EPA-903/9-78-003
June 1978
PHILADELPHIA
PARTICULATE
STUDY
U.S. Environments' Prorcc^an
Region III irfo.ioitisa Reawao*
Center (3PM52)
841 Chestnut Strut
Philadelphia, PA 13167
EPA Report Collection
Information Resource Center
US EPA Region 3
Philadelphia, PA 19107
U.S. ENVIRONMENTAL PROTECTION AGENCY
Region III
Air Programs Branch
Philadelphia, Pennsylvania 19106
I
-------
GCA-TR-78-02-G
PHILADELPHIA PARTICULATE STUDY
Final Report
by
Frank A. Record
Robert M. Bradway
GCA CORPORATION
GCA/Technology Division
Bedford, Massachusetts 01730
June 1978
Contract No. 68-02-2345
Project Officer
William E. Belanger
U.S. ENVIRONMENTAL PROTECTION AGENCY
Region III
Air Programs Branch
Philadelphia, Pennsylvania 19106
-------
DISCLAIMER
This Final Report was furnished to the U.S. Environmental Protection
Agency by GCA Corporation, GCA/Technology Division, Bedford, Massachusetts
01730, in fulfillment of Contract No. 68-02-2345. The opinions, findings,
and conclusions expressed are those of the authors and not necessarily those
of the Environmental Protection Agency or of cooperating agencies. Mention
of company or product names is not to be considered as an endorsement by the
Environmental Protection Agency.
-------
ABSTRACT
This report presents the results of a study to increase understanding of
the principal sources of particulates within the City of Philadelphia, their
dispersion, and the feasibility of their control. The approach was twofold.
One approach involved field experiments to measure the influence of specific
sources through the deployment of hi-vols and other sampling equipment. Most
of this effort was used to investigate the contribution of traffic-related
emissions; some measurements were also made in the vicinity of major construc-
tion (earth-moving) activity. The second approach involved the use of dif-
fusion modeling techniques to calculate the contributions of major categories
of particulate emissions to particulate loadings at th6 various monitoring
sites. The principal modeling tool used in calculating the contribution of
source categories throughout the city was the Air Quality Display Model (AQDM).
A test was also conducted to measure the effectiveness of street washing in
reducing ambient particulate levels.
In the field studies to help define the spatial impact of traffic-related
emissions, measurement were made with conventional hi-vols at trailer-top and
at three heights on an 12.2-meter tower near an intersection, and at rooftop
level (11 meters) on either side of the main street at distances from 15 to
60 meters. Two specialized instruments were used to discriminate between
respirable and nonrespirable particles. One, the EPA Dichotomous Sampler,
also allowed for detailed chemical analysis of the collected particles. The
other, the GCA Ambient Particulate Monitor, determined the short-term (e.g.,
1-hour) changes in respirable and total particulate concentrations. A frac-
tionating head also provided some data on particle size distributions. Auto-
matic traffic counters provided continuous data throughout the sampling
periods.
111
-------
CONTENTS
Abstract iii
Figures vii
Tables xi
Acknowledgment xiii
1. Introduction 1
The Problem 1
Approach 4
Major Findings 7
2. Large Scale Emissions and Meteorological Features of Study Area . 9
Emissions 9
Climatology 9
3. Analysis of Monitoring Network Data 15
AMS Monitoring Network 15
Spatial Variation of TSP Concentrations 17
Seasonal Variation of TSP Concentrations 21
Interrelationships Among Sites 21
Day of the Week and ^recipitation Effects 30
Comparison of Daily TSP Concentrations at SBR and 500 Under
Four Regimes 32
Concentrations of Metals and Sulfates 34
4. Field Study Results 44
Observations Near Traffic Activity 45
Observations Near Construction Activity 74
Analysis of Selected Hi-Vol Filters 76
5. Model Development 100
Treatment of Fugitive Dust in Model Calculations 100
Results of Air Quality Display Model Calculations Ill
6. Summary and Conclusions 127
Nature of the Philadelphia Particulate Problem 127
General Features of TSP Behavior 128
Spatial Distribution of Particulates Near Street Sources . . 129
Particle Sizes 129
Chemical Composition of Particulates 131
Effect of Street Flushing on Ambient Particulate Levels. . . 131
Modeling 132
Prospects for Meeting the NAAQS 132
An Area for Continued Research 134
References 135
-------
CONTENTS (continued)
Appendices
A. Monitoring Site Descriptions and Nearby Traffic Data A-l
B. TSP Correlation Analyses B-l
C. Further Details of the Effect of Precipitation C-l
D. Monthly Average Concentrations of Metals and Sulfates D-l
E. Special Instrumentation E-l
F. February TSP and Lead Concentration F-l
G. February Traffic Volumes G-l
H. Street Flushing Procedure H-l
I. June Traffic Volumes 1-1
J. June TSP and Lead Measurements J-l
K. Particle Size Distributions for Selected Hi-Vol Filters. . . . K-l
L. Lead Emissions From Automotive Sources L-l
vi
-------
FIGURES
Number Page
1 Trend in annual mean TSP concentrations in Philadelphia since
1956 2
2 Annual mean TSP concentrations at two sites in Philadelphia . . 3
3 Metropolitan Philadelphia Interstate Air Quality Control Region 10
4 Location and emission rate of all plants emitting particulate
(1974 emission inventory) 11
5 Sketch of particulate emission density isopleths of 5 and 25
tons/yr/sq. km for the Metropolitan Philadelphia AQCR (ex-
cluding fugitive emissions) 12
6 Wind direction roses based on 4 years of Philadelphia Inter-
national Airport data 14
7 Location of Philadelphia monitoring sites 16
8 Annual geometric mean TSP concentrations in Philadelphia in
1976. Units are yg/m3 20
9 Trend curve at AMS Laboratory superimposed on daily TSP
concentrations 24
10 Average of correlation coefficients between TSP concentrations
at indicated station and all other stations 28
11 Square of the correlation coefficient between TSP levels as a
function of distance, 1976 29
12 Average weekly variation in TSP levels at the AMS Laboratory in
Philadelphia for two precipitation categories during 1974 and
1975. Number of observations for each data point is in
parentheses 31
13 Ratio of TSP concentration at Broad and Spruce to that at 500
South Broad as a function of the concentration at 500 South
Broad 35
vii
-------
FIGURES (continued)
Number Page
14 Relationship between TSP concentrations at Broad and Spruce and
500 South Broad. (Period of observations: September 1974 to
15
16
17
18
19
Relationship between TSP concentrations at Broad and Spruce and
Relationship between TSP concentrations at Broad and Spruce and
500 South Broad on weekdays with precipitation >_ 0.10 in. . . .
Relationship between TSP concentrations at Broad and Spruce and
Relationship between TSP concentrations at Broad and Spruce and
500 South Broad on weekends with precipitation > 0.10 in. . . .
Local neighborhood at Broad and Spruce Streets showing locations
of AMS trailer and monitoring tower
37
38
39
40
46
20 Local neighborhood at 500 South Broad Street showing locations
of high-volume samplers 47
21 Normalized particulate concentrations observed at Philadelphia
site. TSP concentrations above and Pb concentrations in
parentheses below. Units are yg/m3 49
22 Time sequence of TSP and Pb concentrations from 0600 to 2400 EST
at tower site on 17 February 1977 51
23 Total and respirable suspended particulate concentrations measured
by the APM at the tower site during 5 to 14 February 1977 ... 52
24 Traffic volume near tower site during 5 to 14 February 1977 ... 53
25 Comparison of total and respirable suspended particulate concen-
trations measured by the APM at the tower site during 5 to 14
February 1977 54
26 Comparison of APM and hi-vol concentrations (5 to 14 February
1977) 56
27 Respirable/nonrespirable particulate concentration, found by EPA
Dichotomous Sampler, and accompanying meteorological conditions.
Shaded part of bar denotes respirable fraction 58
Vlll
-------
FIGURES (continued)
Number Page
28 Percent of elemental mass found in respirable particles. Aver-
age values for Broad and Spruce are shown at the left of the
vertical dashed line, and average values for 500 South Broad
are shown at the right 62
29 Comparison of dichotomous sampler and hi-vol concentrations
(10 February to 12 March 1977) 66
30 Area of street-flushing experiment 68
31 TSP concentrations and meteorological data for 7 June to 20 June
field program 70
32 Comparison of TSP concentration and traffic volume at Broad and
Spruce Streets 72
33 PCOP-1, 6/16/77 78
34 PCOP-1, 6/16/77 78
35 PCOP-1, 6/16/77 79
36 Tower-A, 6/17/77 79
37 Tower-A, 6/17/77 80
38 Puddle sample, 2000X 80
39 EDAX photograph showing Al, Si, K, Ca, Fe and Ti 81
40 Puddle sample - organic mat (1 mm = 1 pm) 81
41 Typical fly ash particle which contains (by EDAX) major S, Ca;
minor Fe and Zn; low Ti; and low Si and Al. Scale 2 mm = 1 ym 82
42a An opaque particle at 2000X magnification from 500 SB-1, 6/14/77 84
42b EDAX spectrum from the particle in Figure 45a shows peaks for Al,
Si, K, Ca, Fe, and Au peaks from the sample-coating 84
43a An opaque particle at 2000X from TRHV-1, 6/14/77 85
43b EDAX spectrum from th'e particle shows peaks for Al, Si, K, Ca,
Fe (and Au peaks from the sample-coating) 85
44a An opaque particle at 2000X magnification from 500 SB-1, 6/18/77 86
ix
-------
FIGURES (continued)
Number Page
44b EDAX spectrum from the particle shows peaks for Al, Si, K, Ca,
Fe (and Au peaks from the sample-coating) 86
45a An opaque particle at 2000X magnification from Tower-B, 6/18/77 87
45b EDAX spectrum from the particle shows peaks for Al, Si, K, Ca,
Fe (and Au peaks from the sample-coating) 87
46 EDAX spectra for overall filter. In all three cases the peaks
are for Al, Si, K, Ca, Fe (and peaks from the sample-coating).
Spectrum for Tower-B sample was not included and would not be
representative for overall filter composition due to slight
difference in sample preparative coating 88
47 TSP calibration for the New Castle and Philadelphia monitors.
Philadelphia monitors are indicated by their usual identifiers 101
48 Comparison of calculated and observed vertical TSP profiles on
8 February 1977 105
49 Excess TSP concentration at Broad and Spruce above concentra-
tion at 500 South Broau as a function of wind direction . . . 109
50 Annual average TSP concentration contributed by large point
sources, ug/m3 112
51 Annual average TSP concentration contributed by area and small
point sources (nonvehicular), yg/m3 113
52 Annual average TSP concentration contributed by vehicles (tail-
pipe and tire wear only) based on assumed emission rates of
0.21 and 0.12 g/km-vehicle, respectively, ug/m3 114
53 Annual average TSP concentration contributed by reentrained
roadway dust based on an assumed emission rate of 0.42
g/km-vehicle, ug/m3 115
54 Annual average TSP concentration contributed by all inventoried
sources, ug/m3 116
55 Source contributions to annual geometric means at the 500 South
Broad, and Broad and Spruce Street sites in 1976 121
56 Allocation of TSP concentrations to source categories at 13
Philadelphia monitoring sites ..... 122
57 Average annual lead concentration based on an emission rate
of 0.016 g/km-vehicle, ug/m3 126
-------
TABLES
Number Page
1 Annual TSP Geometric Means for 1974, 1975, and 1976 at 13 Phila-
delphia Monitoring Stations 18
2 Summary of TSP NAAQS Violations of the 24-Hour Standards From
January 1974 to June 1977 19
3 Percent of Observations Greater Than 24-Hour Secondary Standard . 22
4 1976 Monthly Geometric Mean TSP Concentrations, yg/m3 23
5 Linear Correlation Coefficients Between 24-Hour TSP Levels at
Philadelphia Monitoring Stations in 1974, 1975, and 1976. ... 25
6 Frequency Distribution of Correlation Coefficients Between TSP
Concentrations at Indicated Station and Remaining 12 Stations
Using 1976 Data 26
7 Average Weekday and Weekend Concentrations for Two Preicipitation
Classes 30
8 Average TSP Concentration Data at the Broad and Spruce and 500
South Broad Sites by Day-of-the-Week and Precipitation Category 33
9 Average Monthly Concentration of Metals and Sulfates in yg/m3 . . 41
10 Monthly Average Lead Concentrations at AFS, SBR and NBR 42
11 Average Results From Elemental Analysis Performed by EPA. Units
12
13
14
15
16
Element Listings Ordered by Concentrations and Concentration
Atomic Ratios, Normalized With Respect to Si .........
Approximate Weights and Percentages of Phases Present
Summary of TSP Concentrations for June Field Program
60
63
64
65
69
xi
-------
TABLES (continued)
Number Page
17 Average Vertical Profile of TSP Concentrations During June Field
Program 71
18 TSP and Meteorological Observations for Construction Activity
Study 75
19 Ranking of Hi-Vol Filters on Arbitrary Scale of Greyness 90
20 Analysis of Hi-Vol Filters by Petrographic Microscopy 91
21 Mean Diameter, Standard Deviation and Number of Particles Counted
for Nonopaque (Mineral) and Opaque Hi-Vol Filter Phases .... 92
22 Results of Benzene Extractions of Selected Filters 94
23 Summary of Infrared Analysis 95
24 Mass Median Diameters of Selected Particulate Samples in ym . . . 98
25 Size Distribution of Opaque and Nonopaque Particles (ym) Nor-
malized to 25 cm2 99
26 Box-Model Estimates of Annual Traffic-Related Contributions to
Particulate Levels at Tower Site as a Function of Wind
Direction 107
27 Correlation Coefficients Between ATSP (Broad & Spruce - 500 S.
Broad) and Wind Speed 108
28 Calculated Annual TSP Concentrations and Contributions by Source
Category for Philadelphia Area in 1974 in yg/m3 117
29 Model Allocation of TSP Concentrations to Source Categories . . . 120
30 Estimated Impact of Local Traffic and Uninventoried Sources . . . 124
31 Comparison of Dichotomous Sampler and Hi-Vol Results 130
xii
-------
ACKNOWLEDGMENT
We wish to express our appreciation to Mr. Robert K. Stevens of the En-
vironmental Services Research Laboratory, U.S. EPA, Research Triangle Park,
North Carolina, who arranged for the use of the dichotomous samplers in the
field program, and for the analysis of the exposed filters, and to the NASA
Lewis Research Center for the loan of the NASA Air Scout. We would also like
to thank members of the Philadelphia Air Management Services Laboratory and
of the City Sanitation Department for their cooperation during the program.
Very special thanks are due Mr. William E. Belanger, Project Officer,
EPA Region III, for his stimulating and considerate guidance throughout the
program, and for his active participation in obtaining and operating special
equipment during the field experiments.
Finally, we express our gratitude to all members of the GCA/Technology
Division staff who worked on the various phases of the program. Major con-
tributions were made by Mr. Victor Corbin, who carried out the computer mo-
deling, by Dr. Charles Spooner, who was responsible for the microscopic anal-
yses of selected filters, by Mr Gordon Deane, who prepared the site descrip-
tions, and by Mr. Theodore Midurski, who was responsible for the traffic data.
We particularly wish to thank the members of the staff who worked on the field
and laboratory phases of the program. The success of the field experiments
could be assured only by the careful operation of the sampling equipment.
Day-to-day operation by the equipment was the principal responsibility of
Mr. Dan Miesse during the February experiments, and of Mr. Lyle Powers during
the June experiments.
xili
-------
SECTION 1
INTRODUCTION
THE PROBLEM
The underlying problem giving rise to the need for the work described in
this report is well illustrated by Figures 1 and 2. Figure 1, prepared by
the City of Philadelphia Air Management Services, shows the striking decrease
of roughly 150 yg/m3 in citywide average TSP levels that has been brought
about in recent years by the control of emissions from major industrial, fuel
combustion, and solid waste disposal sources. However, the citywide trend
curve has now leveled off at about the primary annual standard of 75 yg/m3.
Any additional improvement in air quality awaits a further reduction in par-
ticulate emissions, which is likely to be achieved only as a result of new or
more stringent control measures.
Impressive though this average citywide decrease has been, however, even
the primary national standards are not met uniformly throughout the city. Vio-
lations occur most often in dost, proximity to industrial, construction and
street sources. Of these, the street sources, which are ever present and con-
centrated in central urban areas, have received the most attention in this
study.
Figure 2 depicts the trend in annual mean TSP levels at the downtown,
500 South Broad Street site since 1960. This center-city site, which is at
a height of approximately 10.7 meters (35 feet) on top of a three-story
building at the southern edge of the Central Business District, shows a de-
crease of about 55 yg/m3 in the annual mean between 1966 and 1972 and now
approximates the primary standard. This reduction presumably is a reflection
of decreasing emissions from traditional sources. Figure 2 also shows annual
means for the last 3 years at a neighboring trailer-top site near the inter-
section of Broad and Spruce Streets. Here, at a location strategically se-
lected to measure the impact of street level sources, the standard is exceeded
by 43 yg/m3.
There are two fundamental points illustrated by Figure 2. First, the
apparent attainment or nonattainment of the primary annual particulate standard
within this small urban area which is removed from the direct impact of any ma-
jor traditional source depends simply upon the location of the monitor. Second,
the excess of 43 yg/m3 measured at Broad and Spruce represents the impact of
street level sources. Additionally, it is important to realize that these par-
ticulates are probably not equivalent either in physical characteristics and
-------
I
CO
ec
u
o
i
o
8161
9161
2161
0161
8961
9961
CO V4
OJ
Cfl C
•H O
CU
13
Ctf
^H
•H
J3
PM
O CS
U (U
•U O
C H -H
CU
O JJ
•H ti >w
4J rt o
(8 4J
l-l CO 4J
•U -H C
C w cu
cu w a
o , o,
H CU
P4 iH Q
W -H
H CU ct)
B! -H
tO 0
0) tO
e -H
rt -H iH
itS
tO rd
C T3 CO
•H 0) CU
•H O
t) i-f -rl
QJ & tJ
>-i 3 CU
E-* CO C/l
.uj/6T/'NOllvaiN30NOO
-------
ro
o
CC
H
Ul
200
180
160
140
120
100
o
g 80
_j
z 60
40
20
0
I I I T I
BROAD and SPRUCE-
500 SOUTH BROAD ST.
PRIMARY STANDARD
1 - 1
60 62 64
1 -- 1 - 1 - 1 - 1 - 1 - 1 - ' ' '
66 68
YEAR
70 72 74 76
Figure 2. Annual mean TSP concentrations at two
sites in Philadelphia.
-------
chemical composition, or in their impact upon public health, to an equal con-
centration of the mix of particulates that was measured at rooftop level during
the early and mid-sixties.1
Observations such as these, plus violations of the standards near areas
containing unpaved parking lots, dirt roads, storage piles and other sources
of fugitive particulate emissions have raised a number of issues ranging from
the proper siting of hi-vols to determine compliance with the present particu-
late standards, to more fundamental questions concerning the possible need for
standards which include specifications as to particle sizes (e.g., percent
respirable), and perhaps chemical composition as well. Regardless of the
eventual disposition of such far ranging issues, however, the current require-
ment is to bring about an additional reduction in particulate emissions which
will result in the achievement of the existing standards. The rational selec-
tion and implementation of control plans to bring about this reduction clearly
will require sound estimates of the relative contribution of the various
sources of particulates within the problem area. This is a widely recognized
need in many urban areas.
In this study, the specific TSP problem of the City of Philadelphia has
been addressed by first apportioning the measured particulate loading of the
air to major source categories, and then assessing the feasibility of achieving
the standards through additional control measures.
In an attempt to achieve the maximum benefit from the study, the work was
occasionally redirected during the course of the project and a number of side
issues were investigated. When t^ese side investigations appear to increase
understanding of some phase of the TSP problem, the results are included in
this report. Of particular importance is information dealing with the respirable
fraction of particulates and its chemical composition.
APPROACH
Program goals were pursued through an analysis of historical data, the
collection and analysis of special field data, and by modeling techniques, with
the effort formally being broken down into the following five tasks.
Task 1 - Literature Search and Feasibility Study
The current state of knowledge in the areas of particulate emissions from
nonconventional sources, particle identification technology, and particulate
dispersion modeling was reviewed under this task, and used in the design of
various aspects of the study. An annotated bibliography of selected articles
was provided the Project Officer at the conclusion of this task.
Task 2 - Emission Inventory
The emission inventory for particulates within the Philadelphia AQCR was
reviewed and upgraded by GCA/Technology Division under another contract
(No. 68-02-1376, Task Order No. 24). The major thrust of Task 2 was to provide
better estimates of emissions from nonconventional sources, particularly those
-------
related to traffic and construction activities. The results are reported in
Section 5 of this report where they have been incorporated into the recommended
model for metropolitan Philadelphia.
Task 3 - Ambient Air Analysis
TSP concentrations measured by the AMS Laboratory monitoring network
from January 1974 through 1977 were used to help define the extent and nature
of the Philadelphia particulate problem and to aid in the design of a filter
analysis program and in the selection of sites for special field studies.
During this period, the city TSP monitoring network comprised from 10 to 13
stations. The analysis of the AMS network data is presented in Section 3.
%
Task 4 - Collection and Analysis of Field Data
Two types of field data were collected during the program. One consisted
of 24-hour concentrations of eight metals (Pb, Fe, Mn, Cu, Ni, Cd, Mg and
Al) measured throughout the city. To obtain these data, Spectrograde filters
were supplied to the AMS Laboratory and exposed routinely in the AMS hi-vol
network. All analyses of these filters were carried out by the Laboratory.
The results of these analyses are discussed in Section 3.
The other type of field data was obtained from special field experiments
designed to measure the impact of specific source types on TSP levels. Two
types of sources were investigated, one being major construction activity and
the other being heavily traveled urban streets. The site selected at which
to study the impact of construct!^ was the Southwest Sewage Treatment Plant
where extensive earth moving and related activity connected with the expansion
of the plant were taking place. For this study, conventional hi-vols were
used.
To study the impact of vehicle-related sources on TSP levels, two neighbor-
ing sites on South Broad Street were selected. One was at the intersection
of Broad and Spruce Streets, and the other was at rooftop level three blocks
to the south. The basic configuration of instruments for the street-source
experiments consisted of pairs of hi-vols at three heights on a 12.2 meter
(40 foot) tower adjacent to the AMS Laboratory trailer at the intersection of
Broad and Spruce Streets; a pair of hi-vols on the roof of the AMS trailer;
and five hi-vols at rooftop height (about 10 meters) along a line crossing *
Broad Street and perpendicular to it. In addition, an EPA Dichotomous Sampler
was used at two heights to measure respirable particles and particles in the
range of 3.5 \im to 20 ym aerodynamic diameter.
All samples collected by the Dichotomous Sampler underwent elemental
analysis by EPA. During part of the experiment, a GCA Ambient Particulate
Monitor was used to measure respirable and total particulate concentrations
*
The Dichotomous Sampler was made available for project use by Robert K.
Stevens of The Environmental Services Research Laboratory, U.S. EPA, Research
Triangle Park, North Carolina.
-------
over sampling periods of about 1 hour. Also, a NASA Air Scout designed to
collect airborne particulates as a function of wind direction, was field
tested during the experiment. Traffic counts and meteorological data were
recorded throughout the measurement program. Data obtained with the various
instruments were pooled to establish an average three-dimensional distribution
of particulates, including particle size information, at the study site.
Although the impact of street dust on TSP levels is now generally recog-
nized, at least qualitatively, feasible strategies for reducing the magnitude
of its impact are largely untested. During the measurement program, in what
is believed to be the first major test of the effectiveness of street flushing,
over 100,000 gallons of water were applied to streets in the vicinity of the
Broad and Spruce Street monitors on 3 consecutive days.
A more complete description of the field experiments and the results are
presented in Section 4.
Task 5 - Modeling
The chief purpose of this task was to break down by major source category
the TSP levels which are observed at the various monitoring sites. The prin-
cipal tool used was the Air Quality Display Model (AQDM). Before running this
urban dispersion model, however, the results of field experiments at the Broad
Street sites were tested in line-source models to help select an emission rate
for vehicle-related fugitive dust that would be appropriate for use in the
AQDM.
The most promising procedure tested is to use the urban diffusion model
to calculate "rooftop" concentrations (i.e., concentrations at heights or lo-
cations largely uninfluenced by local fugitive dust sources) and then to adjust
these model concentrations for nearby street-level or other fugitive sources.
In principle, these adjustments can be made either on the basis of empirically-
derived expressions or, if suitable emission factors become available, from
model calculations.
The source categories used in apportioning the TSP levels were:
• Background and uninventoried sources
• Large point sources
• Area and small point sources (excluding motor vehicles)
• Tailpipe and tire wear
• Reentrained dust from paved roads
• Local street emissions
The results of this work are presented in Section 5.
*
The Air Scout was made available by the NASA Lewis Research Center in
Cleveland.
-------
MAJOR FINDINGS
The results of this study confirm that vehicle-related particulates and
other fugitive dust sources impact severely on the air quality of Metropolitan
Philadelphia. Estimates of street-level contributions to annual average TSP
levels measured at an adjacent trailer-top monitor were found to be in excess
of 40 yg/m3. The contribution of vehicle-related emissions to citywide par-
ticulate loadings (rooftop concentrations) could not be accurately separated
from background concentrations but was estimated to be of the order of
20 yg/m3 within the central urban area, with slightly more than half being
from reentrained street dust. Somewhat greater contributions from reentrained
dust are believed to be generated by the same volume of traffic in the more
industrial areas of the city.
Although no attempt was made to evaluate the effectiveness of specific
citywide control strategies, an analysis of source contributions lead to the
following conclusions. Relatively small reductions in any of several source
types could result in the achievement of the primary standard at downtown
rooftop monitors outisde of the principal industrial areas. On the other hand,
substantial additional reductions in emissions will be required to meet the
primary standard at trailer-top height in either commercial or industrial
areas, and at rooftop level in industrial areas. At several sites these re-
ductions must clearly come from vehicle-related sources.
After a 3-day test to determine the effect of intensive street flushing
on TSP levels, concentrations in the vicinity rose to levels approximately
100 yg/m3 higher than expected f -om observations at other locations within the
city.
Concentrations measured concurrently with the EPA Dichotomous Sampler
and with the standard hi-vol made possible an important comparison between the
mass loadings of respirable and larger particles at two heights. At rooftop
level (11 meters) approximately 32 percent of the mass collected by the hi-vol
was composed of respirable particles, 13 percent of particles between 3.5 and
20 ym, and 55 percent of particles greater than 20 y. At the trailer top, the
distribution was 23 percent respirable, 15 percent between 3.5 and 20 ym, and
62 percent greater than 20 ym.
An elemental analysis of particulates collected by the dichotomous sampler
(i.e., particles < 20 ym) showed that the six elements having the highest con-
centration were, in order of decreasing concentration: sulfur, silicon, cal-
cium, lead, iron and aluminum. Of these, sulfur and lead are found principally
in the respirable fraction.
The EPA proposed lead standard of 1.5 yg/m3, based on a monthly average,
was exceeded at three sites within the city. Two of these sites were traffic-
oriented, and the third was near an industrial source. The phase-down schedule
for maximum pooled average lead content of gasoline indicates that the pro-
posed standard will be met at the two traffic-oriented sites in 1978.
-------
The selection of appropriate emission factors for fugitive and reentrained
dust continues to be a major constraint to the successful modeling of particu-
lates. A modeling approach that holds promise is to accept an emission factor
for general use that yields consistently reasonable results throughout the
area for rooftop monitors and for monitors removed from any obvious local par-
ticulate source. Major deviations from these model estimates are then attri-
butable to local sources.
A fuller discussion of the results of the study is provided in Section 6.
-------
SECTION 2
LARGE SCALE EMISSIONS AND METEOROLOGICAL
FEATURES OF STUDY AREA
The City of Philadelphia lies near the center of the Metropolitan
Philadelphia Interstate Air Quality Control Region (AQCR) as shown in
Figure 3. In addition to Philadelphia, which is the fourth largest
Standard Metropolitan Statistical Area in the country, the AQCR also con-
tains Trenton, New Jersey and Wilmington, Delaware, with populations of
one-third and one-half million, respectively. Because of the possibility
of significant particulate transport from such large neighboring population
and industrial centers, emissions from the entire AQCR were included in the
Air Quality Display Model when calculating Philadelphia TSP levels. The
next paragraphs discuss the distribution of particulate emissions within
the AQCR and climatological factors affecting their transport and ambient
concentration.
EMISSIONS
Under Contract No. 68-02-1376, Task Order No. 24, GCA updated the TSP
point source emission inventory for the Metropolitan Philadelphia Inter-
state AQCR and developed a regional TSP area source emission inventory.
Details of this work can be found in the final report for that project,2
but a general idea of the distribution of particulate emissions can be
gained from Figures 4 and 5. Figure 4 shows the location of TSP point
source emissions, and Figure 5 is a sketch of emission density isopleths
of 5 and 25 tons per year per square kilometer. Figure 4 shows a very high
concentration of point sources in Philadelphia and along the Delaware River
and, in addition, numerous sources to the west and north-northwest of the
city. Figure 5 shows somewhat the same geographical distribution of area
source emissions, with a long extension of the 25 ton/yr/sq. km isopleth
southwestward along the Delaware River. It will be seen from the discussion
of prevailing wind directions which follows that many of these sources which
are external to Philadelphia proper are upwind from the city much of the time.
Thus, fine particulates introduced into the atmosphere from these sources can
contribute to the city TSP burden.
CLIMATOLOGY
The major topographical feature of the AQCR is the Delaware River which
separates the Pennsylvania and Delaware portions of the AQCR from the State
of New Jersey. Precipitation, which acts to cleanse the atmosphere and also
suppress the reentrainment of fugitive dust, averages about 102 centimeters
-------
(9
2
ac
o
4500-
4480-
4460-
4440-
4420-
4400-
398O-
3960-
3940 -
I
420
I I I I I I
460 430 500 520 540 960
KM (EASTING)
Figure 3. Metropolitan Philadelphia Interstate Air Quality Control Region.
10
-------
M MtriNM
Figure 4. Location and emission rate of all plants emitting particulate
(1974 emission inventory).
11
-------
4 »00t
44«0-
4440-
s
8 4420-
ii
4310 -
43(0-
4J«0-f.
400
/ /+*
) - f CMHTt« CO. /
- j/
^J
410
440
4«0 4tO 900
KM (CAtTINO)
540
660
Figure 5. Sketch of particulate emission density isopleths of 5 and 25
tons/yr/sq. km for the Metropolitan Philadelphia AQCR (ex-
cluding fugitive emissions).
12
-------
(40 inches) per year. It is fairly evenly distributed throughout the year
with maximum amounts occurring during the late summer months. In the course
of a year, the region is influenced by numerous low pressure centers as they
move across the country or up the coast. Because of the moderating effect
of the Delaware Bay on temperature, and the heat island effect of the city,
snowfall is minimized.
As can be seen from Figure 6, the prevailing wind direction in summer
is from the southwest. In winter, northwesterly winds prevail but south-
westerly winds are still frequent. On an annual basis, southwesterly direc-
tions are the most frequent, with westerly winds being the second most
frequent.
13
-------
WINTER
SUMMER
ANNUAL
N
PERCENT FREQUENCY
OF OCCURRENCE
0 5 10 15
Figure 6. Wind direction roses based on A years of Philadelphia
International Airport data.
14
-------
SECTION 3
ANALYSIS OF MONITORING NETWORK DATA
TSP concentrations measured routinely by the AMS Laboratory monitoring
network from January 1974 through June 1977 were analyzed to define the ex-
tent of the particulate problem and to describe the behavior of TSP levels
within the city. Specifically, the data were summarized so that TSP concen-
trations and violations of the AAQS within the city could be compared roughly
to the distribution of particulate sources and to differences in monitor
siting. Also, the data were examined for seasonal variations and interrela-
tionships among sites which might shed light on source characteristics. To
assist in the development of empirical relationships of potential use in the
modeling of TSP levels, daily observations at the AMS Laboratory were used to
determine the average reduction experienced on weekends and days closely asso-
ciated with precipitation. Concentrations measured at rooftop level at 500
South Broad Street were also compared to concurrent measurements at trailer-
top level at the intersection of Broad and Spruce Streets to see if the con-
tribution of traffic-related particulates varied significantly by day of the
week or precipitation class.
More limited network data of two types were collected as part of the
project effort. One type comprised a small number of particle size distri-
butions determined by the use of a fractionating head attached to a standard
hi-vol which was rotated within the network. The second type comprised con-
centrations of metals determined by the AMS Laboratory from an analysis of
Spectrograde filters supplied under the contract. Summaries of these data
are included in this section.
AMS MONITORING NETWORK
Figure 7 shows the location of the AMS Laboratory TSP monitors at the
end of the study period. Three of these stations, SBR, AFS, and NBR were
placed in service near the end of 1974. Of these three, SBR and NBR are
traffic oriented, with the monitors being located on the roof of trailers
adjacent to Broad Street, and the third is source oriented, being located
directly to the east of a National Lead plant. A fourth site, FRI, was re-
located from the CAMP station adjacent to the Franklin Institute where the
monitor was at a height of about 4 meters, to the roof of the Franklin
Institute, where it is now at a height of about 15 meters. During a 6-month
period, monitors were operated concurrently at the two locations. Details of
the site locations are provided in Appendix A.
15
-------
4445
4440
4433
4430
X
g 4425
z
4420
44! 5
0 I 2
i I I
MILES
44(0
4405
470
475
480
KM. EASTING
4»5
50C
509
Figure 7. Location of Philadelphia monitoring sites.
16
-------
SPATIAL VARIATION OF TSP CONCENTRATIONS
Annual geometric means experienced at the 13 locations during 1974, 1975,
and 1976 are shown in Table 1. The average of the 1975 and 1976 means are
listed in the right-hand column of the table to provide a best estimate of
current TSP levels. During those 2 years, the maximum difference, which
occurred at the source-specific AFS station, was 14 yg/m3. The average con-
centration measured at the remaining 12 locations remained essentially un-
changed, increasing from 82.1 to 82.8 yg/m3. The annual primary standard of
75 yg/m3 was exceeded at eight of the sites and the annual secondary standard
of 60 yg/m3 was exceeded in at least 1 of the 2 years at all but two of the
sites, Belmont and Roxborough, both of which are in the suburban area west
of the main urban activity. The annual mean at a third site, Northeast
Airport, exceeded the secondary slightly in 1975 but was below the standard
in 1976. This site lies north of the urban activity on the edge of a resi-
dential area.
Table 2 summarizes the number of violations of the primary and secondary
24-hour TSP standards over the January 1974 to June 1977 period. During this
period, the primary standard was violated a total of 16 times. Of these 16
violations, five each occurred at Allegheny and the S.E. Water Pollution Con-
trol Plant; two occurred at the CAMP station; two at 500 South Broad Street,
and one at Broad and Spruce. Also, eight occurred in 1974, seven in 1975,
none in 1976, and one in the first half of 1977. The right-hand column of
Table 2 shows the percent of observations greater than the secondary standard
during the combined 3-1/2 year period. The standard was violated most fre-
quently at the two most traffic-oriented sites; i.e., FRI (CAMP) and the trai-
ler at Broad and Spruce (SBR) where the standard was exceeded 25 and 20 per-
cent of the time, respectively. The next three sites in frequency of viola-
tions were ALL, INT, and NBR, all of which are subject to fugitive/fugitive
dust emissions. These five sites also had the highest annual geometric means,
as can be seen from Table 1. Table 2 shows that the secondary standard was
not violated at either BEL and ROX during any of the 3 years and that only
two exceedances occurred in any one year at N/E.
Figure 8 displays the 1976 annual geometric means geographically. The
principal points to be noted at this time are: (1) high concentrations extend
from the southwestern end of the city through the central urban area, and
(2) the concentration falls off rapidly both to the west and to the northeast
of the city. The rapid decrease to the northeast is of particular interest
since the prevailing winds, as shown in Figure 6, are southwesterly during
much of the year. A trajectory from southwest to northeast not only passes
over the main commercial area of the city and much of the industrial area, but
also over a concentration of point sources to the southwest of the city, as
shown in Figure 4. This rapid decrease upon leaving the city indicates that,
although a substantial background of fine particulates undoubtedly enters the
city as a result of external sources, a major part of the particulate burden
of the city is not only internally generated, but is largely confined to the
source area.
17
-------
TABLE 1. ANNUAL TSP GEOMETRIC MEANS FOR 1974, 1975, AND 1976
AT 13 PHILADELPHIA MONITORING STATIONS
Station
code
DEF
ALL
INT
BEL
ROX
N/E
NBR
FRIS
LAB
SBR
S/E
500
AFS
Name
Defense Supply
Allegheny
International Airport
Belmont
Roxborough
Northeast Airport
North Broad
Franklin Institute
AMS Laboratory
South Broad
Southeast Water Pollu-
tion Control Plant
500 South Broad
Aramlngo Fire Station
Geometric mean (yg/m3)
1974
105
(57)
122
(51)
94
(59)
72
(46)
59
(49)
64
(53)
199
(54)
(330)
-
78
(241)
76
(294)
—
1975
88
(59)
98
(57)
97
(57)
59
(39)
53
(57)
62
(50)
105
(52)
79
(60)
(351)
114
(58)
90
(249)
71
(295)
86
(56)
1976
87
(56)
104
(54)
94
(54)
57
(25)
56
(54)
56
(58)
103
(58)
81
(57)
(355)
117
(58)
90
(251)
74
(289)
100
(54)
Average
(1975-1976)
88
101
95
58
55
59
104
80
72
116
90
73
93
CAMP station in 1974; Rooftop station in 1975, 1976.
Note: Number of pairs is enclosed in parentheses.
18
-------
I
O, r-t
19
-------
Figure 8. Annual geometric mean TSP concentrations in Philadelphia
in 1976. Units are yg/m3.
20
-------
SEASONAL VARIATION OF TSP CONCENTRATIONS
Table 3 was prepared to see if there was any strong seasonal pattern in
the frequency of occurrence of high TSP levels that might relate to source
characteristics. In general, the three most traffic-oriented sites, FRI (CAMP),
SBR, and NBR experience the greatest number of concentrations in excess of the
secondary standard during the winter and early spring, and the least number
during the fall. This pattern is one which could be caused by somewhat dirtier
streets and gustier wind conditions during the winter. However, high concen-
trations were also observed rather frequently at these sites during June and
July. Sites DEF, LAB, and AFS have more clearly defined seasonal patterns,
with the greatest number of high concentrations occurring in the late fall and
winter, and the least number in the summer and early fall. On the other hand,
ALL and INT show little seasonal variation, and S/E had the maximum number of
high concentrations during July and August. Although there are these dif-
ferences among the sites, it is not obvious how they relate to seasonal dif-
ferences in emission rates or source types.
Table 4 and Figure 9 present additional evidence for the lack of any well-
defined seasonal trend in TSP levels. Table 4 lists monthly geometric means
at the 13 sites for 1976 and Figure 9 is a plot of the daily TSP concentrations
measured at the AMS Laboratory during 1974 and 1975 upon which has been super-
imposed a trend curve calculated using the Whittaker Henderson technique3 with
"a" set equal to 20. The trend curve has no outstanding features, but an upper
envelope of the raw data points shows that the highest daily concentrations
occur during the late fall and winter months.
INTERRELATIONSHIPS AMONG SITES
Day-to-day variations in TSP concentrations within the AMS network were
examined to see whether or not concentrations at the various sites appeared
to be responding to the same controlling factors to help resolve the question
as to the part played by local sources and the distances over which such
sources make significant contributions.
To study the amount of common variation in the day-to-day changes in
concentration throughout the network, linear correlation coefficients were
calculated between all possible pairs of sites for each of the 3 years.
Table 5 is a matrix of the resulting coefficients. In 1974, the number of
pairs of observations ranged from 31 to 275, and the correlation coefficients
ranged from 0.19 to 0.80. All correlation coefficients except the one between
concentrations at ROX and S/E are significant at the 5 percent level. In 1975,
the number of pairs ranged from 24 to 283, and the correlation coefficients
ranged from 0.27 to 0.91. The lowest coefficient found, and the only one
below the 5 percent significance level, was between the ALL and S/E sites,
which implies an unusually large effect from unrelated local sources at these
two locations. In 1976, the number of pairs ranged from 16 to 283 and the
correlations ranged from 0.37 to 0.95. Table 6 shows the distribution of the
coefficients by class intervals and the average of the coefficients calculated
between the observations at each of the 13 sites and the remaining 12
sites in 1976. The sites with the lowest average correlation were INT
21
-------
9
1
5
H
CO
§
o
w
CO
PH
3
O
S3
1
CN
Z
r3
H
erf
H
O
CO
o
H
I
W
co
o
fa
o
H
z
w
w
PH
m
CO
w
CO
PR
H
CO
fa
o
0
w
-^
CO
erf
PQ
CO
CQ
rJ
U-l
M O
p^ O
[V, pyj
s^^
CL,
Crf <5
fa O
s-/
Crf
ea
w
z
g
erf
^
w
PQ
H
Z
1— 1
,j
rJ
<
fa
W
Q
r"
4J
C
O
s
co m o co O r^ O
i— 1 CM CN
CM CM CO
C ^2 M S-l >% C i— 1
CB 0) crj CX co 3 3
*— ^ £n ^ <£ ^ ^-5 r-^
O O r-l rH
r-l CO
i— 1 O r- 1 r-l
CO ^O CO CM
I—I rH
O r-* o
3 0) O O
<£ CO O ^4
r^~ v^
^O ^d"
rH
CO rH
i — 1
r- O-
CO
0
rH ON
CM OO
00
VD CO
rH
o
OJ CD )-(
Q &
-------
fa
CO
OOCNr-^i—(
CM m
Id
PQ
H
Id
O
r-
oo
C
O
s
23
-------
CO
C
O
o)
M
o
a
en
H
O
•a
at
en
o
I
M
01
CU
CO
M
O
4-1
n)
M
o
at
(-4
3
O
"S
at
CD
M
24
-------
TABLE 5. LINEAR CORRELATION COEFFICIENTS BETWEEN 24-HOUR TSP LEVELS
AT PHILADELPHIA MONITORING STATIONS IN 1974, 1975, AND 1976
Site
DBF -
ALL -
IOT -
BEL -
ROX -
N/E -
NBR -
FRI -
LAB -
SBR -
S/E -
500 -
AFS -
'76
'75
'74
'76
'75
'74
'76
'75
'74
•76
'75
'74
'76
'75
'74
'76
'75
'74
'76
'75
'76
'75
'74
'76
•75
'74
'76
'75
'74
'76
'75
'74
'76
'75
•74
•76
'75
•74
DBF
i.oo
1.00
) .00
0.81
0.61
0.64
0.69
0.56
0.72
0.70
0.72
0.59
0.49
0.64
0.56
0.76
0.65
0.63
0.65
0.66
0 75
0.68
0.74
0.78
0.78
0.79
0.78
0.75
0.89
0.59
0.80
0.79
0.67
0.78
0.77
0.84
ALL
0.81
0.61
0.6't
] .00
1.00
1.00
0.57
0.52
0.50
0.66
0.45
0.50
0.45
0.36
0.44
0.60
0.66
0.51
0.66
0.46
0.75
0.68
0.74
0.70
0.6b
0.62
0.76
0.55
0.81
0.27
0.40
0.85
P. 47
0.62
0.75
0.66
INT
0.69
0.56
0.72
0.57
0.52
0.50
1.00
1 .00
1.00
0.37
O.b6
0.39
0.51
0.51
0.44
0.70
0.56
0.60
0.55
0.41
0.62
0.60
0.56
0.74
0.63
0.54
0.67
0.44
0.61
0.47
0.67
0.65
0.43
0.58
P. 59
0.58
BEL
0.70
0.72
0.59
0.66
0.45
0.50
0.37
0.66
0.39
1.00
1.00
1.00
0.88
0.89
0.56
n.78
0.80
0.55
0.85
0.87
0.91
0.85
0.72
0.78
0.86
0.61
0.81
0.8fa
0.73
0.48
0.53
0.95
0.76
0.62
0.56
0.87
ROX
0.49
0.64
0.56
0.45
0.36
0.44
0.50
0.51
0.44
0.68
0.89
0.56
1.00
1.00
1.00
0.65
0.68
0.68
0.60
0.66
0.67
0.65
0.6',,
0.66
0.78
0.61
0.62
0.64
0.46
0.46
0.19
0.59
0.68
0.52
0.67
0.54
N/E
0.76
0.65
0.63
0.60
0.66
0.51
0.70
0.56
0.60
0.78
0.80
0.55
0.65
0.68
0.68
1.00
1.00
1.00
0.63
0.43
0.70
0.72
0.54
0.85
0.91
0.68
0.71
0.69
0.75
0.54
0.47
0.74
0.65
0.53
0.71
0.73
NBR
0.65
0.66
0.66
0.47
0.55
0.41
0.85
0.87
0.60
0.66
0.63
0.43
1.00
1.00
0.76
0.56
0.78
0.66
0.89
0.72
0.60
0.30
0.75
0.66
0.72
0.62
*
FRI
0.75
0.68
0.74
0.75
0.68
0.74
0.62
0.60
0.56
0.91
0.85
0.72
0.67
0.65
0.60
0.70
0.72
0.54
0.76
0.56
1.00
1 .00
1.00
0.75
0.80
0.70
0.85
0.74
0.78
0.49
0.67
0.83
0.63
0.80
0.68
0.79
LAB
0.78
0.78
0.79
0.70
0.66
0.62
0.74
0.63
0.54
0.78
0.86
0.61
0.66
0.78
0.61
0.85
0.91
0.68
0.78
0.66
0.75
0.80
0.70
1.00
1.00
1.00
0.81
0.81
0.68
0.54
0.54
0.84
0.78
0.68
0.87
0.82
SBR
0.78
0.75
0.76
0.55
0.67
0.44
0.81
0.86
0.62
0.64
0.71
0.69
0.89
0.72
0.85
0.74
0.81
0.81
1.00
1.00
0.74
0.34
0.81
0.65
0.79
0.83
S/E
0.89
0.59
0.80
0.81
0.27
0.40
0.61
0.4/
0.67
0.73
0.48
0.53
0.46
0.46
0.19
0.75
0.54
0.47
0.60
0.30
0.78
0.49
0.67
0.68
0.54
0.54
0.74
0.34
1.00
1.00
1.00
0.70
0.49
0.53
0.84
0.50
500
0.79
ii 67
0.78
".85
0.47
0 62
0.66
0.43
0 58
0.95
0.76
0.62
0.59
ii 68
11.52
0.74
0.65
0.53
0.75
0.66
0.83
0.63
ii 80
',.R4
0. 78
0 68
P. 81
0.65
0.70
0 49
'1.53
i .00
1 .00
1 . 00
|.77
'1.73
AFS
0.77
0.84
0.75
0.66
0.59
0.58
0.56
0.87
0.67
0.54
0.71
0.73
0.72
0.62
0.68
0.79
0.87
0.82
0.79
0.83
0.84
0.50
0.77
0.73
1.00
l.DO
Average
0.74
0.68
0.69
0.70
0.53
0.55
0.61
0.53
0.56
0.75
0.76
0.56
0.60
0.62
0.51
0.72
0.67
0.58
0.70
0.58
0.75
0.68
0.67
0.77
0.75
0.64
0,77
0.67
0.72
0.46
0.53
0.77
0.63
0.63
0.73
0.71
"CAMP station in 1974; roof-top station in 1975.
