903R81005
U.S. EPA Region in
Regional Center for Environmental
Information
1650 Arch Street (3PM52) TR-81-104-01
Philadelphia, PA 19103 January 1981
ASSESSMENT AND UPDATING OF PARTICULATE EMISSIONS DATA
FOR THE SOUTHWEST PENNSYLVANIA INTRASTATE AIR
QUALITY CONTROL REGION
VOLUME I
H. E. Cramer, H. V. Geary and S. F. Saterlie
Prepared for
U. S. ENVIRONMENTAL PROTECTION AGENCY
Region III
Philadelphia, Pennsylvania 19106
Contract No. 68-02-1387
Task Order No. 3
H. E. Cramer company, inc.
UNIVERSITY OF UTAH RESEARCH PARK
POST OFFICE BOX 8049
SALT LAKE CITY, UTAH 84108
-------
903R81005
U.S. EPA Region IH
Regional Center for Environmental
Information
1650 Arch Street (3PM52) TR-81 -104 -01
Philadelphia, PA 19103 January 1981
ASSESSMENT AND UPDATING OF PARTICULATE EMISSIONS DATA
FOR THE SOUTHWEST PENNSYLVANIA INTRASTATE AIR
QUALITY CONTROL REGION
VOLUME I
H. E. Cramer, H. V. Geary and S. F. Saterlie
Prepared for
Regional ( enter for HmimnmenMl Informs ruin
U. S. ENVIRONMENTAL PROTECTION AGENCY rsEPA Reg,™,™
p . TTT 1650 Arch Sl.
Keg ion 111 Philadelphia. PA 19103
Philadelphia, Pennsylvania 19106
Contract No. 68-02-1387
Task Order No. 3
H. G. Cramer company, inc.
UNIVERSITY OF UTAH RESEARCH PARK
POST OFFICE BOX 8049
SALT LAKE CITY, UTAH 84108
-------
ACKNOWLEDGEMENTS
We are indebted to Dr. Peter Finkelstein, EPA Project Officer
and EPA Region III Meteorologist for his technical guidance and very
helpful assistance during all the Phase I activities. The basic par-
ticulate emissions data and the air quality data contained in the report
were provided by the Commonwealth of Pennsylvania Department of Environ-
mental Resources (DER) and the Allegheny County Bureau of Air Pollution
Control. Mr. Gary Triplett, DER Bureau of Air Quality and Noise Control
provided us with emissions data on file in the DER Harrisburg offices
and arranged for our staff members to visit the DER regional offices in
Meadville and Pittsburgh. Mr. Richard Zinn and Mr. Richard H. Baldwin
in the DER Meadville office were most coorperative in providing us with
pertinent emission and process data for particulate sources in their
region. Similarly, Mr. Nicholas Pazuchanics, Mr. Larry Wonders and Mr.
Ken Bowman in the DER Pittsburgh office were of invaluable assistance in
providing us with detailed information on particulate sources and pro-
cesses in their region.
We are also greatly indebted to Mr. Ronald J. Chleboski, Deputy Di-
rector of the Allegheny Bureau of Air Pollution Control, and the following
members of his staff for their cooperation and assistance in compiling
emissions and process data for particulate sources located in Allegheny
County: Mr. Bernard Bloom, Dr. Roger Westman, Dr. Albert Smith, Dr.
Arvid Ek, Mr. Arthur Bulger, Mr. Malcom Lebowitz, Mr. Simon Feigenbaum,
Mr. Robert L. Felt, Mr. Harry J. Elder, Mr. Milton Goldberg, Mr. Carl
Nim, Mr. Donald Graham and Mr. Joseph J. Chirico.
-------
EXECUTIVE SUMMARY
INTRODUCTION
This report describes a two-phased study to update and improve
the particulate emissions inventory for the Southwest Pennsylvania
Intrastate Air Quality Control Region (AQCR). During 1973, the National
Ambient Air Quality Standards (NAAQS) for particulates were violated at
a number of high-volume (hi-vol) monitor locations within the AQCR. The
primary purpose of the work discussed in this report was to develop an
accurate emissions inventory that could be used with appropriate diffusion
modeling techniques to evaluate various emission control strategies for
the attainment and maintenance of the NAAQS. The first phase of the
study consisted of a review and evaluation of the existing (1973) particulate
emissions inventory and a determination of how best to improve the
inventory. The second phase consisted of an updating of the emissions
inventory, the collection and detailed analyses of hi-vol filter samples
in an attempt to determine the sources of the particulates causing
violations of the NAAQS, and the development of particulate diffusion
modeling techniques.
SUMMARY OF PHASE I ACTIVITIES
Phase I began with a compilation of the 1973 particulate emissions
data available for the Southwest Pennsylvania AQCR from the Pennsylvania
Department of Environmental Resources (DER) and the Allegheny County
Bureau of Air Pollution Control (BAPC). A critical and comprehensive
review of the resulting emissions data was then conducted with emphasis
placed on an evaluation of the accuracy and validity of the existing
emissions data. Special attention was given to those sources for which
the particulate emissions had been estimated using emission factors. In
some cases, incorrect emissions factors and/or mathematical errors were
detected and corrected. Appropriate adjustments were also made in the
ii
-------
inventory to reflect recent changes in source configurations and/or types
of control equipment.
Cramer, et al. (1975) had previously developed and applied long-term
and short-term diffusion models to the major sulfur dioxide (S0?) sources
located in and adjacent to Allegheny County, Pennsylvania. Because the
application of these models yielded close correspondence between con-
current calculated and observed annual average and short-term ground-
level SO,., concentrations, we used the updated 1973 particulate emissions
inventory and meteorological data from the Greater Pittsburgh and
Allegheny County Airports with the Cramer, ert al. long-term model to calcu-
late 1973 annual average ground-level particulate concentrations within the
Southwest Pennsylvania AQCR. The annual average concentrations calculated
at hi-vol monitor sites other than those directly impacted by emissions
from large industrial sources accounted for only about 10 percent of
the observed concentrations, even after the effects of emissions from
classical area sources (space heating, vehicle exhaust and tire wear)
were considered. We therefore concluded that the emissions from the
sources contained in the industrial emissions inventory were most likely
not responsible for the major mass fractions of the observed particulate
concentrations at the non-industrial monitor sites. Also, we determined
that the primary objective of the Phase II work should be the identifica-
tion and, if possible, the quantification of the sources responsible for
the major mass fractions of the observed particulate concentrations.
