ESTIMATION OF OZONE EXPOSURES
EXPERIENCED BY OUTDOOR CHILDREN IN
NINE URBAN AREAS USING A
PROBABILISTIC VERSION OF NEM
by
Ted Johnson, Jim Capel, Jill Warnasch Mozier, and Mike McCoy
International Technology Air Quality Services
South Square Corporate Centre One
3710 University Drive, Suite 201
Durham, North Carolina 27707-6208
Contract No. 63-D-30094
Work Assignment No. 0-2
JTN 453212-4
Harvey Richmond, Work Assignment Manager
Nancy Riley, Project Manager
Prepared for
U.S. ENVIRONMENTAL PROTECTION AGENCY
OFFICE OF AIR QUALITY PLANNING AND STANDARDS
RESEARCH TRIANGLE PARK, NORTH CAROLINA 27711
April 1996
-------
CONTENTS
Figures v
Tables vi
Acknowledgment xiv
1. Introduction 1
2. Overview of the Methodology 5
Define study area, population-of-interest,
subdivisions of study area, and exposure period 5
Divide the population-of-interest into an exhaustive
set of cohorts 7
Develop an exposure event sequence for each cohort for
the exposure period 9
Estimate the pollutant concentration and ventilation
rate associated with each exposure event 14
Extrapolate the cohort exposures to the population-of-interest 27
3. The Mass-Balance Model 33
Theoretical basis and assumptions 34
Simulation of microenvironmental ozone concentrations 41
Air exchange rate distributions 44
Window status algorithm 47
4. Preparation of Air Quality Data 51
Selection of representative data sets 51
Treatment of missing values and descriptive statistics 52
5. Adjustment of Ozone Data to Simulate Compliance with Alternative
Air Quality Standards 79
Specification of AQI and estimation of baseline AQI values 80
Estimation of AQI's under attainment conditions 87
Adjustment of one-hour ozone data sets 91
Application of the AQAP's to Philadelphia 94
Special adjustment procedures applied in selected
attainment scenarios 102
HI
-------
CONTENTS (continued)
6. Preparation of Outdoor Children Data Bases
Selection of. time/activity data 1°4
Processing of time/activity data ^
City-specific outdoor children populations 122
7. Ozone Exposure Estimates for Nine Urban Areas 125
Regulatory scenarios 125
Formats of the exposure summary tables 126
Results of analyses 128
Estimates of maximum dose exposures 149
8. Principal Limitations of the pNEM/O3 Methodology 194
Time/activity patterns 195
Equivalent ventilation rates 196
The air quality adjustment procedures 198
Estimation of cohort populations 200
The mass balance model 201
Estimation of ozone exposures for special scenario
associated with attainment of 8H5EX-80
Standard 202
References 205
Appendices
A. Ten Time/Activity Bases Generally Applicable to Air
Pollution Exposure Assessments A-1
B. Monte Carlo Models for Generating Event EVR Values B-1
C. Testing of Monte Carlo Models C-1
D. Sample Output of pNEM/03 Applied to Outdoor
Children (Houston, 1-Hour, Daily Maximum 0.12
ppm Standard [Current NAAQS]) D-1
E. One-Hour Exposure Distributions E-1
F. Estimation of Ozone Exposures in Outdoor Children for
Special 8H10EX-80 Scenario F-1
IV
-------
FIGURES
Number page
1 Page From the Activity Diary Used in the Cincinnati Study 11
2a Eight-Hour Daily Maximum Dose Exposure Distributions for
Outdoor Children Exposed on One or More Days Under
Moderate Exertion (EVR 13-27 Liters/Min-M2) in
Philadelphia, PA 189
2b Eight-Hour Daily Maximum Dose Exposure Distributions of
Total Occurrences for Outdoor Children Exposure Under
Moderate Exertion (EVR 13-27 Liters/Min-M2) in
Philadelphia, PA 189
3a Eight-Hour Daily Maximum Dose Exposure Distributions for
Outdoor Children Exposed on One or More Days Under
Moderate Exertion (EVR 13-27 Liters/Min-M2) in
Houston, TX 190
3b Eight-Hour Daily Maximum Dose Exposure Distributions of
Total Occurrences for Outdoor Children Exposure Under
Moderate Exertion (EVR 13-27 Liters/Min-M2) in
Houston, TX 190
4a Eight-Hour Daily Maximum Dose Exposure Distributions for
Outdoor Children Exposed on One or More Days Under
Moderate Exertion (EVR 13-27 Liters/Min-M2) in
New York, NY 191
4b Eight-Hour Daily Maximum Dose Exposure Distributions of
Total Occurrences for Outdoor Children Exposure Under
Moderate Exertion (EVR 13-27 Liters/Min-M2) in
New York, NY 191
5a Eight-Hour Daily Maximum Dose Exposure Distributions for
Outdoor Children Exposed on One or More Days Under
Moderate Exertion (EVR 13-27 Liters/Min-M2) in
Washington, D.C. 192
5b Eight-Hour Daily Maximum Dose Exposure Distributions of
Total Occurrences for Outdoor Children Exposure Under
Moderate Exertion (EVR 13-27 Liters/Min-M2) in
Washington, D.C. 192
-------
TABLES
Number
1 Characteristics of Study Areas 7
2 Characteristics of Studies Providing Time/Activity Data
for Outdoor Children 12
3 Parameters Associated with Algorithms Used to Estimate
Ozone Concentrations in Microenvironments 16
4 Algorithm Used to Generate Event-Specific Values of
Equivalent Ventilation Rate 22
5 Algorithm for Determining Upper Limit for EVR 25
6 Parameter Values for Algorithm Used to Determine Limits for
Equivalent Ventilation Rates for Outdoor Children 26
7 Population Estimates by Demographic Group and Air
Conditioning Status 30
8 Means, Standard Deviations, and Confidence Intervals
for Estimates of kd(AA/) Provided by Weschler 40
9 Distributions of Air Exchange Rate Values Used in the
pNEM/03 Mass Balance Model 44
10 Probability of Window Status for Day by Air Conditioning
System and Temperature Range 49
11 Probability of Windows Being Open by Clock Hour, Temperature
Range, and Window Status of Preceding Hour (PH) for Residences
With Central Air Conditioning 49
VI
-------
TABLES (continued)
Number
12 Probability of Windows Being Open by Clock Hour, Temperature
Range, and Window Status of Preceding Hour (PH) for Residences
With Window Air Conditioning Units 50
13 Probability of Windows Being Open by Clock Hour, Temperature
Range, and Window Status of Preceding Hour (PH) for Residences
With No Air Conditioning System 50
14 Characteristics of Ozone Study Areas and Monitoring Sites 53
15 Descriptive Statistics for 1991 Data Sets Containing
Hourly-Average Ozone Concentrations Obtained From
Selected Monitoring Sites in the Chicago Study Area 54
16 Descriptive Statistics for 1990 Data Sets Containing
Hourly-Average Ozone Concentrations Obtained From
Selected Monitoring Sites in the Denver Study Area 56
17 Descriptive Statistics for 1990 Data Sets Containing
Hourly-Average Ozone Concentrations Obtained From
Selected Monitoring Sites in the Houston Study Area 57
18 Descriptive Statistics for 1991 Data Sets Containing
Hourly-Average Ozone Concentrations Obtained From
Selected Monitoring Sites in the Los Angeles Study Area 59
19 Descriptive Statistics for 1991 Data Sets Containing
Hourly-Average Ozone Concentrations Obtained From
Selected Monitoring Sites in the Miami Study Area 61
20 Descriptive Statistics for 1991 Data Sets Containing
Hourly-Average Ozone Concentrations Obtained From
Selected Monitoring Sites in the New York Study Area 62
21 Descriptive Statistics for 1991 Data Sets Containing
Hourly-Average Ozone Concentrations Obtained From
Selected Monitoring Sites in the Philadelphia Study Area 64
VII
-------
TABLES (continued)
Number
22 Descriptive Statistics for 1990 Data Sets Containing
Hourly-Average Ozone Concentrations Obtained From
Selected Monitoring Sites in the St. Louis Study Area 66
23 Descriptive Statistics for 1991 Data Sets Containing
Hourly-Average Ozone Concentrations Obtained From
Selected Monitoring Sites .in the Washington Study Area 68
24 Descriptive Statistics for 1991 Data Sets Containing
Eight-Hour Ozone Concentrations Obtained From
Selected Monitoring Sites in the Chicago Study Area 70
25 Descriptive Statistics for 1990 Data Sets Containing
Eight-Hour Ozone Concentrations Obtained From
Selected Monitoring Sites in the Denver Study Area 71
26 Descriptive Statistics for 1990 Data Sets Containing
Eight-Hour Ozone Concentrations Obtained From
Selected Monitoring Sites in the Houston Study Area 72
27 Descriptive Statistics for 1991 Data Sets Containing
Eight-Hour Ozone Concentrations Obtained From
Selected Monitoring Sites in the Los Angeles Study Area 73
28 Descriptive Statistics for 1991 Data Sets Containing
Eight-Hour Ozone Concentrations Obtained From
Selected Monitoring Sites in the Miami Study Area 74
29 Descriptive Statistics for 1991 Data Sets Containing
Eight-Hour Ozone Concentrations Obtained From
Selected Monitoring Sites in the New York Study Area 75
30 Descriptive Statistics for 1991 Data Sets Containing
Eight-Hour Ozone Concentrations Obtained From
Selected Monitoring Sites in the Philadelphia Study Area 76
VIII
-------
TABLES (continued)
Number
31 Descriptive Statistics for 1990 Data Sets Containing
Eight-Hour Ozone Concentrations Obtained From
Selected Monitoring Sites in the St. Louis Study Area 77
32 Descriptive Statistics for 1991 Data Sets Containing
Eight-Hour Ozone Concentrations Obtained From
Selected Monitoring Sites in the Washington Study Area 78
33 Baseline Air Quality Indicators for Nine Cities 84
34 Air Quality Adjustment Procedure Used to Simulate Attainment
of 1H1EX NAAQS (The Expected Number of Daily Maximum
One-Hour Ozone Concentrations Exceeding the Specified Value
Shall Not Exceed One) 88
35 Air Quality Adjustment Procedure Used to Simulate Attainment
of 8H1EX NAAQS (The Expected Number of Daily Maximum
Eight-Hour Ozone Concentrations Exceeding the Specified Value
Shall Not Exceed One) 89
36 Air Quality Adjustment Procedure Used to Simulate Attainment
of 8H5EX NAAQS (The Expected Number of Daily Maximum
Eight-Hour Ozone Concentrations Exceeding the Specified Value
Shall Not Exceed Five) 90
37 Values for Equivalence Relationships 93
38 Determination of Adjustment Coefficients for One-Hour
NAAQS Attainment (1H1EX-120) in Philadelphia 95
39 Descriptive Statistics for Hourly-Hour Data (ppb) for Site
34-005-3001 (District 1, Philadelphia): Baseline and Attainment
of Three Ozone Standards 97
40 Determination of Adjustment Coefficients for Eight-Hour
NAAQS Attainment (8H1EX-80) in Philadelphia 98
41 Determination of Adjustment Coefficients for Eight-Hour
NAAQS Attainment (EH6LDM = 80 ppb) in Philadelphia 1Q1
IX
-------
TABLES (continued)
Number
42 Characteristics of Activity Data for Outdoor Children 106
43 Breathing Rate Categories of Activities in the Cincinnati
Study 1°9
44 Cumulative Breathing Rate Category Probabilities
From the Cincinnati Activity-Diary Study by
Activity Class, Microenvironment, Time of Day Category,
and Event Duration Category 112
45 Activity Classes Assigned to Activity Codes Used in
the California Diary Study 116
46 Activity Classes Assigned to Activity Codes Used in
the Denver Diary Study 119
47 Activity Classes Assigned to Activity Codes Used in
the Valdez Diary Study 120
48 Activity Classes Assigned to Activity Codes Used in
the Washington Diary Study 121
49 Estimated Number of Outdoor Children in Each Study Area 124
50 Number and Percent of Outdoor Children Experiencing
One or More One-Hour Daily Maximum Ozone Exposures
Above 120 ppb at any Ventilation Rate 129
51 Number and Percent of Outdoor Children Experiencing
One or More Eight-Hour Daily Maximum Ozone Exposures
Above 60 ppb at any Ventilation Rate 134
52 Number and Percent of Outdoor Children Experiencing
One or More Eight-Hour Daily Maximum Ozone Exposures
Above 80 ppb at any Ventilation Rate 139
53 Number and Percent of Outdoor Children Experiencing
One or More Eight-Hour Daily Maximum Ozone Exposures
Above 100 ppb at any Ventilation Rate 144
-------
TABLES (continued)
Number Page
54a,b Estimates of One-Hour Maximum Dosage Exposures
Experienced by Outdoor Children in Chicago During
Which Ozone Concentration Exceeded 0.12 ppm and
EVR Equaled or Exceeded 30 Liters Min"1- M'2 150-151
55a,b Estimates of Eight-Hour Maximum Dosage Exposures
Experienced by Outdoor Children in Chicago During
Which Ozone Concentration Exceeded 0.08 ppm and
EVR Ranged From 13 Liters Min"1 M"2 to 27
Liters -Min"1-M"2 152-153
56a,b Estimates of One-Hour Maximum Dosage Exposures
Experienced by Outdoor Children in Denver During
Which Ozone Concentration Exceeded 0.12 ppm and
EVR Equaled or Exceeded 30 Liters Min'1- M'2 154-155
57a,b Estimates of Eight-Hour Maximum Dosage Exposures
Experienced by Outdoor Children in Denver During
Which Ozone Concentration Exceeded 0.08 ppm and
EVR Ranged From 13 Liters Min"1- M'2 to 27
Liters -Min^-M'2 156-157
58a,b Estimates of One-Hour Maximum Dosage Exposures
Experienced by Outdoor Children in Houston During
Which Ozone Concentration Exceeded 0.12 ppm and
EVR Equaled or Exceeded 30 Liters Min"1- M"2 158-159
59a,b Estimates of Eight-Hour Maximum Dosage Exposures
Experienced by Outdoor Children in Houston During
Which Ozone Concentration Exceeded 0.08 ppm and
EVR Ranged From 13 Liters Min"1- M"2 to 27
Liters -Min"1-M"2 160-161
60a,b Estimates of One-Hour Maximum Dosage Exposures
Experienced by Outdoor Children in Los Angeles During
Which Ozone Concentration Exceeded 0.12 ppm and
EVR Equaled or Exceeded 30 Liters Min"1- M"2 162-163
XI
-------
TABLES (continued)
Number
61a,b Estimates of Eight-Hour Maximum Dosage Exposures
Experienced by Outdoor Children in Los Angeles During
Which Ozone Concentration Exceeded 0.08 ppm and
EVR Ranged From 13 Liters MhY1- M'2 to 27
Liters MnY1 M'2 164-165
62a,b Estimates of One-Hour Maximum Dosage Exposures
Experienced by Outdoor Children in Miami During
Which Ozone Concentration Exceeded 0.12 ppm and
EVR Equaled or Exceeded 30 Liters Min'1- M'2 166-167
63a,b Estimates of Eight-Hour Maximum Dosage Exposures
Experienced by Outdoor Children in Miami During
Which Ozone Concentration Exceeded 0.08 ppm and
EVR Ranged From 13 Liters Min'1- M'2 to 27
Liters -Min^-M'2 168-169
64a,b Estimates of One-Hour Maximum Dosage Exposures
Experienced by Outdoor Children in New York During
Which Ozone Concentration Exceeded 0.12 ppm and
EVR Equaled or Exceeded 30 Liters - Min'1- M'2 170-171
65a,b Estimates of Eight-Hour Maximum Dosage Exposures
Experienced by Outdoor Children in New York During
Which Ozone Concentration Exceeded 0.08 ppm and
EVR Ranged From 13 Liters Min'1- M'2 to 27
Liters Min'1-M'2 172-173
66a,b Estimates of One-Hour Maximum Dosage Exposures
Experienced by Outdoor Children in Philadelphia During
Which Ozone Concentration Exceeded 0.12 ppm and
EVR Equaled or Exceeded 30 Liters Min"1- M"2 174-175
67a,b Estimates of Eight-Hour Maximum Dosage Exposures
Experienced by Outdoor Children in Philadelphia During
Which Ozone Concentration Exceeded 0.08 ppm and
EVR Ranged From 13 Liters Min'1- M'2 to 27
Liters Min"1 M"2 176-177
XII
-------
TABLES (continued)
Number
68a,b Estimates of One-Hour Maximum Dosage Exposures
Experienced by Outdoor Children in St. Louis During
Which Ozone Concentration Exceeded 0.12 ppm and
EVR Equaled or Exceeded 30 Liters Min"1- M'2 178-179
69a,b Estimates of Eight-Hour Maximum Dosage Exposures
Experienced by Outdoor Children in St. Louis During
Which Ozone Concentration Exceeded 0.08 ppm and
EVR Ranged From 13 Liters - Min"1- M'2 to 27
Liters -Min"1-M"2 180-181
70a,b Estimates of One-Hour Maximum Dosage Exposures
Experienced by Outdoor Children in Washington, D.C.
During Which Ozone Concentration Exceeded 0.12
ppm and EVR Equaled or Exceeded 30 Liters Min"1- M"2 182-183
71a,b Estimates of Eight-Hour Maximum Dosage Exposures
Experienced by Outdoor Children in Washington, D.C.
During Which Ozone Concentration Exceeded 0.08
ppm and EVR Ranged From 13 Liters Min"1- M"2 to 27
Liters -Min"1-M"2 184-185
XIII
-------
ACKNOWLEDGMENT
In evaluating alternative National Ambient Air Quality Standards (NAAQS), the
U.S. Environmental Protection Agency (EPA) assesses the risks to human health of
air quality meeting each of the standards under consideration. This assessment of
risk requires estimates of the number of persons exposed at various pollutant
concentrations for specified periods of time. Since 1979, IT Air Quality Services
(ITAQS) has assisted EPA in developing various versions of the NAAQS Exposure
Model (NEM) to assist in this process. In 1993, ITAQS developed a probabilistic
version of NEM applicable to ozone (pNEM/O3) and applied it to the general
population residing in each of nine urban areas. In 1994, EPA directed ITAQS to
develop a special version of pNEM/03 applicable to outdoor children and to use it to
estimate the ozone exposures of outdoor children residing in the nine urban areas.
This report summarizes the results of this research effort.
The outdoor children project was managed by Mr. Mike McCoy of ITAQS with
technical direction provided by Mr. Ted Johnson. Mr. Jim Capel of ITAQS
developed the special version of pNEM/O3 and performed all computer runs of the
model. He also developed the input databases listing (1) time/activity data
representative or outdoor children and (2) estimates of the number of outdoor
children in each of the nine study areas.
Ms. Jill Warnasch Mozier and Mr. Jim Capel were the principal authors of
Section 6 and Subsection 8.4 of this report. Mr. Ted Johnson was the principal
author of the remaining sections. Ms. Joan Abernethy typed the report and created
many of the graphs in Section 7.
ITAQS' work on this project was funded under Work Assignment Nos. 0-2,
1-4, and 1-5 of EPA Contract No. 63-D-30094. Mr. Harvey Richmond served as'the
EPA Work Assignment Manager and provided guidance throughout the project. Mr.
Thomas McCurdy provided technical direction and guidance in the development of
pNEM/O3 through March 1994 and has provided technical review since that time.
Ms. Nancy Riley was the EPA Project Manager.
The authors would like to express their appreciation to Ms. Mary Anne
Simpson and Ms. Margaret Brill for their assistance with obtaining 1990 Census
Data at Perkins Library on Duke University campus.
XIV
-------
SECTION 1
INTRODUCTION
Within the U.S. Environmental Protection Agency (EPA), the Office of Air
Quality Planning and Standards (OAQPS) has responsibility for establishing and
revising national ambient air quality standards (NAAQS). In evaluating alternative
NAAQS proposed for a particular pollutant, OAQPS assesses the risks to human
health of air quality meeting each of the standards under consideration.1 This
assessment of risk requires estimates of the number of persons exposed at various
pollutant concentrations for specified periods of time. The estimates may be specific
to an urbanized area such as Los Angeles or apply to the entire nation.
Several researchers2'3 have recommended that such estimates be obtained
by simulating the movements of people through zones of varying air quality so as to
approximate the actual exposure patterns of people living within a defined area.
OAQPS has implemented this approach through an evolving methodology referred
to as the NAAQS Exposure Model (NEM). An early overview of the NEM
methodology is provided in a paper by Biller et al.4 From 1979 to 1988, IT Air
Quality'Services (formerly PEI Associates, Inc.) assisted OAQPS in developing and
applying pollutant-specific versions of NEM to ozone,5 particulate matter,6 and CO.7
These versions of NEM are referred to as "deterministic" versions in that no attempt
was made to model random processes within the exposure simulation.
The deterministic versions of NEM were similar in that each was capable of
simulating the movements of selected segments of an urban population through a
set of environmental settings. Each environmental setting was defined by a
geographic area and a microenvironment. The size and distribution of the
geographic areas were determined according to the ambient characteristics of the
pollutant. Ambient (outdoor) pollutant levels in each geographic area were
1
-------
estimated from either fixed-site monitoring data or dispersion model estimates. To
better utilize fixed-site monitoring data, researchers developed special time series
techniques to fill in missing values and special roll-back techniques to adjust the
monitoring data to simulate conditions under attainment of a particular NAAQS.
Additional details concerning the evolution of the deterministic version of NEM
are provided by Paul et al.8 Critiques of deterministic NEM are included in surveys
of exposure models by Pandian9 and Ryan.10 Two staff papers11'12 prepared by EPA
discuss the use of NEM in evaluating alternative NAAQS for CO and ozone.
In 1988, OAQPS began to incorporate probabilistic elements into the NEM
methodology and to apply the resulting model (pNEM) to the criteria pollutants. The
initial result of this work was an early version of pNEM applicable to ozone
(pNEM/O3). This model used a regression-based relationship to estimate indoor
ozone concentrations from outdoor concentrations. A report by Johnson et al.
describes this model and its application to Houston, Texas13.
An advanced version of pNEM applicable to carbon monoxide (pNEM/CO)
was developed in 1991. This model marked the first time in the evolution of NEM
that a mass balance model was used to estimate indoor pollutant concentrations.
The application of pNEM/CO to Denver, Colorado, has been described by Johnson
et al14.
A new version of pNEM/O3 was developed in early 1992. Unlike the earlier
version of pNEM/O3, the new model uses a mass balance model to estimate indoor
ozone concentrations. A February 1993 report by Johnson et al.15 describes the
new version of pNEM/O3 and summarizes the results of an initial application of the
model to 10 cities.
Subsequent to the February 1993 report, ITAQS made the following
enhancements to pNEM/O3 and its input data bases.
Use of more recent (1990-91) fixed-site monitoring data for estimating
ambient ozone concentrations. The earlier analysis was based on
1981-84 monitoring data.
-------
An increase in the number of fixed-site monitors used to represent
each urban area.
Use of more recent (1990) census data for estimating cohort
populations. The earlier analysis used 1980 census data.
A new methodology for adjusting ambient ozone data to simulate
attainment of one-hour and eight-hour NAAQS.
Revision of the algorithm used to determine limiting values for
equivalent ventilation rate.
Development of origin/destination tables through the use of a new
commuting algorithm.
A report by Johnson et al.16 describes these enhancements and summarizes the
results of applying the enhanced model to nine of the ten cities included in the
previous exposure assessment. Tacoma, Washington, was excluded from the
analysis because of insufficient monitoring data.
In early 1994, EPA directed ITAQS to develop a special version of pNEM/O3
applicable to outdoor workers and to use it to estimate the ozone exposures of
outdoor workers residing in each of the nine areas. A summary of this work can be
found in a report by Johnson et al.17
In a follow-up work effort for EPA, ITAQS developed a second special version
of pNEM/OS applicable to children who tend to be active outdoors (hereafter
referred to as "outdoor children"). This report summarizes the results of applying
this version of pNEM/O3 to outdoor children residing in the nine study areas. The
report is divided into eight sections. Section 2 provides an overview of the
pNEM/O3 methodology and describes in detail how the model was applied to
outdoor children in a specific city (Houston). Section 3 describes the mass balance
model incorporated into pNEM/O3. Section 4 describes the process by which
ambient ozone data sets were selected for use in pNEM/O3. It also describes the
methods used to fill in missing values in these data sets. Section 5 presents the
method used to adjust ambient ozone data to simulate the attainment of proposed
air quality standards. Section 6 describes the methods used to identify time/activity
-------
data representative of outdoor children and to estimate the number of outdoor
children in each urban area. Section 7 provides ozone exposure estimates for each
of the nine cities. The principal limitations of the model are discussed in Section 8.
-------
SECTION 2
OVERVIEW OF THE METHODOLOGY
The general NEM methodology consists of five steps.
1. Define a study area, a population-of-interest, appropriate subdivisions
of the study area, and an exposure period.
2. Divide the population-of-interest into an exhaustive set of cohorts.
3. Develop an exposure event sequence for each cohort for the exposure
period.
4. Estimate the pollutant concentration, ventilation rate, and physiological
indicator (if applicable) associated with each exposure event.
5. Extrapolate the cohort exposures to the population-of-interest and to
individual sensitive groups.
This approach has been followed in developing a probabilistic version of NEM
applicable to ozone (pNEM/03). The remainder of this section provides an overview
of the pNEM/OS methodology as applied to outdoor children. The application of
pNEM/OS to outdoor children in Houston is used as a means of demonstrating
various features of the methodology.
2.1 Define Study Area, Population-of-lnterest, Subdivisions of Study Area,
and Exposure Period
The pNEM/O3 methodology provides estimates of the distribution of ozone
exposures within a defined population (the population-of-interest) for a specified
exposure period. The population-of-interest is typically defined as 1) all residents of
a defined study area or 2) the residents of the study area which belong to a specific
sensitive population. The study area is defined as an aggregation of exposure
-------
districts. Each exposure district is defined as a contiguous set of census tracts or
block numbering areas (jointly referred to as "census units") as defined by the
Bureau of Census for the 1990 U.S. census.
All census units assigned to a particular exposure district are located within a
specified radius of a fixed-site ozone monitor. The pNEM/OS methodology is based
on the assumption that the ambient ozone concentration throughout each exposure
district can be estimated by ozone data provided by the associated fixed-site
monitor.
Table 1 lists the nine study areas defined for the exposure analyses. Each
study area is associated with a major urban area. The table lists the number of
exposure districts and the exposure period for each study area. In each case, the
exposure period is defined as a series of months within a particular calendar year.
The specified months conform to the "ozone season" specified for the urban area by
EPA. The ozone season is the annual period when high ambient ozone levels are
likely to occur. Three ozone seasons appear in Table 1: January through
December, March through September, and April through October. The specified
calendar year is either 1990 or 1991, the selected year being the higher year with
respect to reported hourly ambient ozone concentrations.
In the application of pNEM/03 to Houston, eleven fixed-site monitors were
selected to represent ambient ozone concentrations (see Section 4). An exposure
district was constructed around each monitor through the use of a special computer
program ("DIST90"). This program identified all census units having population
centroids located within 15 km of the monitor. When a census unit was paired with
more than one monitor, the program assigned it to the nearest monitor.
The sum of all census units assigned to the eleven exposure districts defined
the Houston study area. In 1990, the study area consisted of 532 census units and
contained 2,370,512 residents18. A subset of this population, outdoor children, were
designated as the principal population-of-interest.
The Houston ozone season spans the entire calendar year. Consequently,
the Houston exposure period was defined as calendar year 1990.
-------
TABLE 1. CHARACTERISTICS OF STUDY AREAS
Study area
Chicago
Denver
Houston
Los Angeles
Miami
New York
Philadelphia
St. Louis
Washington
Number of
exposure
districts
12
7
11
16
6
12
10
11
11
1990
population3
6,175,121
1,484,798
2,370,512
10,371,115
1,941,994
10,657,873
3,785,810
1,706,778
3,085,419
Exposure period
Year
1991
1990
1990
1991
1991
1991
1991
1990
1991
Months
Apr-Oct
Mar-Sep
Jan-Dec
Jan-Dec
Jan-Dec
Apr-Oct
Apr-Oct
Apr-Oct
Apr-Oct
Number of
outdoor
children
cohorts
360
210
330
480
180
360
300
330
330
aTotal population residing in the exposure districts which comprise the study area.
2.2 Divide the Population-of-lnterest Into an Exhaustive Set of Cohorts
In a pNEM analysis, the population-of-interest is divided into a set of cohorts
such that each person is assigned to one and only one cohort. Each cohort is
assumed to contain persons with identical exposures during the specified exposure
period.
In past pNEM/O3 analyses, cohorts were identified by 1) home district, 2)
demographic group, 3) work district, and 4) residential air conditioning system.15'16'17
Specifying the home and work districts provided a means of linking cohort exposure
to ambient pollutant concentrations. Specifying the demographic group provided a
means of linking cohort exposure to activity patterns that vary with age, work status,
and other demographic variables.
The decision to identify cohorts with respect to the residential air conditioning
system was based on the results of two supplemental analyses by ITAQS. An
-------
analysis19 of data on window openings provided by the Cincinnati Activity Diary
Study (CADS)20 suggested that the time per day that windows are open in a
residence is a function of outdoor temperature and air conditioning system, when
the later is characterized as 1) no air conditioning, 2) room units, or 3) central air.
An analysis21 of data collected by Stock22 during a study of asthmatics in Houston
suggested that indoor ozone levels are significantly higher when windows are open
than when windows are closed. For example, the median ratio of indoor ozone to
outdoor ozone for residences in the Sunnyside section of Houston was 0.89 when
windows were open and 0.09 when windows were closed. The importance of
outdoor ozone concentrations in determining indoor ozone concentrations has also
been reported by Weschler et al.23
The slightly different method was used to identify cohorts for the outdoor
children assessment described in this report. Each cohort was identified by
1. Home district
2. Demographic group
3. Residential air conditioning system
4. Replicate number.
Consistent with the earlier pNEM/O3 analyses, cohorts were identified by home
district, demographic group, and residential air conditioning system. Cohorts were
not identified by work (or school) district, however. Analysts assumed that the
members of each cohort attended schools and worked within the home district;
consequently, additional cohort indices were not required for school and work
locations.
Two demographic groups were specified for the outdoor children assessment:
1. Preteens - ages 6 to 13
2. Teenagers ages 14 to 18.
Outdoor children were defined as children who tend to spend more time outdoors
than the average child. Section 6 provides a more detailed definition of the term
and describes the method used to estimate the number of children belonging to
each demographic group.
8
-------
A new feature was installed in the version of pNEM/O3 applicable to outdoor
children. This feature permits the user to specify a "replication" value (n) such that
the model will produce n cohorts for each combination of home district, demographic
group, and residential air conditioning system. Because pNEM/O3 uses a Monte
Carlo process to construct an activity pattern for each cohort, each of the n cohorts
associated with a particular combination of district, group, and air conditioning system
is associated with a distinct exposure sequence.
The replication feature permits the analyst to divide the population-of-interest
into a larger number of smaller cohorts - a process which decreases the "lumpiness"
of the exposure simulation. For example, a total of 66 cohorts would be defined for
the Houston area based on home district (11 possibilities), demographic group (2
possibilities), and air conditioning system (3 possibilities). The average cohort would
contain 3,042 children [i.e., (200,795 children)/(66 cohorts)]. Specifying a replication
value of 5 increases the number of cohorts to 330 and reduces the average size to
574 children. If all other factors are held constant, exposure estimates based on a
set of 330 cohorts will display a smoother empirical distribution (with more detail in
the upper percentiles) than exposure estimates based on a set of 66 cohorts.
The replication value was set equal to 5 for the analyses described in this
report. Table 1 lists the number of cohorts defined for each of the nine study areas.
2.3 Develop an Exposure Event Sequence for Each Cohort for the Exposure
Period
In the pNEM/03 methodology, the exposure of each cohort is determined by an
exposure event sequence (EES) specific to the cohort. Each EES consists of a
series of events with durations from 1 to 60 minutes. To permit the analyst to
determine average exposures for specific clock hours, the exposure events are
defined such that no event falls within more than one clock hour. Each exposure
event assigns the cohort to a particular combination of geographic area and
microenvironment. Each event also provides an indication of respiration rate. In
typical applications, this indicator is a classification of slow - sleeping, slow - awake,
medium, or fast.
-------
The EESs are determined by assembling activity diary records relating to
individual 24-hour periods into a series of records spanning the ozone season of the
associated study area. Because each subject of a typical activity diary study
provides data for only a few days, the construction of a multi-month EES requires
either the repetition of data from one subject or the use of data from multiple
subjects. The latter approach is used in pNEM analyses to better represent the
variability of exposure that is expected to occur among the persons included in each
cohort.
Previous applications of pNEM/O3 have employed activity diary data obtained
from the CADS20. During this study over 900 subjects completed three-day activity
diaries and detailed background questionnaires. Figure 1 presents a page from the
Cincinnati diary. Each subject was instructed to complete a new diary page
whenever he or she changed location or began a new activity.
In the outdoor children exposure analysis, analysts augmented the CADS data
with diary data from six other time/activity studies (see Table 2 and Appendix A).
Section 6 of this report describes how the data from all seven studies were
assembled and processed to produce a unified time/activity database representative
of outdoor children. The data within this special database were organized by study
subject and 24-hour (midnight-to-midnight) time period. The diary records for one
subject for one 24-hour period were designated a "person-day." The data base
contained 792 person-days, each of which was indexed by the following factors:
1. Demographic group: preteens or teenagers
2. Season: summer or winter
3. Temperature classification: cool or warm
4. Day type: weekday or weekend.
The demographic group index was determined by the age of the child who provided
the diary data. The season and day type indices were based on the calendar date of
the person-day.
10
-------
TIME
AM
PM
A. ACTIVITY (please specify)
B. LOCATION
In transit, car 01
In transit, other vehicle . . 02
Specify
Indoors, your residence ... 03
Indoors, other residence. . . 04
Indoors, office 05
Indoors, manufacturing
facility
06
C. BREATHING RATE
Slow (e.g., sitting) 13
Medium (e.g., brisk walk). . . 14
Fast (e.g., running) 15
Breathing problem 16
Specify
D. SMOKING
I am smoking 17
Others are smoking 18
No one is smoking 19
E. ONLY IF INDOORS
(1) Fireplace in use?
Yes ' 20
No 21
(2) Woodstove in use?
Yes. 22
No 23
(3) Windows open?
Yes 24
No 25
Uncertain 26
Indoors, school 07
Indoors, store 08
Indoors, other 09
Specify
Outdoors, within 10 yards of
road or street 10
Outdoors, other 11
Specify
Uncertain 12
*Enter MIDN for midnight and NOON for noon. Otherwise enter four-digit
time (e.g., 0930 for 9:30 and 1217 for 12:17) and check a.m. or p.m.
Figure 1. Page from the activity diary used in the Cincinnati study.20
11
-------
TABLE 2. CHARACTERISTICS OF STUDIES PROVIDING TIME/ACTIVITY DATA FOR OUTDOOR CHILDREN
Database
name
California - 11
and under
California - 12
and over
Cincinnati
Los Angeles -
elem. school
Los Angeles -
high school
Valdez
Washington
Reference
number(s)
24
25
20
27,28
27,28
29
30
Characteristics
of subjects
Children ages 1 to 11
Ages 12 to 94
Ages 0 to 86
Elementary school
students, 10 to 12 years
High school students, 13
to 17 years
Ages 10 to 72
Ages 18 to 70
Number of
subject-
days
1200
1762
2800
58
66
405
705
Study
calendar
periods
April 1989 -
Feb. 1990
Oct. 1987 -
July 1988
March and
August 1985
Oct. 1989
Sept. and
Oct. 1990
Nov. 1990 -
Oct. 1991
Nov. 1982 -
Feb. 1983
Diary type
Retrospective
Retrospective
Real-time
Real-time*
Real-time8
Retrospective
Real-time
Diary time
period
Midnight to
midnight
Midnight to
midnight
Midnight to
midnight
Midnight to
midnight
Midnight to
midnight
Retrospective
7 p.m. to 7
p.m. (nominal)
Breathing
rates
reported?
No
No
Yes
Yes
Yes
No
No
N>
aStudy employed the Cincinnati diary format.
-------
The temperature classification was based on the daily maximum temperature reported
for the diary study area on that date. The cool range was defined as daily maximum
temperatures below 55° F in winter and temperatures below 84° F in summer.
A distinct EES was developed for each cohort in each of the nine study areas
by applying a computerized sampling algorithm to the time/activity data base. The
algorithm was provided with the sequence of daily maximum temperatures reported
for the associated study area and exposure period (Table 1) and with the list of
cohorts defined for the study area. The temperature data were used to assign each
calendar day in the exposure period to one of the temperature ranges used in
classifying the time/activity data. To construct the EES for a particular cohort, the
algorithm selected a person-day from the time/activity data base for each calendar
day in the specified exposure period according to the demographic group of the
cohort and the season, day type, and temperature classification associated with the
calendar day.
Each exposure event within an EES was defined by 1) district, 2) micro-
environment, 3) breathing rate category, and 4) a set of supplemental variables used
to predict ventilation rate. The district was the home district associated with the
cohort.
Seven microenvironments were defined:
1. Indoors - residence - central air conditioning system
2. Indoors - residence - window air conditioning units
3. Indoors - residence - no air conditioning system
4. Indoors - nonresidential locations
5. Outdoors - near road
6. Outdoors - other
7. In vehicle.
Location codes appearing in the time/activity data base were used to determine the
primary microenvironment location of each exposure event (indoors - residence,
indoors - nonresidential locations, outdoors- near road, outdoors - other, or in vehicle).
The indoors - residence location was subdivided into three microenvironments
13
-------
according to air conditioning (AC) system: central system, window unit(s), or none.
This classification was based on the AC system specified for the cohort's residence.
For example, a cohort designated as residing in a home with central AC would always
be assigned to the microenvironment defined as "indoors - residence - central AC"
when activity diary data indicated the cohort was inside a residence.
Four breathing rate categories were defined according to codes appearing in
the time/activity data base: slow - sleeping, slow - awake, medium, and fast. Each
exposure event was assigned to one of these categories.
Subsection 2.4.3 describes an algorithm which was used to estimate a value of
equivalent ventilation rate for each exposure event. The algorithm determines these
estimates as a function of various "predictor variables." The value of each variable
for each exposure event is determined by the diary data associated with the event.
Appendix B lists these variables and describes in detail how diary data are processed
by the algorithm.
2.4 Estimate the Pollutant Concentration and Ventilation Rate Associated
With Each Exposure Event
In the general pNEM methodology, the EES defined for each cohort is used to
determine a corresponding sequence of exposures, event by event. Each exposure is
defined by a pollutant concentration and a ventilation rate indicator.
2.4.1 Estimation of Pollutant Concentration
In the pNEM/03 analysis, the pollutant concentration during each exposure
event was assumed to be a function of the microenvironment and district associated
with the event. Consequently a continuous season or year-long sequence of hourly
average ozone concentrations was developed for each combination of
microenvironment and district. When an exposure event assigned a cohort to a
particular combination of microenvironment and district, the cohort was assigned the
ozone concentration specified for the corresponding clock hour in the appropriate
microenvironment/district sequence.
14
-------
Each year-long sequence of hourly average ozone values for the indoor and in-
vehicle microenvironments was generated by the mass-balance algorithm described
in Section 3. Briefly, this algorithm estimated the hourly average indoor ozone
concentrations during hour h as a function of the indoor ozone concentration at the
end of the preceding hour (i.e., hour h - 1), the ozone concentration outdoors during
hour h, the air exchange rate during hour h (v), and an ozone decay factor (Fd).
Values for the air exchange rate and the ozone decay factor were sampled from
appropriate distributions 'on a daily basis (Subsections 3.1 and 3.3). Air exchange
rate was permitted to change hourly in the three residential microenvironments
depending on whether windows were assigned a status of "open" or "closed". This
assignment was determined through the use of a probabilistic model (Subsection 3.4)
in which the status during each clock hour was assumed to be a function of AC
system, temperature range, and window status during the previous clock hour.
The outdoor ozone concentration associated with microenvironment m in
district d during hour h was determined by an expression having the general form
Cout(m,d,t,s) = b(m] xCmon(d,t,s} + e(C) , (1)
where Cout(m,d,t,s) is the outdoor (or ambient) ozone concentration in micro-
environment m in district d at time t under regulatory scenario s, Cmon(d,t,s) is the
ozone concentration estimated to occur at the monitor representing district d at time t
under regulatory scenario s, b(m) is a constant specific to microenvironment m, and
e(t) is a random normal variable with mean = 0 and standard deviation = cr(m). A
value for e(t) was selected from a normal distribution with mean = 0 and standard
deviation = er(m) every hour. The value of Cmon(d,t,s) was constant over each clock
hour.
In the application of pNEM/O3 described in this report, b(m) was set equal to
1.056 for all microenvironments. A value of 5.3 ppb (0.0053 pprn) was used as the
value of o(m) for all microenvironments (Table 3). Consequently, each sequence of
hourly ozone values was generated by the expression
Coue(m,d,t,s} = 1.056 x Cmon(d,t,s) +e(t), (2)
15
-------
TABLE 3. PARAMETERS ASSOCIATED WITH ALGORITHMS USED
TO ESTIMATE OZONE CONCENTRATIONS IN MICROENVIRONMENTS
Parameter
b(m)
aim)
Air exchange
rate
Ozone decay
factor
Equation(s)
containing
parameter
1
1
38
38
Microenvironment3
All
All
1 -4, 7
1 -4
7
Parameter value
1.056
5.3 ppb
See Table 9
Normal distribution
Arith. mean = 4.04 h"1
Std. dev. = 1.35 h'1
Minimum = 1.44 h"1
Maximum = 8.09 h"1
72.0 h'1
aMicroenvironments:
1 = Indoors - residence - central air conditioning
2 = Indoors - residence - window units
3 = Indoors - residence - none
4 = Indoors - nonresidential locations
5 = Outdoors - near road
6 = Outdoors - other
7 = In vehicle
where e(t) is a random normal variate with mean = 0 and standard deviation = 5.3
PPb.
The expression is based on the results of regression analyses13 performed by
ITAQS analysts on personal exposure data collected by T. Stock during the Houston
Asthmatic Study22. In these analyses, the dependent variable was five-minute
ozone concentration measured outdoors by a personal exposure monitor (PEM).
The independent variable was the simultaneous ozone concentration (hourly
average value) reported by the nearest fixed-site monitor.
16
-------
An initial regression analysis of 327 paired values yielded an intercept of 0.81
ppb, a slope of 1.042, and set of regression residuals with a standard deviation of
18.5 ppb. The R2 value was 0.544. Because the regression intercept value was
found to be non-significant (p = 0.76), a second regression analysis was performed
in which the regression line was forced through the origin (i.e., intercept = 0). This
analysis yielded a slope of 1.056 and a set of regression residuals with a standard
deviation of 18.5 ppb. The residuals were found to be approximately normal
(skewness = -0.32, kurtosis = 0.87).
Attempts were made to fit more complex regression models to the Stock data.
These models included regression equations using logarithmic transformations of
the variables and regression equations which included the previous PEM value as
an independent variable. These alternative models were found to offer no
significant improvement in performance over the model specified above. Some of
the alternative models were found to be unstable.
The results of this analysis suggested that Equation 2 could be used as a
means of generating five-minute values of Cout(d,t,s), given that e(t) values were
selected every five minutes from a normal distribution with mean equal to 0 and
standard deviation equal to 18.5 ppb. A procedure based on this expression was
used in a previous version of pNEM/O3 to generate five-minute ozone
concentrations for the outdoor microenvironment13.
As the new version of pNEM/O3 required hourly-average outdoor ozone
concentrations rather than five-minute values, the procedure used in the earlier
model was modified so that an hourly-average value of e(t) was selected for each
hour from a normal distribution with mean equal to 0 and standard deviation equal to
5.3 ppb. The use of a smaller standard deviation (5.3 ppb versus 18.5 ppb) for the
hourly-average e(t) terms was based on the statistical principle that the standard
deviation of the average of n values drawn from a distribution with standard
deviation equal to a will tend to have a standard deviation equal to cr/m, where m is
the square root of n. As there are 12 five-minute values in one hour, the value of n
is 12. The corresponding value of m is 3.5, and 18.5 ppb/3.5 = 5.3 ppb.
17
-------
The current version of pNEM/03 provides for two outdoor microenvironments:
No. 5 (outdoors - near road) and No. 6 (outdoors - other). In the pNEM/O3
analyses described in this report, these microenvironments were treated identically;
that is, Equation 2 was used to determine the hourly ozone concentrations in each
outdoor microenvironment. This approach is likely to over-estimate ozone
concentrations in microenvironment No. 5 (outdoors - near road) because it does
not account for potential ozone scavenging by nitric oxides emitted from motor
vehicles. The magnitude of this bias is difficult to quantify because of the scarcity of
research in this area and the inconsistency of research findings. For example, a
study by Rhodes and Holland31 of a single freeway in San Diego found that
downwind ozone concentrations measured near the roadway were less than 28
percent of the ozone concentrations measured simultaneously at more distant
outdoor locations judged to be unaffected by the roadway. However, an analysis21
of outdoor personal exposure data obtained from the Stock study found that the
average ratio of personal ozone concentration to fixed-site ozone concentration was
approximately 1.0 in areas of both low and high traffic density.
2.4.2 The Air Quality Adjustment Model
In Equation 1, Cmon(d,t,s) is the monitor-derived value for district d at time t
under scenario s. The value for this variable was determined by adjusting
monitoring data representing baseline conditions (i.e., 1990 or 1991 air quality)
according to the equation
Cmon(d,t,s) = (a) [Cmon(d, t,e)P O)
where Cmon(d,t,e) is the monitor-derived value for district d under baseline conditions.
The multiplicative factor (a) and the exponent (b) are specific to district and
scenario. Section 5 describes the derivation of Equation 3 and provides examples
of its application to Philadelphia monitoring data.
Equation 3 requires a complete (gapless) year of hourly average Cmon(d,t,e)
values for each district. These data sets were prepared by applying a special
18
-------
interpolation program to the hourly average ozone data reported by each fixed-site
monitor. The interpolation program provided an estimate of each missing value.
The resulting filled-in data sets were assumed to represent baseline conditions at
each monitor.
The interpolation program provides estimates of missing values through the
use of a time series model developed by Johnson and Wijnberg32. The time series
model is based on the assumption that hourly average air quality values can be
represented by a combination of cyclical, autoregressive, and noise processes. The
parameter values of these processes are determined by a statistical analysis of the
reported data.
2.4.3 Equivalent Ventilation Rate
In addition to ozone concentration, an equivalent ventilation rate (EVR) value
was estimated for each exposure event. EVR is defined as ventilation rate divided
by body surface area (BSA). Clinical research by EPA suggests that EVR exhibits
less inter-person variability than ventilation rate for a given level of exertion.33
ITAQS analysts developed a special EVR-generator module for the version of
pNEM/O3 applicable to outdoor children. The module used one of four Monte Carlo
models to generate an EVR value for each exposure event associated with a given
cohort. The applied model varied from event-to-event according to 1) the
demographic group of the cohort (active preteens or active teenagers) and 2) the
type of database (A or B) from which the associated diary data were obtained. The
Type A databases were obtained from five of the studies listed in Table 2
(Cincinnati, Denver, Washington, and the two Los Angeles studies). The Type B
databases included the three remaining studies listed in Table 2 (i.e., the two
California studies and the Valdez study).
The Monte Carlo models were developed through an analysis of data
reported by a research team directed by Dr. Jack Hackney and Mr. William Linn.
The Hackney/Linn team conducted two studies in Los Angeles to obtain ventilation
rate data representative of the typical daily activities of elementary school students
19
-------
and high school students.27'28 The heart rate of each study subject was continuously
monitored as the subject documented his or her activities in a special diary. Separate
clinical trials were conducted in which the heart rate and ventilation rate of each
subject were measured simultaneously. These measurements were used to develop
a "calibration curve" for each subject relating heart rate to ventilation rate.
The calibration curves were used to convert the one-minute heart rate
measurements obtained during each diary period into one-minute ventilation rates.
The ventilation rate values were in turn divided by the subject's estimated body
surface area to produce one-minute EVR values.
The Monte Carlo models were developed by applying a four-step procedure to
each of the one-minute EVR databases. In Step 1. ITAQS processed each one-
minute EVR database to produce a special "event EVR file." Each file provided a
sequence of exposure events keyed to the activities documented by each subject.
The listing for each event included the average EVR for the event and the values of
20 variables which were considered likely to influence EVR values.
In Step 2. ITAQS prepared tables of descriptive statistics for event EVR values
which had been categorized by breathing rate, activity, microenvironment, time of day,
and event duration.34 These statistics provided an initial means for identifying factors
to be considered in developing the EVR prediction algorithms. These factors were
compiled into sets of candidate variables, each set specific to a particular database
type.
In Step 3. ITAQS developed two Monte Carlo models for each database type.
Each model was specific to either preteens or teenagers. The Monte Carlo models
were based on the results of statistical analyses performed on EVR data obtained
from the two Hackney/Linn studies discussed above; i.e., elementary school students
and high school students. Models applicable to the preteens demographic group were
based on analyses of data from the elementary school study; models applicable to
teenagers were based on analyses of data from the high school study. To permit the
use of all seven diary databases listed in Table 2, analysts developed two Monte
20
-------
Carlo models for each demographic group -- one applicable to Type A databases and
one applicable to Type B databases.
Each Monte Carlo model predicted EVR as a function of six or more predictor
variables which constituted a "predictor set." Each predictor set was developed by
first defining a candidate variable set for the database type and then performing
stepwise linear regression analyses to determine which of the candidate variables
were significant predictors of EVR for a particular demographic group. All regression
analyses were performed on the two Hackney/Linn databases, as these were the only
databases available which provided a measurement-based EVR value for each
exposure event. The results of the regression analyses determined the variables to
be included in the predictor set and the coefficients of various terms in the associated
Monte Carlo model.
The best overall predictor variable was found to be LGM, the natural logarithm
of the geometric mean of all event EVR values associated with a subject-day of diary
data. Statistical analysis of the LGM values indicated that the distribution of LGM
values was approximately lognormal.
In addition to LGM, the regression analyses suggested that variables
associated with microenvironment, daytime activities, the exertion level of activities,
day of week, and breathing rate were generally useful in predicting event EVR.
Appendix B provides a listing of these variables and the associated regression
coefficients.
Each regression analysis produced a set of residual values, one for each EVR
value. Statistical analysis of the residuals indicated that 1) the standard deviation of
the residuals varied significantly from subject to subject, and 2) the distribution of the
subject-specific standard deviations was approximately lognormal.
Table 4 presents the general algorithm used to implement each Monte Carlo
model. When this algorithm is applied to an appropriate database, it generates a
sequence of EVR values, one for each event in the database. The EVR value
generated for each individual event is determined by the values of the specified
predictor variables, the regression coefficient associated with each predictor variable,
21
-------
TABLE 4. ALGORITHM USED TO GENERATE EVENT-SPECIFIC VALUES OF
EQUIVALENT VENTILATION RATE
1. Go to first/next person-day i.
2. Determine Monte Carlo model applicable to person-day according to
demographic group of cohort and database type of diary data.