25
-------
CO
O
t_i
H
1
W
U
§
o
PM
CO
H
•*tj
w <
W Q
P vo
W r-.
PQ ON
i— 1
CO
Is !z
W M
M CO
0 CD
t— i
Pn co
Pn 2
W O
O M
U H
*^
ss H
O CO
H
H CM
H
2 S5
2 M
o a
O M
Du §<
O j§
O O
g g.
M O
Pi M
H H
CO -<
M H
O CO
JM Q
CJ W
CD U
cx M
§ s
fr | I |
vO
W
3
Q
0
(1
t-
Q
C
O
Q
O
.
0
o
o
§
o
_
r-
0
o
4J
o
0
•£>
o
o
o
o
in
0
o
0
o
o
o
o
o
^
CO
o
o
0
o
o
AJ
4-t
ooooooooooooo o
H
a.
o
c
•H
c
0
•H
4-1
w
u
c
1
ca
Jr
26
-------
and ROX, each with r = 0.60. The highest average correlation, r = 0.77,
occurred at three locations: the LAB, SBR and 500. The average coefficient
for all 13 stations is 0.72. It is interesting to note that the two stations
(INT and ROX) with the lowest average correlation, are both generally upwind
of the principal urban area. In contrast, the Northeast Airport, although
also removed spatially from the center of the city, is downwind much of the
time and has an average coefficient of 0.72.
More detailed matrices, which include the number of pairs of observations
associated with each correlation coefficient, for the 3 years may be found in
Appendix B (Tables B-l, B-2, and B-3). The averages of the coefficients for
the paired data sets in which a station is included, listed in the right-hand
column of Table 5, have been plotted at the site locations in Figure 10. The
highest average correlations are found along the central core area of the city,
the axis of which extends roughly from southwest to northeast. This effect
would appear to be in part a reflection of a shorter average separation dis-
tance, and in part the result of transport by prevailing southwesterly winds
along the major emissions corridor. The coefficients of 0.76 and 0.75 found
respectively in 1975 and 1976 at BEL do not fit into this general pattern,
but are based on data from incomplete sampling years.
Figure 11, which is a plot of the square of the correlation coefficient
between TSP levels at paired stations versus separation distance, illustrates
the general decrease in common variance with increasing separation distance,
with the principal exception of values which include the station at the North-
east Airport (N/E). The r2 values between this station, which is downwind
from the general urban emission? much of the time, and other stations are about
twice as great as expected from the bulk of the data. The lack of local sources
at this site (see site description, Appendix A) appears to make this station
sensitive mainly to large scale influences and less affected by the small scale
variations which decrease the correlations more rapidly for other station pairs.
Similar plots for 1974 and 1975 are presented in Appendix B (Figures B-l and B-2)
From the results of the preceding correlation analysis and careful exam-
ination of days experiencing either particularly high or low concentrations,
the following conclusions were drawn:
• The highest concentrations observed at individual sites tend to
occur on the same days. These high particulate days are usually
associated with either poor ventilation conditions or with
southerly flow and an apparent concentration of industrial pollu-
tants immediately following the passage of a major high pressure
cell.
• The lowest concentrations observed at individual sites also tend
to occur on the same days. These low particulate days are
usually days with precipitation and frequently occur on weekends.
• The common variation in concentrations measured at two sites is
a function of separation distance, averaging about 70 percent at a
separation distance less than about 5 kilometers (3 miles) and
about 25 percent at distances of about 16 kilometers (10 miles).
27
-------
Figure 10. Average of correlation coefficients between TSP concentrations
at indicated station and all other stations.
28
-------
1.0
0.9
0.8
0.7
0.6
0.5
0.4
0.3
0.2
O.I
0
•
• •
•
•
BEL/NE
4E
DEF/NE
SE/NE
-SBR/NE 9
— FRI/NE y
ROX/NE ^—INT/NE
0
I
4
I
8
1
Km
4 6 8 10 12 14
SEPARATION DISTANCE , mile*
16 18 20
Figure 11. Square of the correlation coefficient between TSP
levels as a function of distance, 1976.
29
-------
Thus, extreme changes in concentration tend to be an areawide phenomenon
brought about in response to the synoptic meteorological condition, while
smaller changes often reflect local conditions. A source-oriented monitor may,
of course, be subject to large variations independent of areawide conditions.
DAY OF THE WEEK AND PRECIPITATION EFFECTS
It is commonly recognized that TSP levels in urban areas are suppressed
by precipitation, and that concentrations also decrease during weekends. In
particular, Lazenka and Weir1* of the Philadelphia Air Management Services
found, using 1972 and 1973 data, that a statistically significant reduction
in concentration occurred for a period of 2 days following precipitation equal
to or greater than 0.25 inches. This reduction averaged about 38 and 25 per-
cent, respectively, for the first and second days after precipitation. Also,
no significant differences were found among the means of the data for the three
sites used in the study (AMS Laboratory, 500 South Broad Street, and the South-
east Water Pollution Control Plant). Similar results have been found in other
cities.5 Thus, accurate short-term prediction of TSP levels by urban modeling
requires a procedure for estimating emission changes and washout effectiveness
with sufficient precision to account for these observed reductions.
Figure 12, prepared from 2 years of data from the AMS Laboratory monitor,
demonstrates rather strikingly the importance of these factors. In preparing
Figure 12, two precipitation categories were used to represent "no-rain" and
"rain" conditions. A no-rain period was defined as one with a trace or less
of precipitation during the 48-hour period ending at the end of the sampling
period, and a rain period was drained as one with at least 0.10 inches of
precipitation during the 48-hour period. Table 7 presents the average weekday
and weekend concentrations for these two precipitation classes. The same aver-
age decrease of 22 yg/m3 or 25 percent occurred between no-rain weekday con-
centrations and either no-rain weekend concentrations or rain weekday concen-
trations. A further average reduction of 13 ug/m3 or 20 percent occurred
between the average concentration of 66 ug/m3 observed during these two con-
ditions and rain weekends.
TABLE 7. AVERAGE WEEKDAY AND WEEKEND CONCENTRATIONS FOR
TWO PRECIPITATION CLASSES
Average
48-hour concentration, Reduction
precipitation, yg/m3
(in.) •yg/m3 Percent
Weekday Weekend
Trace or less
> 0.10
88
66
66
53
22
13
25
20
30
-------
100
90
ro
E
v.
o>
- 80
O
(T
t-
Z
UJ
O
z
o
o
70
60
V)
I-
UJ
o
2 50
UJ
40
TRACE OR LESS
(44)
(46)
(34)
i
Su
M
Tu W Th
DAY OF WEEK
So
Figure 12. Average weekly variation in TSP levels at the AMS Laboratory
in Philadelphia for two precipitation categories during 1974
and 1975. Number of observations for each data point is in
parentheses.
31
-------
A more detailed analysis of the effect of precipitation is provided in
Appendix C. This analysis, which used the 1974, 1975 LAB data, showed that
the overall effect of precipitation from a statistical standpoint was to re-
duce or eliminate days with high concentration; low concentration frequently
occurs on days without precipitation.
The association of precipitation with lowered TSP concentrations results
from a number of interrelated factors. First of all, precipitation is a
cleansing agent and results in the washout and rainout of pollutants; second,
precipitation suppresses fugitive dust emissions; and third, precipitation is
frequently followed by a frontal passage and the arrival of a cleaner air mass.
COMPARISON OF DAILY TSP CONCENTRATIONS AT SBR AND 500 UNDER FOUR REGIMES
The historical TSP data collected by the AMS Laboratory at the Broad and
Spruce and 500 South Broad Street sites between September 8, 1974 and March 26,
1977 were analyzed statistically to guide the model development phase of the
project and to provide a reference against which the success of candidate
modeling techniques might be judged. In particular, an answer was sought to
this question, "How accurately can the Broad and Spruce trailertop concentra-
tion be estimated from the rooftop concentration at 500 South Broad?"
For this analysis, days with concentration data at both locations were
separated into the four categories used earlier in the program to investigate
day-of-the-week and precipitation effects. As before, precipitation was mea-
sured over a 48-hour period ending at the end of the sampling period. The
four categories were:
1. Weekdays with a trace or less of precipitation
2. Weekdays with 0.10 in. or more of precipitation
3. Weekends with a trace or less of precipitation
4. Weekends with 0.10 in. or more of precipitation
Average concentrations at the two sites were then calculated for each
category. The results are given in Table 8(a), and the average difference in
concentration between the two sites for the four categories is given in
Table 8(b).
From Table 8(b) it can be seen that the concentration at Broad and Spruce
averages approximately 43 ug/m3 higher than the concentration at 500 South
Broad, with the maximum difference of 53 ug/m3 occurring on weekdays when the
48-hour precipitation did exceed not a trace, and the minimum difference of
36 yg/m3 occurring on weekends when precipitation equaled or exceeded 0.10 in.
These results are in agreement with the hypothesis that maximum contributions
from street level sources occur with dry streets and high traffic volume.
Two simple assumptions as to the relationship between concentrations at
the two sites were tested. The first was that the concentration at the lower
site was some fixed percentage (greater than 100) of the concentration at the
32
-------
TABLE 8. AVERAGE TSP CONCENTRATION DATA AT THE BROAD
AND SPRUCE AND 500 SOUTH BROAD SITES BY DAY-
OF-THE-WEEK AND PRECIPITATION CATEGORY
(a) Observed concentrations
Site
48-hour precip-
itation class,
Average concen-
tration (yg/m3)
Broad & Spruce
500 South Broad
*
Number of cases it,
(in.)
<_ T
>_ 0.10
<_ T
>_ 0.10
in parentheses
Weekday
137
(35)*
116
(31)
84
(35)
73
(31)
•
Weekend
121
(16)
96
(19)
80
(16)
60
(19)
(b) Concentration differences (Broad and Spruce -
500 South Broad)
48-hour
precipitation
class (in.)
Average TSP, yg/m3
Weekday
Weekend
<_ T
> 0.10
53
43
41
36
33
-------
higher site. The second was that the difference in concentration between the
two heights was independent of the magnitude of the rooftop concentration.
Figure 13 shows that the ratios of concentrations varies with concentration
and ranges from about 2.0 at a rooftop concentration of 40 yg/m3down to per-
haps 1.2 at a rooftop concentration of 200 yg/m3. Figure 14 shows that the
linear regression equation for the prediction of Broad and Spruce Street con-
centration from the rooftop concentration, based on all days, has a slope of
approximately one (1.086), indicating only a very slight increase in the aver-
age difference in concentration as the rooftop concentration increases.
Approximately one-half of the variation in concentration at the two sites is
common (r2 = 0.52). Figures 15 through 18 show similar scatter plots, re-
gression lines, and correlation coefficients for the 4 day-of-the-week/
precipitation categories.
One conclusion to be drawn from this analysis is that the Broad and
Spruce concentration can be roughly approximated by adding 43 yg/m3 to the
observed (or predicted) concentration at 500 South Broad. This estimate
can be refined somewhat by considering the day-of-the-week/precipitation
category. An obvious question that remains is the extent of any improve-
ment that could be expected from the use of wind speed and wind direction
as additional predictors. This topic is addressed briefly in Section 5.
CONCENTRATIONS OF METALS AND SULFATES
Concentrations of metals and sulfates measured by the AMS Laboratory at
the 13 monitoring sites were averaged to obtain monthly concentrations from
July 1976 through June 1977. The number of 24-hour concentrations measured
at the various sites ranged from every 6th day throughout most of the network
to every day at the AMS Laboratory. The monthly averages are tabulated in
Appendix D with the number of observations used in each calculation. At three
sites, LAB, S/E and 500 South Broad Street, concentrations are currently
available only through December 1976.
Table 9 lists the average monthly concentrations of metals and sulfates
calculated from the data in Appendix D. Averaged over the network, sulfates
contribute 11.6 yg/m3 to the particulate loading, with the range being from
8.5 to 13.8 yg/m3. The lowest values were reported for LAB, SBR, S/E and
500, and the highest values were reported for AFS, NBR, DEF, ALL, and INT.
The concentrations of lead are of particular interest because of the EPA
proposed standard of 1.5 yg/m3, based on a monthly average. This proposed
standard was exceeded at three sites: AFS, which is source-oriented and lo-
cated near a National Lead Plant; and the two traffic-oriented, trailer-top
sites on Broad Street, SBR and NBR. It was exceeded on 10 of 12 months at
AFS; on 4 of 11 months at SBR; and on 5 of 12 months at NBR. The monthly
averages, taken from Tables D-7, D-10 and D-13 of Appendix D, are shown in
Table 10. From these monthly averages, it appears that the proposed lead
standard will be met at the two traffic-oriented sites in 1978 as a result of
the phase-down schedule for maximum pooled average lead concentration of gaso-
line.6 This conclusion is based on the following assumptions: (1) there will
be no significant increase in traffic volume, (2) the average lead content of
34
-------
a
*
-------
4.
u7
cc
a.
o
z
o
tr
CD
z
o
LU
u
z
o
o
a.
CO
300
280
260
240
220
200
180
160
140
120
100
80
60
40
20
0
i r
y = l.086x +37.33-
r=0.78
•^-OMITTED
FROM
ANALYSIS _
20 40 60 30 100 120 140 160 180 200 220
TSP CONCENTRATION AT 500 SOUTH BROAD ,
Figure 14. Relationship between TSP concentrations at Broad and Spruce
and 500 South Broad. (Period of observations: September
1974 to March 1977).
36
-------
io
E
*x
O»
4.
tr
a
o
z
o
tr
m
z
O
Z
Ul
u
z
o
a
-------
300
280
260
ro 240
E
? 220
o
3
or
0.
o
2
o
or
CO
cc
K
2
UJ
U
z
o
200
180
160
140
120
100
80
60
40
20
1 I I 1 I
/
Y =l.09x +• 36.70
r =0.76
n =31
OMITTED
FROM
ANALYSIS
_l i
0 20 40 60 80 100 120 140 160 ISO 200 220
TSP CONCENTRATION AT 500 SOUTH BROAD,
Figure 16. Relationship between TSP concentrations at Broad and Spruce
and 500 South Broad on weekdays with precipitation >^ 0.10 in.
38
-------
300
280
260
ro 240
E
o. 220
o
1)
OC
CL
0
z
o
<
o
cr
CD
o
oc.
\-
z
LU
200
180
160
140
!20
100
80
60
O 40
o
a 20
05-
I l
X
_i-
J_
X
_L
_L
J_
0 20 40 60 80 100 I2O 140 160 180 200 220
TSP CONCENTRATION AT 500 SOUTH BROAD,
Figure 17. Relationship between TSP concentrations at Broad and Spruce
and 500 South Broad on weekends with precipitation < T.
39
-------
IO
E
•v*
O>
:t
UJ
or
o.
v>
o
z
<
o
o
or
CD
z
o
Z
UJ
V
z
o
o
=0.94x 438.82
r = 0. 75
n =19
0 20 40 60 80 100 120 140 160 ISO 200 22O
TSP CONCENTRATION AT 500 SOUTH BROAD ,/j.g/m3
Figure 18. Relationship between TSP concentrations at Broad and Spruce
and 500 South Broad on weekends with precipitation > 0.10 in.
40
-------
TABLE 9. AVERAGE MONTHLY CONCENTRATION OF METALS AND SULFATES IN yg/m3
Site
DEF
ALL
INT
BEL
ROX
N/E
NBR
FRI
LAB
SBR
S/E
500
AFS
AVER-
AGES
Pb
0.86
0.67
0.70
0.49
0.43
0.53
1.48
0.75
1.10
1.47
1.13
0.84
3.11
1.04
Fe
1.27
1.39
1.09
0.61
0.61
0.62
1.66
0.80
1.01
1.68
1.47
0.84
1.10
1.09
Mn
0.032
0.222
0.032
0.022
0.025
0.023
0.040
0.026
0.046
0.040
0.055
0.030
0.135
0.056
Cu
0.121
0.137
0.100
0.074
0.052
0.070
0.102
0.0«7
0.179
0.101
0.130
0.116
0.095
0.106
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
Ni
023
021
015
017
015
015
020
019
027
022
028
028
022
021
Cd
0.002
0.004
0.002
0.002
0.002
0.003
0.003
0.002
0.003
0.003
0.004
0.003
0.003
0.003
Mg
0.48
0.37
0.50
0.26
0.23
0.20
0.56
0.29
0.34
0.62
0.47
0.31
0.36
0.38
Al
0,37
0,36
0.50
0.30
0.27
0.22
0.37
0.28
0.37
0.45
0.49
0.33
0.41
0.36
50!+
13.4
13.0
13.1
11.2
11.3
11.7
13.8
12.8
9.7
8.5
9.8
9.2
13.8
11.6
No. of
months
12
12
12
8
12
12
12
12
6
11
6
6
12
41
-------
TABLE 10. MONTHLY AVERAGE LEAD CONCENTRATIONS
AT AFS, SBR AND NBR
AFS
Month
7/76
8/76
9/76
10/76
11/76
12/76
1/77
2/77
3/77
4/77
5/77
6/77
yg/m3
2.84
4.34
2.83
2.52
1.91
3.84
3.55
1.43
3.94
4.11
4.98
1.02
No.*
5
5
4
4
4
4
6
4
5
5
6
4
SBR
yg/m3
1.69
1.41
1.64
1.26
1.40
1.54
1.12
-
1.47
1.84
1.48
1.27
No.
5
5
5
5
5
5
5
0
5
5
6
1
NBR
Ug/m3
1.73
1.38
2.11
1.28
1.28
1.28
0.83
1.40
1.32
1.81
1.82
1.56
*
No.
5
5
5
5
5
5
6
3
5
5
6
4
*
Number of observations.
42
-------
the gasoline used by vehicles in Philadelphia in 1976 and 1977 was no lower
than that of the promulgated schedule (i.e., 1.4 grams per gallon in 1976 and
1.0 gram per gallon in 1977), and (3) that the maximum allowable pooled aver-
age of 0.8 gram per gallon will be met by 1 January 1978. Calculations based
on 22 sets of paired observations at SBR and 500 South Broad Street yielded
values of 1.38 and 1.08 percent, respectively, for the percent of lead in the
total particulate catches.
It is also of interest to note that the two sites most exposed to vehicular
emissions and reentrained dust, SBR and NBR, experience the highest concen-
trations of iron. The site with the third highest concentration of iron was
S/E, perhaps as a result of activities associated with the processing of scrap
iron, or the storage of iron ore materials in the general vicinity.
43
-------
SECTION 4
FIELD STUDY RESULTS
The principal field experiment carried out during the program was designed
to characterize the spatial distribution of particulates in the vicinity of a
major traffic-related source. As noted earlier, recent observations had iden-
tified the area in the vicinity of the 500 South Broad Street site and the in-
tersection of South Broad and Spruce Streets as a promising one for such a
study. Here, within a distance of approximately 275 meters, the rooftop loca-
tion typically has an annual mean approximately equal to the primary standard,
while the trailer-top location experiences an annual mean 50 percent higher,
with the only obvious differences in the sites being the proximity to vehicular
traffic.
Although it was recognized that the effects of the surrounding building
structures would greatly complicate any theoretical study, it was felt that
even the acquisition of data in such a "real-world" problem area would be a
worthwhile goal. Such data would supply needed empirical information and
could, hopefully, provide guidan^ •> in any theoretical modeling that might be
undertaken later. Accordingly, it was decided to undertake the following
three specific project goals at this one central study area: (1) determine
the influence of traffic-related emissions on TSP concentrations as a function
of height, (2) investigate the effect of horizontal distance from the roadway
on rooftop concentrations, and (3) test the effectiveness of street-washing
in reducing reentrained street dust.
In addition to these experiments to monitor traffic-related emissions,
a series of hi-vol measurements were made in the vicinity of the Southwest
Sewage Treatment Plant, where earth moving activities associated with its
construction continued throughout the duration of the entire study period.
As a result of the encouragement and active participation of the Project
Officer, the original design of the field program was altered to include a
number of measurement techniques in addition to the use of hi-vols which had
been specified in the contract scope of work. The inclusion of these tech-
niques made possible the accumulation of a modest data base on the relative
concentration of respirable and nonrespirable particulates at the Broad Street
sites, and on elemental concentrations within these size fractions.
44
-------
OBSERVATIONS NEAR TRAFFIC ACTIVITY
Design of Monitoring Network
Figures 19 and 20 show the streets, building heights and approximate loca-
tions of the monitoring equipment at the two Broad Street sites. At the Broad
and Spruce Street site, a 12.2-meter (40-foot) self-supporting tower was
erected next to the AMS air monitoring trailer (see Figure 19) and outfitted
with six hi-vols at three heights. This allowed operation of the hi-vols on
a 24-hour daily basis without requiring the operator to be present at midnight
to change filters. The effective heights of the hi-vols were 11.2 meters,
9.1 meters and 6.1 meters above ground. The vertical profile was completed
by a pair of hi-vols at an elevation of 4.0 meters on the trailer rooftop. In
addition, an EPA Dichotomous Sampler and a GCA Ambient Particulate Monitor
were operated to provide particle size information and short-term concentration
data. Traffic counts were automatically recorded every 15 minutes at four
locations on Broad and Spruce Streets to define traffic volumes at the inter-
section and on Broad Street in the vicinity of the rooftop monitors.
At 500 South Broad Street, the existing AMS hi-vols were supplemented by
a hi-vol on the front edge of the roof near Broad Street and a hi-vol on the
back edge of the roof away from Broad Street. Meteorological instrumentation
to record wind speed and direction was positioned on an elevated portion of
the roof near the back of the building. Directly across the street from 500
South Broad, on the roof of the Philadelphia Center for Older People (PCOP),
two additional hi-vol sites were established. One was on the front edge near
Broad Street and the other on th" back edge away from Broad Street. This re-
sulted in a horizontal array of tive hi-vols at nearly the same elevation
above Broad Street. The three on the 500 South Broad roof were on the west
side of Broad Street and at distances of approximately 9 meters, 30 meters,
and 58 meters from the street. The hi-vols on the east side of Broad Street
on PCOP were approximately 9 meters and 34 meters from the street. In
addition, during the February sampling period, a NASA Air Scout directional
sampler with its meteorological sensors was operated on the Broad Street edge
of the 500 South Broad Street roof.
A brief description of the three less well-known specialized instruments
used in the program, the EPA Dichotomous Sampler, the GCA Ambient Particulate
Monitor, and the NASA Air Scout, is provided in Appendix E.
Sampling Schedule
The basic hi-vol network, consisting of the vertical array at the tower
site and the horizontal array at rooftop level approximately 275 meters to the
south, was operated from 4 February to 20 February 1977 and from 7 June to
20 June 1977. These observations were supplemented with APM and Dichotomous
Sampler data from the trailer-top roof during the 2-week sampling period in
February, and at the end of this period, the Dichotomous Sampler was moved to
the rooftop at 500 South Broad Street and operated for an additional 2-week
period. The intensive street washing carried out by the City of Philadelphia
45
-------
9m HIGH
S
6m
«—•
N
\
LOCUST STREET
OPEN DECK
GARAGE
-30
m
13m
HIGH
5m
HIGH
}u
Uj
«
k.
Li
SPRUCE STREET
-27m-
14m HIGH
•41m
TOWER EAST APTS
8Zm HIGH
L
7m
i
PARKING LOT
126m
•4fi m
PARKING LOT
/-TRAILER
1 1 ^ — rnwrp
5m ,
Uj
«0
£
^
*
6m
:$>
PARKING
5m
HIGH
6m
HIGH
MANNING ST.
I
o
I"e
-> °- S
z r
g 0
o r
o: 1- (o
O Q. AJ
03 < M
« 37m — H
!*J
^
«0
Q;
ex.
Figure 19. Local neighborhood at Broad and Spruce Streets showing locations
of AMS trailer and monitoring tower.
46
-------
x
0
X
E
to
Z.
o
X
E
o>
_
0
X
E
Z"*
^^
^ (
E
3
J
(9
2
a
<
a.
*"*" T
K- TT OTJ E
53 Q 0
5 Cf ***
z
oc
a.
X
0
X
E
kl
g
Q
1
c>
_L.
*—
i
UJ
2
i
•
<
c_>
Zj
Ul
O X
X &
z i
UJ
E
>• «
H —
z
CL
Ul
t-
CC
u.
<§
"•
5 ^
k- •"' _ .
< a: £
? o ™"
«j
-J O
o
X
a V
\
_j
o
>
1
X
/
]| 1
(O •
CM
(9
E *
•£
'
if
>
a
*
** *fe>
*?>*rf>
*f-
*^
4
X>
y
^Q
1?)
r? 1
1 < 1
1 Ul I
1 y § | 1
1 l ^ 6 1
1 U.
(U
a)
4-J
•a
cfl
O
,c
4J
s
o
o
1 o
4-* •
x ct) to
(9 P
= 13 01
1 OH
O ft
E js e
- o w
•^ S
oO E
•H 3
o
rt A CSI
^. ^ tP »• LJ
^ ^1
^ *A
f *•
\
P
^)
p 60
"'A ^
o-
47
-------
in the vicinity of the monitors took place on the 15th, 16th, and 17th of
June 1977. At the end of the 3rd day of street washing, the hi-vols at Broad
and Spruce were operated on a 4-hour schedule. This 4-hour schedule continued
until 2000 EDT on the 19th.
February Field Measurements
Summary of Hi-Vol Data (TSP and Pb) —
All of the TSP concentrations determined by means of the hi-vols are
tabulated in Appendix F. In addition, lead concentrations are tabulated for
all of the filters exposed during the February experimental program except
those handled by the AMS Laboratory from its regular monitor at 500 South
Broad. Concentrations were determined at GCA/Technology Division using the
Standard Perkin-Elmer method for Metallic Air Pollutants. This consisted of
digesting the filters in hydrochloric acid for 30 minutes, evaporating the ex-
tract nearly to dryness, and diluting the residue with 10 ml HC1, 1 ml HN03
and deionized water to a constant (50 ml) volume. The sample was then analyzed
for lead using Atomic Absorption Spectroscopy and quantified by comparison
with known standards.
Examination of the raw data shows that although the measured vertical
profiles at the tower are sometimes well behaved, that is, show concentrations
decreasing smoothly with height, other profiles are highly irregular. This
may be a reflection of difficulties in handling the hi-vol filters, particu-
larly when servicing the tower-mounted units. At any rate, because of these
irregularities, no attempt has been made to analyze day-to-day profile dif-
ferences. Instead, an average profile has been calculated for the 16-day
period. Differences in sampling times and an occasional missing level have
been accounted for by normalizing concentrations to the bottom tower level.
Similarly, average rooftop concentrations for the 16-day period have been
calculated from the more limited number of these observations by rationing
to average tower concentrations. The resulting average concentrations at the
various monitoring locations are shown in the site sketch presented in
Figure 24. Lead concentrations, calculated in the same way, are shown in
Figure 24 enclosed in parentheses.
In view of the small number of observations and large amount of scatter,
conclusions drawn from Figure 21 should be considered provisional. With this
in mind, the data indicate an increase in concentration at the trailer of
about 38 percent above rooftop levels as a result of street-level sources
(141 ug/m versus 102.5 ug/m3). The rooftop observations indicate concen-
trations about 15 percent higher adjacent to and downwind from Broad Street
than average rooftop concentrations. This increase, which may also be
attributed to street emissions, was not reflected in the average concentra-
tions measured about 80 feet farther from the street. It appears from the
tower measurements that, on the average, particulates are quite uniformly
mixed vertically at this location except for a portion of those contributed
by the nearby street sources which impact most significantly on the hi-vol
on the roof of the trailer.
48
-------
fO « — K
lit
z
z"
t «
W +
s s
* g
o £
QJ >
•U O
•H rH
CO CU
,£>
tfl
•H 03
.C CL)
CX CO
i-H 0)
QJ J2
"Q 4J
CO C
4J C
cd -H
73 CO
(U C
> O
J-i -H
0) 4J
co cO
f^f y i
O 4J
c
CO 0)
C O
o c
•H O
4J O
CO
4-1 PH
C
OJ 13
O C
O
O 0)
CU O
4-> &
ca co
o
•H
co
O •
-H ro
t-i 4-1 0
CO CO -».
CX M bO
a) a) cu
N O (-1
•H C CO
rH O
CO O CO
g4J
PH -H
CU
l-l
49
-------
Average lead concentrations follow the same pattern at the various loca-
tions as the TSP concentrations. The highest average concentration,
1.73 yg/m3, was measured at the trailer site. Concentrations averaged about
1.5 yg/m3 on the tower, and about 1.0 yg/m3 at rooftop level except 1.2 yg/m3
at the Center for Older People at the location nearest Broad Street. These
concentrations of 1.7 and 1.0 yg/m3 measured at the trailer and rooftop sites,
respectively, may be compared with average values of 1.5 and 0.8 yg/m mea-
sured by the AMS Laboratory during the latter half of 1976. Comparison of
average lead concentrations to average TSP concentrations shows that lead com-
prised approximately 1.2 percent of the particulates at the trailer and about
1.0 percent of the partlcualtes at rooftop level.
Consecutive 2-hour hi-vol measurements were made at the four heights at
the tower site from 0600 EST until midnight on February 17th. Again, because
of the irregular behavior of the individual profiles, no attempt has been made
to study height variations. Instead, an average concentration has been calcu-
lated for each 2-hour time period, and this average value plotted in Figure 25
to show the trends in TSP and lead throughout the day. The percent lead has
been calculated from these average values and is shown as a function of time
of day in the lower curve of Figure 22. During this particular 18-hour period,
TSP concentrations were high during the morning and again late in the evening,
with the lowest levels being found from 1600 to 1800 EST and again from 2000
to 2200 EST. On the other hand, the lowest average lead concentrations were
experienced during the morning and early afternoon (1000 to 1400 EST). As a
result, the percent lead reached its lowest level of 0.45 in the morning, and
its highest level of 1.35 in the early evening.
Summary of APM Data—
The GCA Ambient Particulate Monitor was operated on the trailer roof from
the 5th to the 14th of February. The results are shown in Figure 23 where the
concentrations measured by the two channels are plotted. The 1-hour sampling
period used initially was reduced to 45 minutes on February llth. The 2-sigma
concentration accuracy associated with these two sampling periods are approxi-
mately 9 and 12 yg/m3, respectively. Individual TSP observations are plotted
at the top of Figure 23 to show concentration changes throughout the day. The
most frequently reoccurring feature in the daily pattern is the morning
minimum, roughly associated in time with the 0300 to 0700 EST minimum in
traffic volume shown in Figure 24. (Additional traffic data for February
are given in Appendix G. Significant differences from the average diurnal
pattern occurred on Wednesday, February 9th, when the morning minimum was
less marked, and on Saturday and Sunday, February 12th and 13th, when the
concentration fell sharply after noon. The respirable fraction of particu-
lates has been plotted at the bottom of Figure 23. A problem arose with the
respirable channel during the end of the period and, as a result, the data
became less reliable. Approximate correction factors have been applied to
salvage some of these data. Figure 25 is a scatter diagram of the total and
respirable concentrations for all sampling periods. As shown in Figure 25,
most observations indicated a respirable component of over 50 percent.
50
-------
Z
O
t-
oc
X
Itl
O
O
O
06-O8 06-10 10-12 12-14 14-16 16-16 18-20 20-22 22-24
TIME OF SAMPLE
1.4 r
I Z
1.0
JO
°- 0.8
e
o 06
s.
Q.4
0 2
06-O6 06-10 10-12 12-14 I4-I« 16-18 10-20 20-22 22-24
TIME OF SAMPLE
Figure 22. Time sequence of TSP and Pb concentrations from 0600 to
2AOO EST at tower site on 17 February 1977.
51
-------
CD
id
0)
6
CD
c
o
CO
t-l
4-1
c
O)
a
C
O
O
CU
•U
rt
rH
3
a
•H
J-J
(-1
cd
a
0) O
13 4J
C
cu m
ex
CD so
3 C
CD -H
(J
OJ 3
tH T3
cd CU
M 4-)
•H -H
CU W
CO
0) t-l
J-t CU
-a o
C 4-1
td
01
rH JZ
Cfl 4-1
4->
O 4J
H cd
CO
CNJ
Q)
M
3
00
•H
Pn
^ui/6-"
Q3QN3dSnS 1VIO1
52
-------
co
cu
En
00
a
•H
3
(U
ctf
<1)
C
0)
I
a
•H
QNn09HinOS P"0 HiaON'133M1S QVOdS
I33ais
OUJVH1
53
-------
rO
240-
220
200-
!80-
ul 160
o
h-
cr
CL
Q
UJ
0
UJ
a
CO
z>
CO
_J
o
I—
40
120
100
80
60
i i i i i
Oj
CD
CO
PI
<^/«
#
o\o
-O /
:•'-;
tte*"
i I I 111 _l.._J 1 __J L__JL_J L _L_J J 1 L_
"0 20 40 60 80 100 120 140 160 180 200
RESPIRABLE PARTICULATES./ig/m3
Figure 25. Comparison of total and respirable suspended particulate
concentrations measured by the APM at the tower site
during 5 to 14 February 1977.
54
-------
Comparison of APM and Hi-Vol Data—
Figure 26 compares TSP concentrations measured by the APM during the
10-day sampling period with concentrations measured with the standard hi-vol.
Because of timer problems in operating the hi-vol on the roof of the trailer
where the APM was installed, and the lack of sufficiently regular vertical
profiles at the tower to permit downward extrapolation with confidence, the
average of available observations at the tower site has been plotted as the
hi-vol measurement in Figure 26. Corresponding APM measurements are the aver-
age of all 1-hour (or 45 minute) measurements made during the 24-hour period.
During a typical 24-hour period, the APM underwent automatic calibration twice,
resulting in two periods of about 3 hours each when no ambient measurements
were recorded.
Figure 26 shows excellent correlation between the two measurement tech-
niques over the wide range of concentrations experienced during the 10-day
period. Of great interest is the fact that in this highly reactive atmo-
spheric environment the hi-vols are consistently measuring higher concentra-
tions than the APM. This is in marked contrast with results obtained in lab-
oratory and relatively clean ambient air. A possible explanation for this
difference is based on a mass accretion on the hi-vol filters, with respect
to the APM collections, as a result of two principal mechanisms: (a) passive
particle deposition, and (b) chemical reactions between the collected particu-
lates and various gaseous species present in the ambient air. The first
accretion mechanism has been documented recently, whereas the latter one has
received some attention in the past. Two operational characteristics that
distinguish the hi-vol from the APM would lead to the observed difference if
the above-mentioned mechanisms are operative: (a) the hi-vol has a filter
face velocity of about 1/7 of thac of the APM, thus the residence time for the
gases passing through is correspondingly longer, and (b) the hi-vol filters
are left exposed to the accretion mechanisms discussed above for extended
periods of time; i.e., their sampling times are 24 hours versus 1 hour typi-
cally for the APM and, in addition, the hi-vol filters are exposed for an
additional period of time before they are replaced.
These accretion mechanisms in conjunction with the above-discussed oper-
ational characteristics of the two types of samplers are expected to result
in an enhancement of the TSP values as determined by the hi-vol with respect
to those determined with the APM, and the observations corroborate the direc-
tion of this discrepancy.
Concentration differences at outdoor locations could also be expected to
result from differences in the inlet configuration of the two instruments, and
might even be related to the mean wind direction during the period. It has,
for example, been demonstrated by wind tunnel studies that the collection effi-
ciency of the hi-vol for large particles is extremely sensitive to the angle
of the approaching wind.7
Summary of Dichotomous Sampler Data
(Respirable and Nonrespirable Fractions)—
The EPA Dichotomous Sampler was operated by the Project Officer in con-
junction with the field measurement program during the period from 13 February
55
-------
o
>
10
J^
^t
C^)
"
c«»u
220
200
180
160
120
100
80
60
40
20
0
II* //
y
/
• |o
/ «I2 r =0-95
' n = IO
• 14
7*J
6. / *8
TSP,H1.VOL)=9.69+l.39XTSP(ApM)
1 1 1 1 1 1 1 1 | 1 1 I
0 20 40 60 80 100 120 140 160 ISO 200 220 240
TSP(itg/m3) -ARM
Figure 26. Comparison of APM and hi-vol concentrations
(5 to 14 February 1977).
56
-------
to 11 March 1977. The instrument was located on the roof of the trailer at
Broad and Spruce from February 13th through February 25th, and on the roof of
500 South Broad from February 28th through March llth. Eight periods were
sampled at each location; all sampling periods were approximately 24 hours in
length with the exception of one 12-hour period at Broad and Spruce Streets.
For each period, EPA reported the total mass concentration in each of the two
particle size ranges (< 3.5 pm, and 3.5 to 20 ym) and the mass concentration
of 30 elements as determined by X-ray fluoresence spectrometry.
Figure 27 shows the gross results from the dichotomous sampler on a day-
to-day basis, and selected meteorological data for those days. The percent
of the total mass found in the 3.5 to 20 urn size range is shown at the top of
the figure. The percent of "coarse" material varies by about a factor of 2
among the eight samples at each location. The average percent coarse decreases
with height, as expected, being 37 percent at Broad and Spruce and 30 percent
at 500 South Broad. The bar in the middle of Figure 27 divides the total con-
centration into its respirable/nonrespirable (or fine/coarse) fractions. Air-
port wind and precipitation data are shown at the bottom of Figure 30. The
amount of precipitation is indicated by the length of the precipitation bar,
and the period over which the precipitation fell is shown by the width of the
bar. At Broad and Spruce, the heavy rainfall on the 24th appears to have
resulted in a major decrease in the concentrations of both fine and coarse
particulates. It also appears that the concentration of fine particulates re-
mains suppressed longer than the concentration of coarse particulates. At
500 South Broad, the effect of precipitation is less obvious, but perhaps
brought about a decrease in the concentration of coarse particulates. The
data are too limited to justify firm conclusions, however.
Summary of Dichotomous Sampler Data (Elemental Composition)—
It is instructive to compare the concentrations of various elements in
the two particle size categories, and also at the two heights, even though
the sampling periods at the two locations were not the same. This can be done
from Table 11 which presents average concentrations for the eight sampling
periods at Broad and Spruce and for the eight periods at 500 South Broad, and
ratios calculated from these averages. The bottom line of Table 11 was ob-
tained from the total mass concentrations measured by the two dichotomous
sampler collectors. These data show a slight decrease in concentration with
height for the fine particulates and a much greater decrease with height for
the coarse particulates, presumably reflecting a high proportion of coarse
particulates from street level sources, plus the influence of gravitational
settling upon the larger particles of the coarse category.
Much of the information in Table 11 has been rearranged by ordering the
elements according to concentration and the values of the various ratios, and
presented again in Table 12. Brief discussions of some of the interesting
relationships shown by these tables follow.
The top six elements in terms of concentration are, at both locations,
sulfur, silicon, lead, iron, calcium and aluminum. Of these, sulfur has
the highest concentration and, when expressed as SO^, makes up approximately
19 and 25 percent, respectively, of the particulate mass measured by the
dichotomous sampler at Broad and Spruce, and 500 South Broad. Furthermore,
57
-------
PERCENT COARSE
e>u
30
40
30
?n
1
• 1
MEAN _^ | •
• | _ ^ MEA^L
1.1,1,1.1.1.1.1.1 i . i , l •• l i ._!
2/,0 12 14 16 18 20
22 24 26
DAY
28 3/2 4 68 10
12
\" ^ * / / ^ ^ \
LJ—.—L. 4_ _l.i_ I _j J_ j. .L .I 1 j _L_x 1 i I i l i_ l V l
2/K) 12 14 16 18 2O 22 24 26 28 3/2 4 6
DAY
10
Figure 27. Respirable/nonrespirable particulate concentration, found
by EPA Dichotomous Sampler, and accompanying meteorologi-
cal conditions. Shaded part of bar denotes respirable
fraction.