SUMMARY OF PHASE II ACTIVITIES
Hi-Vol Filter Analysis Program: Data Collection
and Filter Analysis Techniques
Approximately 90 particulate samples were collected during
August and September 1976 at the 15 monitor sites shown in Figure I. At
each site, a standard hi-vol sampler was located within about 1 meter of a
iii
-------
BUTLER CO
ARMSTRONG CO.
INDIANA CO
BEAVER CO.
ALLEGHENY CO. ^
©15
WESTMORELAND CO.
WASHINGTON CO.
FAYETTE CO.
GREENE CO.
5 0 5 10 15 20 KILOMETERS
LOCATIONS OF PART1CULATE MONITOR SITES
FIGURE I. Map showing the counties included in the Southwest Pennsylvania
Intrastate Air Quality Control Region and the locations of
monitor sites at which particulate filter samples were collected.
IV
-------
BAPC or DER hi-vol sampler, and both samplers were operated concurrently
for 24 hours (midnight to midnight) on regularly-scheduled sampling days.
The 15 sites, which are identified in Table I, were selected on the basis
of logistical factors as well as the requirement for obtaining filter
samples from industrial, urban and suburban areas. (The proximity of
each site to major industrial sources and the general characteristics of
the area surrounding the site xvere used in assigning the site classifica-
tion.) Spectral grade (Geltnan No. 64948) glass hi-vol filters were used
for all of the particulate samples. The filters were preconditioned,
weighed, sequentially numbered and packaged in plastic in sets of
unfolded pairs by the Coors Spectro-Chemical Laboratory in Golden,
Colorado. Prior to the start of a sample day at each site, one filter
from a paired set was placed in the hi-vol sampler and the other was
retained in the plastic container. Immediately following the sample
day, each exposed filter was removed from the hi-vol sampler, the
unexposed filter from the plastic container was placed on top of the
exposed filter to prevent contamination or loss of particulates, and the
filter pair was sealed in a plastic container. The sealed plastic con-
tainers were collected and returned by air express to the Coors laboratory.
On arrival at Coors, the filter pairs were logged, removed from the
plastic containers, conditioned and weighed. The total weight of the
particulates on each filter pair was obtained by subtracting the
previously-determined weight of the unexposed filter pair.
Hi-vol filter samples were collected on six regularly-scheduled
sample days during August and September 1976. Detailed analyses were
made of the filter samples collected on only three of these sample days
using optical microscopy, scanning electron microscopy (SEM) and energy
dispersive X-ray analysis (EDAX) techniques. Meteorological conditions
on the other three sample days precluded determination of source-receptor
relationships because of the presence of light and variable winds with
many hours of reported calms. The Coors Laboratory prepared the filter
samples for all types of analyses and performed the optical microscopy
v
-------
TABLE I
LIST OF HI-VOL MONITORING SITES
Site
Number
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
Name
Baden
Beaver Falls
Koppel
Brighton Township
Midland
Elco
Downtown
Central Lab
Hazelwood
North Braddock
Duquesne II
Liberty Boro
C lair ton
Greater Pittsburgh Airport
South Fayette
Agency*
DER
DER
DER
DER
DER
DER
BAPC
BAPC
BAPC
BAPC
BAPC
BAPC
BAPC
BAPC
BAPC
Site
Class i£ i cat ion
Industrial
Suburban
Industrial
Suburban
Industrial
Rural
Urban
Urban
Industrial
Industrial
Industrial
Industrial
Indust rial
Rural
Rural
*DER refers to the Pennsylvania Department of Environmental Resour-
ces and BAPC refers to the Allegheny County Bureau of Air Pollution
Control
vi
-------
analyses. The SEM and EDAX analyses of the prepared samplers were per-
formed by the University of Utah Research Institute. The first step in
the sample preparation was the removal of five circular sections, each
of which was 3.8 centimeters in diameter, along a diagonal of each
filter packet (exposed filter and cover filter). A small circular section
measuring 0.64 centimeters in diameter was then removed from the center
of each 3.8-centimeter section. The exposed-filter and cover-filter
portions of the small sections were carefully separated, mounted on SEM
stubs and shipped to the University of Utah Research Institute for SEM
and EDAX analyses. The larger donut-shaped filter sections were used by
the Coors Laboratory for optical microscopy analyses.
The ultimate objective of the optical microscopy and SEM
analyses was to determine the mass distribution by particle size and
generic type of particle on each of the hi-vol filter samples. The EDAX
analyses provided information on the elemental (chemical) compositions of
selected individual particles and of groups of particles. In preparation
for the optical microscopy analyses, samples were cut from each filter
section and placed on glass cover slides. The samples were immersed in
oil with a refractive index of 1.52 so that the glass fibers of the hi-vol
filters would be invisible. The prepared samples were examined under the
microscope using transmitted and reflected light with various polarizations,
The particles within a reference area of fixed size were counted and
classified by size and generic type of particle. Classifications of the
particles by generic type were based on the experience and judgment of
the microscopists and standard particle identification techniques. The
size categories and generic categories used to classify the particles
are defined in Table II. The volume of all particles in a given size
category for a specified generic particle type was obtained from the
appropriate particle count and the average particle diameter for the
size category, assuming the particles to be spherical. By summing these
volumes over all size categories, we obtained the total volume of
particulates in each generic category. The mass distributions for each
VII
-------
TABLE II
PARAMETER VALUES USED FOR ANALYSES
OF FILTER SAMPLES
Particle
Size
Categories
On)
Particle
Diameter*
(imi)
Generic
Ca tegor i es
Particle
Den:- i L y
(fi/cm3)
Standard
Reference
Area
?
