATIONS FOR INDOOR AND OUTDOOR SETTINGS NEAR
UNION SQUARE IN SAN FRANCISCO ON 13 JUNE 1980.
144
-------
0.01 0.05 O.I 0.2 0.5
CUMULATIVE FREQUENCY
20 30 40 50 60 70 80
CONSTRUCTED 6Y THE AUTHORS FROM PAPERS OF:
W*£ PROBABILITY X 2 LOG CYCLES
98 99 99.5 99.8 99.9 99.99
100
OFFICES n = 8)
RESTAURANTS (n = 11)
MISCELLANEOUS
DEPARTMENT STORES
HOME FURNISHING STORES (n = 8)
HOTELS (n = 11)
SERVICE CENTERS (n = 9)
CLOTHING STORES (n = 17
0.1
0.01 0.05 0.1 0.2 0.5 1 2 5 10
20 30 40 50 60 70 80 90 95 98 99 99.5 99.8 99.9 99.99
CUMULATIVE FREQUENCY (%)
FIGURE D-3. CUMULATIVE FREQUENCIES OF CO CONCENTRATIONS FOR EIGHT TYPES OF INDOOR COMMERCIAL
SETTINGS NEAR UNION SQUARE IN SAN FRANCISCO ON 13 JUNE 1980.
145
-------
CONSTRUCTED BY THE AUTHORS FROM PAPERS OF
L/-.C PROBABILITY X 2 LOG CYCLES
0
100
90
80
70
60
50
40
30
20
9
8
7
£
O. c
a. 0
g 4
H
H 3
LJJ
(J
Z
o
O 2
1
0.9
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
CUMULATIVE FREQUENCY "currtL " tsst " "'
D1 0.050.1 0.2 0.5 2 5 10 20 30 40 50 60 70 80 90 95 98 99 99.5 99.899.9 99'9910n
I
-'- -
mm-.
^^: -_-_
_=.--.
it .
----
^
- A '
-j ;
-
I
I
'-',
=3 "
71
1 4
tl
i,i
4 --.
EE3
- '.-":-'
"~"1
r-'fj
^= ^:
;"_ -
-i-.:.;
f-=£: .-
1
TFT
--H 1-
i
i; :
|-:
-:.
i j
:
i 1
1 > I
' \
. ill
I'll
jji
i!! ;
j
t
It
', I
, h
t
T
tf[
i' t -
H
-|
|
f
i i !
i
-
--T
;;
| ;
:i
;!
i
i :
i
I
! i:
i : :
i ; ; :
^_ T
- If
-r -f
^-|
: 1 ;' ! j ji III i : ! ! !
ii; : i i !! ill! ! ; ! i i
: |
h - -^--fT-Hft-Tttt-l
- - - f -- ^ "t -"trt- trtt "T"
1 . - -. - -;-=- : : : i i : ; : i
i.
1 i i !;i i ;i i
: : : | : : i : . ...
I : ! ; i ! : : i : , : ; i ; : i ;
1 f ' : ]
; i ; l ' : ; : : : : ! ; i i : ' j ; ' : ' . :
i : : | . : : . : ' !: : ; : i ; : : i : :
: :; : ' ' : ! ;;::.''! ; | ; . :
;:;. BLOCK 2|:"'!j!:' 1 ' i'M I''!1!-1
: ::; (n = 9) i i :' i : :::' :i
' ;:: : ' : '.. : "'' ^
; ; , y*f ' ' c n"
':::: . : :::f::,:' :."/.:/p.
......,: . / 1
.,,, ^:, /' ' ' f
.::: .:: y.\'. :^ti.\:^
".,'." '"' f ;>/
J ,. . y-- /'
- i : ; ; s : , -ft * | : \ : : i ; :
t - i 1 ' ' t4 f' ' ' -H-i ' ' ' ^r^- 'i"i ' '
i i i i : ;i ::. iif i/ii iii: iii
|
j
,_ _,_+..+.... j.fU-.-H-dL
I : i i4i^__i|_ : |,l
LTIII.iIt..LIi:lii G
i! i U i ii i i
14.
--
tfl
' ' ir'-rn ' ill;
::i;*l::ii:
1
t ]_
1 !:: .. :
r!^|ijttHTrtt|Ttti " :^~
I'/' i:!i iii; iiii iiii iii i i i
III...,,
",'.','], ' ^RLOCK
: i !
15 r
rH I i
|
= 9)
:' : ;.: :: i ; :: : : ..-MBLOCK 4,{n = 11)
; . ; . /I J-t- BLC
. , ; ; __ j -f r T ^rrl" f :"f ', ~ . ^
)CK
11 (
CK 8 (n
CK 1 (n
ini '
- P)
]\\^f^rf-.\l :,/! ! i ^ BLOCK 12 (n = 16)
. : -lip '"' '''',' '
)CI<
" ^ ' jf af"
[,.. .:: " y. .:::: :: :
9! !
:!i: : i : : : ! :::: pif : ,
ii:i :':: iiii iiii iii!
'ttC-I-.-iirii "~ ""
1 44- ii i i i i : : i M i ; _|
I j t :!:' I j
r Hil- " Iii ii )
it t . : - Y . '
r
19 (
\ j
' J
0.01 0.05 0.1 0.2 0.5 1 2 5 10 20 30 40 50 60 70 80 90 95 98 99
i t
[-
; I
>-
M
..-
r
- :
--
-I 4-
::
L
iU
!i
71
H'
--
_
--.-.