58
-------
n
a
*>K^
60
c
w
2r
<;
CO
H
Z
•
PH
W
«
Q
H
S
O
j~j
M
PH
CO
M
CO
«?
f\t
j
"^4
|2
W
a
w
w
^•j
o
Pi
p-l
CO
H
I— H
^3
CO
w
pi
W
<*"^
1
t— 1
, — 1
rJ
PC)
H
n
o
•H
a
o
c
o
«r4
4-1
n
M
4J
e
D
0
a
8
o
n
o
•H
4J
a
u
C
111
C31
O
o
o
a
o
•H
m
4J
C
CO
o
c
0
o
E
C
c*
\
U
1
c
t
E
c
CN
i
IT
f
0
a
c,
4.
t
5
E
u-
r"
\,
V
, — 1
(J
4-
CO
0
0 [
in
^
T)
2
u
cu
en
M
S
u
_s
1
1
>
j
/M
n)
) •"
1 0
1 4-t
1
J
3
J.
s-~ .,
\
vO
O
0
in
r-4
CT*
O
CM
OO
(T:
CM
•*
^
f—(
O
^
o
CM
00
^
vO
O
rH
O
rH
00
in
3
m
^j-
CM
m
OO
in
£H
cno^^ocnfo^o^OrH^j-oooo
rHr^rHCMrHrHr.OOCN^-CM
OCNrHU-lCNOOCMOOO
vOOOO^^DrH^O^vOr-
rHr^CvlvOr^CNOOineMrOCM
OCMrHvDCMOOcOOOO
oo in vD cr^ in oo m «^ cr* «^" rn rH rH CM fM
n A •»
m oo CM rH *£) o
rH -^ CM m rH O CM in O rH CO
r^ in I-H t** CM W3 rH *o PO I-H c^
rH t— 1 rH rH
m
O vt C30 rH CM
O\OOOOOO"»»-HCM OrOrH
in O VD in CM in P*) i^. co m o
rH rH rH rH CM CM
o->
WO&^C3*EH>F£CKZUc5
<± ON co r*» m vo
rH rH r- O O rH
O O rH O rH O
CM 0 0 0 rH 0^
CM CM in rH O rH
O O rH O rH O
m m ^o -j- o m
O O O O r+ O
S S ^ S S 3
O O O O rH O
CM OO
00 C*^ f*»» CM CM if)
m m
•
rH
vo co \£> in m
N en N ° S S
rH
O 00 CM rH
CM
o
r^
M
rv
cn
4J
&
rH
2
o
H
59
-------
(U
CO
CO
O
CO O
o m
H 3
fr Pi
a cu
o g
M -H
r i |"T|
1
W
CJ
a
8
1
CO
a
o
M
H
o2
r .
g CU
o •<-*
a ^
o ^
CO
>< rl
O
Q O
M
'jpylj
W
p
prf
o
to
a
M
H
CO }
h-l
r . QJ
^ CJ
S 2
W ^
i_3 OQ
M
T)
C
cO
CM
ctf
W O
pq pq
H "d
03
0
M
CO
0
O
m
CO
O
pq
CO
o
o
m
cu
o
3
ex
CO
T3
CO
O
H
pq
cO
0
pq
CO
0
m
cu
0
3
J-i
P.
CO
(•8
cO
o
pq
cu
CO
5-1
cO
0
CJ
CU
•5
tw
-U
C
B
CU CO
t— I
W
O 00
C m
o •>
CJ CO
C
CU
E
CU CO
I—I
W
CN
o m
C vD
o
O co
C
6 r-l
cu <;
r-i
W
o oo
•rl in
d o
pd 1 — 1
4J
C
cu
a --i
CU •<•
I-l
W
0 0
•r-l LO
cO vo
CO cO
pq CJ
r^» i~^
O ON
r-l O
CO
P-i pq
o o
r-l r-l
,0. cu
A | r-r .
ON ON
in CM
OO vO
CU CO
Cti CJ
in ON
CM ^>
CN r-l
•\ •%
r-l r-l
CO
CO r-l
O O
r-l r-l
I-l
t*d
O CO
OO vO
O 0
•H r-l
O vO
O OO
r-l 0
1-1 I-l
pq CJ
ON in
CM O
CM CN
I—I
CJ W
CN m
ON vO
CO CO
CO
W pq
vo in
i-l 0
l—l r-l
cfl
OO r-1
CM O
I—I r-l
r— 1 M
m r-i
PM >
00 CN
f^* 1^
3 co
o <;
oo oo
PM
vO vO
r-l r-l
0 O
rl
> pq
r-l 0
CM- CM
0 O
C CU
S PM
O ON
m <)-
0 O
S co
in in
o o
3 C 5-i
O CO CO
r^ oo
-------
86 and 88 percent, respectively, of the sulfur at these two locations is asso-
ciated with particles less than 3.5 ym. The fact that there is no decrease in
the average concentration with height of the "fine" sulfur concentrations
indicates thorough mixing and transport from nonlocal sources. The decrease
in height of the "coarse" sulfur concentrations suggests street level sources
of at least some of these larger particles.
The next most common element at both locations is silicon. In contrast
to sulfur, however, silicon is found predominantly in the coarser size
category, and the concentration falls off rapidly with height for both fine
and coarse particles, indicating heavy contributions from street sources.
The third most common element at Broad and Spruce, and the fourth most
common at 500 South Broad is lead. Lead, like sulfur, is found predominantly
in the fine particulates, but unlike sulfur shows a rapid decrease in concen-
tration with height in this size category, indicating street-level sources.
The remaining three elements of the top six are presumed to be primarily
of mineral origin, and are most plentiful in the coarser particles. The ex-
pected decrease in height shows up in the concentration of iron and aluminum
in both size categories, indicating low level sources. On the other hand, the
concentration of calcium increases slightly with height in the fine particle
size category and decreases only slightly in the coarse particle size category,
suggesting some major contributions from nonstreet-level sources.
Day-to-day changes in the percent of its mass found in particles less than
3.5 ym in aerodynamic diameter -re shown for these six elements, plus bromine,
in Figure 28.
For both sampling locations it is apparent that the major rock-forming
elements are concentrated in the coarse fraction. These elements are aluminum,
silicon, potassium, calcium, titanium, iron, manganese and possibly strontium
and barium. In the absence of direct determination of the mineral phases
present by microscopic examination, one approach in determining whether these
elements can be reasonably assigned to mineral phases involves both comparison
with crustal average values for each element and assignment of the elements
among the likely mineral phases present. This latter step is in some respects
similar to the technique of normative analysis in petrology wherein the
approximate mineral composition of a rock can be modeled from its chemical anal-
sis according to a formal set of "rules." The similarity of approach cannot
be carried too far because the "rules" of normative analysis are intended to
simulate the sequence of crystallization from a silicate melt. Here, we are
simply partitioning the elements among the likely products of chemical
weathering.
In Table 13, selected elemental abundances are presented, normalized with
respect to silicon, on an atomic basis for comparison with the Continental
Crustal Average. There appears to be general agreement in these numbers sup-
porting a geological source for a major portion of the elemental concentrations;
however, it would be erroneous to carry this analysis further since the local
geology assuredly differs from a crustal average. The results generally sup-
port a geological rather than a manmade source.
61
-------
E
rO
*/>
UJ
O
OC
*
' FOUND
i—
z
2
u>
_j
fc
£
D
or
a
100
80
60
3 40
20
0
80
60
- 40
in
20
0
80
60
(/> 40
20
0
80
60
20
0
BO
6O
" "0
JO
0
80
6O
£ "0
20
O
bO
6O
2O
O
1
6.3
.
I 1 1 • I I 1 » f 1 • L 1*1 1 1 1
/IO 12 14 16 18 20 22 24 26
12 0
. • 0 . .
• • • . •
• i> • .
86 0
1. JllLillJLJItllll
13 6
* ,
* . . * «
liiiiit.itiiiiiii
22 3
» . * . • .
* " "
r • ..• • " • *
830
-
-
9
• .*•*"*
63.6
9 8
•
• • * *
2* s/l 4 6 e 10 a
15.7
*
• • • *
|i j illJ.-l.tlJ i 1
893
It 1 11 KLlll 11J
19 2
' . «
Ll 1 110.. Jt 1-ll-lJ
27 8
* *
. • .
|^
... * *• .
84 5
* • .
• • • •
860
L ,_HX, !,,.,„
26 28 V? 4 6 Q 10 t?
Figure 28. Percent of elemental mass found in respirable particles.
Average values for Broad and Spruce are shown at the
left of the vertical dashed line, and average values for
500 South Broad are shown at the right.
62
-------
•H
CO
o
H
H
CJ
W
CO
PS
EC
H
M
o
W
M
H
•^
S
O
S3
^
CO
o
M
H
pJ
CJ
M
£
8
H
<
,
CO
i — 1
w
§
H
-
(0
O
r-T PQ
cd
4J .
O CO
H
o
m
0)
CJ
3
" a.
r-l CO
cfl
4J <£
O
H TJ
Cfl
5-1
pa
TJ
cd
« o
01 h
CO PQ
j_j
Cfl •
o co
u
o
CD
CJ
3
r4
« &
CU CO
CO
!-4 i-3
cfl
o -a
cj cd
PQ
r— I
cfl
4J
c!#
O 4J
d ca
•r-l 3
C o
CJ
4-1
C
0)
6
a)
r-4.
W
VD vO CO **O
OO r-l O CM
CO r-4 in O
O O O O
CT» CM
CO O CM O
O O O O
OO CM O i— 1
O r-. oo CM
o-
o
r-l
X
CO
^«
in
oo
s^f
1
o
r-l
X
»J"
oo
CM
OO
*^J-
1
o
r-l
X
CM
C
o>
d
•i-4
4-1
C
O
ai
H
c
•H •
vj-
CO v£>
4J CTi
ti r-4
0)
M **
<3) r-l
r-4 00
W CM
r-l
Cfl 1
CJ
•H O
0 00
ai CM
J2 r-l
O
*
14-1 a,
o a
0) •>
O 00
fi CM
tfl
*XJ •*
d to
3 4-1
,r> cj
-------
The principal mineral phases one would expect as weathering products are
shown in Table 14. The basic assumption is that most, if not all, the mafic
(iron-and magnesium-rich) minerals have weathered and that the preponderant
mineralogy consists of quartz and clay minerals with iron present as the oxide
(i.e., hematite).
TABLE 14. PRINCIPAL MINERAL PHASES IN WEATHERING PRODUCTS
Mineral
Clay group
Kaolinite
Illite
Montmorillonite
Quartz
Hematite
. , Cation mol ratios
Composition
K Ca Al
A12 Si205 . 2H20 i
K A13 Si3010 • (QH)2 1-3
CaAl2 SisOm. • nH20 - 1 2
Si 02 -
Fe2 03 _ _
Si
1
3
5
1
-
There remains, then, the assignment of each of these elements to the
probable phases present:
Step 1: Suppose all K present is combined with Al and Si in the
mol ratios 1:3:3 to form illite.
Step 2: The remaining Al is combined with the remaining Si in
the ratio 1:1 to form kaolinite (a clay mineral).
Step 3: The remaining Si is used to form quartz.
Step 4: All Fe present is assigned to hematite. It is probable
that Ti, Mn, and to some extent Ca, would be present in
oxide phases.
Following these steps produces the results of Table 15 giving the weight
of each phase based on the element assignment scheme followed and the weight
percent-abundance of each phase among all the mineral phases.
64
-------
TABLE 15. APPROXIMATE WEIGHTS AND PERCENTAGES
OF PHASES PRESENT
Broad and Spruce 500 South Broad
Mineral
ng % mg %
Quartz
Kaolinite
Illite
Hematite
725
649
20
333
42
37
1
19
350
240
13
162
46
31
2
21
Total 1,727 100 774 100
Comparison of Dichotomous Sampler and Hi-Vol Data—
Concentrations measured by hi-vol on the trailer rooftop were available
for comparison with the results obtained with the Dichotomous Sampler on 4
days. In every case, the hi-vol measured substantially higher concentrations
than did the dichotomous sampler with its upper size limit cut-off of approx-
imately 20 yffl. The percent of the hi-vol determined concentration measured by
the dichotomous sampler during these 4 days ranged from 35 to 43 percent with
an average value of 38 percent. If all tower and trailer concentrations are
averaged for a hi-vol value, more in line with the procedure used in making
the APM hi-vol comparison, the range is from 43 to 53 percent.
Comparisons were also made between dichotomous sampler and hi-vol mea-
surements made on the roof at 500 South Broad during the first 2 weeks in
March. In this case, however, the sampling periods do not match exactly since
the dichotomous sampler was operated from noon to noon and the hi-vol from
midnight to midnight. For the first week, the concentration measured by the
dichotomous sampler between noon Tuesday and noon on Saturday averaged 44
percent of the concentration measured by the hi-vol for the 5-day period from
Tuesday through Saturday. The corresponding value for the second week was
46 percent. Thus, there appears to be little if any difference in the percent
of the total particulates captured by the dichotomous sampler at the two heights.
This very limited set of corresponding hi-vol and dichotomous sampler
data is plotted in Figure 29. As with the hi-vol and APM data presented in
Figure 26, there is excellent correlation between the two measurement tech-
niques, but the hi-vols again consistently measure the higher concentrations.
It is also of interest to note that, based on these two different data sets,
the APM gives absolute values that are intermediate between the dichotomous
sampler and the hi-vol. For example, for a hi-vol concentration of 100 yg/m3,
the regression equations yield an APM concentration of 65 yg/m3 and a
dichotomous sampler concentration of either 41 or 45 yg/m3.
65
-------
X
I
220
2OO
ISO
160
140
120
IOO
^
^ 80
(o 60
4O
20
O
R=0.95, N=6 (USING PTS 0+«)
TSP =15.57+2.79 XTSP
(HI-VOL) (D.S.)
R = 0.95, N=6 (USING PTS X + •
KEY §.
0= TRAILER (I HI-VOL)
X = AVE. TRAILER AND TOWER
• =500 S. BROAD (I HI-VOL)
20 40 50 80 100 120 I4O 160 180 200 220
TSP(>,g/m3) DICHOTOMOUS SAMPLER
'Figure 29. Comparison of dichotomous sampler and hi-vol concentrations
(10 February to 12 March 1977).
66
-------
Operation of NASA Air Scout—
The NASA Air Scout was operated by the Project Officer, William E. Belanger,
on five sampling days during the February field program. A discussion of its
operational performance during this period is presented at the end of Appendix E.
June Field Measurements
The principal change in the experimental design of the measurement pro-
gram from that employed in February was the inclusion of 3 consecutive days
of intensive street washing in the vicinity of the special monitors. Also,
no usable respirable and nonrespirable data were collected during the period.
The data from dichotomous samplers installed at the two locations were inva-
lidated because of collection and analysis difficulties, and no GCA APM was
operated during this phase of the program. The primary experimental goals
were to test the effectiveness of street flushing in reducing nearby ambient
particulate levels and to collect summer data for comparison with the data
collected during the winter season.
The flushing was carried out by the City of Philadelphia on 3 consecutive
days between 7:00 a.m. and 6:30 p.m. along the street segments shown in
Figure 30. On the 3rd day, flushing was held up for 2 hours due to a parade.
The amount of water used was as follows: June 15th, 4.3 x 105 liters
(113,000 gallons); June 16th, 4.2 x 105 liters (111,000 gallons); and June 17th,
3.1 x 10s liters (83,000 gallons). At the end of the 3rd day of street
flushing, the hi-vols at Broad and Spruce were operated on a 4-hour schedule
in an attempt to measure the rate of increase in TSP levels resulting from
accumulating street dust. This 4-hour schedule continued until 2000 EDT on
the 19th. Traffic volume data were taken by automatic counters located on
South Broad and Spruce Streets. More details of the street flushing procedures
and traffic volumes are provided in Appendices H and I, respectively. All
TSP concentrations measured during the field program are tabulated in
Appendix J. These data have been summarized in Table 16 to provide 24-hour
concentrations where possible. To do this, individual observations for the
16th and 17th were proportioned and combined as judged appropriate, and one
extreme 4-hour concentration on the 19th was omitted. Also, because of a
special sandblasting event on the 19th, concentrations for two time periods
are provided for that day.
TSP concentrations preceding, during, and following the flushing ex-
periment have been plotted in Figure 31a and the associated precipitation,
wind speed and wind direction data plotted below in Figures 31b, 31c and 31d,
respectively. Figure 31a contains three curves. The Broad and Spruce curve
shows the average concentration for the trailer and tower hi-vols; the x's
indicate average 4-hour concentrations at this location. The 500 South Broad
Street curve shows the average concentration measured at this rooftop loca-
tion; the number of samples contributing to the average ranged from one on
days when only the AMS hi-vol was being operated, to three on days when both
special hi-vols were also operated. The third curve shows the average concen-
tration measured at the two other sites in the city with approximate daily
sampling; i.e., the AMS Laboratory and the Southeast Water Pollution Control
Plant. The periods of intensive street flushing on the 15th, 16th, and 17th
are also indicated in the figure.
67
-------
1
i
t- JOHN
u
a
t-
u
MARKET
N
I 1
J
j
c
•>
ST
CHESTNUT -*•
• "IT""1
SANSOM i
WALNUT
|
__..,r_.
_J J
LOCUST
SPfluCE [
" "1
J
DELANY
— «•
(•••
c
1
1_J|
F KENNEDY BLVC
tvl
H 3 CITY
w a. HALL
UJ
2? K
t-i tn
"[%
)
t
1-
Ul
UJ
a:
t-
V)
t
5 PENN SQ-»
^1
t
J^
!
i
^L
1
t|_
xx/
(-
— »
*—
/-MO
D
x^4/
1
'V
Nl
SI
X
1
a:
UJ
a
z
•«—
MARKET ST t-
^XXX
TORINO
TE
X/Xx
STREET
t
t
t
— *•
LJ
•« —
— »•
•*—
]
]
en
1
UJ
UJ
K
H
1
i
i
JC
ro
I
1
STREET
sTRETr
STREET
STREET
STREET
STREET
STREET
STREET
STREET
STREET WATERING ROUTE
Figure 30. Area of street-flushing experiment.
68
-------
f-4
CQ
m CM rn o
m to -j
QJ tj cn
69
-------
10
0>
z"
o
H
K
Ul
O
O
O
0.
W
Z
o
l
o
UJ
IE
O.
a
z
300
250 -
BROAD U SPRUCE
(AVG, TRAILER TOP 8
I .TOWER)
LAB 8 S/E (AVG)
I I ! i
500 S. BROAD-ROOFTOP (AVG)
i • I
STREET FLUSHING
I I i I
r.'A : ra ! pra
o)TSP Solid lines connect 24-hour concentrations\ x's indicate
4-hour concentrations at Broad and Spruce Streets
0.5
O.4
0.3 -
0.2 -
O.I
I
I)) Precipitation ot Philodelphia Internationol Airport
c)wind Speed at Philadelphia International Airport
z
o
UM
UJ«
O
z
360
270
180
90
0
1 i i ' ; i i
»-*• ' ' ' Ju***,' * '
~ 1 ^™"«l »| 1*
i 1 '• 1
1 i !.\| |
"i,l,l,i,i,
"6/716/8 16/91 6/101 6/11
1 Tu 1 W 1 Th 1 F 1 So
1
1 ',
. V"*
ill
I'll
•* . * ••••
** *.l l.«~| -
*•• • 1 !*
i •,: i
. i..' i i
6/12 16/131 6/1* 6/15"
Su 1 M I Tu 1 W !
i
""^
1 -
1
L j_ . 1 i
6/16 6/l7| 6/fli 6/,9i 6/2C
Th 1 F i So | Su 1 M |
d)w,nd Direction at 5OO South Broad Street (Roof top)
Figure 31. TSP concentrations and meteorological data for
7 June to 20 June field program.
70
-------
Several features of Figure 31a deserve comment. First, citywide concen-
trations, as indicated by the LAB, S/E, and rooftop monitors, were generally
low through the 13th, increased during the 14th and 15th, gradually decreased
during the 16th and 17th, and were down to approximately 75 yg/m3 on the 18th
and 19th. Also, the average concentration at the LAB and S/E sites and the
rooftop concentrations followed each other quite closely. Second, the average
24-hour concentration at the Broad and Spruce Street site did not appear to be
affected significantly by the intensive daytime street flushing which took
place on the 15th and 16th, remaining roughly 40 yg/m3 higher than the average
citywide index. On the 17th, this difference increased considerably. Third,
the concentrations at Broad and Spruce increased dramatically to the highest
levels observed during the field program immediately following the flushing,
while the citywide concentrations remained low. Finally, the effect of dry
sandbasting in the vicinity is shown by the very high concentration of
888 yg/m3 measured over the 4-hour period from 1600 to 2000 on the 19th.
The failure of street flushing to reduce TSP concentrations at the inter-
section, and the unusually high concentrations measured there following the
flushing were both unexpected and at first puzzling results. The most plausible
suggestion made to data is that the vigorous, forced flushing, plus splashing
by motor vehicles, redistributed particulates that had previously become
concentrated adjacent to the curbs, and that many of these redistributed par-
ticulates became airborne as soon as the street became dry. It will also be
recalled that the* flushing operation was less extensive on Friday due to a
parade, giving additional time for reentrainment. Except for the dry sand-
blasting which was noted on the 19th, no other local source that might cause
the high concentrations was observed.
This hypothesis of reentrainment of the redistributed dried particulates
by vehicular traffic is supported by Figure 32 which compares the volume of
traffic entering the intersection at Broad and Spruce Streets in a 4-hour
period with the average TSP concentration measured at that site. Except for
the measurements affected by sandblasting, the correspondence between the two
curves is surprisingly close. Calculation of the linear correlation coeffi-
cient for the 11 pairs yielded a value of +0.79 which is significant at the
1 percent level. The existence of a stronger street-level source than usual
subsequent to the flushing is also confirmed by Table 17 which shows the
average vertical profile at the Broad and Spruce Street site prior to, during,
and after flushing.
TABLE 17. AVERAGE VERTICAL PROFILE OF TSP CONCENTRATIONS
DURING JUNE FIELD PROGRAM
Height
(m)
11
9
6
4
Prior to
street
flushing
6/7 to 6/14*
78
90
78
81
During
Btreet
flunhlng
6/15 to 6/17
153
160
155
158
After
street
flushing
6/18 to 6/19 (1600)
190
186
233
226
Data for 6/8 have been omitted from the average.
71
-------
10
833
l\
O
X
O
O
to
(E
(U
K
z 4
z
UJ
4-HOUR
TRAFFIC-
VOLUME
6/17 6/18
(F) (Sq)
6/19
(Su)
6/20
(M)
I
300
250
z
o
200
o
z
o
o
a
OT
ISO
20- 00- 04- 08- 12- 16- 20- 00- 04- 08- 12- 16- 20-
24 04 08 12 16 20 24 04 08 12 20 0830
SAMPLING PERIOD
100
Figure 32. Comparison of TSP concentration and traffic
volume at Broad and Spruce Streets.
72
-------
In an attempt to relate traffic volume directly to TSP levels during this
period, the following regression equation was calculated between TSP concen-
tration and volume of traffic using the 11 pairs of data:
TSP = 142.5 + 0.051 VOL (1)
where TSP = particulate concentration in yg/m3, and
VOL = average hourly traffic volume entering the intersection.
Equation (1) indicates that an average traffic volume of 100 vehicles per hour
contributed approximately 5 yg/m3 to the particulate loading at the street
intersection during this particular period. It must be remembered that this
relationship occurred under very special circumstances, and that the results
should not, therefore, be considered typical of usual urban conditions.
The June sampling period was characterized initially by the passage of
two storm centers on the 6th and 9th, which left 1.1 and 2.9 centimeters of pre-
cipitation, respectively, and higher than average wind speeds. As pointed out
earlier, TSP levels at this time were low. The period with rising TSP levels
was associated with light winds and precipitation totaling 0.8 centimeters
which occurred late on the night of the 14th and early on the morning of the
15th. The wind direction veered from east to west twice during this period
in response to changes in a generally weak pressure pattern. Slightly stronger
westerly winds continued for the remainder of the sampling period.
Only 5 days of rooftop data are available from which to judge the effect
of street emissions, and because of the anomalous behavior of some of the PCOP
observations it would be unwise to draw general conclusions from this limited
data base. For example, the rooftop concentrations at PCOP are higher than
any observed at Broad and Spruce on June 12th, contrary to previous experience.
Also, although the rooftop concentrations on both buildings increase progres-
sively from the 12th to the 14th to the 16th in general agreement with one
another, the concentration at PCOP-1 was approximately twice that at 500 SB-1
even on the 16th when it was upwind of Broad Street. It is quite likely that
the rooftop concentrations were influenced by a roofing operation that was
taking place across the alley just the south of PCOP on a building of approx-
imately the same height. The tar heater used in connection with this operation
was located in the alley about 8 meters from PCOP and about 8 meters back from
the sidewalk. The heater was used on Thursday, the 16th, the day when a con-
centration of 312 yg/m3was measured by the nearest PCOP hi-vol, but unfor-
tunately the schedule of other activities associated with the roofing was not
well documented.
Lead concentrations during the June sampling program are listed in
Appendix J with the hi-vol concentration data. Concentrations at the inter-
section site averaged about 1.5 yg/m3 and showed very little variation with
height. Rooftop concentrations averaged about half of the intersection con-
centrations until the week of the roofing operations, at which time they become
quite erratic.
73
-------
OBSERVATIONS NEAR CONSTRUCTION ACTIVITY
A series of hi-vol observations was carried out near the Southwest Sewage
Treatment Plant in December of 1976. This site was selected for the study
because of the extensive earth moving and construction activity connected with
the expansion of the plant. The construction site is just north of the Phila-
delphia International Airport primarily on the west side of the approach to
the 1-95 bridge crossing the Schuylkill River just prior to its confluence with
the Delaware River. A hi-vol site was established at the southwest edge of
the construction activity on the existing sewage treatment plant near the grease
burner building and about 6 meters from the western edge of the construction
area. This was the nearest available location to the work area where power
could be obtained. Another hi-vol site was established at the Mingo Creek
pumping station to the north and slightly east of the treatment plant at a dis-
tance of approximately 1.2 kilometers. The Fort Mifflin complex of the Corps
of Engineers provided a hi-vol site to the southeast of and about 1.6 kilo-
meters from the construction area. The existing AMS site near the airport pro-
vided a sampling location about 0.8 kilometers southwest of the area.
Table 18 shows the concentrations measured at these sites during this
special program and provides the average concentration measured at the AMS
Laboratory and Southwest Water Treatment Plant sites for comparison. Meteoro-
logical data from International Airport are also included in the table. It
can be seen from Table 18 that the highest TSP levels in the area were gen-
erally observed at the monitor closest to the construction activitity (South-
west Station). This is particularly evident on 20 and 23 December when levels
in excess of 300 ng/m3 were measured. It is apparent that the high concen-
tration on those 2 days at the Southwest Station were the result of local
source(s) as the other monitors in the network showed much lower TSP levels.
The data from December 16 illustrate, however, that high TSP levels in the
area can be widespread. Although the downwind site, International Ariport, did
report the highest TSP value on that day, perhaps reflecting the increased
burden caused by the construction activity, the TSP observations at the other
sites, including LAB and S/E, on that day were also high. Clearly, these high
citywide concentrations should not be attributed to the construction activity.
74
-------
H
a>
H
>
M
H
%
Z
O
M
H
U
§
H
W
S5
O
O
Pd
o
fe
CO
S5
o
M
H
H ai a
:» rl
u a
q o
a -o IH ^
w q u oo
—I •? O 1>
^ 1> 5 -a
0) H »-<
et TJ
M
a i
§
w -a
HJ O
•8
fl) (fl bJ
^4 *"•»*
a. a to
fi
0
*oc
§
£
a
->i 4J
(Q £B
u
u
O O> -tf -*
r» CM o oo
0000
s s s s
(N rH CM CS
00 00 i-H ^
MD 1
8
i-H
1 t i I
1 1 f i
OO CO 00 rH
if^ -31 O ^O
O"» r*» in t"->
« • w
3 C OJ 3
fS d. 5 iS
«C VO >0 M3
o S o §
3 2 H H
r^. co O
O O O
o o o
•^ CM m
CM CM m
in <• ^
en ON •»
rl rH
0 O
o o o
0 O 0
^ o ^
O ff* *A
r-f iH
O
r-f
00 ^H ^
oo o er>
O \O >^O
O OO U*i
*"*
O (N r-l
o r**. vj
en
en u
a * \C
r-f r-4 f~j
fN CM (N
•H H H
r»
ON
O
8
in
rH
O
o
0
00
1
CM
^
CO
§
VI
,0
00
i-H
CM
r-l
(*^
M3
O
O
VD
CM
O
u-i
0
O
O
c*>
CO
rH
vX>
rH
i-H
sO
x£>
00
pH
^
\Q
i
MO
O
CM
rH
CM
00
O
0
m
CM
-#
sf
O
0
f^.
CM
O
rH
(A
1
VO
CM
rH
en
O
rl
J8
o
II
«
3 a
75
-------
ANALYSIS OF SELECTED HI-VOL FILTERS
During the period of this investigation, several different samples were
collected in addition to the hi-vol filters in an attempt to demonstrate that
some fraction of the material from the roadway could be identified on the fil-
ters. These collections included street sweepings, and suspended solids from
a puddle created during the street flushing experiment.
Sections taken from hi-vol filters were analyzed by scanning electron
microscope (SEM), equipped with an energy-dispersive spectrometer to provide
major and minor element data for phases present. Hi-vol filter sections were
also examined by petrographic microscope to identify opaque and nonopaque
phases of both natural and manmade origin. An image shearing eyepiece was
used to develop particle size distributions for both opaque and nonopaque
phases present on the filters. From these data, the mass median diameter was
calculated.
Additionally, 10 filters were selected to cover a broad range of darkness,
examined microscopically for differences in the amount of opaque and nonopaque
material, and also extracted with benzene for analysis for the major organic
functional groups present.
For the SEM work, an Advanced Metals Research, AMR-1000 instrument was
used which was fitted with an EDAX Model 700 energy-dispersive spectrometer
for elemental analysis. For optical examination of the filters, a Vickers
Model M72C was used. An image shearing eyepiece was used to make the particle
size measurements.
In the discussion which follows, filters have been designated first by the
general location of the hi-vol (i.e., PCOP for the roof of the Philadelphia
Center for Older People, TRHV for the trailer, 500 SB for the roof of the
Public Health Building, and Tower for the tower at Broad and Spruce), and sec-
ond by a number or letter designating the specific hi-vol. For PCOP and 500 SB,
the 1 indicates the hi-vol nearest the street; at the tower, hi-vols A and B
are both at the lowest sampling height, 6 meters.
Results of SEM Analysis
Six filters and one sample of puddle water were selected for SEM analysis.
Two of the filter samples were chosen for their high particulate loadings; two
were chosen for their high concentrations of one or more opaque substances that
could not be identified by optical methods; and two were chosen because of their
low concentrations of opaques. The puddle sample was collected during the
street washing experiment. The six filter samples analyzed were:
High TSP loadings - PCOP-1 6/16/77 0000 to 2400
Tower-A 6/17/77 2000 to 2400
High opaque concentrations - 500 SB-1 6/14/77 0000 to 2400
TRHV-1 6/14/77 0000 to 2400
Low opaque concentrations - 500 SB-1 6/18/77 0000 to 2400
Tower-B 6/18/77 0800 to 1200
76
-------
Figures 33, 34, and 35 (PCOP-1, 6/16/77) —
These figures were taken at 500X and are typical of particulate masses
seen elsewhere on the filter. A combustion product particle is seen at the
upper right in Figure 33 (probably fly ash). All the fly ash was very high
in S. Elements present in this central mass are Si (high), Al, Ca (high),
Fe (low), K (low), Mg (low), Ti (low). In the same sample a mica (muscovite)
was present (high Al, Si, K and platy habit).
Overall, the sample consisted mainly of mineral particles (quartz, mica,
feldspar), combustion products, some elemental sulfur (crystals) (see
Figure 35) and occasional pollen grains. The elemental sulfur may be due to
a local roof-tarring operation that was underway during the sampling.
Figures 36 and 37 (Tower-A, 6/17/77) —
Figure 36 at 10,OOOX* shows a clu
Al and low Zn. (The fibers are glass fiber from the hi-vol filter).
Figure 36 at 10,OOOX* shows a clustered mass containing high S, Ca, some
Figure 37 at 2000X shows a particulate mass containing Al, Si, S, Cl, K,
Ca, and Fe. Some particles were high in Al and low in Si and may be abrasive
materials from a local sandblasting operation. Most particles present have
about equal Si and Al in them and are probably clays.
Figures 38 Through 41 (Puddle Sample)—
The sample of "puddle water" was collected immediately following a street
flushing operation. When examined by petrographic microscope, the solids
present were largely composed of organic-rich "mats" and fibers, which are
probably cellulose. A more detailed examination was made using the SEM. The
sample was prepared by filtration of 1 ml of shaken water onto a 0.1 ym
Nuclepore filter (polycarbonate).
Figure 38 is typical of the mineral fraction in the sample. Most of this
material is probably a clay such as kaolinite (and chlorite mica possibly).
Several flakes of muscovite (potassium mica) were positively identified by
morphology and by nondispersive X-ray fluorescence. In this photo, 2 mm = 1 urn.
Figure 39 shows the EDAX scan of a central portion of the particulate mass.
The "integrated" composition is an aluminum silicate, with K, Ca, Ti, and Fe.
Figure 40 shows a typical organic mat that was also observed by petro-
graphic microscope. These masses were found (by light microscopy) to contain
many cellulose fibers, though many were not visible in the material examined
by SEM. There is one fiber on the right edge of this organic mat. EDAX showed
this material to contain Ca only. Figure 41 shows a typical fly ash particle.
*
Note:
500X:
2,OOOX:
10,OOOX:
0.5 mm = 1 ym.
2 mm = 1 ym.
10 mm = 1 ym.
77
-------
Figure 33. PCOP-1, 6/16/77,
Figure 34. PCOP-1, 6/16/77,
78
-------
Figure 35. PCOP-1, 6/16/77,
Figure 36. Tower-A, 6/17/77,
79
-------
Figure 37. Tower-A, 6/17/77.
Figure 38. Puddle sample, 2000X,
80
-------
Figure 39. EDAX photograph showing Al, Si,
K, Ca, Fe and Ti.
Figure 40. Puddle sample - organic mat (1 mm = 1 ym)
81
-------
Figure 41. This is a typical fly ash particle which contains
(by EDAX) major S, Ca; minor Fe and Zn; low Ti;
and low Si and Al. Scale 2 mm = 1 ym.
The spectra for the particles on all four samples showed the same peaks,
and generally similar intensity ratios. (Compare Figures 42b, 43b, 44b, 45b).
The spectra for the backgrounds of the first three samples were not only similar
to one another (compare Figure 42a,b,c), but also to the spectra for the par-
ticles. However, a low count rate (900 counts/sec) for the particles indicates
the presence of organic phases, perhaps hydrocarbons, with the opaques.
Figure 46a,b,c shows wide scan EDAX spectra over relatively large areas of
filter (ca. 10 ym2). Refer to captions for specific descriptions.
A series of analyses of hi-vol filters was made to determine whether
basic differences in physical appearance of the filters could be correlated
with differences in their microscopic constitution. We took the most obvious
difference, namely, degree of darkness (or greyness) of the filter, which, by
the way, does not correlate invariably with TSP loading. We have attempted
to identify phase or particle size differences microscopically. The major
opaque and nonopaque phases have been identified, their percent mass abundances
estimated and particle sizes measured. For the latter determination we have
used an image shearing eyepiece fitted to a polarizing microscope to develop
particle size histograms for material observed on the filters. From these
histograms the mass median diameters are calculated.
82
-------
Figures 42 Through 46 (500 SB-1, 6/14/77; TRHV-1, 6/14/77;
500 SB-1, 6/18/77; Tower-B, 6/18/77) —
In each case, EDAX spectra were obtained for opaque particles of interest
as well as for the filter background particulate.
The spectra for the particles on all four samples showed the same peaks,
and generally similar intensity ratios (compare Figures 42b, 43b, 44b, 45b).
The spectra for the backgrounds of the first three samples were not only sim-
ilar to one another (compare Figure 42a,b,c), but also to the spectra for the
particles. However, a low count rate (900 counts/sec) for the particles in-
dicates the presence of organic phases, perhaps hydrocarbons, with the opaques.
Figure 46a,b,c shows wide scan EDAX spectra over relatively large areas of
filter (ca. 10 ym2). Refer to captions for specific descriptions.
Results of Optical Microscopy
A series of analyses of hi-vol filters was made to determine whether
basic differences in physical appearance of the filters, such as the degree
of darkness (or greyness) of the filter, could be correlated with differences
in their microscopic constitution. To do this, the major opaque and non-
opaque phases were identified, their percent mass abundances estimated, and
particle sizes measured. For the latter determination an image shearing eye-
piece fitted to a polarizing microscope was used to develop particle size
histogram for material observed on the filters. From these histograms the
mass median diameters were calculated.
In using the image shearing eyepiece to measure particle size, one can
expect some bias where plate-like particles may tend to lie flat on the filter
which would lead to a size overestimate. Image shearing is accomplished in
the "North-South" direction only and while it can be assumed reasonably that
random particle orientation will occur on the filter surface, the dispersion
of particles sizes about the mean (i.e., standard deviation) will increase as
particles become less equidimensional.
As the initial step in determining why some filters were darker than
others, a series of 10 filters over the period June 12 to June 18 and from
February 8 to February 18 were selected. With the exception of the 4-hour
sample from Tower-B, 6/18/77, all others represent 24-hour samples. Tower-B
results are therefore excluded in the analytical comparisons. Table 19 shows
the "greyness" distribution.
We found, by the way, that the degree of "greyness" did not correlate in-
variably with the filter TSP loading.
83
-------
Figure 42a. An opaque particle at 2000X magnification from
500 SB-1, 6/14/77.
Figure 42b. EDAX spectrum from the particle in Figure 42a shows
peaks for Al, Si, K, Ca, Fe (and Au peaks from the
sample-coating).
84
-------
Figure 43a. An opaque particle at 2000X from TRHV-1, 6/14/77,
Figure 43b. EDAX spectrum from the particle shows peaks for Al, Si,
K, Ca, Fe (and Au peaks from the sample-coating).
85
-------
Figure 44a.
An opaque particle at 2000X magnification from 500 SB-1,
6/18/77.
Figure 44b. EDAX spectrum from the particle shows peaks for Al, Si,
K, Ca, Fe (and Au peaks from the sample-coating).
86
-------
Figure 45a. An opaque particle at 2000X magnification from
Tower-B, 6/18/77.
Figure 45b.
EDAX spectrum from the particle shows peaks for Al,
Si, K, Ca, Fe (and Au peaks from the sample-coating)
87
-------
a, EDAX, 500 SB-1, 6/14/77.
b. TRHV-1, 6/14/77.
Figure 46. EDAX spectra for overall filter. In all three cases the
peaks are for Al, Si, K, Ca, Fe (and Au peaks from the
sample-coating). Spectrum for Tower-B sample was not in-
cluded and would not be representative for overall filter
composition due to slight difference in sample preparative
coating.
88
-------
47SEC
36367IMT
EDflX
c. 500 SB-1, 6/18/77.
Figure 46 (continued).
EDAX spectra for overall filter. In all three
cases the peaks are for Al, Si, K, Ca, Fe (and
Au peaks from the sample-coating). Spectrum
for Tower-B sample was not included and would
not be representative for overall filter com-
position due to slight difference in sample
preparative coating.
89
-------
TABLE 19. RANKING OF HI-VOL FILTERS ON ARBITRARY SCALE OF GREYNESS
Lightest
1 2
3 4
500 SB-1
6/12/77
500 SB-1
6/18/77
567
500 SB-1 500 SB-1
6/14/77 2/8/77
Tower-A
2/08/77
500 SB-1
2/18/77
8
TRHV-2
6/15/77
TRHV-1
6/14/77
Darkest
9 10
TRHV-2
2/18/77
Following this ranking, the five June filters were examined by petro-
graphic microscope to identify the mineral and other nonopaque (glass princi-
pally) phases present, and the opaques. We also estimated the percent by volume
of each phase present. From the mass of total suspended particulates and
assumed densities for the phases, their mass was then calculated. We assumed
here a density of 2.6 g/cm3 for the nonopaque (mineral) phases and 1.0 g/cm3
for the opaque phases. This assumption is quite valid for the mineral phases
since the major phase present is quartz (p = 2.6 g/cm3). Feldspar (p = 2.55
to 2.63 g/cm3) and glass are not remarkably different from the density of
quartz. We are less certain in the assumed density of 1.0 g/cm3 for opaques.
However, these phases had indistinct and irregular grain boundaries for the
most part and are most likely organic. Using this density for the few fly
ash particles present is sufficiently accurate for this analysis. The analyti-
cal results are shown in Table 20.
In this analysis some of the mineral phases, especially quartz and feldspar,
were difficult to distinguish from one another. While their optical crystallo-
graphic properties were very different, the small particle size made it impos-
sible to obtain interference figures even using a research-type petrographic
microscope.
The next step involved determination of the size distribution of particles
in all 10 filters. This was done using the image shearing eyepiece and, for
three samples, a Coulter counter technique. The particle size frequency dis-
tributions are contained in Appendix K. The lower limit for the optical deter-
minations using the shearing eyepiece was about 0.5 ym for both nonopaque and
opaque phases. This limit is not so much a restriction imposed by the equip-
ment as a limitation on the ability to discern grain boundaries of the sub-
micron particles. The Coulter counter technique has a practical lower limit
of 1 micron. The sample statistics for these filters, plus those for a sample
of road sweepings, are shown in Table 21.
90
-------
PM
O
O
CO
O
QJ
M
M
CM
^i
S
O
^
pi
H
W
PH
>•*
PQ
OO
pyj
w
H
S
,_j
O
1
M
EC
PL,
O
CO
H
cn
^-t
._]
*^
[Z
5
•
o
CM
PJ
l-J
""^
H
*
PQ
00
rH ^
^- 01
o
H
rH
"3" 1
rH >
^o PG
H
1
00 M
i — 1 CO
^^s.
\0 O
o
m
I
rt
vj- CO
r— |
^^ CO
VO O
in
rH
1
CM PQ
rH CO
•• — ,
^D CD
O
in
/~\
cn
cu
rH
ft
c a
o cd
cu -H en
4»J -O
o) o) >-<
n o 3
0 0
rH 42
1
-^-
CM
^-^
O r- rH /—.
rH r^ rH CM
^—- %_/ v_^ v-x
ro OTi CO rH
o-i in r-T r-^ in
m CM o. CM
CM rH
C? ^
i-O "J*
v_x v-- CU
T3 O
rH O CB
• • d M
. co o
rH LTl Csl iH
S^X ^^^ v^X *^S i^-J
•
00 -.
4-1
CU -H
rH CO
ft C
a cu
cn
-d
t-l CU
3 a
0 D
EC cn
I cn
* -f-
w
0)
cn
CU
4-1
C
cu
J-l
cd
ft
c
•H
4-1
pj
60
•H
CU
^
^
XI
CU
CJ
C
cd
-a
ci
3
FQ
cd
4-1
C
0)
o
cu
•n—
•
a
CJ
— ^,
60
vO
•
CM
II
>-,
4J
•H
en
C!
cu
t3
•o
CU
g
cn
CO
-------
TABLE 21. MEAN DIAMETER, STANDARD DEVIATION AND NUMBER OF PARTICLES COUNTED
FOR NONOPAQUE (MINERAL) AND OPAQUE HI-VOL FILTER PHASES
Nonopaque
a. Optical Microscopy
500 SB-1, 2/8/77
Tower-A, 2/8/77
TRHV-2, 2/18/77
500 SB-1, 2/18/77
TRHV-2, 6/15/77
500 SB-1, 6/12/77
500 SB-1, 6/14/77
500 SB-1, 6/18/77
TRHV-1, 6/14/77
Tower-B, 6/18/77 (4 hr)
500 SB-1, Road Sweepings
b. Coulter Counter
500 SB-1, 2/18/77
TRHV-2, 2/18/77
TRHV-1, 6/14/77
3
3
2
3
5
2
4
5
2
2
16
X
.4
.80
.0
.5
.79
.76
.78
.02
.70
.55
.25
a
6.
3.
1.
2.
3.
1.
3.
5.
2.
2.
17.
6
97
3
8
75
74
51
28
19
66
32
n
100
108
101
128
101
102
100
106
110
171
120
X
7.
3.
5.
6.
4.
2.
3.
6.
3.
5.
19.
Count
0
47
4
0
56
65
28
11
10
25
70
Opaque
a
6.
3.
6.
4.
4.
1.
2.
4.
1.
7.
16.