(c:O
(a) Optical Microscopy Analyses
> 5 < 10
> 10 < 20
> 20 < 50
> 50 < 75
> 75
7.77
15.54
37.02
63.32
88.09
Flyash
Nonmagnetic Iron ux i do
Magnetic Iron Oxide
Quartz
Calcium Carbonate
] .5
4.9
4.9
2.5
2.7
0.00968
(b) SEM Analyses
_> 5 < 10
> 10 < 20
_> 20 < 50
> 50 < 75
1 75
7.77
15.54
37.02
63.32
88.09
Spherical
Irregular
Agglomerate
2.4
2.7
1.5
0.02053
*Ceomctric mean.
vixi
-------
generic category were calculated by multiplying the above particle volume
estimates by an average particle density assigned to each generic
category (see Table II). The total mass of particles on each hi-vol
filter was calculated by multiplying the mass in all generic categories
from the optical microscopy data by the ratio of the total surface area
of a hi-vol filter (423 square centimeters) to the standard reference
area (see Table II) used to obtain the particle counts.
In the SEM analyses, photomicrographs were made of various
areas on the stub filter samples prepared by the Coors Laboratory.
Particles on the photomocrographs were counted and classified by size
and generic cateogry. The particle counts were adjusted to a standard
reference area. As shown in Table II, the SEM partible size categories
were the same as those used in the optical microscopy analyses. However,
the generic categories and the particle densities assigned to these
categories were different. Estimates of the particle volume and mass by
size category and generic category were obtained in the same manner as in
the optical microscopy analyses. In addition, the EDAX analyses provided
information on the elemental compositions of selected individual particles
as well as the average elemental compositions of all of the particles in
the standard reference areas. The individual particles thus analyzed were
identified on the photomicrographs, and the EDAX data on the elemental
compositions of the particles were used with the physical characteristics
revealed by the SEM data to determine the generic classifications.
Various checks were made to test the reliability of the particle
counts, sizing and generic typing as well as the representativeness of the
procedures used to estimate the mass loading of the filters. Samples of
six different hi-vol filters were sent to Walter C. McCrone Associates
for optical microscopy analyses. The particle counts, size distributions
and generic classifications obtained from the independent analyses by
McCrone Associates were in good agreement with the results obtained by
the Coors Laboratory in their analyses of the same filters. We also
-------
investigated the spatial variation of the particle loading on the hi-vol
filters, using a chi-square test described by Conover (1971), which showed
that the hypothesis of a uniform distribution of particulates on the
filters could not be rejected at the 99.5-percent confidence level. In
addition, we performed linear regression analyses of the measured
weights of the particulates on individual hi-vol filters versus the
weights estimated from the optical microscopy and SEM data. The
regression analyses showed that the estimated filter weights based on
the optical microscopy and SEM data for exposed filter samples (without
the cover filter) were, on the average, less than the corresponding
measured filter weights, while the estimated filter weights based on the
analyses of both the exposed and cover filter samples were larger than
the corresponding measured weights.
Hi-Vol Filter Analysis Program: Results
The most important results of the hi-vol filter analyses were
the unexpected presence of many large particles on all of the filters
and the discovery that these large particles were responsible for the major
fraction of the total mass of particulates collected on the filters.
Figure II shows composite cumulative mass distributions developed from
the optical microscopy data and the SEM data for all filter samples.
The composite mass median particle diameter from the optical microscopy
analyses is 68 micrometers if biologicals are included (see the solid
curve in Figure II) and 57 micrometers if biologicals are excluded. The
composite mass median particle diameter obtained from the SEM analyses
is 27 micrometers. With respect to the site classifications, the
optical microscopy data show mass median particle diameters slightly
larger than the composite value at the industrial, urban and suburban
sites. At the rural sites, the optical mass median diameter is about
20 percent smaller than the composite value. The SEM mass median
diameters are also slightly larger than the composite value at the
-------
oo o o o o
o en co N co i?)
(iurt)
o
ro
O
C\J
i
•H
0> H
.d 4J
4J 03
•H
E T3
O
S-l 0)
m 03
CO
T3 E
0)
a. o)
o >
iH -rH
0) 4-1
cn
O
S 6
o a)
03
0) i—I
42 O
03
I
•H
-H
,0
CO
u
-H
CO 0)
4J
O O> 4-J
J2 -H
03 H S
C
O CO
•H • 4-1
4-> to a
3 4J T3
.£> CO
•H T3
}-i
03 W
•H cy^ 4-1
id CL
t3 O
03
U3
a.
QJ o E
> o o
•H 03 M
W O U-l
cO (-1
t-l O 13
3 -H 0)
EEC
3 -H
O rH CO
CO 4-1
-------
industrial, urban and suburban sites, while the mass median diameter at
the rural sites is about 10 percent smaller than the composite value.
Although we cannot be certain, we believe that the relatively
large differences between the optical microscopy and SEM cumulative mass
distributions and mass median diameters are principally explained by the
biases inherent in the two techniques. The refractive index oil used to
prepare the optical microscopy samples tends to cause the particles to
be oriented such that the largest faces are parallel to the horizontal
surfaces of the filter samples. Also, the depth of field in the microscopy
analyses was much larger than the depth of field in the SEM analyses.
Experience gained during the study showed that it was relatively much
easier to identify the presence of large particles in the optical
microscopy analyses. Conversely, we found that the SEM analyses frequently
underestimated the number of large particles present in the filter
samples. Accepting the above statements at face value, it follows that
the actual large particle mass median diameters are probably intermediate
between the optical microscopy and SEM estimates.