3. Model identity determines
MEANLGM: mean of LGM values
SDLGM: standard deviation of LGM values
MU: mean of LSDRES values
SIGMA: standard deviation of LSDRES values
b0: constant
bm: coefficient for variable VARm
Denote the value of bm for variable LGM as b,.
4. Calculate LGM for person-day i:
LGM(i) = MEANLGM + (SDLGM)[Z1(i)]
Z1(i): randomly selected value from unit normal distribution (normal
distribution with mean = 0 and standard deviation = 1).
5. If LGM(i) falls outside range indicated in Table B-7 (Appendix B), discard
value and go to Step 4.
6. Calculate RESSIGMA for person-day i.
LSDRES(i) = MU + (SIGMA)[Z2(i)]
RESSIGMA(i) = Exp[LSDRES(i)]
Z2(i): randomly selected value from unit normal distribution.
7. If LSDRES(i) falls outside range indicated in Table B-6 (Appendix B),
discard value and go to Step 6.
8. Go to first/next event associated with person-day i.
(continued) 22
-------
TABLE 4 (Continued)
9. Read values of variables VAR2, VAR3, ..., VARm for event j of person-day i
from input data file.
10. Calculate residual value for event j of subject i.
RES(i,j) = [RESSIGMA(i)][Z(i,j)]
Z(i,j): randomly selected value from unit normal distribution.
1 1 . Calculate LEVR for event j of person-day i:
LEVR(ij) = b0 + (bOfLGMO)] + (b2)[VAR2(i,j)] + (b3)[VAR3(i,j)] +
(bJtVARJiJ)] + RES(ij)
12. Calculate EVR for event j of person-day i:
EVR(iJ) = Exp[LEVR(i,j)]
13. Write EVR(i.j) to output file.
14. If last event of person-day i, go to Step 1. If not, go to Step 8.
23
-------
an LGM value randomly selected from a study-specific normal distribution, and a
residual standard deviation selected from a subject-specific normal distribution.
Because the algorithm employs Monte Carlo techniques to produce EVR estimates,
each application of the algorithm to a particular time/activity database will produce a
different sequence of exposure estimates. The general algorithm is described in detail
in Appendix B.
In Step 4. ITAQS performed an initial check of the Monte Carlo approach by
applying the EVR-generator algorithm to each of the two Los Angeles databases (see
Appendix C). Each application produced a distribution of event EVR values which
could be compared with the distribution of measurement-derived values. The
modeled and measurement-derived distributions compared favorably with respect to
mean, standard deviation, and percentiles up to the 99th or 99.5th percentiies. At
higher percentiles, the algorithm tended to underestimate EVR for the elementary and
high school databases.
Following these research efforts, ITAQS incorporated the newly-developed
algorithm into an EVR-generator module within the larger pNEM/O3 model. This
module provided an estimate of EVR for each exposure event using the Monte Carlo
model appropriate to 1) the demographic group of the cohort (preteens or teenagers)
and 2) the type of database (A or B) from which the associated diary data were
obtained.
The EVR-generator module also contained an algorithm which established an
upper limit (EVRLIM) for the EVR value assigned to each exposure event. EVRLIM
varied with event duration and was set at a level estimated to be achievable by
members of the cohort who 1) exercised regularly, 2) were motivated to attain high
exertion levels, and 3) were not professional athletes. Joggers would be included in
this group; professional basketball players would not be included.
Table 5 presents the algorithm used to determine EVRLIM. This algorithm is a
variation of "Algorithm B" proposed by Johnson and Adams.35 The algorithm accounts
for the following research findings reported by Erb,36 Astrand and Rodahl,37 and other
researchers.
1. Ventilation rate (VE), oxygen uptake rate (VO2), and the ratio of VE to V02
increase with increasing work rate.
24
-------
TABLE 5. ALGORITHM FOR DETERMINING UPPER LIMIT FOR EVR
1. Obtain values for the following quantities from Table 6.
V02max: maximum oxygen uptake rate
MAXRATIO: maximum ratio of ventilation rate to oxygen
uptake rate
SUBRATIO: submaximal ratio of ventilation rate to oxygen
uptake rate
BSA: body surface area
2. Determine duration of event (t).
3. If t <= 5 minutes, determine the upper limit for EVR (EVRLIM) by the
expression
EVRLIM = (1.2)(VO2max)(MAXRATIO)/BSA.
4. If 5 minutes < t <= 162 minutes, determine the percentage of maximum
oxygen uptake rate that can be maintained for duration t by the expression
PCTV02max = 116.19 - (10.06)[ln(t)].
Next determine the ratio of ventilation rate to oxygen uptake rate by the
expression
RATIO = SUBRATIO +
(MAXRATIO-SUBRATIO)(PCTVO 2max - 65)735.
Finally determine EVRLIM by the expression
EVRLIM = (1.2)(V02max)(PCTV02mJ(RATIO)/(100)(BSA).
5. If t > 162 minutes, determine PCTVO2max by the expression presented in
Step 4 and EVRLIM by the expression
EVRLIM = (1.2)(V02max)(PCTV02max)(SUBRATIO)/(100)(BSA).
25
-------
TABLE 6. PARAMETER VALUES FOR ALGORITHM USED TO DETERMINE
LIMITS FOR EQUIVALENT VENTILATION RATES FOR OUTDOOR CHILDREN
Parameter
acronym
BSA
V02MAX
MAXRATIO
SUBRATIO
Definition
Body surface area, m2
Maximum oxygen uptake rate
(VO2MAX), liters/min
Ratio of ventilation rate (VE) to
oxygen uptake rate (VO2) under
maximum uptake conditions
Ratio of ventilation rate (VE) to
oxygen uptake rate (VO2) under
submaxirnal conditions
Parameter value
Preteens
(ages 6-13)
1.23
2.30
34.5
26.0
Teenagers
(ages 14-18)
1.70
3.49
32.0
22.5
2. A person's maximum VE is determined by his or her maximum oxygen
uptake rate (V02max) and the VEA/O2 ratio in effect under maximum
oxygen uptake conditions (MAXRATIO) such that
'2max'
(MAXRATIO)
3. V02max and MAXRATIO are functions of age, gender, and training,
among other factors.
4. Individuals cannot maintain oxygen uptake rates equal to VO2max for more
than about five minutes.
5. For activity durations greater than five minutes (i.e., t > 5 min), the
percentage of VO2max that can be maintained continuously (PCTVO2max)
decreases as the natural logarithm of the activity duration [ln(t)]
increases.
In determining the EVRLIM value for preteens (ages 6 to 13) applicable to a
particular event duration, the algorithm uses estimates of VO2max, MAXRATIO,
SUBRATIO, and BSA specific to males aged 11 (Table 6). Estimates of EVRLIM
26
-------
provided by Johnson and Adams35 suggest that children in this category are likely to
experience the highest EVR values of all children included in the preteen age group.
In a similar manner, the parameter values listed in Table 6 for children ages 14 to 18
are based on males aged 15.
The reader should note that each of the two sets of parameter values listed in
Table 6 is based on the physiological characteristics of a subset of the specified
demographic group (e.g., males aged 11), but is being applied to all members of the
demographic group (e.g., preteens). Because the EVRLIM of the selected subset is
likely to be higher than average EVRLIM of the demographic group, the use of these
parameter values in the pNEM/O3 simulation will tend to overpredict the occurrence of
high EVR values within each demographic group. This potential bias may be
corrected in future versions of the model by dividing each demographic group into
various subgroups according to age and gender. A separate set of EVRLIM
parameters would have to be developed for each subgroup.
2.4.5 Hourly Average Exposure Sequences
Algorithms within pNEM/03 provided three estimates for each exposure event:
average ozone concentration, average EVR, and the product of average ozone
concentration and EVR (ozone x EVR). These estimates were processed to produce
time-weighted estimates of ozone concentration, EVR, and ozone x EVR for each
clock hour. The result was a year-long sequence of hourly values for each of three
exposure indicators for each cohort. These sequences can be further processed to
determine cohort-specific values for various multihour exposure indicators. Examples
of such indicators include the largest eight-hour daily maximum ozone concentration
and the number of times the hourly-average ozone concentration exceeds 0.12 ppm.
2.5 Extrapolate the Cohort Exposures to the Population-of-lnterest
The cohort-specific exposure estimates developed in Step 4 of the pNEM
methodology (Subsection 2.4) were extrapolated to the general outdoor children
27
-------
population of each study area by estimating the population size of each cohort.
Cohort populations were estimated by the following four-step procedure.
In Step 1, the number of outdoor children residing in each census unit was
estimated by the formula
POPOC(g,c) = ]T (Pig) xPOPC(g,c}\
(4)
where POPOC(g.c) is the number of outdoor children in demographic (age) group g
and census unit c, POPC(g.c) is the number of children in demographic group g who
reside in census unit c, and P(g) is the estimated fraction of children in demographic
group g who are outdoor children. Values for POPC(g,c) were obtained directly from
1990 Bureau of Census data files18 that list population data for age groups by census
unit. Section 6 describes the method used to estimate a value of P(g) for the two
demographic (age) groups used in the outdoor children analysis.
In Step 2, the fraction of homes falling into each of the three air conditioning
categories was estimated by census unit. The fractions associated with each census
unit were determined using 1980 census data, as the 1990 census did not collect air
conditioning data. In cases where the boundaries of a 1990 census unit did not
coincide with 1980 census units, analysts used the fractions associated with the 1980
census unit located nearest to the 1990 census unit.
In Step 3, the outdoor children population of each census unit was multiplied by
the air conditioning fractions to provide an estimate of the number of outdoor children
in each air conditioning category. The estimation equation was
POPOC(g,c,a] =F(c,a) xPOPOC(g,c), (5)
where POPOC(g,c,a) is the population of outdoor children associated with census unit
c and air conditioning system a. F(c,a) is the fraction of housing units in census unit c
with air conditioning system a, and POPOC(g.c) is the number of outdoor children in
demographic group g residing in census unit c (Equation 4). The values of
POPOC(g,c,a) were summed over each exposure district to yield estimates of
28
-------
POPOC(g,d,a), the number of outdoor children in demographic group g within
exposure district d assigned to air conditioning category a. This summation is
explained further in Section 6.3. Table 7 lists the values of POPOC (g.d.a) calculated
for each study area.
As previously discussed, the replication feature was used to create five cohorts
for each combination of demographic group g, exposure district d, and air conditioning
system a. Each of the five cohorts associated with a particular combination of indices
(g, d, and a) received one-fifth of POPOC(g,d,a); that is
POPCOH(g,d,a) = [POPOC(g, d, a) J /5 (6)
where POPCOH(g,d,a) is the population assigned to each cohort.
A special tabulation program in pNEM/03 combined the cohort-specific
estimates of exposure and population to produce histograms and cumulative
frequency tables for various population exposure indicators and averaging times.
Section 7 provides exposure estimates based on existing conditions in each study
area, the attainment of the current NAAQS, and the attainment of each of seven
alternative NAAQS.
29
-------
TABLE 7. POPULATION ESTIMATES BY DEMOGRAPHIC GROUP
AND AIR CONDITIONING STATUS
-
Study area
Chicago
Denver
Houston
Exposure
district
1
2
3
4
5
6
7
8
9
10
11
12
1
2
3
4
5
6
7
1
2
3
4
5
6
7
8
9
10
11
Population estimates by demographic group and air
conditioning status
Preteens
Central
ACa
13,260
2,650
6,890
6,380
9,203
1 1 ,045
4,330
9,980
18,525
10,205
9,950
5,900
3,285
2,215
2,955
675
1,415
305
520
8,525
7,895
1,875
25,645
25,535
3,705
9,645
10,500
4,975
970
3,685
Window
AC
1 1 ,240
8,545
13,680
2,750
29,070
6,420
5,895
9,355
9,290
4,645
3,370
2,540
2,015
600
2,250
595
1,600
1,290
1,625
4,435
650
2,245
2,485
1,850
4,175
2,155
2,310
1,500
2,075
6,400
No
AC
7,410
24,445
9,760
2,440
47,715
2,520
6,390
9,390
5,160
4,115
2,080
1,745
8,430
7,095
1 1 ,045
3,945
7,920
7,285
8,825
1,465
350
1,210
945
465
1,035
515
515
355
1,115
3,790
Teenagers
Central
AC
5,245
1,170
2,930
2,360
3,830
4,420
1,755
4,140
6,780
3,730
3,625
2,640
1,210
860
1,200
305
520
305
685
3,450
2,825
855
9,505
9,635
1,405
3,645
3,925
1,970
385
1,580
Window
AC
4,450
3,670
5,625
1,030
11,760
2,560
2,445
3,730
3,425
1,695
1,245
1,125
715
240
855
270
585
480
655
1,745
235
975
930
720
1,610
805
890
600
830
2,550
No
AC
2,980
9,865
3,935
910
19,040
980
2,600 j
3,835
1,885
1,470
' 755
780
3,005
2,605
3,925
1,765
2,920
2,650
3,370
575
125
525
405
180
415
185
205
140
460
1,515
(continued)
30
-------
Table 7 (Continued)
Study area
Los Angeles
Miami
New York
Exposure
district
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
1
2
3
4
5
6
1
2
3
4
5
6
7
8
9
10
11
12
Population estimates by demographic group and air
conditioning status
Preteens
Central
ACa
5,440
4,415
3,170
4,335
10,165
6,165
2,585
1,195
8,410
7,955
5,780
7,495
19,495
21,840
3,300
13,345
12,190
3,610
11,270
8,320
6,390
11,925
1,030
3,755
780
620
5,045
5,805
2,990
2,465
6,500
5,395
7,150
2,805
Window
AC
7,425
11,855
9,145
9,975
16,735
12,625
5,130
1,870
9,625
8,095
6,215
7,445
4,695
6,635
1,125
4,550
585
1,715
7,920
1,325
13,835
3,725
3,780
21,895
7,570
2,295
11,635
34,070
18,885
30,985
45,975
13,190
20,095
7,605
No
AC
5,695
57,335
53,895
16,775
18,250
17,805
33,220
38,190
17,830
16,965
15,030
9,240
6,510
8,210
1,255
5,075
95
1,010
2,815
120
7,000
1,305
4,810
20,380
5,675
3,890
6,060
83,060
31,750
59,350
41,775
13,290
13,985
6,170
Teenagers
Central
AC
2,120
1,895
1,215
1,870
4,485
2,675
1,055
455
3,540
3,525
2,540
3,365
6,895
8,465
1,155
4,570
4,705
1,415
4,740
3,150
2,675
4,725
455
1,620
325
265
2,330
2,540
1,310
985
2,805
2,370
3,295
1,215
Window
AC
2,860
4,940
3,500
4,105
6,810
5,255
1,920
720
4,125
3,660
2,585
3,080
1,605
2,550
395
1,560
225
670
3,180
590
5,995
1,460
1,635
9,530
3,095
950
5,105
13,750
7,950
12,565
19,900
5,725
8,625
3,215
No
AC
2,195
24,760
20,160
6,880
7,975
7,310
12,160
14,795
7,535
7,540
6,320
3,855
2,280
3,155
440
1,740
35
390
1,130
55
2,800
515
2,070
8,820
2,270
1,630
2,735
32,460
12,595
23,785
17,870
5,535
6,040
2,715
(continued)
31
-------
Table 7 (Continued)
Study area
Philadelphia
St. Louis
Washington
Exposure
district
1
2
3
4
5
6
7
8
9
10
1
2
3
4
5
6
7
8
9
10
11
1
2
3
4
5
6
7
8
9
10
11
Population estimates by demographic group and air
conditioning status
Preteens
Central
ACa
770
8,120
1,645
1,765
4,770
4,410
3,505
4,220
2,305
11,245
3,490
3,435
8,905
10,330
3,295
8,610
7,255
1,965
1,500
740
3,160
2,505
9,310
17,000
9,885
13,575
1,270
6,755
6,555
6,945
3,115
18,105
Window
AC
1,710
12,450
1,420
2,865
4,700
8,740
5,180
16,975
11,965
17,285
3,325
1,705
1,445
765
2,125
1,500
1,820
1,580
2,315
1,915
4,680
2,795
5,590
2,330
1,905
1,770
850
1,315
1,775
800
1,420
1,475
No
AC
1,815
13,465
1,955
2,115
3,785
5,880
3,965
15,685
13,995
7,510
2,605
950
750
460
710
720
930
875
1,510
2,895
4,215
3,335
3,590
1,490
1,785
5,195
900
815
975
405
1,545
1,185
Teenagers
Central
AC
295
3,330
615
685
1,880
1,770
1,590
1,860
980
4,865
1,460
1,350
3,210
4,025
1,265
3,335
2,735
735
515
265
1,285
1,435
4,145
6,325
3,720
5,120
540
2,710
2,555
3,020
1,220
7,940
Window
AC
750
4,750
565
1,105
1,890
3,430
2,220
7,085
4,850
6,725
1,470
675
510
285
910
545
665
570
800
710
1,935
1,555
2,235
850
685
1,865
330
525
695
345
550
580
No
AC
760
5,080
750
815
1,525
2,365
1,675
6,505
5,475
2,915
1,185 I
365
260
175
285
255
340
330
540
1,080
1,700
1,630 I
1 ,490
555
615
2,090
380
315
385
180
555
455
aAC = air conditioning.
32
-------
^SECTION 3
THE MASS-BALANCE MODEL
In the pNEM/03 simulation, the ozone concentration in a particular
microenvironment during a particular clock hour is assumed to be constant. For
indoor and in-vehicle microenvironments, this value is determined by using a mass
balance model to calculate the average ozone concentration for the clock hour
expected under the following conditions:
1. There are no indoor sources of ozone.
2. The indoor ozone concentration at the end of the preceding hour is
specified.
3. The outdoor ozone concentration during the clock hour is constant at a
specified value.
4. The air exchange rate during the clock hour is constant at a specified
value.
5. Ozone decays at a rate that is proportional to the indoor ozone
concentration. The proportionality factor is constant at a specified
value.
The mass balance model employed in these calculations is based on a generalized
mass balance model described by Nagda et al.,38 hereafter referred to as the Nagda
model. As originally proposed, this model assumed that pollutant concentration
decays indoors at a constant rate. For use in pNEM/O3, the Nagda model was
revised to incorporate the alternative assumption that the indoor decay rate is
proportional to the indoor concentration. The Nagda model was further revised to
33
-------
incorporate ozone-specific assumptions concerning various parameter values
suggested by Weschler39 and others.
Subsection 3.1 presents the theoretical basis for the pNEM/O3 mass balance
model and the principal model assumptions. Subsection 3.2 describes the
algorithms which were used to generate hourly values of ozone for the indoor and
in-vehicle microenvironments. Subsection 3.3 presents the procedure used to
determine air exchange rate for the mass balance model. An algorithm for
simulating the opening and closing of windows is described in Subsection 3.4.
3.1 Theoretical Basis and Assumptions
The Nagda model can be expressed by the differential equation
d°in S
~cV n V cV
in = (1 - F ) vC + - mvC - - - i° (7)
( B out ~ in
where Cin = Indoor concentration (units: mass/volume)
FB = Fraction of outdoor concentration intercepted by the enclosure
(dimensionless fraction)
v = Air exchange rate (1/time)
Cout = Outdoor concentration (mass/volume)
S = Indoor generation rate (mass/time)
cV = Effective indoor volume where c is a dimensionless fraction
(volume)
m = Mixing factor (dimensionless fraction)
A = Decay rate (mass/time)
q = Flow rate through air cleaning device (volume/time)
F = Efficiency of the air cleaning device (dimensionless fraction).
In this model, the pollutant decay rate (A) is assumed to be constant. Research by
Nazaroff and Cass40 and by Hayes41 suggests that the decay rate for ozone should
34
-------
be proportional to Cin. Consequently, the pNEM/03 mass balance equation
substitutes the term Fd Cin for the term A/cV in Equation 7. The coefficient Fd is
expressed in units of 1/time.
The following notational changes were made to simplify the equation:
FP
Ve = CV, (9;
Fp is the "penetration factor," and Ve is the "effective volume." The resulting
equation is
Cin = FjC^-r-nNC^-F^C^-S. (10
If the three terms that are proportional to Cin are collected into one term, the
equation can be expressed as
- C- = FvC t + -vfC- , (11)
where
v' = mv+Fd+-^-. (12)
It can be shown that Equation 11 has the following approximate solution:
35
-------
in
in(t) = klCin(t-^t) + k2£ouc+k3, (13)
where
k2 = (FpV/v')
and C^ is the average value of the outdoor concentration over the interval t to t +
At. If GO,, is constant over the interval, then Equation 13 is an exact solution.
The average indoor concentration for hour h, Cjn (h), is given by the
expression
o + a3 (17)
where Cin(h-1) is the instantaneous indoor concentration at the end of the preceding
hour, C^, (h) is the average outdoor concentration for hour h,
aj_ = z(h) , (18)
36
-------
a, = (F0v/v'} [l-z(£)l , (19)
(20]
and
= CL-e-v')/v'.
A steady-state version of the mass balance model can be developed by
solving Equation 11 under the conditions that
-4-Cin = 0 <22)
dt in
and Cout is constant. In this case, the mass balance equation is
n = F vC +_£--v/C. (23)
u * n v '"nnf -,-, Y ^in' \*--J I
which can be rearranged as
Cin = (FpV/Vf)CouC + --. (24)
ve
37
-------
The ratio of indoor concentration to outdoor concentration is
= (Fpv/v<) + S . (25)
Weschler39 has developed a steady-state equation for the indoor/outdoor ratio
which is expressed in his notation as
I/O = Ex/[Ex + kd(A/V}} , (26)
where I = indoor concentration, O = outdoor concentration, Ex = air exchange rate,
kd = deposition velocity, A = surface area, and V = volume. With respect to
Equation 10, Weschler's model implies that there are no indoor sources (S = zero),
no air cleaning devices (F = zero), the penetration factor is unity (Fp = 1), c = 1, and
m = 1. Under these conditions, Equation 10 becomes
= vC-(v+Fd)Cia (27)
dt
and Equation 25 becomes
c I c
^in1 '-out
Weschler's model (Equation 26) and Equation 28 are equivalent if the following
substitutions are made:
Cin = I (29)
'in
38
-------
(30)
v = Ex (31)
= kd(A/V}. (32;
Equation 32 is a particularly useful relationship, as Weschler has identified a number
of studies which suggest that kd(A/V) is relatively constant from building to building.
He suggests that 1.0 x 10~3 sec"1 is a good general estimate of this quantity.
Weschler et al.39 present 14 estimates of kd(A/V) based on data obtained from
specific studies. Nine of these values are based on the observed first-order decay
of ozone in isolated rooms. The remaining five values are based on reported I/O
values and air exchange rates. Table 8 presents means and standard deviations for
the first nine estimates, for the last five estimates, and for all 14 estimates. Two-
sided 95 percent confidence intervals for the means are also provided.
The values in Table 8 can be converted to units of h"1 by multiplying each
value by 3600. Expressed in these units, the mean and standard deviation for the
14 estimates are 4.04 h"1 and 1.35 h"1, respectively. A normal distribution with these
parameters was assumed to represent the distribution of Fd values for the non-
vehicle indoor microenvironments. The value of Fd was not permitted to be less
than 1.44 h"1 or more than 8.09 h"1. The lower bound was based on the smallest
value cited by Weschler39 which was measured in a stainless steel room. The upper
bound corresponds to the 99.87 percentile (i.e., z = 3) of a normal distribution with
mean equal to 4.04 and standard deviation equal to 1.35. The largest value cited by
Weschler et al.39 was 7.2 h'1.
39
-------
The mass balance model was also used to simulate ozone concentrations for
the in-vehicle microenvironment. Ideally, the in-vehicle microenvironment would
have been represented by a distribution of Fd values based on ozone decay rates
measured in a representative sample of motor vehicles. Because of the scarcity of
research concerning ozone decay rates in motor vehicles, ITAQS analysts were not
able to develop such a distribution. Instead, a point estimate of 72.0 h"1 was
assumed for the Fd of the in-vehicle microenvironment. This value was derived by
Hayes41 from an analysis of data for one vehicle presented by Petersen and
Sabersky42. Hayes has used this value in applications of the PAQM exposure
model41.
The use of a point estimate based on a single motor vehicle is likely to
produce a bias in the ozone concentrations estimated for the in-vehicle
microenvironment. The direction of this bias is uncertain.
TABLE 8. MEANS, STANDARD DEVIATIONS, AND CONFIDENCE INTERVALS
FOR ESTIMATES OF kd(A/V) PROVIDED BY WESCHLER
Parameter
Sample size
Mean, sec"1
Standard
deviation, sec"1
Two-sided 95%
confidence
interval, sec"1
Source of kd (A/V) estimate
Observed first-order
decay
9
1.133 x 10"3
0.447 x 10'3
(0.789, 1.477) x 10"3
Reported
I/O values
5
1.098 x 10-3
0.143 x 10'3
(0.920, 1.276) x 10'3
All
14
1.121 x 10"3
0.374 x 10'3
(0.906, 1.335) x 10'3
40
-------
3.2 Simulation of Microenvironmental Ozone Concentrations
Consistent with the theoretical considerations discussed in Subsection 3.1,
the following equation was used to estimate the hourly average ozone concentration
in a particular indoor or in-vehicle microenvironment during hour h:
£.n(h) = a^Cin(h-V+a2£out(h] (33)
where Cjn (h) is the average indoor ozone concentration during hour h, Cin (h-1) is
the instantaneous ozone concentration at the end of the preceding hour, C^ (h) is
the outdoor ozone concentration during hour h,
ax = z(h), (34)
a2 = (v/vO [l-z(h)] , (35)
z(h) = (l-ev)/v', (36)
and
v'-v+fV, (37)
The instantaneous ozone concentration at the end of a particular hour, Cin (h),
was estimated by the equation
Cla(h) = k^C^h-I) +k2£out(h), (38)
where
41
-------
, = e- (39)
JC = (v/vO (1-^), (40)
and v' is determined by Equation 37.
The following algorithm was used to generate a sequence of hourly-average
ozone concentrations for each combination of microenvironment and district.
1. Go to first/next day.
2. Select value of air exchange rate for day from appropriate distribution
or use point estimate. If microenvironment is residential, select one air
exchange value for hours when windows are open and one for hours
when windows are closed. If microenvironment is a nonresidential
building or vehicle, then one air exchange rate is used for all hours of
the day.
3. Select value of decay rate (Fd) for day from appropriate distribution or
use point estimate. If microenvironment is non-vehicular enclosure,
select value of Fd from normal distribution with mean = 4.04 h"1 and
standard deviation = 1.35 h"1. Value is not permitted to be less than
1.44 h"1 or more than 8.09 h"1. If microenvironment is "in vehicle", use
point estimate of 72.0 h"1.
4. Go to first/next clock hour.
5. If microenvironment is residential, use supplementary window algorithm
to determine window status for current hour (open or closed). Window
status determines which air exchange rate determined in Step 2
applies to current hour.
6. Use Equation 33 to determine ozone concentration for current hour
based on air exchange rate specified for hour, outdoor ozone
concentration during hour, and ozone concentration at end of preceding
hour.
7. Use Equation 38 to determine instantaneous ozone concentration at
end of current hour based on air exchange rate specified for hour,
outdoor
42
-------
ozone concentration during hour, and instantaneous ozone
concentration at end of preceding hour. This value is saved for input
into Equation 33 during the next hour.
8. If end of day, go to Step 1. Otherwise, go to Step 4.
Step 2 requires the random selection of an air exchange rate from a specified
distribution. Four enclosure categories were established for this purpose.
Residential buildings - windows open
Residential buildings - windows -closed
Nonresidential buildings
Vehicles.
A survey of the scientific literature determined that there were sufficient data
available to define distributions for only two of the four enclosure categories:
"residential building - windows closed" or "nonresidential building". In each case, a
two-parameter lognormal distribution was found to provide a good fit to the data.
Point (single-valued) estimates were developed for the remaining two enclosure
categories.
Each of the two lognormal distributions was defined by the expression
AER = GM X GSDZ (41)
where AER is the air exchange rate, GM is the geometric mean, and GSD is the
geometric standard deviation. The values for GM and GSD were determined by
fitting lognormal distributions to representative data sets (Subsection 3.3). A value
of AER was selected at random from a particular lognormal distribution by randomly
selecting a value of Z from the unit normal distribution [N(0,1)] and substituting it
into Equation 41. Table 9 lists the values of GM and GSD for the two lognormal
distributions and the values of the point estimates.
The distributions used to determine AER are discussed in more detail in
Subsection 3.3. Subsection 3.4 provides a description of the algorithm used to
determine window status in the residential microenvironments (Step 4).
43
-------
TABLE 9. DISTRIBUTIONS OF AIR EXCHANGE RATE VALUES USED
IN THE pNEM/03 MASS BALANCE MODEL
Enclosure category
Residential building-
windows closed
Residential building-
windows open
Nonresidential building
Vehicle
Air exchange rate distribution
Lognormal distribution
0 Geometric mean = 0.53
0 Geometric standard deviation = 1
0 Lower bound = 0.063
0 Upper bound = 4.47
Point estimate: 6.4
Lognormal distribution
0 Geometric mean = 1.285
0 Geometric standard deviation = 1.
0 Lower bound = 0.19
0 Upper bound = 8.69
704
891
Point estimate: 36
3.3 Air Exchange Rate Distributions
A review of the scientific literature relating to air exchange rates identified 31
relevant references (list available on request). Of these, only a few were found to
contain sufficient data to construct a distribution of air exchange rates relating to a
particular building type such as residence or office. The two most useful studies
were conducted by Grimsrud et al.43 and by Turk et al.44
Residential Locations
Grimsrud et al.43 measured AER's in 312 residences. Reported AER values
ranged from 0.08 to 3.24. ITAQS analyzed these data to determine which of two
distributions (normal versus lognormal) better characterized the data. The lognormal
distribution was found to yield a better fit, as the data were highly skewed. The
fitted lognormal parameters were geometric mean = 0.53 and geometric standard
deviation = 1.704. This distribution was used in pNEM/03 to represent the
44
-------
distribution of AER's in residences with windows closed. Upper and lower limits of
4.47 and 0.063 air changes per hour were established to prevent the selection of
unusually extreme values of AER. These limits corresponded to the substitution of
Z = 4 and Z = -4 in Equation 41 when GM = 0.53 and GSD = 1.704. The upper
bound was 38 percent larger than the largest reported AER (3.24). The lower
bound was 21 percent smaller than the smallest reported AER (0.08).
No comparable data bases were identified which were considered
representative of residences where windows are open. Hayes has used 6.4 h"1 as
the AER value for open windows in applications of the PAQM model.41 This value
was based on an analysis by Hayes45 of a hypothetical building plan with an
assumed "orifice coefficient." Orifice coefficient is defined as the ratio of the
equivalent area of all openings in a building to the building's volume. In support of
this approach, Hayes cites a report by Moschandreas et a!.46 which suggests that
infiltration is proportional to a building's orifice coefficient.
ITAQS analysts considered Hayes's estimate to be the best available
estimate of AER for residences with windows open. Consequently, the AER for
residences with windows open was treated as a point estimate (6.4 h"1) in the
pNEM/O3 analyses described here. Note that the use of an AER estimate
representing a single set of conditions is likely to produce a bias in the ozone
concentrations estimated for this microenvironment. The direction of this bias is
uncertain.
Nonresidential Locations
Turk et al.44 measured AER's in 40 public buildings identified as schools (n =
7), offices (n = 25), libraries (n = 3), and multipurpose buildings (n = 5). The
minimum reported AER was 0.3; the maximum was 4.1. ITAQS analysts fit normal
and lognormal distributions to the data for all 40 buildings and found that the
lognormal distribution produced a slightly better fit, although it had a tendency to
over-predict high values. The fitted lognormal parameters were geometric mean =
1.285 and geometric standard deviation = 1.891.
45
-------
The buildings can be grouped as offices (n = 25) and nonoffices (n = 15).
Lognormal fits to these data sets yielded geometric means and standard deviations
of 1.30 and 1.93 for offices and 1.27 and 1.87 for nonoffices. ITAQS performed a
two-sample t test on the two data sets and found no significant difference in the
means or standard deviations of the data. Consequently, a single lognormal
distribution (geometric mean = 1.285, geometric standard deviation = 1.891) was
used in pNEM/03 for all nonresidential buildings. To prevent the over-prediction of
high AER values, an upper bound of 8.69 was established. This value results when
Z = 3 is substituted into Equation 41 with GM = 1.285 and GSD = 1.891. This value
is over twice the largest AER value (4.1) reported for the 40 buildings and
corresponds to the 99.87 percentile of the specified lognormal distribution. A lower
bound of 0.19 was also established. This value corresponds to a Z value of -3 and
represents the 0.13 percentile of the lognormal distribution.
ITAQS analysts consider the AER data obtained from Turk et al.44 to be
generally representative of buildings with closed windows. Consequently, the
lognormal AER distribution derived from these data may not be applicable to non-
residential buildings which are ventilated by open windows. As comparable data
were not available for non-residential buildings with open windows, analysts applied
the lognormal AER distribution for closed windows to all non-residential buildings.
This approach is likely to under-estimate the ozone exposures of people who
frequently occupy buildings with open windows.
In Vehicle Locations
A point estimate of 36 air changes per hour was used for in-vehicle locations.
This value was obtained from Hayes47 based on his analysis of data for a single
vehicle presented by Peterson and Sabersky42. Hayes notes that the greater AER
observed in vehicles, even with the windows closed, is due to wind effects on the
moving vehicle and the "leakiness" of typical automobiles.
ITAQS analysts considered Hayes's estimate to be the best available
estimate of AER for the in-vehicle microenvironment. Consequently, in-vehicle AER
46
-------
was treated as a point estimate (36 h'1) in the pNEM/O3 analyses described here. It
should be noted that the use of an AER estimate representing a single set of
conditions is likely to produce a bias in the ozone concentrations estimated for this
microenvironment. The direction of this bias is uncertain.
3.4 Window Status Algorithm
The opening and closing of windows in the three residential
microenvironments was simulated by an algorithm which specified a window status
(open or closed) for each clock hour. The algorithm consisted of the following eight-
step procedure.
1. Identify air conditioning system associated with cohort (central, window
units, none).
2. Go to first/next day.
3. Determine average temperature for day from supplementary file.
Identify temperature range which contains this value (below 32, 32 to
below 63, 63 to 75, above 75).
4. Select random number between zero and 1. Compare random number
with probabilities listed in Table 10 for specified air conditioning system
and temperature range. Determine window status for day. If day
status is "windows open all day" or "windows closed all day", set
window status for all clock hours of day as indicated and go to Step 2.
If day status is "windows open part of day", go to Step 5.
5. Go to first/next clock hour.
6. Determine window status of preceding clock hour.
7. Select random number between zero and 1. Compare random number
with probabilities listed in Table 11, 12, or 13 for specified air
conditioning system, clock hour, temperature range, and window status
for preceding hour. If the random number is less than the specified
probability, the window will be open during the clock hour. Otherwise,
the window will be closed.
8. If end of day, go to Step 2. Otherwise, go to Step 5.
47
-------
This algorithm assigns each day to one of three categories: 1) windows closed all
day, 2) windows open all day, and 3) windows open part of day. These
assignments are made according to the air conditioning system associated with the
cohort and the average temperature of the day. If the day assignment is "windows
open part of day", the algorithm assigns window status on an hourly basis for each
of the 24 clock hours in the day. These hourly assignments are made according to
the 1) cohort's air conditioning system, 2) clock hour, 3) average temperature for the
day, and 4) window status of the preceding hour. Both the daily and hourly
assignments are made probabilistically by comparing random numbers to the
probabilities that the specified window status will occur under the stated conditions.
The window status probabilities listed in Tables 10, 11, 12, and 13 were
developed through a statistical analysis of data on window openings obtained from
the CADS.20 This analysis indicated that air conditioning system, temperature, clock
hour, and window status of preceding hour were statistically significant factors
affecting window status.
48
-------
TABLE 10. PROBABILITY OF WINDOW STATUS FOR DAY BY AIR
CONDITIONING SYSTEM AND TEMPERATURE RANGE
Air
conditionin
g system
Central
Room units
None
Temperature
range, °F
Below 32
32 to 62
63 to 75
Above 75
Below 32
32 to 62
63 to 75
Above 75
Below 32
32 to 62
63 to 75
Above 75
Probability of window status for day
Closed all day
1.000
0.851
0.358
0.633
1.000
0.734
0.114
0.160
1.000
0.812
0.095
0.016
Open all day
0
0.009
0.343
0.167
0
0.028
0.505
0.380
0
0.011
0.672
0.823
Open part of day
0
0.140
0.299
0.200
0
0.238
0.381
0.460
0
0.177
0.233
0.161
TABLE 11. PROBABILITY OF WINDOWS BEING OPEN BY CLOCK HOUR,
TEMPERATURE RANGE, AND WINDOW STATUS OF PRECEDING HOUR (PH) FOR
RESIDENCES WITH CENTRAL AIR CONDITIONING
f^\****\*
Clock
hour
1-3
4-6
7-9
10-12
13-15
16-18
19-21
22-24
Probability of windows being open
32°F to 62°F
PH=open
1.000
1.000
0.837
0.679
0.857
0.932
0.646
0.811
PH=closed
0.000
0.005
0.038
0.126
0.149
0.131
0.043
0.036
63°F to 75°F
PH=open
0.978
0.989
0.932
0.865
0.912
0.935
0.892
0.913
PH=closed
0.011
0.000
0.074
0.235
0.240
0.161
0.136
0.101
Above 75 °F
PH=open
0.986
1.000
0.961
0.860
0.923
0.912
0.893
0.909
PH=closed
0.020
0.017
0.094
0.174
0.263
0.000
0.047
0.066
49
-------
TABLE 12. PROBABILITY OF WINDOWS BEING OPEN BY CLOCK HOUR,
TEMPERATURE RANGE, AND WINDOW STATUS OF PRECEDING HOUR (PH) FOR
RESIDENCES WITH WINDOW AIR CONDITIONING UNITS
/-\l _ _l_
Clock
hour
1-3
4-6
7-9
10-12
13-15
16-18
19-21
22-24
Probability of windows being open
32°F to 62°F
PH=open
0.970
0.975
0.864
0.929
0.860
0.859
0.684
0.919
PH=closed
0.006
0.000
0.040
0.121
0.244
0.103
0.063
0.042
63°F to 75°F
PH=open
0.947
0.994
0.934
0.917
0.969
0.956
0.925
0.851
PH=closed
0.007
0.016
0.101
0.303
0.400
0.125
0.176
0.064
Above 75 °F
PH=open
0.974
0.989
0.989
0.849
0.819
0.930
0.902
0.865
PH=closed
0.010
0.017
0.092
0.351
0.152
0.043
0.056
0.121
TABLE 13. PROBABILITY OF WINDOWS BEING OPEN BY CLOCK HOUR,
TEMPERATURE RANGE, AND WINDOW STATUS OF PRECEDING HOUR (PH) FOR
RESIDENCES WITH NO AIR CONDITIONING SYSTEM
Clock
hour
1-3
4-6
7-9
10-12
13-15
16-18
19-21
22-24
Probability of windows being open
32°F to 62°F
PH=open
1.000
1.000
0.950
0.889
0.923
0.848
0.609
0.684
PH=closed
0.015
0.000
0.000
0.200
0.130
0.200
0.067
0.043
63°F to 75°F
PH=open
0.974
1.000
0.868
0.933
1.000
0.964
0.909
0.800
PH=closed
0.031
0.000
0.057
0.400
0.286
0.000
0.500
0.167
Above 75 °F
PH=open
1.000
1.000
1.000
0.875
0.917
0.818
1.000
0.769
PH=closed
0.000
0.000
0.000
0.500
0.000
0.667
0.200
0.500
50
-------
.SECTION 4
PREPARATION OF AIR QUALITY DATA
The pNEM/03 mass balance model requires representative ambient air quality
data for each exposure district in the form of a time series containing one value for
each hour in the specified ozone season. This section describes the procedures
used to select appropriate data sets for the nine study areas. It also describes the
procedure used for filling in missing values in these data sets.
4.1 Selection of Representative Data Sets
To simplify the computer simulation, the ambient ozone concentration
throughout an exposure district was assumed to be a function of the ozone
concentration measured at a single, representative monitoring site located within the
district. Based on guidance from EPA, analysts defined the shape of each exposure
district by first drawing a circle of radius = 15 km with the monitoring site at the
center. If the centroid of a census unit (census tract or block numbering area) was
located within this circle, the census unit was assigned to the exposure district. If a
centroid was located within more than one circle, the census unit was assigned to
the nearest monitor. Note that the monitoring sites selected to represent a city
directly determined the location and shape of the city's exposure districts.
With one exception, the monitoring sites selected for the pNEM/O3 analysis of
outdoor workers were identical to those used in an earlier pNEM/O3 analysis of the
ozone exposure within the general population of the nine study areas. Section 4 of
the report by Johnson et al.16 describes the selection process employed in the
earlier analysis. The exception concerns one of the 12 monitoring sites selected to
represent ambient ozone conditions in the New York study area. This site (identified
by EPA as Site No. 36-061-0063) was selected to represent an exposure district
centered on the southern end of Manhattan Island. Site No. 36-061-0063 was later
51
-------
judged to be unrepresentative of ground-level ozone concentrations in this area of
New York due to the site's high elevation. Consistent with guidance from EPA,
researchers selected the next nearest ozone monitor (No. 36-061-0010) to represent
the Manhattan exposure district in the pNEM/O3 analysis of outdoor children.
Monitor No. 36-061-0010 also represents another exposure district which is centered
on the northern end of Manhattan Island, the actual location of this monitor.
Table 14 lists the number of ozone monitoring sites selected for each study
area. The table also indicates the largest value for the second highest daily
maximum hourly ozone concentration reported by the selected monitors for the
indicated ozone season. It should be noted that the omission of Monitor No. 36-
061-0063 from the New York study area does not affect the value of this air quality
indicator (175 ppb).
4.2 Treatment of Missing Values and Descriptive Statistics
Hourly average ozone data reported by each site were used to estimate the
ambient ozone levels within the associated exposure district. Gaps in the hourly
average ozone data sets were filled in by using a time series model developed by
Johnson and Wijnberg32. The model contains cyclical, autoregressive, and noise
components whose parameters were determined from a statistical analysis of the
reported data.
Tables 15 through 23 provide descriptive statistics for each hourly-average
data set before and after application of the fill-in program. In general, the fill-in
program has little or no effect on the listed percentiles or high values. Whenever
there is a difference in the values for a particular percentile, the filled-in value is
usually lower.
It should be noted that the data sets differ in terms of concentration
resolution. The reported ozone concentration values for all 11 Houston sites and for
15 of the 16 Los Angeles sites are rounded to the nearest 10 ppb. The data for the
other seven cities are rounded to the nearest 1 ppb. All other factors being equal,
the algorithm used to fill in missing values generally performs better when applied to
air quality data of high resolution.
52
-------
TABLE 14. CHARACTERISTICS OF OZONE STUDY AREAS AND MONITORING SITES
Study area
Chicago
Denver
Houston
Los Angeles
Miami
New York City
Philadelphia
St. Louis
Washington, D.C.
Designated
exposure period
s~*
Ozone
season
Apr - Oct
Mar - Sep
Jan - Dec
Jan - Dec
Jan - Dec
Apr - Oct
Apr - Oct
Apr - Oct
Apr - Oct
Year
1991
1990
1990
1991
1991
1991
1991
1990
1991
Number of
counties3
in area
7
6
5
4
2
18
13
7
13
Number of
monitoring
sites
selected
12
7
11
16
6
11"
10
11
11
Largest reported
second high daily
maximum ozone
concentration, ppb
* i ~
129
110
220
310
123
175
156
125
144
aCounties are geographic areas assigned a county code by the Bureau of Census in
Summary Tape File 3 (STF3). A county is counted if any portion is within the study area.
"Monitor No. 36-061-0010 represents two exposure districts.
ITAQS analysts also constructed a data set for each monitor listing eight-hour
running average ozone concentrations based on the filled-in data sets. These data
were used to determine each site's status with respect to various eight-hour NAAQS
under consideration by EPA. Tables 24 through 32 provide eight-hour descriptive
statistics for the monitors selected to represent each city.
53
-------
TABLE 15. DESCRIPTIVE STATISTICS FOR 1991 DATA SETS CONTAINING HOURLY-AVERAGE OZONE
CONCENTRATIONS OBTAINED FROM SELECTED MONITORING SITES IN THE CHICAGO STUDY AREA
Monitor ID
17-031-0001
17-031-0032
17-031-1003
17-031-1601
17-031-4002
17-031-4003
17-031-7002
Monitor
location
Alsip
Chicago
Chicago
Lemont
Cicero
Des Plaines
Evanston
Dis-
trict
code
1
2
3
4
5
6
7
Filled
in?
No
Yes
No
Yes
No
Yes
No
Yes
No
Yes
No
Yes
No
Yes
na
4903
5136
4985
5136
4895
5136
4799
5136
5033
5136
4936
5136
4876
5136
Percentiles, ppb
50
19
19
28
28
19
19
28
28
18
18
23
23
30
30
90
51
50
58
59
51
50
61
60
49
49
53
52
59
58
95
61
61
69
69
63
61
71
71
60
59
63
63
69
69
99
77
77
87
87
81
81
89
89
78
78
80
80
90
90
99.5
83
83
92
92
88
87
98
97
86
86
85
86
97
96
High values,
ppb
Second
104
104
116
116
129
129
126
126
120
120
105
105
115
115
First
108
108
120
120
134
134
152
152
125
125
119
119
123
123
Ol
(continued)
-------
TABLE 15 (Continued)
Monitor ID
17-031-8003
17-043-6001
17-089-0005
17-097-0001
17-097-1002
Monitor
location
Calumet
City
Lisle
Elgin
Deerfield
Waukegan
Dis-
trict
code
8
9
10
11
12
Filled
in?
No
Yes
No
Yes
No
Yes
No
Yes
No
Yes
na
4856
5136
5100
5136
5041
5136
5011
5136
5038
5136
Percentiles, ppb
50
23
24
19
20
26
26
26
26
30
30
90
54
54
49
50
54
54
56
56
61
61
95
64
64
59
59
63
63
67
67
71
71
99
81
81
78
78
82
82
85
85
92
92
99.5
86
86
87
87
91
90
90
90
102
102
High values,
ppb
Second
97
97
116
116
126
126
116
116
119
119
First
109
109
118
118
128
128
124
124
126
126
en
en
'Number of hourly-average ozone concentrations during designated ozone season.
-------
TABLE 16. DESCRIPTIVE STATISTICS FOR 1990 DATA SETS CONTAINING HOURLY-AVERAGE OZONE
CONCENTRATIONS OBTAINED FROM SELECTED MONITORING SITES IN THE DENVER STUDY AREA
Monitor ID
08-001-3001
08-005-0002
08-005-0003
08-013-1001
08-031-0002
08-031-0014
08-059-0002
Monitor
location
Adams Co.
Arapaho Co.
Englewood
Boulder Co.
Denver
Denver
Arvada
a
Dis-
trict
code
1
2
3
4
5
6
7
rar:7;rr.--.T.zi!L.'i..
_.........,..
Filled
in?
No
Yes
No
Yes
No
Yes
No
Yes
No
Yes
No
Yes
No
Yes
=====
=====
na
4322
5136
4047
5136
5036
5136
4458
5136
5063
5136
4453
5136
4908
5136
=====
=====
Percentiles, ppb
50
26
26
40
39
23
23
33
32
17
17
22
22
26
26
90
54
53
63
60
53
54
55
54
40
40
54
53
56
55
"=- -
95
59
58
70
67
62
62
64
63
47
46
62
61
64
64
=====
99
69
68
88
86
76
76
78
77
59
59
77
75
79
79
*---" "
99.5
72
72
93
91
83
83
83
80
64
64
83
81
83
83
=====
=========================
High values,
ppb
Second
87
87
109
109
110
110
102
102
104
104
107
107
115
115
========
First
99
99
111
110
111
111
106
106
120
120
120
120
115
115
=====
en
O5
"Number of hourly-average ozone concentrations during designated ozone season.
-------
TABLE 17. DESCRIPTIVE STATISTICS FOR 1990 DATA SETS CONTAINING HOURLY-AVERAGE OZONE
CONCENTRATIONS OBTAINED FROM SELECTED MONITORING SITES IN THE HOUSTON STUDY AREA
Monitor ID
48-201-0024
48-201-0029
48-201-0046
48-201-0047
48-201-0051
48-201-0059
48-201-0062
Monitor
location
Harris Co.
Harris Co.
Houston
Houston
Houston
Houston
Houston
Dis-
trict
code
1
2
3
4
5
6
7
Filled
in?
No
Yes
No
Yes
No
Yes
No
Yes
No
Yes
No
Yes
No
Yes
na
6865
8760
7689
8760
8138
8760
7970
8760
7999
8760
6941
8760
8072
8760
Percentiles, ppb
50
20
20
20
20
10
10
10
10
20
20
10
10
20
20
90
60
60
50
50
50
50
50
50
50
50
40
40
50
46
95
80
70
70
70
60
60
60
60
70
70
50
50
60
60
99
110
110
100
100
100
100
100
100
110
110
80
70
100
90
99.5
130
120
120
110
120
120
120
120
130
130
90
90
110
110
High values,
ppb
Second
220
220
160
160
200
200
210
210
200
200
140
140
180
180
First
220
220
180
180
230
230
240
240
220
220
190
190
230
230
en
-vl
(continued)
-------
TABLE 17 (Continued)
Monitor ID
48-201-1003
48-201-1034
48-201-1035
48-201-1037
Monitor
location
Deer Park
Houston
Houston
Houston
Dis-
trict
code
8
9
10
11
Filled
in?
No
Yes
No
Yes
No
Yes
No
Yes
na
7685
8760
8098
8760
8300
8760
8086
8760
Percentiles, ppb
50
20
20
10
10
10
10
10
10
90
50
50
50
45
50
50
40
40
95
60
60
60
60
60
60
60
60
99
100
100
90
90
100
100
100
100
99.5
110
110
120
110
120
120
120
120
High values,
ppb
Second
230
230
200
200
230
230
220
220
First
230
230
210
210
230
230
220
220
Ol
00
aNumber of hourly-average ozone concentrations during designated ozone season.
-------
TABLE 18. DESCRIPTIVE STATISTICS FOR 1991 DATA SETS CONTAINING HOURLY-AVERAGE OZONE
CONCENTRATIONS OBTAINED FROM SELECTED MONITORING SITES IN THE LOS ANGELES STUDY AREA
Monitor ID
06-037-0016
06-037-1103
06-037-1301
06-037-1 60V
06-037-1902
06-037-2005
06-037-4002
06-037-5001
Monitor
location
Glendora
Los Angeles
Lynwood
Pico Rivera
Santa Monica
Pasadena
Long Beach
Hawthorne
Dis-
trict
code
1
2
3
4
5
6
7
8
Filled
in?