"Grey"
scale
n
1
38
1
3
18
60
62
20
92
26
74
100
100
100
97
100
356
273
186
255
116
101
median diameter
7
6
10
.8,
.5
.0
7.5
7
7
9
7
8
3
5
3
8
-
-
(ym)
Based on the volume of an equivalent sphere.
92
-------
Benzene Soluble Organics and Infrared Spectra
Each of the 10 filters that underwent microscopic examination was also
extracted with benzene and the infrared spectra of the extracts were examined
to determine major organic functional groups. This was accomplished by fold-
ing a measured portion of the filter into a bundle and placing in a thimble and
soxhlet extractor.
The boiling flask was charged with 80 mis of benzene and the filters were
extracted for 6 hours. At the end of the extraction, the solution was con-
centrated to approximately 5 mis. This was done by removing solvent from the
extraction cup using an aspirating system and collecting the benzene in a
liquid nitrogen trap.
The concentrated extract was filtered through a coarse porosity fritted
glass filter and the original flask and filter were rinsed three times with
small amounts of CHC13. The filtered samples and CHC13 rinses were transferred
to weighed scintillation vials. The remaining solvent was removed by placing
the samples in an oven maintained at approximately 60 C. Drying took about
24 hours. The samples were weighed when dry, reheated and weighed until con-
stant weights were obtained.
Infrared analysis was done by redissolving the sample in a small amount
(about 0.5 ml) of benzene and using a Pasteur pipette to apply the solution to
KBr plates. The plates were then placed under a heat lamp until dry. Spectra
were taken on a Perkin-Elmer 457 Infrared Spectrophotometer.
Table 22 presents the results of the benzene extractions and Table 23
summarizes the infrared spectra analyses. The concentration of benzene soluble
ranged from 0.5 to 5.9 yg/irr while the benzene soluble portion comprised from
0.2 to 5.7 percent of the TSP. Interestingly, the highest TSP levels (253,
164, 149) corresponded to fairly low percent benzene solubles. The highest
benzene soluble content was observed on June 14 at the Broad and Spruce trailer.
This must have been caused by local sources because the 500 South Broad Street
location recorded about one-tenth as much benzene solubles during the same
period.
The intensity of the infrared spectra and even the functional groups
identified therefrom seem to have little correlation with the amount of ben-
zent solubles present nor with the TSP levels. The filter with the highest
benzene soluble content did also exhibit a strong spectrum, but the only other
sample that gave a strong spectrum was from the filter of lowest TSP value and
below average benzene soluble content. The types of organic compounds indi-
cated are very similar for most filters with the exception of 500 South Broad
on February 8, and Broad and Spruce trailer on February 18, both of which
show minimal organics. The benzene extracts of the other filters show
several functional groups indicating a complex mixture of organic compounds.
93
-------
TABLE 22. RESULTS OF BENZENE EXTRACTIONS OF SELECTED FILTERS
Site
500 SB-1
Tower -A
500 SB-1
TRHV-2
500 SB-1
500 SB-1
TRHV-1
TRHV-2
500 SB-1
Tower-B
Date
2-8-77
2-8-77
2-18-77
2-18-77
6-12-77
6-14-77
6-14-77
6-15-77
6-18-77
6-18-77
TSP
(yg/m3)
76
94
108
164
48
82
104
149
74
253
Benzene
yg/m3
3.0
3.0
0.5
1.7
1.5
0.6
5.9
1.5
0.9
0.5
solubles
% of TSP
4.0
3.2
0.5
1.0
3.1
0.7
5.7
1.0
1.2
0.2
94
-------
TABLE 23. SUMMARY OF INFRARED ANALYSIS
Site and date
General intensity of
spectral transmittance
Indicated functional groups
500 SB-1
2-8-77
Tower-A
2-8-77
Very weak spectra
Medium spectra
500 SB-1
2-18-77
Medium spectra
TRHV-2
2-18-77
500 SB-1
6-12-77
Very weak spectra
Strong spectra
500 SB-1
6-14-77
Medium spectra
TRHV-1
6-14-77
Strong spectra
Methylene group
Methyl, methylene group(s) aliphatic
ketone, six or larger membered cyclic
ketone, aliphatic aldehyde, formate,
a-3 unsaturated or aryl ester (m)
aryl, alkyl or vinyl ether (m) alipha-
tic aromatic ethers (m) substituted
aromatics (w).
Methyl, methylene groups (m) aliphatic
ketone, six or larger membered cyclic
ketone, aliphatic aldehyde, formate,
a-B unsaturated or aryl ester(m).
Aryl, alkyl or vinyl ether(m). Ali-
phatic, aromatic ethers(w) substituted
aromatics (w).
Methyl, methylene groups(w) aryl,
alkyl or vinyl ethers (vw).
Aromatics, hetero-aromatics(s)
methyl, methylene groups(s) aliphatic
ketone, six or larger membered cyclic
ketone, aliphatic aldehyde, formate,
ct~3 unsaturated or aryl ester(s).
Alkene conjugated with aromatic ring(w).
Aryl, alkyl or vinyl ether. Aliphatic
ether, aromatic ether(m) substituted
aromatics(m) phthalate ester(w).
Methyl, methylene groups(s) aliphatic
ketone, six or larger membered cyclic
ketone, aliphatic aldehyde, formate,
a-g unsaturated or aryl ester(m) alkene
conjugated with aromatic ring(w)
phthalate ester(w) aryl, alkyl or vinyl
ether (m) aliphatic, aromatic ethers(w)
substituted aromatics(w).
Aromatics, hetero-aromatics(s) methyl,
methylene groups(s) aliphatic ketone,
six or larger membered cyclic ketone,
aliphatic aldehyde, formate, a-|3 un-
saturated or aryl ester(s) alkene con-
jugated with aromatic ring(w) aryl,
alkyl or vinyl ether (m) aliphatic,
aromatic ether(m) substituted aromatics(s)
phthalate ester(w).
(continued)
95
-------
TABLE 23 (continued).
Site and date
General intensity of
spectral transmittance
Indicated functional groups
TRHV-2
6-15-77
500 SB-1
6-18-77
Medium spectra-
not well resolved
below 1300 cirT1
Medium spectra-
well resolved
Tower-B
6-18-77
Medium spectra
Methyl, methylene groups(m) aliphatic
ketone, six or larger membered cyclic
ketone, aliphatic aldehyde, formate,
a-f? unsaturated or aryl ester (w) aryl,
alkyl or vinyl ether(w) aliphatic,
aromatic ethers(w).
Methyl, methylene groups(m) aliphatic
ketone, six or larger membered cyclic
ketone, aliphatic aldehyde, formate,
a-|3 unsaturated or aryl ester(m) aryl,
alkyl or vinyl ether(m) aliphatic
ether(w) substituted aromatics(w).
Methyl, methylene groups(s) aliphatic
ketone, six or larger membered cyclic
ketone, aliphatic aldehyde, formate,
ct-g unsaturated or aryl ester(s) aryl,
alkyl or vinyl ether(s) aliphatic, aro-
matic ether(m) substituted aromatics(w).
96
-------
Discussion of Results
Particle size frequency diagrams for these samples determined optically
are shown in Appendix K. Each figure gives a histogram for the nonopaque or
mineral phases plus glass, and for the opaque phases which would include orga-
nic particles, fly ash and possibly corrosion products. Corrosion products
were not identified specifically in any of the specimens examined, however.
For all samples examined, the particle size distribution remained below about
18 ym and in only a few cases were there larger particles. The reason for
finding such a limited size population is explained below. The percent fre-
quency distribution and the calculated means and standard deviations are
based on the total population observed - the figures present a "window" on
the most prevalent size distribution.
There are several possible sources of bias in particle size determina-
tions by optical means. In performing the measurement, each particle is
viewed on a plane and each particle is sheared in a "north-south" direction.
Therefore, one must assume that the particles are randomly oriented in the
plane of the filter. This assumption is probably valid. One must further
assume that elliptical particles are caught on the filter randomly oriented
rather than with the major axis more or less normal to the plane of the filter.
The fact that some depth filtration occurs within the filter may help in pro-
ducing randomly-oriented particles. Since the shearing measurement is made
in one direction (N-S) only, the measurement will lie between the major and
minor axis lengths for an ellipse or projection of a nonspherical particle.
Therefore, particle populations of low sphericity will exhibit larger standard
deviations about the mean than for spheres. This is, in part, reflected in
the high coefficients of variation for most of the distributions shown in
Appendix K. Bias on the part of the microscopist usually relates to the count-
ing of every particle in a given field. One would expect that the smaller
grains would be missed rather than the larger ones. To avoid this, we have
counted all particles visible in each of many randomly-selected fields. At
high particle densities per field, however, the counting of every particle
becomes extremely tiresome and undoubtedly leads to error. As stated earlier,
we expect that the bias would be toward measuring larger particles, although
comparison with the Coulter counter results would indicate the opposite.
Since the latter technique measures an equivalent spherical diameter, there
may indeed be a preferred orientation with the long axis of particles normal
to the plane of the filter. From the histograms in Appendix K, the mean and
standard deviations were calculated and compared with those obtained by the
Coulter technique (Table 23). Mass median diameters derived from the data in
Appendix K are shown in Table 24.
97
-------
TABLE 24. MASS MEDIAN DIAMETERS OF SELECTED
PARTICULATE SAMPLES IN ym
a) From hi-vol filters (by optical microscopy)
Trailer
Tower (6M)
500 SB-1
Date
Opaque Nonopaque Opaque Nonopaque Opaque Nonopaque
2/8/77
2/18/77 13.2
6/12/77
6/14/77 9.2
6/15/77 12.7
6/18/77
Mean 11.7
11.7 5.0 12.2
6.5 11.7
8.2
8.8 9.6
6.6
10.7 10.2 13.5
7.3 11.2 7.6 ll.O
8.7
6.0
5.6
6.8
8.0
7.0
b) From hi-vol filters (by Coulter Counter)
Date
Trailer 500 SB-1
2/18/77 23
6/14/77 14.5
Mean 18.8
28.5
For size distribution < 20 ym. (See text).
A far more significant problem came to light after most of the smaller
(about 2 cm on the side) filter segments had been examined. We were not
observing particle diameters greater than about 20 to 25 urn, a result at
variance with the dichotomous sampler weight for the two size fractions up
to 20 ym and the hi-vol filter weights. A comparison of these results indi-
cated that a substantial fraction consisted of larger particles. We then
examined much larger filter areas, on the order of 25 cm2 to determine the
few, though mass significant, particles which must have been present. A
number of particles were found for both opaque and nonopaque phases. Un-
fortunately, varying amounts of these larger particles (> 25 ym) had dislodged
from some of the filters and had collected in the envelope. For those filters
which remained intact, the increased counting area gave mass median diameter
results reconciling the hi-vol and dichotomous sampler weighings.
98
-------
The results for the nine reexamined filters are shown in Table 25 for
both opaque and nonopaque phases for size class intervals from 20 to 109 ym
in diameter. The results are normalized to equal areas of 25 cm2.
TABLE 25. SIZE DISTRIBUTION OF OPAQUE AND
NONOPAQUE PARTICLES (ym) NOR-
MALIZED TO 25 cm2
Size class
interval
(ym)
20-29
30-39
40-49
50-59
60-69
70-79
80-89
90-99
100-109
Average number
Opaque
17.8
12.8
5.9
2.7
0.6
0.2
0.2
-
-
of particles
Nonopaque
9.6
6.6
1.8
0.9
0.5
0.1
0.1
-
0.1
Attempts were made to correlate variations in particle size with the
height of the hi-vol sampler and with variations in rainfall. The coeffi-
cients of variation for the particulate measurements are sufficiently large
that no correlations appear warranted with any reasonable statistical cer-
tainty. This is due not only to errors in particle size measurement but also
to the stochastic nature of the process of suspension of roadway dust.
99
-------
SECTION 5
MODEL DEVELOPMENT
The purpose of this task was to develop a modeling procedure for the
Philadelphia area which would provide the best state-of-the-art breakdown
by major source categories of the TSP levels observed within the city. This
information could then be used to evaluate the effectiveness of any additional
control measures that might be judged feasible. The starting point, and basic
tool, for the model development work was the Air Quality Display Model. With
the exception of fugitive dust emissions, the emission inventory used with the
model was one developed under EPA Contract No. 68-02-1376, Task Order No. 24.
Under this task, GCA had updated the TSP point source inventory for the Metro-
politan Philadelphia Interstate AQCR, developed a regional TSP area source
emission inventory, and carried out preliminary TSP modeling. From this
modeling, it was clear that a more refined emission factor for reentrained
street dust was needed. The selection of this emission factor was made follow-
ing exploratory calculations at the tower and 500 South Broad Street sites.
Deviations from the "rooftop" model estimates were then examined in the light
of probable contributions from local sources, with special consideration being
given to the Broad and Spruce Street site.
TREATMENT OF FUGITIVE DUST IN MODEL CALCULATIONS
Although attempts to validate the Air Quality Display Model (AQDM) using
updated emission inventories and observed annual average concentrations for
all AQCR monitors under the earlier contract had been unsuccessful, satis-
factory agreement was obtained between calculated and observed values for a
New Castle - Philadelphia data subset by omitting fugitive dust and accepting
a "background" concentration of 48 ug/m3. Figure 47, which is based on
Figure 15 of the Final Report (EPA 903/9-77-030) submitted under that contract,
shows a plot of these results, and the regression equation. The CAMP station
(monitor FRI) was assumed to be significantly influenced by local parking lot
and traffic sources and was omitted from the regression analysis. Thus, it
appeared that significant improvement in the Philadelphia model could be
achieved by a more realistic handling of roadway emissions. The reintroduction
of these emissions, if more correctly estimated, could be expected to reduce
the background estimate without greatly altering the correlation between
observed and calculated values.
Details of the procedures used in the development of the particulate
emission inventory for the Metropolitan AQCR are given in the Final Report
under Contract No. 68-02-1376 with the exception that the method used to
allocate VMT within Philadelphia County was carried out in more detail. The
100
-------
AFRI
o
o
o
CJ-
o
o
CD
O.
R = 0.869
TSP = 48.2 + 1.01 TSP calc
AlNT
ADEF
O
o.
00
UJo
LU*0
(n
OD
o
o.
O
O
CD.
CM
O
O
10.00 20.00 30.00 40.00 50.00
CRLCULflTED (UG/M3)
Figure 47. TSP calibration for the New Castle and Philadelphia monitors.
Philadelphia monitors are indicated by their usual identifiers.
101
-------
procedure used initially to develop VMT estimates within the whole AQCR was
based on fuel consumption.
To estimate the amount of gasoline and diesel fuel consumed, Highway
Statistics8 was utilized to obtain state totals for on- and off-highway fuel
usage. The state on-highway fuel consumption was then allocated to the coun-
ties based on automobile and truck registration. Gasoline was further allo-
cated to light- and heavy-duty vehicles using the assumption that 89 percent
of the gasoline is burned by light-weight vehicles as recommended in Guide-
lines. Volume 7. For the nine counties in the Delaware Valley Regional
Planning Commission, annual vehicle miles traveled (VMT) were available and
broken down into freeway and nonfreeway VMT. These data were compared to the
VMT calculated from the on-highway fuel consumption in each county. The VMT
was in good agreement except for Philadelphia County where the calculated VMT
was considerably higher than the DVRPC estimates.
The county fuel data were allocated to subcounty grids using the CAASE
methodology. The methodology provides an objective method for allocating
county level fuel data to subcounty grids, and calculating the emissions.
The objective apportionment parameters for each type of fuel are documented
in Guidelines, Volume 8.10
In the original study, the two VMT estimates for Philadelphia county were
brought into approximate agreement by applying an adjustment factor of 0.67
to the estimate based on on-highway fuel consumption to account for city
driving, as suggested in the "OAQPS Guidelines AEROS Manual Series, Volume 2."
The methodology used in the present study involved the conversion of an
available distribution of VMT on a zone basis to the Philadelphia grid. The
following briefly describes the procedure. The grid plot was laid over a
zone map of the same scale. Each zone was subdivided by the estimated per-
centage of its area which fell within each grid. A Philadelphia highway map
was referenced to identify where in each zone highway links were located.
The VMT allocated to each grid was then computed as the product of the follow-
ing three variables: (1) the percent of zonal area located within the grid,
(2) a factor for adjusting for the actual location of highway links within
the zone; and (3) total VMT assigned to the zone. These computations were
performed for both freeway and nonfreeway VMT. In those cases where a grid was
composed of portions of more than one zone, the total grid VMT was the sum of
each allocation.
Emission rates of particulates from roadways are typically expressed in
mass per vehicle (or axle) per unit length of roadway. Although estimates
of emission rates, based on limited observations, are available for a number
of situations, it is recognized that emission rates and the size distribution
of the emitted particles may vary widely within any urban roadway network as
a result of differences in a number of factors. Among the more important of
these factors are: (1) amount, distribution, and physical properties of the
roadway dust and dirt; (2) geometry and exposure of the roadway configuration;
(3) vehicle mix and speed; and (4) meteorological conditions. However, the
necessary quantitative relationships between emission rates and these factors
102
-------
have not yet been sufficiently well established to justify the major effort
that would be required to prepare a detailed spatial emission inventory of
roadway dust. Instead, the relevance of an assumed average emission factor
has been judged by comparing calculated concentrations with the experimental
measurements at Broad and Spruce Streets and with components of annual average
concentrations allocated by source category.
Some idea of the tremendous range in street dust emission rates can be
gained from a Final Report Draft prepared by Midwest Research Institute.
In this report, the following relationship is proposed for the calculation
of emission factors, and values of the independent variables judged to be
appropriate for a number of roadway conditions are given:
e = KLs (2)
where e = emission factor (kg/km vehicle)
K = proportionality constant (vehicle"1)
L = surface loading (kg/km)
s = silt content of the surface material (fraction)
To obtain an emission factor for use in our exploratory calculations, the
following assumptions were made:
1. The particulates of concern have diameters <^ 5 ym.
2. K = 3.00 x 10~5. This number was obtained by averaging
the values for the 37th Street and Fairfax Trafficway
sites given in Table 18 of the MRI Draft Report.
3. L = 165 kg/km. This is the weighted average found for
commercial areas. (Table 1 of the MRI Draft Report.)
Table 1 indicates that this loading should be increased
by a factor of 4.1 for residential areas, and by a factor
of 9.7 for industrial areas.
4. S = 0.085. Reported values of the silt content of surface
dust from paved streets typically range from 5 to 15 percent.
Entering Equation (2), with the above values, results in an emission
factor (e) of 0.42 g/km-vehicle which was used in exploratory calculations
at the Broad Street test site.
Exploratory Calculations at Test Site
From the start of the program, it had been recognized that although the
Broad Street test site is representative of a large class of city intersections,
it deviates greatly from the "ideal" site originally used in the development
and validation of line source models. Further complications in modeling
street-level particulate emissions in downtown areas arise from the difficulty
of either measuring or reliably estimating background concentrations, and in
assigning appropriate emission factors. For these reasons, it is not at all
103
-------
obvious that conventional modeling procedures can be of much use in evaluating
street contributions and control strategies at this, or similar locations.
Such locations are, however, quite typical of downtown traffic-oriented sites
which experience violations of the particulate standards.
To get some quantitative feel for the appropriateness of an emission
factor of 0.42 g/km-vehicle near the test site, and for model calculations, a
number of exploratory calculations were carried out. The approach taken was
to assume that the street contribution to concentrations measured at the tower
is simply the difference between the tower concentrations and the rooftop con-
centration measured at 500 South Broad Street. It was further assumed that
the street contribution came from both Broad and Spruce Streets and that emis-
sions were directly proportional to traffic volume. Current information was
judged to be insufficient to warrant making adjustments in the emission rates
of either reentrained street dust or tailpipe particulate emissions according
to vehicle speed or operating mode. The exploratory calculations at the Broad
and Spruce Street intersection included the application of a Gaussian line
source model and a box model first for a 24-hour period, and then for a 1-year
period. The "cut-section" submodel of HIWAY was used to calculate the average
annual impact of reentrained dust at rooftop level at 500 South Broad Street
and at the Philadelphia Center for Older People.
February 8, 1977, a day with westerly winds and the requisite hourly
traffic and meteorological data, was chosen for the 24-hour exploratory calcu-
lations. The emission rate of 0.42 g/km-vehicle, developed from data in the
draft MRI report * was used for reentrained dust. AP-42 emission factors of
0.21 and 0.12 g/km-vehicle were used respectively for tailpipe exhaust and
tire wear. It was assumed that the source height of these emissions was 1.0
meters above street level and that az was equal to 1.5 meters at the source.
An adaptation of the HIWAY model (Intersection-Midblock model) was run for
each hour of the day for four receptor heights on the tower and the results
averaged over the 24-hour period. The solid curve in Figure 50 is the result.
The rapid decrease in concentration with height is to be expected in view of
the assumed vertical dimension at the source and the fact that no adjustment
in stability class was made to account for additional turbulence induced by
the building structures.
Using Figure 48, a comparison can be made between the calculated amount
of traffic-related particulates and the observed amount as indicated by the
difference in concentration between the tower and the rooftop at 500 South
Broad Street. This concentration difference, estimated to be 18 yg/m3 over
the height of the tower, is indicated in the figure by the dashed vertical
line. The approximate agreement between calculated and observed flux of par-
ticulates by the tower is shown by the close equivalence of the two shaded
areas. However, the fact that concentrations at the top of the tower were
substantially higher than those at the equivalent rooftop height indicate that
the street contribution extended to even greater heights. This suggests that
the true vehicular-related emission factor was somewhat greater than the
assumed 0.75 g/km-vehicle (0.42 + 0.21 + 0.12).
104
-------
8
LJ
OBSERVED DIFFERENCE
BETWEEN 500 S. BROAD
AND TOWER
I V /—CALCULATED
|_ \f CONCENTRATION
AT INTERSECTION
40
30
20
o
UJ
X
10
0
0 20 40 60 80 100
TSP CONCENTRATION >
Figure 48. Comparison of calculated and observed vertical TSP
profiles on 8 February 1977.
105
-------
The quite thorough vertical mixing that takes place within a layer at
least as deep as the tower suggests that conditions near the intersection might
be more realistically represented by a box-type model instead of a Gaussian
model. To apply a box model on February 8th, the following assumptions were
made:
• The height of the box is equal to the height of the tower (12 m).
• The box is flushed by the average rooftop wind speed (2.3 m/sec).
• Traffic-related emissions are contributed at a rate of 0.75 g/km-
vehicle by all vehicles entering the intersection of Broad and
Spruce Streets during the day (39,000 vehicles).
With these assumptions, the following calculation provides an estimate
of the average concentration of traffic-related particulates within the box
on this day:
Mass added
Concentration = —rr-^
Volume
39,000 x 750/24 * 3600 .„ , 3
= 12 x 2.3 x 1 ' 12 ^/m
which is in fair agreement with the estimate of 18 yg/m3 obtained from the
observations. The use of a deeper box than the assumed 12 meters would, of
course, reduce the estimated concentration proportionately. Agreement between
the two estimates can be achieved by increasing the vehicle-related emission
rate by 50 percent (e.g., by increasing the emission rate for reentrained dust
by 89 percent).
To obtain some further idea of the suitability of the 0.42 g/km-vehicle
emission rate for reentrained dust in combination with the Intersection-
Midblock Model, the model was applied to the tower site over a 1-year period,
arbitrarily using a 2-meter receptor height. For this calculation, the STAR
meteorological summary for 1974 was used. Since no distinction is made as to
hour of the day in the STAR summary, the average daily traffic volume was
assumed to have been distributed evenly over a 24-hour period. Also, traffic
counts from the February 1977 field program were assumed to represent average
travel during the year. The result of these calculations was an annual aver-
age concentration of 37 yg/ra3. This is in good agreement with the annual
differences of 43 and 44 yg/m3, respectively, experienced between the 500 South
Broad Street site and the Broad and Spruce Street site in 1975 and 1976.
However, judging by the concentration profile shown in Figure 50, the calcu-
lated 2-meter concentration should have exceeded the annual difference by
about a factor of two for good agreement. Leaving tailpipe and tire wear
emissions unchanged, this could be accomplished by increasing the emission
factor for resuspended dust by a factor of 3.
The box model was also applied to the tower site on an annual basis making
the assumption that emissions from each of the adjacent streets affect concen-
trations at the tower site when the wind blows across the street and toward
106
-------
the tower. Specifically, it was assumed that: (1) emissions from Spruce
Street fully affect concentrations at the tower site when the average wind
direction is from ESE through South to W, inclusive, and (2) emissions from
South Broad Street fully affect concentrations at the tower site when the
average wind direction is from SSW through West to N, inclusive. Using a
12-meter box height, wind speed and direction data from the Philadelphia
International Airport, and a vehicle-related emission factor of 0.75 g/km-
vehicle gave the results shown in Table 26 with an average annual contribution
of 6 yg/ms. Traffic volumes on Spruce and Broad Streets were assumed to be
12,000 and 27,000 vehicles per day, respectively, and the wind speed - wind
direction bivariate distribution was based on all observations made during
the year. Thus, the possible effect of diurnal variations in traffic flow
and meteorology on the resulting estimates was not investigated.
The estimates of vehicle-related contributions calculated on an annual
basis are clearly far below the 43 yg/m3 value given by the difference between
the annual average concentrations observed at the trailer and the roof of
500 South Broad Street. Before attributing all of the underestimate to too
low an emission rate, however, daily concentration differences between the two
sites were examined for speed and directional effects in conformance with the
model predictions.
TABLE 26. BOX-MODEL ESTIMATES OF ANNUAL TRAFFIC-RELATED
CONTRIBUTIONS TO PARTICULATE LEVELS AT TOWER
SITE AS A FUNCTION OF WIND DIRECTION
Wind direction
N
NNE
NE
ENE
E
ESE
SE
SSE
;V
Concentration
(yg/m3)
6
-
-
-
-
3
3
3
Wind direction
S
SSW
SW
WSW
W
WNW
NW
NNW
*
Concentration
(yg/m3)
3
8
8
12
9
5
6
6
*
Based on an emission rate of 0.75 g/km-vehicle (0.42, reentrained
dust; 0.21, tailpipe; 0.12, tire wear).
Table 27 gives linear correlation coefficients calculated between the
difference in concentration at the two sites and the average airport wind
speed for the day for four precipitation - time of week categories.
107
-------
TABLE 27. CORRELATION COEFFICIENTS BETWEEN ATSP (BROAD &
SPRUCE - 500 S. BROAD) AND WIND SPEED
48-Hour
precipitation
r
n
Weekday
0.10 in.
- 0.42 - 0.04
35 31
Weekend
£ T >_ 0.10 in.
- 0.22 - 0.22
16 19
Of these, only the correlation coefficient between wind speed and ATSP
on weekdays with a trace or less of precipitation is significant at the 5
percent level. Thus, at least on a 24-hour basis, the magnitude of the
vehicle-related contribution bears little relationship to the airport wind
speed.
The difference in concentration at the two sites has been plotted against
wind direction by category in Figure 49. Of the four plots, Figure 49a,
comprising weekdays with precipitation equal to or less than a trace, shows
a directional effect upon the concentration difference. With winds from
south through west, the concentration at Broad and Spruce averaged about
30 yg/m3 higher than with winds from the west, north and east. This suggests
a greater impact from traffic-related sources with southwesterly flow, as
would be expected from the location of the street intersection with respect
to the Broad and Spruce Street monitor. With this exception, and that of rainy
weekends with northeasterly winds when the difference decreases to about
20 yg/m3, concentrations at Broad and Spruce average roughly 40 yg/m3 higher
than those at 500 South Broad Street regardless of wind direction. This is
approximately equal to the combined contribution calculated by the box model
from both South Broad and Spruce Streets with a wind speed of 0.7 m/sec.
The same emission rate (0.42 g/km-vehicle) was also used with the "cut-
section" submodel of HIWAY to calculate the average annual impact of re-
entrained dust at rooftop level at 500 South Broad Street and the Philadelphia
Center for Older People. The resulting concentrations were: 2.6 yg/m3 at
500 South Broad Street, and 9.4 yg/m3 at the Philadelphia Center for Older
People. The difference in average concentration between the two sides of
Broad Street reflect the prevailing westerly winds. Although there are no
annual values with which these model estimates can be compared, the estimates
are in approximate agreement with the average downwind values experienced
during the February field program.
Conclusions
In summary, a comparison of the results of test-site calculations to above
observed concentrations suggests that:
• When physical site characteristics are similar to those at
the Broad and Spruce Street intersection, box-type models are
108
-------
10
£
^
0.
I4O
120
100
60
60
40
20
o
i i i i i ill
• • *
*
• • . %
.'.'••' ' "
DIRECTION OF INTERSECTION-n * *
1 1 i i i i ii
40 80 120 160 200 240 280 320 360
WIND DIRECTION, degrees
(o) WEEKDAY, PRECIPITATION ,< T
•o
E
3
OL
(/)
H
I4O
120
too
80
60
4O
20
o
-?o
II 1 1 i 1 1 1
-
• *
• • -
•
• • • *
• I. • ,
"• •
-, 1 1 1 1 1 1 1 1
O 40 60 120 ISO 200 240 280 320 360
WIND DIRECTION, degrees
(b) WEEKDAY, PRECIPITATION, >O.IO degrees
Figure 49. Excess TSP concentration at Broad and Spruce above concen-
tration at 500 South Broad as a function of wind direction.
109
-------
K>
140
120
100
E
^
a 8°
Q.
cn 60
<
40
20
0 40 8O 120 160 200 240 280 320 360
WIND DIRECTION, degrees
(c) WEEKEND, PRECIPITATION < T
10
140
120
100
E
^» 80
^
W 60
H-
<
40
20
0 40 8O 120 160 20O 240 280 320 360
WIND DIRECTION, degrees
(d ) WEEKEND, PRECIPITATION, > 0.10 inches
Figure 49 (continued)
Excess TSP concentration at Broad and Spruce above
concentration at 500 South Broad as a function of
wind direction.
110
-------
as useful in estimating local contributions from steet-level
particulate sources as more sophisticated models.
• Concentrations near such intersections are not greatly dependent
upon the prevailing wind direction because of locally induced
turbulence from buildings, exhaust heat, and passing vehicles.
• Where building heights are uniform, the "cut-section" submodel
of HIWAY offers a means of calculating the contribution of
street-level sources to rooftop particulate levels.
• For downtown Philadelphia, an emission factor for reentrained
street dust of about 0.42 g/km-vehicle appears to be appropriate
for rooftop receptors. For short travel distances (e.g., at the
Broad and Spruce Street trailer), this emission factor should be
increased by a factor of perhaps 2 or 3.
RESULTS OF AIR QUALITY DISPLAY MODEL CALCULATIONS
TSP Concentrations
The AQDM was used to calculate annual average TSP concentrations, and
components of the annual average according to source category, at 5-kilometer
grid intersections throughout a 35 x 40 kilometer rectangular area centered
on Philadelphia. The results obtained using the emission inventory described
earlier in this section, an emission factor for reentrained dust from paved
roadways of 0.42 g/km-vehicle, and 1974 meteorological data are given in
Figures 50 through 54 and in Table 28. The results are presented in the fol-
lowing order:
• Figure 50 - Contribution from major point sources.
• Figure 51 - Contribution from nonvehicular area and small
point sources.
• Figure 52 - Contribution from vehicles (tailpipe and tire
wear only).
• Figure 53 - Contribution from reentrained roadway dust.
* Figure 54 - Total contribution from all inventoried sources.
Table 28 lists the contributions by components and the total for each
grid intersection.
As would be expected, the maximum contributions from all source categories
tend to occur within the principal urban area, with the major axis of the
central concentration isopleth oriented in a southwest-northeast direction.
Large point sources have two areas of major impact, one in the southern part
of the city, and one to the east (see Figure 51). The relative contributions
of the various source categories, averaged throughout the area, are also
111
-------
10
/
4445
4440 -
12
4435-7 TIO
4415
4410
-8
.7_A
^C j. 9 . J. Q j. 7
4405
470
475
480
490
KM, EASTING
495
500
505
Figure 50. Annual average TSP concentration contributed
by large point sources, vg/m3.
112
-------
4445
4440 -
- 7
44O5
470 475 480 485 490
KM, EASTING
495
500
505
Figure 51. Annual average TSP concentration contributed by area
and small point sources (r.onvehicular),
113
-------
4445 r
485 490
KM, EASTING
Figure 52. Annual average TSP concentration contributed by vehicles
(tailpipe and tire wear only) based on assumed emission
rates of 0.21 and 0.12 g/km-vehicle, respectively, yg/m3,
114
-------
4445
4440
470 475 480 485 490 .495 500 505
KM, EASTING
Figure 53. Annual average TSP concentration contributed by reentrained
roadway dust based on an assumed emission rate of 0.42
g/km-vehicle, pg/m3.
115
-------
4445
4440 -
4435-25
4430
z
X
§ 4425
z
3
4420
4415
4410
4405
470
475
4QO
490
KM, EASTING
495
500
28
505
Figure 54. Annual average TSP concentration contributed by all
inventoried sources,
116
-------
TABLE 28. CALCULATED ANNUAL TSP CONCENTRATIONS AND
CONTRIBUTIONS BY SOURCE CATEGORY FOR
PHILADELPHIA AREA IN 1974 IN yg/m3
VTM coordinates
470 4405
4410
4415
4420
4425
4430
4435
4440
4445
475 4405
4410
4415
4420
4425
4430
4435
4440
4445
480 4405
4410
4415
4420
4425
4430
443*)
4440
4445
485 4405
4410
4415
4420
4425
4430
4435
4440
4445
490 4405
4410
..415
4420
4425
4430
-.435
4440
4445
Large point
sources
only
6
8
7
6
6
6
7
7
6
6
7
7
7
7
7
10
9
6
6
7
9
11
9
9
9
11
7
7
10
22
17
11
12
9
8
7
7
9
14
1"
20
1!
y
8
7
Area and
small point
sources
5
7
6
6
6
6
7
8
5
6
7
8
a
9
6
8
8
7
7
9
11
12
11
11
9
9
7
8
10
13
15
15
13
11
10
8
8
10
13
15
16
14
12
10
9
Tall pipe
and
tire wear
4
•)
5
5
5
5
5
6
4
4
5
6
7
7
6
6
5
5
5
5
7
8
8
7
6
6
5
6
6
8
11
9
8
8
7
6
6
7
8
10
10
9
b
ft
7
Reentrainment _ ,
. Total
dust
5
s
b
6
b
6
6
8
5
5
6
8
9
9
8
8
6
6
6
6
9
10
10
9
8
8
6
8
8
10
14
11
10
10
9
8
8
9
10
13
13
11
1C
10
9
20
26
24
23
23
23
25
29
20
21
25
29
31
32
29
32
28
24
24
27
36
41
38
36
32
34
25
29
34
53
57
46
43
38
34
29
29
35
45
52
59
45
39
36
32
(continued)
117
-------
TABLE 28 (continued).
VTM coord Inn ten
495
500
505
4405
4410
4415
4420
4425
4430
4435
4440
4445
4405
4410
4415
4420
4425
4430
4435
4440
4445
4405
4410
4415
4420
4425
4430
4435
4440
4445
Average
Pi- re
•ent
Large point
sources
only
6
8
10
12
16
15
10
9
8
6
7
8
9
11
12
12
11
10
8
9
8
8
9
11
12
12
11
9.4
27
Area and
small point
sources
8
9
11
13
14
14
12
10
9
7
8
9
10
11
12
11
10
8
6
7
8
9
9
10
9
9
8
9.5
28
Tall pipe
and
tire wear
6
7
8
9
9
9
8
7
7
6
7
8
8
9
8
8
7
6
6
6
7
7
7
8
7
7
7
6.8
20
Reentralnment
dust
8
9
10
11
11
11
10
9
9
8
9
10
10
11
10
10
9
8
8
8
9
9
9
10
9
9
9
8.7
25
Total
28
33
39
45
50
49
40
35
33
27
31
35
37
42
42
41
37
32
28
30
32
33
34
39
37
37
35
34.4
118
-------
shown in Table 28. These averages indicate that vehicle-related emissions
are responsible for 45 percent of the total accounted for by the emission in-
ventory, and that the balance is almost equally divided between large point
sources and the area and small point sources.
It is now possible to judge the appropriateness of the assumed 0.42 g/km-
vehicle emission rate used for paved roadways, if one assumes that the emis-
sions of all of the regularly inventoried sources, and the resulting model
estimates of their contributions to the particulate loading throughout the
network are reasonably well defined. For example, the annual arithmetic mean
concentration at the western edge of the city near the Belmont and Roxborough
stations is about 62 yg/m . Figure 54 shows that the total contribution from
all inventoried sources, including reentrained roadway dust in this part of
the city was about 36 yg/m3. The difference of 26 yg/m3, left for background
and uninventoried sources, is a reasonable amount to be assigned to such
sources. On the other hand, if the emission factor for reentrained roadway
dust were four times as great (still less than the average estimate for resi-
dential areas),11 the reentrained contribution would be increased from about
9 yg/m3 (see Figure 53) to 36 yg/m3, leaving nothing to be assigned to back-
ground and other uninventoried sources. Thus, the assumed value of 0.42 g/km-
vehicle appears to be a suitable first approximation for an emission rate for
general application in calculating citywide TSP levels. The difference between
this value and the higher rates measured by MRI suggests that large losses of
even fine particulates occur as they move from the near-roadway locations used
in measuring emission factors to more remote monitoring locations, possibly
due to the screening and cleansing effects of vegetation. It is also an indi-
cation of the extreme complexity of the problem of experimentally determining
suitable areawide emission factors for reentrained dust for use in air quality
dispersion models.
The AQDM was run again with the 1974 emission inventory and an emission
factor of 0.42 g/km-vehicle for reentrained dust, but with 1976 meteorological
data. In this case, the various source contributions were calculated for each
monitoring location. Table 29 shows the results. The model contributions to
the arithmetic mean concentrations presented in Table 29 have been scaled to
approximate geometric mean values and used to prepare Figures 55 and 56. At
500 South Broad Street, as shown in Figure 55, the four modeled source cate-
gories contribute 45 yg/m3 out of the observed 74 yg/m3, leaving 29 yg/m3 to
be attributed to background and uninventoried sources. This is only slightly
greater than the value estimated for the western edge of the city. The excess
43 yg/m3 measured at Broad and Spruce is attributed to local low-level sources,
principally reentrained street dust and other vehicular emissions.
In Figure 56, the model estimates of the four contributing source cate-
gories have been plotted cumulatively on top of an assumed background level
of 29 yg/m3 at all 13 monitoring sites. The top of each calculated bar
(dashed line) can be compared with the observed concentration (solid line),
and the difference attributed to the local impact from reentrained dust, other
vehicle-related emissions, and perhaps very local uninventoried sources. At
Allegheny, the 82 yg/m3 contribution calculated by the model from a grain
loading facility approximately 300 meters from the monitor and with an estimated
emission rate of 100 tons/year has been omitted from the point source estimate.
119
-------
TABLE 29. MODEL ALLOCATION OF TSP CONCENTRATIONS TO SOURCE CATEGORIES
Nonlocal
Site
DEF
ALL
INT
BEL
ROX
N/E
NBR
FRI
LAB
SBR
S/E
500
AFS
Major Area and
point small
sources point source
26
*
97
8
10
8
12
12
14
17
15
17
15
19
12
15
11
11
10
11
14
15
14
14
13
14
17
Tailpipe
and
tire wear
8
10
6
8
6
8
9
10
9
9
8
9
10
Observed
„ . , annual
Reentrained _ .. . ,,
Total arithmetic
street
, mean
dust
10
13
8
10
8
10
11
13
11
11
10
11
13
56
135
33
39
32
41
46
52
51
49
48
49
59
94
111
102
62
62
62
112
88
82
124
96
80
107
•K ~
82 ug/nr attributed by model to a nearby grain loading facility.
Note: 1974 inventory; 1976 meteorological and TSP data. Units are
yg/m3.
120
-------
120 r
100
10
z
o
Ul
o
z
o
o
a.
V)
80
60
40
20
LOCAL
REENTRAINED DUST,
and
VEHICULAR
EMISSIONS, and
4^
3
"UNINVENTORIED SOURCES
-J TAILPIPE and
z< TIREWEAR
*3 REENTRAINED
DUST
AREA and SMALL
POINT SOURCES
MAJOR POINT
SOURCES
BACKGROUND
and
UNINVENTORIED
SOURCES
8
i n
i v
i Tt
14
29
t
UJ
«
"* ™~ 'X ' '" T!
H QC
] W
< ^
O
01 o
03 g
0
3C ft'
H 03
O
to
0
o
to
i r i
i
Figure 55. Source contributions to annual geometric means at the
500 South Broad, and Broad and Spruce Street sites in
1976.
121
-------
140
LOCAL IMPACT'
REENTRAINED OUST,
VEHICULAR EMISSION*.
UNINVENTORIED
SOURCES
TAILPIPE I
AND I unil
TIREWEAR >LOCAL
REENTRAINEOJ
OUST )
AREA AND
SMALL POINT
SOURCES
MAJOR POINT
SOURCES
SITE
Figure 56. Allocation of TSP concentrations to source categories
at 13 Philadelphia monitoring sites.
122
-------
Figure 56 shows quite acceptable agreement between calculated concentra-
tions using a 29 yg/m background level and observed concentrations at ROX,
BEL, N/E, 500, LAB, FRI and DEF. The first three of these sites are exterior
to the main urban area and uninfluenced by major street-level emissions. The
next three sites are centrally located, but the monitors are at rooftop level;
and the last of the seven sites, DEF is within the enclosed area at the Defense
Supply Depot. Thus, the sites where large contributions appear to have been
made from local traffic or other uninventoried sources are S/E, INT, AFS, NBR,
ALL, and SBR. The magnitude of these contributions are given in Table 30.
The highest concentration listed in Table 30, 43 yg/m3, was observed at
SBR, the site with the greatest amount of local traffic. The second highest
concentration, 35 yg/m3, was observed at INT where the volume of traffic in-
creased an unknown but substantial amount with the construction of 1-95, and
which is located in a part of the city characterized by numerous fugitive
dust sources, including earth moving. The third highest concentration was at
NBR, where most of the 32 yg/m3 observed at this site can be attributed to
street sources. The 26 yg/m3 at ALL can be attributed to a combination of
traffic on Allegheny Avenue, emissions from the nearby grain elevator which
were deliberately omitted from the point source modeling, and other fugitive
dust emissions. Local traffic is insignificant at the S/E site and probably
has a relatively small impact at AFS because of the height of the monitor.
The 17 yg/m3 and 16 yg/m3 values at these two sites may well be a result of
general fugitive and fugitive dust emissions in the area which have not been
adequately taken into account in the emission inventory. For example, it is
quite likely that street loadings and resulting fugitive dust emissions in the
industrial and less developed areas of the city are substantially greater than
those in the CBD and residential areas. Any such increase in the emission
factor for paved roads above the 0.42 g/km-vehicle used in the-AQDM calcula-
tions would bring the observed and calculated concentrations for S/E, INT,
AFS and ALL into better agreement.
In summary, Figure 56 is an attempt to place in perspective the contri-
butions of the major classes of particulates to the annual concentrations
measured throughout the city. At downtown rooftop monitors outside of the
principal industrial areas and at trailer top monitors in outlying residential
areas, observed concentrations are adequately explained by contributions from
inventoried point and nonvehicle-related area sources, tailpipe and tire wear,
and reentrained street dust emitted at a rate of 0.42 g/km-vehicle plus a
background concentration of 29 yg/m3. Relatively small reductions in any or
all of these sources could result in the achievement of the primary standard
at these sites. On the other hand, at trailer top monitors within either
commercial or industrial areas, or at rooftop level in industrial areas, sub-
stantial additional reductions in emissions will be required to meet the primary
standard, and at several sites, these reductions must clearly come from vehicle-
related sources. This is particularly true at the two Broad Street trailer
sites where the estimated concentrations remaining after the elimination of
all contributions from point and nonvehicle-related area sources are still in
excess of the standard. It may also be true at the Allegheny and International
Airport sites as well, but at these locations the impact of other fugitive
dust and fugitive emissions needs further clarification.