Table III shows the average particulate concentration by site
classification and generic particle category. As might be expected, the
highest ambient particulate concentrations are found at industrial sites
and the lowest ambient particulate concentrations are found at rural
sites. On the average, the contributions of combustion products (flyash)
to ambient particulate concentrations at industrial, urban/suburban and
rural sites are 68, 50 and 36 percent, respectively. Similarly, the
average contributions of iron oxide particles to ambient particulate
concentrations at industrial, urban/suburban and rural sites are 7, 4
and 2 percent, respectively. Thus, particles of industrial origin
(combustion products and iron oxide) significantly affect ambient
particulate concentrations at all types of sites, but are of greatest
xii
-------
TABLE III
COMPOSITE SUMMARY OF AVERAGE PARTICULATE
CONCENTRATIONS BY SITE CLASSIFICATION
GENERIC PARTICLE CATEGORIES
Parameter
Average Filter Weight* (rag)
Average Air Concentration* (yg/m )
Combustion Products
Iron Oxide
Biologicals
Quartz and CaCO
TOTAL
Parameter Value
Industrial
Sites
350 ± 170
130 ± 70
14 ± 19
30 ± 40
18 ± 16
190 ± 100
Urban /Suburban
Sites
160 ± 80
70 ± 20
5 ± 4
12 ± 11
30 ± 30
140 ± 50
Rural
Sites
190 ± 70
40 ± 20
2 ± 3
30 ± 40
30 ± 30
110 ± 50
*The average filter weight or ambient air concentration plus or minus one
standard deviation.
xiix
-------
importance at industrial sites. Non-industrial particles (biologicals,
quartz and calcium carbonate) account for over 50 percent of ambient
particulate concentrations at rural sites and for 25 to 30 percent of
ambient particulate concentrations at industrial and urban/suburban sites.
A relatively large effort was made in the SEM and EDAX
analyses of selected large particles to identify the emission sources for
these particles. While much information was obtained on the geometry,
physical properties and the chemical elements present in the particles,
it was not generally possible to relate this information to specific
sources. One of the problems is that large quantities of furnace slag
have been produced over a period of many years by the steel industry in
the Southwest Pennsylvania AQCR. This slag material has been widely
used in the construction of buildings and roads and has also been
accumulated in large storage piles. We performed a statistical analysis,
using the EDAX' data, which ranked ten elements characteristic of the
particles collected on the filters according to the relative concentra-
tion of these elements in the individual particles. A comparison of this
ranking with a similar ranking of elements present in slag samples
obtained from various steel production facilities in the area showed
high correlations with basic oxygen and blast furnace slags. We
estimated that, on the average, about 20 percent of the total mass on
the filter samples had a chemical composition characteristic of furnace
slags.
It is important to note that the optical microscopy, SEM and
EDAX analyses described above were restricted to the coarse particle
mode (particle diameters greater than 3 micrometers) of the bimodal
mass distribution characteristics of urban areas (Whitby, et al., 1972
and others). The small particles of the fine particle mode (diameters
less than about 1 micrometer) are not detectable by optical techniques.
However, the results of the optical and SEM filter analyses showed that
the contributions of particles with diameters less than about 5 micrometers
to the total filter mass loadings were small. Because the current NAAQS
xiv
-------
for suspended particulates consider only the total weight collected on
hi-vol filters and do not specify particle size, we believe that the
neglect of particles smaller than 5 micrometers in diameter does not
significantly affect the results of the study given above.
Phase II Emissions Inventory and Diffusion-Model Calculations
The three principal differences between the Phase I and Phase
II particulate emissions inventories were: (1) the addition to the
particulate emissions from the coke-oven quenching process of the solids
contained in the quench water; (2) the addition of area source emissions
of dust resuspended from paved roads by motor vehicle traffic; and, (3)
the conversion of the coke-oven emissions parameters to the form required
to implement new procedures for calculating plume rise for coke ovens.
Additionally, the emissions inventory was updated to reflect changes
between 1973 and 1975, and some of the major particulate sources in
Ohio and West Virginia were added to the inventory to improve the
accuracy of the diffusion model calculations.
We used the 1975 particulate emissions inventory and Greater
Pittsburgh Airport meteorological data with the Cramer, e_t al., (1975)
long-term and short-term diffusion models to calculate both 1975 annual
average ground-level particulate concentrations and 24-hour average
ground-level particulate concentrations for two of the days included in
the special sampling program described above. In general, the correspon-
dence between calculated and observed 1975 annual average concentrations
was much better than obtained using the 1973 emissions inventory,
especially in the vicinity of large industrial sources. The agreement
between calculated and observed 24-hour average concentrations for the
two sample days was not as good as the agreement between calculated and
observed 1975 annual average concentrations. The difficulties in calcu-
lating 24-hour average particulate concentration patterns which agree
with the observed concentrations may in part be attributed to the fact
xv
-------
that the average particulate emissions from the major sources were used
in the model calculations because the actual emissions on the two sample
days were not available and are unknown. Direct comparisons of calculated
and observed 24-hour average concentrations were also complicated by
uncertainties in the observed concentrations and in the meteorological data.
For example, the concurrent 24-hour average concentrations measured by
the colocated hi-vol samplers at the 15 locations shown in Figure I
differed by as much as 50 micrograms per cubic meter. Also, the hourly
wind-direction observations from the Greater Pittsburgh Airport are inade-
quate for specifying accurate plume trajectories.
The results of both the annual and 24-hour concentration
calculations provided circumstantial evidence that: (1) unknown sources
not included in the industrial particulate emissions inventory significantly
affect many of the hi-vol monitor sites; (2) the entrainment of dust
from paved roads by motor vehicle traffic probably is one of the most
significant of the unknown sources; (3) advection of particulates from
urban and industrial areas outside of the Southwest Pennsylvania AQCR
may affect ambient particulate concentrations within the AQCR, especially,
along the western edge of the AQCR; and, (4) the particulate emissions
from the quenching process have been overestimated and/or the hi-vol
samplers do not retain a significant fraction of the very small particu-
lates contained in the vaporized quench water.
SUMMARY
The principal result of the study described in this report is
that particles with diameters larger than 25 micrometers account for
over 50 percent of the mass loading on hi-vol filters in the Southwest
Pennsylvania AQCR. In general, these large, particles appear to be of
industrial origin. The estimated average contribution of industrial
particles to ambient particulate concentrations ranges from about 75
percent in industrial areas to about 40 percent in rural areas. The
xvi
-------
elemental compositions of many of these large particles are correlated
at the 97.5-percent confidence level with the elemental compositions of
the slag which is produced by the steel industry and is found throughout
Southwest Pennsylvania in slag dumps. This slag also has been extensively
used in road and building construction. The processes by which slag and
other large particles which appear to be of industrial origin, but which
are not contained in the industrial emissions inventory, are released to
the atmosphere and transported to hi-vol monitor sites are unknown.