No
Yes
No
Yes
No
Yes
No
Yes
No
Yes
No
Yes
No
Yes
No
Yes
na
8416
8760
8356
8760
8478
8760
8523
8760
8179
8760
8344
8760
8377
8760
8465
8760
Percentiles, ppb
50
20
20
10
10
10
10
10
10
26
25
10
10
20
20
20
20
90
80
80
50
50
40
40
60
60
65
64
70
70
40
40
50
50
95
110
110
70
70
50
50
80
80
80
79
100
100
50
50
60
60
99
180
180
120
110
80
80
130
130
114
112
160
160
70
70
80
80
99.5
200
200
130
130
90
90
160
160
131
128
170
170
80
80
90
90
High values,
ppb
Second
310
310
170
170
130
130
250
250
191
191
220
220
100
100
110
110
First
320
320
190
190
160
160
260
260
191
191
230
230
110
110
110
115
CD
(continued)
-------
TABLE 18 (Continued)
Monitor ID
06-059-0001
06-059-1003
06-059-3002
06-059-5001
06-065-8001
06-071-1004
06-071-4003
06-071-9004
Monitor
location
Anaheim
Costa Mesa
Los Alamitos
La Habra
Rubidoux
Upland
Redlands
San
Bernardino
Dis-
trict
code
9
10
11
,12
13
14
15
16
Filled
in?
No
Yes
No
Yes
No
Yes
No
Yes
No
Yes
No
Yes
No
Yes
No
Yes
na
8473
8760
8358
8760
8442
8760
8492
8760
8521
8760
8408
8760
8374
8760
8514
8760
Percentiles, ppb
50
10
10
30
30
20
20
20
15
20
20
10
10
30
30
20
13
90
50
50
50
50
50
50
60
53
90
80
70
70
90
90
80
80
95
60
60
60
60
60
60
70
70
110
110
100
90
120
120
110
110
99
100
100
80
80
90
90
110
110
160
160
160
160
180
180
160
160
99.5
110
110
90
90
100
100
130
130
180
180
180
180
190
190
170
170
High values,
ppb
Second
200
200
140
140
150
150
190
190
240
240
240
240
250
250
240
240
First
250
250
170
170
170
170
210
210
240
240
270
270
250
250
250
250
aNumber of hourly-average ozone concentrations during designated ozone season.
-------
TABLE 19. DESCRIPTIVE STATISTICS FOR 1991 DATA SETS CONTAINING HOURLY-AVERAGE OZONE
CONCENTRATIONS OBTAINED FROM SELECTED MONITORING SITES IN THE MIAMI STUDY AREA
Monitor ID
12-011-0003
12-011-2003
12-011-8002
12-025-0021
12-025-0027
12-025-0029
a
Monitor
location
Broward Co.
Pompano
Beach
Dania
Dade Co.
Dade Co.
Dade Co.
1 '--'," "-
Dis-
trict
code
1
2
3
4
5
6
Filled
in?
No
Yes
No
Yes
No
Yes
No
Yes
No
Yes
No
Yes
na
8624
8760
8664
8760
8732
8760
8470
8760
8486
8760
8576
8760
======
Percentiles, ppb
50
22
22
23
23
26
26
21
21
28
28
21
21
~- V."!"
90
42
42
41
41
43
43
41
41
44
44
39
39
95
48
48
46
46
49
49
46
46
49
49
45
44
_-
99
59
59
58
58
61
61
57
57
58
57
54
54
99.5
63
63
64
63
64
64
64
63
65
64
58
58
=====
High values,
ppb
Second
93
93
91
91
95
95
123
123
90
90
85
85
First
94
94
96
96
100
100
124
124
95
95
90
90
:'.===
'Number of hourly-average ozone concentrations during designated ozone season.
-------
TABLE 20. DESCRIPTIVE STATISTICS FOR 1991 DATA SETS CONTAINING HOURLY-AVERAGE OZONE
CONCENTRATIONS OBTAINED FROM SELECTED MONITORING SITES IN THE NEW YORK STUDY AREA
Monitor ID
09-001-0017
34-013-0011
34-017-0006
34-027-3001
34-039-5001
36-001-0080
36-061-0010
Monitor
location
Greenwich
Newark
Bayonne
Morris Co.
Plainfield
Bronx Co.
New York
City
Dis-
trict
code
1
2
3
4
5
6
7,8
Filled
in?
No
Yes
No
Yes
No
Yes
No
Yes
No
Yes
No
Yes
No
Yes
na
4882
5136
5033
5136
4968
5136
4691
5136
4986
5136
4422
5136
4893
5136
Percentiles, ppb
50
29
29
18
18
24
24
39
39
19
20
12
13
14
14
90
61
60
52
52
64
64
75
73
55
55
36
36
43
42
95
75
74
67
67
81
80
88
86
69
68
47
45
58
57
99
110
110
92
92
109
108
111
111
90
90
68
67
87
87
99.5
120
118
97
97
116
116
118
118
97
96
72
72
95
95
High values,
ppb I
Second
147
147
123
123
166
166
137
137
115
115
92
92
151
151
First
161 I
161 I
132
132 I
167 I
167 I
139 I
139 I
120 I
120 I
94 I
94
155 I
155 |
en
ro
(continued)
-------
TABLE 20 (Continued)
Monitor ID
36-061-0063
36-081-0004
36-085-0067
36-103-0002
36-119-2004
Monitor
location
New York
City
Queens Co.
Richmond
Co.
Babylon
White Plains
Dis-
trict
code
b
9
10
11
12
Filled
in?
No
Yes
No
Yes
No
Yes
No
Yes
No
Yes
1 ' '-"-
na
4912
5136
4912
5136
4086
5136
4884
5136
4975
5136
=====
Percentiles, ppb
50
41
41
20
20
28
29
30
30
27
27
=====
90
82
82
57
57
67
62
67
67
62
61
95
96
95
72
72
81
77
81
80
78
78
-
99
122
122
105
105
106
103
111
110
107
107
=====
99.5
130
130
115
115
116
111
121
120
116
116
=====
High values,
ppb
Second
175
175
162
162
169
169
175
175
145
145
First
177
177
174
174
178
178
217
217
152
152
~"""-
en
CO
BNumber of hourly-average ozone concentrations during designated ozone season
bOriginally assigned to District 8. Replaced by Monitor No. 36-061-0010.
-------
TABLE 21. DESCRIPTIVE STATISTICS FOR 1991 DATA SETS CONTAINING HOURLY-AVERAGE OZONE
CONCENTRATIONS OBTAINED FROM SELECTED MONITORING SITES IN THE PHILADELPHIA STUDY AREA
Monitor ID
34-005-3001
34-007-0003
34-007-1001
34-015-0002
42-017-0012
42-045-0002
42-091-0013
Monitor
location
McGuire AFB
Camden
Camden
Gloucester
Bristol
Chester
Norristown
-- - " -i- -
Dis-
trict
code
1
2
3
4
5
6
7
Filled
in?
No
Yes
No
Yes
No
Yes
No
Yes
No
Yes
No
Yes
No
Yes
_
na
4939
5136
4998
5136
4989
5136
5001
5136
4986
5136
5085
5136
4907
5136
=====================^====^==========^
Percentiles, ppb
50
35
34
28
28
36
36
33
33
28
28
30
30
26
26
90
72
72
70
70
76
76
74
73
70
70
67
67
67
66
95
88
88
84
84
89
89
87
87
84
84
78
78
78
77
99
117
117
115
114
112
112
115
115
111
110
103
103
99
98
99.5
126
124
120
120
117
117
125
125
119
118
108
108
106
105
High values,
ppb
Second
156
156
143
143
146
146
151
151
139
139
125
125
125
125
First
156
156
148
148
149
149
151
151
144
144
135
135
127
127
(continued)
-------
TABLE 21 (Continued)
Monitor ID
42-101-0014
42-101-0023
42-101-0024
Monitor
location
Philadelphia
Philadelphia
Philadelphia
Dis-
trict
code
8
9
10
Filled
in?
No
Yes
No
Yes
No
Yes
na
4900
5136
4786
5136
4984
5136
Percentiles, ppb
50
30
30
20
20
30
30
90
70
70
50
50
70
70
95
80
80
70
70
80
80
99
100
100
90
90
110
110
99.5
110
110
100
100
110
110
High values,
ppb
Second
140
140
130
130
130
130
First
140
140
130
130
140
140
CD
cn
aNumber of hourly-average ozone concentrations during designated ozone season.
-------
TABLE 22. DESCRIPTIVE STATISTICS FOR 1990 DATA SETS CONTAINING HOURLY-AVERAGE OZONE
CONCENTRATIONS OBTAINED FROM SELECTED MONITORING SITES IN THE ST. LOUIS STUDY AREA
Monitor ID
17-163-0010
29-183-1002
29-189-0001
29-189-0006
29-189-3001
29-189-5001
29-189-7001
Monitor
location
East St.
Louis
St. Charles
Affton
St. Louis Co.
Clayton
Ferguson
St. Ann
Dis-
trict
code
1
2
3
4
5
6
7
Filled
in?
No
Yes
No
Yes
No
Yes
No
Yes
No
Yes
No
Yes
No
Yes
na
4963
5136
4587
5136
4218
5136
5038
5136
5042
5136
5026
5136
5036
5136
Percentiles, ppb
50
19
19
23
27
28
29
24
24
24
24
18
18
29
29
90
48
48
55
55
62
59
48
48
53
54
42
42
58
58
95
57
57
66
66
75
72
55
55
65
65
48
47
70
70
99
73
73
90
90
93
90
70
69
83
83
61
61
92
92
99.5
83
82
102
98
100
99
75
75
93
92
64
64
96
96
High values,
ppb
Second
116
116
125
125
120
120
99
99
125
125
75
75
130
130
First
124
124
125
125
127
127
100
100
127
127
80
80
135
135
CD
05
(continued)
-------
TABLE 22 (Continued)
Monitor ID
29-510-0007
29-510-0062
29-510-0072
29-510-0080
=====
Monitor
location
St. Louis
St. Louis
St. Louis
St. Louis
=====
Dis-
trict
code
8
9
10
11
---
Filled
in?
No
Yes
No
Yes
No
Yes
No
Yes
-" n -
na
5008
5136
4928
5136
4830
5136
5044
5136
=====
Percentiles, ppb
50
18
18
24
24
18
18
24
24
=====
90
44
44
53
53
40
40
53
53
95
52
52
63
63
48
48
64
65
=
99
69
69
82
82
64
64
86
86
""-
99.5
74
74
89
89
72
72
94
94
High values,
ppb
Second
96
96
108
108
100
100
117
117
:
First
96
96
111
111
110
110
129
129
during designated ozone season.
-------
TABLE 23. DESCRIPTIVE STATISTICS FOR 1991 DATA SETS CONTAINING HOURLY-AVERAGE OZONE
CONCENTRATIONS OBTAINED FROM SELECTED MONITORING SITES IN THE WASHINGTON STUDY AREA
Monitor ID
11-001-0017
11-001-0025
24-031-3001
24-033-0002
24-033-8001
51-013-0020
51-059-0018
Monitor
location
Washington
Washington
Rockville
Greenbelt
Suitland-
Silver Hills
Arlington
Co.
Mt. Vernon
Dis-
trict
code
1
2
3
4
5
6
7
Filled
in?
No
Yes
No
Yes
No
Yes
No
Yes
No
Yes
No
Yes
No
Yes
na
4928
5136
5031
5136
4881
5136
5034
5136
4997
5136
5034
5136
4897
5136
Percentiles, ppb
50
19
19
24
24
29
29
30
30
31
31
28
28
30
30
90
54
54
61
60
69
68
74
74
69
68
68
68
71
71
95
64
64
72
71
79
79
87
87
81
81
80
79
83
83
99
82
82
90
90
100
99
110
109
102
102
102
102
106
105
99.5
91
90
99
99
103
103
115
114
108
108
107
107
111
111
High values,
ppb
Second
137
137
144
144
135
135
148
148
139
139
142
142
126
126
First
147
147
148
148
137
137
153
153
144
144
148
148
142
142
O5
OO
(continued)
-------
TABLE 23 (Continued)
Monitor ID
51-059-1004
51-059-5001
51-510-0009
51-600-0005
Monitor
location
Seven
Corners
McLean
Alexandria
Fairfax
===================
Dis-
trict
code
8
9
10
11
--1 '-"-- "'-
Filled
in?
No
Yes
No
Yes
No
Yes
No
Yes
---- - ' - ---
na
4951
5136
5037
5136
4916
5136
4947
5136
=============
Percentiles, ppb
50
33
33
27
27
22
22
33
32
=====
90
71
71
63
63
54
54
66
66
=====
95
86
86
73
74
65
65
77
76
.
99
110
109
95
95
84
84
97
96
=====
99.5
119
116
104
101
95
94
107
106
=======
High values,
ppb
Second
174
174
137
137
131
131
131
131
-
First
178
178
138
138
132
132
132
132
=" L-r
CD
CD
"Number of hourly-average ozone concentrations during designated ozone season.
-------
TABLE 24. DESCRIPTIVE STATISTICS FOR 1991 DATA SETS CONTAINING EIGHT-HOUR OZONE
CONCENTRATIONS OBTAINED FROM SELECTED MONITORING SITES IN THE CHICAGO STUDY AREA
""
Monitor ID
17-031-0001
17-031-002
17-031-1003
17-031-1601
17-031-4002
17-031-4003
17-031-7002
17-031-8003
17-043-6001
17-089-0005
17-097-0001
17-097-1002
========
Monitor
location
Alsip
Chicago
Chicago
Lemont
Cicero
Des
Plaines
Evanston
Calumet
City
Lisle
Elgin
Deerfield
Waukegan
District
code
1
2
3
4
5
6
7
8
9
10
11
12
==================================S====== ======^=======:==
Percentiles, ppb
50
20
28
19
28
18
24
30
24
20
26
26
31
'---'-- " -' ' -
90
46
54
46
57
45
48
55
49
45
50
52
58
=============
95
54
63
55
66
54
57
64
58
53
58
61
66
==============
99
69
80
71
82
70
72
83
74
70
74
77
84
99.5
75
84
76
88
75
77
86
78
79
82
83
88
:
High values, ppb
Second
94
106
101
108
95
93
101
90
98
106
101
104
"
First
95
107
101
109
95
95
102
90
98
106
103
106
=====
-------
TABLE 25. DESCRIPTIVE STATISTICS FOR 1990 DATA SETS CONTAINING EIGHT-HOUR OZONE
CONCENTRATIONS OBTAINED FROM SELECTED MONITORING SITES IN THE DENVER STUDY AREA
Monitor ID
08-001-3001
08-005-0002
08-005-0003
08-013-1001
08-031-0002
08-031-0014
08-059-0002
Monitor
location
Adams Co.
Arapaho Co.
Englewood
Boulder Co.
Denver
Denver
Arvada
District
code
1
2
3
4
5
6
7
Percentiles, ppb
50
26
38
24
33
18
23
26
=====
90
47
56
48
50
35
47
50
95
52
62
54
57
41
52
57
- - -'-' -
99
60
76
65
68
51
62
68
=====
99.5
63
80
70
71
54
64
72
High values, ppb
Second
72
87
83
83
84
77
95
First
74
87
83
85
85
80
96
-------
TABLE 26. DESCRIPTIVE STATISTICS FOR 1990 DATA SETS CONTAINING EIGHT-HOUR OZONE
CONCENTRATIONS OBTAINED FROM SELECTED MONITORING SITES IN THE HOUSTON STUDY AREA
Monitor ID
48-201-0024
48-201-0029
48-201-0046
48-201-0047
48-201-0051
48-201-0059
48-201-0062
48-201-1003
48-201-1034
48-201-1035
48-201-1037
Monitor
location
Harris Co.
Harris Co.
Houston
Houston
Houston
Houston
Houston
Deer Park
Houston
Houston
Houston
======
District
code
1
2
3
4
5
6
7
8
9
10
11
50
21
21
14
15
21
14
17
19
16
15
12
=====
=============================3==^^
Percentiles, ppb
90
50
49
42
42
48
33
41
46
41
42
39
95
64
61
53
55
61
41
52
56
54
56
51
99
92
89
84
82
92
60
79
84
81
86
81
=============
99.5
104
96
95
96
105
71
90
92
90
97
92
=================
- '" " ' "'!.' '-!_L.
High values, ppb
Second
149
124
151
156
167
110
154
139
144
156
160
==============
First
150
124
152
164
170
112
155
140
146
157
164
=n =
to
-------
TABLE 27. DESCRIPTIVE
CONCENTRATIONS OBTAINED
STATISTICS FOR 1991 DATA SETS CONTAINING EIGHT-HOUR OZONE
FROM SELECTED MONITORING SITES IN THE LOS ANGELES STUDY AREA
Monitor ID
06-037-0016
06-037-1103
06-037-1301
06-037-1601
06-037-1902
06-037-2005
06-037-4002
06-037-5001
06-059-0001
06-059-1003
06-059-3002
06-059-5001
06-065-8001
06-071-1004
06-071-4003
06-071-9004
Monitor
location
Glendora
Los Angeles
Lynwood
Pico Rivera
Santa Monica
Pasadena
Long Beach
Hawthorne
Anaheim
Costa Mesa
Los Alamitos
La Habra
Rubidoux
Upland
Redlands
San
Bernardino
District
code
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
Percentiles, ppb
50
24
14
12
12
27
18
17
21
17
25
25
17
24
16
30
19
90
70
47
34
51
58
62
35
46
42
47
50
50
76
61
86
74
95
95
60
41
67
69
84
42
51
52
55
59
62
97
84
110
96
99
135
85
62
97
93
120
56
67
77
71
75
90
139
124
152
135
99.5
150
92
67
111
101
130
61
76
85
76
80
100
155
134
162
146
High values, ppb
Second
181
120
86
142
155
165
82
96
119
101
97
129
194
164
197
192
First
182
120
89
146
155
166
83
99
119
102
99
132
196
165
197
192
CO
-------
TABLE 28. DESCRIPTIVE STATISTICS FOR 1991 DATA SETS CONTAINING EIGHT-HOUR OZONE
CONCENTRATIONS OBTAINED FROM SELECTED MONITORING SITES IN THE MIAMI STUDY AREA
Monitor ID
12-011-0003
12-011-2003
12-011-8002
12-025-0021
12-025-0027
12-025-0029
Monitor location
Broward Co.
Pompano Beach
Dania
Dade Co.
Dade Co.
Dade Co.
District
code
1
2
3
4
5
6
Percentiles, ppb
50
22
22
25
21
27
21
90
39
39
42
37
43
37
95
44
44
47
43
47
42
99
54
52
56
52
55
51
99.5
56
54
59
55
58
53
High values, ppb
Second
76
71
71
77
77
73
First
77
72
72
79
80
73
-------
TABLE 29. DESCRIPTIVE STATISTICS FOR 1991 DATA SETS CONTAINING
CONCENTRATIONS OBTAINED FROM SELECTED MONITORING SITES IN THE
EIGHT-HOUR OZONE
NEW YORK STUDY AREA
Monitor ID
09-001-0017
34-013-0011
34-017-0006
34-027-3001
34-039-5001
36-001-0080
36-061-0010
36-061-0063
36-081-0004
36-085-0067
36-103-0002
36-119-2004
Monitor
location
Grenwich
Newark
Bayonne
Morris Co.
Plainfield
Bronx Co.
New York City
New York City
Queens Co.
Richmond Co.
Babylon
White Plains
District
code
1
2
3
4
5
6
7, 8
a
9
10
11
12
Percentiles, ppb
50
29
19
25
39
21
14
15
41
21
29
30
27
90
57
46
58
70
50
32
39
79
51
58
62
58
95
67
59
72
82
61
41
50
90
64
71
73
70
99
95
82
95
100
80
56
73
113
90
95
97
94
99.5
103
89
103
109
88
59
79
122
99
101
104
105
High values, ppb
Second
125
102
112
125
109
69
102
133
119
135
129
125
First
126
103
112
125
109
71
102
135
119
136
129
127
en
aOriginally assigned to District 8. Replaced by Monitor No. 36-061-0010.
-------
TABLE 30. DESCRIPTIVE STATISTICS FOR 1991 DATA SETS CONTAINING EIGHT-HOUR OZONE
CONCENTRATIONS OBTAINED FROM SELECTED MONITORING SITES IN THE PHILADELPHIA STUDY AREA
Monitor ID
34-005-3001
34-007-0003
34-007-1001
34-015-0002
42-017-0012
42-045-0002
42-091-0013
42-101-0014
42-101-0023
42-101-0024
Monitor
location
McGuire AFB
Camden
Camden Co.
Gloucester
Bristol
Chester
Norristown
Philadelphia
Philadelphia
Philadelphia
District
code
1
2
3
4
5
6
7
8
9
10
Percentiles, ppb
50
34
28
37
33
28
30
26
31
21
27
90
67
64
71
68
64
62
60
65
49
61
95
80
76
81
80
76
72
70
76
60
72
99
107
101
103
105
100
92
92
96
79
97
99.5
114
109
107
113
104
98
98
100
86
103
High values, ppb
Second
138
129
124
135
115
113
118
125
112
116
First
141
131
125
135
116
114
118
127
114
116
CD
-------
TABLE 31. DESCRIPTIVE STATISTICS FOR 1990 DATA SETS CONTAINING
CONCENTRATIONS OBTAINED FROM SELECTED MONITORING SITES IN THE
EIGHT-HOUR OZONE
ST. LOUIS STUDY AREA
Monitor ID
17-163-0010
29-183-1002
29-189-0001
29-189-0006
29-189-3001
29-189-5001
29-189-7001
29-510-0007
29-510-0062
29-510-0072
29-510-0080
Monitor
location
East St. Louis
St. Charles
Affton
St. Louis Co.
Clayton
Ferguson
St. Ann
St. Louis
St. Louis
St. Louis
St. Louis
District
code
1
2
3
4
5
6
7
8
9
10
11
Percentiles, ppb
50
20
26
30
24
25
19
29
19
25
19
25
90
43
50
54
44
49
39
54
40
48
37
50
95
51
59
64
50
58
44
64
48
57
43
60
99
66
78
80
62
76
54
81
60
73
56
76
99.5
70
85
85
67
80
56
84
65
77
61
83
High values, ppb
Second
98
110
100
85
93
62
101
76
89
83
99
First
99
110
103
86
94
63
104
77
91
85
100
-------
TABLE 32. DESCRIPTIVE STATISTICS FOR 1991 DATA SETS CONTAINING EIGHT-HOUR OZONE
CONCENTRATIONS OBTAINED FROM SELECTED MONITORING SITES IN THE WASHINGTON STUDY AREA
Monitor ID
11-001-0017
11-001-0025
24-031-3001
24-033-0002
24-033-8001
51-013-0020
51-059-0018
51-059-1004
51-059-5001
51-510-0009
51-600-0005
Monitor
location
Washington
Washington
Rockville
Greenbelt
Suitland S.H.
Arlington Co.
Mt. Vernon
Seven
Corners
McLean
Alexandria
Fairfax
District
code
1
2
3
4
5
6
7
8
9
10
11
Percentiles, ppb
50
20
25
30
31
32
29
30
33
28
23
33
90
48
55
62
68
63
61
65
66
56
50
61
95
58
64
71
79
73
72
75
77
65
59
70
99
73
79
88
96
90
91
92
97
81
75
88
99.5
78
85
93
102
94
97
99
102
89
82
96
High values, ppb
Second
120
114
113
129
124
127
110
147
115
111
110
First
120
117
113
131
125
128
112
147
115
111
111
GO
-------
SECTION 5
ADJUSTMENT OF OZONE DATA TO SIMULATE COMPLIANCE
WITH ALTERNATIVE AIR QUALITY STANDARDS
In applying pNEM/03 to a particular study area, the analyst typically defines
the air quality conditions within the area as representing (1) baseline conditions or
(2) conditions in which the area just attains a specific NAAQS. This section
describes the procedures used to develop monitor-specific ozone data sets
representing baseline and attainment conditions in each of the nine study areas.
Fixed-site monitoring data for the years 1990 and 1991 were used to
represent baseline conditions for each of the nine study areas. Special air quality
adjustment procedures (AQAP's) were used to adjust the baseline data to simulate
conditions in which each study area just attains a specific NAAQS. EPA identified
the following NAAQS formulations for assessment:
1. One hour daily maximum - one expected exceedance (1H1EX): the
expected number of daily maximum one-hour ozone concentrations
exceeding the specified value shall not exceed one.
Standard levels: 120 ppb (the current NAAQS for ozone), 100 ppb
2. Eight-hour daily maximum - one expected exceedance (8H1EX): the
expected number of daily maximum eight-hour ozone concentrations
exceeding the specified value shall not exceed one.
Standard levels: 70 ppb, 80 ppb, 90 ppb, 100 ppb
3. Eight-hour daily maximum - five expected exceedances (8H5EX): the
expected number of daily maximum eight-hour ozone concentrations
exceeding the specified value shall not exceed five.
Standard levels: 80 ppb, 90 ppb
A separate AQAP was developed for each of the three classes of NAAQS (1H1EX,
8H1EX, and 8H5EX).
Each AQAP consisted of the following four steps:
79
-------
1. Specify an air quality indicator (AQI) to be used in evaluating the status
of a monitoring site with respect to the NAAQS of interest.
2. Determine the value of the AQI for each site within the study area
under baseline conditions.
3. Determine the value of the AQI under conditions in which the air
pollution levels within the study area have been reduced or increased
until the site with the highest pollution levels just attains a specified
NAAQS.
4. Adjust the one-hour values of the baseline data set associated with
each site to yield the AQI value determined in Step 3. The adjusted
data set should retain the temporal profile of the baseline data set.
Subsection 5.1 discusses the specification of appropriate AQI's (Step 1) and the
determination of baseline AQI values (Step 2). Subsection 5.2 presents the
methods used to estimate AQI's under attainment conditions (Step 3). Subsection
5.3 describes the procedures used in Step 4 to adjust one-hour data to simulate
significant reductions in ozone levels within a study area. More detailed
descriptions of these procedures can be found in Appendices A and B of a report by
Johnson et al.16 Subsection 5.4 provides examples in which the procedures
described in Subsection 5.3 were applied to Philadelphia. Subsection 5.5 presents
an alternative procedure which analysts used to adjust one-hour data to simulate
small changes (decreases or increases) in ozone levels within a study area. This
procedure was applied to Denver, Chicago, and Miami for all NAAQS formulations.
5.1 Specification of AQI and Estimation of Baseline AQI Values
The following AQI's were selected for evaluating the 1H1EX, 8H1EX, and
8H5EX standards.
1H1EX: the characteristic largest daily maximum one-hour ozone
concentration
8H1EX: the characteristic largest daily maximum eight-hour ozone
concentration (except for Denver, in which the observed second
highest daily maximum was used, as explained in Subsection
5.5)
80
-------
8H5EX: the observed sixth largest daily maximum eight-hour ozone
concentration.
Note that a statistical AQI (the characteristic largest value) was generally specified
for the 1H1EX and 8H1EX standards, whereas a deterministic AQI (the observed
sixth largest value) was used for the 8H5EX standards. Analysts elected to use
statistical AQI's for the 1H1EX and 8H1EX standards because such indicators are
less affected by anomalous high values than the corresponding deterministic AQI
(the second highest observed value). A statistical indicator was not considered
necessary for the 8H5EX standards, as the sixth highest observed value is relatively
unaffected by anomalous high values.
The characteristic largest value (CLV) of a distribution is that value expected
to be exceeded once in n observations. If F(x) is the cumulative distribution of x,
then
F(x) = 1 - - (42)
n
when x is the CLV.
Selection of an appropriate cumulative distribution to fit data is important in
determining a reasonable CLV- Two distributions that often provide close fits to
ambient air quality data are the Weibull and the lognormal. The Weibull distribution
is defined as
F(x) = 1 - exp [-(*)*] (43)
o
where 6 is the scale parameter and k is the shape parameter. The lognormal
distribution is defined as
F(x) =
_L f^exp (-tV2) dt (44}
V27T J
where
81
-------
_ In x - fj. (45)
a
and In x is distributed normally with mean jj and variance a2. As discussed in
previous reports, the Weibull distribution generally provides a better fit to hourly
average ozone data.15
The hourly average values reported by a single monitoring site during a
specified ozone season form a time series xt (t = 1, 2, 3, ..., n). If the hourly
average time series is complete, it will contain n = (24)(N) values, where N is the
number of days in the ozone season. From this time series a second time series of
daily maximum 1-hour values can be constructed.
Assume that a Weibull distribution with parameters 6 and k provides a good
fit to the empirical distribution of hourly average values, if one disregards
autocorrelation, the value expected to be exceeded once in n = (24)(N) hours can
be estimated as
CLVOH = 5 [In (24) (N)]i/k. <46)
This is the characteristic largest one-hour value. If we again disregard
autocorrelation, the daily maximum 1-hour value expected to be exceeded once in N
days can be estimated as
CLVOHDM= 6{-ln[l - ()1/24]}1^. (47)
This is the characteristic largest daily maximum one-hour value. For 7-month and
12-month ozone seasons, N is equal to 214 and 365, respectively. For these values
of N, CLVOH and CLVOHDM are virtually indistinguishable in value over the range
in k
values typically found in ozone data (0.6 < k < 2.5). For example, the following
values were calculated using 6 = 40 ppb.
82
-------
0.6
1.4
2.5
0.6
14
2.5
1428
185
94
1580
193
97
1428
185
94
1580
193
97
N k CLVOH CLVOHDM
214
365
The CLVOH and CLVOHDM values match to the nearest ppb. Consequently,
the expression
CLVOHDM = 5 [ln(24) (N) ]1/k (48)
can be used as an alternative to Equation 47 for calculating CLVOHDM. The quantity
calculated by Equation 48, hereafter denoted by CLV1, was selected as the AQI to be
used in evaluating the status of a monitoring site with respect to a particular 1H1EX
standard.
A data set containing one-hour concentration values can be processed to
determine a corresponding data set containing eight-hour running average values. If
a Weibull distribution is fit to the eight-hour data, one can determine a characteristic
largest eight-hour value by the equation
CLVEH = 5 [ln(24)W) ]i/k, (49)
where 6 and k are the Weibull parameters for the eight-hour fit. Based on the
argument made above for one-hour data, this value should be approximately equal to
the characteristic largest daily maximum eight-hour value (CLVEHDM) of the data set.
For simplicity, the term CLV8 is hereafter used to refer to the quantity calculated by
Equation 49. CLV8 was selected as the AQI to be used in evaluating attainment
status with respect to a particular 8H1EX standard.
Table 33 lists the data sets selected to represent baseline conditions in each of
the nine cities under analysis. Table 33 also provides estimates of CLV1 and CLV8
83
-------
TABLE 33. BASELINE AIR QUALITY INDICATORS FOR NINE CITIES
City
Chicago
Denver
Houston
Year
1991
1990
1990
District
1
2
3
4
5
6
7
8
9
10
11
12
1
2
3
4
5
6
7
1
2
3
4
5
6
7
8
9
10
11
Ozone concentration, ppb
CLV1
109
124
123
134
120
111
119
104
122
127
122
131
91
116
114
103
98
117
109
224
182
241
224
227
180
208
207
231
235
232
CLV8
94
107
106
114
99
97
106
92
106
111
106
111
74
94
85
86
79
78
94
162
137
161
171
179
131
165
143
154
171
167
EH6LDM
78
86
77
90
78
79
89
78
82
83
85
91
67
84
73
74
56
65
75
116
110
110
107
124
86
104
99
101
116
107
(Continued)
84
-------
Table 33 (Continued)
City
Los Angeles
Miami
New York
Year
1991
1991
1991
District
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
1
2
3
4
5
6
1
2
3
4
5
6
7a
8a
9
10
11
12
Ozone concentration, ppb
CLV1
321
185
148
271
215
248
116
136
198
153
167
216
264
266
261
249
90
97
93
105
96
87
158
121
153
143
123
97
141
141
162
170
183
148
CLV8
207
133
99
166
162
172
85
104
121
101
100
134
209
184
215
204
74
74
72
82
80
72
135
112
133
134
113
75
108
108
131
143
140
137
EH6LDM
170
109
75
129
115
146
64
84
94
81
87
110
167
146
180
165
60
60
64
59
65
57
108
91
113
105
88
64
83
83
104
101
107
105
(Continued)
85
-------
Table 33 (Continued)
City
Philadelphia
St. Louis
Washington
Year
1991
1990
1991
District
1
2
3
4
5
6
7
8
9
10
1
2
3
4
5
6
7
8
9
10
11
1
2
3
4
5
6
7
8
9
10
11
Ozone concentration, ppb
CLV1
167
149
153
162
145
134
135
140
131
141
124
141
131
103
122
78
124
100
114
103
119
134
135
130
143
141
143
135
169
141
134
145
CLV8
142
136
128
138
120
118
123
128
116
126
100
116
106
87
97
65
103
79
91
84
104
110
113
113
128
119
123
118
143
120
112
123
EH6LDM
116
113
111
115
107
101
102
104
90
102
73
88 i
87
68
81
59
87
67
80
64
86 I
80 I
88
95
106
98
100
104
102
91
85
100
Districts 7 and 8 in New York are represented by the same ozone monitor (Monitor
No. 36-061-0010).
86
-------
based on Weibull fits to the upper two percent of each data set. These values were
used as estimates of CLV1 and CLV8 representing baseline conditions.
As previously indicated, the sixth largest daily maximum 8 hour value
(denoted EH6LDM) was used to evaluate the status of a monitoring site with respect
to a particular 8H5EX standards. Table 33 lists the baseline value of this AQI for
each site in the nine cities under analysis.
5.2 Estimation of AQI's Under Attainment Conditions
Tables 34, 35, and 36 provide the step-by-step procedures followed in
implementing the AQAP's developed respectively for 1H1EX, 8H1EX, and 8H5EX
NAAQS. In general, analysts assumed that the i-th ranked site (ranking determined
by baseline AQI) will undergo a change in its AQI value proportional to the change
required for the highest ranked site to exactly attain the specified standard. The
ranking assigned to a particular site under attainment conditions was determined by
the site's average ranking over five years, rather than the site's ranking under
baseline conditions. Consequently, the site ranked highest under baseline
conditions was not necessarily the highest ranked site under attainment conditions.
Evaluation of representative ozone data suggested that a site's future ranking could
be better predicted from its long-term average rank than from a single year's
ranking.
Steps 1 through 4 in each table comprise the procedures used to estimate the
value of an attainment AQI value for each site in a particular city. Each attainment
AQI was converted to a corresponding characteristic one-hour largest value under
attainment (ACLV1). For 1H1EX standards (Table 34), the value of ACLV1
determined by Step 4 was used without further adjustment as the value of ACLV1
required in subsequent steps. For 8H1EX standards (Table 35), the value of ACLV8
determined in Step 4 was converted to the required ACLV1 value through the use of
an equivalence relationship (Step 5). The equivalence relationship was
ACLVl = (RATIO 1) (ACLV8) (53)
where RATIO1 varied with urban area (Table 37).
87
-------
TABLE 34, AIR QUALITY ADJUSTMENT PROCEDURE USED TO SIMULATE
ATTAINMENT OF 1H1EX NAAQS (THE EXPECTED NUMBER OF
DAILY MAXIMUM ONE-HOUR OZONE CONCENTRATIONS EXCEEDING THE
SPECIFIED VALUE SHALL NOT EXCEED ONE)
1. Determine the following quantities.
CLV1(i,j): the CLV1 of i-th ranked site in City j for the "baseline"
or "start" year.
MAXCLV1G): the largest CLV1 of all sites in City j for the
baseline year.
AMAXCLV1Q'): the largest CLV1 value permitted under the
proposed 1-hr NAAQS.
2. Select five years prior to the baseline year and determine the value
of CLV1 (or related air quality indicator) at each site m in City j for
each year. Rank these values by city and year. Let RANK(m,j,y)
indicate the rank of site m in city j in year y. Let MEANRANK(m j)
indicate the mean value of RANK(m,j,y) over the five years. Rank
the MEANRANK(mj) values and let RELRANK(mj) indicate the
relative rank of MEAN RAN K(m,j).
3. Calculate an adjusted CLV1 for the i-th ranked site in City j by the
expression
ACLVKi.j) = [CLVl(i,j)}[AMAXCLVl(j)]/[MAXCLVl(j)]. (50)
4. If RELRANK(mj) = i, then m will be the i-th ranked site in City j
under attainment. That is,
ACLV1(m,j) = ACLV1(i,j) if RELRANK(mj) = i.
5. The 1-hour data at Site m under attainment will be determined by
adjusting the 1-hour data at Site m in the baseline year. A Weibull
distribution fit to the adjusted data will have a CLV1 equal to
ACLV1(i,j) where i = RELRANK(mj). Subsection 5.3 provides a
method for estimating the parameters of this distribution and for
making the adjustment.
-------
TABLE 35. AIR QUALITY ADJUSTMENT PROCEDURE USED TO SIMULATE
ATTAINMENT OF 8H1EX NAAQS (THE EXPECTED NUMBER OF DAILY
MAXIMUM EIGHT-HOUR OZONE CONCENTRATIONS EXCEEDING
THE SPECIFIED VALUE SHALL NOT EXCEED ONE)
1. Determine the following quantities.
CLV8(i,j): the eight-hour CLV of i-th ranked site in City j for
the "baseline" or "start" year.
MAXCLV8(j): the largest CLV8 of all sites in City j for the
baseline year.
AMAXCLV8(j): the largest CLV8 value permitted under the
proposed 8-hr NAAQS.
2. Select five years prior to the baseline year and determine the value
of CLV8 (or related air quality indicator) at each site m in City j for
each year. Rank these values by city and year. Let RANK(m,j,y)
indicate the rank of site m in city j in year y. Let MEANRANK(mj)
indicate the mean value of RANK(m,j,y) over the five years. Rank
the MEANRANK(mj) values and let RELRANK(m,j) indicate the
relative rank of MEAN RAN K(m,j).
3. Calculate an adjusted CLV8 for the i-th ranked site in City j by the
expression
ACLV8(i,j) = [CLV8(i.,j)][AMAXCLV8(j)]/[MAXCLV8(j)]. (51)
4. If RELRANK(m,j) = i, then m will be the i-th ranked site in City j
under attainment. That is,
ACLV8(m,j) = ACLV8(i,j) if RELRANK(mj) = i.
5. Using Equation 53, estimate the CLV1 associated with each
ACLV8(m,j) value. Denote this value as ACLV1(m,j).
6. The 1-hour data for Site m under attainment of the 8-hr NAAQS will
be determined by adjusting the 1-hour data for Site m in the
baseline year. A Weibull distribution fit to the adjusted data will
have a CLV1 equal to ACLV1(i,j) where i = RELRANK(mj).
Subsection 5.3 provides a method for estimating the parameters of
this distribution and for making the adjustment.
89
-------
TABLE 36. AIR QUALITY ADJUSTMENT PROCEDURE USED TO SIMULATE
ATTAINMENT OF 8H5EX NAAQS (THE EXPECTED NUMBER OF DAILY
MAXIMUM EIGHT-HOUR OZONE CONCENTRATIONS EXCEEDING THE
SPECIFIED VALUE SHALL NOT EXCEED FIVE)
1. Determine the following quantities.
EH6LDM(i,j): the EH6LDM of the i-th ranked site in City j for
the baseline year,
MAXEH6LDMG): the largest EH6LDM of all sites in City j for the
baseline year.
AMAXEH6LDMG): the largest EH6LDM value permitted under the
proposed 1-hr NAAQS.
2. Select five years prior to the baseline year and determine the value
of EH6LDM (or related air quality indicator) at each site m in City j
for each year. Rank these values by city and year. Let
RANK(mj,y) indicate the rank of site m in city j in year y. Let
MEANRANK(mj) indicate the mean value of RANK(m,j,y) over the n
years. Rank the MEANRANK(mj) values and let RELRANK(mj)
indicate the relative rank of MEANRANK(mj).
3. Calculate an adjusted EH6LDM for the i-th ranked site in City j by
the expression
AEH6LDM(i,j) = [EH6LDM(i,j)][(AMAXEH6LDM(j)/[MAXEH6LDM(j)]. (52)
4. If RELRANK(m,j) = i, then m will be the i-th ranked site in City j
under attainment. That is,
AEH6LDM(m,j) = AEH6LDM(i,j) if RELRANK(mj) = i.
5. Using Equation 54, estimate the CLV1 associated with each
AEH6LDM(m,j) value. Denote this value as ACLV1(m,j).
6. The 1-hour data for Site m under attainment of the 8H5EX NAAQS
will be determined by adjusting the 1-hour data for Site m in the
baseline year. A Weibull distribution fit to the adjusted data will
have a CLV1 equal to ACLV1(i,j) where i = RELRANK(mj).
Subsection 5.3 provides a method for estimating the parameters of
this distribution and for making the adjustment.
90
-------
A similar method was employed for 8H5EX standards (Table 36). The value
of AEH6LDM determined in Step 4 was converted to the required ACLV1 value
through the use of an equivalence relationship (Step 5). In this case, the
equivalence relationship was
ACLVI = (RATI02) (AEH6LDM) (54)
where RATI02 varied with city (Table 37).
Through these procedures, a distinct ACLV1 value was assigned to each site
for each standard under evaluation. This ACLV1 value was subsequently used to
construct an attainment one-hour data set using the procedures described in
Subsection 5.3.
5.3 Adjustment of One-Hour Ozone Data Sets
After a site's attainment ACLV1 value was determined, the baseline one-hour
data set associated with the site was adjusted hour-by-hour to create an attainment
one-hour data set. A two-stage adjustment procedure was employed. In the first
stage, the baseline one-hour data were adjusted to produce an initial attainment
data set that had the specified ACLV1 value. In the second stage, the initial data
set was "fine-tuned" to produce a final attainment data set having the exact AQI
value specified for the site.
5.3.1 Initial Adjustment for All Standards
The initial adjustment equation was
yt = (a) (xt)b (55}
where xt was the baseline ozone concentration for hour t and y, was the attainment
ozone concentration for hour t. The terms a and b were "adjustment coefficients"
specific to the site and to the standard being attained.
The adjustment equation was based on the general assumption that Weibull
distributions would provide good fits to the one-hour data sets under baseline and
91
-------
attainment conditions. A Weibull distribution can be completely characterized
through the use of a shape parameter (k) and a scale parameter (
-------
TABLE 37. VALUES FOR EQUIVALENCE RELATIONSHIPS
City
Chicago
Denver
Houston
Los Angeles
Miami
New York
Philadelphia
St. Louis
Washington
RATIO 1a
1.155
1.234
1.374
1.444
1.248
1.178
1.132
1.226
1.179
RATIO2b
1.441
1.453
2.091
1.846
1.513
1.436
1.367
1.506
1.450
aRATIO1 = (ACLV1)/(ACLV8).
bRATIO2 = (ACLV1)/(EH6LDM).
5.3.2 Final adjustment for Eight-hour Standards
When applied to the 8H1EX standards, the initial adjustment procedure
described above produced a one-hour data set with a CLV1 value that exactly
matched the specified CLV1. Because the assumed relationship between CLV1 and
CLV8 was only an approximation, the CLV8 value of the adjusted data set did not
always match the attainment CLV8 value specified for the site. Consequently,
analysts made a final "fine-tuning" adjustment to the one-hour data to obtain the
exact CLV8 value specified. The following final adjustment equation was used.
Adjusted yt = (yt)(Target attainment CLV8)/(lnitial attainment CLV8) (61)
In this equation, yt is the one-hour value for hour t after the initial adjustment
procedure (Equation 55). The "initial attainment CLV8" is the CLV8 value of this
data set. The "target attainment CLV8" is the attainment CLV8 value assigned to
the site by the procedure summarized in Table 35.
A similar fine-tuning procedure was employed for the 8H5EX standards. The
final adjustment equation was
Adjusted y, = (y,)(Target attainment EH6LDM)/(lnitial attainment EH6LDM) (62)
93
-------
The "initial attainment EH6LDM" is the EH6LDM value of the site after the initial
adjustment (Equation 55). The "target attainment EH6LDM" is the attainment
EH6LDM value assigned to the site by the procedure summarized in Table 36.
5.4 Application of the AQAP's to Philadelphia
To test the reasonableness of the AQAP's described above, each was initially
applied to Philadelphia. Three attainment scenarios were evaluated:
1H1EX-120: One-hour daily maximum, one expected exceedance of 120 ppb
8H1EX-80: Eight-hour daily maximum, one expected exceedance of 80 ppb
8H5EX-80: Eight-hour daily maximum, five expected exceedances of 80
ppb.
In each case, baseline conditions were represented by filled-in 1991 ozone data
obtained from the 10 monitoring sites listed in Table 21.
5.4.1 Attainment of1H1EX-120 Standard
The AQAP summarized in Table 34 was applied to Philadelphia for the
purpose of simulating the attainment of the 1H1EX-120 ppb standard. Table 38
presents the results of each step. In this example, baseline conditions in
Philadelphia were assumed to be represented by 1991 ozone data as reported by
the 10 monitoring sites listed for Philadelphia in Table 21.
Analysts initiated the AQAP by fitting a Weibull distribution to the filled-in 1991
one-hour data set associated with each Philadelphia monitoring site. Each fit
produced estimates of the Weibull parameters (k and 6} and the CLV1. The largest
CLV1 for 1991 was associated with District 1 (167 ppb).
To exactly attain the specified NAAQS, the largest CLV1 must equal 120 ppb.
Consequently, Equation 48 (Step 3, Table 34) was implemented as
94
-------
TABLE 38. DETERMINATION OF ADJUSTMENT COEFFICIENTS FOR ONE-HOUR NAAQS
ATTAINMENT (1H1EX-120) IN PHILADELPHIA
District
1
2
3
4
5
6
7
8
9
10
Weibull fit to 1991 1-hr data
k
1.69
2.21
1.96
1.81
2.28
2.23
1.93
2.14
1.74
2.26
6
46.9
56.4
51.0
49.3
56.6
51.2
44.3
51.2
38.1
54.5
CLV1
167
149
153
162
145
134
135
140
131
141
1-hr NAAQS attainment parameters3
Adjusted
CLV1
120
107
110
116
104
96
97
101
94
101
Reassigned
CLV1
107
110
116
120
104
101
94
101
96
97
k'
2.494
2.896
2.546
2.346
3.139
3.194
3.101
3.106
2.809
3.369
6'
45.27
52.44
49.95
48.09
52.51
51.60
47.06
50.62
44.74
51.31
Adjustment
coefficients
a
3.336
2.417
2.420
2.377
2.800
3.305
4.447
3.361
4.694
3.511
b
0.678
0.763
0.770
0.772
0.726
0.698
0.622
0.689
0.619
0.671
(D
cn
aAssumes maximum CLV1 equals 120 ppb.
-------
ACLVl(i,j)*[CLVl(i,j)} ( 120/167 } = [CLVl(i,j] ] (0.719) . (63)
Applying this expression to each 1991 CLV1 produced 10 ACLVI's representing
attainment .conditions. These values are listed in the column labeled "adjusted
CLV1." These values were then reassigned to the Philadelphia districts according to
the five-year ranking determined for each district. Thus, the largest adjusted CLV1
(120 ppb) was assigned to District 4 because District 4 had the highest five-year
ranking. Similarly, the second largest adjusted CLV1 (116 ppb) was assigned to
District 3 because District 3 had the second highest five-year ranking.
In this example, the five-year ranking of each site was determined by
analyzing second-high daily maximum one-hour ozone concentrations reported by
the site over a recent five-year period. Second-high daily maximum values were
used in this step rather than CLVTs because they were easier to obtain from
standard EPA reports.
Analysts next used Equations 56 and 57 to estimate site-specific values for k'
and 6', the values of the Weibull parameters under attainment conditions. For
District 1, the substitution of k = 1.69, ACLV1 = 107 ppb, and n = 5136 produced the
estimates k' = 2.494 and 6' = 45.27 ppb. These values were substituted into
Equations 59 and 60 to produce the values of the adjustment coefficients listed in
Table 38 for District 1 (a = 3.336 and b = 0.678).
A one-hour ozone data set representing attainment conditions was
constructed for each site by applying Equation 55 to the baseline one-hour data set
for the site. Table 39 provides descriptive statistics for the baseline and attainment
data sets associated with District 1.
5.4.2 Attainment of 8H1EX-80 Standard
To evaluate the AQAP for 8H1EX standards, the procedure summarized in
Table 35 was applied to Philadelphia for the purpose of simulating the attainment of
the 8H1EX-80 standard. The results are presented in Table 40. As in the previous
example, baseline conditions for Philadelphia were represented by 1991 ozone data.
96
-------
TABLE 39. DESCRIPTIVE STATISTICS FOR HOURLY-HOUR DATA (PPB)
FOR SITE 34-005-3001 (DISTRICT 1, PHILADELPHIA): BASELINE AND
ATTAINMENT OF THREE OZONE STANDARDS
Statistic
Number of values
Mean
Standard deviation
Minimum
5th percentile
10th percentile
25th percentile
50th percentile
75th percentile
90th percentile
95th percentile
99th percentile
99.5 percentile
99.8 percentile
99.9 percentile
Maximum
Baseline
5136
38
25
0
4
8
19
34
51
72
87
117
124
137
143
156
Attainment of indicated standard
1H1EX-120
5136
37
18
0
9
14
25
36
48
61
69
84
88
94
' 97
102
8H1EX-80
5136
36
14
0
12
17
27
36
45
54
60
71
73
77
78
82
8H5EX-80
5136
34
16
0
7
13
23
34
43
54
61
74
77
82
84
89
Analysts initiated the AQAP by fitting a Weibull distribution to the filled-in 1991
one-hour data set associated with each Philadelphia monitoring site. Each fit
produced estimates of the Weibull parameters (k and 6) and the CLV1. As in the
previous example, the largest CLV1 for 1991 was associated with District 1 (167
ppb).
Analysts next estimated a baseline CLV8 for each site by fitting a Weibull
distribution to the running-average eight-hour data associated with each Philadelphia
monitoring site. The largest CLV8 was 142 ppb (District 1).
To exactly attain the specified NAAQS, the largest CLV8 must equal 80 ppb.
Consequently, Equation 49 (Step 3, Table 35) was implemented as
ACLV8(i,j) = [CLV8U.J)] (80/142) = (CLV8 (i , j) ] (0 . 563 ) . (64)
97
-------
TABLE 40. DETERMINATION OF ADJUSTMENT COEFFICIENTS FOR EIGHT-HOUR NAAQS
ATTAINMENT (8H1EX-80) IN PHILADELPHIA
CD
CD
District
1
2
3
4
5
6
7
8
9
10
Weibull fits to 1991 data
1-h k
1.69
2.21
1.96
1.81
2.28
2.23
1.93
2.14
1.74
2.26
1-h 6
46.9
56.4
51.0
49.3
56.6
51.2
44.3
51.2
38.1
54.5
CLV1
167
149
153
162
145
134
135
140
131
141
CLV8
142
136
128
138
120
118
123
128
116
126
8-hr NAAQS attainment
parameters8
Adjuste
d
CLV8
80
77
72
78
68
66
69
72
65
71
Reassigned
CLV8
72
77
78
80
72
69
65
71
66
68
Equivalent
CLV1
82
87
88
91
82
78
74
80
75
77
1-h Weibull
pgrameters
k'
3.173
3.725
3.339
3.057
4.119
4.237
3.941
3.960
3.518
4.359
6'
41.45
49.01
46.44
44.89
48.41
47.08
42.70
46.75
40.60
47.06
Adjustment
coefficients
a
5.339
4.481
4.618
4.465
5.183
5.932
6.671
5.572
6.708
5.922
b
0.533
0.593
0.587
0.592
0.554
0.526
0.490
0.540
0.495
0.518
"Assumes maximum CLV8 equals 80 ppb.