123
-------
TABLE 30. ESTIMATED IMPACT OF LOCAL TRAFFIC AND UNINVENTORIED SOURCES
Nearby traffic
Concen- Height „.
I) 1 ^1 I r-lTl ("* Q
Site tration above ground Street m __. *
/ / 3\ , , Traffic count from monitor
(.yg/m ; (m; (veh/day) (m)
S/E 17 1 Not Applicable
INT 35 4 Island Ave. 6,000 (1972) 11
(No data since beginning of 1-95 construction)
AFS
NBR
ALL
SBR
16
32
26
43
9
4
4
4
Aramingo Ave.
Huntington Ave.
North Broad St.
Butler St.
Germantown Ave.
Allegheny Ave.
South Broad St.
Spruce St.
18,500
2,100
6,000
2,200
6,400
5,200
23,500
7,400
(1976)
(1974)
(1975)
(1975)
(1975)
(1976)
(1977)
(1977)
19
10
7
6
16
18
9
6
Average daily traffic at INT; average weekday traffic at other locations.
Lead Concentrations
Background—
Lead is emitted principally from automotive exhaust systems and therefore
serves as an excellent tracer for exhaust particulates. Furthermore, since
the emission rate of fine lead particles from motor vehicles can be reasonably
well estimated, and ambient lead concentrations have been routinely measured
within the AMS laboratory network, it is instructive to compare model esti-
mates with observed values. At the suggestion of the Project Officer, current
literature was reviewed to develop a tentative lead emission rate for the
Philadelphia area which was then used with the AQDM to calculate ambient
concentrations.
A review of the available literature indicated that sufficient research
has not been conducted to fully characterize the relationships between vehicles
operating in a specific mode and the corresponding lead (Pb) emission rate.
Sufficient research has been conducted, however, to provide reasonably sound
estimates of lead emissions under a number of general operating conditions.
A summary of the background information used in developing the emission rate
used in the model calculations is provided in Appendix D.
Application to the Philadelphia TSP Study—
To develop an estimate of the airborne lead emission rate for the Phila-
delphia region, several assumptions were required. First, it was assumed
that the driving patterns within tha entire region are uniform and that the
124
-------
overall pattern can be described as primarily city and suburban driving, where
the overall average speed is in the range of 30 to 35 miles per hour. Second,
it was assumed that the nationwide average fuel economy value for 1977 esti-
mated by EPA adequately reflects the Philadelphia vehicle fleet. This value
is 13.3 miles per gallon times a factor of 0.9 (to reflect a higher portion
of city driving) or 12 miles per gallon. Third, it was assumed that the
pooled average lead content of gasoline sold in the Philadelphia region is
consistent with the original EPA phase-down requirements for 1977; this value
is 1.0 grams per gallon. Finally, it was assumed that the area is sufficiently
large so that a wide range of driving modes occur and, hence, the appropriate
emission factor is 0.7. The lead emission rate, then, can be determined from
Equation (D-l), Appendix D:
V C
_ Pb Pb _ (0.7)(1.0 grams/gallon) = 0.058 grams
^Pb R 12 vehicle-miles/gallon veh-mile
G
The value 0.058 grams per vehicle-mile represents the total lead emission
rate. For model application, however, we are interested only in the fraction
that becomes airborne. Figure D-l shows that about 45 percent of the total
lead particles emitted from vehicles with over 28,000 miles accumulated are
less than 10 ym. Thus it is reasonable to assume here that the airborne lead
component is comprised of about 45 percent of the total lead emission. The
assumed airborne lead emission rate /Q , \ for Philadelphia, then, is:
I pba/
Qpb = (0.45)(Qpb) = (0.45)(0.058 g/mi-vehicle) =
a
0.026 g/mi-vehicle
or
0.016 g/km-vehicle
When this 1977 emission rate is used with the AQDM model and 1974 STAR
meteorological data, the annual average lead concentrations shown in Fig-
ure 57 result. Appropriate levels for 1976 can be obtained by multiplying
these values by the average scheduled gasoline lead content ratio for the 2
years (1.4/1.0). Multiplying this scaling factor by the maximum concentration
of 0.5 yg/m3 found in Figure 57 yields 0.7 ug/m3 as the maximum concentration
to be expected at "rooftop" heights, or at locations removed from the direct
influence of street or industrial sources. These model estimates of annual
average concentrations can be compared with the 3-month average values for
July, August and September of 1976 given in Section 3 of this report. The
two rooftop monitors free from the influence of major industrial sources and
near the area of maximum concentration, as indicated by the model, are at
Franklin Institute and 500 South Broad Street, where 0.79 and 0.74 yg/m3 were
measured, respectively. It is also of interest to compare the annual average
indicated by the model at the Roxborough and Northeast Airport sites with the
3-month average observed values. At Roxborough the model yields a concentra-
tion of 0.42 yg/m3 versus an observed 3-month concentration of 0.57 yg/m3; at
the Northeast Airport the model yields 0.56 yg/m3 and the observed concen-
tration was 0.53 yg/m3.
125
-------
4449
4440 •
0.2 0.2 0.2 0.9 O.3 0.3 10.3 0.3
440
480
««5 4*0
KM, EAST I WO
4»8 500 00ft
Figure 57. Average annual lead concentration based on an emission rate
of 0.016 g/km-vehicle, yg/m3.
126
-------
SECTION 6
SUMMARY AND CONCLUSIONS
The principal results of this study, grouped topically, are summarized
below.
NATURE OF THE PHILADELPHIA PARTICULATE PROBLEM
Citywide average TSP concentrations in Philadelphia dropped abruptly from
about 240 yg/m3 in 1956 to about 165 yg/m3 in 1958. This rapid decrease was
followed by a gradual downward trend, accelerated between 1966 and 1968,
through 1972. For the last 5 years, the citywide average concentration has
exhibited only minor year-to-year variations and appears to have leveled off
just slightly above the primary annual standard of 75 yg/tn3. However, this
standard is still being exceeded at 8 of the 13 monitoring stations, and annual
concentrations less than the secondary standard of 60 yg/m3 are found at only
three monitoring stations. These three stations are located in the western
and northeastern parts of the city and are removed from the centers of urban ac-
tivity. Furthermore, although high annual concentrations and rather frequent
violations of the 24-hour secondary standard also occur at sites subject to
more general sources of fugitive dust or specific industrial or commercial
sources, the highest annual concentrations are found at sites that are located
in close proximity to busy streets (i.e., sources of vehicular emissions and
reentrained dust).
Of particular significance is the fact that the apparent attainment of
the primary annual standard within the central business district depends
largely upon the placement of the monitor. Specifically, concentrations mea-
sured at rooftop level (11 meters) on the Public Health Building at 500 South
Broad Street yield an annual average of about 75 yg/m , while concentrations
measured on the roof of the trailer at the intersection of Broad and Spruce
Streets average about 116 yg/m . Sensitivity to measurement height near
street sources has also been documented at the Franklin Institute site, and
in other cities. This demonstrated dependency of TSP concentration upon moni-
tor height points out the importance of proper sitings when determining com-
pliance with the standards. In this connection, it should be noted that the
additional particulate loading acquired as one approaches street level differs
both in physical characteristics and chemical composition from that measured
at rooftop level, and that, as pointed out by Belanger,1 the primary standard
for TSP is based on studies involving rooftop monitors and monitors removed
from any major source of fugitive dust.
127
-------
Although fundamental questions are now being raised concerning the pos-
sible need for standards which include specifications as to particle sizes
and chemical composition, the current requirement is still to reduce emissions
sufficiently to bring avout the achievement of existing standards. The se-
lection and implementation of feasible and effective control plans to bring
about this reduction can be dene only if sound estimates of the relative con-
tributions of the various sources of particulates to the problem areas are
available.
GENERAL FEATURES OF TSP BEHAVIOR
An analysis of day-to-day changes in TSP levels within the monitoring
network showed the following characteristics:
• Extreme changes in concentration tend to occur in response to
major changes in synoptic meteorological conditions and are
therefore citywide in character; however, a source-oriented
monitor may be subject to large variations in concentration
independent of areawide conditions. On the other hand, small
changes at the various sites often are unrelated to one another
and therefore appear to be the result of changing local
conditions.
• Days with high concentrations are usually associated with either
poor ventilation conditions or with southerly flow and an
apparent concentration of industrial pollutants.
• Days with very low concentration are usually days with precip-
itation, and frequently occur on weekends. At the AMS Labora-
tory, the average concentration observed on a "rain" weekday
was 25 percent less than on a "non-rain" weekday; the average
concentration on a no-rain weekend was also 25 percent less
than on a no-rain weekday. A further reduction of 20 percent
occurred on rain weekends. In this study a no-rain period was
defined as one with a trace or less of precipitation during
the 48-hour period ending at the end of the sampling period, and
a rain period was one with at least 0.10 inches of precipitation
during the 48-hour period. The association of precipitation
with lowered TSP concentrations results from a number of inter-
related factors. First of all, precipitation is a cleansing
agent and results in the washout and rainout of pollutants;
second, precipitation suppresses fugitive dust emissions; and
third, precipitation is frequently followed by a frontal passage
and the arrival of a cleaner air mass.
• To date, TSP concentration at the Broad and Spruce trailer have
averaged approximately 43 yg/m3 higher than the concentration
on the roof at 500 South Broad Street (height 11 meters). The
maximum average difference of 53 yg/m3 occurred on weekdays
when the 48-hour precipitation did not exceed a trace, and the
minimum average difference of 36 yg/m occurred on weekends
when precipitation equaled or exceeded 0.10 in. These results
128
-------
are in agreement with the hypothesis that maximum contributions
from street-level sources occur with dry streets and high traffic
volume. The difference in concentration between these two
neighboring sites shows little, if any, dependence upon the average
24-hour airport wind speed. On weekdays with precipitation
equal to or less than a trace, the concentration difference
between the two sites was about 30 yg/m higher with winds
from the south through west (i.e., when the trailer was down-
wind from the intersection) than with winds from the west,
north and east. At other times, the difference appeared to
be unrelated to the airport wind direction.
• The common variation in concentration measured at two sites
is a function of separation distance, averaging about 70
percent at a separation distance less than 5 kilometers and
about 25 percent at distances of about 16 kilometers.
SPATIAL DISTRIBUTION OF PARTICULATES NEAR STREET SOURCES
At the Broad and Spruce Streets intersection, street-level emissions
usually become quite uniformly mixed with height (to at least 11 meters) be-
cause of locally induced turbulence from buildings, exhaust heat and passing
vehicles. Concentrations measured at a height of 11 meters on the tower
averaged about 88 percent of those measured at the top of the trailer in
February, and about 96 percent of those at the trailer in June, except imme-
diately following the street-flushing experiment. At an equivalent height on
the roof of the building at 500 South Broad Street, concentrations averaged
about 72 percent of those at the trailer during the February experiments.
Annual average concentrations at 500 South Broad during 1975 and 1976 were
62 and 63 percent, respectively, of those at the Broad and Spruce Streets
trailer.
Rooftop concentrations 9 meters downwind from Broad Street averaged about
15 percent higher than upwind concentrations during the February experiments.
Any increase at a downwind distance of 34 meters fell within the measurement
precision of the hi-vol. Variations in rooftop concentrations during the
June experiments appeared to be dominated by nontraffic-related local sources,
anc the effect of reentrained street dust and vehicular emissions could not be
isolated at either distance.
PARTICLE SIZES
During the project, various pieces of information relating to the particle
size distribution of suspended particulates were acquired. Some of this in-
formation was obtained from direct measurements and some was inferred. More
importantly, some of the data appears to be conflicting. However, with the
primary exception of data obtained with the fractionating-head hi-vol, a fairly
consistent picture emerges and is presented in this summary.
The most extensive relevant set of data collected makes a comparison of
hi-vol and dichotomous sampler observations possible and thus permits a mass
329
-------
breakdown into three approximate size categories. They are: < 3.5 ym, from
3.5 to 20 urn, and > 20 ym. Table 31 presents a composite picture of the
spatial distribution determined from measurements made during February and
March. Dichotomous sampler data at the two sites were acquired with one
instrument, operated on an approximate every-other-day schedule for about 2
weeks on the roof of the trailer, and for a similar period on the roof at
500 South Broad.
TABLE 31. COMPARISON OF DICHOTOMOUS SAMPLER AND
HI-VOL RESULTS
Concentration
(yg/m3)
Height (m) Dichotomous Sampler Hi-Vol
Location
Fine Coarse Total
Top of tower
Rooftop
Trailertop
11
11
4
30
34
13
23
43
57
130
95
149
The fine fraction of particulates has an upper 50 percent cutpoint of
3.5 ym and is considered respirable, while the coarse fraction is composed
roughly of the particles from 3.5 to 20 ym. At the rooftop (500 SB) the
respirable fraction comprises 70 percent of the total measured by the dicho-
tomous sampler; at the trailertop the respirable fraction is 60 percent of
that total. Perhaps of greater interest, however, is a comparison of the
concentration of respirable particulates with the concentration determined by
the hi-vol. In this case, the respirable fraction represents 32 percent of
the mass measured by the hi-vol at rooftop-level and only 23 percent of that
measured by the hi-vol at trailertop. Also, assuming a cutoff of 20 ym by
the dichotomous sampler, approximately 55 percent of the particulate mass
collected by the hi-vol at rooftop level was composed of particles of aero-
dynamic diameter greater than 20 ym. At the trailertop location, this per-
centage increased to approximately 62 percent.
Direct measurements of the size distribution of particles collected on
10 selected hi-vol filters were made by means of optical microscopy. Sep-
arate size distributions were made for opaque and nonopaque particles. In
this study, the mass median diameter of the opaque and nonopaque particles
averaged about 11 ym and 7 ym, respectively. The mass median diameters of
these 10 samples did not appear to be height dependent. Measurements of par-
ticle diameter were also made using 3 of these 10 filters and a Coulter
counter. Using this technique, estimates of the mass median diameter for the
three samples were 14.5, 23, and 28.5 ym. These Coulter counter results are
in better agreement with the results of the hi-vol - dichotomous sampler
comparison which indicated that a little over half of the mass collected by
the hi-vol was composed of particles greater than 20 ym in aerodynamic
diameter.
130
-------
CHEMICAL COMPOSITION OF PARTICULATES
Elemental concentrations determined from the dichotomous samples showed
that the six elements having the highest concentrations were, in order of de-
creasing concentration: sulfur, silicon, calcium, lead, iron and aluminum.
Of these six elements, sulfur had the highest concentration and, when expressed
as S0i+, made up 19 and 25 percent, respectively, of the mass at the Broad and
Spruce, and 500 South Broad locations. Eighty-seven percent of the sulfur was
found in the fine fraction. There was a reduction in sulfur concentration of
only 2 percent with height, indicating thorough mixing and transport from non-
local sources.
Of the remaining five elements, all but lead are presumed to be primarily
of mineral origin and are most abundant in the coarser fraction, with the per-
cent fine ranging from 8 to 14 percent. Also, with the exception of calcium,
their concentrations decrease substantially with height (from 36 to 48 percent),
indicating strong contributions from local fugitive dust sources. The third
most common element at Broad and Spruce, and the fourth most common at 500
South Broad Street, is lead. Lead, like sulfur, is found predominantly in the
fine particulates, but unlike sulfur, shows a rapid decrease in concentration
with height (37 percent), in agreement with the hypothesis that motor vehicles
are its principal source.
Lead concentrations were also determined throughout the city by the AMS
Laboratory. The proposed standard of 1.5 yg/m3, based on a monthly average,
was exceeded at three sites: the Aramingo Fire Station, which is source-
oriented and located near a National Lead Plant; and the two traffic-oriented,
trailer-top sites on Broad Street. However, it appears that the proposed stan-
dard can be met at the two traffic-oriented sites in 1978 by lowering the max-
imum allowable pooled average lead content of gasoline to the 0.8 grams per gal-
lon specified for 1 January 1978 in The Federal Register of September 28, 1976.
Ten hi-vol filters were also selected for extraction with benzene and sub-
sequent infrared analysis. The benzene soluble content of the TSP varied from
0.5 to 5.9 yg/m3 but the variations do not seem to correlate with TSP levels
or with height above street level. The infrared spectra indicate complex
mixtures of organic compounds in most of the benzene extracts.
EFFECT OF STREET FLUSHING ON AMBIENT PARTICULATE LEVELS
Intensive street flushing on 3 consecutive days between the hours of
7:00 a.m. and 6:30 p.m. in the vicinity of the Broad and Spruce Street moni-
toring site not only failed to reduce 24-hour concentrations but appeared to
increase concentrations dramatically immediately following the flushing oper-
ation. Concentrations at the monitoring site rose to levels which were
roughly 100 yg/m3 higher than would have been expected from observations at
other locations within the city.
This phenomenon may be explained by assuming that the vigorous, forced
flushing, plus splashing by motor vehicles, redistributed particulates that
had previously become concentrated adjacent to the curbs, and that many of
these redistributed particulates became airborne as soon as the street became
131
-------
dry. This hypothesis is supported by a strong relationship between traffic
volume and TSP concentrations during this period.
MODELING
The selection of appropriate emission factors for fugitive and reentrained
dust continues to be one of the most perplexing problems associated with par-
ticulate modeling. Enough measurements have been made to confirm that road-
way emission rates and the size distribution of the emitted particles within
an urban area vary widely both spatially and temporally. Generalizing from
these measurements frequently leads to unreasonable results. For example,
the introduction of emission factors for roadway dust particles <_ 5 ym in
aerodynamic diameter developed by MRI for residential areas leads, through the
use of the Air Quality Display Model, to excessively high estimates for the
contribution of reentrained dust within the Metropolitan Philadelphia AQCR.
If these emission factors are, in fact, generally applicable, it would appear
that large losses of these particulates occur as they move from the roadway
to monitoring locations, due, perhaps, to the screening and cleansing effects
of vegetation and to mechanisms such as agglomeration and settling.
A modeling approach that holds promise, and the one that was adopted in
this study, is to accept an emission factor for general use that yields con-
sistently reasonable results throughout the area for rooftop monitors and
for monitors removed from any obvious local particulate source. Major devia-
tions from these "rooftop" model estimates are attributed to local sources.
An attempt can then be made to quantify these local effects either by modeling
or by the use of empirically-derived expressions.
Near built-up intersections, -such as at the Broad and Spruce Street
location, box-type models appears to be useful because of the high degree
of mixing that occurs as a result of locally-induced turbulence.
For Philadelphia, an appropriate emission factor for use in calculating
rooftop concentrations is about 0.42 g/km-vehicle. For trailer-top concen-
trations near major streets, this emission factor should probably be in-
creased by a factor of 2 or 3.
PROSPECTS FOR MEETING THE NAAQS
Two objectives of this report have been to put into perspective the recent
history of TSP levels within the City of Philadelphia and to identify the major
source categories currently contributing to these levels. Between 1956 and the
early 1970's, emission reductions from the so-called traditional sources
lowered particulate levels by approximately 150 yg/m3, from a citywide annual
average of about 225 yg/m3 to roughly 75 yg/m3. According to the model results
of this study, these sources now contribute about 27 yg/'m* at 500 South Broad
Street and perhaps 35 yg/m3 at locations where the larger point sources impact
most heavily. This implies that a reduction of the order of 80 percent has
already taken place in the joint contributions of traditional large point
sources, and small point and area sources. As a limiting case, if all of the
reduction is attributed to reduced emissions from large stationary sources,
132
-------
similar reasoning indicates that these sources have been cleaned up by about
90 percent. Thus, a great deal has been accomplished. The most feasible and
productive emission reductions, however, have undoubtedly already been achieved
and, except where identifiable point-source problems still exist, it is now
necessary to investigate possible reductions from other major source categories
if the standards are to be achieved throughout the city.
At downtown rooftop monitors outside of the principal industrial areas,
this study has indicated that vehicle-related emissions are responsible for
about 43 percent of the total particulate loading accounted for by the emission
inventory, and that the balance is almost equally divided between large point
sources, and area and small point sources. A total of 29 yg/m3 has been
assigned to background or other sources. The primary standards could probably
be achieved in these areas by relatively small reductions in any or all of
these source categories. On the other hand, at trailer-top monitors near
heavily traveled streets, an additional contribution of as much as 43 ug/m3 is
believed to be made by reentrained street dust and vehicular sources. If these
estimates are correct, the concentration that would remain at these locations
after eliminating all contributions from point and nonvehicle-related area
sources would still be in excess of the standards. At such sites, therefore,
achievement of the standards will require a very major reduction in street
level emissions in addition to any other reductions that might be brought
about. Finally, at specific locations in the principal industrial areas the
impact of individual sources contribute significantly to violations. At these
locations the situation is also frequently aggravated by a greater than aver-
age contribution from street and other fugitive dust sources.
These results point to the need for additional controls on selected sources.
It is important to note, however, that the current particulate standard attain-
ment status of some areas of Philadelphia has recently become uncertain because
of the deletion of monitors which were considered to be too close to major
streets, and the planned addition of a number of more appropriately located mo-
nitors. Thus the magnitude of needed controls must now be redefined, and will
depend, in part, on future citywide monitoring results. Potential additional
controls could conceivably be obtained by any number of strategies, each of
which could be expected to result in a small improvement, but which taken to-
gether might provide not only any reductions that may be required to meet the
primary standard, but those needed to meet the. secondary standard as well.
Such measures include control of fugitive emissions from industry and vehicles,
control of small oil-burning sources normally included in the area source inven-
tory (possibly by upgrading oil burner service) and any additional controls
which could be applied to large stationary sources. Some additional reductions
might also be obtained in the secondary "background" by control of S02 and N02
emissions over large areas because these are precursors to secondary particulate
matter. In addressing the point source problem, the source-receptor file of
the AQDM can be used to identify the contribution of individual sources to the
annual mean concentration at problem sites. The feasibility of imposing addi-
tional controls on these sources and their effectiveness in improving air qua-
lity can then be analyzed. The problems encountered in selecting and imple-
menting effective procedures for the control of fugitive dust emissions are well
known. A discussion of possible cor.crol strategies is given in "Guideline for
Development of Control Strategies in Areas With Fugitive Dust Problems."
133
-------
(EPA-450/2-77-029, October 1977). The one strategy-effectiveness test carried
out during this program was the vigorous street flushing experiment which re-
sulted in increased particulate levels at the nearby monitors.
In conclusion, it is becoming generally accepted that it is inappropriate
to monitor TSP levels at low elevations adjacent to heavily traveled streets
and highways for comparison with the health-related primary standard. The
early adoption of a policy which allows states to adjust or ignore low-level
TSP monitoring, such as that carried out at the two Broad Street sites, for
purposes of designating attainment status and control strategy development
would move the Philadelphia particulate problem from a position where feasible
controls appear to be totally inadequate to one which,, though challenging,,
holds a real possibility for solution.
AN AREA FOR CONTINUED RESEARCH
Although this study was initiated because of the need to neet the NAAQS
for particulates as measured by the standard hi-vol, ic was recognized that
the detailed knowledge of urban particulates required for the effective develop-
ment and testing of attainment strategies would require the use of more sophis-
ticated instrumentation. Because of this recognized need, the hi-vols used
in the field measurement program were supplemented by instruments designed to
provide at least some information concerning the physical and chemical nature
of the airborne particulates. Additionally, it was hoped that the acquisition
of field data of this type might ultimately contribute to the appropriate use
of some future particulate standard which either directly or indirectly took
into consideration the size distribution and chemica?, composition of suspended
particles.
From a comparison of the results obtained by different monitoring devices
during this study, it appears that very basic conclusions may be affected by
the type of instrument employed. If this is so, the selection from currently
available instrumentation should be made only after carefully considering the
purpose of the monitoring program and any special characteristics of the candi-
date instruments. Although in principle the choice of instruments can be made
on the basis of sampling time, particle size cut, and the type of chemical or
physical analysis desired, if any, the actual results may be influenced by the
particular instrument selected, and even by the nature of the atmosphere being
measured as well.
Further progress requires the continuing development and application of
research-type instrumentation for the detailed analysis of the particulate
problem, plus the standardization of reliable instrumentation for network use
to replace, or at least supplement, the hi-voL so that more meaningful health
effects studies can be carried out and a more solidly based primary standard
developed.
134
-------
REFERENCES
1. Belanger, W. E. An Investigation of the Siting of High Volume Air Samplers
for Determination of Compliance with the Particulate Air Quality Standard.
EPA Region III. Unpublished Manuscript, May 1977,
2. Emission Inventory and Sulfur Dioxide Alternatives for the Metropolitan
Philadelphia Region. Final Report, August. 1977. GCA Corporation,
GCA/Technology Division, Bedford, Mass. (EPA 903/9-79-030), Contract
No. 68-02-1376, Task'Order No. 24.
3. Spirtas, R. and H. J. Levin. Patterns and Trends in Levels of Suspended
Particulate Matter. J Air Pollut Contr Assoc. 21:329-333. June 1971.
4. Lasenka, C. and T. Weir. Rain Particulate Study. City of Philadelphia
Air Management Services. Unpublished Preliminary Document. January 4,
1974.
5. National Assessment of Particulate Problem. GCA/Technology Division under
EPA Contract No. 68-02-1376, Task Order No. 18.
6. Federal Register, Control of Lead Additives in Gasoline, Vol. 41, No. 189,
U.S. Environmental Protection Agency. September 28, 1976.
7. Wedding, D. B., A. R. McFarland and J. E. Cermak. Large Particle Collec-
tion Characteristics of Ambient Aerosol Samplers. Environmental Science
and Technology, Vol. 11, No. 4. April 1977.
8. Highway Statistics. U.S. Department of Transportation, Federal Highway
Administration, Washington, D.C. 1974.
9. Guidelines for Air Quality Maintenance Planning and Analysis, Volume 7:
Projecting County Emissions. U.S. EPA, OAQPS, Research Triangle Park,
North Carolina. January 1975.
10. Guidelines for Air Quality Maintenance Planning and Analysis, Volume 8:
Computer-Assisted Area Source Emissions Gridding Procedure. U.S. Environ-
mental Protection Agency, Office of Air Quality Planning and Standards,
Research Triangle Park, North Carolina. September 1974.
11. Midwest Research Institute. Quantification of Dust Entrainment from
Paved Roadways. Draft Final Report, EPA Contract No. 68-02-1403, Task
Order No. 25. March 4, 1977.
135
-------
APPENDIX A
MONITORING SITE DESCRIPTIONS AND
NEARBY TRAFFIC DATA
Key to traffic data:
AAWT = Average Annual Weekday Traffic
AADT = Average Annual Daily Traffic
AMPH = Morning Peak Hour Traffic
PMPH = Afternoon Peak Hour Traffic
(Site descriptions were prepared in November 1976)
A-l
-------
CO
W
H
Crt
O
Z
s
O
H
O
&
PL,
H
s
PM
M
^^
»-3
f— 4
PH
•
rH
1
W
3
H
u
4J
ID
M-t
4)
0
rV
U
Q
rH
n
rl
4J
U
3
6
I
B
rl
3
3
0
01
Pi
0)
rl
CO
rH
s
•o
B
OJ
0)
B
O)
J3
00
0)
l-t
r-l
^
a
a)
e*
0)
rH
rH
•*
r-3
rH
0)
U.
4J
(/]
3
"8
i
B
rl
3
3
to
O
a
cj
•H
u
4->
0
a
'H
U-l
o
U
,
s
T3
B
CO
rH
cn
w
4J
IH
O
O-
•H
rH
CO
B
o
>H
4J
a
E
o>
E
M
M
rH
CO
•H
B
01
T)
•iH
in
01
I
B
cO
.a
3
J3
3
c/l
1
O O
• rl
0
•
01
4J
c
1
rH
T3
B
CO
•0
at
•a
rl
o
fc.
B
o
•H
4J
a
to
IH
4J
rH
•H
4-t
a
o>
CO
rJ
W
CO
r-t
<0
B
01
73
'H
cn
01
1
B
ra
rl
3
JD
3
t/1
o
III
4J
01
01
4J
to
0>
rH
E
CO
01
a
TJ
B
CO
CO
W
O
4J
CO
0>
4J
rH
•H
4=
60
O
M
O
«
X
§
rH
CO
E
01
•rt
cn
01
1
B
to
14
3
3
O
•
s
a
0
0)
T3
C
CO
>
^
B
CO
O
4J
O
e-
-H
^J
cn
CO
01
rC
4J
O
W
z
rH
^H
CJ
i
O
U
1
E",
4J
>H
U
IH
01
4J
E
01
O
to
4J
OJ
01
01
4J
rH
3
CO
•B
CO
•8
o
CO
•a
CO
o
rl
ca
j:
4J
^O
a:
s
rH rH
.343
O 0
^ ^
0 0
u o
1 1
tr, ^.
4J 4J
•H -H
CJ CJ
rl M
01 01
4J 4J
B B
CU 01
O tj
*r o
rH
X X
CO CO
11
.
4-1 W
CO to
.C J2
0 0
CSJ CN
a
o
^"S
§ 0
O pi
a> ai
3 3
•H vH
4-- 4-1
W (A
0 C
C C
•H -H
3 3
cfl n}
M M
tn h
rH
e
rH
CO
4J
B
0)
•a
>rt
cn
01
rl
1
B
IH
3
3
vO
4-1
00
B
•H
a
rJ~
W
rH
O
rH
CT
o
u
CO
IH
2
3
jH
5
3
3
u
I
Q
U
1
p^
4J
CJ
rl
01
4J
B
(U
U
O
cn
4J
01
01
rl
4-1
01
CJ
3
rl
O.
to
"O
B
CO
T)
CO
O
rl
pa
Tl
rl
n
3
o
w
ai
CO
to
u
rl
g
g
U
1
^
4J
CJ
rl
01
4J
B
(U
CJ
O
•a-
cn
01
01
rl
4-1
C/t
01
0
B
O
,
H
4J
3
rH
rH
O
0-
rl
01
4J
$ 2
w a.
w
To
rH
CJ
14
s
1
CJ
1
.^
4J
CJ
rl
01
4J
a
0)
o
rH
rH
.
to
"^3
cO
0
rl
CO
J2
4-1
3
0
to
0
o
m
•a
M
ca
j.
u
§
to
8
o
o
m
rH
a
•H
IH
4J
cn
3
"S
«r-t
1
^
4J
O
r.
01
4J
a
01
u
o
Os
4J
B
O
4J
60
B
4-1
B
£
•o
B
CO
«
0>
O
00
B
•H
B
3
rl
B
o
•rt
4J
CO
4J
cn
01
rl
•H
rV
O
60
I
H
i!
to
5!
4J
«4H
00
II
s
rH
t
tO
ON
a
•H
at
g
IH
1
£
3
|
«
A-2
-------
DEFENSE SUPPLY DEPOT (DEF)
20th St. and Oregon Avenue
UTM Coordinates
Northing: 4418.32 Easting: 483.85
General Site Description
General industrial trend site; moderate industry in the area; probably
some local influence from operations and activities at the Defense Depot
(to east of monitor) and also from trains on hill to west of monitor.
Placement of Monitor
Elevation above ground: 4 m, on trailer
Exposure: good all directions
Immediate Surroundings (see Figure A-l)
Fairly open
Height of buildings: one story
Roads: paved, light to moderate dirt
Parking areas: paved, light to moderate dirt
Curbs: none
Sidewalk: none
Traffic volume: very light (service road)
Speed: 50 to 65 kph
Type of traffic: cars, trucks
Traffic signals: none
Other ground level surfaces: railroad tracks, small amount of grass
Nearby Sources
Type Description Distance Direction
*
Fugitive dust Dirt pile, apparently worked 45 m NNE
Gravel beds of RR tracks 10 to 45 m N to W
Fugitive emissions None evident
Nearby stacks Diesel locomotives on tracks 45 m W
Recent activity None evident
*
None apparent during site survey in moderate wind.
A-3
-------
STORAGE BUILDING
1 STORY
STORAGE BUILDING
I STORY
STORAGE BUILDING
I STORY
STORAGE BUILDING
I STORY
H\ INI I || MM IMIIIII | | I H
STORAGE BUILDING
I STORY
Figure A-l. Defense Supply Depot, monitor location.
A-4
-------
ALLEGHENY (ALL)
Allegheny Avenue and Delaware River
UTM Coordinates
Northing: 4425.17 Easting: 491.67
General Site Description
General industrial trend site; moderate industry in the area; probably
some local influence from Bunge Grain elevator and dirt pile to the west,
and from the Franklin Smelting and Refining Company to the Northeast
Placement of Monitor
Elevation above ground: 4 m
Exposure: good all directions
Immediate Surroundings (see Figure A-2)
Open
Height of buildings: one to two stories, storage silos
Roads: paved, light dirt
Parking areas: paved, light dirt
Curbs: 8 cm high
Sidewalk: None
Traffic volume: light
Speed: 50 to 65 kph
Type of traffic: cars, trucks
Traffic signals: none
Other ground level surfaces: some bare dirt and gravel
Nearby Sources
Type Description Distance Direction
Fugitive dust Possible from dirt pile 55 m WNW
Fugitive emissions Possible from Bunge Elevator 150 m W to NW
Nearby stacks None evident
Recent activity Road may be new 20 m N
Parking area for firefighters 6m S
may be new
A-5
-------
4
/
TO BUNOE
ELEVATOR
I
"TO i-9s
jVafflc Data
AAWT « 5200 vpd
AADT = 4900 vpd
AMPH = 500 vph
PMPH = 400 vph
O O O O O
STOWAGE SILOS
O O O O O
o
LOADING 1
PILE ^^^\
- — x -
""" -*•
-"" *s^-
""EF
^*N
PAVED
PARKING
MARINE
FIRE
FIGHTING
„—
' .
Ss TRAILER
with
HI -VOL
PAVE
;r~^— -
DELAWARE RIVER
Figure A-2. Allegheny Avenue, monitor location.
A-6
-------
PHILADELPHIA INTERNATIONAL AIRPORT (INT)
Island Rd., East of Airport Circle
UTM Coordinates
Northing: 4414.79 Easting: 480.26
General Site Description
Light industry to commercial neighborhood; trend site but with local
influences due to traffic on paved roads and activity at maintenance
shop
Placement of Monitor
Elevation above ground: 4 m, on trailer
Exposure: tree to NNW only a few feet from monitor and rising 15 m
above it; otherwise good
Immediate Surroundings (see Figure A-3)
Fairly open
Height of buildings: one to two stories
Roads: paved, light to moderate on Island Ave., very heavy on service
road and along curbs
Parking areas: paved on side of Island Ave.; dirt, gravel, and grass off
of service road.
Curbs: 9 cm on Island Ave., no gutter
Sidewalk: 1 m wide, dirty
Traffic volume: moderate to heavy
Speed: 50 to 65 kmh
Type of traffic: cars and trucks
Traffic signals: lights 90 m on either side
Other ground level surfaces: grass, bare dirt, gravel, trees, mound
of dirt 9 m to south
Nearby Sources
Type Description Distance Direction
Fugitive dust Dirt pile -3m high, apparently 9m SE
worked on one side, no dust seen
from it; gravel and dirt piles on
service road-dust seen from it
during visit.
Fugitive emissions Sand and gravel storage, junk dis- 30 m S-SE
posed behind maintenance shop
Nearby stacks None evident 90 m E
A-7
-------
Type Description Distance Direction
Recent activity Island Ave. has been opened to 10 m W
more traffic, road appears
cleaner, new median strips
A-8
-------
&
&&
&
i
/
Traffic Data
1972 AADT = 6,000 vpd
No data available since
beginning of 1-95 con-
struction in the area.
ORAVEU
STORAOE
Figure A-3. International Airport, monitor location.
A-9
-------
BELMONT FILTER STATION (BEL)
Ford Rd. and Monument Rd.
UTM Coordinates
Northing: 4427.83 Easting: 481.20
General Site Description
On a reservoir in a residential neighborhood with some commercial in-
fluence; trend site; good for describing air quality being transported
into the city limits from the west; possible from being situated on the
ground in field of high grass and weeds.
Placement of Monitor
Elevation above ground: 0 m, on ground
Exposure: except for location on ground, exposure good all directions
Immediate Surroundings (see Figure A-4)
Fairly open
Height of buildings: one story
Roads: roads around reservoir paved, service road is broker cement
and gravel
Parking areas: same as above
Curbs: not applicable
Sidewalk: not applicable
Traffic volume: very light
Speed: 30 kph
Type of traffic: too small to matter
Traffic signals: none
Other ground level surfaces: grassy field
Nearby Sources
Type Description Distance Direction
Fugitive dust Wind-induced reentrainment from 9 m N-NE
paved surface - probably minimal
Fugitive emissions None
Nearby stacks None except space heating
Recent activity None apparent
A-10
-------
Traffic Data
Ford Road between Belmont and Monument
1977 AAWT = 7300 = 7300 vpd
1977 AADT = 7100 vpd
AMPH = 505 vph
PMPN = 580 vph
Belmont Ave. between Conshocken and Monument
1977 AAWT = 12,700 vpd
1977 AADT = 12,000 vpd
AMPH - 1050 vph
PMPH = 1050 vph
Monument Road between BeTmont and Ford
1977 AAWT = 16,900 vpd
1977 AADT = 16,000 vpd
AMPH « 1350 vph
PMPH « 1350 vph
Figure A-4. Belmont Filter Station, monitor location.
A-ll
-------
ROXBOROUGH FILTER STATION (ROX)
Dearnley and Eva Streets
UTM Coordinates
Northing: 4433.12 Easting: 479.48
General Site Description
Urban trend site in a light residential suburb of the city; should
provide a fairly good measure of the concentration in the air entering
the region.
Placement of Monitor
Elevation above ground: 4 m, on AMS monitor
Exposure: trees to SW are high and close to monitor, otherwise good
Immediate Surroundings (see Figure A-5)
Open
Height of buildings: one to two stories
Roads: paved
Parking areas: paved
Curbs: not applicable
Sidewalk: not applicable
Traffic volume: very light
Speed: 8 to 55 kph
Type of traffic: cars
Traffic signals: none
Other ground level surfaces: trees, grass, dirt track
Nearby Sources
Type Description Distance Direction
Fugitive dust Dirt track on reservoir 20 m E
natural reentrainment from cement 30 m NE
Fugitive emissions None evident
Nearby stacks Only school and space heating
Recent activity None evident
A-12
-------
-------
NORTHEAST AIRPORT (N/E)
Grant Ave. and Ashton Rd.
UTM Coordinates
Northing: 4436.02 Easting: 498.97
General Site Description
A primarily downwind urban trend site located in a residential area
with some commercial activity; fairly remote from any usual in-
fluences except those at the airport.
Placement of Monitor
Elevation above ground: 4 m, on AMS trailer
Exposure: good all directions, a few trees nearby
Immediate Surroundings (see Figure A-6)
Open
Height of buildings: one to two stories but too far away
Roads: paved, light to heavy dirt
Parking areas: paved
Curbs: none
Sidewalk: none
Traffic volume: minimal
Speed: 30 kph
Type of traffic: cars, planes, trucks
Traffic signals: none
Other ground level surfaces: grass, trees, runway, some dirt and
gravel area
Nearby Sources
Type Description Distance Direction
Fugitive dust Natural reentrainment from pavement 10 m NE
and gravel area
Helicopter and plane-induced 10 m NE
reentrainment
Fugitive emissions None evident
Nearby stacks: None
Recent activity None evident
A-14
-------
GRASSY AREA
GRAVEL~
,H ^ —
# &
TRAILER
•RECREATION
AREA
No traffic data available.
GRASSY AREA
PAVED AIRPLANE PARKING
Figure A-6. Northeast Airport, monitor location.
A-15
-------
NORTH BROAD (NBR)
Broad and Butler Streets
UTM Coordinates
Northing: 4428.73 Easting: 487.03
General Site Description
Urban trend site in commercial/residential area on main street;
probably receives significant impact from vehicular activity on
paved roads.
Placement of Monitor
Elevation above ground: 4 m, on AMS trailer
Exposure: good all directions
Immediate Surroundings (see Figure A-7)
Built up
Height of buildings: two or five stories
Roads: paved, moderately dirty on North Broad, fairly heavy on Butler
and other side streets
Parking areas: paved
Curbs: 1 to 3 inches with gutter
Sidewalk: 3.5 to 5.5 m, moderately dirty
Traffic volume" heavy
Speed: 25 to 55 kph
Type of traffic: cars, trucks, buses
Traffic signals: at Broad and Butler, Broad and Germantown, Germantown
and Butler
Other ground level .'surfaces: trees, open area to west
Nearby Sources
Typ_e Description Distance Direction
Fugitive dust Vehicular-induced, primarily from 15 m E
Broad
Natural reentrainment from unpaved 45 m WNW
area
Fugitive emissions None evident
Nearby stacks Space heating
Recent activity None apparent
A-16
-------
OPEN
AREA,
DIRTY PAVE
PROBABLY
USED FOR
PARKING
3 STORY
2 STORY
-n STOPO1
BUTLER STREET LI6HT2
H
-ZOm—•
North Broad St. s/o Germantown
1975 AWT =6,000
Butler w/o Germantown
1975 AWT = 2,200
Germantown s/o North Broad
1975 AWT = 6,400
HI-VOL
CHURCH
2 WAY
PAVED
PARKIN6
8
o
l
Figure A-7. North Broad, monitor location.
A-17
-------
FRANKLIN INSTITUTE (FRI)
20th St. and Parkway
UTM Coordinates
Northing: 4422.8 Easting: 485.1
General Site Description
Commercial CBD site; good for monitoring urban trend away from local
influences; some interference with exposure from parapets.
Placement of Monitor
Elevation above ground: 15.2 m
Exposure: interference from parapets and observatory (see Figure A-8).
/
Immediate Surroundings (see Figure A-9)
Built up
Height of buildings: three to 10 stories
Roads: paved, light dirt
Parking areas: paved, light dirt
Curbs: 7 to 10 cm
Sidewalk: light dirt
Traffic volume: heavy
Speed: 40 to 65 kph
Type of traffic: cars, buses
Traffic signals: lights at every corner of block
Other ground level surfaces: trees and grass in park and around building.
Nearby Sources
Type Description Distance Direction
Fugitive dust Vehicular-induced reentrainment
but probably minimal
Fugitive emissions None evident
Nearby stacks Railroad terminal 0.8 km SW
Recent activity Paved Institute parking lot 30 m WSW
Built townhouses) _ .
„ . . , > Summer of 75 45 m W
Put in airplane f
Bicentennial activities 45 to 90 m N to W
A-18
-------
N
~9 m HIGHER
Traffic Data
Ben Franklin Pkwy. between 20th and
Locjan Circle
1976 AAWT = 31 ,500 vpd
1976 AADT = 29,800 vpd
AMPH = 3,650 vph
PMPH = 2.850 vph
20th St. between Vine and Race
1971 AWT = 8,300 vpd
Race St. between 20th and 21st Sts.
1970 AWT = 2,000 vpd
21st St. between Winter and Race
T97T AWT = 11 ,400 vpd
~4.5m HIGHER
r
6.5m
1
-*.
li.
O I
O U>
DC J
O
Z 6
0 CM
_J )
^~~
-r^
i
•
"
j
i
i
i
i
^
ISm
40m
L-C
4.8m 3.9m
»— PARAPET
Zm HIGH
30m WIDE
-HI-VOLS
Figure A-8. Franklin Institute, placement of monitor on roof.
A-19
-------
FRANKLIN
INSTITUTE
MUSEUM
Figure A-9. Location of Franklin Institute.
A-20
-------
AMS LABORATORY (AMS)
1501 E. Lycoming Street
UTM Coordinates
Northing: 4428.54 Easting: 491.61
General Site Description
Urban trend site in dense residential area on fringe of center city;
possible small impact from traffic on paved roads; on path of pre-
vailing wind from CBD.
Placement of Monitor
Elevation above ground: 4.6 m, on wooden walkway on flat, gravel on tar
roof
Exposure: good all directions.