However, there is circumstantial evidence that the entrainment of dust
from paved roads by motor vehicle traffic is one of the primary mechan-
isms. The estimated average contribution of non-industrial particles
(biologicals, quartz and calcium carbonate) to ambient particulate
concentrations ranges from about 25 percent at industrial sites to about
55 percent at rural sites. At many sites, the presence in the hi-vol
filter samples of certain elements or generic particle categories
appears to be determined in part by the mean x^ind direction during the
sampling period, a result that suggests that specific sources are
responsible for the presence of these elements or generic particle
categories. However, the data generally were far too limited to define
any specific source-receptor relationships.
The updated industrial particulate emissions inventory for the
Southwest Pennsylvania AQCR, when used with appropriate diffusion
modeling techniques, yields calculated annual average particulate
concentrations that are in reasonable agreement with the concentrations
observed in the vicinity of large industrial sources. However, the
annual average concentrations calculated in downtown areas of heavy
traffic and in rural areas tend to be well below the observed concen-
trations. The calculated 24-hour average concentrations show the same
trends as the calculated annual average concentrations, but the uncer-
tainties in the calculated 24-hour concentrations arising from factors
such as uncertainties in short-term particulate emissions and meteorolo-
gical data are larger than the uncertainties in the calculated annual
concentrations.
xvii
-------
TABLE OF CONTENTS
VOLUME I
Section 1 Ti_tl^ Page
ACKNOWLEDGEMENTS i
EXECUTIVE SUMMARY . ii
1 INTRODUCTION 1-1
2 WORK ACCOMPLISHED DURING PHASE I 2-1
2.1 Review of Existing Emissions Inventory 2-1
2.2 Diffusion-Model Calculations 2-7
2.3 Existing Particulate Air Quality 2-30
2,4 Comparison of Calculated and Observed
1973 Annual Average Particulate Concen-
trations 2-44
2.5 Summary of Work Accomplished in Phase I;
Recommended Phase II Work Objectives and
Major Activities 2-51
3 DESCRIPTION OF THE PHASE II PARTICULATE
SAMPLING PROGRAM 3-1
3.1 Description of the Sampler Locations 3-1
3.2 Meteorological Conditions on the Six
Hi-Vol Sample Days 3-4
3.3 Sampling Procedures 3-8
3.4 Summary of Filter Samples, Weights
and Particulate Concentrations 3-22
3.5 Sample Preparation 3-22
3.6 Optical Microscopy Analysis Procedures 3-25
3.7 Scanning Electron Microscopy Analysis
Procedures 3-30
3.8 Study Reliability 3-42
4 RESULTS OF THE OPTICAL AND SEM/EDAX ANALYSES
OF THE HI-VOL FILTER SAMPLES 4-1
4.1 Particle Mass Distributions 4-1
4.2 Elemental Analyses 4-15
4.3 Wind-Direction Effects on Hi-Vol Filter
Mass Loadings 4-32
4.4 Differences in Filter Samples Between
Monitor Site Classifications 4-39
4.5 Summary of the Results of the Filter Analyses 4-39
XVlll
-------
VOLUME I (Continued)
Section Title
REVISION OF THE PHASE I EMISSIONS INVENTORY
5.1 Emissions from Industrial Sources and
Source Complexes
5.2 Emissions from Classical Area Sources
5.3 Reentrainment and Resuspension of Road
Dusts by Motor Vehicles
RESULTS OF THE PHASE II DIFFUSION MODEL CALCU-
LATION FOR PARTICULATE EMISSIONS FROM INDUSTRIAL
SOURCES
6.1 Introduction
6.2 Results of the Annual Average Concentrations
for 1975
6.3 Results of the 24-Hour Concentration Calcu-
lations for Selected Sampling Days
COMPARISON OF CALCULATED AND OBSERVED PARTICULATE
CONCENTRATIONS
7.1 Annual Concentrations for 1975
7.2 Twenty-Four Average Concentrations for
16 and 28 August 1976
REFERENCES
Page
5-1
5-1
5-2
5-12
6-1
6-1
6-3
6-17
7-1
7-1
7-4
8-1
Appendix
A
MATHEMATICAL MODELS USED TO CALCULATE GROUND-LEVEL
CONCENTRATIONS
A.I Introduction
A. 2 Plume Rise Formulas
A.3 Short-Term Concentration Model
A.4 Long-Term Concentration Model
A.5 Application of the Short-Term and Long-Term
Concentration Models in Comolex Terrain
A-l
A-l
A-7
A-9
A-16
A-21
SEASONAL AND ANNUAL FREQUENCY DISTRIBUTIONS OF WIND
SPEED AND WIND DIRECTION B-l
INDUSTRIAL PARTICULATE EMISSIONS INVENTORY C-l
XXX
-------
TABLE OF CONTENTS
VOLUME II
Appendix Title Page
ADDITIONAL DETAILS OF THE HI-VOL FILTER ANALYSIS
PROGRAM D-l
xx
-------
SECTION 1
INTRODUCTION
Efficient management of air resources in urban and industri-
alized areas requires a detailed knowledge of source-receptor relation-
ships. In the case of total suspended particulates, it is extremely
difficult to define these relationships because there are both natural
and man-made sources of particulates. Man-made sources of particulates
in the atmosphere include stack and fugitive emissions from stationary
industrial sources, particulate emissions resulting from motor vehicle
operations, and windblown dust from unpaved roads and cultivated fields.
Wind erosion, pollen and forest fires are among the natural causes of
particulates in the atmosphere. With the exception of particulate
emissions from industrial stacks, it is often difficult to quantify the
emissions associated with the various natural and man-made sources of
particulates in the atmosphere. Additionally, the contributions of the
various particulate sources to observed ambient particulate concentrations
are generally not known.