H f"
-:- r
- --^
- --
-
-_
:_- ::
99.5 99.8 99.9
'
t-TT
f- J 1
i :
i I ;
1 ' '
I j
- . j
f ' M
T
: "
: :-
-
*r
-
99
90
80
70
60
50
40
30
20
10
9
8
7
6
5
1 0
(-
3 K
Z
LJJ
O
o
2 0
1
0.9
0.8
0.7
0.6
0.5
0.4
0.3
02
0.1
99
CUMULATIVE FREQUENCY (%)
FIGURE D-4. CUMULATIVE FREQUENCIES OF CO CONCENTRATIONS FOR EIGHT CITY BLOCKS IN
SQUARE DISTRICT OF SAN FRANCISCO ON 13 JUNE 1980.
THE UNION
146
-------
100
90
80
70
60
50
40
30
20
0"1 0.05 0.1 0.2 0.5 1 2
CUMULATIVE FREQUENCY
10 20 30 40 50 60 70 80
CONSTRUCTED BY THE AUTHORS FROM PAPERS OF
K*£ PROBABILITY X 1 LOG CYCLES
90 95 98 99 99.5 99.8 99.9 99.99
10
8
1
~ 6
I 5
O 4
t-
<
cc
i- 3
z
LU
u
z
o
O J
0.9
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
m
-1
n
ii
to
H-
:IFPT
i
rtt
UNION SQUARE AREA,
SAN FRANCISCO (n = 23)
(11 April 1980)
,
/PALO ALTO (n = 24)
. (24 Jan 1980)
-LOS ANGELES (n = 19)
(27 March 1980)
:MOUNTAIN VIEW (n
(13 March 1980)
= 11)
CHINATOWN-FINANCIAL DISTRICT,
SAN FRANCISCO (n = 18)
(11 April 1980)
j
Efctt
Tt"
100
90
80
70
60
50
40
30
20
10
9
8
7
6
LLJ
CJ
O
2 o
0.9
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0.01 0.05 0.1 0.2 0.5 1 2 5 10 20 30 40 50 60 70 80 90 95 9J 99 99.5 99.8 99.9 99.99
CUMULATIVE FREQUENCY (%)
FIGURE D-5 CUMULATIVE FREQUENCIES OF CO CONCENTRATIONS FOR SELECTED INDOOR COMMERICAL SETTINGS
AT FIVE GEOGRAPHIC LOCATIONS.
147
*D.S. GOVEMMEBT PRIHTIHa oyWOE : 1984 0-421-082/51Z
-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse bejore completing)
REPORT NO.
EPA-600/4-84-019
3. RECIPIENT'S ACCESSION NO.
TITLE AND SUBTITLE
FIELD SURVEYS OF CARBON MONOXIDE IN COMMERCIAL
SETTINGS' USING PERSONAL EXPOSURE MONITORS
5. REPORT DATE
February 1984
6. PERFORMING ORGANIZATION CODE
AUTHOR(S)
Peter G. Flachsbart and Wayne R. Ott
8. PERFORMING ORGANIZATION REPORT NO.
PERFORMING ORGANIZATION NAME AND ADDRESS
Department of Urban and Regional Planning
University of Hawaii at Manoa
Honolulu, Hawaii 96822
10. PROGRAM ELEMENT NO.
A114
11. CONTRACT/GRANT NO.
A-1598-NNLX
12. SPONSORING AGENCY NAME AND ADDRESS
Office of Mobile Source Air Pollution Control
Office of Research and Development
U.S. Environmental Protection Agency
Washington, D.C. 20460
13. TYPE OF RE PORT AND PERIOD COVERED
Final
14. SPONSORING AGENCY CODE
15. SUPPLEMENTARY NOTES
16. ABSTRACT
This study employed miniaturized personal exposure monitors (PEMs) to measure car-
bon monoxide (CO) in 588 different commercial settings (e.g., retail stores, office
buildings, hotels, restaurants) in five California cities. Altogether, 5000 CO obser-
vations were made by recording the instantaneous instrument readings at 1-minute inter-
vals as the investigators walked along sidewalks and into buildings. For 11 of 15 sur-
vey dates, two investigators walked side-by-side, permitting two adjacent PEMs to be
compared. Quality assurance tests for 1706 pairs of values showed a very high degree
of agreement.
CO levels for indoor commercial settings were similar to those measured outdoors on
sidewalks, apparently because the pollutant seeps into the structures from traffic out-
side. Although indoor levels usually were above 0 ppm, they seldom were above 9 ppm
(the National Ambient Air Quality Standard for an 8-hour exposure), unless some indoor
source (e.g., enclosed parking garage) could be identified. For example, an office
building with high CO levels from its garage was "hot" in the sense that CO permeated
the upper floors, exposing many office workers to concentrations above 9 ppm, well
above ambient levels outside. Indoor settings, without their own sources of CO, were
sufficiently similar in concentrations to be treated as a class, although levels did
vary slightly from date to date. CO levels on outdoor streets did not vary greatly on
different sides of the street, on corners and faces of blocks, and intersections.
17.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b.IDENTIFIERS/OPEN ENDED TERMS
COS AT I Field/Group
Carbon Monoxide
Personal Monitors
Indoor Air Quality
Exposure Assessments
Field Surveys
18. DISTRIBUTION STATEMENT
Release unlimited
19. SECURITY CLASS (This Report)
Unclassified
162
20. SECURITY CLASS (This page)
Unclassified
22. PRICE
EPA Form 2220-1 (9-73)
------- |