-------
Analysts applied .this expression to each 1991 CLV8 to obtain 10 ACLVS's
representing attainment conditions. These values are listed in the column labeled
"adjusted CLV8." These values were then reassigned to the Philadelphia districts
according to the five-year ranking determined for each district. The resulting
assignments are listed in Table 40 under the heading "reassigned CLV8."
Each reassigned CLV8 was then converted into an equivalent attainment
CLV1 using Equation 53 with the RATIO1 value for Philadelphia (1.132). For
example, the reassigned CLV8 for District 1 (72 ppb) was multiplied by 1.132 to
produce an equivalent attainment CLV1 of 82 ppb.
Analysts next used Equations 56 and 57 to estimate site-specific values for k1
and 6', the values of the Weibull parameters for one-hour data under attainment
conditions. For District 1, the substitution of k = 1.69, ACLV1 = 82 ppb, and n =
5136 produced the estimates k' = 3.173 and 6' = 41.45 ppb. These values were
substituted into Equations 59 and 60 to produce the values of the adjustment
coefficients listed in Table 40 for District 1 (a = 5.339 and b = 0.533). These
coefficients were then substituted into Equation 55 to produce an initial one-hour
data set approximating attainment conditions.
The one-hour data were processed to produce a corresponding 8-hour
running average data set. A Weibull distribution was next fit to the adjusted eight-
hour data for the site to determine an initial attainment CLV8. Analysts then used
Equation 61 to make the final "fine-tuning" adjustment to the one-hour data
necessary to achieve the target CLV8 specified for the site (72 ppb). The resulting
one-hour data set was assumed to represent attainment conditions for District 1.
Table 39 provides descriptive statistics for this data set. Attainment data sets were
developed in a similar manner for each of the other Philadelphia monitoring sites.
5.4.3 Attainment of 8H5EX-80 Standard
The AQAP for 8H5EX standards (Table 36) was applied to Philadelphia for
the purpose of simulating the attainment of the 8H5EX-80 standard. The results are
presented in Table 41.
99
-------
As in the two previous examples, baseline conditions for Philadelphia were
represented by 1991 ozone data. Analysts began the AQAP by fitting a Weibull
distribution to the filled-in 1991 one-hour data set associated with each Philadelphia
monitoring site. Each fit produced estimates of the Weibull parameters (k and 6)
and the CLV1. The largest CLV1 for 1991 was associated with District 1 (167 ppb).
Analysts next determined a baseline EH6LDM value for each site by first
calculating all eight-hour daily maximum concentrations in the associated one-hour
data set and then identifying the sixth largest value. The largest EH6LDM was 116
ppb (District 1).
The largest EH6LDM value permitted under the 8H5EX-80 standard is 80
ppb. As the largest baseline EH6LDM was 116 ppb, Equation 52 (Table 36) was
expressed as
AEH6LDM(i,j}=[EH6LDM(i,j}} (80/116)=[EH6LDM(i,j)](0.563). (65)
Analysts applied this expression to each 1991 EH6LDM to obtain 10 AEH6LDM's
representing attainment conditions. These values are listed in the Table 41 column
labeled "adjusted EH6LDM." Analysts next reassigned the values to the
Philadelphia districts according to the five-year ranking determined for each district.
The resulting assignments are listed in Table 41 under the heading "reassigned
EH6LDM."
Each reassigned EH6LDM was then converted into an equivalent attainment
CLV1 using Equation 54 with the RATIO2 value for Philadelphia (1.367). In the
case of District 1, the reassigned EH6LDM (74 ppb) was multiplied by 1.367 to
produce an equivalent attainment CLV1 of 101 ppb.
Analysts next used Equations 56 and 57 to estimate site-specific values for k'
and <$', the values of the Weibull parameters for one-hour data under attainment
conditions. For District 1, the substitution of k = 1.69, ACLV1 =101 ppb, and n =
5136 produced the estimates k' = 2.626 and 8 = 44.62 ppb. These values were
substituted into Equations 59 and 60 to produce the values of the adjustment
coefficients listed in Table 41 for District 1 (a = 3.750 and b = 0.644). These
100
-------
TABLE 41. DETERMINATION OF ADJUSTMENT COEFFICIENTS FOR EIGHT-HOUR NAAQS
ATTAINMENT (EH6LDM = 80 ppb) IN PHILADELPHIA
District
1
2
3
4
5
6
7
8
9
10
Parameters of 1991 data
1-h k
1.69
2.21
1.96
1.81
2.28
2.23
1.93
2.14
1.74
2.26
1-h 6
46.9
56.4
51.0
49.3
56.6
51.2
44.3
51.2
38.1
54.5
CLV1
167
149
153
162
145
134
135
140
131
141
EH6LDM
116
113
111
115
107
101
102
104
90
102
8-hour NAAQS attainment parameters
Adjusted
EH6LDM
80
78
77
79
74
70
70
72
62
70
Rank
5
2
1
3
4
6
8
9
10
7
Reassigned
EH6LDM
74
79
80
78
77
72
70
70
62
70
Equivalent
CLV1
101
108
109
107
105
98
96
96
85
96
1-hour Weibull
parameters
k'
2.626
2.954
2.708
2.615
3.106
3.300
3.038
3.277
3.136
3.408
6'
44.62
52.24
49.37
47.10
52.64
51.16
47.38
49.88
42.89
51.15
Adjustment
coefficients
a
3.750
2.556
2.869
3.171
2.721
3.581
4.261
3.817
5.689
3.608
b
0.644
0.748
0.724
0.692
0.734
0.676
0.635
0.653
0.555
0.663
"Assumes maximum EH6LDM equals 80 ppb.
-------
coefficients were then substituted into Equation 55 to produce an initial one-hour
data set approximating attainment conditions.
The one-hour data were processed to produce a corresponding 8-hour
running average data set. These data were analyzed to determine an initial
attainment EH6LDM. Analysts then employed Equation 62 to make the final "fine-
tuning" adjustment to the one-hour data necessary to achieve the target attainment
EH6LDM specified for the district (101 ppb). The resulting tuned data set was
assumed to represent attainment conditions for District 1. Table 39 presents
descriptive statistics for this data set. Attainment data sets were developed in a
similar manner for each of the other Philadelphia monitoring sites.
5.5 Special Adjustment Procedures Applied in Selected Attainment Scenarios
The AQAP's described above were developed by comparing the ozone data
reported by a site in a high ozone year with ozone data reported by the same site in
a low ozone year. Consequently, the AQAP's are expected to perform best when
used to simulate a significant reduction in the ozone levels at a site. The results of
an analysis of AQAP performance by ITAQS suggested that the AQAP's described
above may produce unrealistic data sets for Denver, Chicago, and Miami when used
to simulate a small reduction in ozone levels or when used to simulate an increase
in ozone levels. For this reason, ITAQS used a different set of AQAP's for all
attainment scenarios in the Chicago, Denver, and Miami study areas. The Chicago
scenarios generally required small decreases in ozone levels to exactly meet the
specified attainment conditions. The Denver and Miami scenarios required small
changes in both directions.
In the alternative AQAP's for the 1H1EX-120 and the 1H1EX-100 scenarios,
the procedures summarized in Table 34 were followed to the point in Step 5 where
the reader is directed to Section 5.3. The procedures in Section 5.3 were not
employed to adjust the one-hour data; instead, each value of the adjusted data set
was estimated by the expression
yt = (c] (xt] (66)
102
-------
where xt was the -.baseline ozone concentration for hour t and y, was the attainment
ozone concentration for hour t. The value of c was determined by the expression
C= (ACLVl) / (CLVl) (67)
where ACLV1 is the characteristic largest one-hour value of the site before
adjustment and ACLV1 is the characteristic largest one-hour value assigned to the
site in Step 4 to represent attainment conditions.
In a similar manner, the alternative AQAP's for the 8H1EX-70, 8H1EX-80,
8H1EX-90, and 8H1EX-100 scenarios followed the procedures summarized in Table
35 to the point in Step 6 where the reader is directed to Section 5.3. Again, the
procedures in Section 5.3 were not employed to adjust the one-hour data. Instead,
an initial estimate of each value of the adjusted data set was estimated according to
Equations 66 and 67. For Chicago and Miami, ACLV1 was the characteristic largest
one-hour value assigned to the site in Step 5 of Table 35. For Denver, however, the
observed second highest daily maximum value was assigned to the site in Step 5 of
Table 35, instead of the ACLV1. The observed second highest daily maximum was
used as the air quality indicator in Denver because it provided a better
representation of the data than the characteristic largest one-hour value provided.
The alternative adjustment procedure for all three cities was completed by applying
Equation 61 to the data to make a final "fine-tuning" adjustment.
The alternative AQAP's for the 8H5EX-80 and 8H5EX-90 scenarios followed
the steps listed in Table 36 to the point in Step 6 where the reader is directed to
Section 5.3. The applicable procedures in Section 5.3 were again omitted; instead,
Equations 66 and 67 were employed to make an initial estimate of each value of the
adjusted data set. In Equation 67, ACV1 was the characteristic largest one-hour
value assigned to the site in Step 5 of Table 36. The adjustment procedure was
completed by using Equation 62 to make the final fine tuning adjustment.
103
-------
SECTION 6
PREPARATION OF OUTDOOR CHILDREN DATA BASES
As previously described in Section 2 of this report, a special version of
pNEM/OS was used to estimate the exposures of outdoor children residing in nine
study areas under various air quality scenarios. In these exposure assessments,
the outdoor children in each study area were represented by a collection of cohorts.
The distribution of ozone exposures across the outdoor children population of each
study area was equal to the sum of the exposures of the individual cohorts.
To simulate the ozone exposures of a particular cohort, the pNEM/O3 model
required an exposure event sequence for the cohort and an estimate of the number
of people represented by the cohort. The exposure event sequence was
constructed by sampling a special time/activity database containing activity diary
data obtained from seven studies. Analysts estimated the population of each cohort
by applying a percentage to the total population of children in each of the nine study
areas. These percentages were determined from the activity diary data and
represented that part of the total population of children which would be considered
active outdoors. This section describes the procedures employed to create the
time/activity database, to construct an exposure event sequence for each cohort,
and to estimate the number of children in each cohort.
6.1 Selection of Time/Activity Data
Previous applications of pNEM/O3 have employed activity diary data obtained
from the CADS20. In the outdoor children exposure analysis, analysts augmented
the CADS data with diary data from six other time/activity studies (see Table 2).
These seven studies are a subset of 10 studies identified by Johnson et al.48 as
generally appropriate for use in exposure assessments. The remaining three
studies listed by Johnson et al. (Denver, Los Angeles - outdoor workers, and Los
104
-------
Angeles - construction workers) did not provide any data representative of outdoor
children. Appendix A provides a brief description of each of the 10 studies.
Under the direction of EPA, ITAQS developed a procedure for identifying
outdoor children among the subjects of the seven time/activity studies listed in Table
2. First, analysts identified the codes (designated "microenvironment" codes) used
in each study to indicate diary entries associated with outdoor microenvironments.
A subject was designated an active child if the subject was associated with at least
one person-day of diary data in which the child spent a specified amount of time
outdoors.
The specified amount of time outdoors varied by season and
weekend/weekday designation. A child was defined as "outdoor" if
During a winter weekday the child had at least one diary day where
he/she spent 120 minutes or more outdoors, or
During a winter weekend the child had at least one diary day where
he/she spent 180 minutes or more outdoors, or
During a summer day (weekday or weekend) the child had at least one
diary day where he/she spent 270 minutes or more outdoors.
For this analysis, summer was defined as June, July, and August, and winter as all
other months. This procedure produced a pool containing 479 outdoor children with
792 person-days of activity diary data (Table 42).
6.2 Processing of Time/Activity Data
In a typical pNEM analysis, the ozone exposure of each cohort is determined
by the cohort's exposure event sequence. An exposure event sequence consists of
a series of person-days with each person-day further divided into a series of
exposure events. Each exposure event specifies a start time, an event duration, a
microenvironment, a breathing rate category, and a home district location. Exposure
event sequences are constructed by sampling person-days from a prepared
time/activity database according to a set of selection rules.
105
-------
TABLE 42. CHARACTERISTICS OF ACTIVITY DATA FOR
OUTDOOR CHILDREN
Study
Cincinnati
Washington, D.C.
California -12 and over
California - 1 1 and under
Los Angeles - Elementary
School
Students
Los Angeles - High School
Students
Valdez
Total
Number of person-days
384
3
54
257
38
47
9
792
Number of persons
130
3
54
257
13
13
9
479
106
-------
In the special pNEM/O3 analysis of outdoor children described here, each
exposure event sequence was constructed by sampling a time/activity database
containing 792 person-days of diary data drawn from seven studies. To create this
database, analysts first defined a standard data format which met the input
requirements of the exposure model. The diary data obtained from each study were
then converted into an equivalent data set with the specified format.
The standard format was designed to easily accommodate CADS data, as
data from this study had been used in the majority of previous pNEM analyses. As
three of the seven studies selected for the outdoor children analysis employed the
CADS diary (Cincinnati, Los Angeles - elementary students, and Los Angeles - high
school students), data from these studies required minimal processing to be
included in the time/activity database. The data obtained from the remaining four
studies (Washington, California - 12 and over, California - 11 and under, and
Valdez) required significant processing. None of these four studies characterized
diary entries according to a breathing rate category. Consequently, researchers
developed a Monte Carlo technique to assign breathing rate categories to diary
entries obtained from the these four studies.
The Monte Carlo technique employed assignment probabilities which varied
according to four event descriptors: activity type, microenvironment, time of day, and
duration. These descriptors were identified by Johnson et al.34 as influencing
exertion levels associated with diary events. To estimate assignment probabilities
relative to these descriptors, each event in the CADS database was categorized
according to the following indices:
Activity class
A: high probability of fast breathing rate
B: moderate probability of fast breathing rate
C: low probability of fast breathing rate
D: sleeping
107
-------
Microenvironment
1: Indoors - residence
2: Indoors - other
3: Outdoors
4: In vehicle
Time of day
1: 0700 to 1659
2: 1700 to 0659
Duration
1: 0 to 20 minutes
2: Greater than 20 minutes
The rnicroenvironment classification was determined by the location code (e.g.,
school) associated with the event in the CADS database. The time of day
classification was determined by the start time of the event.
The activity classification consisted of three waking classes (A, B, and C) and
one sleeping class (D). Activities were assigned to these classes based on the
likelihood that the activity would be associated with a fast breathing rate, with
Classification D being reserved for sleeping activities. Table 43 matches each
CADS activity code to one of the four activity classes (A, B, C, or D). These
matchups are based primarily on the results of an analysis of the CADS database
performed by Johnson in 199214.
ITAQS created a data group for each of the 48 combinations of activity class,
rnicroenvironment, time of day, and duration which could be specified using only the
three non-sleeping activity classes (A, B, and C). Each diary entry in the CADS
database was assigned to one of the 48 groups. Within each data group, the diary
entries were further identified by breathing rate category (slow, medium, or fast).
Table 44 lists the number of diary entries in each of the 48 groups which
were placed in each of the three breathing rate categories and the corresponding
cumulative fractions. For example, the group identified as Activity Class = A,
108
-------
TABLE 43. BREATHING RATE CATEGORIES OF ACTIVITIES IN THE
CINCINNATI STUDY
Activity
Code
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
Description of Activity
All destination - oriented travel
(including walking)
Income - related work
Day - care
Kindergarten - 12th grade
College or trade school
Adult education and special training
Homework
Meal Preparation and cleanup
Laundry
Other indoor chores
Yard work and outdoor chores
Child care and child - centered
activities
Errands and shopping
Personal care outside home (doctor,
hair dresser)
Eating
Sleeping
Other personal needs
Religious activities
Meetings of clubs, organizations,
committees, etc.
Other collective participation
Breathing
Rate
Category
B
B
C
C
C
C
C
C
B
B
A
C
C
C
C
D
C
C
C
C
(continued)
109
-------
Table 43 (continued)
Activity
Code
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
Description of Activity
Spectator sports events
Movies, concerts, and other
entertainment events outside home
Cafe, bar, tea room
Museums and exhibitions
Parties and receptions
Visiting friends
Recess and physical education
Active sports and games outside
school, including exercises and
aerobics
Hunting, fishing, hiking
Jogging or bicycling
Taking a walk
Artistic creations, music, and hobbies
Other active leisure
Reading
Television or radio
Conversation and correspondence
Relaxing, reflecting, thinking (no visible
activity)
Other passive leisure
Asthma attack
Other sudden illness or injury
Breathing
Rate
Category
B
C
C
B
B
C
A
A
A
A
A
C
A
C
C
C
C
C
C
C
(continued)
110
-------
Table 43 (continued)
Activity
Code
43
44
45
Description of Activity
Interview
Wakeup
Baby crying
Breathing
Rate
Category
C
C
A
Microenvironment = 1, Time of Day = 1, and Duration = 1 contained 418 events
(see first entry in Table 44). These 418 events were apportioned among the three
breathing rate categories as follows:
Breathing Rate Number
Fraction
Slow
Moderate
Fast
262
122
34
0.63
0.29
0.08
Cumulative Fraction
0.63
0.92
1.00
In this example, 63 percent of the events were characterized as slow, 29 percent as
moderate, and 8 percent as fast.
Researchers developed a Monte Carlo algorithm to assign breathing rate
categories to events obtained from the four diary studies which did not report
breathing rate categories. Each event from one of these studies was indexed
according to activity class (Tables 45 through 48), microenvironment, time of day,
and duration. The algorithm generated a random number for each event which was
compared to the cumulative fractions listed in Table 44 for the particular combination
of indices.
For example, the random number generated for an event identified as Activity
Class = A, Microenvironment = 1, Time of Day = 1, and Duration = 1 would be
compared to the cumulative fractions listed in the first row of Table 44. If the
random number was between 0 and 0.63, the algorithm would assign a slow
breathing rate to the event. The algorithm would assign a moderate breathing rate
to events with
111
-------
TABLE 44. CUMULATIVE BREATHING RATE CATEGORY PROBABILITIES FROM THE CINCINNATI
ACTIVITY-DIARY STUDY BY ACTIVITY CLASS, MICROENVIRONMENT, TIME OF DAY CATEGORY, AND EVENT
DURATION CATEGORY
Activity
class
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
Micro-
environment
1
1
1
1
2
2
2
2
3
3
3
3
4
4
4
Time of day
category
1
1
2
2
1
1
2
2
1
1
2
2
1
1
2
Event
duration
category
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
Cumulative probability of assigning breathing rate
categories (number of events used to determine
percentage)
Low (2)
0.63 (262)
0.78 (589)
0.60 (152)
0.66 (327)
0.20 (25)
0.25 (56)
0.20 (10)
0.24 (47)
0.33 (367)
0.29 (536)
0.32 (235)
0.27 (336)
1.00 (2)
1.00 (3)
0.00 (0)
Medium (3)
0.92 (122)
0.97 (141)
0.89 (74)
0.93 (138)
0.63 (55)
0.64 (90)
0.80 (29)
0.79 (105)
0.86 (599)
0.88 (1,071)
0.88 (413)
0.86 (757)
NAa (0)
NA(0)
NA (0)
High (4)
1.00 (34)
1.00 (26)
1.00 (28)
1.00 (34)
1.00 (48)
1.00 (81)
1.00 (10)
1.00 (40)
1.00 (163)
1.00 (229)
1.00 (87)
1.00 (173)
NA (0)
NA (0)
NA (0)
(continued)
-------
TABLE 44 (Continued)
Activity
class
A
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
Micro-
environment
4
1
1
1
1
2
2
2
2
3
3
3
3
4
4
4
Time of day
category
2
1
1
2
2
1
1
2
2
1
1
2
2
1
1
2
Event
duration
category
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
Cumulative probability of assigning breathing rate
categories (number of events used to determine
percentage)
Low (2)
1.00 (1)
0.71 (757)
0.80 (1,582)
0.75 (448)
0.86 (691)
0.68 (970)
0.90 (3,762)
0.80 (313)
0.81 (824)
0.63 (5,854)
0.53 (361)
0.67 (3,876)
0.72 (330)
1.00 (5,264)
1.00 (1,848)
1.00 (3,358)
Medium (3)
NA (0)
0.99 (298)
1.00 (382)
1.00 (150)
1.00(110)
0.99 (449)
1.00 (401)
0.99 (74)
0.99 (184)
0.99 (3,366)
0.96 (298)
0.99 (1,902)
0.98 (118)
NA (5)
NA (1)
NA (0)
High (4)
NA (0) I
1.00 (7)
NA (4)
NA (1)
NA (3)
1.00(12) I
NA (14)
1.00(5)
1.00 (7) I
1.00(87)
1.00 (25)
1.00 (41)
1.00 (8)
NA (0)
NA (0)
NA (0)
(continued)
-------
TABLE 44 (Continued)
Activity
class
B
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
Micro-
environment
4
1
1
1
1
2
2
2
2
3
3
3
3
4
4
4
Time of day
category
2
1
1
2
2
1
1
2
2
1
1
2
2
1
1
2
Event
duration
category
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
Cumulative probability of assigning breathing rate
categories (number of events used to determine
percentage)
Low (2)
1.00 (1,266)
0.99 (7,807)
0.99 (8,024)
0.99 (6,644)
1.00 (10,861)
0.96 (2,559)
0.98 (4,032)
0.97 (894)
0.99 (1,236)
0.91 (505)
0.95 (419)
0.94 (331)
0.94 (480)
1.00 (12)
1.00 (10)
1.00 (13)
Medium (3)
NA (2)
1.00 (87)
1.00 (48)
1.00 (75)
NA (41)
1.00 (117)
1.00 (88)
1.00 (32)
1.00 (17)
1.00 (46)
0.99 (17)
0.99 (18)
1.00 (30)
NA (0)
NA (0)
NA (0)
High (4)
NA (0)
NA (0)
NA (0)
NA (1)
NA (1)
NA (2)
NA (0)
NA (0)
NA (0)
NA(2)
1.00 (3)
1.00 (2)
NA (2)
NA (0)
NA (0)
NA (0)
(continued)
-------
TABLE 44 (Continued)
Activity
class
C
Micro-
environment
4
Time of day
category
2
Event
duration
category
2
Cumulative probability of assigning breathing rate
categories (number of events used to determine
percentage)
Low (2)
1.00 (5)
Medium (3)
NA (0)
High (4)
NA(0)
"Not applicable.
-------
TABLE 45. ACTIVITY CLASSES ASSIGNED TO ACTIVITY CODES USED
IN THE CALIFORNIA DIARY STUDY
Activity
code
Description of activity
Activity
class
1
2
3
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
Work - income related at-tand away-from-home
Unemployment - job search, welfare activities
Travel during work
Other paid work - second job, part-time youth job
Eating at work - lunch, coffee while working
Activities at work - before and after work day - i.e.
conversations
Breaks - coffee breaks
Travel to/from work or job-search travel
Food preparation - cooking, serving, preserving
Food cleanup - cleaning table, dishes
Cleaning house - mainly indoor
Outdoor cleaning - yard work, garbage, snow, etc.
Clothes care - laundry, other clothes care
Car repair/maintenance - oil, tires, body work, etc.
General repairs: indoor, outdoor, carpentry, painting
Plant care - outdoor garden, houseplants
Pet and animal care - domestic, feeding livestock
Other household - garage sale, packing, groceries, chores
Baby care - feeding, etc. to children < 4
Child care - children between 5 and 17
Helping/teaching - children with homework, hobbies
Talking/reading - discipline (to children), conversing,
listening
Indoor playing with baby, children
Outdoor playing - playing, coaching children
Medical care - child
Other child care - coordinating non-school activities,
Babysitting
Dry cleaning activities - pick up/drop off
Travel related to child care (including walking)
Everyday shopping
Durable good/house shopping
Personal care services
Medical appointments
Government/financial services (errands too)
Car repair services - buying gas, etc.
Other repairs - errands for: clothes, appliances
B
B
B
B
C
C
C
B
C
C
B
A
B
B
B
B
B
B
C
C
C
C
C
B
C
C
C
B
C
C
C
C
C
C
C
(continued)
116
-------
TABLE 45 (Continued)
Activity
code
Description of activity
Activity
class
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
54
55
56
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
Other services - lawyer, video pick up, etc. (errand related)
Errands
Travel related to goods and services
Washing - personal hygiene
Medical care - at home
Help and care - to relatives, i.e., moving neighbors
Meals at home
Meals out (friends' or at restaurant)
Night sleep
Naps/sleep
Dressing, grooming
Not ascertained activities
Travel related to personal care
Students' classes
Other classes - lectures, professional, tutor
Doing homework - reading, studying, research
Using library
Other education
Travel related to education
Work for professional/union organizations
Work for special interest identity organizations
Work for political party and civic participation
Work for volunteer/helping organizations
Work for religious groups
Religious practice
Work for fraternal organizations
Work for child/youth/family organizations
Work for other organizations
Travel related to organizational activity
Sports events - attending as spectator
Miscellaneous events - circus, fairs, rock concerts
Movies
Attending theater
Visiting museums
Visiting - socializing with friends
C
C
B
C
C
B
C
C
D
D
C
C
B
C
C
C
C
C
B
C
C
C
C
C
C
C
C
C
B
B
B
C
C
B
C
(continued)
117
-------
TABLE 45 (Continued)
Activity
code
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
Description of activity
Parties and picnicking
Bars/lounges
Other social events
Travel related to event/social activities
Active sports
Outdoor leisure - hunting, fishing, boating, camping, etc.
Walking/biking/hiking/jogging, etc.
Hobbies - photography, scrapbooks, etc.
Domestic crafts - knitting, sewing, quilting
Art - sculpture, painting, potting drawing
Music/drama/dance/active leisure
Games - card, board, computer
Computer use
Travel related to active leisure
Radio use
TV use
Records/tapes
Read books
Reading magazines/not ascertained
Reading a newspaper
Conversations
Letters, writing, paperwork
Other passive leisure
Travel related to passive leisure
Activity
class
B
C
C
B
A
B
A
C
C
C
A
C
C
B
C
C
C
C
C
C
C
C
C
C
118
-------
TABLE 46. ACTIVITY CLASSES ASSIGNED TO ACTIVITY CODES USED
IN THE DENVER DIARY STUDY
Activity
code
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
Description of activity
All travel
Work (income-related) and study
Cooking
Laundry
Other indoor chores and child care
Yard work and other outdoor activities
Errands and shopping
Eating
Sleeping
Other personal needs
Social, political or religious activities
Cafe or pub
Walking, bicycling, or jogging (not in transit)
Other leisure activities
Uncertain of applicable code
No entry in diary
Interview
Final entry
Autolog value (i.e., hourly value automatically logged by
PEM)
Begin breath sample
End breath sample
Activity
class
B
C
C
B
B
A
C
C
D
C
C
C
A
C
C
NA
C
C
C
C
C
119
-------
TABLE 47. ACTIVITY CLASSES ASSIGNED TO ACTIVITY CODES USED
IN THE VALDEZ DIARY STUDY
Activity
code
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
99
Description of activity
Cooking
Eating
Driving car, truck, bus
Driving boat
Driving plane
Driving other
Biking
Sedentary activity
Physical activity
At school
Grooming, dressing
Socializing
Shopping, errands
Going to bed
Getting out of bed
Exercising
Walking
At work
Fishing
Pumping gasoline
Not specified
Playing
At dock
Interview
Activity
class
C
C
B
B
B
B
A
C
A
C
C
C
C
D
C
A
B
B
B
B
C
B
B
C
120
-------
TABLE 48. ACTIVITY CLASSES ASSIGNED TO ACTIVITY CODES USED
IN THE WASHINGTON DIARY STUDY
Activity
code
1
2
3
4
5
6
7
8
9
11
12
13
14
15
16
17
18
19
21-36
77
86
87
88
89
Description of activity
Transit, travel
Work, business meeting
Cooking
Laundry
Inside house - chores
Outside house - chores
Errands, shopping, etc.
Personal activities
Leisure activities
Sleeping
School, study
Eating, drinking
Sports and exercise
Church, political meetings, etc.
Inside house - miscellaneous
In parking garage or lot
Outside, not otherwise specified
Doctor or dentist office
Same as activities 1-16 including suspected sleep
Same as activity 87 including suspected sleep
Dummy start diary
Start diary
End diary
Any other activity
Activity
class
B
B
C
B
B
A
C
C
C
D
C
C
A
C
C
B
B
C
C
C
C
C
random numbers between 0.63 and 0.92; similarly, fast breathing rates would be
assigned to events with random numbers between 0..92 and 1.00.
The cumulative fractions listed in Table 44 were used by the Monte Carlo
algorithm to process all diary events associated with waking activities. When the
activity code for a diary entry indicated that the subject was sleeping during the
event (i.e., activity class = D), the algorithm always assigned the fourth breathing
rate category (sleeping) to the event.
As indicated above, 792 person-days of diary data representing 479 outdoor
children were processed and combined into a database suitable for input into
121
-------
pNEM/O3. Subsection 2.3 describes the algorithm used by pNEM/O3 to sample this
database and construct an exposure event sequence for each cohort.
6.3 City-Specific Outdoor Children Populations
In applying pNEM/03 to the outdoor children in a study area, analysts
employed Equation 6 in Subsection 2.5 to estimate the number of children
represented by each cohort. This equation in turn required the estimation of a value
for the P(g) term in Equation 4. P(g) was defined as the fraction of children in
demographic group g who were "outdoor children." The demographic group was
either preteens (children ages 6 to 13) or teenagers (children 14 to 18).
In the analyses described in this report, P(g) was assumed to be constant
across all cohorts belonging to demographic group g, regardless of study area. P(g)
was estimated by the expression
P(g) = [POPOC(g,ddb]} / (POPC(g, ddb) } (63)
where
P(g) = the fraction of outdoor children in demographic
group g.
POPOC(g.ddb) = the number of children in demographic group g
from the diary data bases (ddb) that were classified
as "outdoor children."
POPC(g.ddb) = the total number of children in demographic group g
from the diary data bases (ddb).
The values of POPOC(g.ddb) and POPC(g.ddb) were obtained from an analysis of
time/activity databases obtained from three of the studies listed in Table 2:
California - 11 and under, California - 12 and over, and Cincinnati. Each of these
studies employed a random selection procedure to enroll a relatively large number
of subjects.
Considered together, the three studies provided diary data for 771 preteens
and 258 teenagers. Of the 771 preteens, 361 (46.8 percent) were judged to be
active outdoors according to the criteria discussed in Subsection 6.1. In a similar
122
-------
manner, 80 of the 258 teenagers (31.0 percent) were judged to be active outdoors.
Consequently, analysts set P(g) equal to 0.468 for preteens and 0.310 for
teenagers. These estimates were multiplied by census-derived estimates for the
total number of preteens and teenagers in each study area to produce the estimates
listed in Table 49. The populations of individual cohorts were estimated using
Equations 4 through 6.
123
-------
TABLE 49. ESTIMATED NUMBER OF OUTDOOR
IN EACH STUDY AREA
CHILDREN
Study area
Chicago
Denver
Houston
Los Angeles
Miami
New York
Philadelphia
St. Louis
Washington, DC
Demographic
group
Preteens
Teenagers
Total
Preteens
Teenagers
Total
Preteens
Teenagers
Total
Preteens
Teenagers
Total
Preteens
Teenagers
Total
Preteens
Teenagers
Total
Preteens
Teenagers
Total
Preteens
Teenagers
Total
Preteens
Teenagers
Total
Total number
of children
722,861
433,639
1,156,500
165,679
93,934
259,613
309,886
180,013
489,899
1,216,936
737,950
1,954,886
203,346
124,050
327,396
1,180,573
742,235
1,922,808
419,237
255,194
674,431
197,617
115,360
312,977
301,827
185,767
487,594
Multiplier
[P(g)j
0.468
0.310
-
0.468
0.310
-
0.468
0.310
-
0.468
0.310
-
0.468
0.310
-
0.468
0.310
-
0.468
0.310
-
0.468
0.310
-
0.468
0.310
-
Estimated number
of outdoor children
338,290
134,420
472,710
77,540
29,125
106,665
144,995
55,800
200,795
569,515
228,775
798,290
95,155
38,455
133,610
552,515
230,085
782.600
196,215
79,105
275,320
92,480
35,770
128,250
141,265
57,595
198,860
124
-------
SECTION 7
OZONE EXPOSURE ESTIMATES FOR NINE URBAN AREAS
The enhanced pNEM/03 methodology described in this report was applied to
the nine urban areas listed earlier in Table 1. The result of each application was a
set of 18 exposure summary tables for each regulatory scenario under evaluation.
This section describes the scenarios that were analyzed, provides a guide to the
interpretation of output tables, and summarizes the principal results of each
exposure assessment.
7.1 Regulatory Scenarios
The following regulatory scenarios were examined in applying
pNEM/OS to each study area.
Baseline Ambient ozone conditions were represented by unadjusted fixed-
site monitoring data as reported for the exposure period listed in
Table 1. These data were assumed to represent ambient ozone
levels typical of "as is" air quality conditions.
1H1EX One hour daily maximum - one expected exceedance: the
expected number of daily maximum one-hour ozone
concentrations exceeding the specified value shall not exceed
one.
Standard levels: 100 ppb, 120 ppb (the current NAAQS for
ozone)
8H1EX Eight-hour daily maximum - one expected exceedance: the
expected number of daily maximum eight-hour ozone
concentrations exceeding the specified value shall not exceed
one.
Standard levels: 70 ppb, 80 ppb, 90 ppb, 100 ppb
125
-------
8H5EX Eight-hour daily maximum - five expected exceedances: the
expected number of daily maximum eight-hour ozone
concentrations exceeding the specified value shall not exceed
five.
Standard levels: 80 ppb, 90 ppb,
Section 5 describes the procedures used to adjust baseline data to simulate
attainment of 1H1EX, 8H1EX, and 8H5EX standards.
7.2 Formats of the Exposure Summary Tables
Appendix D contains exposure summary tables for the outdoor children
population obtained from a sample application of pNEM/O3 to Houston. The tables
are organized according to the following table formats. (Note that the table numbers
listed under each format refer to the tables in Appendix D.)
Number of people - cumulative exposures (or doses) by EVR range
These tables list estimates by ozone concentration and EVR range. Each
table entry lists the number of outdoor children who experienced one or more ozone
exposures (or doses) during which the ozone concentration was at or above the
level indicated by the row label and the average EVR was within the range indicated
by the column heading. Separate tables provide estimates for one-hour exposures
(Table 1 in Appendix D), one-hour daily maximum exposures (Table 1A), one-hour
daily maximum doses (Table 1B), eight-hour daily maximum exposures (Table 4),
and eight-hour daily maximum doses (Table 4A).
Number of people -- cumulative seasonal mean exposures
Table 7 in Appendix D lists estimates by ozone concentration only. Each
entry lists the number of outdoor children who were associated with a seasonal
mean exposure at or above the ozone level indicated by the row label. The
seasonal mean is calculated as the average of the eight-hour daily maximum ozone
exposures occurring from April to October, inclusive.
Number of occurrences exposures (or doses) by EVR range
These tables list estimates arranged by ozone concentration range and EVR
range. Each table entry lists the number of times an outdoor child experienced an
ozone exposure during which the ozone concentration was within the range
126
-------
indicated by the row label and the average EVR was within the range indicated by
the column heading. There are separate tables for one-hour exposures (Table 2 in
Appendix D), one-hour daily maximum exposures (Table 2A), one-hour daily
maximum doses (Table 2B), eight-hour daily maximum exposures (Table 5), and
eight-hour daily maximum doses (Table 5A).
Number of occurrences -- seasonal mean exposures
Table 8 in Appendix D presents estimates by ozone range only. Each entry
lists the number of times an outdoor child experienced a seasonal mean exposure at
or above the ozone level indicated by the row label. The seasonal mean is
calculated as the average of the eight-hour daily maximum ozone exposures
occurring from April to October, inclusive.
Number of people -- highest exposures (or doses) by EVR range
Each of these tables lists estimates arranged by ozone concentration and
EVR range. Each entry indicates the number of outdoor children who experienced
their maximum ozone exposure under conditions in which the ozone concentration
was at or above the level indicated by the row label and the average EVR was
within the range indicated by the column heading. There are separate tables for
one-hour daily maximum exposures (Table 3 in Appendix D) and eight-hour daily
maximum exposures (Table 6).
Number of people -- cumulative daily maximum doses by number of days
These tables provide estimates arranged by ozone concentration and number
of days per year. Each entry lists the number of outdoor children who experienced
a daily maximum dose at or above the indicated ozone concentration for the
specified number of days. Separate tables are provided for daily maximum one-
hour doses (Table 9 in Appendix D), daily maximum eight-hour doses (Table 10),
daily maximum one-hour doses with EVR of 30 liters x min'1 x m"2 or greater (Table
11), and daily maximum eight-hour doses with EVR ranging from 13 liters x min"1 x
m'2 to 27 liters x min-1 x m'2 (Table 12).
Regardless of format, each table in Appendix D provides footnotes identifying
the study area and regulatory scenario. The footnotes also indicate the number of
exposure districts in the study area, the first and last days of the ozone season, and
the number of days in the ozone season.
127
-------
7.3 Results of Analyses
The pNEM/03 model incorporates a number of stochastic (random) elements
which directly affect the exposure estimates produced by the model. Consequently,
exposure estimates are likely to vary from run to run. To better characterize this
variability, ITAQS ran the model 10 times for each combination of study area and
regulatory scenario. Tables 50 through 53 provide means and ranges for selected
exposure indicators based on these runs.
Table 50 illustrates the general format used in Tables 50 through 53. This
table presents estimates for the number and percentage of outdoor children
experiencing one or more one-hour daily maximum ozone exposures above 120 ppb
at any ventilation rate. The first row in the table lists results for the Chicago study
area under the baseline scenario. Of the estimated 472,710 outdoor children in the
Chicago study area, 252,914 (53.50 percent) are estimated to have experienced the
specified exposure conditions based on the mean of the 10 runs. The estimates
associated with individual runs range from 233,862 (49.47 percent) to 288,683
(61.07 percent). Tables 51, 52, and 53 employ the same format to present
estimates for the number and percentage of outdoor children who experience one or
more eight-hour daily maximum ozone exposures above 60 ppb, 80 ppb, and 100
ppb, respectively, at any ventilation rate.
A review of the estimates in Tables 50 through 53 indicates that exposures
are generally higher under baseline conditions than under any one of the standards.
Denver and Miami show some exceptions to this generalization; exposures under
the current NAAQS, the 8H1EX-100 and the 8H1EX-90 scenarios are higher than
exposures under baseline conditions. St. Louis also displays this reversal under the
current NAAQS and 8H1EX-100 scenarios for outdoor children experiencing one or
more eight-hour daily maximum ozone exposures above 60 ppb at any ventilation
rate. In each of these cases, the ambient ozone levels permitted by the regulatory
scenario are higher than the ambient levels which occur under baseline conditions.
Consequently, the adjustment of baseline data to exactly meet the current NAAQS,
for example, produces an increase in ozone exposure.
128
-------
TABLE 50. NUMBER AND PERCENT OF OUTDOOR CHILDREN EXPERIENCING ONE OR MORE
ONE-HOUR DAILY MAXIMUM OZONE EXPOSURES ABOVE 120 PPB AT ANY VENTILATION RATE
i" 1
Study Area
Chicago
Denver
l
Number of
Persons at Risk
472,710
106,665
Regulatory
Scenario
Baseline
Current NAAQS
111 1 EX- 100
81 HEX- 100
8HIEX-90
8HIEX-80
811IEX-70
8II5EX-90
8II5EX-80
Baseline
Current NAAQS
1H1EX-100
81 11 EX- 100
8IHEX-90
8I1IEX-80
8IIIEX-70
81I5EX-90
8H5EX-80
Mean
Number of
Persons Exposed
252,914
86,918
0
216,080
47,557
168
0
236,130
64,579
21,438
33,358
0
7 1 ,923
4 1 ,7 1 0
9,907
0
45,140
11,370
Percent of
Total
53.50
18.39
0.00
45.71
10.06
0.04
0.00
49.95
13.66
20.10
31.27
0.00
67.43
39.10
9.29
0.00
42,32
10.66
Ranj
Number of Persons
Exposed
233,862 - 288,683
56,585 - 120,993
0 - 0
185,954 - 246,754
18,498 - 90,111
0- 1,179
0- 0
212,570 - 256711
45,242 - 85,735
14,167 - 28,750
18,235 -42,539
0- 0
64,180 - 77,056
37,443 - 45,247
6,139 - 13,347
0- 0
34,287 - 54,906
3,557- 16,282
*^si"* i \r\ i i_
'°
Percent
of Total 1
49.47 - 61.07
11.97 - 25.60
0.00 - 0.00
39.34 - 52.20
3.91 - 19.06
0.00 - 0 25
0.00 - 0.00
44.97 - 54.31
9.57- 18.14
13.28 - 26.95
17.10 - 39.88
0.00 - 0.00
60.17 - 72.24
35.10 - 42.42
5.76 - 12.51
0.00 - 0.00
32.14 - 51.48
3.33 - 15.26
CD
(continued)
-------
TABLE 50 (Continued)
Study Area
Houston
Los
Angeles
Number of
Persons at Risk
200,795
798,290
Regulatory
Scenario
Baseline
Current NAAQS
1 HI EX- 100
8H1EX-100
8H1EX-90
81UEX-80
8H1EX-70
8H5EX-90
8H5EX-80
Baseline
Current NAAQS
IIIIEX-IOO
81 11 EX- 100
8H1EX-90
8H1EX-80
81I1EX-70
8H5EX-90
8H5EX-80
Mean
Number of
Persons Exposed
200,425
35,892
29
112,215
35,416
5,875
0
143,166
64,555
713,214
16,198
57
162,639
62,926
14,179
109
90,405
21,448
Percent of
Total
99.82
17.87
0.01
55.89
17.64
2.93
0.00
71.30
32.15
89.34
2.03
0.01
20.37
7.88
1.78
0.01
11.32
2.69
Range
Number of Persons
Exposed
199,136 - 200,795
20,888 - 48,332
0- 293
96,504 - 116,943
27,644 - 44,145
0-8,510
0- 0
132,566 - 153,049
52,319 - 75,290
695,388 - 734,039
12,235 - 20,532
0- 572
147,820 - 172,915
49,301 - 71,960
8,794 - 18,974
0- 1,088
80,115 - 106,528
18,666 - 24,647
Percent
of Total
99.17- 100.00
10.40 - 24.07
0.00 - 0.15
48.06 - 58.24
13.77 -21.99
0 - 4.24
0.00 - 0.00
66.02 - 76.22
26.06 - 37.50
87.11 -91.95
1.53 -2.57
0.00 - 0.07
18.52 - 21.66
6.18 - 9.01
1.10 - 2.38
0.00 -0.14
10.04 - 13.34
2.34 - 3.09
CO
o
-------
TABLE 50 (Continued)
Sliuiy Area
Miami
New
York
Number of
Persons at Risk
133,610
782,600
Regulatory
Scenario
Baseline
Current NAAQS
1H1EX-IOO
81 11 EX- 100
8H1EX-90
81 1 1 EX-80
8H1EX-70
8H5EX-90
8H5 EX-80
Baseline
Current NAAQS
IIIIEX-IOO
8! 11 EX- 100
8H1EX-90
8111 EX-80
8111EX-70
8H5EX-90
8M5EX-80
Mean
Number of
Persons Exposed
4,374
20,364
3,141
83,937
29,808
6,884
927
107,339
37,518
541,114
34,132
76
93,837
19,208
1,413
0
89,581
10,561
Percent of
Total
3.27
15.24
2.35
62.82
22.31
5.15
0.69
80.34
28.08
69.14
4.36
0.01
1 1 .99
2.45
0.18
0.00
11.45
1.35
Range
Number of Persons
Exposed
2,554 - 5,754
13,756 - 23,459
24 - 5,778
74,556 - 101,859
23,390 - 36,802
3,248 - 9,979
0-2,318
99,591 - 111,275
24,651 - 50,133
500,315 - 567,283
26,297 - 44,525
0-756
83,238 - 102,323
7,548 - 31,145
0 - 8,246
0 - 0
81,382 - 97,255
5,028 - 17,657
Percent
of Total
1.91 -4.31
10.30- 17.56
0.02 - 4.32
55.80 - 76.24
17.51 - 27.54
2.43 - 7.47
0.00- 1.73
74.54 - 83.28
18.45 - 37.52
63.93 - 72.49
3.36 - 5.69
0.00 - 0.10
10.64 - 13.07
0.96 - 3.98
0- 1.05
0.00 - 0.00
10.40 - 12.43
0.64 - 2.26
-------
TABLE 50 (Continued)
Study Area
Philadelphia
St. Louis
Number of
Persons at Risk
275,320
128,250
Regulatory
Scenario
Baseline
Current NAAQS
1HIEX-100
8H1EX-100
8H1EX-90
8H1EX-80
81 1 \ EX-70
8I15EX-90
8H5EX-80
Baseline
Current NAAQS
1H1EX-100
8M1EX-100
8H1EX-90
8IIIEX-80
81 11 EX-70
8H5EX-90
8H5EX-80
Mean
Number of
Persons Exposed
269,385
12,933
0
19,781
112
0
0
15,892
0
45,807
15,609
322
10,315
3,000
0
0
32,638
3,686
Percent of
Total
97.84
4.70
0.00
7.18
0.04
0.00
0,00
5.77
0.00
35.72
12.17
0.25
8.04
2.34
0.00
0.00
25.45
2.87
Range
Number of Persons
Exposed
265,362 -271,485
6,943 - 18,949
0- 0
13,831 - 29,354
0- 573
0- 0
0- 0
4,281 -29,174
0-0
42,107 - 51,554
12,517 - 19,294
0- 1,451
8,535 - 14,573
994 . 4,496
0- 0
0-0
26,123 - 39,788
951 -7,302
Percent
of Total
96.38-98.61
2.52 - 6.88
0.00 - 0.00
5.02 - 10.66
0.00 - 0.21
0.00 - 0.00
0.00 - 0.00
1.55 - 10.60
0.00 - 0.00
32.83 - 40.20
9.76 - 15.04
0.00 - 1.13
6.65 - 11.36
0.78 - 3.51
0.00 - 0.00
0.00 - 0.00
20.37-31.02
0.74 - 5.69
CO
ro
-------
TABLE 50 (Continued)
Study Area
Washington
D.C.
======:
Number of
Persons at Risk
198,860
Regulatory
Scenario
Baseline
Current NAAQS
IH1EX-100
8H1EX-100
8H1EX-90
8H1EX-80
8H1EX-70
8U5EX-90
8H5EX-80
^
Mean
Number of
Persons Exposed
190,259
14,796
38
16,268
4,915
0
0
43,941
2,657
T=-^ .