The work performed by the H. E. Cramer Company, Inc. under
Contract No. 68-02-1387, Task Order No. 3 with the U. S. Environmental
Protection Agency relates to particulate emissions and the impact of
these emissions on ambient air quality in the Southwest Pennsylvania
Intrastate Air Quality Control Region (AQCR). Existing air quality data
obtained from approximately 35 high volume air sampling stations, many
of which are located near large industrial sources of particulates,
indicate total particulate loadings in excess of the 24-hour and annual
standards. Also, there are serious deficiencies in existing knowledge
of the sources of particulate emissions, especially with regard to fugi-
tive emissions which are thought to contribute importantly to the
existing high particulate loadings revealed by the high-volume air
sampling data. In order to establish reasonable emission control measures
1-1
-------
that will achieve the reduction in pollution levels needed to attain
the national air quality standards for particulates in the Southwest
Pennsylvania AQCR, it is necessary to identify the major particulate
sources and assess their contributions to ambient air quality.
The Southwest Pennsylvania AQCR is comprised of the nine Penn-
sylvania counties shown in Figure 1-1. Although Lawrence County is
not part of the Southwest Pennsylvania AQCR, it has been included in
the study and is shown in Figure 1-1 because sources in Lawrence County
are believed to affect the ambient air quality in Beaver County and pos-
sibly Butler County. The Pennsylvania Department of Environmental Re-
sources is responsible for enforcing air pollution control regulations
throughout the area shown in Figure 1-1 with the exception of Allegheny
County. In Allegheny County, the Department of Environmental Resources
has delegated its responsibility to the Allegheny County Bureau of Air
Pollution Control. The Allegheny County Bureau of Air Pollution Control
offices are in Pittsburgh. The Department of Environmental Resources
maintains a central office in Harrisburg, Pennsylvania and regional of-
fices in Meadville and Pittsburgh.
The work under the contract, which is described in this report,
was divided into two major phases:
Phase I - Review and evaluation of present air quality and
emissions data relative to suspended particulates to deter-
mine the most efficient course of action to bring the emissions
inventory to an acceptable level of accuracy.
Phase IT - Updating of the emissions inventory; collection
and analysis of ambient air quality data; and, the develop-
ment and validation of a particulate diffusion model for
the Southwest Pennsylvania AQCR.
-------
WESTMORELAND CO
WASHINGTON CO
10 15 20 KILOMETERS
FIGURE 1-1. Map showing the counties that comprise the Southwest Pennsyl-
vania Intrastate Air Quality Control Region. Although not a
part of the Southwest Pennsylvania AQCR, Lawrence County is
included in the figure because sources with the County are be-
lieved to affect the ambient air quality in the Southwest Penn-
sylvania AQCR.
1-3
-------
SECTION 2
WORK ACCOMPLISHED DURING PHASE 1
The work under Phase I was divided into three major tasks:
• Review and evaluation of the existing particulate
emissions data for the Southwest Pennsylvania AQCR
• Review and evaluation of the existing air quality data
for the Southwest Pennsylvania AQCR
• Determination of specific Phase II activities for
updating the particulate emission inventory for the
Southwest Pennsylvania AQCR and bringing the emissions
data to an acceptable level of accuracy
This section of the report describes in detail the work effort under these
tasks and the results achieved during Phase I.
2.1 REVIEW OF EXISTING EMISSIONS INVENTORY
Particulate emissions data for the Southwest Pennsylvania AQCR
are available from the Commonwealth of Pennsylvania Department of Environ-
mental Resources (DER) and the Allegheny County Bureau of Air Pollution
Control (BAPC). The BAPC emissions data refer to sources located within
Allegheny County while the DER emissions refer to sources in the other
counties within and adjacent to the Southwest Pennsylvania AQCR (see
Figure 1-1).
2-1
-------
2.1.1 Review and Updating of the Department of Environ-
mental Resources Particulate Emissions Inventory
The central office of the Department of Environmental Resources
at Harrisburg, Pennsylvania provided us with a copy of their computerized
emissions inventory for the Southwest Pennsylvania Intrastate Air Quality
Control Region. This copy of the emissions inventory, dated 25 July
1974, is an organized list of the sources within each of the counties in
the Air Quality Control Region. The format of the inventory provided
for the following'technical data entries:
• Work shifts per day
• Hours per work shift
• Work days per week
• Work days per year
• Months of peak loads or process
• Lists of boilers
• Waste disposal information
• Lists of process types and equipment
• Normal and peak processing loads
• Process exhaust volumes
• Emission types
• Emission rates
• Basis for emission rates
• Stack parameters
• Pollution control equipment
• Control equipment efficiency
• Installation date of control equipment
• Total emissions from each source
The initial review of the computerized particulate emissions
inventory resulted in numerous questions pertaining to individual sources.
These questions were referred to the engineering staff of the Department
of Environmental Resources. During visits to the local offices of the
2-2
-------
Department of Environmental Resources at Meadville and Pittsburgh made
in September 1974, each of the sources under their jurisdiction was
discussed, the emissions inventory for each source was reviewed and, as
far as possible, missing technical data in the inventory were filled in.
After these visits with the local engineering staffs of the Department
of Environmental Resources, a critical review of the resulting particu-
late emissions inventory was conducted with special emphasis placed on
estimating the accuracy and validity of the reported data. Special
attention was paid to those sources for which the particulate emissions
inventory had been generated through the use of emission factors. In
some cases, incorrect emissions factors and/or mathematical errors were
detected and corrected. Appropriate adjustments were also made in the
inventory to reflect recent changes in source configuration and/or types
of equipment discovered in reviewing the notes taken during the technical
discussions with the local engineering staffs of the Department of Environ-
mental Resources.
As part of the critical review, the particulate emissions data
for the 25 largest particulate sources located in the nine counties
under the jurisdiction of the Department of Environmental Resources were
given to an independent consultant for evaluation. In his evaluation,
the consultant considered the type of source, the process, the age of
the facility and the reported emissions for the source in relation to
the reported emissions from other similar sources. The conclusion of the
consultant was that the emissions data for the 25 sources were in general
reasonable but that a number of specific details needed to be checked
and clarified.
A second series of technical visits with the engineering staff
of the Department of Environmental Resources at Meadville and Pittsburgh
was made during January 1975. During these visits, all suspected errors
and suggested changes were discussed and resolved with the engineering
staffs.