Immediate Surroundings (see Figure A-10)
Park on the block and then built up
Height of buildings: one to two stories
Roads: paved, moderate dirt
Parking areas: AMS - small paved area; parking on side of street
Curbs: 2 minimal
Sidewalk: 2 m wide, dirty
Traffic volume: moderate to heavy
Speed: 5 kph
Type of traffic: cars, trucks
Traffic signals: at Castor and Lycoming; Castor and Hunting Park Ave.
Other ground level surfaces: trees, grass, playing field.
Nearby Sources
Type Description Distance Direction
Fugitive dust Reentrainment from paved roads, 20 to 35 m W
primarily Castor Ave.
Fugitive emissions None
Nearby stacks Space heating
Recent activity None evident
A-21
-------
H
HUNTING PARK AVENUE
'" '
2 STORY
HOUSES
58
2 STORY
HOUSES
aa
m **
m-
X
BL06
• Clltor Avt
l9>2 AHT -
.
Lycomlng S
— • Park. 197
__ . . • Hunting Pa
i
t§ 3.8
s
s
!p
-------
500 SOUTH BROAD (500)
500 S. Broad Street
UTM Coordinates
Northing: 4421.35
General Site Description
Easting: 485.81
Urban trend site in CBD commercial area, located up and away from
local traffic influence
Placement of Monitor
Elevation above ground: 11 m on gravel roof
Exposure: 2.4 m high building extensions to north and west
(see Figure A-ll)
Immediate Surroundings (see Figure A-12)
Built up
Height of buildings: three to five stories
Roads: paved, light to moderate dirt
Parking areas: paved to north and south of building; paved but dirty
to west of building.
Curbs: 5 cm high, no real gutter
Sidewalk: 20 foot width on Broad St., light to moderate dirt
Traffic volume: heavy
Speed: 40 to 55 kph
Type of traffic: cars, trucks, buses
Traffic signals: on each corner around the building
Other ground level surfaces: few trees
Nearby Sources
Type
Fugitive dust
Description
Vehicular-induced reentrainment -
probably small
Distance Direction
35 m E
Fugitive emissions None evident
Nearby stacks
Recent activity
Low residential, commercial space
heating
Possible recent construction of
Philadelphia Center for Older
People and new sidewalk around
Center
Work on apartment house
50 m
3 blocks
E
A-23
-------
SOUTH BROAD
\
SIDEWALK 6m
1
0
z
ft
a.
o
UJ
.>
a.
^ T f rn
"W~ oO m
• r n
i
i
i
. 8 foot HIGH
BUILDING EXTENSION
'
1
I
I
1
1
1
1
1
i
1
6m
I
i
* m "
23
+%-,
14
__1
6m
,
_— ^
— *
im
A /
A
4
m
J
o
ac
a>
o
o
>•
UJ
-7.5m-*| «n
>
-HI -
VOLS
O
E
u>
e>
z
DIRTY PAVED PARKING
Figure A-ll. 500 S. Broad Street, placement of monitor on roof.
A-24
-------
1.
It)
J3
O
t-
• O
' O
I O
o
s-
co
o
E
T
E
a>
x
o
X
E
ooo
0.(-
ox
1 1
E
8
»-
(S
z
a:
<
a.
UJ
1
o
ll
UJ
O X
Z 0
5 x
UJ
E
>- 10
H -
z
(T
UJ
IT
u.
— S*-H
§
£"
*~ a! x
UJ ° —
*-> CL z
!§ *
"• _J
UJ ^
0
_l
5E
a
•a
o
o
•a
•H
^
tf
^••IHIH
X
(9
X
E
in
*™
^
^>
\ \
V
7A
^
cfl
o
PQ
O
O
0)
M
3
00
•H
A-25
-------
SOUTH BROAD (SBR)
Broad and Spruce Streets
UTM Coordinates
Northing: 4421.65 Easting: 486.00
General Site Description
Urban trend site in CBD commercial area; probably receives a significant
impact from vehicular activity on paved roads; also natural reentrain-
ment from parking lot, streets, and sidewalks.
Placement of Monitor
Elevation above ground: 4.6 m, on AMS trailer
Exposure: good all directions
Immediate Surroundings (see Figure A-13)
Built up
Height of buildings: 14 to 82 m
Roads: paved, moderate dirt
Parking areas: paved, light to moderate dirt
Curbs: 0 to 10 cm, with gutter
Sidewalk: 4.5 to 7.5 m, moderately dirty
Traffic volume: heavy
Speed: 40 to 50 kph, stop and go
Type of traffic: cars, trucks, buses
Traffic signals: at Broad and Spruce (40 seconds, 20 seconds)
Other ground level surfaces: none to report
Nearby Sources
Type Description Distance Direction
Fugitive dust Vehicular-induced reentrainment on
Broad and Spruce Streets
See Fienfe A—1S
Natural reentrainment from streets, B
parking lot, sidewalk
Fugitive emissions None
Nearby stacks Space heating only
Recent activity None obvious
A-26
-------
Figure A-13.
Broad St. n/o Spruce
AT5T • 23,000 vpd
• Broad St. s/o Spruce
1977 AWT • 24,000 vpd
• Spruce St. e/o Broad
T977 AWT » 7,600 vpd
• Spruce St. w/o'Broad
T977 AWT « 7,300 vpd
South Broad at Spruce, monito" location,
A-27
-------
S.E. WATER POLLUTION CONTROL PLANT (S/E)
Front St. and Packer Ave.
UTM Coordinates
Northing: 4417.23 Easting: 487.23
General Site Description
General industrial impact site located in moderate to heavy industrial
area; some fugitive dust likely from field next to hi-vols.
Placement of Monitor
Elevation above ground: 0 m, on concrete slab on ground
Exposure: Reasonably good all directions, AMS "trailer" 6m to east.
Immediate Surroundings
Open
Height of buildings: one story at treatment plant
Roads: service road paved
Parking areas: (at treatment plant) paved
Curbs: 5 cm
Sidewalk: none
Traffic volume: minimal
Speed: 25 to 30 kph
Type of traffic: cars
Traffic signals: none
Other ground level surfaces: grass; bare dirt, railroad tracks to west,
Nearby Sources
Type Description Distance Direction
Fugitive dust Open dirt field - windblown 8m w
Open dirt field - windblown 275 m SE
Fugitive emissions None evident
Nearby stacks Treatment facility stack - 90 m N
black smoke
Industry along river NW
Recent activity None evident except possibly the
clearing of the field mentioned
above
A-28
-------
ARAMINGO FIRE STATION (AFS)
Aramingo Ave. and Huntingdon Street
UTM Coordinates
Northing: 4425.2 Easting: 489.9
General Site Description
In a moderate to heavy industrial area that is also high density
residential; site is primarily source specific (National Lead)
Placement of Monitor
Elevation above ground: 30 feet
Exposure: interference from high parapets on all sides
(see Figure A-14).
Immediate Surroundings (see Figure A-15)
Built up
Height of buildings: two to three stories
Roads: paved, moderate dirt
Parking areas: side of street only
Curbs: 0 to 5 cm, some gutter
Sidewalk: moderately dirty
Traffic volume: medium to heavy
Speed: 25 to 35 mph
Type of traffic: cars, trucks
Traffic signals: stop light all directions
Other ground level surfaces: playing field to north
Nearby Sources
Type Description Distance Direction
Fugitive dust Vehicular-induced reentrainment - 15 m N to W
probably small contribution given
height
Possible natural from playing field 90 m N
Fugitive emissions Possibly from National Lead but none 25 to 90 m S to W
observed on Saturday
Nearby stacks National Lead 25 to 90 m S to W
Recent activity Possibly resurfacing of Aramingo Ave. 15 m W to N
A-29
-------
4*
PARAPET
IVENT 7m
O
5 m
-•^2
I m
SKYLIGHT
18m
VENT
O
AERIAL VIEW
r
2.7m
SKYLIGHT
f*- PARAPET
CROSS- SECTION
Figure A-14. Aramingo Fire Station, placement of monitor on roof.
A-30
-------
2 STORY
HOUSES
* Hutitlngton between Annlngo tnd Richmond
1974 AST • 2.
100 vpd
• Ar«m1nao bttwetn Oakdtle ind Lthtjh
1978 ADT • 18.500
2 STORY HOUSES
Figure A-15. Aramingo Fire Station, neighborhood.
A-31
-------
SW WATER POLLUTION CONTROL PLANT
UTM Coordinates
Northing: Easting:
General Site Description
Moderate industrial site located for the purpose of measuring the impact
of the construction of the pollution control facility.
Placement of Monitor
Elevation above gorund: 0 m
Exposure: 7.5 m building about 4.5 m to NW; otherwise good
Immediate Surroundings (see Figure A-16)
Fairly open
Height of building: 4.5 to 7.5 m
Roads: paved, moderate dirt
Parking areas : not relevant
Curbs: 8 cm
Sidewalk: none
Traffic volume: minimal
Speed: 15 to 30 kph
Type of traffic: cars
Traffic signals: none
Other ground level surfaces: grass, bare dirt at construction site
Nearby Sources
Type Description Distance Direction
Fugitive dust Construction (ground moving) activity 10 m S
Construction activity - building 90 m W
Fugitive emissions From construction site (see picture) 15 m SW
Nearby stacks None evident
Recent activity (above construction)
A-32
-------
No traffic data available.
SILOS
A
HI-VOL
OREA8E
BURNER
BUILOINO
PAVED SERVICE ROAD
HOLDING POND
Figure A-16. SW Water Pollution Control Plant, monitor location.
A-33
-------
MINGO CREEK PUMPING STATION
UTM Coordinates
Northing: - Easting: -
General Site Description
Outskirts of moderate industrial area, located for the purpose of mea-
suring impact of construction of SW water treatment plant to south.
Placement of Monitor
Elevation above ground: 0 m
Exposure: generally good, 4 to 9 m building about 15 m to W
Immediate Surroundings (see Figure A-17)
Open
Height of buildings: only one building - 4 and 9 m high
Roads: paved with gravel on top, dirt road
Parking areas: not relevant
Curbs: none
Sidewalk: none
Traffic volume: minimal (< ADT)
Speed: 15 to 30 kph
Type of traffic: probably cars
Traffic signals: none
Other ground level surfaces: grass, trees, bare dirt and gravel
Nearby Sources
Type Description Distance Direction
Fugitive dust Natural 20 m S
Vehicular-induced reentrainment A5 m S
from Penrose Ave. Bridge
Fugitive emissions Sand and gravel area - potential 180 m SE
emissions
Nearby stacks Petrochemical complexes 0.4 km NE
Recent activity None evident
A-34
-------
No traffic data available.
*
/!
>
FENCE
RR
TRACKS'
9m 4 5m
HIGH HIGH
V.
DIRTY PAVED
A-«- HI -VOL
V
l_
DIRT ROAD
PENROSE AVENUE BRIDGE ~IOOft. HIGH
TO
SCHUYLKILL
R?VER
Figure A-17. Mingo Creek Pumping Station, monitor locations.
A-35
-------
CORPS OF ENGINEERS
UTM Coordinates
Northing:
*
General Site Description
Easting:
Light industrial area, locate- on the Corps of Engineers facility for the
purpose of measuring the impact of the construction at the SW water
treatment plant; possible problem of fugitive dust from on-site activity
Placement of Monitor
Elevation above ground: 0 m
Exposure: 1.2 m platform 4.5 m to north of monitor
Immediate Surroundings (see Figure A-18)
Fairly open
Height of buildings: two stories
Roads: dirt and gravel in immediate vicitniy
Parking areas: dirt and gravel
Curbs: none
Sidewalk: not applicable
Traffic volume: very light
Speed: 15 to 30 kph
Type of traffic: cars, trucks
Traffic signals: none
Other ground level surfaces: mounds of dirt including dredging
disposal area, bare dirt.
Nearby Sources
Type
Fugitive dust
Description
Distance
Direction
Natural reentrainment from bare dirt General area All
Vehicular-induced reentrainment 3 to 30 m E to S
Fugitive emissions Activity in area
Nearby stacks None evident
Recent activity None evident
A-36
-------
DREDGING DISPOSAL AREA
TO
CONSTRUCTION
ACTIVITY
t
BLDG.
I BLDG.
Figure A-18. Corps of Engineers, monitor location.
A-37
-------
APPENDIX B
TSP CORRELATION ANALYSES
B-l
-------
TABLE B-l. LINEAR CORRELATION COEFFICIENTS BETWEEN 24-HOUR TSP LEVELS
AT 10 PHILADELPHIA MONITORING STATIONS (1974)
Site
DEF
ALL
INT
BEL
ROX
N.E
FRI
LAB
S/E
500
Note:
DEF
1.00
0.64
(47)
0.72
(53)
0.59
(43)
0.56
(45)
0.63
(49)
0.74
(52)
0.79
(53)
0.80
(37)
0.78
(45)
Number
ALL
0.64
(47)
1.00
0.50
(49)
0.50
(40)
0.44
(42)
0.51
(46)
0.74
(47)
0.62
(47)
0.40
(36)
0.62
(39)
INT
0.72
(53)
0.50
(49)
1.00
0.39
(43)
0.44
(47)
0.60
(51)
0.56
(52)
0.54
(55)
0.67
(38)
0.58
(47)
of pairs is
BEL
0.59
(43)
0.50
(40)
0.39
(43)
1.00
0.56
(37)
0.55
(41)
0.72
(43)
0.61
(44)
0.53
(31)
0.62
(36)
enclosed
ROX
0.56
(45)
0.^-4
(42)
0.44
(47)
0.56
(37)
1.00
0.68
(41)
0.61
(43)
0.61
(46)
*
0.19
(33)
0.52
(41)
N/E
0.63
(49)
0.51
(46)
0.60
(51)
0.55
(41)
0.68
(41)
1.00
0.54
(48)
0.68
(49)
0.47
(33)
0.53
(41)
FRI
0.74
(52)
0.74
(47)
0.56
(52)
0.72
(43)
0.60
(43)
0.54
(48)
1.00
0.70
(50)
0.67
(34)
0.80
(42)
LAB
0.79
(53)
0.62
(47)
0.54
(55)
0.61
(44)
0.61
(46)
0.68
(49)
0.70
1.00
0.54
(225)
0.68
(275)
S/E
0.80
(37)
0.40
(36)
0.67
(38)
0.53
(31)
(33)
0.47
(33)
0.67
(34)
0.54
(225)
1.00
0.53
(22'8)
500
0.78
(45)
0.62
(39)
0.58
(47)
0.62
(36)
0.52
(41)
0.53
(41)
0.80
(42)
0.68
(275)
0.53
(228)
1.00
in parentheses.
Asterisk indicates that correlation coefficient is not significant
at the 5 percent level.
B-2
-------
TABLE B-:>. LINEAR CORRELATION COEFFICIENTS BETWEEN 24-HOUR TSP LEVELS
AT 13 PHILADELPHIA MONITORING STATIONS (1975)
Si to
I)KF
ALL
INT
BKL
ROX
N/E
NBR
FR1
LAB
SBR
S/K
500
AFS
UEF
I. 00
0.61
(54)
0.56
(56)
0. 72
(29)
0.64
(56)
0.65
(49)
0.66
(51)
0.68
(53)
0. 78
(57)
0. 75
(57)
0.59
(40)
0.67
(49)
o . a/i
(54)
ALL
0.61
(54)
1.00
0.52
(52)
0.45
(27)
0.36
(52)
0.66
(45)
0.47
(50)
0.68
(49)
0.66
(54)
0.55
(54)
0.27*
O8)
0.47
(46)
0.66
(52)
INT
0.56
(56)
0.52
(52)
1.00
0.66
(29)
0.51
(55)
0.56
(48)
0.41
(49)
0.60
(52)
0.63
(55)
0.44
(55)
0.47
(39)
0.43
(47)
0.58
(52)
BEL
0.72
(29)
0.45
(27)
' 0.66
(29)
1.00
0.89
(28)
0.80
(24)
0.87
(27)
0.85
(27)
0.86
(36)
0.86
(29)
0.48
(29)
0.76
(34)
0.87
(26)
ROX
0.64
(56)
0.36
(52)
0.51
(55)
0.89
(28)
1.00
0.68
(48)
0.66
(49)
0.65
(52)
0.78
(55)
0.64
(55)
0.46
(38)
0.68
(46)
0.54
(52)
N/E
0.65
(49)
0.66
(45)
0.56
(48)
0.80
(24)
0.68
(48)
1.00
0.43
(43)
0.72
(47)
0.91
(48)
0.69
(48)
0.54
(36)
0.65
(41)
0.73
(46)
NBR
0.66
(51)
0.47
(50)
0.41
(49)
0.87
(27)
0.66
(49)
0.43
(43)
1.00
0.56
(47)
0.66
(51)
0.72
(51)
0.30*
(34)
0.66
(42)
0.62
(49)
FRI
0.68
(53)
0.68
(49)
0.60
(52)
0.85
(27)
0.65
(52)
0.72
(47)
0.56
(47)
1.00
0.80
(58)
0.74
(52)
0.49
(43)
0.63
(50)
0.79
(50)
LAB
0.78
(57)
0.66
(54)
0.63
(55)
0.86
(36)
0.78
(55)
0.91
(48)
0.66
(51)
0.80
(58)
1.00
0.81
(56)
0.54
(240)
0.78
(283)
0.82
(53)
SBR
0.75
(57)
0.55
(54)
0.44
(55)
0.86
(29)
0.64
(55)
0.69
(48)
0.72
(51)
0.74
(52)
0.81
(56)
L.OO
0.34
(39)
0.65
(48)
0.83
(54)
S/E
0.59
(40)
0.27*
(38)
0.47
(39)
0.48
(29)
0.46
(38)
0.54
(36)
0.30*
(34)
0.49
(43)
0.54
(240)
0.34
(39)
1.00
0.49
(237)
0.50
(36)
500
0.67
(49)
0.47
(46)
0.43
(47)
0.76
(34)
0.68
(46)
0.65
(41)
0.66
(42)
0.63
(50)
0.78
(283)
0.65
(48)
0.49
(237)
1.00
0.73
(45)
AFS
0.84
(54)
0.66
(52)
0.58
(52)
0.87
(26)
0.54
(52)
-0.73
(46)
0.62
(49)
0.79
(50)
0.82
(53)
0.83
(54)
0.50
(36)
0.73
(45)
1.00
NOTE: Number of pairs is enclosed in parentheses. Asterisk indicates that correlation
coefficient is not significant at the 5 percent level.
B-3
-------
TABLE B-3. LINEAR CORRELATION COEFFICIENTS BETWEEN 24-HOUR TSP LEVELS AT 13
PHILADELPHIA MONITORING STATIONS (1976)
Site
DBF
ALL
INT
BEL
ROX
N/K
NBR
FRI
TAB
S»R
S/K
500
AFS
DBF
1.00
0.81
(50)
0.69
(53)
0.70
(25)
0.49
on
0.76
(54)
0.65
(54)
0.75
(53)
0.78
(55)
0.78
(54)
0.89
(38)
0.79
(42)
0.77
(50)
ALL
0.81
(50)
1.00
0.57
(48)
0.66
(22)
0.45
(48)
0.60
(52)
0.65
(52)
0.75
(51)
0.70
(52)
0.76
(52)
0.81
(36)
0.85
(39)
0.75
(49)
INT
0.69
(53)
0.57
(48)
1.00
0.37
(25)
0.50
(49)
0.70
(53)
0.55
(53)
0.62
(51)
0.74
(53)
0.67
(52)
0.61
(36)
0.65
(39)
0.59
(48)
BEL
0.70
(25)
0.66
(22)
*
0.37
(25)
1.00
0.88
(22)
0.78
(25)
0.85
(25)
0.91
(24)
0.78
(25)
0.81
(23)
0.73
(16)
0.95
(18)
0.56
(22)
ROX
0.49
(51)
0.45
(48)
0.50
(49)
0.88
(22)
1.00
0.65
(52)
0.60
(52)
0.67
(52)
0.66
(53)
0.62
(53)
0.46
(38)
0.59
(40)
0.67
(48)
N/E
0.76
(54)
0.60
(52)
0.70
(53)
0.78
(25)
0.65
(52)
1.00
0.63
(58)
0.70
(55)
0.85
(56)
0.71
(56)
0.75
(40)
0.74
(43)
0.71
(52)
NBR
0.65
(54)
0.66
(52)
0.55
(53)
0.85
(25)
0.60
(52)
0.63
(58)
1.00
0.76
(55)
0.78
(56)
0.89
(56)
0.60
(40)
0.75
(43)
0.72
(52)
FRI
0.75
(53)
0.75
(51)
0.62
(51)
0.91
(24)
0.67
(52)
0.70
(55)
0.76
(55)
1.00
0.75
(55)
0.85
(55)
0.78
(39)
0.83
(44)
0.68
(51)
LAB
0.78
(55)
0.70
(52)
0.74
(53)
0.78
(25)
0.66
(53)
0.85
(56)
0.78
(56)
0.75
(55)
1.00
0.81
(56)
0.68
(248)
0.84
(283)
0.87
(52)
SBR
0.78
(54)
0.76
(52)
0.67
(52)
0.81
(23)
0.62
(53)
0.71
(56)
0.89
(56)
0.85
(55)
0.81
(56)
1.00
0.74
(40)
0.81
(43)
0.79
(52)
S/E
0.89
(38)
0.81
(36)
0.61
(36)
0.73
(16)
0.46
(38)
0.75
(40)
0.60
(40)
0.78
(39)
0.68
(248)
0.74
(40)
1.00
0.70
(232)
0.84
(37)
500
0.79
(42)
0.85
(39)
0.66
(39)
0.95
(18)
0.59
(40)
0.74
(43)
0.75
(43)
0.83
(44)
0.84
(283)
0.81
(43)
0.70
(232)
1.00
0.77
(42)
AFS
0.77
(50)
0.75
(49)
0.59
(48)
0.56
(22)
0.67
(48)
0.71
(52)
0.72
(52)
0.68
(51)
0.87
(52)
0.79
(52)
0.84
(37)
0.77
(42)
1.00
Correlation Is not significant at 5 percent level.
B-4
-------
1.0
0.9 -
0.8
\
\
1- \
0.7
0.6
0.5
0.4
0.3
0.2
O.I
-------
1.0
0.9
0.8
0.7
0.6
"uO.5
0.4
0.3
0.2
O.I
LA6/NE
. *
• •
• BEL/NE
• FRI/NE
•SBR/NE
• ROX / N E
• 500/ «DEF/NE
• NE
INT/NE
•
J 1 1 ,1
j 1 i i
0 2 4 6 8 10 12 14
SEPARATION DISTANCE, MILES
16 18
Figure B-2. Square of the correlation coefficient between TSP levels as
a function of distance (1975).
B-6
-------
APPENDIX C
FURTHER DETAILS OF THE EFFECT OF PRECIPITATION
ON TSP LEVELS AT THE AMS LABORATORY
In the investigation of the effect of precipitation on TSP levels, com-
puter plots were prepared of 24-hour concentrations versus precipitation summed
over three different time periods. The plots were prepared from TSP concen-
trations measured at the AMS Laboratory roof site during 1974 and 1975 and
airport precipitation amounts. Because of the very significant reduction in
average concentrations noted at this site during weekends, weekday and weekend
data have been separated for most of the plots. The data have also been
stratified according to the distribution of precipitation during the 48-hour
period ending at the end of the particulate sampling period in four of the
plots. The plots are displayed as Figures C-l through C-10.
Inspection of Figure C-l shows that concentrations above the secondary
standard very rarely occurred when any significant amount of precipitation
fell during the 48-hour period, and that only one value in excess of this
standard occurred when the 48-hour precipitation exceeded about 0.3 inches.
Similarly, Figure C-4 shows that no violations occurred when precipitation on
the day of observation exceeded about 0.14 inches. In general, the greater
the precipitation, the lower the maximum observed concentration. However, it
is also apparent that low concentrations are frequently observed on days with
little or no precipitation during the 48-hour period. Thus, the overall
effect of precipitation, from a statistical standpoint, is to reduce the number
of, or eliminate, days with high concentration. In Figures C-4 and C-5, and
C-5 and C-6, respectively, concentrations have again been plotted against
the 48-hour and 24-hour precipitation amounts, but the data have also been
separated by weekdays and weekends. In Figures C-7 through C-10, the weekday
and weekend observations have been separated according to the distribution of
precipitation during the 48-hour period. In Figures C-7 and C-8, only days
with no precipitation on the day of the observation are included; the amount
of precipitation indicated is that measured on the previous day. In Figures
C-9 and C-10, only days with no precipitation on the day prior to the ob-
servation are included and the amount of precipitation indicated was measured
on the observation day.
To illustrate differences in the distributions of TSP concentrations
measured during dry periods with those associated with periods of precipita-
tion more clearly, Figures C-ll, C-12, and C-13 have been prepared from the
computer plots. In preparing these figures, observations made when the precip-
itation was measurable but less than approximately 0.22 inches have arbitrarily
been eliminated.
C-l
-------
o
o
o
o
o
o
CO
o
o
X
I
o
7
o.
z t-
o
UJ O
a. o
* . *
»i #
»
* ••#»*#
*
* * # w #
4 * (M
^£££«?2«i*1ti*S;J?Sa22T;JC;!!lS*M
O O O O Q O
O -O fM CO •*• O
• ••••!•
1ft U\ *O ig t*« CO
1 » + . r_|
1 1 -f "J
I ! «
1 1 >— '
i * ! C
1 1 o- O
! ;«? 3
! ! m «)
1 1 -H
1 T a,
1 I -H
j I? o
1 1 rf\ M
J P.
( 1
I i
i M
! 3
i L °
! i°. f
* 1 m oo
<+
i 1 ^
«M»^«M j __, MM K, _, ^ _>-,.-,,-.« ~ ~ _H „ M 1 ^
1 1 3
»
i » . cu
* i CM >>
i ! e
t °
» * # i *rl
# 1 -0 •*->
• » i" *
i * i *\j
J 4-1
* I C
* 1 9
* * 1 *" Pn
» i cn
* * * 1 H
* IM * * M
* » * # i c 3
* * i ^ O
» » * * T -• J3
* * » | 1
* ' t-i
r i3
j * * # » * °
• '» » OJ 1 ~0 "V
*.f • » I*, jk
1 » m * 1 o 4-1
» » 1 * **** » I r-.
* i « i s
1 * » N M N * 1 -l I-. + « «-l M 11 » >-. W M W * H. H M « + .Q .^
O O O 'O $
•0 N «D * S
• • -» • • » •
CO (T > p x
o- f* m * • w
C-2
-------
O O O O
O -Q -g CO
• • * «
lO lf> ^O O I
t- • +
o a. 1
V}
II '
h- r
•« X (M »
uj O <*J
of o • * »
2- ° ,
U. + » » * » * o- r~ m
o o a o o o o
^ O -O OJ *J 'V
t 1 # »
I 1 • I
j I |
i i t
*
i i i
* i i
• i *
• i
i i i
1 ! 1
i i i
i i i
1 1 !
! ! T
j i i
! : i
! i i
i i i
! ! i
! ! * !
i i i
i i i
1 _ j M M -. Ml- 1
1 1 1
! i !
j j |
i i I
i
i i i
i I *
i i * ** * !
i i i
1 1 « * * T
i i i
i i * i
1 ! *i
1* i i
! » i
„ i
* i i
! I * T
i * *
i i * * i
* N *
i # i
* i
* » T
* « i
» 1
* +
» * j
» (M j
# » * 1
» » 1 * * +
1 * * * 1
* 1 » * j
* 1 *
1 * » M # I
•t t% * . * M
«* l**m*« » I
. * ! N N*. *»5 * i
* i5?N ,* * .*8 « I
• IN* i
» * 1 » N (n N »*(M 1
*« «* Oil »N**CM* +
* j * * * * in i
* • * # » N » ***ni»N i
* * i*nj rvj » » » ficvj m*« • i
«NC('O-*c>ir-(OO>^)me»><^h-f~O'c>'«oo>o>o>o>»inooo>*»»» +
? g S S «? S §
N «o n o< v o •*
vi •* o- •• r- »> i * M
N
rn
*
o
00
«1
•*
fT.
(NJ
o
m
IA
CM
<0
•"«
rj
fr>
•
"^
o
^
00
o
m
o
o
•
o
s~^
^T
o
co
rt
3
a)
cu
ti
x«x
c.
o
•H
cfl
J-l
>r_{
a
•H
^j
01
v^
a
3
O
1
oo
co
3
co
0)
>
c
o
"^
CO
M
4-J
c
0)
^
o
(J
p ,
CO
E-i
j_,
^
O
rf^
1
3
O
*
t
^*»
Twent
CN
i
o
CD
M
t>0
•H
to
C-3
-------
00
IS
0.
•o
r-
V
^
•s
a>
0
<
J
Ul
0
rf
a
M
u
g
ox
§
o
I
1
CD
m
a:
— •
t/i
1
u
•A Ul
r- en
V CO
81
t
Ul
K
3..
(/I
O *"
M —
< >
28
o-
^
j
ii
§
i-
•
m
J-
N
,
CO
^ .
(M
H
fsi
0*
<-*•
.
•a
«O
^
•
t-f
•
in
r-t
r^
•
0
t '
•*
O
m
0
m
• i
o
a o o o
e N -t «
2 * * 2
^
•
,
•
*
•
•
•
^
h
•
h
*
•
.
*
*
,4MMWM4M*M^I**»«!MJ
o c o e
« i -t ID
O O O
CO O (M
5 2 ™
»
«MM1_I_««««M1-,M«M>-,
'
* *
*
*
,
<
*
*
t
t
N * 4
* * *
j*»t ***R«S:«,
o o o
M o IB
M rt
0 p O O
-t 5 « / o
S 5 « . 3
* <
4
•
'
i
H
*
*
•
* 4
* * *
. *
if
*
.
* # *
<
«_,««MM«««M,-l«M«««H.
N
* •
N
•
S
^ •
(M
0>
> i
fH
.
>r
\/\
•
co
N
' Ji
:
S
• «
-
•
f-
'5
.
f^
m
o
•0
•
O
O
• o
7;
iH
C
o
CO
C
2*
^
4J
e
-------
o
UJ
O i
* I
• I
CO
>%
n)
X
r-
c
•tt
± -
LU &
CL •
U1
T
O
>*
•
^ '•!
3
*•*
H-
4
H-
1-4
a c
»~* c
o •
it ra
t
a.
Q.
<_> (*•
uj •
•r ,-1
Q.
l/»
ut
O
>V 1-4
' ) if*
«J. •
•— _J
r-t
i
l
LD
O •/)
»— t (7s
1— *
net
t—
y
UJ
o
y:
C rr
'J "O
• 4
V , 0
-I
<
* H
0
a.
to
t-
7.
"* -T
O -«
0 *
~ o
a
0
i i ;
; , , '
t- i '. *"
1 i ' !
1 ! '
• i ' i
11 i i
iii , !
• , ;
i '
i , ;
i '
i ;
i ' i
1 • ••
i
i
i ,
i
1 «
1 !
1
11 1
1 ' ' ' :
1 i
;l ' I '
•i i ! '
i
:' ' ,
] • ; '
! ' i * '
1 ' ' !
1 : '
1
1
. » '
i '
1 ! ,
1 i i
i 'I '
11:
1
!
'
i
,1 • , *
1 ,
1-
1 . '
' !
1
ii :
i
i
i ' '
1\
.
1 i
1 ,
1 ' '
1 1
1 '
I '
i *
i *
'I 1 . '
i ' , ; i '
! : '
i •
II: ; i
> ! i «
' ! I '
; ' !l i • *• ' i
. .! i » i ! *•
- j ! * . ! ;
i . ; • , ' *
i i i
» i i <
i i *
^ t ' .
1 i « *
1 ' i» J * »
1 . * ,
i '
1 » : * » » 1
* # 1 » » « * «
i ; ; • *
* '1 l * * f
»I*»^»IM!»I«I»I
• + »"i»HIH»-« + IH*-(>Hl-l + t-l»-IM*M + »'oso^Oko^o^otosm
^
i
u
cu
C-5
-------
o
O'
in
•H,' >
- O
VI M -1
>• V W
IU *
SI w 5
S 8*. °'
& g"
afr-4 •-•
H. Z
«e < i •*
C tuo «
ac Q i
aj o— O
H U.
juO
^ iS
a. O ac
UJ fl£
O 111
« »-
i
•
•
!
;
;
1
•
1
1
*
» *
•*«** »CM •»»»Nm-4v«tni
§O O O i
•e N «
2 S r~ ;io ,'l
N rt -> •»
i I : i
: i
; '
i i
1
, |
i
i ! •
'
i j
1
1
•
! ' i '
• #
i
1
i
*
i ' ' i
: *• '* •
! *
i
<
! ' 1
* ' !
,» *
# ' * t
1 !
i *
' | ^
: ' 1 *
.• -r *: t . '
: * *
>m « f- mf- m e^3 «ojo> r> «o^ e> »,]
Ot O O,
^t O fO'
! f*- '» • ' «o
^i t^ ! CK
1 I r
i ' :
, !
!
,
I
,
i !
1 , • j
#
i
,
: * i
»
1
• j
i
,
1
j
; , «
#
'
* »
i » :
^
*
I
1 ,
•
i
.
i
M«MWW1-,l-1WWM$1W,,«MM.-1
,
*
*
«
» * *
» < •
: « ; * *
* ;
* . .
* * '
* »
M ' ' #
* * »
* i »•
» » * »
* N t* ** *
# # # ***
* * » » (M « * »
*»> * * *»*» *
*«ift IIM» Si5 5
MMIHM4
0000
CM (0 * O
O^ O O , »•*
,h- . m '•«• IM
l^X
0)
in S
' 1
r- §
^ >*
• ""
N «)
01
CO
S r°
. 0
N
CJ
o
c
^ *H
• 4J
" efl
.pi
(X
•rl
« CU
* J-f
•" a
CO
3
co
CO 1-1
C
U
2 «»
'3 tl
"" CJ
I
[(I
en §
OD ^
t
H
M
«' f
Wl
o
14-1
« 1
^ ^"S
O "^
0 .
1 •
.0 ^
u
1 3,
£M O
LL V)
C-6
-------
,
1 I '
' ' i
! o
"•
«r +* *
-^ 1
OC CO |
u-> r- |
1M 0. . +
-1 j
£ i
I 1
UJ O +
o z! i
< M. I
a. OS j
1 1 , in I
* *
ol ""* 1
' P ' 1
•O M, 1
>. M * 1
(M lit M, j
•x , ad I
K ' £| 1
O 1
1
' MI •v* 1
tjj (M 1 M
ti] • +
CC -i 1
Q- 1 1
~ !
t/^ +
w 1
Ol 1
ol l*v 1
°i ° '
~ "*' j
'
^
1
•» 1
! m 1
* +
Kl C 1
I; 'i
"•* i
6' ITl I
t-i * 4
^' O 1 *""
f£- 1
1- 1
^ i
•n +
— Z 1
*r o c~ I
c- o «t 1
— v c 1
rvi -r 1
• N «, 1
,i O CJ +
i
n I
• i T 1
»- » +
1
u. n c I
cy "~i * •*•
<> -1 c t
I 1
I
u J C* • *
• i r
s
\
t-t
-4 »—
O
<
"*
MM
. -
•
O
0)
•
sO
' M
•-4
MM
'
1
|
!
'* 4
1-4 *-
;
i
i
|
1
t-t !_.
MM
|
M
IMH
O
e
•
^4
0
*~l 1-4
!
1
*
! ^
o
i &
\~
j
^
hH H"
MH
5
*
» <•<•
g
S
^* ^
MM
:
tf
# *
: #
Ml
•
1 '
: i*.
i » + •
1 , 1 M
1 1
1 i 1 1
; t
i ; i .. L
; • ""
1 i
i i
1 !
! : i IS
i T-
i
i
i
i ' ;
i , ' • 1 -
. 1 m
,!..; „:„!"
i • . ' ' • ; If
: ' ' 2
i j i i
i ; . !il-
1 : < .
i i
i * +
! ' !
! ' i '1*
* il°
: » , |
1 ' • • +
i ' '
i » ,i
i i < i IT
1 ' | » * . \\ >-
1 I c
I ' * i
i ' ' ,i
' i i
l +
' ; ' t
' 1 -c
t 1 ^
, ' «• +
1 1 C
1 ' 1
1 1
1 * i
1 1
1 * * i
I Mr-
# . i f
1 ; +
1 * 1C
I * i
I 1 i
« i
* , I
I
1 i : ' ' i •-
* » +
#, » ' ' 1 C
tt 1
* r- |
# # '|
# +
1 f 1
*! ' ••»* 'i»]#' ,lc
>o tw ^ ro1 »t CT^I «j p^ ^ f*- i/* f-i ^ m ^g f^; pj # ] +
1 C C f O ' OJ
j- •<> cr c
, 0^ t^ r- MJ
' 1
.
i^P
a
o
CO
t3
g
_s
cu
cu
•jj
>s
C0
T3
C
O
•H
• i
E
0)
CO
o
1 o
Q
o
..J
4-1
CO
4J
•H
CX.
•H
1
'
o
.^
4->
: «d
M
1 (3
CU
o
s
o
CJ
PL,
C/3
H
M
3
: o
43
1
1
I
• r>
4J
fj
(U
EH
'
* •
o
a)
M
•H
fe
C-7
-------
«-*
(-4
S
a
•o
V
r-
03
0
fr*
.J
IU
^
_1
S
>
IU
•n
o
a.
t-
UJ
or:
o
u.
UJ
CD
x
4
O
Z
a
o
a:
a.
04
2
ac
i/l
0
rf
O
.-.
•O UJ
f- 1-
^ M
X 03
S<
_l
II
UJ
^
o a
«/>
5 *~
H<* **
K Z
IU O
a. a
o —
u_
Ul O
15
is
Ho
•4
• •
-*
.
•H
•
rM
,_,
H
wf
O
„••
1
03
00
• H
0
.
m
*:.
°
i
»n
• -
°
•
•H
^
O
in
t*
o
S
o
•
O O i' O
o IA o >n
(^ co 'A f*
•-* r-i ^^ <-i
>•
K
>
h
!.
k
•
^
k
*
#
*
* *
* *
• »
• tl-ll-ll-ll-l + l-tl-ll-tM + t-lt-ll-ll-l + lj-l*
0 0 0 0
o in o in
M eo in »t
*4 r* ** 94
O O 0
Q ijfS O
^ •
\ o
j
*
o
1 00
Io
+
s
+ •
1 °
1
*
» + «
o
t*\
» » m
• ' •
o
i>
. •
* 0
*
«
r-4
» • O
KM » • O
o
o
R
C
O
C
o
"^
4-*
(0
4-1
a
•H
U
(U
(X
O
C
0)
M
0
y-i
yj
••^
^
«
*T3
c
o
C
o
*rt
0)
4-1
a
o
0)
M
a.
co
3
(0
M
>
4-1
C
J»
•
0
0)
l-l
3
60
^
P^
O
co
cfl
(U
0)
Jj
g
•H
4-t
E
cu
co
(f)
O
14-1
o
CO
C-3
-------
e w j o
o o o o
•< i- ,
1- f -J (M
HI +
^ 1
CD
O OJ 1
-0 1
a. . +
£ " !
^ 1
-> n I
r- u • ^
~ UJ -. 1
*» ac
» a
O 1
4
IM «
5 ^ i
t£ 1
1
— 1
(/) 4
SO
0
OC *
U 0>
- °'l
1
1
;
fi.
0
:
0
•0
• •
0
'
„
•O UJ (0
Ct; *..
5co o
•V03
§3 •
II
IU rvj
H~ • '
< O
a a.
I/)
Z f~
M —
£i CK
UJ O 0
oca • •
O— 0
u.
UI O
*-• a z
i is
-J Z 19
ui oc
O UJ
#
1
•
§000
o o o
• • • t
-« t- m <>
r- « * S
4
^
(
<
* . s *
§ § §
• tfv »* ^
ft 0 «B
* + •
1 -
.!
i
+
* i
i
i «->
!"'
i
i
i N
1 0
* + •
1 -"
i
1
i
L
1 10
1 O
* 1
* i
i
is
« •
|0
1
1
„«._„«««««««.««««««!
1 N
1 m
# f •
• 0
#
* '
* \t\
1 10
1 O
» i
* i
* * 1
i ID
» 1 -«
» * •
O
**
*
4» * *
« » * » N» o
* + •
O O O O
o o o o
• • . « ft
P a '5 s
T)
C
•rl
CX
O
OJ
CO
n
C
O
C
0
C
4)
U
I ITS
o W O
JK M-l
*-> O
§ *
^^
00
(U
60
C-9
-------
c
o
S
UJ
£ .
o
u
UJ
^ K r"
5 I -
o -*•
UJ • 4
It <-l I
S
oc u\
U N
T
CD t
O I
2 "'I
i I
o o !
t-i 00 I
I- • *
z
Ul
u
— z .
-
if f-1
CX *v »-i
O
a.
»-*
O
O Q
II
UJ
g. °?
zJ2 !
a
UJ O <-<
or Q • •
o — o
Ul
51 u.
* Ul O
I/)
U
Ul
Is
oae
z o
a
UJ
35
o
o
o-
o
IM
O O 0 O
o
o> o •"" MWI-ll-l!4wiM».4l-«4l-«l-»l-
O O O O
to o- q N
I- in * N
r»
* * •
1 p*
I
I
4
i
» 1 ^
i . > i
# » ; •» * i t
* , 1
IN. 1 »
* * * IN
* -f •
rw 1 c
» * * * 1
(M 1
* * i
IM IM * * j
N * 1
* * IS
* # + •
<^1*>-*»^^tW*»-t»M^4»-l4>-tMMM4 «O
CO
•H
._j
•M
o
(U
M
P.
O
a
h^
td
T)
fl
0
4J
observa
g
§
•H
4-1
ct)
4J
(X
•rl
0
CU
£j
P-
CO
(0
j^
CU
c
o
•H
4J
S
C
CU
U
C
o
0
p^
$2
^J
3
|
1
§
<4-l
(^
^
S
IH
o
O
o-
o
IM
•t
m
e
HI .
o
«D
-o
o>
o
R
b
§
O
ca
>,
(fl
Oi
CJ
CU
i,
C-10
-------
*M
Ul
(9
a.
•o
t-
^
N
f>- '
O
>•
<
0
i
>
UJ
ee
a
z
0
a.
^4
U
a.
i
^
>-
o
UJ
-J
a.
z
<
V)
a
Ul
ac
a
UJ
-i
2
^
*A
ft
<
Q
§
a
u
UJ
oc
a
s
T
CM
a
^
u
or
a
^
V)
t/i
2
u
*-
fO
1
z
o
K
/76)
CONCENTRE
<-" >-
IM .J
si
it
m
a a
V)
K Z
< X
Ul o
t£ a
o —
U-
< x
z o
a:
ut
Ul >-
-J <
t* 0
u~ v\
•
09
• '
»*
t '
o
'
^
CO
0
^
0-
•o
o
IA
o
>t
«T
* <
o
•
O
O
•
p-
0
*
o
O 0 0 O
o m « •>
O.^O (** O
•-• «-4 •-» ^4
•
1,
•
K
h
11
:
1
.
i
,
•
'
>
,
^
,
•
•
*
H
.
•
'
•
.
'
I ' .
i i
I
Y i
1
« * »
o o o o
o m -o er-
o -o m o
M 2 2 2
1
! o
' «
1
,
(
p
t
,
I1
,
i
[
i
i
i
.
i
,
i i
•* M M +
s
CO
1
o
*
J«
00
t
i
I
t
1
,
] .
'
1
'
,
1
'
.'
\*.