Percent of
Total
95.67
7,44
0.02
8.18
2.47
0.00
0.00
22.10
1.34
Range
Number of Persons
Exposed
183,960- 192,795
10,855 - 18,513
0-381
14,184 - 18,189
901 - 10,267
0-0
0-0
40,217 - 46,946
706 - 5,678
Percent
of Total
92.51 -96.95
5.46-9.31
0.00 - 0.19
7.13 -9.15
0.45 - 5.16
0.00 - 0.00
0.00 - 0.00
20.22 - 23.61
0.36 - 2.86
CO
-------
TABLE 51. NUMBER AND PERCENT OF OUTDOOR CHILDREN EXPERIENCING ONE OR MORE
EIGHT-HOUR DAILY MAXIMUM OZONE EXPOSURES ABOVE 60 PPB AT ANY VENTILATION RATE
Study Area
Chicago
Denver
Number of
Persons at Risk
472,710
106,665
Regulatory
Scenario
Baseline
Current NAAQS
1H1KX-100
81 11 EX- 100
8I11HX-90
8HIEX-80
8JIIHX-70
8IISEX-90
8IISEX-80
Baseline
Current NAAQS
1II1EX-IOO
81 11 EX- 100
8M1EX-90
8M1EX-80
8M1EX-70
81I5KX-90
8I15EX-80
Mean
Number of
Persons Exposed
472,621
464,191
320,239
472,710
462,228
362,463
152,443
472,492
466,817
99,449
93,808
68,166
106,206
101,820
88,046
49,132
104,362
91,092
Percent
of Total
99.98
98.20
67.75
100.00
97.78
76.68
32.25
99.95
98.75
93.23
87.95
63.91
99.57
95.46
82.54
46.06
97.84
85.40
Range
Number of Persons
Exposed
471,820 -472,710
458,510 - 469,167
300,967 - 341,007
472,710 - 472,710
458,566 - 464,920
340,715 - 380,222
129,118 - 186,260
471,354 -472,710
462,632 -471,279
97,015 - 102,785
89,684 - 97,462
60,492 - 77,765
104,657 - 106,665
99,305 - 104,465
85,311 - 89,437
41,986 - 52,135
103,247 - 105,326
85,715 - 94,992
Percent
of Total
99.81 - 100.00
97.00 - 99.25
63.67 - 72.14
100.00 - 100.00
97.01 - 98.35
72.08 - 80.43
27.31 - 39.40
99.71 - 100.00
97.87 - 99.70
90.95 - 96.36
84.08 - 91.37
56.71 - 72.91
98.12 - 100.00
93.10 - 97.94
79.98 - 83.85
39.36 - 48.88
96.80 - 98.74
80.36 - 89.06
(continued)
-------
TAULIi 5J (Continued)
Study Area
Houston
Los
Angeles
Number of
Persons at Risk
200,795
798,290
Regulatory
Scenario
Baseline
Current NAAQS
11 11 EX- 100
SHI EX- 100
8HIEX-90
8H1EX-80
8H1EX-70
8USEX-90
8H5EX-80
Baseline
Current NAAQS
IIIIEX-IOO
8H1EX-100
8H1EX-90
8HIEX-80
8H1EX-70
8IISEX-90
8I15EX-80
Mean
Number of
Persons Exposed
200,795
189,773
122,526
195,897
180,552
132,992
58,555
196,664
181,226
789,497
248,727
156,847
341,341
284,248
227,175
115,220
270,811
206,669
Percent
of Total
100.00
94.51
61.02
97.56
89.92
66.23
29.16
97.94
90.25
98.90
31.16
19.65
42.76
35.61
28.46
14.43
33.92
25.89
Range
Number of Persons
Exposed
200,795 - 200,795
182,702 - 195,230
114,768 - 130,997
190,738 - 199,486
174,556 - 184,568
108,727 - 143,553
40,618 - 70,084
194,422 - 200,093
173,936 - 185,528
782,143 - 794,073
239,525 - 264,995
149,423 - 166,488
329,109-359,623
277,015 -296,505
219,415 -239,119
100,530 - 122,490
251,328 -285,161
198,978 - 215,761
Percent
of Total
100.00 - 100.00
90.99 - 97.23
57.16 - 65.24
94.99 - 99.35
86.93 - 91.92
54.15 - 71.49
20.23 - 34.90
96.83 - 99.65
86.62 - 92.40
97.98 - 99.47
30.00 - 33.20
18.72 - 20.86
41.23 - 45.05
34.70 - 37.14
27.49 - 29.95
12.59 - 15.34
31.48 - 35.72
24.93 - 27.03
Ol
(continued)
-------
TABLE 51 (Continued)
Study Area
Miami
New
York
Number of
Persons at Risk
133,610
782,600
Regulatory
Scenario
Baseline
Current NAAQS
1H1EX-100
8H1EX-100
8H1EX-90
8H1EX-80
8H1EX-70
8H5EX-90
8H5EX-80
Baseline
Current NAAQS
1111 EX- 100
8H1EX-100
811IEX-90
8H1EX-80
8HIEX-70
8H5EX-90
8H5EX-80
Mean
Number of
Persons Exposed
73,725
117,572
57,107
131,107
119,964
91,237
37,762
133,582
128,818
741,850
525,369
316,317
609,108
572,823
364,261
170,627
596,391
437,680
Percent
of Total
55.18
88.00
42.74
98.13
89.79
68.29
28.26
99.98
96.41
94.79
67.13
40.42
77.83
73.19
46.54
21.80
76.21
55.93
Range
Number of Persons
Exposed
59,528 - 81,301
110,490 - 124,216
45,695 - 69,183
126,359 - 133,610
115,156 - 127,540
79,426 - 101,375
33,108 -42,797
133,327 - 133,610
123,328 - 131,309
721,627-762,270
486,814 - 550,873
292,975 - 346,748
598,403 - 610,915
554,906 - 594,127
350,011 - 383,466
164,012 - 178,692
585,302 - 603,551
386,569 - 465,269
Percent
of Total
44.55 - 60.85
82.70 - 92.97
34.20 - 51.78
94.57 - 100.00
86.19 - 95.46
59.45 - 75.87
24.78 - 32.03
99.79 - 100.00
92.30 - 98.28
92.21 - 97.40
62.20 - 70.39
37.44 - 44.31
76.46 - 78.06
70.91 - 75.92
44.72 - 49.00
20.96 - 22.83
74.79 - 77.12
49.40 - 59.45
CO
CD
(coniiiiued)
-------
TABLli 51 (Continued)
Study Area
Philadelphia
St. Louis
Number of
Persons at Risk
275,320
128,250
Regulatory
Scenario
Baseline
Current NAAQS
luiEx-ioo
8111 EX- 1 00
8H1EX-90
81 -HEX -80
8H1EX-70
8H5EX-90
8II5EX-80
Baseline
Current NAAQS
UI1 EX- 100
8H1EX-100
8HIEX-90
8II1EX-80
8H1EX-70
8II5EX-90
8II5EX-80
Mean
Number of
Persons Exposed
275,320
275,320
270,747
275,320
274,390
252,092
102,407
274,718
263,383
112,768
121,279
96,669
116,855
103,510
75,937
25,087
122,468
105,291
Percent
of Total
100.00
100.00
98.34
100.00
99.66
91.56
37.20
99.78
95.66
87.93
94.56
75.38
91.12
80.71
59.21
19.56
95.49
82.10
Range
Number of Persons
Exposed
275,320 - 275,320
275,320 - 275,320
265,871 - 272,985
275,320 - 275,320
272,309 - 275,320
246,303 - 260,476
92,371 - 110,040
272,927 - 275,320
256,376 - 269,006
110,241 - 117,523
120,699 - 121,710
92,319 - 101,910
115,031 - 118,142
100,256 - 105,577
69,941 - 78,998
21,772 - 28,812
120,991 - 123,809
100,020 - 109,146
Percent
of Total
100.00 - 100.00
100.00 - 100.00
96.57 - 99.15
100.00 - 100.00
98.91 - 100.00
89.46 - 94.61
33.55 - 39.97
99.13 - 100.00
93.12 - 97.71
85.96-91.64
94.11 - 94.90
71.98 - 79.46
89.69 - 92.12
78.17 - 82.32
54.53 - 61.60
16.98 - 22.47
94.34 - 96.54
77.99 - 85.10
CO
-vl
(continued)
-------
TABLE 51 (Continued)
Study Area
Washington
D.C.
Number of
Persons at Risk
198,860
Regulatory
Scenario
Baseline
Current NAAQS
1HIEX-100
8H1EX-100
8H1EX-90
8H1EX-80
8H1EX-70
8H5EX-90
8H5EX-80
Mean
Number of
Persons Exposed
198,860
198,714
181,013
198,701
196,054
148,596
41,670
198,730
191,006
Percent
of Total
100.00
99.93
91.03
99.92
98.59
74.72
20.95
99.93
96.05
Range
Number of Persons
Exposed
198,860 - 198,860
198,237 - 198,860
171,334 - 188,581
197,272 - 198,860
191,130 - 198,079
133,641 - 161,442
38,770 - 44,329
198,223 - 198,860
187,596- 194,134
Percent
of Total
100.00 - 100.00
99.69 - 100.00
86.16 - 94.83
99.20 - 100.00
96.11 -99.61
67.20 - 81.18
19.50 - 22.29
99.68 - 100.00
94.34 - 97.62
oo
00
-------
TABLE 52. NUMBER AND PERCENT OF OUTDOOR CHILDREN EXPERIENCING ONE OR MORE
EIGHT-HOUR DAILY MAXIMUM OZONE EXPOSURES ABOVE 80 PPB AT ANY VENTILATION RATE
Study Area
Chicago
Denver
Number of
Persons at Risk
'172,710
106,665
Regulatory
Scenario
Baseline
Current NAAQS
1HIGX-100
811 1 EX- 100
8II1EX-90
81 1 1 EX-80
8H1EX-70
8II5EX-90
8H5EX-80
Baseline
Current NAAQS
IIIIEX-IOO
81 11 EX- 100
81HEX-90
8H1 EX-80
81IIEX-70
8M5EX-90
8115EX-80
Mean
Number of
Persons Exposed
339,451
147,277
3,662
269,575
116,934
6,549
0
313,605
118,124
20,046
32,176
712
68,815
39,927
5,669
0
33,438
8,745
Percent
of Total
71.81
31.16
0.77
57.03
24.74
1.39
0.00
66.34
24.99
18.79
30.17
0.67
64.52
37.43
5.31
0.00
31.35
8.20
Range
Number of Persons
Exposed
316,734 - 357,026
122,337 - 171,046
0 - 5,035
254,658 - 302,935
103,478 - 146,028
3,050 - 10,420
0- 0
287,400 - 333,762
88,860 - 150,054
15,972 - 25,258
28,246 - 36,327
0 - 2,090
64,388 - 72,155
34,455 - 46,335
3,114 - 8,651
0- 0
25,110 - 39,257
4,118 - 12,141
Percent
of Total
67.00 - 75.53
25.88 - 36.18
0.00 - 1.07
53.87 - 64.08
21.89 - 30.89
0.65 - 2.20
0.00 - 0.00
60.80 - 70.61
18.80 - 31.74
14.97 - 23.68
26.48 - 34.06
0.00 - 1 .96
60.36 - 67.65
32.30 - 43.44
2.92 - 8.11
0.00 - 0.00
23.54 - 36.80
3.86 - 11.38
00
CD
(continued)
-------
TABLE 52 (Continued)
Study Area
Houston
Los
Angeles
Number of
Persons at Risk
200,795
798,290
Regulatory
Scenario
Baseline
Current NAAQS
1H1EX-100
8111 EX- 100
8H 1 EX-90
8H1EX-80
811 1 EX-70
81 15 EX-90
8I15EX-80
Baseline
Current NAAQS
1H1EX-100
8H1EX-IOO
81! 1 EX-90
8H1EX-80
81 11 EX-70
81 15 EX-90
8IT5EX-80
Mean
Number of
Persons Exposed
198,249
41,968
2,248
98,802
39,607
8,125
536
116,698
45,770
672,461
23,164
0
164,153
93,508
13,222
0
105,039
35,601 '
Percent
of Total
98.73
20.90
1.12
49.21
19.73
4.05
0.27
58.12
22.79
84.24
2.90
0.00
20.56
11.71
1.66
0.00
13.16
4.46
Range
Number of Persons
Exposed
196,879 - 199,745
23,580 - 63,386
36 - 5,200
84,776 - 108,945
29,391 - 50,040
4,177- 17,118
0- 1,280
103,422 - 133,271
33,559 - 62,760
634,085 - 690,933
16,961 - 29,087
0- 0
154,585 - 178,854
84,751 - 101,011
8,270 - 19,039
0-0
96,547 - 117,422
29,745 - 47,489
Percent
of Tolal
98.05 - 99.48
11.74 - 31.57
0.02 - 2.59
42.22 - 54.26
14.64 - 24.92
2.08 - 8.53
0.00 - 0.64
51.51 - 66.37
16.71 - 31.26
79.43 - 86.55
2.12 --3.64
0.00 - 0.00
19.36 - 22.40
10.62 - 12.65
1.04 -2.38
0.00 - 0.00
12.09 - 14.71
3.73 - 5.95
(conlA nuecY)
-------
TABLli 52 (Continued)
Study Area
Miami
New
York
Number of
Persons at Risk
133,610
782,600
Regulatory
Scenario
Baseline
Current NAAQS
1II1KX-100
811 1 EX- 100
8H1EX-90
8H1EX-80
8H1CX-70
8H5EX-90
8H5EX-80
Baseline
Current NAAQS
1H1EX-100
8H1EX-100
81 1 1 EX-90
8II1EX-80
8FIIEX-70
81I5EX-90
8H5EX-80
Mean
Number of
Persons Exposed
702
17,344
634
65,918
27,354
3,014
0
85,434
42,953
578,118
130,169
8,490
244,838
104,028
13,339
0
174,570
61,157
Percent
of Total
0.53
12.98
0.47
49.34
20.47
2.26
0.00
63.94
32.15
73.87
16.63
1.08
31.29
13.29
1.70
0.00
22.31
7.81
Range
Number of Persons
Exposed
0 - 3,533
11,189 - 26,615
0 - 5,592
55,549 - 73,943
17,549 - 37,699
265 - 7,313
0 - 0
77,113 - 94,261
36,515 -48,555
560,849 - 595,247
110,739 - 146,820
6,135 - 12,438
227,164 -262,501
86,852 - 122,881
7,926 - 1 8,090
0 - 0
161,793 - 184,845
48,906 - 69,286
Percent
of Total
0.00 - 2.64
8.37 - 19.92
0.00 - 4.19
41.58 - 55.34
13.13 - 28.22
0.20 - 5.47
0.00 - 0.00
57.71 - 70.55
27.33 - 36.34
71.66 - 76.06
14.15 - 18.76
0.78 - 1.59
29.03 - 33.54
11.10 - 15.70
1.01 - 2.31
0.00 - 0.00
20.67 - 23.62
6.25 - 8.85
(continued)
-------
TABLE 52 (Continued)
Study Area
Philadelphia
St. Louis
Number of
Persons at Risk
275,320
128,250
Regulatory
Scenario
Baseline
Current NAAQS
IHIEX-IOO
8111 EX- 100
8H1EX-90
8H1EX-80
8H1EX-70
8H5EX-90
8H5EX-80
Baseline
Current NAAQS
1HIEX-100
81 11 EX- 100
811 1 EX-90
8H1EX-80
8H1EX-70
8H5EX-90
8MSEX-80
Mean
Number of
Persons Exposed
273,481
164,718
28,861
174,478
63,313
8,464
0
137,731
33,892
57,054
56,463
4,889
49,925
13,045
1,030
0
62,735
20,986
Percent
of Total
99.33
59.83
10.48
63.37
23.00
3.07
0.00
50.03
12.31
44.49
44.03
3.81
38.93
10.17
0.80
0.00
48.92
16.36
Range
Number of Persons
Exposed
271,175 - 275,320
146,501 - 180,412
20,147 - 35,055
154,686- 189,712
46,707 -76,541
4,641 - 14,705
0- 0
126,576 - 154,987
22,344 - 39,860
51,550 - 64,321
51,280 - 62,789
1,501 - 7,414
45,837 - 63,350
8,384 - 19,013
0 - 2,563
0- 0
55,410 - 68,971
16,192 - 25,192
Percent
of Total
98.49 - 100.00
53.21 - 65.53
7.32 - 12.73
56.18 - 68.91
16.96 - 27.80
1.69 - 5.34
0.00 - 0.00
45.97 - 56.29
8.12 - 14.48
40.19- 50.15
39.98 - 48.96
1.17 - 5.78
35.74 - 49.40
6.54 - 14.82
0.00 - 2.00
0.00 - 0.00
43.20 - 53.78
12.63 - 19.64
ro
-------
J ABLli 52 (Continued)
Study Area
Washington
D.C.
Number of
Persons at Risk
198,860
Regulatory
Scenario
Baseline
Current NAAQS
IHIEX-IOO
8H1EX-100
8H1EX-90
8H1EX-80
8H1EX-70
8H5EX-90
8H5EX-80
Mean
Number of
Persons Exposed
192,494
76,159
11,975
86,540
31,298
7,428
36
120,335
34,928
Percent
of Total
96.80
38.30
6.02
43.52
15.74
3.74
0.02
60.51
17.56
Range
Number of Persons
Exposed
185,648 - 196,208
62,438 - 83,945
6,928 - 15,803
79,074 - 97,301
26,450 -35,911
4,875 - 12,364
0- 357
112,787 - 132,359
28,095 - 39,613
Percent
of Total
93.36 - 98.67
31.40 - 42.21
3.48 - 7.95
39.76 - 48.93
13.30- 18.06
2.45 - 6.22
0.00 - 0.18
56.72 - 66.56
14.13 - 19.92
CJ
-------
TABLE 53. NUMBER AND PERCENT OF OUTDOOR CHILDREN EXI'I
EIGHT-HOUR DAILY MAXIMUM OZONE EXPOSURES ABOVE 100 PPB
KIENCING ONE OR MORE
Study Area
Chicago
Denver
Number of
Persons at Risk
472,710
106,665
Regulatory
Scenario
Baseline
Current NAAQS
11IIEX-IOO
81 HEX- 100
8IIIHX-90
8IIIEX-80
8IIIEX-70
8II5EX-90
8IISEX-80
Baseline
Current NAAQS
1HIEX-100
8H1EX-IOO
8H1EX-90
811 1 EX-80
8HIEX-70
8M5HX-90
8H5liX-80
Mean
Number of
Persons Exposed
20,005
107
0
7,653
0
0
0
17,772
29
0
111
0
6,760
0
0
0
632
0
Percent
of Total
4.23
0.02
0.00
1.62
0.00
0.00
0.00
3.76
0.01
0.00
0.10
0.00
6.34
0.00
0.00
0.00
0.59
0.00
==r
Range
Number of Persons
Exposed
16,490 - 25,496
0 - 528
0- 0
4,041 - 12,019
0-0
0- 0
0 - 0
12,768 - 27,884
0- 294
0- 0
0- 789
0-0
3,325 - 10,870
0-0
0- 0
0- 0
54 - 1,765
0-0
Percent
of Total
3.49 - 5.39
0.00 - 0.11
0.00 - 0.00
0.85 - 2.54
0.00 - 0.00
0.00 - 0.00
0.00 - 0.00
2.70 - 5.90
0.00 - 0.06
0.00 - 0.00
0.00 - 0.74
0.00 - 0.00
3.12 - 10.19
0.00 - 0.00
0.00 - 0.00
0.00 - 0.00
0.05 - 1.65
0.00 - 0.00
.
-------
1 AiiLli 53 (Continued)
Study Area
Houston
Los
Angeles
Number of
Persons at Risk
200,795
798,290
Regulatory
Scenario
Baseline
Current NAAQS
IIUEX-IOO
8111 EX- 100
8H1EX-90
8H1EX-80
8H1EX-70
8H5EX-90
8HSEX-80
Baseline
Current NAAQS
1111 EX- 100
8H1EX-100
81 1 1 EX-90
8H1EX-80
8HIEX-70
8H5EX-90
8M5EX-80
Mean
Number of
Persons Exposed
165,332
481
0
13,408
1,740
74
0
25,945
2,748
457,507
0
0
9,642
0
0
0
2,285
0
Percent
of Total
82.34
0.24
0.00
6.68
0.87
0.04
0.00
12.92
1.37
57.31
0.00
0.00
1.21
0.00
0.00
0.00
0.29
0.00
Range
Number of Persons
Exposed
157,637 - 173,080
0 - 3,207
0 - 0
5,852 - 24,394
0 - 4,801
0- 737
0- 0
15,942 -40,102
130 - 7,391
441,832 - 472,777
0-0
0-0
5,577 - 17,963
0-0
0- 0
0- 0
0 - 5,038
0- 0
Percent
of Total
78.51 - 86.20
0.00 - 1 .60
0.00 - 0.00
2.91 - 12.15
0.00 - 2.39
0 - 0.37
0.00 - 0.00
7.94 - 19.97
0.06 - 3.68
55.35 - 59.22
0.00 - 0.00
0.00 - 0.00
0.70 - 2.25
0.00 - 0.00
0.00 - 0.00
0.00 - 0.00
0.00 - 0.63
0.00 - 0.00
Ol
-------
TABLE 53 (Continued)
Study Area
Miami
New
York
Number of
Persons at Risk
133,610
782,600
Regulatory
Scenario
Baseline
Current NAAQS
IH1EX-100
8H1EX-100
8M1EX-90
8H1EX-80
8H1EX-70
8H5EX-90
8H5EX-80
Baseline
Current NAAQS
1HIEX-IOO
8111 EX- 100
8I11EX-90
8H1EX-80
8H1EX-70
8H5EX-90
8H5EX-80
Mean
Number of
Persons Exposed
0
0
0
3,715
0
0
0
11,250
1,157
284,741
905
0
11,074
0
0
0
13,010
0
Percent
of Total
0.00
0.00
0.00
2.78
0.00
0.00
0.00
8.42
0.87
36.38
0.12
0.00
1.42
0.00
0.00
0.00
1.66
0.00
Range
Number of Persons
Exposed
0-0
0 - 0
0 - 0
0 - 8,585
0- 0
0- 0
0- 0
5,625 - 22,866
0-4,108
262,297 - 315,800
0 - 2,030
0- 0
2,353 - 17,134
0- 0
0-0
0- 0
4,273 - 18,656
0- 0
Percent
of Total
0.00 - 0.00
0.00 - 0.00
0.00 - 0.00
0.00 - 6.43
0.00 - 0.00
0.00 - 0.00
0.00 - 0.00
4.21 - 17.11
0.00 - 3.07
33.52 - 40.35
0.00 - 0.26
0.00 - 0.00
0.30 - 2.19
0.00 - 0.00
0.00 - 0.00
0.00 - 0.00
0.55 -2.38
0.00 - 0.00
CD
-------
TAULE 53 (Continued)
Study Area
Philadelphia
St. Louis
Number of
Persons at Risk
275,320
128,250
Regulatory
Scenario
Baseline
Current NAAQS
11I1EX-100
811 1 EX- 100
8H1EX-90
8H1EX-80
811 1 EX-70
8H5EX-90
8II5EX-80
Baseline
Current NAAQS
1H1EX-100
8H1EX-100
8II1EX-90
8H1EX-80
81-11 EX-70
8H5EX-90
8H5EX-80
Mean
Number of
Persons Exposed
193,608
3,654
0
6,763
42
0
0
2,842
0
3,932
1,196
0
741
68
0
0
3,383
0
Percent
of Total
70.32
1.33
0.00
2.46
0.02
0.00
0.00
1.03
0.00
3.07
0.93
0.00
0.58
0.05
0.00
0.00
2.64
0.00
Range
Number of Persons
Exposed
181,224 -205,873
1,271 - 9,757
0 - 0
4,255 - 10,039
0- 423
0- 0
0- 0
152 - 6,260
0- 0
2,268 - 6,338
0 - 3,000
0- 0
186- 1,815
0- 547
0- 0
0-0
1,193 - 6,252
0- 0
Percent
of Total
65.82 - 74.78
0.46 - 3.54
0.00 - 0.00
1.55 - 3.65
0.00 - 0.15
0.00 - 0.00
0.00 - 0.00
0.06 - 2.27
0.00 - 0.00
1.77 -4.94
0.00 - 2.34
0.00 - 0.00
0.15 - 1.42
0.00 - 0.43
0.00 - 0.00
0.00 - 0.00
0.93 - 4.87
0.00 - 0.00
(continued)
-------
TABLE 53 (Continued)
Study Area
Washington
D.C.
Number of
Persons at Risk
198,860
Regulatory
Scenario
Baseline
Current NAAQS
1H1EX-100
8H1EX-100
8H1EX-90
8H1EX-80
8HIEX-70
8H5EX-90
8H5EX-80
Mean
Number of
Persons Exposed
98,432
5,294
0
6,437
658
0
0
9,788
75
Percent
of Total
49.50
2.66
0.00
3.24
0.33
0.00
0.00
4.92
0.04
Range
Number of Persons
Exposed
88,321 - 113,524
657- 10,154
0-0
2,026- 11,735
0 - 4,335
0- 0
0-0
7,053 - 13,301
0- 355
Percent
of Total
44.41 - 57.09
0.33 - 5.11
0.00 - 0.00
1.02 - 5.90
0.00- 2.18
0.00 - 0.00
0.00 - 0.00
3.55 - 6.69
0.00 - 0.18
00
-------
7.4 Estimates of Maximum Dose Exposures
Each ozone exposure estimated by pNEM/03 includes a value for ozone
concentration and a value for EVR. The product of ozone concentration and EVR
provides an indication of ozone dose. The "daily maximum dose" is assumed to
occur each day during the period when this product is highest. Consistent with this
concept, pNEM/03 provides dose estimates for two averaging times: the one-hour
maximum daily dose and the eight-hour daily maximum dose. Analysts selected two
specific exposure indicators from these model outputs for further evaluation:
The number of outdoor children who experienced one or more one-
hour maximum daily dosage exposures during which the ozone
concentration exceeded 0.12 ppm (120 ppb) and the EVR equaled or
exceeded 30 liters min"1- m"2.
The number of outdoor children who experienced one or more eight-
hour maximum daily dosage exposures during which the ozone
concentration exceeded 0.08 ppm (80 ppb) and the EVR ranged from
13 liters min'1 m"2 to 27 liters min"1 m"2.
Tables 54a through 71 b present a summary of the exposure estimates based on
these two indicators. The tables are grouped in pairs by study area; for example,
Tables 54a,b and 55a,b present the one-hour and eight-hour dose estimates,
respectively, for Chicago. Note that the values listed in each table consist of mean
values and ranges based on 10 runs of pNEM/O3. Each table provides a separate
set of estimates for each of the nine air quality scenarios discussed previously.
Tables 58a and 58b illustrate the general format used in all of the one-hour
tables. The statistics in the first row are the 10-run mean estimates (by scenario)
for the number of outdoor children in Houston who experienced one or more one-
hour maximum daily dosage exposures during which the ozone concentration
exceeded 0.12 ppm and the EVR equaled or exceeded 30 liters min'1- m'2. Under
baseline conditions, 18,374 outdoor children are estimated to have experienced the
specified exposure. According to the value listed in the second row, 18,374 children
149
-------
TABLE 54a. ESTIMATES OF ONE-HOUR MAXIMUM DOSAGE EXPOSURES EXPERIENCED
BY OUTDOOR CHILDREN IN CHICAGO DURING WHICH OZONE CONCENTRATION
EXCEEDED 0.12 ppm AND EVRa EQUALED OR EXCEEDED 30 LITERS- MIN'1- M2
Statistic"
Mean Estimate of the Number of Outdoor Children
Percent of Total Outdoor Children Population
Range in this percentage for 10 runs
Mean Estimate of Person-Occurrences
Percent of Total Person-Occurrences
Range in this percentage for 10 runs
Mean Estimate of Occurrences/Person Exposed
Percentage exposed for indicated number of days
1 Day
2 Days
>2 Days
Regulatory scenario
Baseline
3,651
0.77
0.00-2.76
3.651
d
0.00-0.01
1.00
100.00
0.00
0.00
-'" * ' i I,.
1H1EX-120C
390
0.08
0.00-0.58
390
d
e
1.00
100.00
0.00
0.00
1H1EX-100
0
0.00
0
0.00
-
-
01
o
aEquivalent ventilation rate = (ventilation rate)/(body surface area).
bMean or range for 10 runs of pNEM/O3.
"Current NAAQS.
dLess than 0.01 percent.
eAII values less than 0.01 percent.
-------
TABLE 54b. ESTIMATES OF ONE-HOUR MAXIMUM DOSAGE EXPOSURES EXPERIENCED
BY OUTDOOR CHILDREN IN CHICAGO DURING WHICH OZONE CONCENTRATION
EXCEEDED 0.12 ppm AND EVRa EQUALED OR EXCEEDED 30 LITERS- MIN'1- M'2
Statistic"
Mean Estimate of the Number of Outdoor Children
Percent of Total Outdoor Children Population
Range in this percentage for 10 runs
Mean Estimate of Person-Occurrences
Percent of Total Person-Occurrences
Range in this percentage for 10 runs
Mean Estimate of Occurrences/Person Exposed
Percentage exposed for indicated number of days
1 Day
2 Days
>2 Days
Regulatory scenario
8H1EX-100
1,526
0.32
0.00-1.23
1,526
d
0.00-0.01
1.00
100.00
0.00
0.00
8H1EX-90
138
0.03
0.00-0.29
138
d
e
1.00
100.00
0.00
0.00
8H1EX-80
0
0.00
0
0.00
-
-
8H1EX-70
0
0.00
0
0.00
-
-.
8H5EX-90
1,268
0.27
0.00-1.48
1,268
d
0.00-0.01
1.00
100.00
0.00
0.00
8H5EX-80
456
0.10
0.00-0.58
456
d
e
1.00
100.00
0.00
0.00
en
aEquivalent ventilation rate = (ventilation rate)/(body surface area).
bMean or range for 10 runs of pNEM/O3.
dLess than 0.01 percent.
eAII values less than 0.01 percent.
-------
TABLE 55a. ESTIMATES OF EIGHT-HOUR MAXIMUM DOSAGE EXPOSURES EXPERIENCED
BY OUTDOOR CHILDREN IN CHICAGO DURING WHICH OZONE CONCENTRATION
EXCEEDED 0.08 ppm AND EVRa RANGED FROM 13 LITERS- MIN1- M'2 TO 27 LITERS- MIN'1- M'2
Statistic"
Mean Estimate of the Number of Outdoor Children
Percent of Total Outdoor Children Population
Range in this percentage for 10 runs
Mean Estimate of Person-Occurrences
Percent of Total Person-Occurrences
Range in this percentage for 10 runs
Mean Estimate of Occurrences/Person Exposed
Percentage exposed for indicated number of days
1 Day
2 Days
3 Days
>3 Days
Regulatory scenario
Baseline
128,451
27.17
18.08-34.59
157,505
0.16
0.10-0.20
1.23
81.26
15.65
2.71
0.38
1H1EX-1200
41.435
8.77
6.94-11.38
45,280
0.04
0.03-0.06
1.09
91.85
7.33
0.83
0.00
1H1EX-100
527
0.11
0.00-0.27
527
d
e
1.00
100.00
0.00
0.00
0.00
en
NJ
"Equivalent ventilation rate = (ventilation rate)/(body surface area).
"Mean or range for 10 runs of pNEM/O3.
°Current NAAQS.
dLess than 0.01 percent.
"All values less than 0.01 percent.
-------
TABLE 55b. ESTIMATES OF EIGHT-HOUR MAXIMUM DOSAGE EXPOSURES EXPERIENCED
BY OUTDOOR CHILDREN IN CHICAGO DURING WHICH OZONE CONCENTRATION
EXCEEDED 0.08 ppm AND EVRa RANGED FROM 13 LITERS- MIN1- M2 TO 27 LITERS- MIN'1- M2
Statistic"
Mean Estimate of the Number of Outdoor Children
Percent of Total Outdoor Children Population
Range in this percentage for 10 runs
Mean Estimate of Person-Occurrences
Percent of Total Person-Occurrences
Range in this percentage for 10 runs
Mean Estimate of Occurrences/Person Exposed
Percentage exposed for indicated number of days
1 Day
2 Days
3 Days
>3 Days
Regulatory scenario
8H1EX-100
93,077
19.69
15.29-24.08
115,045
0.11
0.08-0.15
1.24
79.68
17.96
1.82
0.55
8H1EX-90
31,445
6.65
3.74-12.81
33,033
0.03
0.02-0.06
1.05
94.47
5.53
0.00
0.00
8H1EX-80
1,570
0.33
0.00-1.14
1,570
d
0.00-0.01
1.00
100.00
0.00
0.00
0.00
8H1EX-70
0
0.00
0
0.00
-
-
8H5EX-90
109,660
23.20
16.82-32.71
134,610
0.13
0.10-0.18
1.23
80.32
17.02
1.81
0.85
8H5EX-80
34,651
7.33
4.34-11.34
36,388
0.04
0.02-0.06
1.05
96.31
2.63
1.07
0.00
cn
aEquivalent ventilation rate = (ventilation rate)/(body surface area).
bMean or range for 10 runs of pNEM/O3.
dLess than 0.01 percent.
eAII values less than 0.01 percent.
-------
TABLE 56a. ESTIMATES OF ONE-HOUR MAXIMUM DOSAGE EXPOSURES EXPERIENCED
BY OUTDOOR CHILDREN IN DENVER DURING WHICH OZONE CONCENTRATION
EXCEEDED 0.12 ppm AND EVRa EQUALED OR EXCEEDED 30 LITERS- MIN'1- M2
Statistic"
Mean Estimate of the Number of Outdoor Children
Percent of Total Outdoor Children Population
Range in this percentage for 10 runs
Mean Estimate of Person-Occurrences
Percent of Total Person-Occurrences
Range in this percentage for 10 runs
Mean Estimate of Occurrences/Person Exposed
Percentage exposed for indicated number of days
1 Day
2 Days
>2 Days
Regulatory scenario
Baseline
34
0.03
0.00-0.32
34
d
e
1.00
100.00
0.00
0.00
1H1EX-1200
12
0.01
0.00-0.11
12
d
e
1.00
100.00
0.00
0.00
1H1EX-100
0
0.00
0
0.00
-
-
en
aEquivalent ventilation rate = (ventilation rate)/(body surface area).
bMean or range for 10 runs of pNEM/O3.
cCurrent NAAQS.
dLess than 0.01 percent.
eAII values less than 0.01 percent.
-------
TABLE 56b. ESTIMATES OF ONE-HOUR MAXIMUM DOSAGE EXPOSURES EXPERIENCED
BY OUTDOOR CHILDREN IN DENVER DURING WHICH OZONE CONCENTRATION
EXCEEDED 0.12 ppm AND EVRa EQUALED OR EXCEEDED 30 LITERS MIN'1- M'
Statistic6
Mean Estimate of the Number of Outdoor Children
Percent of Total Outdoor Children Population
Range in this percentage for 10 runs
Mean Estimate of Person-Occurrences
Percent of Total Person-Occurrences
Range in this percentage for 10 runs
Mean Estimate of Occurrences/Person Exposed
Percentage exposed for indicated number of days
1 Day
2 Days
>2 Days
Regulatory scenario
8H1EX-100
818
0.77
0.00-2.07
818
d
0.00-0.01
1.00
100.00
0.00
0.00
8H1EX-90
71
0.07
0.00-0.42
71
d
e
1.00
100.00
0.00
0.00
8H1EX-80
32
0.03
0.00-0.30
32
d
e
1.00
100.00
0.00
0.00
8H1EX-70
0
0.00
-
0
0.00
-
-
8H5EX-90
246
0.23
0.00-1.49
246
d
0.00-0.01
1.00
100.00
0.00
0.00
8H5EX-80
87
0.08
0.00-0.39
87
d
e
1.00
100.00
0.00
0.00
Ol
en
Equivalent ventilation rate = (ventilation rate)/(body surface area).
bMean or range for 10 runs of pNEM/O3.
dLess than 0.01 percent.
eAII values less than 0.01 percent.
-------
TABLE 57a. ESTIMATES OF EIGHT-HOUR MAXIMUM DOSAGE EXPOSURES EXPERIENCED
BY OUTDOOR CHILDREN IN DENVER DURING WHICH OZONE CONCENTRATION
EXCEEDED 0.08 ppm AND EVRa RANGED FROM 13 LITERS- MIN1- M2 TO 27 LITERS- MIN1- M'2
Statistic"
Mean Estimate of the Number of Outdoor Children
Percent of Total Outdoor Children Population
Range in this percentage for 10 runs
Mean Estimate of Person-Occurrences
Percent of Total Person-Occurrences
Range in this percentage for 10 runs
Mean Estimate of Occurrences/Person Exposed
Percentage exposed for indicated number of days
1 Day
2 Days
3 Days
>3 Days
Regulatory scenario
Baseline
4,877
4.57
2.07-6.86
5,456
0.02
0.01-0.04
1.12
8744
12.56
0.00
0.00
1H1EX-120C
10,961
10.28
5.46-16.38
12,533
0.05
0.03-0.08
1.14
86.64
12.69
0.56
0.11
1H1EX-100
378
0.35
0.00-1.65
378
d
0.00-0.01
1.00
100.00
0.00
0.00
0.00
en
o
aEquivalent ventilation rate = (ventilation rate)/(body surface area).
bMean or range for 10 runs of pNEM/O3.
cCurrent NAAQS.
dLess than 0.01 percent.
-------
TABLE 57b. ESTIMATES OF EIGHT-HOUR MAXIMUM DOSAGE EXPOSURES EXPERIENCED
BY OUTDOOR CHILDREN IN DENVER DURING WHICH OZONE CONCENTRATION
EXCEEDED 0.08 ppm AND EVRa RANGED FROM 13 LITERS MIN'1 M'2 TO 27 LITERS- MINT1- M'2
Statistic"
Mean Estimate of the Number of Outdoor Children
Percent of Total Outdoor Children Population
Range in this percentage for 10 runs
Mean Estimate of Person-Occurrences
Percent of Total Person-Occurrences
Range in this percentage for 10 runs
Mean Estimate of Occurrences/Person Exposed
Percentage exposed for indicated number of days
1 Day
2 Days
3 Days
>3 Days
Regulatory scenario
8H1EX-100
32,522
30.49
26.34-34.52
55,103
0.24
0.21-0.29
1.69
54.26
26.87
14.31
4.56
8H1EX-90
16,734
15.69
10.78-21.65
20,018
0.09
0.05-0.14
1.20
80.65
19.07
0.28
0.00
8H1EX-80
2,020
1.89
0.42-3.73
2,020
0.01
0.00-0.02
1.00
100.00
0.00
0.00
0.00
8H1EX-70
0
0.00
0
0.00
-
-
8H5EX-90
10,844
10.17
5.46-14.27
13,575
0.06
0.04-0.08
1.25
79.68
14.53
4.64
1.15
8H5EX-80
2.309
2.16
0.15-5.78
2,437
0.01
0.00-0.03
1.06
95.35
4.65
000
0.00
Ol
aEquivalent ventilation rate = (ventilation rate)/(body surface area).
bMean or range for 10 runs of pNEM/O3.
-------
TABLE 58a. ESTIMATES OF ONE-HOUR MAXIMUM DOSAGE EXPOSURES EXPERIENCED
BY OUTDOOR CHILDREN IN HOUSTON DURING WHICH OZONE CONCENTRATION
EXCEEDED 0.12 ppm AND EVRa EQUALED OR EXCEEDED 30 LITERS- MIN"1- M~2
Statistic11
Mean Estimate of the Number of Outdoor Children
Percent of Total Outdoor Children Population
Range in this percentage for 10 runs
Mean Estimate of Person-Occurrences
Percent of Total Person-Occurrences
Range in this percentage for 10 runs
Mean Estimate of Occurrences/Person Exposed
Percentage exposed for indicated number of days
1 Day
2 Days
>2 Days
Regulatory scenario
Baseline
18,374
9.15
6.50-11.64
18,666
0.03
0.02-0.03
1.02
98.71
1.29
0.00
1H1EX-120C
299
0.15
0.00-1.05
299
d
e
1.00
100.00
0.00
0.00
1H1EX-100
0
0.00
0
0.00
-
-
en
co
aEquivalent ventilation rate = (ventilation rate)/(body surface area).
bMean or range for 10 runs of pNEM/O3.
°Current NAAQS.
dLess than 0.01 percent.
8AII values less than 0.01 percent.
-------
TABLE 58b. ESTIMATES OF ONE-HOUR MAXIMUM DOSAGE EXPOSURES EXPERIENCED
BY OUTDOOR CHILDREN IN HOUSTON DURING WHICH OZONE CONCENTRATION
EXCEEDED 0.12 ppm AND EVRa EQUALED OR EXCEEDED 30 LITERS- MIN'1- M'2
Statistic11
Mean Estimate of the Number of Outdoor Children
Percent of Total Outdoor Children Population
Range in this percentage for 10 runs
Mean Estimate of Person-Occurrences
Percent of Total Person-Occurrences
Range in this percentage for 10 runs
Mean Estimate of Occurrences/Person Exposed
Percentage exposed for indicated number of days
1 Day
2 Days
>2 Days
Regulatory scenario
8H1EX-100
1,452
0.72
0.00-2.91
1,452
d
0.00-0.01
1.00
100.00
0.00
0.00
8H1EX-90
358
0.18
0.00-0.64
358
d
e
1.00
100.00
0.00
0.00
8H1EX-80
74
0.04
0.00-0.37
74
d
e
1.00
100.00
0.00
0.00
8H1EX-70
0
0.00
0
0.00
-
-
8H5EX-90
799
0.40
0.00-1.05
799
d
e
1.00
100.00
0.00
0.00
8H5EX-80
452
0.23
0.00-1.41
452
d
e
1.00
*
100.00
0.00
0.00
en
-------
TABLE 59a. ESTIMATES OF EIGHT-HOUR MAXIMUM DOSAGE EXPOSURES EXPERIENCED
BY OUTDOOR CHILDREN IN HOUSTON DURING WHICH OZONE CONCENTRATION
EXCEEDED 0.08 ppm AND EVRa RANGED FROM 13 LITERS- MIN1- M'2 TO 27 LITERS- MIN'1- M2
Statistic1"
Mean Estimate of the Number of Outdoor Children
Percent of Total Outdoor Children Population
Range in this percentage for 10 runs
Mean Estimate of Person-Occurrences
Percent of Total Person-Occurrences
Range in this percentage for 10 runs
Mean Estimate of Occurrences/Person Exposed
Percentage exposed for indicated number of days
1 Day
2 Days
3 Days
>3 Days
Regulatory scenario
Baseline
137.146
68.30
60.51-75.57
345,211
0.47
0.40-0.54
2.52
28.18
28.11
21.68
22.03
1H1EX-120C
11,457
5.71
3.65-9.95
11,550
0.02
0.01-0.03
1.01
99.48
0.52
0.00
0.00
1H1EX-100
383
0.19
0.00-0.74
38^
d
e
1.00
100.00
0.00
0.00
0.00
0)
o
"Equivalent ventilation rate = (ventilation rate)/(body surface area).
bMean or range for 10 runs of pNEM/O3.
cCurrent NAAQS.
dLess than 0.01 percent.
eAII values less than 0.01 percent.
-------
TABLE 59b. ESTIMATES OF EIGHT-HOUR MAXIMUM DOSAGE EXPOSURES EXPERIENCED
BY OUTDOOR CHILDREN IN HOUSTON DURING WHICH OZONE CONCENTRATION
EXCEEDED 0.08 ppm AND EVR8 RANGED FROM 13 LITERS- MIN1- M'2 TO 27 LITERS- MIN 1- M2
Statistic"
Mean Estimate of the Number of Outdoor Children
Percent of Total Outdoor Children Population
Range in this percentage for 10 runs
Mean Estimate of Person-Occurrences
Percent of Total Person-Occurrences
Range in this percentage for 10 runs
Mean Estimate of Occurrences/Person Exposed
Percentage exposed for indicated number of days
1 Day
2 Days
3 Days
>3 Days
Regulatory scenario
8H1EX-100
38,544
19.20
12.28-25.77
47,923
0.07
0.04-0.09
1.24
79.31
17.57
2.66
0.47
8H1EX-90
13,798
6.87
4.59-11.46
14,552
0.02
0.01-0.03
1.05
95.06
4.94
0.00
0.00
8H1EX-80
2,640
1.31
0.00-3.77
2,640
d
0.00-0.01
1.00
100.00
0.00
0.00
0.00
8H1EX-70
204
0.10
0.00-0.64
204
d
e
1.00
100.00
0.00
0.00
0.00
8H5EX-90
49,320
24.56
15.76-30.46
58,454
0.08
0.05-0.10
1.19
82.47
15.83
1.65
0.05
8H5EX-80
16,331
8.13
4.61-12.24
17,113
0.02
0.01-0.04
1.05
95.93
4.07
0.00
0.00
aEquivalent ventilation rate = (ventilation rate)/(body surface area).
bMean or range for 10 runs of pNEM/O3.
dLess than 0.01 percent.
eAII values less than 0.01 percent.
-------
TABLE 60a, ESTIMATES OF ONE-HOUR MAXIMUM DOSAGE EXPOSURES EXPERIENCED
BY OUTDOOR CHILDREN IN LOS ANGELES DURING WHICH OZONE CONCENTRATION
EXCEEDED 0.12 ppm AND EVRa EQUALED OR EXCEEDED 30 LITERS- MIN'1- M'2
Statistic"
Mean Estimate of the Number of Outdoor Children
Percent of Total Outdoor Children Population
Range in this percentage for 10 runs
Mean Estimate of Person-Occurrences
Percent of Total Person-Occurrences
Range in this percentage for 10 runs
Mean Estimate of Occurrences/Person Exposed
Percentage exposed for indicated number of days
1 Day
2 Days
>2 Days
Regulatory scenario
Baseline
132,648
16.62
11.62-22.08
175,884
0.06
0.05-0.08
1.33
74.18
19.19
6.62
1H1EX-1200
114
0.01
0.14
114
d
e
1.00
100.00
0.00
0.00
1H1EX-100
0
0.00
0
0.00
-
-
CD
ro
aEquivalent ventilation rate = (ventilation rate)/(body surface area).
bMean or range for 10 runs of pNEM/O3.
cCurrent NAAQS.
dLess than 0.01 percent.
"All values less than 0.01 percent.
-------
TABLE 60b. ESTIMATES OF ONE-HOUR MAXIMUM DOSAGE EXPOSURES EXPERIENCED
BY OUTDOOR CHILDREN IN LOS ANGELES DURING WHICH OZONE CONCENTRATION
EXCEEDED 0.12 ppm AND EVRa EQUALED OR EXCEEDED 30 LITERS-MIN'1-M2
Statistic"
Mean Estimate of the Number of Outdoor Children
Percent of Total Outdoor Children Population
Range in this percentage for 10 runs
Mean Estimate of Person-Occurrences
Percent of Total Person-Occurrences
Range in this percentage for 10 runs
Mean Estimate of Occurrences/Person Exposed
Percentage exposed for indicated number of days
1 Day
2 Days
>2 Days
Regulatory scenario
8H1EX-100
6,029
0,76
0.32-1.31
6,029
d
e
1.00
100.00
0.00
0.00
8H1EX-90
622
0.08
0.00-0.27
622
d
e
1.00
100.00
0.00
0.00
8H1EX-80
114
0.01
0.00-0.14
114
d
e
1.00
100.00
0.00
0.00
8H1EX-70
0
0.00
0
0.00
-
-
8H5EX-90
2.418
0.30
0.00-0.84
2,532
d
e
1.05
94.87
5.13
0.00
8H5EX-80
332
0.04
0.00-0.14
332
d
e
1.00
100.00
0.00
0.00
O)
CO
aEquivalent ventilation rate = (ventilation rate)/(body surface area).
bMean or range for 10 runs of pNEM/O3.
dLess than 0.01 percent.
eAII values less than 0.01 percent.
-------
TABLE 61 a. ESTIMATES OF EIGHT-HOUR MAXIMUM DOSAGE EXPOSURES EXPERIENCED
BY OUTDOOR CHILDREN IN LOS ANGELES DURING WHICH OZONE CONCENTRATION
EXCEEDED 0.08 ppm AND EVRa RANGED FROM 13 LITERS- MIN1- M2 TO 27 LITERS-MIN 1-M
-2
Statistic"
Mean Estimate of the Number of Outdoor Children
Percent of Total Outdoor Children Population
Range in this percentage for 10 runs
Mean Estimate of Person-Occurrences
Percent of Total Person-Occurrences
Range in this percentage for 10 runs
Mean Estimate of Occurrences/Person Exposed
Percentage exposed for indicated number of days
1 Day
2 Days
3 Days
>3 Days
Regulatory scenario
Baseline
496,472
62.19
59.20-64.60
3,365.193
1.15
1.12-1.19
6.78
24.51
12.75
9.00
53.73
1H1EX-120C
7,584
0.95
0.30-1.55
7,698
d
e
1.02
98.77
1.23
0.00
0.00
1H1EX-100
0
0.00
0
0.00
-
-
aEquivalent ventilation rate = (ventilation rate)/(body surface area).
bMean or range for 10 runs of pNEM/O3.
cCurrent NAAQS.
dLess than 0.01 percent.
eAII values less than 0.01 percent.
-------
TABLE 61 b. ESTIMATES OF EIGHT-HOUR MAXIMUM DOSAGE EXPOSURES EXPERIENCED
BY OUTDOOR CHILDREN IN LOS ANGELES DURING WHICH OZONE CONCENTRATION
EXCEEDED 0.08 ppm AND EVRa RANGED FROM 13 LITERS MIN'1 M'2 TO 27 LITERS- MIN'1- M2
Statistic"
Mean Estimate of the Number of Outdoor Children
Percent of Total Outdoor Children Population
Range in this percentage for 10 runs
Mean Estimate of Person-Occurrences
Percent of Total Person-Occurrences
Range in this percentage for 10 runs
Mean Estimate of Occurrences/Person Exposed
Percentage exposed for indicated number of days
1 Day
2 Days
3 Days
>3 Days
Regulatory scenario
8H1EX-100
90,651
11.36
9.02-13.10
147.532
0.05
0.04-0.06
1.63
58.60
26.61
9.45
5.34
8H1EX-90
33.994
4.26
3.22-6.53
41.661
0.01
0.01-0.02
1.23
80.53
16.98
2.20
0.29
8H1EX-80
4,634
0.58
0.14-1.07
4,634
d
e
1.00
100.00
0.00
0.00
0.00
8H1EX-70
0
0.00
-
0
0.00
-
-
0.00
0.00
0.00
0.00
8H5EX-90
48,837
6.12
5.63-7.05
70.829
0.02
0.02-0.03
1.45
67.02
23.79
6.24
2.95
8H5EX-80
11,486
1.44
0.60-2.46
13.800
d
e
1.20
82.94
12.40
4.66
0.00
CD
Ol
aEquivalent ventilation rate = (ventilation rate)/(body surface area).
bMean or range for 10 runs of pNEM/O3.
dLess than 0.01 percent.
eAII values less than 0.01 percent.
-------
TABLE 62a. ESTIMATES OF ONE-HOUR MAXIMUM DOSAGE EXPOSURES EXPERIENCED
BY OUTDOOR CHILDREN IN MIAMI DURING WHICH OZONE CONCENTRATION
EXCEEDED 0.12 ppm AND EVRa EQUALED OR EXCEEDED 30 LITERS-MIN'1-M'2
Statistic"
Mean Estimate of the Number of Outdoor Children
Percent of Total Outdoor Children Population
Range in this percentage for 10 runs
Mean Estimate of Person-Occurrences
Percent of Total Person-Occurrences
Range in this percentage for 10 runs
Mean Estimate of Occurrences/Person Exposed
Percentage exposed for indicated number of days
1 Day
2 Days
>2 Days
Regulatory scenario
Baseline
0
0.00
0
0.00
-
-
1H1EX-1200
27
0.02
0.00-0.20
27
d
e
1.00
100.00
0.00
0.00
1H1EX-100
0
0.00
0
0.00
-
-
O)
CD
aEquivalent ventilation rate = (ventilation rate)/(body surface area).
bMean or range for 10 runs of pNEM/O3.
Current NAAQS.
dLess than 0.01 percent.
"All values less than 0.01 percent.
-------
TABLE 62b. ESTIMATES OF ONE-HOUR MAXIMUM DOSAGE EXPOSURES EXPERIENCED
BY OUTDOOR CHILDREN IN MIAMI DURING WHICH OZONE CONCENTRATION
EXCEEDED 0.12 ppm AND EVRa EQUALED OR EXCEEDED 30 LITERS- MIN'1- M'2
Statistic"
Mean Estimate of the Number of Outdoor Children
Percent of Total Outdoor Children Population
Range in this percentage for 10 runs
Mean Estimate of Person-Occurrences
Percent of Total Person-Occurrences
Range in this percentage for 10 runs
Mean Estimate of Occurrences/Person Exposed
Percentage exposed for indicated number of days
1 Day
2 Days
>2 Days
Regulatory scenario
8H1EX-100
850
0.64
0.00-2.10
850
d
0.00-0.01
1.00
100.00
0.00
0.00
8H1EX-90
140
0.10
0.00-1.05
140
d
e
1.00
100.00
0.00
0.00
8H1EX-80
0
0.00
0
0.00
-
-
8H1EX-70
0
0.00
0
0.00
-
-
8H5EX-90
1.306
0.98
0.00-2.07
1.306
d
0.00-0.01
1.00
100.00
0.00
0.00
8H5EX-80
28
0.02
0.00-0.20
28
d
e
1.00
100.00
0.00
0.00
O)
"Equivalent ventilation rate = (ventilation rate)/(body surface area).
bMean or range for 10 runs of pNEM/O3.
dLess than 0.01 percent.
eAII values less than 0.01 percent.