2-3
-------
2.1.2 Review and Updating of the Allegheny County Parti-
ticulate Emissions Inventory
During an initial meeting with officials of the Allegheny
County Bureau of Air Pollution Control, it was revealed that detailed
emissions inventory files were maintained by individual engineers in
the Bureau assigned the responsibility for sources on the basis of source
type. That is, individual engineers are responsible for all sources of
a single type (i.e., power plants, cement products, refractories, com-
mercial boilers, etc.). For this reason, the most practicable method
for us to use in developing an updated emissions inventory for Allegheny
County was to review each source in each staff engineer's files. The
Allegheny County Bureau of Air Pollution Control provided a list of the
major sources located within the County and the name of the responsible
engineer. Each engineer maintained a file on each source which included
copies of permits, variance schedules, applications, source tests and other
information that he had obtained pertaining to the source. Each engineer
was interviewed to obtain the pertinent data contained within his files
and to discuss his personal knowledge of the source's characteristics,
operating schedules, control equipment and emission rates. The results
of any source tests that had been conducted, as well as emission factor
estimates or personal knowledge of the emissions from individual sources
were also obtained. Several visits to Pittsburgh and the Allegheny
County Bureau of Air Pollution Control were necessary to conduct the
inverviews with the engineering staffs. The forms shown in Figures
2-1 and 2-2 indicate the source information and emissions data that were
in the files of the Allegheny County Bureau of Air Pollution Control.
2.1.3 Updated Particulate Emissions Inventory for the
Southwest Pennsylvania Intrastate AQCR
During the review of the emissions data supplied by the
Allegheny County Bureau of Air Pollution Control and the Department of
2-4
-------
2
a:
0
u_
-
V) U.
Ul 75
O hj
O UJ Q_
< K Jj UJJ
o 5 3 K =
H ilj — ' O U.
2*
UJ
o.
o
u.
(J
UJ
a.
to
to o:
(/> UJ UJ UJ
UJ h- — X CO
L> - v.
< U ^
2-
3 Ll_ 3 Q > ^ z z
O < u u
2-5
-------
t—
z
UJ
5
LU
_J
0_
0_
"J
^J
tf\
\J)
(T
£
4*f
1
<
a
SOURCE
>
o
<
o
o
o
4
u.
o
UJ
\-
^
2
&—
en
UJ
o
Ul
z
X
ce
UJ
UJ
o
0
J
o
I
0
UJ 1.1
. UJ
KI-C
<< F
1**
"*'£
peg
a:
DIAMETER
INNER AT TOP
(METERS)
?5
2^1
KIUJ
S^1
UJ O
u ~.
J7j ^ O
Q.U-1
«So
<«0)
UJ
>
0 •£
03 ? 2
•«s
ts-
"^5
Ul
I
X
u
<
K
^3
U
U
*I
U.
0
Ul
^~
<:
Z
h-
CO
UJ
O
Ul
z
2
(£
UJ
UJ
Q
O
O
I
a
Wu
K£-
PARTICUL^
EMISSION RA
(TONS/YF
c!
t
cc
u
I/)
UJ
Q
UJ
0
a
3
0
to
iCu7 COS CO
OK UJt CO
<4 -^ ** 1,1
or -^ a: UJ
1-5 Cu O
to* tz O
i §i g
uj — 'uj 0_
o
cc
}~
•z.
o
u
<->
z
cc
l-
u
Ul
a.
4-1
C
cu
c.
«
ce
3
o
Ul
o
o
ce
a.
<
cc
UJ
a.
o
o
z
I "j§
Ju.
<0
cc
n
12z
xo
GJ
o
o
LO
r-J
I
tJ
erf
3
c;
cc
<
>
z
o
IUJ
u.
o
cc
o
t-
4
z
SIO
YR;
OWED E
(TO
O
cc
I-
z
o
o
-------
Environmental Resources, approximately 510 sources or source complexes
were considered. The Department of Environmental Resources supplied
information pertaining to approximately 300 sources or source complexes,
and the Allegheny County Bureau of Air Pollution Control supplied infor-
mation on approximately 210 sources or source complexes. After isolated
sources and source complexes with emission rates less than 22 tons per
year were eliminated from further consideration, approximately 200
sources and source complexes remained. These comprise the updated
particulate emissions inventory for industrial sources in the Southwest
Pennsylvania AQCR.
Gross estimates of the particualte emissions from domestic
heating, vehicle exhaust and tire wear were obtained using emission
factors. The vehicle mileage data were supplied by EPA Region III and
are given in Table 2-1. Particulate emissions resulting from domestic
heating were estimated on the basis of heat requirements and the popula-
tion of the area. Liberal allocations of fuel oil (40 percent) and coal
(10 percent) as domestic heating fuels supplementing natural gas were
used, which should yield maximum particulate emission rates from domestic
heating. Tables 2-2 and 2-3 present the estimated particulate emissions
due to vehicle traffic and domestic heating, respectively. These emissions
are quite small compared to the particulate emissions from industrial
sources. As pointed out in Section 2.4, using the above emission rates
in our long-term area source model yields estimates of the maximum
annual particulate concentration contributed by both vehicle emissions
and domestic heating of the order of 1 microgram per cubic meter.