I*
*
*
H IH >-c IH » M
o
m
in
to
o
N
f*
1
* f
1
i
*
I
I ;
,
i
(
i
1
in
* '
-4 M IM ^ K4 M »
O
•
• 0
C
O
•H
4-1
n)
•H
a
•H
CJ
0)
t t
a
o
^
«
^
rt
C
o
•H
4J
n)
a>
JQ
o
JPH
0
o
*H
^ *
(1)
4J
•H
Cu
U
0}
M
a
CO
3
co
^
>
C
o
•H
4J
aj
4-1
^)
CJ
g
O
PM
CO
H
^
Jj
O
f
1 O
14-1
1
Lt
C
4)
rS
o
rH
1
U
-------
30 p
u
« £0
&
ft
>
U 15
UJ
'0
V* 48-HR PRECIP.>0.22in
\
NO PRECIP. IN 48hr$
0 20 40 60 80 100 120 140 160 180 200 220
TSP CONC.,/ig/m3
(a) WEEKDAYS
r
35
30
: 25
»
J
[ 2C >-
S "0
NO PRECIP. IN 48 hrt
48-HR PRECIP. > 0.22 In
20 40 60 80 100 120 140 160 180 200 220
TSP CONC. ,Mg/m3
(b) WEEKENDS
Figure C-l-1.
Comparison of TSP concentrations on days
without precipitation in a 48-hour period
to concentrations on days with 48-hour
precipitation > 0.22 inches
C-12
-------
30 r
u
h»
« £0
*
i '8
a 10
V«-24-HR PRECIP. >0.22 in
NO PRECIP. DURING
OBSERVATION
0 20 40 60 80 100 120 140 160 160 200 220
TSP CONC.,/ifl/m3
(a) WEEKDAYS
U
33 r
30
25
20
13
S
DC
~\
h«-24-HR PRECIP. > 0.22 in
NO PRECIP. DURING
OBSERVATION
0 20 40 60 80 100 120 140 160 180 200 220
TSP CONC.,yug/m3
(b) WEEKENDS
Figure C-12. Comparison of TSP concentrations on days
without precipitation to concentrations
on days with precipitation > 0.22 inches
C-13
-------
35
30
c 25
v
o
20
Il5
2 'o
QC
b.
5
PRECIP. > 0.22 in. ON DAY
OF OBSERVATION.NONE DAY
BEFORE
PRECIP. > 0.22 in. DAY BEFORE
OBSERVATION, NONE DAY OF
OBSERVATION
20 40
60 80
TSP
100 120
CONC.
140 160 180 200
Figure C-13.
Comparison of TSP concentrations on days with
precipitation > 0.22 inches following days
without precipitation to concentrations on
days without precipitation following days
with >0.22 inches of precipitation (week-
-»
days only)
C-14
-------
APPENDIX D
MONTHLY AVERAGE CONCENTRATIONS
OF METALS AND SULFATES
D-l
-------
00
a
CO
w
CO
CO
w
2
o
CO
o
M
H
H
W
I
w
rJ
H
z
o
s
W
CO
C
O
M-l -r-l
O -U
O rJ
2 CU
CO
rO
O
df
CO
o?
00
"O
o
3
e
cu
£
fj
4-1
a
o
S
m
vO
r-i
r-i
vO
CN
O
CN
en
O
O
0
o
0
m
CN
0
o
vO
»^-
r-l
o
en
CN
O
o
ON
ON
O
OO
r-
0
vO
p^
^^.
^
en
r.
r-
r— 1
CN
en
O
en
in
O
i— i
0
o
o
00
r-l
O
o
ON
CN
r-l
O
00
CN
0
o
ON
CN
i— 1
o-
i— I
r-l
VO
r^
^^
OO
•*
O
-------
00
0
00
is
M
rJ
^
H
CO
w
H
<^
[~T!
l_3
1 — >
CO
Q
5
**•
CO
5;
H
W
fa
O
CO
M
H
«
H
2
S
2
O
^_2
w
e>
w
£>
<^
D-3
-------
CO
oo C
a o
"•**. '1 j >r4
oo o -u
3. ca m
Z OH
(-4 Z OJ
CO
H .£>
Z 0
r-l
H oo
O *d"
OO CO r-4
w
H
< CO
pti oj *^
13 O
O CO
Z 60 •*
O
CO
<; co
H 0
o
o
in
CO r-4
Z -HO
M O
H
« oo
H vO
Z D i-i
CJ> O
z
o
O i— '
CO
w Co
2
w
^^ r«H
0)
>* fa r-l
H j-
o
•
I—I
CM
in
•
0
r-».
r^.
^^
i—4
•*
CM
ON
O
-*
0
vO
vO
O
CO
O
O
O
r**,
i— i
O
0
CM
m
o
o
CM
CO
o
0
CM
o
•
r-l
o
r*^
•
0
r^
i-»
•**^_
CM
o-
1-
ON
1
l
m
vO
0
^j-
0
o
o
o
CM
0
o
^
*J-
o
o
vO
CM
O
O
CO
o
•
1— 1
^
VO
O
r^.
r*^.
>^^
CO
m
r-
r-i
r— i
oo
vO
0
I-l
vO
0
CO
o
o
o
r-4
1-1
0
0
^
f-».
o
0
m
*^-
o
o
CO
CO
•
r-4
o
ON
O
1 —
r**.
••^
-*
vO
"*
D-4
-------
no
,6
00
2
M
p-l
W
H
"*•
OO
w
H
pti
>J
^
CO
Q
5-
CO
_!
^J
H
§
O
2
O
i— i
H
c2
£H
J5
W
o
2
o
w
0
Si
w
*35
^H
•J
5B
H
O
S
-d-
I
Q
W
B>
C
O
0 4J
(0
O M
2 J-
•
o
NO
p^
•^^^
l-l
r-t
in
o
0
CM
vO
CM
•
o
00
CM
o
CO
o
o
o
00
C*J
o
o
1-^
in
o
•
o
vO
r-t
o
0
o
vO
0
00
s^
•
0
vO
p^
^^^
CM
1— 1
vO
VO
o
r-l
in
r-t
•
o
vO
r-t
0
CO
o
o
o
m
o
•
o
CM
vO
o
•
o
p-
r-l
o
•
o
^j-
CO
•
0
1— 1
CO
•
o
r*"*
P-*
•"•x,,
1— 1
D-5
-------
00
J3
60
Z
M
X
O
H
<
CO
W
H
fa
CO
O
<
S
H
Z
0
S
.
in
1
Q
S
(0
C
O
H ] trj
O 4J
ca
o n
Z -l
0 0
0 0
t-l CM
vo in
0 0
O 0
m m
i— 1 CM
o o
0 O
CM r-l
CO vD
0 0
CO CO
D-6
-------
CO
co C
e ,°
00 O 4-1
3- to in
Z 0 M
M Z <1>
CO
~~s o
Z
H r-
O 0
w
H
O
co
Q m
Zj 60 rH
O
CO
^^ r— 4
H 0
W T3 O
0
O
00
CO O
Z "H O
M 0
H
K -Ct-
H m
Z 3 O
o o
• o
O vO
rH
W CO
o s
2
> vO
D-7
-------
45
bO
fO
en
w
tu
g
V)
Q
CO
O
to
O
t-i
H
H
Z
_
§
O
w
o
u
O
I
Q
pa
CO
a
o
>4-l -r-l
O 4-1
CO
O M
Z 0)
CO
O
C?
CO
<=>*
oo
-o
•l-l
C
a
O)
,0
P-.
(•"*.
4-1
C
o
a
m
CM
CO
r-l
CO
CM
O
i— 1
>*
0
CM
O
O
o
vO
i— I
O
o
m
ON
O
0
O
CO
o
•
0
ON
vO
i— <
CO
1 —
•
l-l
vO
r^.
*•**,
'"*
m
ON
00
1— 1
ON
CM
0
ON
CO
O
CO
O
O
O
m
i— i
o
0
t— 1
ON
O
O
.
^•^
•^
vO
-------
m
_e
00
3-
Z
l_l
K-l
g
H
CO
w
5
s
•5
CO
o
CO
^
H
w
s
O
CO
a
r-l
ft!
H
z
0
o
w
o
3
w
J>
<;
2
H
Z
O
^
•
00
1
p
rJ
CO
C
O
U-l -rt
O 4->
CO
O r<
01
X)
o
o
CO
0?
00
•H
D
c
s
0)
f»
p_l
j:
4-J
C
o
s
in
m
CO
i— i
o
CM
o
o
CO
0
I— 1
0
o
o
CM
o
o
CM
CN
i— I
0
ON
CN
O
•
O
vO
ON
O
i— 1
oo
•
0
vO
p*.
•^^
r-
co
CO
ON
i— 1
Vfr
m
o
CN
CO
o
CN
O
O
O
r-l
0
o
ON
in
r-l
O
o
CO
o
•
o
r-l
0
1— 1
ON
ON
•
0
vO
r^.
•^^
00
m in
co o
r-~ co
r-l
m o
CO CM
0 O
00 r-l
CN CN
O 0
CN r-l
O O
O 0
O O
-* Is*
ff\ i —I
\* 1 f^
O 0
o o
m *
pis.
***^
.
CO
O
r-
CO
o
CM
O
0
O
i— 1
O
o
CN
O
r-l
0
i— 1
^J-
0
0
-^
o
1— 1
m
o
r-i
f-x.
fN^
*^^
•*
VO
r-l
vO
i— 1
r^
D-9
-------
en
,6
60
2
M
3
H
^
•
W
H
fe
W
Q
^3
C/l
^
H
§
O
W
z
o
M
H
(2
H
2
W
O
0
Q^
W
i
>
^J
^H
_l
SB
H
Z
O
X
T
o
W
«
W
C
O
0 4J
cfl
" r*
O t-i
2
-------
ro
B
00
3u
Z
M
orf
PQ
CO
H
co
w
H
fa
i-J
CO
Q
^
CO
H
H
J£
fa
O
CO
z
o
M
H
H
Z
fa1
CJ
Z
O
o
w
o
SI
w
0)
vo ^
D-ll
-------
CO
B
00
3.
Z
r-l
W
CO
H
-"
BS
H
Z
O
S
1
O
W
PQ
CO
C
O
U-l -H
O 4J
CO
• ^
O r-l
Z 0)
CO
.n
o
o"
w
^
00
*^3
O
»r-t
Z
3
O
C
s
CD
fa
i-Q
PI
4-1
C!
O
s
r-i
CM
m
1-1
i— i
^j.
•
!— 1
VO
OO
•
o
vO
^^,
00
m
CN
1-1
ON
r— 1
vO
0
^
CO
o
<;}.
O
o
•
o
o\
I-l
o
1
o
^J.
r— 1
t— 1
•
o
00
^3-
o
•
0
vD
CM
•
r"H
CM
i— 1
•
r-l
vO
•^^x.
ON
m
CN
«*
f^
f^
CO
o
1— 1
•*
0
^
0
o
o
CM
CO
o
•
o
^
1— 1
1— 1
•
o
I—I
s^-
o
o
o
CM
^^
vO
o
1— 1
r-.
o
T~H
CO
CM
rH
CO
1-1
oo
vO
o
^
r-
o
CO
o
o
•
o
^
CO
o
I
o
CO
00
o
•
o
ON
vO
o
•
0
CM
ON
1— t
CN
CO
•
1-1
r--
i—4
vO
CN
m
vO
CO
CO
o
D-12
-------
CO
e
60
3.
Z
^_l
O
0
H
CO
w
H
S
i-J
CO
§
^J
C/3
H
IS
pL|
O
CO
Z
O
M
H
H
Z
w
u
z
o
o
w
0
p2
w
5j
s
33
H
Z
O
^£2
CM
t-l
1
Q
W
3
CO
d
0
14-1 -H
O 4-1
CO
• r»
Z 0)
(0
,JO
O
J-
O
CO
°t
60
S
•o
o
• H
Z
3
U
d
s
0)
fa
.0
PU
.d
4-1
d
o
S
r--
CM
vO
•
Ov
CO
CO
O
CO
CM
•
O
CO
o
o
•
o
r--
i— i
0
•
o
^
t-l
t— 1
•
o
•^
CM
O
•
0
f^
f-
o
lf\
V£>
•
o
vO
t--
^^
f»
H
vO
CM
CM
O
in
CM
o
^
0
o
•
o
1-1
CM
O
•
O
1— 1
1— 1
1— 1
•
0
1— 1
CO
o
•
o
00
00
0
f*^
00
•
0
vO
r-
*»^
ON
r^
CM
r-.
•
vO
ON
CO
O
^
CM
•
O
^.
O
o
•
o
00
CM
o
•
o
oo
o
I—I
•
o
CO
CO
o
•
o
o\
00
o
o
•
o
^D
fx.
^x^
o
^_l
CO
CM
o
•
CM
1-1
ON
CO
O
o
m
•
o
CO
o
o
•
o
CO
CO
o
•
o
o
1— 1
1— 1
•
0
I— 1
CO
o
•
o
CM
O\
0
t-l
ON
•
0
vO
r-.
**^
j_i
CO
CM
CM
•
ON
,4-
CM
O
^
CO
o
o-
o
0
•
0
CM
«^f
o
•
0
CM
CO
1— 1
•
o
0
CO
o
•
o
in
1"-
o
00
ON
•
0
\o r*"» r** r**
r-- t** r-* r*-
***x. ***•» **-^ **"^
CM r-l CN CO
( )
CO
CO
CO
O
CO
o
o
00
CM
O
vO
O
CO
o
oo
o
•vl-
00
V
00
cfl
1-1
m
vo
D-13
-------
e
00
z
M
trt
fa
Q
CO
H
fa
O
§
H
2
Id
U
§
O
ta
o
w
s
I
Q
a
CO
a
o
MH -H
0 4J
CO
CO
ja
o
J"
o
2
y
-0
-H
Jjj
CD
fa
fN
PH
^C
4-1
C
O
s
m
o
ON
oo
CM
0
^
CM
O
CM
o
o
0
rH
CO
o
o
rH
o-
r-l
o
CO
VD
o
o
CO
CM
•
t— 1
^.
oo
•
CN
VO
r*.
*»^
r-.
m
CM
VO
i— i
^
CM
O
vO
CM
O
CM
O
o
o
00
I—I
o
o
CM
CO
1— 1
o
ON
*^-
o
o
00
0
i— >
^.
CO
•
•*
vo
1^
"^^
00
f^»
"-^
ON
"*
VO
00
r-l
^
CO
O
VO
CO
O
CO
O
o
o
oo
CM
O
o
o
1-1
o
0
0
o
oo
ON
•
0
CN
in
•
CM
vO
1^
~^
O
i— I
**
OO
vO
CN
CO
0
vO
CO
o
\o
o
0
o
m
CN
0
O
00
CM
r-l
O
o-
CO
0
O
CN
1-1
•
1— 1
rH
ON
•
r-l
vD
f-
^^
i— 1
rH
"*
CM
t-H
CM
r-l
m
o
CN
m
o
CM
O
o
o
ON
CM
O
O
CM
•st
0
o
oo
oo
o
o
VD
i — i
•
i— I
^
oo
•
CO
\D
p>*
*^^
CM
vD
VD
OO
i— 1
CO
CO
o
m
CO
0
,-j.
o
o
o
r-
rH
0
o
vD
«3-
O
O
m
CO
O
O
o-
f-»*
•
0
m
m
•
CO
p^.
p^
*^.
1— 1
~*
m
CM
r— 1
m
CO
0
r-.
co
0
i— i
o
o
o
o
CM
O
O
vo
CO
O
O
VD
OO
O
O
^
rH
•
i— 1
CO
^J-
•
(— 1
^.
f^.
^^
CM
m
st
vO
i— 1
i
1
rH
CO
o
^
o
o
o
ON
i— 1
0
o
o
o
o
vO
^j-
o
o
o
00
•
o
^J.
ON
•
CO
r—
r*^
•^»^
CO
in
i^
CO
i— i
^
VO
0
vO
s^
o
CO
o
o
0
CO
1— 1
o
o
CO
r-
0
o
m
vD
in
0
VD
CM
•
i— 1
i— 1
rH
•
"*
p^.
^.
•^^^
-*
^O
"*^
in
rH
VD
r^
o
CO
in
o
,_4
O
o
o
^
rH
O
O
o
0
rH
o
vD
r*^
rH
O
^j.
vD
•
rH
00
ON
•
-d-
r^.
r--
in
"*
0
o
rH
^
CN
O
^
CN
O
CN
O
O
0
^
1— 1
o
o
ON
m
o
0
vD
^j-
O
0
^
vD
0
CN
O
rH
r**
1 —
vO
oo
CO
rH
i— 1
"*
O
vD
CO
O
CO
O
O
O
CN
CN
O
O
m
ON
O
O
in
CO
rH
O
0
rH
•
i— 1
rH
rH
•
CO
0)
cxO
n)
0)
D-14
-------
APPENDIX E
SPECIAL INSTRUMENTATION
SPECIAL INSTRUMENTATION
A brief description of three of the less well known particulate monitors
follows.
EPA Dichotomous Sampler
The dichotomous sampler utilizes virtual impaction to separate the res-
pirable and nonrespirable fractions. Air is drawn into the inlet at a rela-
tively high flow rate to ensure collection of particles up to 20 ym aerodynamic
diameter. The inlet has a deflector rim on the bottom of a rain shield to
direct particles into the sampler during high wind conditions. At the bottom
of the inlet a portion of the total sample flow is isokinetically sampled at
a rate of 14 liters/min and transported to the two-stage virtual impactor.
A servosystem and pump maintain a flow rate of 13.6 liters/min to the fine
particle collector and 0.4 liters/min to the large particle collector. This
results in particles in the range of 3.5 to 20 ym aerodynamic diameter being
collected on one membrane filter and particles less than 3.5 ym aerodynamic
diameter being collected on another membrane filter. At the conclusion of the
sampling period, which is typically 24 hours, the filters are removed and the
concentration of the respirable and nonrespirable fractions are determined
gravimetrically. The membrane filters are then available for chemical analysis
free from the interferences inherent with glass fiber filters.
GCA Ambient Particulate Monitor (APM)
The APM also utilizes a high flow rate at the inlet of the instrument
and a configuration such that the conditions are similar to that of a hi-vol.
The APM uses beta attenuation sensing for mass determination and since the
sensitivity depends on small area collection, a flow rate of 9 liters/min is
isokinetically withdrawn via two probes. One probe has a cyclone which removes
the nonrespirable fraction (i.e., greater than 3.5 ym) and the other probe
collects the total particulate. The flow rates into the two probes are kept
constant either by means of a specially designed sonic venturi-type nozzle or
by an active feedback-type mass flow control system.
The aerosol to be measured is collected on a reinforced glass fiber fil-
ter tape and analyzed by beta radiation. In practice, an area of the tape is
tared by a beta source/detector pair, the tape is indexed by a stepping motor
to the collection port where the sample is collected, and after collection for
the selected sampling period, is indexed to a second beta source/detector
E-l
-------
pair where the transmitted beta radiation is again measured. The appropriate
calculations are performed by a microprocessor and the TSP concentration for
each fraction is printed on a tape. The sensitivity of the system allows col-
lection times much shorter than the conventional hi-vol without seriously
affecting the measurement error. During this study, sampling periods of about
1 hour were used.
NASA Air Scout
The Air Scout was developed by scientists at the Lewis Research Center of
NASA to permit collection of airborne particulate samples as a function of
wind direction. It includes a recording wind speed and direction system, and
the wind direction is also fed to the circuitry of the sampling device.
Eight individual electrically-actuated poppet valves, positioned in a
circular pattern in a housing above a hi-vol, control air flow through the
ports. Each port represents one of the eight directions of the compass. A
solenoid actuates each valve and is connected electrically to a set of con-
tacts in the wind vane corresponding to the direction assigned to the valve.
Thus each valve, when open, permits collecting a particulate sample in a dis-
crete location on a filter slide when it is in sampling position over the
housing. The sample thus collected represents particulates in the air coming
from a 45° arc of the compass. The central hole in the housing does not have
a control valve. It is open continuously so that the sample collected on the
filter slide can provide a measure of the TSP. The solenoids opening the
valves have sufficient force to overcome the suction of the hi-vol which runs
continuously during the sampling period.
A full discussion of the operational performance of the instrument during
the February field program is provided in the paper by William E. Belanger
which follows.
E-2
-------
Report on Operation of the
NASA AIR SCOUT
William E. Belanger, P.E.
U.S. Environmental Protection Agency
Region III
July 1977
E-3
-------
SECTION 1
INTRODUCTION
In February of 1976, the United States Environmental Protection Agency
(EPA) began a study of particulate matter in Philadelphia. This study is a
part of a national effort to discover the reason for nonattainment of the
national ambient air quality standard for particulate matter. It was the
suspicion of EPA that the violations of the standard were being caused, at
least in part, by dust generated by automotive traffic and by material re-
suspended by traffic. A major part of that investigation is an experiment
conducted on Broad Street in center city Philadelphia.
Since the only method for determination of compliance with the standard
in the high-volume air sampler (hi-vol), the experiment is designed around
use of the hi-vol for determination of the cause of the nonattainment. In
addition, several nonstandard methods were used in conjunction with the
hi-vols, mostly to gain maximum insight into the problem. These nonstandard
methods are the EPA dichotomous sampler, the GCA automatic particle monitor
and the NASA Air Scout. This report deals only with the operation of the Air
Scout.
The primary objective in design of the Air Scout is to achieve a direc-
tional resolution of 45 in a time-averaged particulate sample while maintain-
ing some degree of similarity to the hi-vol. The Air Scout achieves this by
drawing the sample air up under a large roof, then through nine filters. The
large central filter remains active throughout the sampling period while eight
smaller filters are controlled by poppet valves. The valves are controlled
by a wind-vane, thus each small filter samples air only when the wind is from
a particular direction. It is intended that this method would result in eight
direction-specific samples which may then be analyzed to reveal the direction
to the source of the particulate. Because of its similarity to the hi-vol and
because of the nearby heavy traffic on Broad Street, the Air Scout seemed a
logical means to investigate the contribution of Broad Street to measured
particulate levels.
Several features of the Air Scout were not utilized during this experiment.
An automatic filter change mechanism is provided which will allow up to 12
sampling periods to be completed without operator intervention. This feature
was not used because of the possibility that material might accumulate on in-
active filters, thus contaminating the results. In addition, there was a
possibility that rough handling by the filter change mechanism might cause a
loss of material from exposed filters (filters are dropped about a foot into
E-4
-------
a holding bin). Since EPA had no previous experience with the Air Scout, we
decided that manual operation with a 24-hour sampling period would minimize
possible sources of variance.
One other major area of difference was the method of filter analysis.
NASA utilized a central glass fiber filter which is weighed to determine total
mass in micrograms per cubic meter. The eight small controlled filters were,
however, a cellulose material and were analyzed only for trace elements (not
weighed). Since EPA desired to resolve particulate directionality on a weight
basis, this procedure was not followed. Instead, all filters were made from
Gelman "spectro-grade" glass fiber filters. All filters were then weighed
with the intent of eventually doing some sort of elemental analysis. This
procedure revealed several problems which may be generic to the Air Scout, and
others which may be due to the weighing procedure.
It should be noted that the precision attainable from a weight measurement
was investigated before this method was chosen. Three cassettes were loaded
with prepared and tared filters just as they would have been if the filters
were to be used for sampling. The filter cassettes were then covered and
stored for a day.
After this period of nonsampling, the filters were demounted from the
cassettes and then reweighed after desiccation. The result of this experiment
showed that most of the filters reweighed within 0.0001 gram of the original
weight with a few off by 0.0002 gram. Differences were observed with about
equal percentage of loss and gain, so there was no reason to believe that
there was any problem with a gravimetric method. (A typical sample weight
after 24-hour exposure would be about 0.001 to 0.003 gram, so the uncertainty
due to weighing errors would be less than ± 10 percent of the measured valve.)
After careful cleaning, the sampler was run on 5 sampling days for
24-hour periods concurrent with the operation of five other high-volume
samplers in the same location. Sampling occurred midnight-to-midnight on
2/10/77, 2/13/77, 2/16/77, 2/18/77 and 2/20/77. The sampling location is
given in Figure 1. After sampling, filters were removed from the Air Scout
the next day and covered to prevent accumulation of dust. Samples were then
returned to the laboratory for desiccation and weighing.
Several problems appeared in the interpretation of the results. These
problems are discussed in depth in the next section, and conclusions and re-
commendations are given in a final section. The main problem experienced was
a tendency for filters with only minimal air flow to exhibit relatively large
mass accumulation. No filter exposed in the field exhibited any weight loss
as did those tested for accuracy of the weighing procedure. The conclusion
from analysis of the data is that some material drawn under the sample roof by
the sampler motor or by ambient wind is too coarse to be directed to a par-
ticular small filter by air flow. Material, therefore, settles out randomly
and contaminates the small controlled filters during the sampling period.
E-5
-------
o
X
E
10
X
(9
X
E
Ok
X
u>
X
E
i
£
u>
«-
z
s
(£
a.
UI
1
j
o
13
UI
(9 X
Z (9
5 5
UJ
E
v 2
2
X
/
TJ
O
O
H
O
I
•H
a)
c
cd
a
o
E
O
CM
S/v-3
*^
**
j\
>$,
v
\~l 1
1 2 1
I .. Q X 1
1 — ^ O
^^ ? 1 .^fl
? to
1 6 1
1 u. «o cvi .
U
n
O
_
T
V
/
1 ° S " ' I/
^ L,
' ui If
L ° JJ
^ .
^7^
^ T^
^4.
1* 0-
<£.
%
L
^•••H
X
S
X
10
V
^^
p
V
•fc
ai
0)
•s
o
M
PQ
4J
O
CO
o
o
(U
M
S,
E-6
-------
SECTION 2
METHOD OF ANALYSIS FOR AIR SCOUT DATA
After weighing the exposed Air Scout filters, the result is a mass col-
lected on each of eight controlled filters (with wind-vane actuated solenoids)
and one uncontrolled central filter. Mass on the filters can then be converted
to a concentration in yg/m3 by dividing by the volume of air passing through
the filter. This volume is determined by the equation:
V = K F T
n n
where V = air volume through the filter
n
F = average total flow through the instrument
Tn = the time the filter was on
K = the percent of the total flow passing through the "on"
filter at any given time
Concentrations is then given by:
Cn = AMn/Vn
AM
where n is mass collected on the filter.
The factor, K, was measured by NASA as 0.23 and was calculated by mass ratio
to range between 0.18 and 0.31.
Preliminary calculation of concentrations yielded results which showed a
strong dependence of concentration on the percent "on" time of the associated
solenoid valve. This dependence is illustrated by Figure 2. This result is
not reasonable nor acceptable since dominant wind directions would then auto-
matically yield low concentrations. Statistical methods of removing this
trend were rejected since any such method, applied to all data points, would
remove directional information which is the object of the sampling. There
were insufficient data points to perform any manipulations based on individual
wind directions. It was, therefore, decided to seek a physical explanation of
the reason for the observed trend, and to correct for it in a manner which
would apply equally to all filters, and so would not favor any given wind
direction or remove the information sought.
E-7
-------
0)
CO
g
0)
t-i
o
co
O
4-1
0)
a
0
O.
o
c
o
g
O
c
o
o
tt-4
o
0)
o
c
0)
-a
g
P.
0)
p
0)
bO
•H
NOUVyiN30NOD JO 801
E-8
-------
Two possible causes were postulated for the observed trend, (1) perhaps
there was a small air flow through all filters not accounted for in the eyeball
interpretation of the wind direction chart record, or an error in the reading
technique, or (2) perhaps some mass was deposited on the filters which was not
controlled by the solenoid valves either through leakage through the valves
or through the inability of the small air flows involved to direct the large
mass particles to the filters.
The first combination of possibilities was investigated by calculating
the flow correction that would be necessary for each filter to achieve the
average concentration measured that day. While this obviously would remove
all directional information if done to each filter individually, it was hoped
that a trend would indicate the required correction. Figure 3 is a graph of
the required change in "on" time which would be added to the estimated "on"
time for each filter to achieve the average concentration that day. This chart
would need a gain of approximately 10 percent in order to equal the average
concentration, with filters with greater "on" time requiring a decrement equal
to about half the estimated "on" time for large times. While a small correc-
tion of this type would be acceptable, the large values indicated here are
untenable since solenoid leakage and random "on" periods during variable winds.
could not account for this magnitude of difference. Reading errors on the
strip chart would also not be this severe. Since random "on" periods were
observed on the chart, 1 percent has arbitrarily been added to all "on" times
to prevent division by zero.
The second possibility was investigated by plotting filter mass collected
against "on" time. This plot (Figure 4) shows a clear tendency of filter mass
to increase with solenoid "on" time, and also shows a definite positive in-
tercept at zero "on" time. This intercept was estimated to be about 0.0012
grams of mass. The average of all filter masses below 5 percent "on" time was
0.00114 grams, and this value was chosen as the correction factor. The results
of this subtraction of a constant mass is shown in Figure 5. It can be seen
that there is a large spread in calculated concentrations below 10 percent
"on" time due to division of mass by low flow volumes, but the trend above
10 percent "on" time has been effectively removed. It can also be seen that
calculated concentrations are now much lower since the measurement is only that
part of the mass which is due to operation of the solenoid valves. The rela-
tively large constant mass has been removed. This mass averages 76 percent of
the collected mass on the controlled filters, leaving only 24 percent explained
by air flow through the filters. While seemingly unreasonable at first glance,
the dynamics of large particles and the low flows through the controlled fil-
ters would seem to allow this result, since the "respirable" fraction of the
particulate is also in this order of magnitude in relation to total suspended
particulate. The sampler would then be expected to make its main differentia-
tion in this range, with less as particles get larger. It is also important
to note that this technique does not alter the directional information.
Final results of the 5 sampling days are given graphically by Figure 6.
Only readings where the solenoid was "on" more than 10 percent of the time
have been included because of the excessive variance in results at these low
"on" times. Each day at each direction has been charted individually to yield
a composite feel for concentrations from that direction. One sigma confidence
E-9
-------
§
CO
3
CO
cfl
t-l
u
ti
a>
o
o
co
rt
00
CU
cd
a*
QJ
(1)
4J
0
'aaainoaw ami MOIJ IVNOUIOOV
E-10
-------
•s
fi
o
o
o
I!
CO
CO
o
•H
J-J
4-1
CO
(U
00
OJ
CU
0)
6
c
o
fi
0>
o
^1
cu
p.
co
CO
CO
CO
cu
4-1
•H
•H
~4"
CU
a
•H
E-ll
-------
J
c
o
e
0)
G
l-l
M
s,
Noiivaj.N3DNOo Q3isnrav jo son
E-12
-------
limits derived from the standard deviation of the distribution of mass on fil-
ters below 5 percent "on" time are also indicated in Figure 5. These confi-
dence limits are variable due to the division of filter mass by volume of air
passing through the filter. It is expected that this subtracted mass is the
major source of variance in the method, hence no additional variance is assumed.
In fact, the only other known source of variation is the weighing technique
which is an order of magnitude smaller.
Qualitatively, the results show no significant (in a statistical sense)
difference between wind directions, but the data exhibits the highest "partial"
concentration with wind from the east, where Broad Street is located. This is
interesting in that the great bulk of stationary source emissions are to the
south and west of the sampler, and no elevated readings are shown from these
directions.
Since a large number of filters have been excluded due to excessive
variance (below 10 percent "on" time), it is desirable to further analyze the
data based on lumped filter masses. Since it is statistically incorrect to
average two concentrations when the flow through each has been different (i.e. ,
different sample sizes), the method employed is to add the masses collected
and the flow volumes through all filters in each direction. The analysis is
an analogy to operating single filters for 5 days without a filter change.
Since five filters are involved in each direction, five times the "static"
mass has been subtracted from the aggregate. This has a beneficial effect in
reducing the variance by \/5> since five samples from the population are now
involved. Results are shown in Figure 7.
It can be seen readily that a feature is apparent here which has been un-
revealed in the individual days due to the large number of points excluded.
An elevated trend in mass is evident in the directions Northeast through South-
east, with an isolated, and high variance, elevated point to the Northwest.
This trend is still not statistically significant, but is interesting.
Further lumping of filters using the directions North-to-Southeast
(traffic oriented) and South-to-Northwest (industrial) does reveal a signif-
icant difference (see Figure 8)• This lumping would be equivalent to
operating a hi-vol pair for 5 days, controlled by a wind-vane set for these
two pairs of quadrants. For the directions North-to-southeast, the partial
concentration is 35.6 ± 9.8 yg/m3. For the directions South-to-Northwest,
the partial concentration is 15.5 ±3.2 yg/m3. These are one-sigma confidence
limits, but even using two-sigma limits the results would be significant.
One must note here that these sigmas are derived from theoretical consider-
ations, and are not based on the variation between readings in the population
since these are subject to other complicating influences, mainly the variation
in pollution concentrations from day to day.
E-13
-------
1 1 1 1 1 1
1 • 1
1 '
rmtmm
\ m 1 *+-*-
1 • 1
1 • 1 i
' * •' \
1 • I * '
1 * ' • i
r • 1
I m
1 •
1 m 1
r • 1
1 • I H
' * 1 1 •
I • 1
1 • 1
1 • 1
1 m 1
1 • 1
L. 1 1 | | |
M
CO
cd
* 5
a
CO
H
Cfl
> -0
•H
>
,_J
•M
T3
8 5
" M
O
U-l
a
o
••J
*M
z ^
"1 «
B S
S -o
0 5
S 1 *
* 0
w c
o
4-J
cfl
4-1
0)
ill c
0
CJ
N
6. Partial
CO
limits.
if idence
M
O
o
4J
•H
1 § § § 2 9 S ° |
E-14
-------
1 1 1 1 1 1
1 ~ 1
1 • 1
1 • 1
1 • 1
1 1
L._. ,. m {
\ m ]
1 • i
« ^ i
L... A I
1 • | -
1 •
- . 1 i 1 1 1 i
1
»
S
1
u
£
2
o
i
Ul
jtf
X
X
' S 8 ? 8 8 2 o
>>
CU
-d
•H
co
co
S
01
i i
+J
CO
00
CU
00
4?
•
c
0
4-1
4J
CU
CJ
§
o
rH
n)
•rf
4J
l-l
CU
4->
•H
CO
0
g.
§
o
(0
13
1
cu
>
•H
CU
•H
Fn
•
Jj
O
«4-l
CU
4-1
ct)
00
0)
00
00
id
IVUMVd
E-15
-------
O
to
o
in
O
co
M
O
4-1
O
OJ
co
c
O
•H
4->
O
0)
l-l
•H
O
S
i
i
O
14-1
ti
O
•rl
4-1
rt
OJ
O
C
O
O
hi
M
td
'rl
4J
M
n)
PL.
CO
o
I
u
CO
0)
M
S,
•H
E-16
-------
SECTION 3
OBSERVATIONS AND RECOMMENDATIONS
Several problems arose in the operation of the Air Scout which make it
of marginal value in an investigation of particulate matter. These may be
classified as operational difficulty and problems may be summarized as diffi-
culty in preparing the large number of small filters for the cassettes, me-
chanical stability of the filters, particulate distributions on the filters,
and a fundamental problem in the rapid response of the wind direction control.
The Air Scout utilizes a system of cassettes where nine filters are ex-
posed simultaneously. In the case of this exercise, since sampling periods
were 24 hours, the Air Scout uses 9 filters per day or 45 filters over the
course of the experiment. These filters must be assembled by hand using slide
mounts, cut sections of filter and a press. In addition, each filter must be
individually desiccated and tared if a gravimetric method of analysis is to be
used. This manual process is quite time consuming and does not lend itself
well to the usual routine involved in air monitoring. It would be highly
desirable if a better means were available to mount the small filters in the
cassettes without the troublesome slide frames. The frames themselves are a
potential source of error if a gravimetric method of analysis is used since
their affinity for water is not known. In addition, the cassettes are cut to
very close tolerances resulting in some difficulty in installing the small
filters. There is a strong tendency for a finger to slip while installing the
small filters, thus ruining a tared filter. There was also a tendency for
fiberglass to flake off the outside of the slide frames, especially at the
rounded corner. This was a worrisome detail which did not seem to cause a
significant problem when "dummy" cassettes were prepared and analyzed.
When the first cassette was installed on the Air Scout according to the
instructions provided and the motor was started, the first major operational
problem developed. Close inspection revealed that the small filters, so
laboriously prepared, were no longer in their mounts. The vacuum had neatly
pulled them down into the space below the cassette, totally separating them
from their frames. Apparently, the cellulose filter material used by NASA is
somewhat stronger than glass fiber. A solution to this problem was to install
small screens in the cassettes below the filters. This effectively prevented
the filters from being pulled down by the motor vacuum, but failed to prevent
the filters from popping up due to the opening of the valves. It was therefore
necessary to install screen reinforcements top and bottom with an unknown effect
on sampling characterisitcs. Since any effect would be the same on all fil-
ters, it is not expected that this arrangement would interfere with the direc-
tionality of the device. It is abundantly clear, however, that there is a
E-17
-------
distinct upward motion of air upon opening of the poppet valves, and the effect
of any reverse air flow on the collected sample is not known.
Coupled with the problems with the filters is the problem shown by the
data. One could infer that 75 percent of the mass collected on the filters is
not related to air flow through the filters. This problem is not obvious from
examination of the filters, since the color of the low-flow filters would not
lead one to expect the large mass gain. Two explanations are possible, one
being a gain of water in the plastic frames used to hold the small filters.
This effect would, however, tend to affect total mass readings and the Air
Scout yielded lower total suspended particulate readings than a side-by-side
hi-vol. A more likely explanation is that the small air flow through the con-
trolled filters is not sufficient to direct the movement of large particles
once they enter the dead space under the roof. It must be recognized that the
hi-vol is also subject to this problem, but directionality is not attempted
using hi-vols. In addition, this class of instrument is subject to the entry
of large particles under the roof by action of wind independent of the motor,
as demonstrated by a great deal of recent investigation.
This theory is substantiated by the observation that the fine particulate
in the atmosphere is a dark color, while the coarse material tends to be of
light color. This is readily apparent when examining samples taken by the EPA
dichotomous sampler. This observation would tend to agree with the light color
of the low-flow filters and would support coarse particles as the source of
the problem.
A suggested solution to the above problems might be to eliminate the
poppet valves altogether and go to some form of air flow control which also
covers the top of the filters when they are not supposed to be sampling. One
possible solution might be a rotating cover which is cut to expose only one
filter at a time. This could perform the simultaneous functions of control of
air flow and covering of the inactive filters. Of course, a rubbing air seal
could lead to problems of contamination, but proper choice of materials; i.e.,
teflon, would minimize this problem due to abrasive action on the seals. The
assembly could be driven by a servo system from the wind vane.
A final problem is theoretical rather than experimental. The EPA Regional
Meteorologist volunteered this opinion upon examination of the system. The
problem is that turbulent flow in the atmosphere is random. Particles sus-
pended in the air do not travel in a straight line, but meander around with
local air currents. Thus the instantaneaous path of a particle has little
relation to its source. It is only when air flow is averaged that an indi-
cation of the flow direction is available, and it is this average that deter-
mines the transport of air from place to place. The instant response of the
solenoid valves would therefore tend to respond to local turbulence and
"assign" particles to directions other than their true origin. This effect
is compounded by the delay inherent in the large chamber under the roof. It
would be more theoretically correct to use some sort of time averaging system
to control the sampler rather than the instantaneous system provided.
E-18
-------
APPENDIX F
FEBRUARY TSP AND LEAD CONCENTRATIONS
F-l
-------
•j§
2
8
erf
Pi
9
Pd
Pu
j!ylj
&
erf
pq
H
Pn
0
5
Prf
O
rrt
nJ
^^
i
H
i-U
^-i
PQ
P
£5
S
™jj
fc§
S
M
r i
tS
E
u
O /-s
25 r/i
O W
U (J
PM
Q S
^^ "^
w co
§8
<«: w
PM s*
CO CM
H "-'
•
^•^
1
fe
PQ
H
uu
N*.
f 1
1^
CM
-^
CM
m
C i
CM
6
4.
#
1
!
<•
(
*i
^^
"B
00
u
u
c
o
s~*
a
oo
s —
•
U
c
0
o
r*1
*i^
00
.-i
.
c
o
(_>
B
00
a
•
c
0
u
"^5
3
C
0
o
u
J
0
^ J^
5 -rH
i..c ~"
X
c
c
o
u
CJ
o
XI
d.
A.
t/)
H
lO
Cl.
P-.
Vi
H
C-.
C/l
H
J3
PM
tn
H
JQ
P-i
e/3
I
i
«jc>ONr^ So^f^Sr^
W C_)
1 t t 1 t
r-< r— * t— 4 I— *
0 00 CD S 1 1 1 1 1
i—4 r-t i~t
U, O « tM
OO vD O CM f> iT-J -^
C) O •-< •~4 O O O
*
f--. r >. t^ CTN rn CM | ir\ if\
UJ (_J < t-H
"^
m o rj OJ , .
^ ^ ^ 1 1 ! 1 1
^
#
r^-- vO rn CT\ o
^ oo o cr\ 1 | | oo 1
t-^ t— <
UH Q CQ r-4
CM vo tn oo | | | | |
Ch O OJ — t Oi \O -41
V)
-—< ^ CM
t-< III
(U CM I-H CQ PQ PQ
M IM h r-4 | | crt {/i (/I
-H PM a. it
HHHH a^&u^iAtA
V
01
4J
w
V
fi
AJ
(0
4)
«
ft)
6
(Q
T3
0>
a>
•S
3
C
(0
O
• H
J
n
u
V) 4J
*— ( CO
o
> -o
1 eg
•* o
JS ^
en
rH
3 in
•o
•H O
•H " S
•a oi oi
C > 3
•H >H C
4J «H
01 CO u
-o c3
S n
i2
OQ (LI
c • .2
co u) a
>
U] CO 'H
4J M JC
Q" 8 „
ul -C 01
• o c
U O OO 'H
•H ^ B
*• iJ r-l
pa a a a.
u eg x
- o (2
-------
/•-N
•o
H
fC
B
00
3
t
u
B
0
.0
B4
0.
E^
4-1
«0
.as
X M^
0 •- B
a. .L:
a.
•f.
n
4J
to
CJ
O
J
O oo CM m
1 1 1 1 1
rj r-* rM CM
tf CO
— • CM cn O I I 1 ON |
CM CM CM CM F-4
,^-x
(JL, Q CQ CM
^^
2 ^ 2 n j j ^ > 2
* -i
£S^S ^2^Q^*2
s-*.
UJ U < r-t
^ "°' "^ 1 1 t 1 1 1
(N r-H ^-<
oo o o m
CM (-4 r-4 | | | | 1— t I
pt- Q CQ
^ ^ ^ ^
03
5 CM
^ III
* o o o o o
OOh OUOOO
H H H H ft- 0- In J< ^
,^
O
_f
£
U
01
CO 4J
fH C/3
O
1- 60
•* o e
• ^4 'rH
•00) B
B >
• rt •* O
o) a
no) e
5> n
MO) m
•O «
^> 1-1 .-* m
CM •-<
w^ 01 T3 ^-*
C W C
O a.
•-> U 0
i-l CU 00 O
»- E CM
•a .H r-i
co B — <
w a a o
o> * e u
f"O *J ttt
O 01 .1-1
E M V4 *—* ^>
PQ 4J • i-« CN
•O CO T5 1~-
B 0) CN 6
CD W 01 U O
X! <0 B
ta O 4-1 t~* o m
o a. CM
W CO b CM O
4-1 CU tO
- CO 14 VJ O B
CO CU 4J r-H M
- B B —" -H «-i
CJ O >~t t-l
•HO) UH B B
- w V 3 »
oo « a B a:
u U. B
- O —t O <0 1
< 1-4 (-1 OS
•O CN
M o. a) i I •-<
b O M 1
4> 4J Ot O O 1-1
0) O § r i r-t r t
iJ as B :< i i-i
-------
T3
0)
S
4J
C
0
0
1"*^
1
fa
3
^
2
^•4
^
^
—"
r 1
e-l
S
,_.
e
00
c
o
u
°J3
CO
3.
C
0
U
PL,
Pb
CO
£
Ci<
to
,p
00
_j
c:
o
U
•r
00
U
c
o
U
£
CO
H
PH
pt
CO
H
0)
a
•••* ,c
X 00 —
o ^ f ;
V* UJ •—
U.