-------
TABLE 63a. ESTIMATES OF EIGHT-HOUR MAXIMUM DOSAGE EXPOSURES EXPERIENCED
BY OUTDOOR CHILDREN IN MIAMI DURING WHICH OZONE CONCENTRATION
EXCEEDED 0.08 ppm AND EVRa RANGED FROM 13 LITERS MIN 1 M'2 TO 27 LITERS MIN'1 M
Statistic"
Mean Estimate of the Number of Outdoor Children
Percent of Total Outdoor Children Population
Range in this percentage for 10 runs
Mean Estimate of Person-Occurrences
Percent of Total Person-Occurrences
Range in this percentage for 10 runs
Mean Estimate of Occurrences/Person Exposed
Percentage exposed for indicated number of days
1 Day
2 Days
3 Days
>3 Days
Regulatory scenario
Baseline
625
0.47
0.00-2.17
625
d
0.00-0.01
1.00
100.00
0.00
0.00
0.00
1H1EX-120C
5,709
4.27
1.35-8.60
5,867
0.01
0.00-0.02
1.03
97.75
2.25
0.00
0.00
1H1EX-100
149
0.11
0.00-0.56
149
d
e
1.00
100.00
0.00
0.00
0.00
OD
aEquivalent ventilation rate = (ventilation rate)/(body surface area).
bMean or range for 10 runs of pNEM/O3.
cCurrent NAAQS.
dLess than 0.01 percent.
eAII values less than 0.01 percent.
-------
TABLE 63b. ESTIMATES OF EIGHT-HOUR MAXIMUM DOSAGE EXPOSURES EXPERIENCED
BY OUTDOOR CHILDREN IN MIAMI DURING WHICH OZONE CONCENTRATION
EXCEEDED 0.08 ppm AND EVRa RANGED FROM 13 LITERS MIN'1 M'2 TO 27 LITERS MIN'1 M2
Statistic"
Mean Estimate of the Number of Outdoor Children
Percent of Total Outdoor Children Population
Range in this percentage for 10 runs
Mean Estimate of Person-Occurrences
Percent of Total Person-Occurrences
Range in this percentage for 10 runs
Mean Estimate of Occurrences/Person Exposed
Percentage exposed for indicated number of days
1 Day
2 Days
3 Days
>3 Days
Regulatory scenario
8H1EX-100
24,674
18.47
1 1 .26-28.49
30,049
0.06
0.03-0.08
1.22
81.34
14.44
4.22
0.00
8H1EX-90
9,106
6.82
1.99-12.02
10,352
0.02
0.01-0.04
1.14
87.89
12.11
0.00
0.00
8H1EX-80
1,040
0.78
0.00-4.61
1,040
d
0.00-0.01
1.00
100.00
0.00
0.00
0.00
8H1EX-70
0
0.00
0
0.00
-
-
8H5EX-90
33,040
24.73
17.18-30.34
41,357
0.08
0.06-0.10
1.25
78.43
18.54
2.56
0.47
8H5EX-80
15,672
11.73
8.21-16.76
16,042
0.03
0.02-0.05
1.02
98.05
1.95
0.00
000
O)
CD
aEquivalent ventilation rate = (ventilation rate)/(body surface area).
"Mean or range for 10 runs of pNEM/O3.
dLess than 0.01 percent.
-------
TABLE 64a. ESTIMATES OF ONE-HOUR MAXIMUM DOSAGE EXPOSURES EXPERIENCED
BY OUTDOOR CHILDREN IN NEW YORK DURING WHICH OZONE CONCENTRATION
EXCEEDED 0.12 ppm AND EVRa EQUALED OR EXCEEDED 30 LITERS-MIN'1-M'2
Statistic"
Mean Estimate of the Number of Outdoor Children
Percent of Total Outdoor Children Population
Range in this percentage for 10 runs
Mean Estimate of Person-Occurrences
Percent of Total Person-Occurrences
Range in this percentage for 10 runs
Mean Estimate of Occurrences/Person Exposed
Percentage exposed for indicated number of days
1 Day
2 Days
>2 Days
Regulatory scenario
Baseline
9,979
1.28
0.60-2.26
10,295
0.01
0.00-0.01
1.03
96.19
3.81
0.00
1H1EX-120C
164
0.02
0.18
164
d
e
1.00
100.00
0.00
0.00
1H1EX-100
0
0.00
0
0.00
-
-
"Equivalent ventilation rate = (ventilation rate)/(body surface area).
bMean or range for 10 runs of pNEM/O3.
'Current NAAQS.
dLess than 0.01 percent.
eAII values less than 0.01 percent.
-------
TABLE 64b. ESTIMATES OF ONE-HOUR MAXIMUM DOSAGE EXPOSURES EXPERIENCED
BY OUTDOOR CHILDREN IN NEW YORK DURING WHICH OZONE CONCENTRATION
EXCEEDED 0.12 ppm AND EVRa EQUALED OR EXCEEDED 30 LITERS-MIN'1- M2
Statistic15 .
Mean Estimate of the Number of Outdoor Children
Percent of Total Outdoor Children Population
Range in this percentage for 10 runs
Mean Estimate of Person-Occurrences
Percent of Total Person-Occurrences
Range in this percentage for 10 runs
Mean Estimate of Occurrences/Person Exposed
Percentage exposed for indicated number of days
1 Day
2 Days
>2 Days
Regulatory scenario
8H1EX-100
1,612
0.21
0.00-1.05
1,612
d
e
1.00
100.00
0.00
0.00
8H1EX-90
0
0.00
0
0.00
-
-
8H1EX-80
0
0.00
0
0.00
-
-
8H1EX-70
0
0.00
0
0.00
-
-
8H5EX-90
172
0.02
0.00-0.12
172
d
e
1.00
100.00
0.00
0.00
8H5EX-80
0
0.00
0
0.00
-
-
aEquivalent ventilation rate = (ventilation rate)/(body surface area).
bMean or range for 10 runs of pNEM/O3.
dLess than 0.01 percent.
"All values less than 0.01 percent.
-------
TABLE 65a. ESTIMATES OF EIGHT-HOUR MAXIMUM DOSAGE EXPOSURES EXPERIENCED
BY OUTDOOR CHILDREN IN NEW YORK DURING WHICH OZONE CONCENTRATION
EXCEEDED 0.08 ppm AND EVRa RANGED FROM 13 LITERS MIN'1 M'2 TO 27 LITERS- MIN'1- M2
" ' " ' ' "".:.".'."." - - ' "--ii1 ""-- ' j.m. .. . -. . .
Statistic"
Mean Estimate of the Number of Outdoor Children
Percent of Total Outdoor Children Population
Range in this percentage for 10 runs
Mean Estimate of Person-Occurrences
Percent of Total Person-Occurrences
Range in this percentage for 10 runs
Mean Estimate of Occurrences/Person Exposed
Percentage exposed for indicated number of days
1 Day
2 Days
3 Days
>3 Days
"" ' ' ''" ' ---' -.'- i .. ,_^._ , ., , ,,,, ,,. ....... _^_
Regulatory scenario
Baseline
321;060
41.02
37.70-43.20
722,616
0.43
0.41-0.46
2.25
41.74
24.12
16.27
17.87
1 ==
1H1EX-120C
42,144
5.39
3.89-6.34
52,207
0.03
0.02-0.04
1.24
79.90
16.48
3.32
0.29
- . , ...,,,_,, ,, , , , ,,_._
1H1EX-100
1,940
0.25
0.03-0.56
1,940
d
e
1.00
100.00
0.00
0.00
0.00
"Equivalent ventilation rate = (ventilation rate)/(body surface area).
bMean or range for 10 runs of pNEM/O3.
°Current NAAQS.
dLess than 0.01 percent.
eAII values less than 0.01 percent.
-------
TABLE 65b. ESTIMATES OF EIGHT-HOUR MAXIMUM DOSAGE EXPOSURES EXPERIENCED
BY OUTDOOR CHILDREN IN NEW YORK DURING WHICH OZONE CONCENTRATION
EXCEEDED 0.08 ppm AND EVRa RANGED FROM 13 LITERS MIN 1 M'2 TO 27 LITERS MIN'1 M'2
Statistic"
Mean Estimate of the Number of Outdoor Children
Percent of Total Outdoor Children Population
Range in this percentage for 10 runs
Mean Estimate of Person-Occurrences
Percent of Total Person-Occurrences
Range in this percentage for 10 runs
Mean Estimate of Occurrences/Person Exposed
Percentage exposed for indicated number of days
1 Day
2 Days
3 Days
>3 Days
Regulatory scenario
8H1EX-100
108,048
13.81
11.31-18.03
152,315
0.09
0.07-0.11
1.41
68.51
23.05
6.52
1.92
8H1EX-90
29,435
3.76
2.37-5.74
32,104
0.02
0.01-0.03
1.09
91.84
7.52
0.64
0.00
8H1EX-80
1,410
0.18
0.00-0.51
1,410
d
e
1.00
100.00
0.00
0.00
0.00
8H1EX-70
0
0.00
0
0.00
-
-
8H5EX-90
68,244
8.72
6.76-10.30
93,597
0.06
0.04-0.08
1.37
69.33
25.41
4.52
0.74
8H5EX-80
16,372
2.09
1.00-3.12
16,938
0.01
0.00-0.01
1.03
96.47
353
0.00
0.00
-vl
CO
"Equivalent ventilation rate = (ventilation rate)/(body surface area).
bMean or range for 10 runs of pNEM/O3.
dLess than 0.01 percent.
eAII values less than 0.01 percent.
-------
TABLE 66a. ESTIMATES OF ONE-HOUR MAXIMUM DOSAGE EXPOSURES EXPERIENCED
BY OUTDOOR CHILDREN IN PHILADELPHIA DURING WHICH OZONE CONCENTRATION
EXCEEDED 0.12 ppm AND EVRa EQUALED OR EXCEEDED 30 LITERS- MIN'1- M'2
Statistic"
Mean Estimate of the Number of Outdoor Children
Percent of Total Outdoor Children Population
Range in this percentage for 10 runs
Mean Estimate of Person-Occurrences
Percent of Total Person-Occurrences
Range in this percentage for 10 runs
Mean Estimate of Occurrences/Person Exposed
Percentage exposed for indicated number of days
1 Day
2 Days
>2 Days
Regulatory scenario
Baseline
9,676
3.51
1.26-5.78
10,136
0.02
0.01-0.03
1.05
94.74
5.26
0.00
1H1EX-120C
81
0.03
0.00-0.15
81
d
e
1.00
100.00
0.00
0.00
1H1EX-100
0
0.00
0
0.00
-
-
"Equivalent ventilation rate = (ventilation rate)/(body surface area).
bMean or range for 10 runs of pNEM/O3.
°Current NAAQS.
dLess than 0.01 percent.
eAII values less than 0.01 percent.
-------
TABLE 66b. ESTIMATES OF ONE-HOUR MAXIMUM DOSAGE EXPOSURES EXPERIENCED
BY OUTDOOR CHILDREN IN PHILADELPHIA DURING WHICH OZONE CONCENTRATION
EXCEEDED 0.12 ppm AND EVRa EQUALED OR EXCEEDED 30 LITERS- MIN1- M2
Statistic6
Mean Estimate of the Number of Outdoor Children
Percent of Total Outdoor Children Population
Range in this percentage for 10 runs
Mean Estimate of Person-Occurrences
Percent of Total Person-Occurrences
Range in this percentage for 10 runs
Mean Estimate of Occurrences/Person Exposed
Percentage exposed for indicated number of days
1 Day
2 Days
>2 Days
Regulatory scenario
8H1EX-100
139
0.05
0.00-0.21
139
d
e
1.00
100.00
0.00
0.00
8H1EX-90
0
0.00
0
0.00
-
-
-
8H1EX-80
0
0.00
0
0.00
-
-
-
8H1EX-70
0
0.00
0
0.00
-
-
-
8H5EX-90
0
0.00
0
0.00
-
-
-
8H5EX-80
0
0.00
0
0.00
-
-
-
01
"Equivalent ventilation rate = (ventilation rate)/(body surface area).
bMean or range for 10 runs of pNEM/O3.
dLess than 0.01 percent.
eAII values less than 0.01 percent.
-------
TABLE 67a. ESTIMATES OF EIGHT-HOUR MAXIMUM DOSAGE EXPOSURES EXPERIENCED
BY OUTDOOR CHILDREN IN PHILADELPHIA DURING WHICH OZONE CONCENTRATION
EXCEEDED 0.08 ppm AND EVRa RANGED FROM 13 LITERS- MIN1- M2 TO 27 LITERS-MIN'1-M
-2
Statistic6
Mean Estimate of the Number of Outdoor Children
Percent of Total Outdoor Children Population
Range in this percentage for 10 runs
Mean Estimate of Person-Occurrences
Percent of Total Person-Occurrences
Range in this percentage for 10 runs
Mean Estimate of Occurrences/Person Exposed
Percentage exposed for indicated number of days
1 Day
2 Days
3 Days
>3 Days
Regulatory scenario
Baseline
186,273
67.66
65.59-70.18
580,171
0.98
0.89-1.10
3.12
21.59
22.11
20.98
35.32
1H1EX-1206
58,377
21.20
18.67-24.29
79,318
0.13
0.12-0.17
1.36
74.07
19.01
4.41
2.51
1H1EX-100
7,500
2.72
1.11-4.44
7.698
0.01
0.01-0.02
1.03
97.07
2.93
0.00
0.00
en
"Equivalent ventilation rate = (ventilation rate)/(body surface area).
bMean or range for 10 runs of pNEM/O3.
cCurrent NAAQS.
-------
TABLE 67b. ESTIMATES OF EIGHT-HOUR MAXIMUM DOSAGE EXPOSURES EXPERIENCED
BY OUTDOOR CHILDREN IN PHILADELPHIA DURING WHICH OZONE CONCENTRATION
EXCEEDED 0.08 ppm AND EVRa RANGED FROM 13 LITERS- MIN 1- M2 TO 27 LITERS-MIN1-M
-2
Statistic"
Mean Estimate of the Number of Outdoor Children
Percent of Total Outdoor Children Population
Range in this percentage for 10 runs
Mean Estimate of Person-Occurrences
Percent of Total Person-Occurrences
Range in this percentage for 10 runs
Mean Estimate of Occurrences/Person Exposed
Percentage exposed for indicated number of days
1 Day
2 Days
3 Days
>3 Days
Regulatory scenario
8H1EX-100
68,765
24.98
20.88-30.23
98.111
0.17
0.14-0.20
1.43
69.98
19.55
8.87
1.60
8H1EX-90
17,971
6.53
4.32-10.65
20,342
0.03
0.02-0.06
1.13
87.77
11.87
0.37
0.00
8H1EX-80
1,634
0.59
0.12-2.27
1,634
d
0.00-0.01
1.00
100.00
0.00
0.00
0.00
8H1EX-70
0
0.00
0
0.00
-
-
8H5EX-90
50,283
18.26
15.05-21.82
65,900
0.11
0.08-0.14
1.31
75.47
20.44
2.45
1.64
8H5EX-80
6,182
2.25
0.91-4.83
7,087
0.01
0.00-0.02
1.15
85.64
14.36
0.00
0.00
aEquivalent ventilation rate = (ventilation rate)/(body surface area).
bMean or range for 10 runs of pNEM/O3.
"Less than 0.01 percent.
-------
TABLE 68a. ESTIMATES OF ONE-HOUR MAXIMUM DOSAGE EXPOSURES EXPERIENCED
BY OUTDOOR CHILDREN IN ST. LOUIS DURING WHICH OZONE CONCENTRATION
EXCEEDED 0.12 ppm AND EVRa EQUALED OR EXCEEDED 30 LITERS MIN'1- M'2
Statistic"
Mean Estimate of the Number of Outdoor Children
Percent of Total Outdoor Children Population
Range in this percentage for 10 runs
Mean Estimate of Person-Occurrences
Percent of Total Person-Occurrences
Range in this percentage for 10 runs
Mean Estimate of Occurrences/Person Exposed
Percentage exposed for indicated number of days
1 Day
2 Days
>2 Days
Regulatory scenario
Baseline
550
0.43
0.00-1.39
550
d
0.00-0.01
1.00
100.00
0.00
0.00
1H1EX-1200
107
0.08
0.00-0.41
107
d
e
1.00
100.00
0.00
0.00
1H1EX-100
o
0.00
o
0.00
_
_
CO
"Equivalent ventilation rate = (ventilation rate)/(body surface area).
bMean or range for 10 runs of pNEM/O3.
cCurrent NAAQS.
dLess than 0.01 percent.
"All values less than 0.01 percent.
-------
TABLE 68b. ESTIMATES OF ONE-HOUR MAXIMUM DOSAGE EXPOSURES EXPERIENCED
BY OUTDOOR CHILDREN IN ST. LOUIS DURING WHICH OZONE CONCENTRATION
EXCEEDED 0.12 ppm AND EVRa EQUALED OR EXCEEDED 30 LITERS-MIN'1-M'2
Statistic"
Mean Estimate of the Number of Outdoor Children
Percent of Total Outdoor Children Population
Range in this percentage for 10 runs
Mean Estimate of Person-Occurrences
Percent of Total Person-Occurrences
Range in this percentage for 10 runs
Mean Estimate of Occurrences/Person Exposed
Percentage exposed for indicated number of days
1 Day
2 Days
>2 Days
Regulatory scenario
8H1EX-100
145
0.11
0.00-1.13
145
d
0.00-0.01
1.00
100.00
0.00
0.00
8H1EX-90
0
0.00
-
0
0.00
-
-
_
_
-
8H1EX-80
0
0.00
-
0
0.00
-
-
_
_
-
8H1EX-70
0
0.00
-
0
0.00
-
-
8H5EX-90
112
0.09
0.00-0.27
112
d
e
1.00
100.00
0.00
0.00
8H5EX-80
0
0.00
0
0.00
_
-
CD
aEquivalent ventilation rate = (ventilation rate)/(body surface area).
bMean or range for 10 runs of pNEM/O3.
dLess than 0.01 percent.
eAII values less than 0.01 percent.
-------
TABLE 69a ESTIMATES OF EIGHT-HOUR MAXIMUM DOSAGE EXPOSURES EXPERIENCED
BY OUTDOOR CHILDREN IN ST. LOUIS DURING WHICH OZONE CONCENTRATION
EXCEEDED 0 08 ppm AND EVRa RANGED FROM 13 LITERS-MIN1 M'2 TO 27 LITERS-MIN'1-M2
Statistic"
Mean Estimate of the Number of Outdoor Children
Percent of Total Outdoor Children Population
Range in this percentage for 10 runs
Mean Estimate of Person-Occurrences
Percent of Total Person-Occurrences
Range in this percentage for 10 runs
Mean Estimate of Occurrences/Person Exposed
Percentage exposed for indicated number of days
1 Day
2 Days
3 Days
>3 Days
Regulatory scenario
Baseline
20,607
16.07
12.25-19.96
25,729
0.09
0.07-0.13
1.25
78.96
18.06
2.80
0.19
1H1EX-1200
20,205
15.75
12.14-18.23
28,249
0.10
0.07-0.14
1.40
70.78
22.97
3.37
2.89
1H1EX-100
1.264
0.99
0.00-2.32
1,264
d
0.00-0.01
1.00
100.00
0.00
0.00
0.00
CO
o
"Equivalent ventilation rate = (ventilation rate)/(body surface area).
bMean or range for 10 runs of pNEM/O3.
cCurrent NAAQS.
dLess than 0.01 percent.
-------
TABLE 69b. ESTIMATES OF EIGHT-HOUR MAXIMUM DOSAGE EXPOSURES EXPERIENCED
BY OUTDOOR CHILDREN IN ST. LOUIS DURING WHICH OZONE CONCENTRATION
EXCEEDED 0.08 ppm AND EVRa RANGED FROM 13 LITERS- MIN1- M2 TO 27 LITERS- MIN'1- M2
Statistic"
Mean Estimate of the Number of Outdoor Children
Percent of Total Outdoor Children Population
Range in this percentage for 10 runs
Mean Estimate of Person-Occurrences
Percent of Total Person-Occurrences
Range in this percentage for 10 runs
Mean Estimate of Occurrences/Person Exposed
Percentage exposed for indicated number of days
1 Day
2 Days
3 Days
>3 Days
Regulatory scenario
8H1EX-100
19,910
15.52
12.67-17.48
24,730
0.09
0.07-0.11
1.24
81.45
14.23
2.99
1.32
8H1EX-90
3,980
3.10
1 .74-4.62
4,163
0.02
0.01-0.02
1.05
94.38
5.62
0.00
0.00
8H1EX-80
19
0.01
0.00-0.15
19
d
e
1.00
100.00
0.00
0.00
0.00
8H1EX-70
0
0.00
-
0
0.00
-
-
_
_
_
-
8H5EX-90
28,486
22.21
18.94-30.58
39,291
0.14
0.13-0.19
1.38
70.26
21.80
7.37
0.58
8H5EX-80
6,844
5.34
2.21-8.38
7.496
0.03
0.01-0.04
1.10
91.59
7.96
0.45
0.00
CD
"Equivalent ventilation rate = (ventilation rate)/(body surface area).
bMean or range for 10 runs of pNEM/O3.
dLess than 0.01 percent.
eAII values less than 0.01 percent.
-------
TABLE 70a. ESTIMATES OF ONE-HOUR MAXIMUM DOSAGE EXPOSURES EXPERIENCED
BY OUTDOOR CHILDREN IN WASHINGTON D.C. DURING WHICH OZONE CONCENTRATION
EXCEEDED 0.12 ppm AND EVRa EQUALED OR EXCEEDED 30 LITERS- MIN'1- M'2
Statistic"
Mean Estimate of the Number of Outdoor Children
Percent of Total Outdoor Children Population
Range in this percentage for 10 runs
Mean Estimate of Person-Occurrences
Percent of Total Person-Occurrences
Range in this percentage for 10 runs
Mean Estimate of Occurrences/Person Exposed
Percentage exposed for indicated number of days
1 Day
2 Days
>2 Days
Regulatory scenario
Baseline
3,626
1.82
0.49-3.54
3,626
0.01
0.00-0.02
1.00
100.00
0.00
0.00
1H1EX-1200
36
0.02
0.00-0.18
36
d
e
1.00
100.00
0.00
0.00
1H1EX-100
0
0.00
0
0.00
-
-
oo
M
aEquivalent ventilation rate = (ventilation rate)/(body surface area).
bMean or range for 10 runs of pNEM/O3.
Current NAAQS.
dLess than 0.01 percent.
"All values less than 0.01 percent.
-------
TABLE 70b. ESTIMATES OF ONE-HOUR MAXIMUM DOSAGE EXPOSURES EXPERIENCED
BY OUTDOOR CHILDREN IN WASHINGTON D.C. DURING WHICH OZONE CONCENTRATION
EXCEEDED 0.12 ppm AND EVRa EQUALED OR EXCEEDED 30 LITERS- MIN'1- M2
Statistic6
Mean Estimate of the Number of Outdoor Children
Percent of Total Outdoor Children Population
Range in this percentage for 10 runs
Mean Estimate of Person-Occurrences
Percent of Total Person-Occurrences
Range in this percentage for 10 runs
Mean Estimate of Occurrences/Person Exposed
Percentage exposed for indicated number of days
1 Day
2 Days
>2 Days
Regulatory scenario
8H1EX-100
0
0.00
0
0.00
-
-
8H1EX-90
0
0.00
0
0.00
-
-
8H1EX-80
0
0.00
0
0.00
-
-
8H1EX-70
0
0.00
0
0.00
-
-
8H5EX-90
321
0.16
0.00-0.66
321
d
e
1.00
100.00
0.00
0.00
8H5EX-80
0
0.00
0
0.00
-
-
oo
OJ
aEquivalent ventilation rate = (ventilation rate)/(body surface area).
bMean or range for 10 runs of pNEM/O3.
dLess than 0.01 percent.
eAII values less than 0.01 percent.
-------
TABLE 71 a. ESTIMATES OF EIGHT-HOUR MAXIMUM DOSAGE EXPOSURES EXPERIENCED
BY OUTDOOR CHILDREN IN WASHINGTON D.C. DURING WHICH OZONE CONCENTRATION
EXCEEDED 0.08 ppm AND EVRa RANGED FROM 13 LITERS- MIN1- M2 TO 27 LITERS- MIN'1- M'2
Statistic"
Mean Estimate of the Number of Outdoor Children
Percent of Total Outdoor Children Population
Range in this percentage for 10 runs
Mean Estimate of Person-Occurrences
Percent of Total Person-Occurrences
Range in this percentage for 10 runs
Mean Estimate of Occurrences/Person Exposed
Percentage exposed for indicated number of days
1 Day
2 Days
3 Days
>3 Days
Regulatory scenario
Baseline
111,887
56.26
52.88-61.87
256,997
0.60
0.57-0.63
2.30
36.81
28.12
17.97
17.10
1H1EX-120C
25,108
12.63
9.10-15.10
35,069
0.08
0.07-0.10
1.40
72.01
16.14
10.03
1.81
1H1EX-100
2,926
1.47
0.18-2.93
2,998
0.01
0.00-0.01
1.02
97.65
2.35
0.00
0.00
Equivalent ventilation rate = (ventilation rate)/(body surface area).
bMean or range for 10 runs of pNEM/O3.
°Current NAAQS.
-------
TABLE 71 b ESTIMATES OF EIGHT-HOUR MAXIMUM DOSAGE EXPOSURES EXPERIENCED
BY OUTDOOR CHILDREN IN WASHINGTON D.C. DURING WHICH OZONE CONCENTRATION
EXCEEDED 0 08 ppm AND EVRa RANGED FROM 13 LITERS MIN 1 M2 TO 27 LITERS MIN'1 M
Statistic11
Mean Estimate of the Number of Outdoor Children
Percent of Total Outdoor Children Population
Range in this percentage for 10 runs
Mean Estimate of Person-Occurrences
Percent of Total Person-Occurrences
Range in this percentage for 10 runs
Mean Estimate of Occurrences/Person Exposed
Percentage exposed for indicated number of days
1 Day
2 Days
3 Days
>3 Days
8H1EX-100
31 ,209
15.69
13.92-17.68
47,544
0.11
0.09-0.14
1.52
66.31
21.13
8.00
456
8H1EX-90
10,353
5.21
2.37-7.14
12,219
0.03
0.02-0.04
1.18
83.03
13.56
2,65
0.76
=============================================
Regulatory scenario
8H1EX-80
2,043
1.03
0.00-2.35
2,043
d
0.00-0.01
1.00
100.00
0.00
0.00
0.00
8H1EX-70
36
0.02
0.00-0.18
36
d
e
1.00
100.00
0.00
0.00
0.00
8H5EX-90
44,110
22.18
18.20-27.89
57,474
0.14
0.11-0.18
1.30
75.76
20.29
2.43
1.52
8H5EX-80
8,284
4.17
0.97-8.37
8.635
0.02
0.00-0.04
1.04
95.41
4.59
0.00
0.00
CD
cn
aEquivalent ventilation rate = (ventilation rate)/(body surface area).
bMean or range for 10 runs of pNEM/O3.
dLess than 0.01 percent.
eAII values less than 0.01 percent.
-------
comprise 9.15 percent of the total outdoor children population in Houston. Entries in
the third row indicate that the percentage values ranged from 6.50 to 11.64 percent
over the 10 runs.
The fourth row in Tables 58a and 58b lists 10-run mean estimates for the
number of person-occurrences in which an outdoor child in Houston experienced a
one-hour maximum daily dosage exposure during which the ozone concentration
exceeded 0.12 ppm and the EVR equaled or exceeded 30 liters min'1- m'2. As
each child can experience more than one person-occurrence during the Houston
ozone season, the estimated number of person-occurrences can exceed the number
of persons exposed at the specified levels. Under baseline conditions, for example,
the 10-run mean for person-occurrences (18,666) is larger than the number of
exposed children listed in the first row (18,374).
The total possible number of one-hour person-occurrences is equal to
73,290,175 -- the product of the number of Houston outdoor children (200,795) and
the number of days in the Houston ozone season (365). According to the value
listed in the fifth row of Table 58a, 18,666 person-occurrences is 0.03 percent of the
total possible number of person-occurrences; that is, 18,666/73,290,175 = 0.03
percent. Entries in the sixth row of Table 58a indicate that the percentage values
ranged from 0.02 to 0.03 percent over the 10 runs.
The seventh row in the table lists the ratio of person-occurrences to people
exposed based on 10-run means. Under baseline conditions, the ratio is
18,666/18,374 or 1.02.
The last three rows in Table 58a provide exposure frequency statistics for
outdoor children who experienced the specified exposure conditions on at least one
day. Of the 18,374 outdoor children exposed under baseline conditions, 98.71
percent were exposed for one day only while 1.29 percent were exposed for exactly
two days. No one was exposed for more than two days.
Tables 59a and 59b use this same general table format to present eight-hour
daily maximum dosage estimates for Houston. The statistics in the first row are the
10-run mean estimates (by scenario) for the number of outdoor children in Houston
186
-------
who experienced one or more eight-hour maximum daily dosage exposures during
which the ozone concentration exceeded 0.08 ppm and the EVR ranged from 13
liters min"1- m~2 to 27 liters min"1- m'2. Under baseline conditions, 137,146
outdoor children are estimated to have experienced the specified exposure. This
value is equivalent to 68.30 percent of the total outdoor children population in
Houston. The percentage values ranged from 60.51 to 75.57 percent over the 10
runs.
The fourth row in Tables 59a and 59b lists 10-run mean estimates for the
number of person-occurrences in which an outdoor child in Houston experienced an
eight-hour maximum daily dosage exposure during which the ozone concentration
exceeded 0.08 ppm and the EVR ranged from 13 liters min"1- m"2 to 27 liters
min"1- m"2. The 10-run mean for person-occurrences under baseline conditions
(345,211) is more than 2.5 times the number of exposed children listed in the first
row (137,146).
Consistent with the one-hour analysis, the total possible number of eight-hour
person-occurrences is equal to 73,290,175 - the product of the number of Houston
outdoor children (200,795) and the number of days in the Houston ozone season
(365). According to the value listed in the fifth row of Table 59a, 345,211 person-
occurrences is 0.47 percent of the total possible number of person-occurrences.
The baseline percentage values ranged from 0.40 to 0.54 percent over the 10 runs.
The seventh row in Table 59a lists the ratio of person-occurrences to people
exposed based on 10-run means. Under baseline conditions, the ratio is
345,211/137,146 or 2.52.
Table 59a concludes with four rows listing exposure frequency statistics for
outdoor children who experienced the specified exposure conditions on at least one
day. Of the 137,146 outdoor children exposed under baseline conditions, 28.18
percent were exposed for one day only, 28.11 percent were exposed for exactly two
days, and 21.68 percent were exposed for exactly three days. The remaining 22.03
percent of the outdoor children were exposed for more than three days.
187
-------
Figures 2a through 5b are graphs showing eight-hour daily maximum dose
exposures for outdoor children under various air quality scenarios. Two graphs are
provided for each of four study areas (Philadelphia, Houston, New York, and
Washington). The graphs use two indicators to characterize ozone exposure:
Number of children experiencing eight-hour daily maximum-dose
exposures on one or more days under moderate exertion conditions,
Number of occurrences in which a child experiences an eight-hour daily
maximum dose exposure under moderate exertion conditions.
Moderate exertion conditions are defined as an EVR level between 13 and 27 liters-
min"1 m"2.
Figure 2a presents results for the first indicator (number of children) based on
applications of pNEM/03 to Philadelphia. Nine distributions are plotted on the
graph: one for baseline ("as is") conditions; two for one-hour, one-exceedance
standards (1112 and 1110); four for eight-hour, one-exceedance standards; and two
for eight-hour, five-exceedance standards. The first digit in the code for each
standard indicates the averaging time; the second digit specifies the number of
exceedances. The last two digits indicate the ozone concentration of the standard
expressed in pphm. For example, 8508 indicates an eight-hour five exceedance
standard with ozone concentration equal to 8 pphm or 0.08 ppm.
The ordinate (y coordinate) of each point on the graph shows the number of
children with one or more eight-hour daily maximum dose exposures equal to or
above the ozone concentration indicated by the point's abscissa (x coordinate). In
Figure 2a, the "as is" curve is associated with the highest number of children
exposed when the specified ozone concentration falls between 0.05 ppm and 0.14
ppm. The nine curves tend to converge at lower and higher ozone concentrations.
In a similar manner, the 8107 standard tends to be associated with lowest number
of children exposed when the specified ozone concentration falls between 0.03 and
0.07 ppm.
Figures 2a through 5b provide eight-hour daily maximum dose distributions
for exposures occurring under moderate exertion conditions (EVR values between
188
-------
FIGURE 2a. EIGHT-HOUR.DAILY MAXIMUM DOSE EXPOSURE DISTRIBUTIONS FOR
OUTDOOR CHILDREN EXPOSED ON ONE OR MORE DAYS UNDER MODERATE
EXERTION (EVR 13-27 LITERS/MIN-M2) IN PHILADELPHIA, PA
300
0.02 0.04
0.06 0.08 0.1 0.12 0.14
CONCENTRATION, PPM
0.16 0.18
0.2
FIGURE 2b. EIGHT-HOUR DAILY MAXIMUM DOSE EXPOSURE DISTRIBUTIONS OF
TOTAL OCCURRENCES FOR OUTDOOR CHILDREN EXPOSURE UNDER MODERATE
EXERTION (EVR 13-27 LITERS/MIN-M2) IN PHILADELPHIA, PA
16,000
ASIS
«
1112
-*-
8109
-*-
8108
811(3
8508
8107
1110
8509
\ /
0.02 0.04 0.06 0.08 0.1 0.12 0.14
CONCENTRATION, PPM
0.16 0.18 0.2
189
-------
FIGURE 3a. EIGHT-HOUR DAILY MAXIMUM DOSE EXPOSURE DISTRIBUTIONS FOR
OUTDOOR CHILDREN EXPOSED ON ONE OR MORE DAYS UNDER MODERATE
EXERTION (EVR 13-27 LITERS/MIN-M2) IN HOUSTON, TX
250
ASIS
1112
8109
-*
8108
8508
-*
8107
1110
8509
_
~ \
0.02 0.04
0.06 0.08 0.1 0.12 0.14
CONCENTRATION, PPM
0.16 0.18 0.2
FIGURE 3b. EIGHT-HOUR DAILY MAXIMUM DOSE EXPOSURE DISTRIBUTIONS OF
TOTAL OCCURRENCES FOR OUTDOOR CHILDREN EXPOSURE UNDER MODERATE
EXERTION (EVR 13-27 LITERS/MIN-M2) IN HOUSTON, TX
25,000
0.02 0.04
0.06 0.08 0.1 0.12 0.14
CONCENTRATION, PPM
0.16 0.18 0.2
190
-------
FIGURE 4a. EIGHT-HOUR DAILY MAXIMUM DOSE EXPOSURE DISTRIBUTIONS FOR
OUTDOOR CHILDREN EXPOSED ON ONE OR MORE DAYS UNDER MODERATE
EXERTION (EVR 13-27 LITERS/MIN-M2) IN NEW YORK, NY
0.02 0.04
0.06 0.08 0.1 0.12 0.14
CONCENTRATION, PPM
0.16
FIGURE 4b. EIGHT-HOUR DAILY MAXIMUM DOSE EXPOSURE DISTRIBUTIONS OF
TOTAL OCCURRENCES FOR OUTDOOR CHILDREN EXPOSURE UNDER MODERATE
EXERTION (EVR 13-27 LITERS/MIN-M2) IN NEW YORK, NY
50,000
ASIS
1112
-+
8109
-*
8108
8110
8508
-%r
8107
1110
8509
0.02 0.04 0.06 0.08 0.1 0.12 0.14
CONCENTRATION. PPM
0.16 0.18 0.2
-------
FIGURE 5a. EIGHT-HOUR DAILY MAXIMUM DOSE EXPOSURE DISTRIBUTIONS FOR
OUTDOOR CHILDREN EXPOSED ON ONE OR MORE DAYS UNDER MODERATE
EXERTION (EVR 13-27 LITERS/MIN-M2) IN WASHINGTON, D.C.
ASIS
0.02 0.04
0.16 0.18
0.06 0.08 0.1 0.12 0.14
CONCENTRATION, PPM
FIGURE 5b. EIGHT-HOUR DAILY MAXIMUM DOSE EXPOSURE DISTRIBUTIONS OF
TOTAL OCCURRENCES FOR OUTDOOR CHILDREN EXPOSURE UNDER MODERATE
EXERTION (EVR 13-27 LITERS/MIN-M2) IN WASHINGTON, D.C.
12,000
0.02 0.04
0.06 0.08 0.1 0.12 0.14
CONCENTRATION, PPM
0.16 0.18 0.2
192
-------
13 and 27 liters- min"1- m"2). Appendix E provides graphs for one-hour exposures
for two other EVR ranges of interest to EPA: 16 to 30 liters- min"1- m"2 and 30+
liters- min"1- m'2.
193
-------
SECTION 8
PRINCIPAL LIMITATIONS OF THE pNEM/O3 METHODOLOGY
The pNEM/03 methodology was developed specifically to meet the
requirements of OAQPS for a computer-based model capable of simulating the
ozone exposures of specific population groups under alternative NAAQS. In
addition to meeting these needs, the designers of pNEM/O3 have attempted to
create a model which is flexible in application and easy to upgrade. The model was
deliberately constructed as a collection of stand-alone algorithms organized within a
modular framework. For this reason, analysts can revise individual algorithms
without the need to make major changes to other parts of the model.
The structure of each algorithm in pNEM/O3 is largely determined by the
characteristics of the available input data. For example, the algorithm used to
construct a season-long exposure event sequence for each cohort is constrained by
the fact that none of the available time/activity studies provides more than three
days of diary data for any one subject. To make maximum use of the available
diary data, the pNEM/O3 sequencing algorithm constructs each exposure event
sequence by sampling data from more than one subject. The other pNEM/O3
algorithms are similarly designed to make best use of available data bases.
In evaluating the exposure estimates presented in this report, the reader
should note that the model has a number of limitations which may affect its
accuracy. These limitations are usually the result of limitations in the input data
bases. The available data were typically collected for purposes other than use in a
population exposure model. Consequently, these data frequently represent special
sets of conditions which differ from those assumed by pNEM/O3. In these
situations, analysts must exercise a certain degree of judgement in adapting the
data for use in pNEM/03.
194
-------
This section presents a brief discussion of the principal limitations in the
pNEM/O3 methodology as applied to outdoor children. The limitations are
organized according to five major components of the model: time/activity patterns,
equivalent ventilation rates, air quality adjustment, the mass balance model, and the
estimation of cohort populations.
Section 7 presented pNEM/O3 exposure estimates based on the assumption
that a specified urban area just attained a particular standard. One of the standards
under review (designated 8H5EX-80) stated that the expected exceedance rate for
daily maximum 8-hour ozone concentrations above 80 ppb shall not be greater than
five. To simulate this standard, the ozone data reported by the historical "high
ozone" monitor for a specified year was adjusted so that the sixth highest daily
maximum concentration equaled 80 ppb. Researchers assumed that this approach
represented average attainment conditions when compliance was determined over a
three-year period.
Subsection 8.6 presents results of alternative exposure assessments in which
the data for the high ozone monitor were adjusted to permit 10 exceedances of 80
ppb. This scenario represents a reasonable upper-bound for the number of
exceedances that could occur in any one year under the 8H5EX-80 standard when
16 exceedances are permitted to occur over a three-year period.
8.1 Time/Activity Patterns
In the general pNEM/03 methodology, the exposure-related activities of each
cohort are represented by a multi-day exposure event sequence which spans a
specified ozone season. Each sequence is constructed by an algorithm which
selects 24-hour (midnight-to-midnight) activity patterns from a specially prepared
database. This database contains data from one or more time/activity studies in
which subjects recorded their daily activities in diaries.
In the application of pNEM/O3 to outdoor children, the time/activity database
consisted of diary data obtained from 479 subjects identified as outdoor children.
The database contained 792 person-days of data, an average of slightly less than
two days per subject. These data should adequately characterize the spectrum of
activity patterns associated with outdoor children.
195
-------
The subjects who contributed to this database may not provide a balanced
representation of U.S. outdoor children. The majority of subjects (97 percent)
resided in either the State of California (337 subjects) or in Cincinnati (130 subjects).
Three subjects resided in Washington, DC, and nine subjects resided in Valdez, AK.
Random selection protocols were used in the selection of the 453 subjects
who participated in the Cincinnati, Washington, and California studies. The
remaining 26 subjects participated in the two Los Angeles studies and were solicited
using non-random protocols.
Analysts used time/activity data obtained from these 479 subjects to represent
the activities of outdoor children in nine study areas. Only two of these study areas
(Washington and Los Angeles) were locations of diary studies which contributed
time/activity data to the analysis. Although the algorithm which constructs exposure
event sequences attempts to account for effects of local climate on activity, it is
unlikely that this adjustment procedure corrects for all inter-city differences in
children's activities. Time/activity patterns are likely to be affected by a variety of
local factors, including topography, land-use, traffic patterns, mass transit systems,
and recreational opportunities.
As discussed previously, the average subject provided less than two
days of diary data. For this reason, the construction of each season-long exposure
event sequence required either the repetition of data from one subject or the use of
data from multiple subjects. The latter approach was used in the outdoor children
pNEM/03 analyses to better represent the variability of exposure expected to occur
among the children included in each cohort. The principal deficiency of this
approach is that it may not adequately account for the day-to-day repetition of
activities common to individual children. Using activities from different subjects may
underestimate multiple occurrences of high exertion and/or outdoor exposure for
those segments of the population who engage in repetitive outdoor activities.
8.2 Equivalent Ventilation Rates
The application of pNEM/03 to outdoor children marks the first use of a
newly-developed algorithm for estimating EVR values. The algorithm applies one of
four Monte Carlo models to each exposure event, the selected model depending on
196
-------
the demographic group of the cohort and on the type of database (A or B) which
provided the time/activity data. The parameters of these Monte Carlo models were
determined from an analysis of EVR data obtained from two diary studies conducted
by J. Hackney and associates in Los Angeles. These studies are referred to as
"Los Angeles - elementary school" and "Los Angeles - high school" in Table 2.
A total of 39 subjects participated in the two Los Angeles studies. Because of
the small sample size, the resulting EVR database may not accurately represent the
variability of EVR across the population. In addition, the database may not provide
sufficient data to adequately characterize age-specific differences in EVR. For
example, none of the Los Angeles subjects was below the age of 10 or above the
age of 17. The two demographic groups defined for the analysis (preteens and
teenagers) included children from age 6 to age 18. Consequently, the Monte Carlo
models developed from the Los Angeles EVR data may not adequately characterize
the younger children (6 to 9 years of age) in the preteens group and the older
children (18 years of age) in the teenagers group.
As discussed in Subsection 2.4.3, EVR values were not permitted to exceed
an upper bound determined by the EVR limiting algorithm for the specified
demographic group and event duration. For preteens, this bound was set equal to
the maximum EVR value attainable by boys aged 11 who exercise regularly and
who are motivated to reach a high ventilation rate. Note that this bound is likely to
be too high for other members of the preteens demographic group who differ with
respect to age, gender, exercise regime, and motivation. For similar reasons, the
EVR bounds set for teenagers are too high for many members of that demographic
group. In general, the EVR limiting algorithm will tend to permit more high EVR
values to occur in the pNEM/O3 simulation than would occur in the actual outdoor
children population. This potential bias may be corrected in future versions of
pNEM/O3 by distinguishing outdoor children cohorts by gender, age, and physical
conditioning. The parameters of the EVR limiting algorithm would be varied
according to these factors to yield a reasonable upper EVR limit for each cohort.
197
-------
The algorithm used to estimate EVR requires that each exposure event be
assigned to one of the four breathing rate categories. These assignments were
readily available in the time/activity data obtained from the Cincinnati and Los
Angeles studies, as the subjects of these studies entered this information directly
into their diaries. Information on breathing rate category was not provided by the
diaries used in the remaining time/activity studies. Consequently, the Monte Carlo
procedure described in Subsection 6.2 was used to assign breathing rate categories
to the time/activity data obtained from these studies. The Monte Carlo procedure
was based on the untested assumption that the probabilistic relationships between
activity type (e.g., yard work) and breathing rate category calculated from the
Cincinnati time/activity data could be applied to time/activity data from other studies.
8.3 The Air Quality Adjustment Procedures
Section 5 presents a summary of the procedures used to adjust baseline
ozone monitoring data to simulate conditions expected when a study area just
attains a specified NAAQS. These procedures assume that 1) the Weibull
distribution provides a good fit to most ozone data, and 2) the parameters of the
Weibull distribution fitting data from a particular monitoring site will change over time
in a predictable fashion. The adjustment procedures include equations for predicting
the values of the Weibull parameters under future attainment conditions.
The prediction equations were developed through a statistical analysis of
ozone data obtained from selected monitoring sites which have experienced
moderate reductions in ozone levels during the 1980's. It should be noted that none
of the selected monitoring sites reported ozone reductions of the magnitude required
to bring Los Angeles into compliance with any one of the NAAQS under evaluation.
For this reason, the prediction equations may not produce accurate estimates for the
Weibull distribution parameters when applied to Los Angeles ozone data.
Researchers have recently performed a series of tests to evaluate the air
quality adjustment procedures with respect to moderate reductions in ozone levels.
"in a technical letter, Johnson51 describes the general test procedure and its
198
-------
application to six pNEM/OS study areas (Chicago, Washington, Houston, Los
Angeles, New York, and Philadelphia). Analysts first selected a year representing
relatively low ozone concentrations for each city. The air quality adjustment
procedure for each of the three NAAQS formulations (1H1EX, 8H1EX, and 8H5EX)
was applied to the baseline ozone data for the study area with the goal of simulating
the ozone levels observed during the "low ozone" year. The resulting "estimated"
ozone concentrations for the low ozone year were compared with the actual ozone
levels reported for the low ozone year. The comparisons were performed using
selected percentiles of the cumulative ozone distributions (estimated and observed)
associated with each fixed-site monitor.
The test results suggest that the air quality adjustment procedures perform
adequately in the upper-tail region (90th percentile and above) of the ozone
distribution, the region that determines the exposures of most concern in pNEM/03
analyses. The results also show that the air quality adjustment procedures may
significantly over-estimate ozone concentrations in the lower portions of the
distribution. This problem is probably the result of using a somewhat "stiff1 two-
parameter distribution (the Weibull) to characterize one-hour ozone data.
Researchers may achieve better results by using a more flexible three-parameter
distribution, although this approach would likely require a more complicated air
quality adjustment procedure.
The air quality adjustment procedure is based on an assumption that the
attainment status of a particular city can be determined by a single year of
monitoring data. For example, the current status of Philadelphia is determined by
ozone monitoring data for 1991. This single year of monitoring data is then adjusted
to exactly meet a specified NAAQS. It should be noted that the pNEM/O3 approach
to determining attainment status differs somewhat from the actual method used by
EPA to determine attainment status. EPA typically examines three recent years of
monitoring data for a particular city and calculates a multi-year air quality indicator
(e.g., the fourth highest daily maximum one-hour ozone concentration for the three-
year period). The air quality indicator determined by this method is likely to differ
199
-------
from the air quality indicator determined in a pNEM/03 analysis from a single year
of data. As the direction of the difference is random, the degree of adjustment
applied to a city by pNEM/03 may be greater than or less than the adjustment
required to bring the city into compliance based on three years of data.
8.4 Estimation of Cohort Populations
Subsection 6.3 of this report describes the procedure used to estimate cohort
populations. An outdoor children cohort is defined by demographic group, home
district, and air conditioning status. For the outdoor children analysis, only two
demographic groups were defined: preteen (children 6 to 13) and teenager
(children 14 to 18). In addition, two major assumptions were employed in order to
estimate the population of each cohort. First, an outdoor child was defined as a
child who spent a specified amount of time outdoors, dependent on season and
weekend or weekday designation. The time criteria were determined somewhat
subjectively in an effort to include a sufficient number of person-days of diary data to
adequately represent the variability of activities among children, while at the same
time insuring that these criteria were rigorous enough to select only data which
represented children who spent noticeably more time outdoors than the "average"
child. Analysts evaluated several alternative time criteria before selecting the
specific criteria employed in the model (see subsection 6.1).
The second major assumption employed in the estimation of cohort
populations is the assumption that the ratio of outdoor children to all children is
constant across all cohorts belonging to a certain demographic (age) group,
regardless of study area. In actuality, it would be expected that this ratio would vary
by geographic region due to climate differences, by home district (whether rural,
suburban, or urban), by finer age demarcations, and perhaps even by gender. No
attempt was made to account for these factors, as applicable research and census
data do not currently exist.
200
-------
8.5 The Mass Balance Model
The pNEM/OS methodology uses the mass balance model described in
Section 3 to estimate ozone concentrations for the following enclosure categories:
Residential buildings - windows closed
Residential buildings - windows open
Nonresidential buildings
Vehicles.
The mass balance model provides hourly average ozone concentrations for each
enclosure category as a function of outdoor ozone concentration, AER, and ozone
decay rate.
In the application of pNEM/O3 to outdoor children, the outdoor ozone
concentration required by the mass balance model was always derived from fixed-
site monitoring data. These data were representative of local conditions and were
considered to be relatively reliable.
The AER values for residential buildings with closed windows were obtained
from a lognormal distribution fit to AER data from 312 residences. These data were
considered to be generally representative of housing in urban areas in the U.S.
No comparable databases were identified which were considered
representative of residences with open windows. Consequently, analysts
represented this enclosure category with a point estimate developed by Hayes45.
Analysts are uncertain as to the accuracy and general applicability of this estimate.
The AER values for nonresidential buildings were obtained from a lognormal
distribution fit to AER data from 40 buildings provided by Turk et al.44 This sample
may be too small to adequately characterize nonresidential buildings in the U.S. It
should also be noted that the Turk data are likely to represent only buildings with
closed windows. Consequently, the lognormal distribution derived from the Turk
data is likely to under-estimate the ozone exposures of people who frequently
occupy nonresidential buildings with open windows.
201
-------
A point estimate of 36 air changes per hour was used for the AER of
vehicles. This value was obtained from Hayes47 based on his analysis of data
reported by Peterson and Sabersky42 for a single vehicle. The use of a point
estimate is considered unrealistic as it does not account for varying ventilation
conditions within a particular motor vehicle or the variability in AER from vehicle to
vehicle.
Analysts also used a point estimate for the ozone decay rate of vehicles.
This value was based on data from a single automobile and may be biased.
Ozone decay rates for residential and nonresidential buildings were sampled
from a normal distribution. This distribution was based on decay rate data for a
relatively small number of buildings assembled by Weschler et ai.39 These data may
not adequately represent the variability of ozone decay rates among urban buildings
in the U.S.
8.6 Estimation of Ozone Exposures for Special Scenario Associated With
Attainment of 8H5EX-80 Standard
Section 7 presents the results of a series of exposure assessments using
pNEM/O3 in which the ozone levels within a specified study area have been
adjusted to meet a particular formulation of the ozone NAAQS. One of the
standards under review (designated 8H5EX-80) states that the expected
exceedance rate for daily maximum 8-hour ozone concentrations above 80 ppb shall
not be more than five. To evaluate this standard, ITAQS adjusted the ozone
monitoring data representing each study area using the AQAP described in Section
5. As a result of this procedure, the ozone data reported by each monitor was
adjusted so that the sixth highest daily maximum 8-hour concentration equaled a
specified air quality indicator (AQI). The sixth highest value of the historical "high
ozone" monitor was adjusted to equal 80 ppb.
This adjustment procedure is intended to limit the average exceedance rate of
the high ozone monitor to five exceedances of 80 ppb per year, based on a single
year of monitoring data. EPA has recently begun to evaluate an alternative form of
this standard which limits the average value of the fifth highest daily maximum 8-
202
-------
hour concentration to 80 ppb (here designated 8H5AVG-80). Under this standard,
there is no explicit limit to the number of exceedances that can occur in a given
year. However, a recent analysis by EPA found that very few ozone monitors report
more than 10 exceedances during a single year in an area that meets the 8H5AVG-
80 standard over a three-year period. As a result of this analysis, EPA directed
ITAQS to develop a procedure for adjusting the monitoring data in an area to
simulate conditions in which 10 exceedances occur at the historical high-ozone
monitor. These data were then be used in a pNEM/O3 analysis to estimate the
ozone exposures that could occur under these conditions. Subsection 8.6.1 briefly
describes the AQAP developed by ITAQS. Subsection 8.6.2 provides exposure
estimates for seven study areas.
8.6.1 The Air Quality Adjustment Procedure
The AQAP for the 10 exceedance scenario is similar to that used for adjusting
ozone data to simulate attainment of an 8H5EX standard. In essence, the data are
adjusted to meet an 8H10EX-80 standard, i.e., the expected number of daily
maximum eight-hour ozone concentrations exceeding 80 ppb shall not exceed ten.
The procedure is outlined in Table 1 of the letter in Appendix F. Note that
supplementary material concerning Step 6 of the procedure can be found in Section
5.3 of this report.
Section 5.4 of this report describes the application of an AQAP for the
8H5EX-80 standard to Philadelphia. The new procedure described in this letter is
essentially identical to the procedure in Section 5.4 when one makes the following
substitutions throughout the discussion: substitute 11th highest value for sixth
highest value and substitute RATIOS for RATIO2. Table 2, in Appendix F, lists
values of RATIOS by study area.
The adjustment procedure was applied to the ozone monitoring data which
have been used in previous pNEM/O3 analyses of seven study areas: Chicago,
203
-------
Houston, Los Angeles, New York, Philadelphia, St. Louis, and Washington, D.C.
The two remaining pNEM/03 study areas (Denver and Miami) were omitted from the
analysis because the ozone levels in these cities were relatively low with respect to
the levels permitted by the 8H5AVG80 standard.
8.6.2 Exposure Estimates for Selected Study Areas
The pNEM/03 model incorporates a number of stochastic (random) elements
which directly affect the exposure estimates produced by the model. Consequently,
exposure estimates are likely to vary from run to run. Consistent with earlier
analyses, ITAQS ran the model 10 times for each of the seven study areas. Tables
3 through 10, in Appendix F, provide means and ranges for selected exposure
indicators based on these runs. In each case, the exposure estimates apply to the
population group previously designated as "outdoor children" and use the adjusted
ozone data described above. The exposure indicators are defined in Sections 7.2
and 7.4 of this report.
In Tables 3 through 10, .of Appendix F, the attainment scenario is described
in terms of a "8H10EX-80" scenario, as the ozone monitoring data were adjusted to
simulate attainment of this indicator. In using this designation, it is understood that
the scenario is actually intended to represent a special high-ozone situation that
could occur during a single year when a 8H5AVG-80 standard is attained over a
three-year period.
Appendix F provides a detailed discussion of the exposure estimates in
Tables 3 through 10. The overall pattern of results indicates that ozone exposures
expected under the 8H10EX-80 scenario always exceed those of the 8H5EX-80
scenario and almost always are less than those under the 8H5EX-90 scenario.
204
-------
REFERENCES
1. Richmond, H. M. and T. McCurdy, 1988, "Use of Exposure Analysis and Risk
Assessment in the Ozone NAAQS Review Process," Paper No. 88-121.3,
presented at the 81st Annual Meeting of APCA, Dallas, Texas.
2. Ott, W. R., 1982, "Concepts of Human Exposure to Air Pollution,"
Environment International. Vol. 7, pp. 179 - 196.
3. Duan, N., 1982, "Models for Human Exposure to Air Pollution," Environment
International. Vol. 8, pp. 305 - 309.
4. Biller, W. F., T. B. Feagans, T. R. Johnson, G. R. Duggan, R. A. Paul, T.
McCurdy, and H. C. Thomas, 1981, "A General Model for Estimating
Exposure Associated With Alternative NAAQS," Paper No. 81-18.4, presented
at the 74th Annual Meeting of the Air Pollution Control Association, Dallas,
Texas, June.
5. Paul, R. A. and T. McCurdy, 1986, "Estimation of Population Exposure to
Ozone," Paper No. 86-66.2, presented at the 79th Annual Meeting of the Air
Pollution Control Association, Dallas, Texas, June.
6. Johnson, T. R. and R. A. Paul, 1981, "The NAAQS Exposure Model (NEM)
Applied to Particulate Matter," prepared by PEDCo Environmental, Inc. for the
Office of Air Quality Planning and Standards, U. S. Environmental Protection
Agency, Research Triangle Park, North Carolina.
7. Johnson, T. R. and R.A. Paul, 1983, "The NAAQS Exposure Model (NEM)
Applied to Carbon Monoxide," EPA Report No. 450/5-83-003, prepared by
PEDCo Environmental, Inc. for the Office of Air Quality Planning and
Standards, U. S. Environmental Protection Agency.
8. Paul, R. A., T. R. Johnson, and T. McCurdy, 1988, "Advancements in
Estimating Urban Population Exposure," Paper No. 88-127.1, presented at the
81st Annual Meeting of the Air Pollution Control Association, Dallas, Texas,
June.
9. Pandian, M. D., 1987, "Evaluation of Existing Total Human Exposure Models,"
EPA-600/4-87-004, U. S. Environmental Protection Agency, Las Vegas,
Nevada.
205
-------
10. Ryan, P. B., 1991 "An Overview of Human Exposure Modeling," Toxicology
and Industrial Health. Vol. 1, No. 4, pp. 453-474.
11. McKee, D., H. Richmond, P. Johnson, and T. McCurdy, 1984, "Review of the
NAAQS for Carbon Monoxide: Reassessment of Scientific and Technical
Information," EPA-450/5-84-004, U. S. Environmental Protection Agency,
Office of Air Quality Planning and Standards, Research Triangle Park, North
Carolina.
12. McKee, D., P. Johnson, T. McCurdy, and H. Richmond, 1989, "Review of the
National Ambient Air Quality Standard for Ozone: Assessment of Scientifrc
and Technical Information," EPA-450/2-92-001, U. S. Environmental
Protection Agency, Research Triangle Park, North Carolina.
13. Johnson, T. R., R. A. Paul, J. E. Capel, and T. McCurdy, 1990, "Estimation of
Ozone Exposure in Houston Using a Probabilistic Version of NEM," Paper
No. 90-150.1, presented at the 83rd Annual Meeting of the Air and Waste
Management Association, Pittsburgh, Pennsylvania.
14. Johnson, T. R., J. E. Capel, E. Olaguer, and L. Wijnberg, 1992, "Estimation of
Carbon Monoxide Exposures and Associated Carboxyhemoglobin Levels in
Denver Residents Using a Probabilistic Version of NEM," prepared by IT Air
Quality Services for the Office of Air Quality Planning and Standards, U. S.
Environmental Protection Agency, Research Triangle Park, North Carolina.
15. Johnson, T., J. Capel, E. Olaguer, and L Wijnberg, 1993, "Estimation of
Ozone Exposures Experienced by Urban Residents Using a Probabilistic
Version of NEM," prepared by IT Air Quality Services for the Office of Air
Quality Planning and Standards, U.S. Environmental Protection Agency,
Research Triangle Park, North Carolina, February.
16. Johnson, T., J. Capel, and M. McCoy, 1993, "Estimation of Ozone Exposures
Experienced by Urban Residents Using a Probabilistic Version of NEM and
1990 Population Data," Draft Report, prepared by IT Air Quality Services for
the Office of Air Quality Planning and Standards, U.S. Environmental
Protection Agency, Research Triangle Park, North Carolina, September.
17. Johnson, T., J. Capel, M. McCoy, and J. Warnasch, 1994, "Estimation of
Ozone Exposures Experienced by Outdoor Workers in Nine Urban Areas
Using a Probabilistic Version of NEM", Draft Report, prepared by IT Air
Quality Services for the Office of Air Quality Planning and Standards, U.S.
Environmental Protection Agency, Research Triangle Park, North Carolina
July.
206
-------
18. Bureau of Census, 1990, "1990 Census of Population and Housing, Equal
Opportunity File," Washington, D.C.
19. Johnson, T. R., 1989, Letter to Tom McCurdy, Office of Air Quality Planning
and Standards, U.S. Environmental Protection Agency, Research Triangle
Park, North Carolina, June 19.
20. Johnson, T. R., 1987, "A Study of Human Activity Patterns in Cincinnati,
Ohio," prepared by PEI Associates, Inc. for Electric Power Research Institute,
Palo Alto, available from Ted Johnson, IT Corporation, 3710 University Drive,
Durham, North Carolina 27707.
21. Johnson, T. R., 1989, Letter to Tom McCurdy, Office of Air Quality Planning
and Standards, U.S. Environmental Protection Agency, Research Triangle
Park, North Carolina, July 28.
22. Stock, T. H., D. J. Kotchman, C. F. Contant, et al., 1985, "The Estimation of
Personal Exposures to Air Pollutants for a Community-Based Study of Health
Effects in Asthmatics - Design and Results of Air Monitoring," Journal of the
Air Pollution Control Association, Vol. 35, p. 1266.
23. Weschler, C. J., H. C. Shields, and D. V. Naik, 1989, "Indoor Ozone
Exposures," Journal of the .Air Pollution Control Association. Vol. 39, p. 1562.
24. Wiley, J. A. et al., 1991, "Study of Children's Activity Patterns," Research
Division, California Air Resources Board, Sacramento, California, September.
25. Wiley, J. A. et al., 1991, "Activity Patterns of California Residents," Research
Division, California Air Resources Board, Sacramento, California, May.
26. Johnson, T. R., 1984, "A Study of Personal Exposure to Carbon Monoxide in
Denver, Colorado," EPA-600/54-84-014, U.S. Environmental Protection
Agency, Research Triangle Park, North Carolina, March.
27. Spier, C. E. et al., 1992, "Activity Patterns in Elementary and High School
Students Exposed to Oxidant Pollution," Journal of Exposure Analysis and
Environmental Epidemiology, Vol. 2, pp. 277-293.
28. Linn, W. S., D. Shamoo, and J. Hackney, 1992, "Documentation of Activity
Patterns in High-Risk Groups Exposed to Ozone in the Los Angeles Area,"
Tropospheric Ozone in the Environment II. editor, R. Berglund, Air and Waste
Management Association, Pittsburgh, PA.
207
-------
29. Goldstein, B. D. et al., 1992, "Valdez Air Health Study," Alyeska Pipeline
Service Company, Anchorage, Alaska, June.
30. Hartwell, T. D. et al., 1984, "Study of Carbon Monoxide Exposure of
Residents of Washington, D. C., and Denver, Colorado," EPA-600/54-84-031,
U.S. Environmental Protection Agency, Research Triangle Park, North
Carolina, March.
31. Rhodes, C. E. and D. M. Holland, 1981, "Variations of NO, N02, and 03
Concentrations Downwind of a Los Angeles Freeway," Atmospheric
Environment. Vol. 15, p. 243.
32. Johnson, T. R. and L. Wijnberg, 1981, "Time Series Analysis of Hourly
Average Air Quality Data," Paper No. 81-33.5, presented at the 74th Annual
Meeting of the Air Pollution Control Association, Philadelphia, Pennsylvania.
33. McDonnell, W. F., D. H. Horstman, and M. J. Hazuch, 1983, "Pulmonary
Effects of Ozone Exposure During Exercise: Dose-Response
Characteristics," Journal of Applied Physiology. Vol. 54, p. 1345.
34. T. Johnson and M. McCoy, 1994, "A Monte Carlo Approach to Generating
Equivalent Ventilation Rates in Population Exposure Assessments," available
from Dr. Will Ollison, American Petroleum Institute, 1220 L Street, N.W.,
Washington, D.C. 20005.
35. Johnson, T. R. and W. C. Adams, 1994, "An Algorithm for Determining
Maximum Sustainable Ventilation Rate According to Gender, Age, and
Exercise Duration," available from Ted Johnson, IT Corporation, 3710
University Drive, Durham, North Carolina 27707.
36. Erb, B. D., 1981, "Applying Work Physiology to Occupational Medicine,"
Occupational Health Safety. Vol. 50, pp. 20-24.
37. Astrand, P. 0. and K. Rodahl, 1977, Textbook of Work Physiology. 2nd ed.
McGraw-Hill, New York, New York.
38. Nagda, N. L., H. E. Rector, and M. D. Koontz, 1987, Guidelines for Monitoring
Indoor Air Quality. Hemisphere Publishing Corporation, Washington, D. C.
39. Weschler, C. J., H. C. Shields, and D. V. Nike, 1992, "Indoor Ozone: Recent
Findings," Tropospheric Ozone in the Environment II. editor, R. Burglund, Air
and Waste Management Association, Pittsburgh, PA, pp. 681 - 700.
208
-------
40. Nazaroff, W.W., Cass, G.R., 1986, "Mathematical Modeling of Chemically
Reactive Pollutants in Indoor Air," Environmental Science and Technology,
Vol. 20, pp. 880-885.
41. Hayes, S.R., 1989, "Estimating the Effect of Being Indoors on Total Personal
Exposure to Outdoor Air Pollution," Journal of the Air Pollution Control
Association. Vol. 39, No. 11, pp. 1453-1461.
42. Peterson, G.A. and R.H. Sabersky, 1975, "Measurements of Pollutants Inside
an Automobile," Journal of the Air Pollution Control Association. Vol. 25, No.
10, pp. 1028-1032, October.
43. Grimsrud, D. T., M. H. Sherman, and R. C. Sondregger, 1982, "Calculating
Infiltration: Implications for a Construction Quality Standard," Proceedings of
the ASHRAE/DDE Conference: Thermal Performance of the Exterior
Envelopes of Buildings II. Las Vegas, Nevada.
44. Turk, B. H., D. T. Grimsrud, J. T. Brown, K. L Geisling-Sobotka, J. Harrison,
and R. J. Prill, 1989, "Commercial Building Ventilation Rates and Particle
Concentrations," ASHRAE Transactions. Vol. 95, Part 1.
45. Hayes, S. R. and G. W. Lundberg, 1985, "Further Improvement and
Sensitivity Analysis of an Ozone Population Exposure Model," Report No.
SYSAPP-85/061, Systems Applications, Inc., San Rafael,, California.
46. Moschandreas, D. J., J. Zabransky, and D. J. Peltan, 1981, "Comparison of
Indoor and Outdoor Air Quality," Report No. EA-1733, Electric Power
Research Institute.
47. Hayes, S. R., 1991, "Use of an Indoor Air Quality Model (IAQM) to Estimate
Indoor Ozone Levels." Journal of the Air and Waste Management Association.
Vol. 41, pp.161-170.
48. T. Johnson, M. McCoy, J. Capel, L. Wijnberg, and W. Ollison, 1992, "A
Comparison of Ten Time/Activity Databases: Effects of Geographic Location,
Temperature, Demographic Group, and Diary Recall Method," Proceedings
of the 1992 International Conference and Course on Tropospheric Ozone. Air
and Waste Management Association, Pittsburgh, Pennsylvania.
49. Shamoo, D. A., et al., "Activity Patterns in a Panel of Outdoor Workers
Exposed to Oxidant Pollution," Journal of Exposure Analysis and
Environmental Epidemiology. Vol. 1, pp. 423-438.
209
-------
50. Linn, W. S., C. E. Spier, and J. D. Hackney, 1993, "Activity Patterns in
Ozone-Exposed Construction Workers," Journal of Occupational Medicine and
Toxicology. Vol. 2, pp. 1-14.
51. Johnson, T. R., 1995, Letter to Harvey Richmond, Office of Air Quality
Planning and Standards, U.S. Environmental Protection Agency, Research
Triangle Park, North Carolina, August 4.
210
-------
APPENDIX A
TEN TIME/ACTIVITY DATABASES GENERALLY APPLICABLE
TO AIR POLLUTION EXPOSURE ASSESSMENTS
A-1
-------
In 1993, Johnson et al.48 conducted a literature review to identify all time/activity
databases which would be appropriate for use in pNEM exposure assessments. The
survey identified ten databases with adequate data characteristics. Eight of the
databases relate to five individual urban areas: Cincinnati, Ohio; Denver, Colorado;
Los Angeles, California; Valdez, Alaska; and Washington, D.C. The remaining two
databases relate to the entire State of California. In the discussion that follows, each
database will be identified according to the associated geographical area. When a
geographical area is associated with more than one database, each database will be
further distinguished according to the sampled population (e.g., Los Angeles - outdoor
workers).
California
The California Air Resources Board conducted two state-wide time/activity
studies24'25 to provide a large pool of activity pattern data suitable for use in estimating
environmental exposures. The first study, referred to hereafter as the "California - 12
and over" study, was conducted between October 1987 and July 1988. During the
study, interviewers collected one day of activity data from each of 1762 California
residents over the age of 11. The second study ("California - 11 and under") was
conducted from April 1989 through February 1990. The study gathered one day of
activity data from each of 1200 children ages 11 and under. Both studies employed
retrospective telephone interviews to obtain a record of each subject's activities during
the preceding day.
Cincinnati
The Cincinnati Activity Diary Study20 was conducted by the Electric Power
Research Institute during March and August 1985 to provide an extensive database
for evaluating human exposure to air pollution. The sampled population included all
residents of a three-county area in and around Cincinnati, Ohio. Each subject
recorded his or her activities over a three-day period in a real-time diary and
completed a detailed background questionnaire. The 487 March subjects provided
A-2
-------
1401 subject-days of diary data; the 486 August subjects provided 1399 subject-
days. Activity diary data collected during the Cincinnati study have been used by
the U.S. Environmental Protection Agency (EPA) in various applications of the
pNEM/CO14 and pNEM/O316 exposure models.
Denver and Washington
The U.S. Environmental Protection Agency conducted studies of adults (18 to
70 years) in Denver, Colorado, and Washington, D.C., during the winter of 1982 -
1983 for the purpose of collecting representative data on personal exposure to
carbon monoxide. In the Denver study26, each of 454 subjects carried a personal
exposure monitor (PEM) and a real-time time/activity diary for two 24-hour periods.
Each subject also provided a breath CO sample at the end of each monitored period
and completed a detailed background questionnaire. The Washington study30
employed a similar protocol to obtain data for a single 24-hour period from each of
908 subjects. Activity diary data from these two studies have been used in
conjunction with data from the Cincinnati study in applications of EPA's pNEM/CO
exposure model14.
Los Angeles
Between 1989 and 1991, a research team headed by Dr. Jack Hackney
conducted four activity diary studies in Los Angeles with funding provided by the
American Petroleum Institute. The first of these, the "Los Angeles - outdoor worker
study", was conducted during the summer of 1989.49 Each of 20 outdoor workers
wore a heart rate monitor for a three-day period during which the worker recorded
his or her activities in a real-time diary identical to that used in the Cincinnati study.
In October of 1989, the outdoor worker study was expanded to include 20
healthy elementary school children. During this phase of the Los Angeles study,
referred to here as the "Los Angeles - elementary school" study, each child wore a
heart-rate monitor for two or three days and recorded his or her activities in the real-
time Cincinnati diary. Approximately 58 subject-days of activity data were collected.27
A-3
-------
A third phase of the Los Angeles study (the "Los Angeles - high school"
study) was conducted during September and October 1990.27 During this phase, 66
subject-days of real-time activity data were collected from 19 students between the
ages 13 and 17 using the Cincinnati diary.
The Hackney research team conducted a fourth study in Los Angeles
between July and November 1991. Each of 19 construction workers between the
ages of 23 and 42 wore a heart rate monitor during a typical work day. The
Cincinnati diary was used to record each subject's activities during this period. The
study protocol differed from the other Los Angeles studies in that each diary was
completed by a trained observer rather than by the subject. The observer monitored
each subject's activities visually and by two-way radio. This approach produced
unusually detailed diaries of high accuracy.50
Valdez
The Valdez Air Health Study29 was undertaken by the Alyeska Pipeline
Service Company in response to concerns expressed by the citizens of Valdez,
Alaska, regarding their potential exposure to certain volatile organic compounds
(VOCs). Between November 1990 and October 1991, 405 subjects aged 10 to 72
years were interviewed and requested to report their daily activities for an earlier 24-
hour period. In addition to the activity data, researchers collected extensive data on
personal exposures to VOC's, ambient air quality, and meteorological conditions.
A-4
-------
APPENDIX B
MONTE CARLO MODELS FOR GENERATING EVENT
EVR VALUES
Database Types
In the pNEM/03 methodology, each cohort is represented by an exposure
event sequence. The sequence is constructed from data obtained from studies in
which subjects recorded their activities over 24-hour periods (person-days) in
diaries. Table B-1 lists the seven studies which provided diary data for the
application of pNEM/O3 to outdoor children. Appendix A provides brief descriptions
of these seven studies and three other studies which have been used in other
pNEM applications.
Three of the studies listed in Table B-1 (Cincinnati, Los Angeles - elementary
students, and Los Angeles - high school students) employed a diary which used the
page format shown in Figure B-1. This page format (referred to as the "Cincinnati"
format) provided data which could be used to directly classify each exposure event
with respect to five microenvironments and four breathing rates:
Microenvironments Breathing Rates
Indoors - residence Sleeping
Indoors - other Slow
Outdoors - near road Medium
Outdoors - other Fast
In vehicle
The databases obtained from the three studies which used this format were
designated Type A1 databases. Time/activity data from Type 1 databases generally
can be used "as is" in pNEM/O3 assessments.
One of the studies listed in Table B-2 (Washington) employed the diary page
format shown in Figure B-2. This format supports the use of the five
B-1
-------
Table B-1. Characteristics of studies associated with the seven time/activity databases.
Database name
California - 11
and under
California - 12
and over
Cincinnati
Los Angeles -
elem. school
Los Angeles -
high school
Valdez
Washington
Database
type
B
B
A1
A1
A1
B
A2
Characteristics
of subjects
Children ages 1 to 11
Ages 12 to 94
Ages 0 to 86
Elementary school
students, 10 to 12 years
High school students,
13 to 17 years
Ages 10 to 72
Ages 18 to 70
Number of
subject-
days
1200
1762
2800
58
66
405
705
Study
calendar
periods
April 1989 -
Feb. 1990
Oct. 1987 -
July 1988
March and
August 1985
Oct. 1989
Sept. and
Oct. 1990
Nov. 1990 -
Oct. 1991
Nov. 1982 -
Feb. 1983
Diary type
Retrospective
Retrospective
Real-time*
Real-time*
Real-time8
Retrospective
Real-time
Diary time
period
Midnight to
midnight
Midnight to
midnight
Midnight to
midnight
Midnight to
midnight
Midnight to
midnight
Varying
24-h period
7 p.m. to
7 p.m.
(nominal)
Breathing
rates
reported?
No
No
Yes
Yes
Yes
No
No
Ni
9Study employed the Cincinnati diary format.
-------
TIME
AM
PM
A. ACTIVITY (please specify)
B. LOCATION
In transit, car 01
In transit, other vehicle . . 02
Specify
Indoors, your residence ... 03
Indoors, other residence. . . 04
Indoors, office 05
Indoors, manufacturing
facility 06
Indoors, school 07
Indoors, store 08
Indoors, other 09
Specify .
Outdoors, within 10 yards of
road or street 10
Outdoors, other 11
Specify
Uncertain 12
*Enter MIDN for midnight and NOON for noon. Otherwise enter four-digit
time (e.g., 0930 for 9:30 and 1217 for 12:17) and check a.m. or p.m.
C. BREATHING RATE
Slow (e.g., sitting) 13
Medium (e.g., brisk walk). . . 14
Fast (e.g., running) 15
Breathing problem 16
Specify
D. SMOKING
I am smoking 17
Others are smoking 18
No one is smoking 19
E. ONLY IF INDOORS
(1) Fireplace in use?
Yes 20
No 21
(2) Woodstove in use?
Yes 22
No 23
(3) Windows open?
Yes 24
No 25
Uncertain 26
Figure B-1. Blank page from Cincinnati activity diary.
B-3
-------
Table B-2. Database types.
Type AT: Time/activity data acquired using the "Cincinnati" diary. Type
A1 data support the use of four breathing rate categories
(sleeping, slow, medium, and fast) and five
microenvironments (indoors - residence, indoors - other,
outdoors - near road, outdoors - other, and in vehicle).
Applicable studies: Cincinnati, Los Angeles - elementary
school, and Los Angeles - high school.
Type A2: Time/activity data acquired using the "Denver/Washington"
diary. Type A2 data support the use of the five Type 1
microenvironments. Breathing rate data (not available in the
reported data) were developed through a Monte Carlo
algorithm.
Applicable study: Washington.
Type B: Time/activity data acquired using other diary formats. Type B
data support the use of four microenvironments: indoors -
residence, indoors - other, outdoors, and in vehicle.
Breathing rate data (not available in the reported data) were
developed through a Monte Carlo algorithm.
Applicable studies: California - 11 and under, California - 12
and over, and Valdez.
B-4
-------
TIME FROM MONITOR
A. ACTIVITY
B. LOCATION
In transit 1
Indoors, residence 2
Indoors, office 3
Indoors, store 4
Indoors, restaurant 5
Other indoor location 6
Specify:
Outdoors, within 10 yards of road
or street 7
Other outdoor location
Specify:
. 8
Uncertain 9
C. ADDRESS (if not in transit)
D. ONLY IF IN TRANSIT
(1) Start address
(2) End address
(3) Mode of travel:
Walking 1
Car 2
Bus 3
Truck 4
Train/subway 5
Other 6
Specify:
ONLY IF INDOORS
(1) Garage attached to building?
Yes 1
No 2
Uncertain 3
(2) Gas stove in use?
Yes 1
No 2
Uncertain 3
ALL LOCATIONS
Smokers present?
Yes 1
No 2
Uncertain 3
Figure B-2. Blank-page from Washington activity diary.
B-5
-------
microenvironments listed above but provides no data on breathing rate. Because
the pNEM/O3 methodology requires that each exposure event be characterized by
breathing rate, ITAQS analysts developed a Monte Carlo technique to estimate
breathing rate indirectly from other information provided by the diary. The
technique, which is described in Subsection 6.2, randomly assigns a breathing rate
to each exposure event based on selection probabilities which vary with activity
class, microenvironment, time of day, and duration. The selection probabilities are
based on a statistical analysis of the Cincinnati time/activity database. When this
technique was applied to the Washington database, the technique produced an
"augmented" database which was consistent in format with the Type A1 databases
described above. The augmented database for Washington was referred to as a
Type A2 database.
The diaries employed by the remaining three studies in Table B-1 (California -
children, California - adults, and Valdez) do not permit analysts to identify outdoor
locations near roadways. Consequently, only four microenvironments were used to
categorize data from these studies: indoors - residence, indoors - other, outdoors,
and in vehicle. The California and Valdez diaries also omitted breathing rate
information; consequently, the Monte Carlo technique described above was used to
randomly assign a breathing rate to each exposure event. The resulting augmented
databases were referred to as Type B databases.
Table B-2 provides a brief summary of the characteristics of each database
type. In the discussion that follows, Types A1 and A2 are discussed jointly as Type
A.
GENERAL PROCEDURE FOR DEVELOPMENT OF MONTE CARLO MODELS
ITAQS analysts developed a special EVR-generator module for the version of
pNEM/O3 applicable to outdoor children. The module used one of four Monte Carlo
models to generate an EVR value for each exposure event associated with a given
cohort. The applied model varied from event-to-event according to (1) the
B-6
-------
demographic group of the cohort (preteens or teenagers) and (2) the type of
database (A or B) from which the associated diary data were obtained.
The Monte Carlo models were based on the results of statistical analyses
performed on EVR data obtained from the two Los Angeles studies listed in Table
B-1: elementary school students and high school students. Models applicable to
the preteens demographic group were based on analyses of data from the
elementary school study; models applicable to teenagers were based on analyses
of data from the high school study. To permit the use of all seven diary databases
listed in Table B-1, analysts developed two Monte Carlo models for each
demographic group -- one applicable to Type A databases and one applicable to
Type B databases.
Each Monte Carlo model predicted EVR as a function of six or more predictor
variables which constituted a "predictor set." Each predictor set was developed by
performing stepwise linear regression analyses on one of the two Los Angeles
databases. Each of the Los Angeles databases consisted of a collection of "person-
days," each person-day containing the data obtained from one subject during one
24-hour period. The data for each person-day were organized into a sequence of
exposure events. Each exposure event was characterized by an average EVR
value and by a value for each of the 24 variables listed in Table B-3 (as applicable).
Exploratory statistical analyses by Johnson and McCoy34 identified these variables
as good candidate variables for the regression analyses.
Two series of regression analyses were performed on the Los Angeles
databases. The first series treated each Los Angeles database as being a Type A
database with five microenvironments. Each of these regression analyses was
performed using the Type A candidate variables listed in Table B-3. The second
series treated the Los Angles databases as being Type B databases with four
microenvironments.
B-7
-------
Table B-3. Candidate variables used in stepwise linear regression analyses.
Variable
Explanation
Candidate
variable group
B
CD
00
LGM
SLEEP
SLOW
MEDIUM
FAST
DUR1
DUR2
DUR3
DUR4
DUR5
DUR6
DUR7
INDOOR
OUTDOOR
OUTOTHER
VEH
MALE
WEEKDAY
HIGHTOWK
DAYACT
TRAVEL
HIGHACT
LOWACT
WORK
Natural logarithm of geometric mean of event EVR values for individual subject
SLEEP=1 if breathing rate = sleeping, 0 otherwise
SLOW=1 if breathing rate = slow, 0 otherwise
MEDIUM=1 if breathing rate = medium, 0 otherwise
FAST=1 if breathing rate = fast, 0 otherwise
DUR1=1 if duration < 5 minutes, 0 otherwise
DUR2=1 if 6 < duration < 10 minutes, 0 otherwise
DUR3=1 if 11 < duration < 20 minutes, 0 otherwise
DUR4=1 if 21 < duration < 30 minutes, 0 otherwise
DUR5=1 if 31 < duration < 45 minutes, 0 otherwise
DUR6=1 if 46 < duration < 60 minutes, 0 otherwise
DUR7=1 if duration > 60 minutes, 0 otherwise
INDOOR = 1 if event occurs in an indoor microenvironment, 0 otherwise
OUTDOOR = 1 if event occurs in an outdoor microenvironment, 0 otherwise
OUTOTHER = 1 if event occurs in the outdoors - other microenvironment, 0 otherwise
VEH = 1 if event occurs in a vehicle microenvironment, 0 otherwise
MALE = 1 if subject is male, 0 otherwise
WEEKDAY = 1 if event occurs on a weekday, 0 otherwise
HIGHTOWK = 1 if daily maximum temperature exceeds 79°F and event occurs in
outdoor microenvironment and activity code = 2 (work), 0 otherwise
DAYACT = 1 if event begins between 7:00 a.m. and 4:59 p.m., 0 otherwise
TRAVEL = 1 if activity code is 1 (travel), 0 otherwise
HIGHACT = 1 if activity code is 11, 27, 28, 29, 30, 31, or 33; 0 otherwise
LOWACT = 1 if activity code is 10, 12, 16, 23, 34, 35, 37, or 44; 0 otherwise
WORK = 1 if activity code = 2 (work), 0 otherwise
-------
These regression analyses were performed using the Type B variables listed
in Table B-3. With the exception of the continuous variable LGM, each of the
variables listed in Table B-3 is a binary "dummy" variable. A dummy variable equals
one when specified conditions are met and equals zero under all other conditions.
Among the variables listed in Table B-3 are variables which indicate breathing rate,
event duration, microenvironment, subject gender, time of day, day of the week, and
temperature. Several variables classify activities according to level of exertion, work
status (work/non-work), and travel status (travel/non-travel).
The continuous variable LGM is equal to the natural logarithm of the
geometric mean of all event EVR values associated with a subject. Analyses of
variance performed on the two Hackney/Linn data sets indicated that inter-subject
variability with respect to average EVR was a major source of variability in the event
EVR values.34 LGM is an indicator of average subject EVR which can be related
directly to In(EVR), the dependent variable defined for the regression analyses.
The results of each stepwise regression analysis were used to (1) identify
significant predictor variables and (2) estimate the coefficients of a regression
equation which included only the significant variables. The regression equation had
the general form
ln[EVR(i,j)] = b0 + (bOEVAR^ij)] + (b2)[VAR2(i,j)] + ... + (bJIVARJij)] + e(ij) (1)
where EVR(i.j) is the EVR value for event j associated with subject i; b0, b,, b2, ...,
bm are constants; VAR^ij), VAR2(i,j) VARm(ij) are the values of the predictor
variables for the event; and e(i,j) is the residual.
RESULTS OF STEPWISE LINEAR REGRESSION ANALYSES
Tables B-4 and B-5 present the results of the stepwise regression analyses
performed on the elementary and high school databases, respectively. As indicated
above, the dependent variable in each regression analysis was In(EVR), i.e., the
natural logarithm of the average EVR for the event. Each table lists the results for
two regression analyses -- one using the Type A variables and one using Type B
B-9
-------
variables. The results listed for each regression analysis include the variables
selected by the regression procedure as being significant predictors of In(EVR), the
regression coefficient associated with each variable, the p value associated with the
coefficient, and the cumulative R2 value that resulted when the variable was added
to the regression equation.
Table B-4 provides the results of the regression analyses performed on the
elementary school database. The regression analysis of Type A variable set
selected seven variables for the regression equation (LGM, OUTDOOR, FAST,
DAYACT, SLEEP, WEEKDAY, and HIGHACT). The cumulative R2 value for all
seven variables is 0.7083. This value indicates that the regression equation
explains 70.83 percent of the variation in In(EVR).
The regression analysis performed on the Type B variable set selected the
same seven variables: LGM, OUTDOOR, FAST, DAYACT, SLEEP, WEEKDAY,
and HIGHACT. Consequently, the regression equations associated with Types A
and B are identical.
Table B-5 presents regression results for the high school database. The
regression procedure selected 12 variables from the Type A variable set. The first
three variables to be selected were LGM, OUTOTHER, and HIGHACT. These three
variables had a cumulative R2 value of 0.3833. The cumulative R2 value for all 12
variables is 0.4596.
The regression procedure selected 12 variables from the Type B variable set.
The first three variables were LGM, OUTDOOR, and HIGHACT (cumulative R2 =
0.3848). The cumulative R2 for all 12 variables is 0.4547. A comparison of the
B-10
-------
Table B-4. Results of stepwise linear regression analyses performed on elementary
school data set.
Candidate
variable set
A and B
Selected
variable3
Constant
LGM
OUTDOOR
FAST
DAYACT
SLEEP
WEEKDAY
HIGHACT
Regression
coefficient
-0.08174
0.98606
0.12156
0.16111
0.07188
-0.17393
0.04674
0.05962
p value
0.1544
0.0000
0.0000
0.0000
0.0001
0.0021
0.0062
0.0159
Cumulative R2
0.0000
0.6600
0.6812
0.6939
0.7002
0.7037
0.7063
0.7083
3HIGHTOWK not applicable to this data set.
B-11
-------
Table B-5. Results of stepwise linear regression analyses performed on high
schools data set.
Candidate
variable set
A
B
Selected
variable3
Constant
LGM
OUTOTHER
HIGHACT
SLOW
DAYACT
DUR7
LOWACT
WEEKDAY
DUR5
FAST
DUR6
VEH
Constant
LGM
OUTDOOR
HIGHACT
SLOW
DAYACT
LOWACT
DUR7
WEEKDAY
FAST
DUR5
DUR6
INDOOR
Regression
coefficient
0.16385
0.91365
0.11198
0.15447
-0.07989
0.08175
-0.13637
-0.08749
0.05873
-0.09184
0.14685
-0.10629
-0.04650
0.10646
0.91721
0.11621
0.15820
-0.07636
0.08131
-0.08556
-0.13256
0.05898
0.16330
-0.09306
-0.10331
0.04298
p value
0.0158
0.0000
0.0000
0.0000
0.0000
0.0000
0.0000
0.0000
0.0000
0.0000
0.0001
0.0001
0.0255
0.1451
0.0000
0.0000
0.0000
0.0000
0.0000
0.0000
0.0000
0.0000
0.0000
0.0000
0.0000
0.0000
Cumulative R2
0.0000
0.3063
0.3505
0.3833
0.4038
0.4209
0.4325
0.4408
0.4462
0.4503
0.4545
0.4582
0.4596
0.0000
0.3063
0.3495
0.3848
0.4014
0.4171
0.4276
0.4353
0.4406
0.4454
0.4500
0.4536
0.4547
aHIGHTOWK not applicable to this data set.
B-12
-------
results presented in Tables B-4 and B-5 finds that six variables appear in both
tables: LGM, OUTDOOR, DAYACT, HIGHACT, WEEKDAY, and HIGH. LGM is
always the first variable selected. LGM contributed 0.660 to the cumulative R2 value
in Table B-4. In Table A-5, adding LGM increased the R2 value by 0.306.
These results suggest that variables associated with average subject EVR
(LGM), outdoor microenvironments (OUTDOOR), daytime activities (DAYACT), the
exertion level cf activities (HIGHACT), day of week (WEEKDAY), and breathing rate
(HIGH) are particularly useful in predicting event EVR.- Note that the duration
variables tended to be relatively insignificant predictors. Adding DUR6 or DUR7 to
the regression equation never increased the cumulative R2 value by more than
0.015.
THE DISTRIBUTION OF REGRESSION RESIDUALS
Each regression analysis produced a set of residual values, one for each
EVR value. Researchers performed a series of exploratory data analyses in which
they attempted to find patterns in the residuals which could be used to characterize
random effects in the Monte Carlo approach. Statistical analysis of the residuals
indicated that (1) the standard deviation of the residuals varied significantly from
subject to subject and (2) the distribution of the subject-specific standard deviations
was approximately lognormal.
Based on these findings, researchers assumed that the residual term in
Equation 1 could be represented by a normally distributed random variable with
mean equal to zero and standard deviation equal to SDRES. The value of SDRES
was assumed to vary with subject and to be lognormally distributed among subjects;
i.e., the natural logarithm of SDRES [LSDRES = In(SDRES)] is normally distributed
with mean = MU and standard deviation = SIGMA. The values of MU and SIGMA
were specific to the data set undergoing the regression analysis.
Consistent with these assumptions, analysts performed the following
statistical analysis of the residuals obtained from each regression analysis:
B-13
-------
1. Classify residuals by subject.
2. Calculate the standard deviation of residuals associated with each
subject (SDRES).
3. Calculate a value of LSDRES for each subject where LSDRES is the
natural logarithm of each SDRES value obtained in Step 2.
4. Calculate the mean (MU) and the standard deviation (SIGMA) of the
LSDRES values determined for all subjects.
Table B-6 lists the values of MU and SIGMA determined in Step 4.
THE DISTRIBUTION OF LGM VALUES
As indicated above, researchers found that the LGM variable was the single
best predictor of In(EVR) in each regression analysis. An analysis of the LGM
values associated with the subjects in each of the Los Angeles studies indicated
that the distribution of LGM values for each study was approximately normal. The
parameters of these normal distributions were estimated by the following procedure.
1. Classify the event EVR values by subject.
2. Calculate In(EVR) for each event.
3. Calculate the mean of the In(EVR) values associated with each subject
(LGM).
4. Calculate the arithmetic mean and arithmetic standard deviation of the
LGM values determined for all subjects in the database.
The arithmetic mean and standard deviation values determined in Step 4 are
listed in Table B-7.
B-14
-------
Table B-6. Distribution of LSDRES values.
Database
Elementary
school
High school
Number of
subjects
16
19
Regression model
producing residuals
A and B
A
B
Parameters of normal
distribution fit to subject
LSDRES values3
MU =
mean
-1.6068
-1.4662
-1.4586
SIGMA = standard
deviation
0.4450
0.2997
0.3054
Wilk-Shapirob statistic
for LSDRES
0.9540
0.9845
0.9834
Range
Minimum
-2.3066
-2.0472
-2.0503
Maximum
-0.7251
-0.7822
-0.7634
Ol
"LSDRES is the natural logarithm of the standard deviation of the regression residuals associated with one subject
bAn indicator of normality (1.0 = normal distribution).
-------
Table B-7. Distribution of LGM values.
Database
Elementary school
High school
Number of
subjects
16
19
Parameters of normal distri-
bution fit to subject LGM values8
MEANLGM =
mean
2.3629
2.1621
SDLGM =
standard deviation
0.4324
0.1890
Wilk-Shapirob statistic
for LGM
0.9630
0.9597
Range
Minimum
1.3713
1.8135
Maximum
3.0517
2.5603
CD
8LGM is the natural logarithm of the geometric mean of the event EVR values associated with one subject.
bAn indicator of normality (1.0 = normal distribution).
-------
GENERAL ALGORITHM FOR EXECUTING THE MONTE CARLO MODEL
The EVR-generator module contained four Monte Carlo models, one for each
combination of demographic group and database type. The module processed the
exposure event sequence of each cohort as a series of person-days. An EVR value
was generated for each event in the first person-day using the Monte Carlo model
which matched the demographic group of the cohort and the database type (A or B)
of the person-day. The module then generated an EVR value for each event in the
second person-day using the Monte Carlo model which matched the new conditions.
The process continued until an EVR was assigned to each exposure sequence.
Table B-8 presents the general algorithm incorporated into the EVR-generator
module. The algorithm begins by processing the first (or next) person-day in a
particular exposure event sequence. The algorithm checks the cohort for
demographic group and the source of the diary data for database type. Based on
this information, the algorithm identifies the applicable Monte Carlo model for the
person-day. Associated with each Monte Carlo model are values for the following
parameters:
MEANLGM: mean of the LGM values
SDLGM: standard deviation of LGM values
MU: mean of LSDRES values
SIGMA: standard deviation of LSDRES values
b0: constant
bm: coefficient of VARm
These values are held constant for each person-day i.
The algorithm determines a value of LGM(i) for person-day i by randomly
selecting a value from a normal distribution with mean = MEANLGM and standard
deviation = SDLGM (Table B-7). LGM(i) values are not permitted to fall outside the
range indicated in Table B-7.
The algorithm also determines a value of LSDRES(i) for person-day i. This
value is randomly selected from a normal distribution with mean = MU and standard
deviation = SIGMA (Table B-6); the value is not permitted to fall outside the range
B-17
-------
indicated in Table B-6. The value of LSDRES(i) is exponentiated to produce a
corresponding value of RESSIGMA(i).
The algorithm reads the data listings for each exposure event j associated
with person-day i to determine the values of the variables VAR^ VAR2, .... VARm.
The algorithm also determines a residual value [RES(ij)] for each event j by
randomly selecting a value from a normal distribution with mean = 0 and standard
deviation = RESSIGMA(i). The equation in Step 11 of Table B-8 is then used to
determine a value for LEVR(ij). This value is exponentiated to determine a value of
EVR for the event. The algorithm steps through each event associated with the first
person-day and then processes the next person-day. The process continues until
all person-days in the exposure event sequence have been completed.
Appendix C presents the results of initial efforts to test this algorithm.
B-18
-------
Table B-8. Algorithm used to generate event-specific values of equivalent
ventilation rate.
1. Go to first/next person-day i.
2. Determine Monte Carlo model applicable to person-day according to
demographic group of cohort and database type of diary data.
3. Model identity determines
MEANLGM: mean of LGM values
SDLGM: standard deviation of LGM values
MU: mean of LSDRES values
SIGMA: standard deviation of LSDRES values
b0: constant
bm: coefficient for variable VARm
Denote the value of bm for variable LGM as bv
4. Calculate LGM for person-day i:
LGM(i) = MEANLGM + (SDLGM)[Z1(i)]
Z1(i): randomly selected value from-unit normal distribution (normal
distribution with mean = 0 and standard deviation = 1).
5. If LGM(i) falls outside range indicated in Table B-7, discard value and go
to Step 4.
6. Calculate RESSIGMA for person-day i.
LSDRES(i) = MU + (SIGMA)[Z2(i)]
RESSIGMA(i) = Exp[LSDRES(i)]
Z2(i): randomly selected value from unit normal distribution.
7. If LSDRES(i) falls outside range indicated in Table B-6, discard value and
go to Step 6.
8. Go to first/next event associated with person-day i.
(continued) B-19
-------
Table B-8 (Continued)
9. Read values of variables VAR2, VAR3, .... VARm for event j of person-day i
from input data file.
10. Calculate residual value for event j of subject i.
RES(ij) = [RESSIGMA(i)][Z(i,j)]
Z(i,j): randomly selected value from unit normal distribution.
1 1 . Calculate LEVR for event j of person-day i:
LEVR(ij) = b0 + (b^LGMG)] + (b2)[VAR2(i,j)] + (b3)[VAR3(i,j)] +
(bJIVARJi j)] + RES(ij)
12. Calculate EVR for event j of person-day i:
EVR(i,j) = Exp[LEVR(i,j)]
13. Write EVR(ij) to output file.
14. If last event of person-day i, go to Step 1. If not, go to Step 8.
B-20
-------
APPENDIX C
TESTING OF MONTE CARLO MODELS
At the time of this report (October 1994), the two Los Angeles databases
(elementary school and high school) provided the only means of testing the
reasonableness the Monte Carlo approach described in Appendix B. These were
the only databases available which included high quality time/activity data together
with EVR values determined from heart rate measurements. This appendix
summarizes the results of initial efforts to test the Monte Carlo approach using these
two databases.
APPLICATION OF THE ALGORITHM TO THE HACKNEY/LINN DATABASES
Table B-8 in Appendix B presents an algorithm which can be used to
generate an EVR value for each event in a time/activity database, given that the
database is Type A or Type B. Both of the Los Angeles diary studies (elementary
school and high school) produced Type A databases. Consequently, the application
of the algorithm to these databases should provide an indication of model
performance with respect to Type A databases.
The algorithm was applied to the elementary school database in the following
manner. Researchers used the regression results listed in Table B-4 for Candidate
Variable Set A to determine the set of predictor variables, the coefficient of each
variable, and the constant. The selected predictor variables were LGM, OUTDOOR,
FAST, DAYACT, SLEEP, WEEKDAY, and HIGHACT. The constant was -0.082, the
coefficient for LGM was 0.986, the coefficient for OUTDOOR was 0.122, and so on.
The resulting EVR generator equation was
In(EVR) = -0.082 + (0.986)(LGM) + (0.122)(OUTDOOR) +
(0.161)(FAST) + (0.072)(DAYACT) + (-0.174)(SLEEP)
+ (0.047)(WEEKDAY) + (0.060)(HIGHACT) + e.
C-1
-------
This equation was applied to each event listed in the elementary school database.
The values of OUTDOOR, FAST, DAYACT, SLEEP, WEEKDAY, and HIGHACT for
each event were determined by diary entries associated with the event. The value
of LGM was constant for each of the 16 subjects, but was allowed to vary among
subjects. The LGM value for each subject was randomly selected from a normal
distribution with mean = 2.3629 and standard deviation = 0.4324, the normal
distribution specified in Table B-7 for elementary school students.
The value of e was selected from a normal distribution with mean = 0 and
standard deviation = SDRES. The value of SDRES was constant for each subject.
Subject-specific SDRES values were selected from a lognormal distribution defined
by the parameters MU = -1.6068 and SIGMA = 0.4450. These parameter values
were obtained from Table B-6 (elementary school).
Table C-1 provides descriptive statistics for the event EVR values generated
by three applications (runs) of the model to the elementary school database. The
results vary from run to run because of the random elements incorporated into the
Monte Carlo algorithm. Table C-1 also presents the average of the three runs and
descriptive statistics for the observed event EVR values. A comparison of the three-
run model averages with the corresponding observed statistics indicates good
agreement (less than a 10 percent difference) with respect to arithmetic mean,
standard deviation, and percentiles up to the 99th percentile. The model
underestimates the 99.5th percentile (36.32 I- min"1- m"2 versus 48.18 I- min"1- m'2)
and the maximum value (52.70 I- min"1- m"2 versus 86.04 I- min"1- m"2).
This analysis was repeated for the high school database. In this case, the
«
EVR generator equation included a constant (0.16385) and 12 variables. The first
C-2
-------
Table C-1. Descriptive statistics for modeled and observed event EVR values
(elementary school database).
Statistic3
Number of event
EVR values
Arithmetic mean
Arithmetic std. dev.
Skewness"
Kurtosis"
Minimum
25th percentile
50th percentile
75th percentile
90th percentile
95th percentile
98th percentile
99th percentile
99.5th percentile
Maximum
Modeled data
Run 1
870
13.57
7.11
1.67
4.78
4.07
8.44
11.63
16.81
23.55
27.79
33.15
36.03
40.27
67.05
Run 2
870
11.94
5.95
1.05
1.14
2.03
7.64
10.10
15.37
20.31
23.26
27.36
30.19
31.63
41.23
Run 3
870
12.84
6.18
1.44
3.67
2.57
9.20
11.65
15.48
20.30
24.16
32.13
35.86
37.05
49.81
Average of
three statistics
870
12.78
6.41
1.39
3.20
2.89
8.43
11.13
15.89
21.39
25.07
30.88
34.03
36.32
52.70
Observed
data
870
12.45
6.53
3.71
30.71
2.80
8.64
11.22
15.21
19.72
21.98
27.36
30.52
48.18
86.04
aUnits are liters-min"1-m"2 unless otherwise indicated.
"Dimensionless.
C-3
-------
grouping in Table B-5 lists these variables and the associated coefficients. For
example, the table indicates that OUTOTHER is one of the variables and that its
coefficient is 0.11198. Consistent with Table B-7, LGM values for the high school
database were selected from a normal distribution with mean equal to 2.1621 and
standard deviation equal to 0.1890. MU was set equal to -1.4662; SIGMA was
0.2997 (Table B-6).
Table C-2 presents descriptive statistics for three applications of the algorithm
to the high school database, averages of these statistics, and descriptive statistics
for the observed EVR values. The modeled and observed data compare favorably
with respect to the mean, standard deviation, and percentiles up to the 99th
percentile. The model underestimates the 99.5th percentile (21.28 ! min'1- m'2
versus 28.81 I- min'1- m2) and the maximum value (31.61 I- min"1- nY2 versus 48.67
I- min'1- m"2).
The reader will note that the tests discussed above consisted of applying the
algorithm to the same databases from which the algorithm's parameters were earlier
derived. Although these tests provide a test of the general performance of the EVR
algorithm, they do not constitute a true validation of the approach. To be properly
validated, the algorithm should be applied to other Type A databases which have
measured EVR values. As previously indicated, the two Los Angles studies
produced the only Type A databases with measurement-derived EVR values
applicable to the two demographic groups of interest.
C-4
-------
Table C-2. Descriptive statistics for modeled and observed event EVR values (high
school database).
Statistic3
Number of event
EVR values
Arithmetic mean
Arithmetic std. dev.