0
U
o
- -
CO §
UJ O
r-4 r— *
S 0
r— < r—*
^ a
r) in
0 ^<
U 0
r-l O
0 00
rO O
^H i— *
U. Q
.-< a>
ai a>
3 3
0 0
r~
O — I O O
vO f- .0 >0 vO
in r^ in -^
OOO O
-H- un i»i
00 C^ S CO 0
! 1 1 1 I
1 1 1 1 1
r- £T> vo *O
000 O
"h
00 O CT* O^ OO
l~*
*E
^-« < eg
i 1 I
CN *-* po rt »
1 1 OO t/5 C/l
a, a.
So o o o
U O O O
O* O* lA u"i u"\
Ul
0
1
4J
QJ
(U
tO 4J
O
> T)
I
•H •,-(
4J
U 0}
*J t— '
0} (U
K
« V
a> w
•o «
>c4
/"N r- (
(S
vrf1 (U
"O (Q
g
eg o.
Q oJ
^-*
2 1
4) •
.0 -0 *J
a a 4;
3 o O
• H 41
« 4J K
CO « HI
U
• O — i
< •-"
t)
(o a. ai
l-i O P
V U 41
u o S
41 O 3
•4 o£ C3
to
4)
O
z
/-^
CM
^
s
CN
K
O
0)
o
c
01
-------
^
§
•iH
c
o
U
1
w
h4
<«
E"1
o
r 1
r i
2
CM
00
C^
(N
*
r~-
r-*
•~x.
CN
4
^— -
^E3
M
*~'
.
U
c:
o
^~s
r"i
"•^1
00
3.
U
c
o
o
^
c
ftO
I
c
o
o
m
a
oO
»
c
o
U
V
J
t
i *•*
3 ,c
D •tS"
•+ O> s-
a- ,c
;x
t;
R
, i
*-j
(Q
(J
O
-.,
P,
VI
n
^
fc
p-t
^
en
H
,0
U.
w
H
3
o o r-t r-* ^ > o O
fi c*^ -^ o o1* r^ f*-- o
1.1 o -t; at a) -H pu PL<
S33n OOpoo
OOOM uOooo
,_,
o
^
•H
x:
..
4.)
a>
a>
Ul J-l
rH t/1
o
> -o
1 a
•H O
X t-i
ta
(0
3 «
."2 o
^ •*-"
•H
T)
• H >H
*J r-f
flj (U •
C ^ CM
M I
• H T) U,
tfi 0)
a> w ai
T3 0) •-<
^V ^-H 10
IN H
^^ Q;
"O « W
a P-
o
r^ ID
— ft< O
^^ o
we o
M eg
4) • *J
.0 T3 W «
S CO 01
3 o 01 c
C IJ >- 3
to u 00
•O w 4)
c • *>
10 W O)
tn o •" e
o -H
•. m u r-*
u ta a,
u
U 01
- C C *-"
U 0 3
•r4 0> O
— 4J fcj ,C
aa « ra 1
u p
«,;" ^ ""* ^
•O
« Cu Ol 1
M O tJ
4) ^j c; i~~»
WOE -^
JS«§ § +~
W
0)
F-5
-------
sg
o
o
ffS
n
-3
w
Pn
>•<
28
5
pa
W
o
M
o
CO
o
1
M
a
^"*
PQ
O /-s
W S
U {H
^* PH
CO g
O pq
r i [-^*
2 r^.
r_j [
z
U 2;
a o
Z
O CO
O £4
< S
W <<
H^ vJ
O Prf
5 R
X U) —
O -r-i L
a. x: "^
a.
c
o
4J
n
o
vf O °O ^£>
r-< CM O — <
O tT^ <^~> O
^H ^-1 r-H (N
U, Q CO -— <
r^i m CN r4
0 ^ ^ ^
rH fv| CN i-H
Ul U < CN
C n -4 o
r-» o *-« -H
so oo >— • tn
C I fN r-»
[J-. f-J CO '-i
r-- r-j m — * r-J fN CN
U] U < CN
-31 \O O
i — i in c^i
^C^ i/^ O
i— * r~4 | C*1
I*- 0 *-i
r-t (J^ ^p <}•
•" '
1-
a>
M l_, H ,— 1
(U (U
O
o
o
CN
o
O
o
CN
O
O
o
00
"
o
o
CO
0
4-1
o
o
^7
00
u
a
0
S"
a,
8
^
CO
00
^— '
.
o
c:
3
me
'^^
00
^
0
o
o
XI
04
CO
a,
8
"S
&0
,
u
§
CJ
-o
ex.
a.
0)
B 4-1
•H X
X t>0^
o >H £j
a. .c ^^
<"
c
0
• H
4J
u
o
^> »o -H tr.
(NJ rH CS r-»
CSJ »-* ^^ ^H
b. O P3 rl
»-( 00 O IA
r-* O r-< f-*
{?% m i-t oo
(N O O O
t-H ' >-H t«4 r-l
W CJ «C r^J
^ 00 00
1
in in co
r-- CN en
- -• - i
^ (~) ff)
•-t CO o <; CN
>-^ O sO -41
*— <
u
«
O O O W
Hi H H H
F-6
-------
APPENDIX G
FEBRUARY TRAFFIC VOLUMES
Automatic traffic counters were installed at four locations in the vicin-
ity of the field experiment on 25 January 1977 and operated until 20 February
1977. One counter measured the total volume on Spruce Street adjacent to the
tower site, one measured the total volume on Broad Street south of Spruce Street,
and the two remaining counters measured the north- and southbound volumes on
Broad Street north of Spruce Street. Table G-l lists the daily volumes mea-
sured by the four counters during the period from 26 January to 20 February
1977. Of principal interest is the large volume of traffic carried by Broad
Street, and the substantial amount of traffic carried by Spruce Street. The
intersection of Broad and Spruce is signalized, and vehicles on Spruce Street,
which carries one-way traffic to the west, queue beside the AMS Laboratory
trailer and tower site. It is also of interest that the volume of traffic
during this sampling period remained high during the weekends. On the average,
at Spruce Street there was little if any decrease on Saturday and a slight
increase on Sunday. Traffic volume on Broad Street south of Spruce showed
little change on Saturday, but a decrease of about 20 percent on Sunday. Traf-
fic counts are available for only one weekend for Broad Street, north of Spruce.
Compared to the average weekday volume at this location, the Saturday volume
was about 5 percent lower, and the Sunday volume was about 11 percent lower.
Hour-by-hour detail at Spruce Street and at Broad Street north of Spruce
during the period from the 5th through the 14th of February when the GCA Am-
bient Particulate Monitor was in operation on the roof of the AMS trailer at
the tower site was presented in Section IV, Figure 27. The diurnal pattern of
traffic at both locations shows a pronounced minimum occurring roughly from
about 0300 to 0700 EST. This minimum is followed by a rapid rise in the volume
which then remains high for the remainder of the day. A prominent secondary
peak occurs at the end of the afternoon, but evening traffic continues at a
high level until after midnight at both locations. Figure G-l shows an hour-
by-hour breakdown of traffic by direction and day of the week at the Broad
Street, north of Spruce, location. During weekdays the split between north
and southbound traffic is nearly even throughout most of the day, but during
the late evening and early morning northbound traffic is the heavier. On
Saturdays, northbound traffic is considerably heavier than southbound from
0700 to 1100 EST, and again from 1400 to 1500 EST, but is much lighter from
about 2200 in the evening to 0500 in the morning. The striking feature of
Sunday traffic on the 13th of February is the north-south split during the
morning: northbound traffic is as great as it is during the remainder of the
week, but southbound traffic is greatly reduced, following the Saturday morn-
ing trend, but with lower volume.
G-l
-------
1
(A
PQ
W
I*
oo
i—i
I
m
CM
co
w
H
M
co
O
H
H
O
H
M
Z
M
O
fa
O
PQ
<:
H
Street
: Spruce,
•ections
T3 O -H
flj 'O
O J3
p5 3 4J
O O
CO J3
Street
! Spruce,
rections
^O O *H
cfl ""O
0 X!
J-l 4J J3
PQ M 4-1
2 2
C -O
Street
: Spruce,
ibound
T3 O 4J
O J3 O
M 4-1 CO
PQ M
§
Street
: Spruce,
ibound
*o o ^j
o .c o
££ c
o
c
Street
f Broad,
aound
u co
3 4J CU
t-i co U
cx oj
CO CU
cu
t-i
n)
Q
>* "4-i CU
CD O U
O rH
CM m
oo -a-
1 -1 rH | | |
CM CM
CO vO
-0- 0
vO f***
1 * M 1 I 1
" "
in in
CM CM
oo cj\
1 ^ 1 1 1
1 coco ' ' '
m r- r^ o
-* CO O rH
CO CM CTi in
CM CM rH CM
CM o co <)• m
"* vO CM ^^» CT\
r^ ^>» (-^* ^o r**
in vo r>- oo CTN o
CM CM CM CM CM CO
rH
CO
CO . )-i . . .
CU T3 3 -H 4J C
3 OJ f: l-l ™
CN CN
-------
1000
800
600
400
200
NORTHBOUND
OO 02 04 06 08 10 12 14 16 18 20 22 24
HOUR
(a)SATURDAY 12 FEBRUARY I9T7
5
>s
V)
UJ
-1
o
I
UJ
2
8OO
600
400
200
;\«- NORTHBOUND
00 02 04 06 08 10 12 14 16 18 20 22 24
HOUR
Ib) SUNDAY 13 FEBRUARY 1977
IOOO
800 •
600
40O - ^
200
NORTHBOUND
-A f
-SOUTHBOUND
"00 02 04 06 06 10 12 14 16 18 20 22 24
HOUR
-------
APPENDIX H
STREET FLUSHING PROCEDURE
H-l
-------
STREET FLUSHING PROCEDURE
Details of the street flushing operation are described in a memorandum by
Bruce Gledhill, Construction Engineer, City of Philadelphia, which is quoted
in part, below.
"The streets to be flushed were the following: Broad Street from Walnut
to South, Spruce Street from 13th to 15th, Locust Street from Broad to 13th.
Broad Street was to be kept continuously wetted. Locust and Spruce Streets
were to be flushed every 2 hours. Al Davis, Waste Collection Supervisor from
Area II, was assigned to supervise the flushing operations. During most of
the project four flushers were employed, two 2,000 gallon, two 1,000 gallon.
Headquarters was set up at the corner of Broad and Rodman Streets. Several
fire hydrants within the area were used to fill the flushers.
Flushing operations started at 7:00 a.m. from South Street. Flusher
proceeded northbound on Broad Street utilizing both front and side nozzles.
Upon reaching Walnut Street flushing was discontinued as the flusher pro-
ceeded around City Hall onto Broad Street southbound. Flushing resumed as
the vehicle crossed Walnut Street and continued to South Street. After this
the flusher would proceed to the nearest open fire hydrant for refilling. A
second flusher was dispatched as the first returned. One of the smaller flush-
ers was dispatched to wet the cross streets.
The initial operating procedure had to be amended shortly into the
project. Rush-hour and noontime traffic caused severe delays to the normal
round trip-time of 10 to 15 minutes. Flushers were dispatched every 15 min-
utes even if the prior unit had not returned. On some occasions traffic became
so heavy that all units were out at one time. A fifth flusher was employed
for a short period to alleviate this problem. It was soon observed that
flushing Locust and Spruce Streets once every 2 hours would be insufficient
because of the warm weather and heavy traffic. The operating procedure was
amended on the first morning to increase the flushing of cross streets to
once an hour.
After these initial adjustments the operation performed smoothly through
the first 2 days. On Friday, June 17, the Old Newsboy's Day Parade was
conducted on Chestnut Street. This event totally stopped vehicle traffic
within the area from 12:00 noon until 2:00 p.m. During this 2-hour period,
the flushers were dispatched sporadically and then only on Broad Street
from South to Spruce. The last flusher was dispatched at 6:00 p.m. every day.
Total water used is as follows:
First day 6/15 - 113,000 gallons
Second day 6/16 - 111,000 gallons
Third day 6/17 - 83,000 gallons
Project total 307,000 gallons
H-2
-------
APPENDIX I
JUNE TRAFFIC VOLUMES
1-1
-------
1
w
55
|"7*
*->
cn
rH
X
H=5
q
(A
H
W
52
^>
l—j
^
O
0
M
W
P-i
W
13
0
S5
M
§
P
S5
0
H
CJ
w
cn
(v!
W
H
H^
H
i
H
CO
W
CJ
:=>
Pi
PM
cn
H"
w
3
H
cn
Cj
o
PQ
W
pr{
H
H
•<
H
^
•
rH
1
M
W
i-J
9
H
Q)
Q
3
rH
O
**
rH
Cfl
4-1
O
H
x-x
CO
cu
CJ
•H
r*
CU
>4H
0
CO
to
o
0
o
^
0)
1
rH
O
^
rH
•H
cfl
T3
f — 1
CO
4J
O
H
^
C
O CO
•H CU
4-1 r-H
CJ CJ
CU iH
CO J2
M CU
CU >
rt '| )
•H 0
00 CO
C -
•H O
H O
CU O
4J rH
Pi
CU
4J TJ
CU CO
CU O
4-1 PQ
cn
Q) O
CJ
M CO
p, (U
cn >
4J TJ
O) CO
01 O
4-i PQ
cn
MH
eu o
o
3 4-1
M CO
O. CO
cn cu
cu
o
cu >•<
"> "O
CQ CO
T3 CU
CO C
CU TJ
3 CU
H S
***** "*•*»
QJ 0)
I 1
r- 00
iH vD CM in \O 00 O
in r-* in r^» rH CM
-K
K.O O^ *»3"
ON O rH
rH rH
*
CM T3 >i >~> Cfl CO
CO Ctf rJ Cfl Cfl TJ CU
r4 T3 3 13 13 CO C
3 -H 4J p! G 4) TJ
J3 M cfl 3 o 3 CU
H Pft en cn S H Is
"^•«. **"^ ^^. ^^ "^^. ^^* *^
cu cu cu cu cu cu cu
C C c C C P! C
3333333
ON o rH CM m -d~ in
rH rH H rH rH H
CM
f^
m
CM
CM
rH
1
vO
m
CM
-*
^.
CM
>N
cfl
13
CO
M
3
H
•*^..
CU
§
>-j
vO
rH
^f
oo
CM
CM
CM
rH
1
ON
VO
CM
*
in
CM
r*>
Cfl
•O
•H
^
pLj
*^Vh
cu
§
>-)
r^.
rH
CM
cn
en
*
m
o
rH
1
~*
en
CM
*
CM
CM'
eN>
!>•>
cfl
13
M
3
cfl
cn
••^
CU
1
oo
rH
CM
r^
CM
•K
VO
|-v
1
0
O
CM
*
CM
ON
rH
^
CO
13
rj
3
CO
-x^
CU
c
3
ON
H
o
CD
3
M
O
CU
4J
cfl
E
• H
4-1
CO
1-2
-------
se
o
05
fn
CO
1
SB
M
or:
O
a
i
^•fe
«
S3
o
H
CJ
W
CO
ffi
w
H
55
H
H
W
i§
H
CO
w
o
§
PH
co
A
H
W
a
CO
9
§
PQ
W
B
H W
§
co <->
py ^^^
S €N
go
> H
Cj [i]
M 25
PM ^
pfi r*v
H rH
CN
1
S
ft
§
CO
u
o
H
O}
•H
O
o
•H
CO
hi
H
J
a u
01 a
0 01
01 u
c
•H
U TJ
ai co
Ol O
M hi
u eo
cn
IM
Q) Q
O
3 U
hi a
en" 3!
4J T3
01 <0
01 O
U (0
cn
SM
0) O
o
24J
CO
a. «
cn oi
01
01 a.
01 cn
hi
i-l VM
cn o
•o X
CO V
O hi
hi Q
ca ,S
01
4J U
Ol 3
01 M
hi a
u cn
cn
•o o
*
0 JC
« 3
S
3
•H
JJ
CO
•a
o>
&
8
CO
o
0
CN
8
CN
O
O
>0
o
o
so
8
CN
O
o
g
w
CN
CO
•o
hi
lu
Ol
">
t*~
~*
8
m
1
iH
*-T
8
8
-*
8
3
o
S
8
s*
CO
•O
3
CO
cn
01
§
>->.
eo
rH
1 1
CN in
8 8
00 -»
,_,
8 8
oo •*
,_,
o o
o 5
rH CN
O O
O O
o o
8 8
§ rH
O O
§o
0
3 §
X >.
CO CO
TJ -O
hi hi
3 3
£ £
01 01
§ §
•-) *•>
oo oo
8 8
CN P»
sO SO
O O
o o
00 O
rH CN
§0
o
rH CN
§0
0
m oo
8 8
1 1
rH CN
0 0
§Q
0
CN SO
rH rH
X X
CO CO
•0 TJ
hi hi
3 3
cn cn
01 01
§ §
1 r.)
CO 00
r-l rH
i §
sO vO
8 8
m CN
CN CN
8 8
*n CN
CN CN
§o
0
PO S
o o
0 O
OS 00
8
•» o
CN O
o 3
0 °
8 u
o o
CN O
X O
<0 —
"Z a
0 1!
co 3
cn cn
H) u
§ §
<-> •->
00 ON
iH rH
8 8
CN cn
O 0
0 5
r- eo
o o
o S
r» co
8 8
o o
o S
sO CN
O O
O O
§ .H
0 O
o o
o S
3 S
CO CO
3 3
cn cn
I i
1 "5
Os ON
rH rH
8 8
•» rH
m so
§O
0
in os
rH rH
8 8
m os
rH rH
O O
go
^
8 8
00 O
8 8
rH CN
0 0
8 8
rH rH
CO CO
"a "§
§ §
CO C/l
i i
!-) >•>
Os OS
rH rH
8
in
cn
rH
O
o
CO
sr
8
eo
^
§
OS
O
o
m
co
°
§
o
o
g
ts
CM
X X
CO CO
•O ^
1 £
01 Ol
§ 3
"1 •-,
0
rH CN
a
1
o
"S
1
m
«"
u
1-3
-------
APPENDIX J
JUNE TSP AND LEAD MEASUREMENTS
J-l
-------
o
z
l-l
g
o
V)
f
1
M
SB
5H
Q /-v
W CO
3 w
» .J
->
•
1
W
iJ
5
H
t_<
«<
vO
O
•£>
0>
^
00
•a
r-.
""^
g
^
1
>
c
w
c
f
-t
_
(1
0
00
.1
u
c
0
jf^
(T)
00
1
^
J
1
I f^
( AO^
' -^ E
i-X ""
L,
g
•H
4J
«
U
O
JO
(X
PM
(/I
l^
^
fr«
Cu
en
H
_£, .
CL.
Ul
f
CL.
CL.
H*
CU
PM
H
s
3
fn \D
1 1 1 1 1
f-4 i-. .— 1
v£> >£> CN ^ OO
r- r- l~- 00 1 1 1 <*> 1
IK O « i-i
•-I in r-l vO vO ON in
— i ^ ^-4 o o o o
osvorqpi cntMinn^
UJ U < CN
^ CT* ^^ *o
«' X J J 1 1 1 1 1
o o *-• *-*
O 00 -3" -tf O
i—4
&4 a CQ ^
otnoo^
5 1
U *(
• - B
0 0 g
° V °
(S i— I
o 3
4-1 *H
s !
o j:
I c?
V. >^
3
•0 13
C
a in
o g
8 *
o
0 §
O J3
8 •
80
IM
>W ^
•H OB
Lw 0
0) r-l
a. A
9l
•H M
(— I
1 *
i
i
«"° +^
J-2
-------
at
j
c
4-1
<3
o
0
>t*^
•H
4,
W
S
S*
5
»n
^-.
a
*°
CO
(— 1
*o
CM
"-*
^^
rn
a
00
. 1
u
c
o
u
"a
^.
00
o
o
(J
*— ,
' B
00
^
o
c
o
CJ
^^
m
-•^
00
3
u
c
o
o
u
0
3 4->
< "M'
D -H
*4 0) ^
g
U
0*
(X.
l/l
H
IX
t-t
JD
OL,
0.
cn
X)
0-
a,
H
^
a
-/
o o _i _t
1 1 t 1 1
r^ Csl CM i « ooooo
O O l-c UUOOO
HHHH O,BL.iiSiA>n
•o
ftt
a
•H
B
O
O
J-3
-------
t)
3
.5
c
o
u
1
a
H
2
tO
00
rH
•o
v^
^>
tO
r-t
vO
,r
00
-1
o
c
3
^—- -,
,0
00
.
u
c
o
o
o.
"kO
1111 x « x ' «•
o ~ o o
1 I 1 1 oo r- r-» \o oo
cs
1 1 1 ^ 1 1 1 1 1
4.
i i i £ i i i " i
rsj
^^
O 00 i-l O O i-l CM
C*J •-* Csl CN l-l CM rH
* * *
"^ m vO «^ "^ r— 1 «^ ^ |
U. Q CO r-
**^
i—t CTt tO "^
r-4
Sg
i—4 3 CM
H III
OJ CM t-l 03 DO 03
H K V" i-l 1 1 M U> M
^ 5 ft fl o o o <™t Q
HHHH ft. a. S i?N iX
p»
tO
c
o
2
CM
o
xO
r-4
*^
tO
CN
0 «
O o r-
r-l f«l
* -f-
J-4
-------
o
z
M
pcS
t-3
o
co
Q
>
1
M
33
Q
b3 ^^
2 co
t3 M
CO »J
CO S
5* 3
co
z oe:
0 »
M O
H SB
OJ
1
^
w
H
CO
d
-0
00
vO
CO
*XJ
00
s
^
»-t
vO
0
o
»-*
0
0
o
CM
p-t
0
o
i-»
o
o
o
03
0
o
o
00
0
o
o
o
o
1
o
Q
4-1
O
O
o
o
1
CM
O
*"*
O
8
CM
s-^
"a
"M
.
u
B
o
O.
CO
PU
H
^-^
n
^
00
.
o
0
o
^^
00
. I
o
3
Ou
CO
X
H
o.
oq
IX
in
H
"B
"ob
•-^
tj
a
o
o
£•
CU
I/)
t-»
,-T
"a
00
3.
o
B
O
(J
(X
CO
CM
t/5
H
m
B u
.3 &
K 00-^^
o •H a
W QJ >v
O.
c
o
jj
a
X
^j
O GO
t-i CM ^-4 •-•
C^ CM CM tfs
^-* 00 r-« i-4
CM CM CM CM
W O < CM
rH O^ *O •rf
t^
V
t-i l-t M •-*
U ^) 0)
-------
^
o
Cv
C^
10
OO
i-4
*o
00
•0
0
0
o
AJ
o
S
o
I
0
o
o
§
0
o
o
0
§
cs
0
0
C-4
0
1-1
a
ft
m
0
00
.
o
c
o
U
»—
•<
g
4J
1
o CM & o
»M i— < o i— <
es m i— i m
O (N m -^
C*J IA •— ( r-*
/-.
UJ O < fN
CN r* m r CM
ft
-------
•a
o
00
o
0
4-1
o
o
o
CSI
o
o
o
(M
o
4-1
o
$
f— <
o
3
r-4
o
4-1
o
0
A sO 00 *9
r-4 r-t *$ t+)
In Q (O rH
s^
* • • *
CM CM CO m
r- o -* o
»-4
r^ oo a* o
r-i
lM Q PO .-i
ȣ> CO O ON
O O ^ O
n r^- r-i co
oo a\ en o
r-l rH n CN
w u < c^
"^^*
00 CO CO Ov
• • • •
o o o o
O vO «N O>
\O u~\ Wt \O
r-4 ^< fvj «-4
PL, O CQ i— <
*—( O> vO > *
O O O *
H H H H
J-7
-------
APPENDIX K
PARTICLE SIZE DISTRIBUTIONS FOR SELECTED
HI-VOL FILTERS
K-l
-------
J tO N CJ
o
1 n H n
l« b e
£
1 1 1
1 1 i 1 1 1 1 i 1 1 1
1 1 1 1 1 1 1 1
<0
IO
ro
CM
0>
00
(O
IO
IO
CM
V)
Ul
.
3 O> (0 O
J H ii n
0|K be
c
•O
10
CM E
— o
•
It
~" MJ
co
01
O4
<0
O.
O
C a
0) CO
to v
o
O r-
C '
o
O 3 CO
,£> 60
•rl C
00 UJ M -H
I 4-> CX
rr M
co
o ro
4_»
M O
tO O
PL, in
3)
IO
10 CJ
10
8
8
»
K-2
-------
2 *
-------
•»
w
M
4
X
a.
[
gON U
2 00 0) 00
Irdrd °
1 II H H
g 1* b c
JC
1
1
—
«•
_
[
«M
C
™*
_
^^^
1
L,
1
r
L
1 .
^~!:
__J.
!
1
«•
1 1 l i 1 i 1 1 i 1 j i 1 i 1 1 i g i i 1 1 i I 1 I i i 1 1 1 1 1 ^
0 K) CM CM — —
N
M*
5
[
M
w
|
a.
S^ « 2
o>o fd 2
J II H II
Of* be r
t,
**
C
j
1
[
r
!.
i
_j 4
r
L.
|"
^•— J ™
1 _
J "
u -
1
I
1
1
1
? 8 S 8 « 2 * c
«
^
<0
— co
cu
« 1
•— a.
cu
i cr
a.
.» o
•o 2
— 5 T3
* c
• 3
oi 6 cu
— 0 3
J: cr
.2 «
c ex
— • o
c
o" §
"" Ul C
20
_ -H
jt- ^f 4J
w — 3
0 I
•H
•0 UJ -u
i co
™* •!-! .
O -a r»!
K t cu --
K N 00
^t °rl *^^
J (0 CM
«0 „ .
Z H <
< -3 i
K) Ul 4-i cu
«0 O
PM H
rn
lO "^
(U
« 1
*^ »H
o
K-4
-------
Ul -
i
d»
a
w
2 O ro ~~
£ ' • ^^
O W — — „
1 H H II
X
•
-
-
-
[
_
[
""
—
-
[
r
i
I
•
'— i
L
r>OK>Oioowc
i ro CM CM — —
—
£-
(O
10
1
IO
CM
^""*
0
0)
00
K
.
10
<«•
fO
CM
f*l
0 *
3
a.
iu <3- _ O
* ^ ^^
O IO U) _
J n M M
0|M b C
[
L
C
c
c
i •
i
1
*y
^ ,
1
i
r
1
r1:
1
1 ~-
••
*1 -
—
Illllltlllllllllllllllllllllllllll
? S 8 8 12 2 « c
w
!<.
»
co
01
*> a)
"" ^3
Cu
* 0)
~" cr
CO
^ O-
w 2 o
~" * T>
• ^
CM E *
"~ O Q)
.5 cr
Ecd
_ P-
» o
£ §
"i «
5 o
A ^ "^
** •• jj
0 s
•H
O) IU i-j
"^ rn
o .2 .
K >- ^ r^
flC a)«^
< N 00
t -rl rH
.« W *^
W CM
Z iH
U ° ^
10 z £fe
cd S
e^i H
*
IO V^
0)
Nd
bO
•H
>
|U03J»d
K-5
-------
I 11II
I H H II
1- b c
I
ro
CM
0>
u>
10
t
10
CM
w
HI
I
CL
o
J M H M
O|M b c
Q
1 1 1 1 1 n
10 O tf> O 10 O
1C K) CM CM — —
|UtOJ*d
N E
— o
DE
l"
UJ
Z
o><
CO
cd
0)
3
cr
p-
o
rt
2 01
er
«
ex
o
0
§
(0
•o
ce
o.
10
CM
on
^
•H
00 r*^
•H r~-
T3 -^.
oo
0) rH
N »-».
•H CM
CO
-------
hi
o>
N _
1 M M ||
g 1- b c
s
f
I
I
r<
I
(0
M
to
n
K> fO CM M — —
3
O.
o\« b c
888852
[
J
CO
(0 CO
— rt
§
a4
0)
P.
o
M E id
— o
hi CU
I *
= 6.2.
0*1
Ul
U -H
_J »-l
O to
m
UJ U CN
•« -H i
rt S
&* H
0)
a
•H
K-7
-------
-t
822
M
T
X
O oi — m
J H « M
0|H b •
c
n o
S 8 8 8 2 2
111»i ii ii i ij»>
M O
ca
-------
[
C
Ii Illlllltllllllllllllllillllllll I
to
10
O
M
0)
0>
.
10 M N
H n it
O|K
CO
0)
co
.n rt
•O J3
&
cfl
cx
« °
"•'si
-I
= I
oT
HS"
*
Us
QJ
-------
O|M b
CO
a)
to
co
•s
0)
cr
03
a.
o
"S
n)
QJ
a-
co
ft
o
(3
o
c
o
•H
•H «
M r-
4-1 f-»
w ^.
•H ca
T3 rH
0) v£>
N
•rH «
W rH
<1J M
rH CO
O
•r( O
•u o
M m
CO
P4
ON
60
K-10
-------
w
III
a. _
w _
•% ^
Is *
5 s. -. o
o N «• —
1 H H II
§ IK b c
ae
*"
J
_
1
J
C
-
^
r—
\
•—I
I
r1
J-
1 '-
S-
••^i
j.
1
"-) :
noiooioomc
o 10 CM M - -
00
N
° j{
•
fll
W
1 -
X
^° 2 « H
•*•"***»
0 K» - M
2 H w M r
0|M be l
f
[
m
«•
•
i
•
r
\
I
I
r«
r
4
1
f1
L
r
r1
•M mi
' i -
> o » o » o ft c
> IO Al cv| _ —
•
<0 ^
— 0)
CD
ct)
10 "p.
0>
13
^ ^
Z «)
o.
o
» 2 "S
— Z C!
• «
• cu
N E 3
— 0 a-
O O<
•2 0
= E. §
C c
2" d
|4J O
2 I!
0»< 4^
0 £
«w «
o "^ !^
S ^" N <•
ac -H ^
^ (0 ~-—
ft- cu
U> <-i -
« U rH
5 -H i
** « si
* o
rH
1
10
cu
3
W g
>
K-ll
-------
w
1 -
I"- "
2 « w -
& —I —I ^"
X " *^* "••
1 " M ||
I 1- b e 1
-
1
J
-
-
[
-
-
•
-
j.
I
1 —
1
E
1
|
]
j
,, , , Ij.l.H,, ..III.! 1
•> o 10 O too me
K
CO
r-
0 s
to "
III .
3
0.
k. ** * Ml "
*" N Q»O»O«C
> IO CM eg — —
w
fw
co
— to
JC
a.
1 0)
3
cr
•o 2 |
• T)
M i §
•»> 0
.5 |
yj C
P»- c
Id
M» #«
*< §
O 4J
3
•3 3
O tn •
_ *^... 'H r*-
N ^ *b r-^
< O 00
«0 co ^o
•b ...
^tf
K-12
-------
E
M
(0
a. ui
w 2
u o
M
<\
«>
V
i i
3
UI
N
-------
o
~
x o "
M « <
Ul
N
55
o
o ?
o
s
o
* w
o *»
S *
M
to
<
I
u.
O
UJ
60
• H
§
t-l
CO
<1>
o
•H
ct)
CX
C
O
• H
stri
CO
CO
c«
CO
1—I
I
60
•H
O
M
O
N
OO>fl»f-»«
K>
CM
K-14
-------
i l l
i! 3
in
O
Ul
8
z
si
• !
8 °
" X
»:
S 5
ss
u.
o
Ul
»
u
oc
QJ
t-l
60
•H
§
I-l
CO
0)
o
• H
S-l
c«
A,
C
O
• H
4-1
3
42
•i-l
(-1
4-1
(0
• i-l
•a
CO
CO
cu
1-1
bo
K-15
-------
Ul
O
• 402
X K » "
ti t- < •
1 1 1 1
s
•I
* -
O O
10
O *
O *
s *
O „,
" i
a s
w
0
_ Ul
3
M
• H
CO
QJ
r-l
O
•H
4-J
U-l
O
c
O
• 1-1
CO
•H
CO
co
CD
M
•H
O • •
* "> N
'M313MVIQ
K-16
-------
-*••-
N T O i
N «?5
X M £ O
N
811 II
< •
S S
« e
O ^
:§
5*
o :
* X
O K
I *
Ul
Ul
u
It
O
M
« N 10 «
LPl
3
60
e
O
CO
0)
cfl
p.
O
• H
S-l
4-)
10
• i-l
T3
co
03
cd
00
• H
K-17
-------
-
w
3
J
tgfi
x J £ O
X K " "
_
•
^
:
< < •
4
"d *
^ •
^
^
-
^
<
<
»
^ *
• o
•
1 1 1 1 1 1 1 1 1 1
*
i
SI
I
s
o
0
*
O
2
8
o
o
10
o
-------
tal
-3
8
8
" Ul
N
8 »
o
«s
o o
«
_ M
8 3
o
M
X
w
o
- ff
00
•H
ta
I
l-i
m
CO
-------
Ul
o
_ so
s
Ul
o
« 5
s
o
«:
o o
aN<0A
O
(M
K-20
-------
hi
3
wo
N T O '
N «
-------
« •• — ' — — '
-
Ul
.sf
E5?5
lt*
ot
0>
o
at
o
o
o
*
o
A
O
o
ft
0
M
0
""
0
j O OfflWNtOlOVlO N —
l» N ~
Itl
M
O
tal
M
g
^
O
w
1C
o
X
X
H
V)
M
J
M
5}
2
^
O
UJ
tf
X
UJ
u
K
111
•
O
r-l
1
OJ
60
•H
|
)-<
U-l
CO
-------
-
Ul
Is I i
§s!|
t»o»
9
«
-
*
< "•
« ~
•
^ 0
V-
""*•.
A
4 ^ _
^
•
<
—
i i i i i i i i i i
•»
01
0>
0»
CD
0»
tf)
«
o
0»
o
00
o
N
O
o
in
o
0
10
o
N
O
•
^
O
3 O Ofl»«»-W«»M N —
ON -
W
3
o
i
M
O
<
O
w
•e
o
s
X
M
M
U
«)
3
31
o
V
^
^
u
U
K
bl
s-**
t-4
i-(
1
Q)
i-i
M)
• H
B
U-l
\«>
CO
(1)
1— 1
o
•H
4J
td
(X
14-4
o
c
o
J3
•i-(
l |
CO
13
CO
CO
cfl
S
•
CN
0)
3
60
•H
K-23
-------
s
£
I K K
(B LU UI
o < 3
o at at
» i- i-
N" N* N
r- N N
N 10
D • o
S
S
o
o
D
D o
D
D
o
M
•a
4
o
a
M
0)
3
O
O
M-l
CO
O
t-l
CO
0.
C
O
)-l
4-1
CO
CO
03
CNI
I
00000 O
OO>«DN<0tt
CM
bO
• H
K-24
-------
APPENDIX L
LEAD EMISSIONS FROM AUTOMOTIVE SOURCES
Characterizations of lead emissions from automotive sources generally
show that, under most operating conditions, only a percentage of the total
lead in the fuel burned is exhausted to the atmosphere; the fractions of total
lead that is not exhausted is retained in engine deposits, lubricating oil,
and the exhaust system.
These characterizations also show that the lead emission rate is func-
tionally dependent on the operating mode as well as the lead content of the
fuel burned. For example, limited testing has indicated that the lead emis-
sion rate during acceleration is much higher than during cruising. In fact,
several studies have shown that the rate that lead is emitted during acceler-
ation from 0 to 60 miles per hour under wide open throttle (WOT) conditions,
is about 10 times the lead-in-fuel consumption rate.1"3 This phenomenon is
thought to be a result of the purging effect of the exhaust gases on lead
deposits in the engine and exhaust system. It is noted, however, that emis-
sion data for a typical acceleration rate have apparently not been developed.
Analyses have been conducted of the particle size distribution of lead
emitted from automobile exhausts.5 On the basis of these distributions, total
lead emissions can be disaggregated into two components - airborne and
settleable. The fraction of particles smaller than 10 ym is generally con-
sidered to be airborne. Figure L-l shows the particle size distribution for
the airborne fraction of total lead emissions from an auto operating under
typical city driving conditions;5 note that the size distribution is func-
tionally related to accumulated vehicle mileage. It is estimated that curve A
on Figure L-l would most closely reflect an "average" condition with respect
to the accumulated mileage of in-use vehicles. Several analyses have been
made of the settleable fraction of lead emitted from motor vehicles. In one
study,6 lead concentrations in topsoil adjacent to a roadway section were
measured at various perpendicular distances from the roadway; the results are
shown here in Figure L-2.
*
This would reflect a very extreme condition where the acceleration rate would
be, for instance, about three to four times the average rate for vehicles
departing from a traffic signal.
L-l
-------
9.
8.
7.
d.
5.
4,
3. --
2. --
.2--
i r
Lead Particle Size Distribution
tmitted Fnn An Automobile Under
City-Type Driving, 1968
90
28,000 Accumulated Miles
21.000 Accumulated Miles
16,000 Accumulated Miles
5,000 Accumulated Miles
I Less Tnan Stated Particle Size
Figure L-l. Particle size versus percent less than stated particle
size by accumulated mileage (thousands of miles).
(Source: Reference 5).
L-2
-------
2 4 6 6 10
METERS FROM EDGE OF ROAD
Figure L-2. Lead concentration in topsoil at various distances
from the roadway. (Source: Reference 6).
Since gasoline is the source of lead emitted from motor vehicles, the
emission rate is directly proportional to the lead content of the fuel.
Currently, the amount of lead in an average gallon of gasoline is scheduled to
be limited to 0.8 grams per gallon in 1978 and 0.5 grams per gallon in sub-
sequent years. Under the originally promulgated schedule,7 the respective
maximum pooled average lead content per 1975, 1976, and 1977 was to be 1.7,
1.4, and 1.0 grams per gallon.^
It also follows that the emission rate will be directly proportional
to the quantity of gasoline burned. Numerous studies have been conducted
to estimate the "average" fuel consumption rate for the U.S. vehicle fleet
for the current and future years. Federal standards11 have been established
for post-1977 model year vehicles that specify the minimum fuel economy that
the vehicles within each model year must attain.
Fuel consumption is strongly affected by the type of driving mode.
Maximum fuel economy is generally attained at cruising speeds of 30 to 40
miles per hour. The least economical driving pattern consists of stop-and-go
driving in congested areas. The effect of various driving modes on gasoline
consumption has been fairly well defined.8"10 Techniques have been developed
for estimating average fuel consumption rates for various model year vehicles
for a broad range of travel speeds and driving patterns.^
LEAD EMISSION RATE
From the previous discussion it is obvious that lead emissions are re-
lated to vehicle speed and driving mode, vehicle model year, the rate of fuel
The weighted average lead content of leaded gasoline burned by noncatalyst
vehicles and unleaded gasoline burned by catalyst vehicles.
It is noted that even so-called unleaded gasoline contains about 0.05 grams
of lead per gallon.
L-3
-------
consumption, and the lead content in the fuel. Several studies have shown
that the rate of lead emitted by vehicles traveling on typical urban and sub-
urban roadways is about 28 to 45 percent of the total amount of lead in the
fuel used.1"^ For large areas where a wide range of driving modes is likely,
it has been shown that about 70 percent of the total lead in the burned fuel
is exhaust ed.1"1* From the information available, then, it is possible to
define an emission factor for lead, and to identify other factors which, in
total, can provide the basis for developing an estimate of ' the lead emission
rate for an area or a particular roadway section. This rate can be expressed
as:
F C
Pb Pb ..,
where Q = the lead emission rate, in grams per vehicle-mile (s)
F , = the lead emission factor for the area or type of facility,
expressed as the ratio of the quantity of lead exhausted
to the quantity of lead in the burned fuel,
Cp, = pooled lead content of the gasoline, in grams per gallon,
R = the rate of gasoline consumption for the area or facility,
expressed as vehicle-mile (s) per gallon.
The quantity of lead emitted (Pb ) for an area or particular facility
would be:
Pbe - Qpb (VMT) (L-2)
where VMT = the vehicle-miles traveled over the facility or within the area of
concern.
L-4
-------
REFERENCES
1. Hirschler, D. A., L. G. Gilbert, F. W. Lamb, and L. M. Niebylski.
Particulate Lead Compounds in Automobile Exhaust Gas. Industr. Eng.
Chem. 49. 1957.
2. Hirschler, D. A., and L. Gilbert. Nature of Lead in Automobile Exhaust
Gas. Archives of Environmental Health, Volume 8. February 1964.
3. Ter Haar, G. L., D. L. Lenane, J. N. Hu, and M. Brandt. Composition,
Size, and Control of Automotive Exhaust Particulates. Journal of Air
Pollution Control Association. Volume 22, Number 1. January 1972.
4. Wilson, James H. Unpublished Paper. U.S. Environmental Protection
Agency, Office of Air Quality Planning and Standards. Research
Triangle Park, North Carolina. 1977.
5. Habibi, Kamran. Characterization of Particulate Matter in Vehicle
Exhaust. Environmental Science and Technology, Volume 7, Number 3.
March 1973.
6. Pierson, W. R., and W. Brachaczek. Particulate Matter Associated With
Vehicles on the Road. Society of Automotive Engineers, Paper
Number 760039. 1976.
7. Federal Register. Control of Lead Additives in Gasoline. Volume 41,
Number 189. U.S. Environmental Protection Agency. September 28, 1976.
8. Austin, T. C. et al. Passenger Car Fuel Economy Trends Through 1976.
Automobile Engineering Meeting. Detroit, Michigan. October 1975.
9. U.S. Environmental Protection Agency. A Report on Automotive Fuel
Economy. Washington, D.C. February 1974.
10. U.S. Environmental Protection Agency. Factors Affecting Automotive
Fuel Economy. Office of Air and Waste Management. Washington, D.C.
May 1976.
11. Act of December 22, 1975. 89 Sta. 15 USC 2002.
L-5
-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completing)
1 RtPORT NO
903/9-78-003
3. RECIPIENT'S ACCESSION-NO.
4. TITLE AND SUBTITLE
5. REPORT DATE
Philadelphia Particulate Study
Final Report
6. PERFORMING ORGANIZATION CODE
April, 197ft (apprnvpd)
7 AUTHOR(S)
Robert M. Bradway
Frank A. Record
8. PERFORMING ORGANIZATION REPORT NO.
GCA-TR-78-02G
9. PERFORMING ORGANIZATION NAME AND ADDRESS
GCA/Technology Division
Burlington Road
Bedford, Massachusetts 01730
10. PROGRAM ELEMENT NO.
11. CONTRACT/GRANT NO.
68-02-2345
12. SPONSORING AGENCY NAME AND ADDRESS
U.S. Environmental Protection Agency
Region III
6th & Walnut Streets
Philadelphia, PA 19106
13. TYPE OF REPORT AND PERIOD COVERED
Final
14. SPONSORING AGENCY CODE
15. SUPPLEMENTARY NOTES
16. ABSTRACT
This report presents the results of a study to increase understanding of the
principal sources of particulates within the City of Philadelphia, their dispersion,
and the feasibility of their control. The approach was twofold. One approach in-
volved field experiments to measure the influence of specific sources through the
deployment of hi-vols and other sampling equipment. Most of this effort was used
to investigate the contribution of traffic-related emissions; some measurements
were also made in the vicinity of major construction (earth-moving) activity. The
second approach involved the use of diffusion modeling techniques to calculate
the contributions of major categories of particulate emissions to particulate loadings
at the various monitoring sites. The principal modeling tool used in calculating the
contribution of source categories throughout the city was the Air Quality Display
Model (AQDM). A test was also conducted to measure the effectiveness of street
washing in reducing ambient particulate levels.
17.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
Air Pollution
Particulate
Urban
Traffic
Street-Canyon
Lead
b.IDENTIFIERS/OPEN ENDED TERMS C. COS AT I Field/Group
Hi-Vol
Dichotomons Sampler
APM (GCA Corp.)
DISTRIBUTION STATEMENT
Unlimited
19. SECURITY CLASS (ThisReport)
None
21. NO. OF PAGES
288
20. SECURITY CLASS (Thispage)
None
22. PRICE
EPA Form 2220-1 (9-73)
L-6
------- |