Skewness0
Kurtosisb
Minimum
25th percentile
50th percentile
75th percentile
90th percentile
95th percentile
98th percentile
99th percentile
99.5th percentile
Maximum
Modeled data
Run 1
2055
9.30
2.82
1.00
2.13
2.47
7.37
8.95
10.78
12.71
14.58
16.97
18.50
19.21
27.66
Run 2
2055
9.55
3.31
2.59
15.53
3.31
7.52
8.99
10.89
13.15
14.98
17.42
21.67
24.68
43.02
Run 3
2055
9.34
2.99
0.92
1.24
3.36
7.21
8.85
11.05
13.21
14.81
17.33
18.42
19.94
24.14
Average of
three statistics
2055
9.40
3.04
1.50
6.30
3.05
7.37
8.93
10.91
13.02
14.79
17.24
19.53
21.28
31.61
Observed
data
2055
9.21
3.75
3.30
22.07
3.73
6.96
8.41
10.59
13.27
15.51
18.25
20.80
28.81
48.67
aUnits are liters min'1- m'2 unless otherwise indicated.
"Dimensionless.
C-5
-------
APPENDIX D
SAMPLE OUTPUT OF pNEM/O3 APPLIED TO
OUTDOOR CHILDREN (HOUSTON, 1-HOUR, DAILY MAXIMUM
0.12 PPM STANDARD [CURRENT NAAQS])
D-1
-------
Table 1.
Cumulative Numbers of People at Hourly 03 Exposures
during 03 Season by Equivalent Ventilation Rate
03 Level
Equalled or Equivalent Ventilation
Exceeded, ppm <15 15-24 25-29
.401+
.381
.361
.341
.321
.301
.281
.261
.241
.221
.201
.181
.161
.141
.121
.101
.081
.061
.041
.021
.001
0.000
0
0
0
0
0
0
0
0
0
0
0
0
0
0
32985
138239
199071
200795
200795
200795
200795
200795
0
0
0
0
0
0
0
0
0
0
0
0
0
0
16177
62718
159375
198912
200795
200795
200795
200795
0
0
0
0
0
0
0
0
0
0
0
0
0
0
737
12837
38956
120332
164154
177364
191899
192185
Rate, l/nin-m**2
30-34 35+
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1361
12074
40966
126712
159649
169544
169621
0
0
0
o
0
0
0
0
0
0
0
0
0
0
0
223
1726
21311
86 588
128260
149209
149209
ANT
0
0
0
0
0
0
0
0
0
0
0
0
0
0
46741
147325
199877
200795
200795
200795
200795
200795
Study Area = HOUSTON 1 1H NAAQS
No. exposure districts =
First day of 03 season =
Last day of 03 season -
No. days in 03 season =
Active Children
11
1
365
365
D-2
-------
Table 2.
Occurrences of People at Hourly Exposures
During 03 Season by Equivalent Ventilation Rate
03 Interval
ppo
.401+
.381 -.400
.361 -.380
.341 -.360
.321 -.340
.301-. 320
.281 -.300
.261-. 280
.241 -.260
.221 -.240
.201 -.220
.181 -.200
.161 -.180
.141-. 160
.121-. 140
.101-. 120
.081 -.100
.061-. 080
.041-. 060
.021 -.040
.001 -.020
0.000
, Equivalent Ventilation Rate
<1S 15-24 25-29
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
35566.
289456.
2092982.
11045318.
54341980.
214922779.
1187831007.
100749878.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
18215.
87374.
S01015.
2952005.
13332051.
38215530.
115687023.
7746919.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
737.
13095.
32032.
230470 .
845236.
1828990 .
3584640.
230200 .
, l/oin-m**2
30-34
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
1361.
10713.
39110.
206308.
493109.
880108.
30271.
35+
0.
0.
0.
o.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
223.
1503.
24266.
106006.
198571.
337160.
AWT
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
54518.
391509.
2638245.
14291171.
68831581.
255658979.
1308319938.
20991. 108778253.
Study Area = HOUSTON 1 1H MAAQS
No. exposure districts =
First day of 03 season =
Last day of 03 season =
No. days in 03 season =
Active Children
11
1
365
365
1758964200.
D-3
-------
Table 1A.
Cumulative Numbers of People at Ihr Dally Max. Exposure
During 03 Season by Equivalent Ventilation Rate
03 Level
Equalled or ' Equivalent Ventilation
Exceeded, ppm <15 15-24 25-29
.401+
.381
.361
.341
.321
.301
.281
.251
.241
.221
.201
.181
.161
.141
.121
.101
.081
.061
.041
.021
.001
0.000
0
0
0
0
0
0
0
0
0
0
0
0
0
0
32985
130318
198968
200795
200795
200795
200795
200795
0
0
0
0
0
0
0
0
0
0
0
0
0
0
14514
53011
147265
193853
200795
200795
200795
200795
0
0
0
0
0
0
0
0
0
0
0
0
0
0
737
7259
24443
98331
149552
159632
159632
159632
Rate, l/a>in-B**2
30-34 35+
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1361
10252
27522
78056
118578
122138
122138
0
0
0
0
0
o
0
0
0
0
0
0
0
0
0
223
685
4467
38071
55979
63185
63185
ANT
0
0
0
0
0
0
0
0
0
0
0
0
0
0
46741
147325
199877
200795
200795
200795
200795
200795
Study Area - HOUSTON 1 1H NAAQS
No. exposure districts =
First day of 03 season =
Last day of 03 season -
Ho. days in 03 season =
Active Children
11
1
365
365
D-4
-------
Table 2A.
Occurrences of People at Ihr Dal.ly Wax. Exposure
During 03 Season by Equivalent Ventilation Rate
03 Interval, Equivalent Ventilation Rate
ppm ' <15 15-24 25-29
.401+
.381 -.400
.361 -.380
.341-. 360
.321 -.340
.301 -.320
.281 -.300
.261 -.280
.241-. 260
.221-. 240
.201 -.220
.181 -.200
.161-. 180
.141-. 160
.121-. 140
.101-. 120
.081-. 100
.061 -.080
.041 -.060
.021 -.040
.001 -.020
0.000
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
34071 .
177630.
1147474.
4861607.
17066439.
27047686.
8661407.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
15794.
52623.
316691.
1339265.
4311111.
5767317.
1346967 .
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
737.
7517.
17407 .
115031.
297854.
350094.
53064.
0.
, l/nin-B**2
30-34
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
1361.
8891.
21835.
61978.
118233.
12471.
0.
35+
0.
0.
0.
0.
0.
0.
0..
0.
0.
0.
0.
0.
0.
0.
0.
223.
462.
3782.
37902.
26420.
8831.
0.
ANT
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
50602.
239354.
1490925.
6341520.
21775284.
33309750.
10082740.
0.
Study Area = HOUSTON 1 1H NAAQS Active Children
No. exposure districts = 11
First day of 03 season = 1
Last day of 03 season = 365
No. days in 03 season = 365
73290175.
D-5
-------
Table IB.
Cumulative Numbers of People at 1-Hr Daily Max. Dose
During 03 Season by 1-Hr 03 and EVR.
03 Level
Equalled or Equivalent Ventilation
Exceeded, ppm <15 15-24 25-29
.401 +
.381
.361
.341
.321
.301
.281
.261
.241
.221
.201
.181
.161
.141
.121
.101
.081
.061
.041
.021 -
.001
0.000
0
0
0
0
0
0
0
0
0
0
0
0
0
0
17621
99632
192813
200795
200795
20079S
200795
200795
0
0
0
0
0
0
0
0
0
0
0
0
0
0
11048
51453
153532
198912
200795
200795
200795
200795
0
0
0
0
0
0
0
0
0
0
0
0
0
0
737
12837
38052
109423
161301
175587
178108
178108
SSZSSSSSSSSSSSS'XSS ZSSSSZ
Rate, l/min-n**2
30-34 35+
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1361
11625
37937
123489
158838
160947
160947
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
223
1726
21311
85558
117307
127902
127902
S=SSSSSBSS
ANY
0
0
0
0
0
0
0
0
0
0
0
0
0
0
28669
128518
198747
200795
200795
200795
200795
200795
Study Area » HOUSTON 1 1H MAAQS
Ko. exposure districts =
First day of 03 season =
Last day of 03 season =
No. days in 03 season =
Active Children
11
1
365
365
D-6
-------
Table 2B.
Occurrences of People at 1-Hr Daily flax. Dose
During 03 Season by 1-Hr 03 and EVR.
03 Interval,
PP<»
.401+
.381 -.400
.361 -.380
.341 -.360
.321-. 340
.301-. 320
.281-. 300
.261-. 280
.241 -.260
.221 -.240
.201-. 220
.181 -.200
.161-. 180
.141-. 160
.121-. 140
.101-. 120
.081 -.100
.061-. 080
.041 -.060
.021 -.040
.001-.020'
0.000
Equivalent Ventilation Rate
<15 1S-24 25-29
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
18707.
129380.
746113.
3453558.
13639830.
24904343.
9397037.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
12328.
51506.
303758.
1466922.
5422353 .
8553473.
2547775.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
737.
12100.
28178.
173323.
545593.
794265.
224343 .
0.
, l/fflin-o**2
30-34
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
1361.
10264.
35606.
170182.
278722.
61142.
0.
35+
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
223.
1503.
22538.
89751.
141828.
51433.
0.
ANY
0
0
0
0
0
0
0
0
0
0
0
0
0
0
31772
194570
1089816
5151947
19867709
34672631
12281730
0
Study Area = HOUSTON 1 1H NAAQS Active Children
No. exposure districts = 11
First day of 03 season = 1
Last day of 03 season = 365
No. days in 03 season = 365
73290175.
D-7
-------
Table 3.
Number of People at Their Highest Ihr Daily flax. Exposure
During 03 Season by Ventilation Rate Categories
__=w=___=====.
03 Level
Equalled or
Exceeded, ppo
.401 +
.381 -.400
.361 -.380
.341 -.360
.321 -.340
.301 -.320
.281 -.300
.261-. 280
.241-. 260
.221 -.240
.201-. 220
.181 -.200
.161-. 180
.141-. 160
.121-. 140
.101-. 120
.081 -.100
.061 -.080
.041-. 060
.021 -.040
.001-. 020
0.000
Equivalent Ventilation
<15 15-24 25-29
0
0
0
0
0
0
0
0
0
0
0
0
0
0
32985
97333
68650
1827
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
14514
38497
94254
46588
6942
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
737
6522
17184
73888
51221
10080
0
0
Rate, l/min-B**2
30-34 35+
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1361
8891
17270
50534
40522
3560
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
223
462
3782
33604
17908
7206
0
ANT
0
0
0
0
0
0
0
0
0
0
0
0
0
0
46741
100584
52552
918
0
0
0
0
Study Area = HOUSTON 1 1H NAAQS
No. exposure districts -
First day of 03 season =
Last day of 03 season =
No. days in 03 season =
Active Children
11
1
365
365
D-8
-------
Table 4.
Cumulative Numbers of People at 8-Hr Daily Max. Exposure
During 03 Season by 8-Hr Equivalent Ventilation Rate
03 Level
Equalled or 8hr
Exceeded, ppm <15
.201 +
.191
.181
.171
.161
.151
.141
.131
.121
.111
.101
.091
.081
.071
.061
.041
.021
.001
0.000
0
0
0
0
0
0
0
0
0
0
0
8727
34682
118314
191737
200795
200795
200795
200795
Equivalent
15-24
0
0
0
0
0
0
0
0
0
0
0
0
2907
29241
85482
152651
163895
171251
171251
Ventilation Rate, l/aiin-m**2
25-29 30-34 35+
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1492
11745
38761
50236
50236
0
0
0
0
0
0
0
0
0
0
0
0
0
300
300
300
300
300
300
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
AJTT
0
0
0
0
0
0
0
0
0
0
0
8727
37589
125362
195230
200795
200795
200795
200795
Study Area = HOUSTON 1 1H NAAQS
No. exposure districts -
First day of 03 season =
Last day of 03 season =
Ho. days in 03 season =
Active Children
11
1
365
365
D-9
-------
Table 5.
Occurrences of People at 8-Hr Daily Max. Exposure
During 03 Season by 8-Hr Equivalent Ventilation Rate
03 Interval,
ppm
.201 +
.191-. 200
.181-. 190
.171-. 180
.161-. 170
.151-. 160
.141-. 150
.131-. 140
.121-. 130
.111-. 120
.101-. 110
.091-. 100
.081 -.090
.071 -.080
.061 -.070
.041 -.060
.021 -.040
.001-. 020
0.000
8b.r Equivalent Ventilation Rate, l/nin-m**2
<15 15-24 25-29 30-34 35+
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
8727-
26437.
135233.
454366.
5272620.
27544063.
29954696.
194.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
2907.
29070.
108833 .
1128984.
4750842.
3817777-
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
1492.
10253.
30S4S.
12836 .
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
300.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
ANY
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
8727.
29344.
164603.
564691 .
6411857.
32325450.
33785309.
194.
Study Area = HOUSTON 1 1H NAAQS
Ho. exposure districts =
First day of 03 season =
Last day of 03 season =
Ho. days in 03 season =
Active Children
11
1
365
365
73290175.
D-10
-------
Table 4A.
Cumulative Numbers of People at 8-Hr Daily Max. Dose
During 03 Season by 8-Hr 03 aod 8-Hr EVR.
03 Level
Equalled or
Exceeded , ppo
.201+
.191
.181
.171
.161
.151
.141
.131
.121
.111
.101
.091
.081
.071
.061
.041
.021
.001
0.000
8hr
<15
0
0
0
0
0
0
0
0
0
0
0
8727
27626
112823
190333
200795
200795
200795
200795
Equivalent
15-24
0
0
0
0
0
0
0
0
0
0
0
0
2907
29534
85920
154503
173945
178604
178604
Ventilation
25-29
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1699
12620
45921
59947
59947
Rate,
30-34
0
0
0
0
0
0
0
0
0
0
0
0
0
300
300
300
300
300
300
l/min-m**2
35+
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
- 0
0
0
0
ANT
0
0
0
0
0
0
0
0
0
0
0
8727
30533
119871
194323
200795
200795
200795
200795
Study Area = HOUSTON 1 1H NAAQS
No. exposure districts *
First day of 03 season =
Last day of 03 season -
No. days in 03 season =
Active Children
11
1
365
365
D-11
-------
Table SA.
Occurrences of.People at 8-Hr Daily Max. Dose
During 03 Season by 8-Hr 03 and 8-Hr EVR
03 Interval,
ppa
.201+
.191-. 200
.181-. 190
.171-. 180
.161-. 170
.151-. 160
.141-. ISO
.131 -.140
.121-. 130
.111-. 120
.101-. 110
.091-. 100
.081 -.090
.071 -.080
.061-. 070
.041 -.060
.021-. 040
.001 -.020
o.doo
8hr Equivalent Ventilation Rate, l/min-B**2
<1S 15-24 25-29 30-34 35+
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
8727.
19088.
133884.
404172.
4834807.
25707664.
30043074.
1965.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
2907.
29363.
107590.
1206173.
5267757.
5450445 .
931.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
1699.
12424.
38409.
18796.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
300.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
ANT
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
8727.
21995.
163547.
513461.
6053404.
31013830.
35512315.
2896.
Study Area = HOUSTON 1 1H NAAQS Active Children
No. exposure districts = 11
First day of 03 season = 1
Last day of 03 season = 365
No. days in 03 season = 365
73290175.
D-12
-------
Table 6.
Number of People at Their Highest 8-Hr Daily Max. Exposure
During 03 Season by 8-Hr Ventilation Rate Categories
03 Level
Equalled or
Exceeded, pp
.201+
.191-. 200
.181-. 190
.171-. 180
.161-. 170
.151-. 160
.141-. 150
.131-. 140
.121-. 130
.111-. 120
.101-. 110
.091-. 100
.081 -.090
.071 -.080
.061 -.070
'.041-.060
.021 -.040
.001-. 020'
0.000
======= =====!
8hr
m <15
0
0
0
0
0
0
0
0
0
0
0
8727
25955
83632
73423
9058
0
0
0
=============
Equivalent
15-24
0
0
0
0
0
0
0
0
0
0
0
0
2307
26334
56241
67169
11244
7356
0
============
Ventilatiot
25-29
0
0
0
0
0
0
0
0
0
0
' 0
0
0
0
1492
10253
27016
11475
0
===========
i Rate, :
30-34
0
0
0
0
0
0
0
0
0
0
0
0
0
300
0
0
0
0
0
========
L/nin-m**2
35+
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
=============±
AHY
0
0
0
0
0
0
0
0
0
0
0
8727
28862
87773
69868
5565
0
0
0
=======
Study Area = HOUSTON 1 1H NAAQS Active Children
Ho. exposure districts » 11
First day of 03 season = 1
Last day of 03 season = 365
No. days in 03 season = 365
D-13
-------
Table 7.
Cumulative Numbers of People at 8-Hr Daily Max.
Seasonal Mean (April to October) Exposure
03 Level
Equalled or
Exceeded , ppm
.071+
.066
.061
.056
.051
.046
.041
.036
.031
.026
.021
.011
.001
0.000
0
0
0
0
0
0
0
293
16543
171987
200514
200795
200795
200795
Stud? Area = HOUSTON 1 1H KAAQS Active Children
No. exposure districts = 11
First day of 03 season = 1
Last day of 03 season - 365
No. days in 03 season - 365
D-14
-------
Table 8.
Occurrences of People at 8-Hr Daily Max.
Seasonal Mean (April to October) Exposure
03 Interval ,
ppn
.071 +
.066 -.070
.061-. 065
.056-. 060
.051-. 055
.046 -.050
.041 -.045
.036 -.040
.031-. 035
.026-. 030
.021-. 025
.011-. 020
.001 -.010
0.000
0
0
0
0
0
0
0
293
16250
155444
28527
281
0
0
Stud? Area * HOUSTON 1 1H NAAQS Active Children
No. exposure districts = 11
First day of 03 season = 1
Last da? of 03 season - 365
No. days in 03 season - 365
D-15
-------
Table 9.
Number of People at Dally Max Dose that Exceed
Specified 1-HR 03 Levels 1 or Wore Tines per Year
03 Level
Equalled or
Exceeded, ppo
.401 +
.381
.361
.341
.321
.301
.281
.261
.241
.221
.201
.181
.161
.141
.121
.101
.081
.061
.041
.021
.001
0.000
1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
25566
62512
9014
0
0
0
0
0
2
0
0
0
0
0
0
0
0
0
0
0
0
0
0
3103
40156
16381
0
0
0
0
0
Days / Yeai
3
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
20619
4226
0
0
0
0
0
r
4
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
4494
17004
0
0
0
0
0
5
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
737
30921
645
0
0
0
0
>5
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
121201
200150
200795
200795
200795
20079 S
Study Area = HOUSTOH 1 1H NAAQS
Ho. exposure districts =
First day of 03 season =
Last day of 03 season =
No. days in 03 season -
Active Children
11
1
365
365
D-16
-------
Table 10.
Number of People at Dally Hax 8-HR Dose that Exceed
Specified 8-hr 03 Levels 1 or More Times per Tear
03 Level
Equalled or
Exceeded , ppo
.201+
.191
.181
.171
.161
.151
.141
.131
.121
.111
.101
.091
.081
.071
.061
.041
.021
.001
0.000
1
0
0
0
0
0
0
0
0
0
0
0
8727
30344
61949
24028
0
0
0
0
]
2
0
0
0
0
0
0
0
0
0
0
0
0
189
44102
31750
0
0
0
0
Days / Yeai
3
0
0
0
0
0
0
0
0
0
0
0
0
0
11353
41910
0
0
0
0
4
0
0
0
0
0
0
0
0
0
0
0
0
0
2278
40920
0
0
0
0
5
0
0
0
0
0
0
0
0
0
0
0
0
0
189
30957
0
0
0
0
>5
0
0
0
0
0
0
0
0
0
0
0
0
0
0
24758
200795
200795
200795
200795
Study Area = HOUSTON 1 1H XAAQS
No. exposure districts «
First day of 03 season =
Last day of 03 season *
No. days in 03 season =
Active Children
11
1
365
365
D-17
-------
Table 11.
Number of People that Exceed Specified 03 Levels
at 1-HR Daily Max Dose 1 or More Tines per Tear
with Ventilation Rates of 30 or Higher
03 Level
Equalled or
Exceeded, ppm
.401 +
.381
.361
.341
.321
.301
.281
.261
.241
.221
.201
.181
.161
.141
.121
.101
.081
.061
.041
.021
.001
0.000
1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1584
13351
44585
41485
21033
20265
20265
1
2
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
10758
S2718
20887
8764
8764
Days / lea:
3
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1798
13963
7862
10460
10460
r
4
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
26927
27412
24171
24171
5
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
4936
16794 "
17606
17606
>5
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1705
69081
83308
83308
Study Area = H00STON 1 1H NAAQS
Ho. exposure districts =
First day of 03 season =
Last day of 03 season -
No. days in 03 season -
Active Children
11
1
365
365
D-18
-------
Table 12.
Xunber of People that Exceed Specified 8.HR 03 Levels
at Daily Hax 8-HR Dose 1 or More Tines per Tear
with 8 Hour Ventilation Rates from 13 through 27
03 Level
Equalled or
Exceeded , ppm
.201 +
.191
.181
.171
.161
.151
.141
.131
.121
.111
.101
.091
.081
.071
.061
.041
.021
.001
0.000
1
0
0
0
0
0
0
0
0
0
0
0
6055
8962
37727
40668
19611
1437
785
785
2
0
0
0
0
0
0
0
0
0
0
0
0
0
13413
39607
11717
3421
236
236
Days / Yea
3
0
0
0
0
0
0
0
0
0
0
0
0
0
1179
15988
3357
3914
3247
3247
r
4
0
0
0
0
0
0
0
0
0
0
0
0
0
0
13904
1033
16988
1663
1663
5
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1180
4014
10298
8055
8055
>5
0
0
0
0
0
0
0
0
0
0
0
0
0
0
293
141100
164421
186809
186809
Study Area * HOUSTON 1 1H NAAQS
Ho. exposure districts =
First day of 03 season =
Last day of 03 season =
Mo. days in 03 season =
Active Children
11
1
365
365
D-19
-------
APPENDIX E
ONE-HOUR EXPOSURE DISTRIBUTIONS
E-1
-------
250
FIGURE E-1. ONE-HOUR EXPOSURE DISTRIBUTIONS FOR OUTDOOR
CHILDREN EXPOSED ON ONE OR MORE DAYS UNDER HEAVY
EXERTION (EVR 30+ LITERS/MIN-M2) IN PHILADELPHIA, PA
ASIS
1112
8109
8108
-B-
8110
-e-
8508
8107
1110
8509
\ /
0.02 0.04
0.06 0.08 0.1 0.12 0.14
CONCENTRATION, PPM
0.16 0.18
FIGURE E-2. ONE-HOUR EXPOSURE DISTRIBUTIONS OF TOTAL
OCCURRENCES FOR OUTDOOR CHILDREN EXPOSURE UNDER HEAVY
EXERTION (EVR 30+ LITERS/MIN-M2) IN PHILADELPHIA, PA
0.02 0.04 0.06 0.08 0.1 0.12 0.14
CONCENTRATION, PPM
0.16 0.18 0.2
E-2
-------
FIGURE E-3. ONE-HOUR EXPOSURE DISTRIBUTIONS FOR OUTDOOR
CHILDREN EXPOSED ON ONE OR MORE DAYS UNDER HEAVY
EXERTION (EVR 30+ LITERS/MIN-M2) IN HOUSTON, TX
200
ASIS
1112
8109
8108
8508
\l/
/T\
8107
1110
850S
0.02 0.04 0.06 0.08 0.1 0.12 0.14 0.16 0.18 0.2
CONCENTRATION, PPM
FIGURE E-4. ONE-HOUR EXPOSURE DISTRIBUTIONS OF TOTAL
OCCURRENCES FOR OUTDOOR CHILDREN EXPOSURE UNDER HEAVY
EXERTION (EVR 30+ LITERS/MIN-M2) IN HOUSTON, TX
3.000
CO
Q
CO
O
2,500
2,000 -
CO
LU
O
LU
CC.
K
O
O
O
CO
1,500 -
1,000 -
500 -
ASIS
1112
-+
8109
-*
8108
8110
x\
8508
xT\
8107
1110
8509
0.02
0.04 0.06 0.08 0.1 0.12 0.14 0.16 0.18 0.2
CONCENTRATION, PPM
E-3
-------
FIGURE E-5. ONE-HOUR EXPOSURE DISTRIBUTIONS FOR OUTDOOR
CHILDREN EXPOSED ON ONE OR MORE DAYS UNDER HEAVY
EXERTION (EVR 30+ LITERS/MIN-M2) IN NEW YORK, NY
700
ASIS
1112
8109
8108
81 U)
/^
8508
-*-
8107
1110
8509
0 0.02 0.04 0.06 0.08 0.1 0.12 0.14
CONCENTRATION, PPM
0.16 0.18
FIGURE E-6. ONE-HOUR EXPOSURE DISTRIBUTIONS OF TOTAL
OCCURRENCES FOR OUTDOOR CHILDREN EXPOSURE UNDER HEAVY
EXERTION (EVR 30+ LITERS/MIN-M2) IN NEW YORK, NY
0.02 0.04
0.06 0.08 0.1 0.12 0.14
CONCENTRATION, PPM
0.16 0.18
0.2
E-4
-------
FIGURE E-7. ONE£-HOUR EXPOSURE DISTRIBUTIONS FOR OUTDOOR
CHILDREN EXPOSED ON ONE OR MORE DAYS UNDER HEAVY
EXERTION (EVR 30+ LITERS/MIN-M2) IN WASHINGTON, D.C.
ASIS
1112
8109
8108
8110
s\
8508
8107
1110
8509
0.02 0.04
0.06 0.08 0.1 0.12 0.14
CONCENTRATION. PPM
0.16 0.18 0.2
FIGURE E-8. ONE-HOUR EXPOSURE DISTRIBUTIONS OF TOTAL
OCCURRENCES FOR OUTDOOR CHILDREN EXPOSURE UNDER HEAVY
EXERTION (EVR 30+ LITERS/MIN-M2) IN WASHINGTON, D.C.
1,200
0.02 0.04
0.06 0.08 0.1 0.12 0.14
CONCENTRATION, PPM
0.16 0.18
0.2
E-5
-------
FIGURE E-9. ONE-HOUR EXPOSURE DISTRIBUTIONS FOR OUTDOOR
CHILDREN EXPOSED ON ONE OR MORE DAYS UNDER MODERATE
EXERTION (EVR 16-30 LITERS/MIN-M2) IN PHILADELPHIA, PA
300
ASIS
1112
8109
8108
&
8110
O-
8508
-*-
8107
1110
8509
0.02 0.04
0.06 0.08 0.1 0.12 0.14
CONCENTRATION, PPM
0.16 0,18 0.2
FIGURE E-10. ONE-HOUR EXPOSURE DISTRIBUTIONS OF TOTAL
OCCURRENCES FOR OUTDOOR CHILDREN EXPOSURE UNDER MODERATE
EXERTION (EVR 16-30 LITERS/MIN-M2) IN PHILADELPHIA, PA
100,000
oo
Q
<
co
Z)
o
CO
111
o
LU
OL
-------
FIGURE E-11. ONE-HOUR EXPOSURE DISTRIBUTIONS FOR OUTDOOR
CHILDREN EXPOSED ON ONE OR MORE DAYS UNDER MODERATE
EXERTION (EVR 16-30 LITERS/MIN-M2) IN HOUSTON, TX
ASIS
1112
-4-
8109
8108
8508
-*
8107
£\
1110
8509
0 0.02 0.04 0.06 0.08 0.1 0.12 0.14 0.16 0.18 0.2
CONCENTRATION, PPM
FIGURE E-12. ONE-HOUR EXPOSURE DISTRIBUTIONS OF TOTAL
OCCURRENCES FOR OUTDOOR CHILDREN EXPOSURE UNDER MODERATE
EXERTION (EVR 16-30 LITERS/MIN-M2) IN HOUSTON, TX
160,000
0.02 0.04
0.06 0.08 0.1 0.12 0.14
CONCENTRATION, PPM
0.16 0.18 0.2
E-7
-------
FIGURE E-13. ONE-HOUR EXPOSURE DISTRIBUTIONS FOR OUTDOOR
CHILDREN EXPOSED ON ONE OR MORE DAYS UNDER MODERATE
EXERTION (EVR 16-30 LITERS/MIN-M2) IN NEW YORK, NY
ASIS
1112
8109
8108
-B-
8110
8508
-*
8107
1110
8509
\ /
0.02 0.04
0.06 0.08 0.1 0.12 0.14
CONCENTRATION, PPM
0.16 0.18 0.2
FIGURE E-14. ONE-HOUR EXPOSURE DISTRIBUTIONS OF TOTAL
OCCURRENCES FOR OUTDOOR CHILDREN EXPOSURE UNDER MODERATE
EXERTION (EVR 16-30 LITERS/MIN-M2) IN NEW YORK, NY
300,000
CO
Z 250,000
I
O
E 200,000
CO
LU
O
g 150,000
cr
a:
o 100,000
9
O
CO
a:
UJ
a.
50,000 -
0.02 0.04 0.06 0.08 0.1 0.12 0.14
CONCENTRATION, PPM
0.16 0.18 0.2
E-8
-------
250
FIGURE E-15. ONE-HOUR EXPOSURE DISTRIBUTIONS FOR OUTDOOR
CHILDREN EXPOSED ON ONE OR MORE DAYS UNDER MODERATE
EXERTION (EVR 16-30 LITERS/MIN-M2) IN WASHINGTON, D.C.
ASIS
1112
8109
-*-
8108
811C)
/\
8508
8107
1110
8509
0.16 0.18 0.2
0
0 0.02 0.04 0.06 0.08 0.1 0.12 0.14
CONCENTRATION, PPM
FIGURE E-16. ONE-HOUR EXPOSURE DISTRIBUTIONS OF TOTAL
OCCURRENCES FOR OUTDOOR CHILDREN EXPOSURE UNDER MODERATE
EXERTION (EVR 16-30 LITERS/MIN-M2) IN WASHINGTON, D.C.
0.02 0.04
0.06 0.08 0.1 0.12 0.14
CONCENTRATION. PPM
0.16 0.18 0.2
E-9
-------
APPENDIX F
ESTIMATION OF OZONE EXPOSURES IN OUTDOOR CHILDREN
FOR SPECIAL 8H10EX-80 SCENARIO
F-1
-------
INTERNATIONAL
TECHNOLOGY
CORPORATION
February 23, 1996
IT Project No. 763997-7
Mr. Harvey Richmond
U.S. Environmental Protection Agency
OAQPS, MD-12
RTF, North Carolina 27711
Estimation of Ozone Exposures in Outdoor Children for Special 8H10EX-80 Scenario
Dear Harvey:
Under Work Assignment 2-7 of EPA Contract No. 68-D3-0094, IT Air Quality Services
(ITAQS) has performed a sensitivity analysis using the outdoor children version of
pNEM/03. In this analysis, ITAQS examined the ozone exposures that would occur in
each of seven study areas when ozone levels meet a special set of conditions: the number
of daily maximum eight-hour concentrations exceeding 80 ppb equals 10. This letter
provides a summary of the procedures used in this sensitivity analysis and summarizes the
results.
Background
The Office of Air Quality Planning and Standards (OAQPS) has conducted a series of
exposure assessments using pNEM/03 in which the ozone levels within a specified study
area have been adjusted to meet a particular formulation of the ozone NAAQS. One of the
standards under review (designated 8H5EX-80) states that the expected exceedance rate for
daily maximum 8-hour ozone concentrations above 80 ppb shall not be more than five. To
evaluate this standard, ITAQS adjusted the ozone monitoring data representing each study
area using the Air Quality Adjustment Procedure (AQAP) described in recent pNEM/O3
project reports. As a result of this procedure, the ozone data reported by each monitor was
adjusted so that the sixth highest daily maximum 8-hour concentration equaled a specified
air quality indicator (AQI). The sixth highest value of the historical "high ozone" monitor
was adjusted to equal 80 ppb.
Regional Office
South Square Corporate Center One 3710 University Drive. Suite 201 Durham, North Carolina 27707-6208
919-493-3661 FAX: 919-493-1773
IT Corporation Is a wholly owned subsidiary ot International Technology Corporation
-------
Mr. Harvey Richmond 2 February 23. :996
This adjustment procedure is intended to limit the average exceedance rate of the hi^h
ozone monitor to five exceedances of 80 ppb per year, based on a single year of monitoring
data. EPA has recently begun to evaluate an alternative form of this standard which limits
the average value of the fifth highest daily maximum 8-hour concentration to 80 ppb (here
designated 8H5AVG-80). Under.this standard, there is no explicit limit to the number of
exceedances that can occur in a given year. However, a recent analysis by Warren Freas of
OAQPS found that very few ozone monitors report more than 10 exceedances during a
single year in an area that meets the 8H5AVG-80 standard over a three-year period.° As a
result of this analysis, EPA directed ITAQS to develop a procedure for adjusting the
monitoring data in an area to simulate conditions in which 10 exceedances occur at the
historical high-ozone monitor. These data would then be used in a pNEM/03 analysis to
estimate the ozone exposures that could occur under these conditions. The next section of
this letter briefly describes the AQAP developed by ITAQS.
The Air Quaiirv Adjustment Procedure
The AQAP for the 10 exceedance scenario is similar to that used for adjusting ozone data
to simulate attainment of an 8H5EX standard. In essence, the data are adjusted to meet an
8H10EX-80 standard, i.e., the expected number of daily maximum eight-hour ozone
concentrations exceeding 80 ppb shall not exceed ten. The procedure is outlined in Table 1
of this letter. Note that supplementary material concerning Step 6 of'the procedure can be
found in Section 5.3 of the ITAQS project report describing the application of pNEM/03 to
outdoor children.
Section 5.4 of the outdoor children report describes the application of an AQAP for the
8H5EX-80 standard to Philadelphia. The new procedure described in this letter is
essentially identical to the procedure in Section 5.4 when one makes the following
substitutions throughout the discussion: substitute llth highest value for sixth highest value
and substitute RATIOS for RATI02. Table 2 lists values of RATIOS by study area.
The adjustment procedure was applied to the ozone monitoring data which have been used
in previous pNEM/03 analyses, of seven study areas: Chicago, Houston, Los Angeles,
New York, Philadelphia, St. Louis, and Washington, D.C. The two remaining pNEM/03
study areas (Denver and Miami) were omitted from the analysis because the ozone levels in
these cities were relatively low with respect to the levels permitted by the 8H5AVG80
standard.
F-3
-------
Mr. Harvey Richmond 3 February 23, 1996
TABLE 1. AIR QUALITY ADJUSTMENT PROCEDURE USED TO SLMULATE
SPECIAL ATTAINMENT CONDITIONS (CONDITIONS EQUIVALENT TO
ATTAINMENT OF 8H10EX STANDARD)
1. Determine the following quantities.
EHllLDM(iJ): the llth largest eight-hour daily maximum concentration
of the i-th ranked site in City j for the baseline year.
MAXEH1 lLDM(j): the largest EH11LDM of all sites in City j for the
baseline year.
AMAXEH11LDMG): the largest EH11LDM value permitted under the standard
(i.e., 80 ppb).
2. Select five years prior to the baseline year and determine the value of
EH11LDM at each site m in City j for each year. Rank these values by city and
year. Let RANK(m,j,y) indicate the rank of site m in city j in year y. Let
MEANRAhfK(m,j) indicate the mean value of RANK(m,j,y) over the five years.
Rank the MEANRANK(mj) values and let RELRANK(mj) indicate the relative
rank of MEANRANK(m,j).
3. Calculate an adjusted EH11LDM for the i-th ranked site in City j by the
expression
AEHllLDM(ij) = [EHllLDM(iJ)][AiV£AXEHllLDMG)]/[MAXEHllLDMO')].
4. If RELRANK(mJ) = i, then m will be the i-th ranked site in City j under
attainment. That is,
AEHIlLDM(m,j) = AEHllLDM(ij) if RELRANK(mj) = i.
5. ' Use the equation
ACLV1 =(RATI03)(AEH11LDM)
to estimate the characteristic largest one-hour value (CLV1) associated with each
AEH1 lLDM(m,J) value. Denote this value as ACLVl(mj). Values of RATIOS
are listed by city in Table 2.
6. The one-hour data for Site m are adjusted so that a Weibull distribution fit to the
adjusted data will have a CLV1 equal to ACLVl(ij) where i = RELRANK(mJ).
Subsection 5.3 of the outdoor children pNEM/O3 report provides a method for
estimating the parameters of this distribution and making the adjustment.
F-4
-------
Mr. Harvey Richmond ' 4 February 23, 1996
TABLE 2. RATIOS VALUES BY STUDY AREA
City
Chicago
Denver
Houston
Los Angeles
Miami
New York
Philadelphia
St. Louis
Washington
RATIOS'
1.583
1.627
2.346
1.945
1.697
1.647
1.465
1.598
1.596
'RATIOS = (ACLV1)/(AEH11LDM)
Exposure Estimates for Selected Studv Areas
The pNEM/03 model incorporates a number of stochastic (random) elements which directly
affect the exposure estimates produced by the model. Consequently, exposure estimates are
likely to vary from run to run. Consistent with earlier analyses, ITAQS ran the model 10
times for each of the seven study areas. Tables 3 through 10 provides means and ranges
for selected exposure indicators based on these runs. In each case, the exposure estimates
apply to the population group previously designated as "outdoor children" and use the _
adjusted ozone data described above. The exposure indicators are defined in Sections 1.2
and 7.4 of the pNEM/03 project report for outdoor children. ^
In Tables 3 through 10 the attainment scenario is described in terms of a "8H10EX-80"
scenario, as the ozone monitoring data were adjusted to simulate attainment of this
indicator In using this designation, it is understood that the scenario is actually intended to
represent a special high-ozone situation that could occur during a single year when a
8H5AVG-80 standard is attained over a three-year period.
Table 3 lists exposure estimates for number and percent of outdoor children experiencing
one or more one-hour daily maximum ozone exposures above 120 ppb at any ventilation
rate These results for the 8H10EX-80 scenario can be compared with similar estimates for
. other scenarios in Table 50 of the pNEM/03 project report for outdoor children. The
vleslo H?o1x-80 Led for each city in Table 3 fall between the corresponding values
for 8H5EX-80 and 8H5EX-90 in Table 50, regardless of study area.
F-5
-------
TABLE 3. NUMBER AND PERCENT OF OUTDOOR CHILDREN EXPERIENCING ONE OR MORE
ONE-HOUR DAILY MAXIMUM OZONE EXPOSURES ABOVE 120 PPB AT ANY VENTILATION RATE UNDER
8H10EX-80 SCENARIO
Study Area
Chicago
Houston
Los Angeles
New York
Philadelphia
St. Louis
Washington, DC
Number of Persons
at Risk
472,710
200,795
798,290
782,600
275,320
128,250
198,860
Mean
Number of
Persons Exposed
169,006
127,114
41,507
66,393
553
23,331
24,811
Percent of
Total
35.75
63.31
5.20
8.48
0.20
18.19
12.48
Range
Number of Persons
Exposed
137,422 - 213,679
120,022 - 132,678
33,105 - 46,365
56,842 - 73,325
0 - 3,244
19,971 - 29,932
16,941 - 30.047
Percent
of Total
29.07 - 45.20
59.77 - 66.08
4.15 - 5.81
7.26 - 9.37
0.00 - 1.18
15.57 - 23.34
8.52 - 15.11
-------
TABLE A. NUMBER AND PERCENT OF OUTDOOR CHILDREN EXPERIENCING ONE OR MORE
EIGHT-HOUR DAILY MAXIMUM OZONE EXPOSURES ABOVE 60 PPB AT ANY VENTILATION RATE
UNDER THE 8HIOEX-80 SCENARIO
Study Area
Chicago
Houston
Los Angeles
New York
Philadelphia
St. Louis
Washington, DC
Number of Persons at
Risk
472,710
200,795
798,290
782,600
275,320
128,250
198,860
Mean
Number of
Persons Exposed
471,354
175,837
223,914
593,320
263,827
113,782
195,024
Percent of
Total
99.71
87.57
28.05
75.81
95.83
88.72
98.07
Range
Number of Persons
Exposed
467,714 - 472,710
168,175 - 184,677
217,662 - 232,082
582,353 - 600,824
259,451 - 268,140
111,825 - 116,372
189,346 - 197,510
Percent
of Total
98.94 - 100.00
83.75 - 91.97
27.27 - 29.07
74.41 - 76.77
94.24 - 97.39
87.19 - 90.74
95.22 - 99.32
-------
TABLE 5. NUMBER AND PERCENT OF OUTDOOR CHILDREN EXPERIENCING ONE OR MORE
EIGHT-HOUR DAILY MAXIMUM OZONE EXPOSURES ABOVE 80 PPB AT ANY VENTILATION RATE
UNDER THE 8H10EX-80 SCENARIO
Study Area
Chicago
Houston
Los Angeles
New York
Philadelphia
St. Louis
Washington, DC
Number of Persons at
Risk
472,710
200,795
798,290
782,600
275,320
128,250
198,860
Mean
Number of
Persons Exposed
243,097
95,348
55,361
158,065
85,648
41,380
86,127
============;
Percent of
Total
51.43
47.49
6.93
20.20
31.11
32.27
43.31
======:
Range
Number of Persons
Exposed
215,145 -278,767
85,124 - 109,717
51,975 -62,295
145,057 - 173,013
74,059 - 99,292
37,006 - 45,013
79,912 - 94,154
' _
Percent
of Total
45.51 - 58.97
42.39 - 54.64
6.51 - 7.80
18.54 - 22.11
26.90 - 36.06
28.85 - 35.10
40.19 - 47.35
i========
CD
-------
TABLE 6. NUMBER AND PERCENT OF OUTDOOR CHILDREN EXPERIENCING ONE OR MORE
EIGHT-HOUR DAILY MAXIMUM OZONE EXPOSURES ABOVE 100 PPB AT ANY VENTILATION RATE
UNDER THE 8H10EX-80 SCENARIO
ji
CD
Study Area
Chicago
Houston
Los Angeles
New York
Philadelphia
St. Louis
Washington, DC
Number of Persons at
Risk
472,710
200,795
798,290
782,600
275,320
128,250
198,860
Mean
Number of
Persons Exposed
10,210
19,023
114
7,706
0
872
2,535
Percent of
Total
2.16
9.47
0.01
0.98
0.00
0.68
1.27
Range
Number of Persons
Exposed
2,736 - 18,662
7,284 - 27,127
0- 1,139
3,881 - 11,406
0- 0
133 - 1,794
381 - 4,924
Percent
of Total
0.58 - 3.95
3.63 - 13.51
0.00 - 0.14
0.50 - 1.46
0.00 - 0.00
0.10 - 1.40
0.19 - 2.48
-------
TABLE 7. ESTIMATES OF ONE-HOUR MAXIMUM DOSAGE EXPOSURES EXPERIENCED
BY OUTDOOR CHILDREN UNDER THE 8H10EX-80 SCENARIO DURING WHICH OZONE
CONCENTRATION EXCEEDED 0.12 ppm AND EVR" EQUALED OR EXCEEDED 30 LITERS MIN'1 M2
Statistic11
Mean Estimate of the Number of Outdoor Children
Percent of Total Outdoor Children Population
Range in this percentage for 10 runs
Mean Estimate of Person-Occurrences
Percent of Total Person-Occurrences
Range in this percentage for 10 runs
Mean Estimate of Occurrences/Person Exposed
Percentage exposed for indicated number of days
1 Day
2 Days
>2 Days
Study area
Chicago
806
0.17
0.00 - 1.23
806
c
0.00 - 0.01
1.00
100.00
0.00
0.00
Houston
1,731
0.86
0.00 - 2.06
1,924
c
0.00 - 0.01
1.11
89.40
10.60
0.00
Los Angeles
1,200
0.15
0.00 - 0.51
1,200
c
d
1.00
100.00
0.00
0.00
"Equivalent ventilation rate = (ventilation rate)/(body surface area).
bMean or range for 10 runs of pNEM/O3.
°Less than 0.01 percent.
JA1I values less than 0.01 percent.
-------
TABLE 8. ESTIMATES OF ONE-HOUR MAXIMUM DOSAGE EXPOSURES EXPERIENCED
BY OUTDOOR CHILDREN UNDER THE 8H10EX-80 SCENARIO DURING WHICH OZONE
CONCENTRATION EXCEEDED 0.12 ppm AND EVR" EQUALED OR EXCEEDED 30 LITERS- MIN ' M2
Statislicb
Mean Estimate of the Number of Outdoor Children
Percent of Total Outdoor Children Population
Range in this percentage for 10 runs
Mean Estimate of Person-Occurrences
Percent. of Total Person-Occurrences
Range in this percentage for 10 runs
Mean Estimate of Occurrences/Person Exposed
Percentage exposed for indicated number of days
1 Day
2 Days
>2 Days
Study Area
New York
0
0
-
-
Philadelphia
0
0
-
-
St. Louis
85
0.07
0.00 - 0.55
85
c
d
1.00
100.00
0.00
0.00
Washington, DC
0
0
-
-
-n
"Equivalent ventilation rale = (ventilation rale)/(body surface area).
bMean or range for 10 runs of pNEM/01.
cLess than 0.01 percent.
''All values less than 0.01 percent.
-------
TABLE 9. ESTIMATES OF EIGHT-HOUR MAXIMUM DOSAGE EXPOSURES EXPERIENCED
BY OUTDOOR CHILDREN UNDER THE 8H10EX-80 REGULATORY SCENARIO DURING WHICH OZONE
CONCENTRATION EXCEEDED 0.08 ppm AND EVR» RANGED FROM 13 LITERS-MIN'1 M2 TO
11 LITERS '
Statistic11
Mean Estimate of the Number of Outdoor Children
Percent of Total Outdoor Children Population
Range in this percentage for 10 runs
Mean Estimate of Person-Occurrences
Percent of Total Person-Occurrences
Range in this percentage for 10 runs
Mean Estimate of Occurrences/Person Exposed
Percentage exposed for indicated number of days
1 Day
2 Days
3 Days
>3 Days
Study Area
Chicago
80,968
17.13
13.22 -21.52
94,630
0.09
0.08 - 0.12
1.17
84.44
13.73
1.83
0.00
============
Houston
40,022
19.93
14.60 -28.41
49,775
0.07
0.04 - 0.09
1.24
79.86
16.53
3.21
0.40
===========================B====
Los Angeles
25,566
3.20
2.50 -.3.92
32,992
0.01
0.01 - 0.02
1.29
76.77
18 83
1 L> . O _J
3 20
J . ±~ \J
1.20
=============================
jn
"Equivalent ventilation rate = (ventilation rate)/(body surface area).
bMean or range for 10 runs of pNEM/O3.
-------
TABLE 10. ESTIMATES OF EIGHT-HOUR MAXIMUM DOSAGE EXPOSURES EXPERIENCED
BY OUTDOOR CHILDREN UNDER THE 8H10EX-80 SCENARIO DURING WHICH OZONE
CONCENTRATION EXCEEDED 0.08 ppm AND EVR' RANGED FROM 13 LITERS MIN ' M2 TO
27 LITERS-
Statistic1"
Mean Estimate of the Number of Outdoor Children
Percent of Total Outdoor Children Population
Range in this percentage for 10 runs
Mean Estimate of Person-Occurrences
Percent of Total Person-Occurrences
Range in this percentage for 10 runs
Mean Estimate of Occurrences/Person Exposed
Percentage exposed for indicated number of days
1 Day
2 Days
3 Days
>3 Days
Study Area
New York
55,012
7.03
5.89 - 8.34
69,198
0.04
0.03 - 0.05
1.26
77.82
18.54
3.53
0.10
Philadelphia
27,866
10.12
6.93 - 14.92
32,216
0.05
0.04 - 0.08
1.16
84.63
14.17
1.20
0.00
Si. Louis
11,354
8.85
5.24 - 12.45
12,914
0.05
0.03 - 0.06
1.14
86.91
11.73
1.35
0.00
Washington, DC
27,087
13.62
9.75 - 19.51
31,118
0.07
0.06 -0.10
1.15
85.37
13.45
1.18
0.00
I
CO
"Equivalent ventilation rate = (ventilation rate)/(body surface area).
bMean or range for 10 runs of pNEM/O3.
-------
Mr. Harvey Richmond 13 February 23, 1996
Table 4 lists exposure estimates for the number and percent of outdoor children
experiencing one or more eight-hour daily maximum ozone exposures above 60 ppb at any
ventilation rate. These results are comparable to the estimates in Table 51 of the outdoor
children project report. For each study area, the 8H10EX-80 estimates in Table 4 fall
between the estimates for 8H5EX-80 and 8H5EX-90 in Table 51.
The pattern holds for Tables 5 and 6. In both tables, the exposure estimates for the
8H10EX-80 scenario fall between the corresponding estimates for 8H5EX-80 and 8H5EX-
90 in Section 7 of the project report for outdoor children.
Each ozone exposure estimated by pNEM/03 includes a value for ozone concentration and
a value for equivalent ventilation .rate (EVR). The product of ozone concentration and EVR
provides an indication of ozone dose. The "daily maximum dose" is assumed to occur each
day during the period when this product is highest. Consistent with this concept, pNEM/O3
provides dose estimates for two averaging times: the one-hour maximum daily dose and
the eight-hour daily maximum dose. Analysts have previously evaluated two specific dose
indicators:
o The number of outdoor children who experienced one or more one-hour
maximum daily dosage exposures during which the ozone concentration
exceeded 0.12 ppm (120 ppb) and the EVR equalled or exceeded 30 liters
min'1 m"2.
o The number of outdoor children who experienced one or more eight-hour
maximum daily dosage exposures during which the ozone concentration
exceeded 0.08 ppm (80 ppb) and the EVR ranged from 13 liters min"' m"2 to
27 liters min'1 m"2.
Tables 7 and 8 provide exposure estimates for the first of these two exposure indicators.
Exposure estimates for the second exposure indicator are presented in Tables 9 and 10.
When the one-hour dose estimates in Tables 7 and 8 for the 8H10EX-80 scenario are
compared with similar estimates for other scenarios in the project report, the 8H10EX-80
values are found to always equal or exceed the 8H5EX-80 estimates. The 8H10EX-80
estimates are less than the corresponding 8H5EX-90 estimates for all study areas except
Houston. A similar evaluation of the eight-hour dose estimates in Tables 9 and 10
indicates that the 8H10EX-80 values fall between the corresponding 8H5EX-80 and
8H5EX-90 estimates for all seven study areas.
The overall pattern of results indicates that ozone exposures expected under the 8H10EX-80
scenario always exceed those of the 8H5EX-80 scenario and almost always are less than
those under the 8H5EX-90 scenario.
F-14
-------
Mr. Harvey Richmond
14 February 23, 1996
I hope that you find these results useful. Please call me if you have any questions or
comments.
Sincerely,
IT Corporation
Ted Johnson
cc: J. Capel
J. Mozier
T. Palma
F-15
------- |