2.2 DIFFUSION-MODEL CALCULATIONS
In order to evaluate the adequacy of current emissions data
and to assess the relationship between reported particulate emissions
and observed air quality, we calculated 1973 annual average ground-level
particulate concentrations for the following areas:
2-7
-------
TABLE 2-1
1972 VEHICLE MILEAGE FOR SELECTED AREAS IN THE SOUTHWEST
PENNSYLVANIA INTRASTATE AIR QUALITY CONTROL REGION
Area
Pittsburgh
Allegheny County
Butler County
Armstrong County
Westmoreland County
Washington County
Beaver County
Daily Vehicle Mileage
Light Duty
Vehicles
3,458, 169
13,578,293
2,086,405
1,029,916
4,931,992
3,015, 601
1,905,955
Heavy Duty
Vehicles
177, 722
450,671
34,133
44,779
138,280
62, 049
80, 675
Diesel
Vehicles
66, 646
171,627
12,800
17,475
51,215
24,820
30,253
Total
3,702, 537
14,200, 591
2,133,338
1,092,170
5,121,487
3,102,470
2,016, 883
2-8
-------
TABLE 2-2
ANNUAL PARTICULATE EMISSIONS FROM VEHICLES FOR SELECTED
AREAS OF THE SOUTHWEST PENNSYLVANIA AQCR (tons/km2)
Area
Pittsburgh
Allegheny County
Butler County
Armstrong County
Westmoreland County
Washington County
Beaver County
Light Duty
Vehicles1
5.22
1.56
0.22
0.11
0.40
0.29
0.36
Heavy Duty
Vehicle s^
0.42
0.08
0.00
0.01
0.02
0.01
0.02
Diesel
Vehicles-^
0.22
0.04
0.00
0.00
0.01
0.01
0.01
Total
5.86
1.69
0.22
0.12
0.43
0.31
0.40
Emission Factors from EPA Publication AP-42:
0. 54 g/mi
"0. 85 g/mi
1.18 g/mi
2-9
-------
TABLE 2-3
PARTICULATE EMISSIONS FROM DOMESTIC HEATING FOR SELECTED
AREAS OF THE SOUTHWEST PENNSYLVANIA AQCR
Area
New Castle
Beaver
Pittsburgh
Population
Density
(km2)
369.4
369.4
2383.4
Particulate Emissions (tons/km")
Fuel Oil
0.54
0.53
1.59
Natural
Gas
0.20
0.19
0.32
Propane
0.04
0.04
0.00
Coal
0.27
0.27
1.75
Total
0.50
0.53
3.66
2-10
-------
» New Castle
• Beaver Valley
• Pittsburgh
The emissions and meteorological data used in the calculations are
briefly discussed in Sections 2.2.1 and 2.2.2, respectively. In Section
2.2,3, the calculation procedures are described and the results of the
calculations presented. Calculated and observed concentrations are
compared in Section 2.2.4.
2.2.1 Source and Emissions Data
Tables 2-4, 2-5 and 2-6 list the major particulate sources and
source complexes in the New Castle, Beaver Valley and Pittsburgh areas,
respectively. The tables also show, for each source or source complex,
the number of individual sources used in the concentration calculations,
the Universal Transverse Mercator (UTM) coordinates of the source or
center of the source complex and the total annual particulate emissions
in tons per year. Emissions from classical area sources such as motor
vehicles and non-industrial fugitives, as well as emissions from industrial
sources located outside the Southwest Pennsylvania AQCR, were not con-
sidered in the calculations.
2.2.2 Meteorological Data
Meteorological inputs required by the long-term concentration
model described in Appendix A were principally obtained from 1973
seasonal distributions of wind speed and wind direction, classified
according to the Pasquill stability categories, for the Greater
Pittsburgh and Allegheny County Airports. These seasonal distributions
were developed from surface weather observations using the definitions
2-11
-------
TABLE 2-4
SOURCES, LOCATIONS AND PARTICULATE EMISSION RATES USED
TO CALCULATE 1973 ANNUAL AVERAGE GROUND-LEVEL
CONCENTRATIONS IN THE NEW CASTLE AREA
Source Details
Source
Number
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
Source Name
Babcock and Wilcox
(Koppel)
Pennsylvania Power
Medusa Portland Cement
Bessemer Cement
Van Port Stone
Shenango China
Fenati Brick
Elwood Steel Casting
Pentex
New Castle Refractories
American Metallurgical
Products
Gennaro Asphalt
New Castle Foundry
Blair Strip Steel
W. R. Grace
Number of
Individual
Sources
5
6
38
39
2
16
9
2
4
5
3
2
3
1
2
UTM Coordinates
X
(m)
557,473
553, 105
556,874
543,049
554,014
554,077
558,024
560,684
557,581
553, 659
557,676
541,922
554, 238
557,292
552, 939
Y
(m)
4,520,452
4,531,767
4,525,289
4,535, 803
4,532,237
4, 539, 546
4,539, 637
4,523,068
4,536, 581
4,536, 089
4,536,427
4,540, 823
4,536, 679
4,537, 689
4,535,436
Total
Particulate
Emissions
(tons/year)
35. 8
7054.8
4522.4
2385.5
270.0
558.3
492. 6
90.2
32.4
24.7
103.8
18.6
10.0
13.8
20.8
TOTAL INDIVIDUAL SOURCES 137
2-12
-------
TABLE 2-5
SOURCES, LOCATIONS AND PARTICULATE EMISSION RATES USED
TO CALCULATE 1973 ANNUAL AVERAGE CONCENTRATIONS
IN THE BEAVER VALLEY AREA
Source Details
Source
Number
1
2
3
4
5
6
7
8
9
10
11
12
13
14
Source Name
Jones and Laughlin
Steel Corporation
Crucible, Inc.
St. Joe Minerals Corp.
Zinc Smelting Division
Sinclair-Koppers Co.
Babcock & Wilcox Co.
Armstrong Cork Co.
Ambridge Municipal
Auth. Boro
Ashland Oil, Inc.
Armco Steel Corp.
MacKintosh-Hemphill
Div. , E.W. Bliss Co.
The Phoenix Glass Co.
Mayfield Foundry, Inc.
Wilofsky Bros. , Inc.
Darlington Brick &
Clay Products Co.
Number of
Individual
Sources
42
38
31
66
4
20
1
3
12
2
6
1
1
4
UTAI Coordinates
X
(m)
564,299
545,595
556,112
554, 530
555,264
557,097
564,819
562,314
565,095
546,277
560,696
556, 399
558, 009
547,281
Y
(m)
4,494,976
4,498,168
4,502,248
4,500,263
4,515,255
4,511,476
4,494,641
4,504, 580
4,495,353
4,497,987
4,504, 381
4,513,845
4,511, 853
4,517,420
Total
Particulate
Emissions
(tons/year)
13,924.9
2,806.3
1,243. 1
985.9
308. 9
188.4
125.8
109. 6
90. 7
79. 5
78.2
51.0
51.0
43. 8
2-13
-------
TABLE 2-5 (Continued)
Source Details
Source
Number
15
16
17
18
19
20
21
22
23
24
Source Name
Babcock & Wilcox Co.
(Koppel)
Mercer Lime & Stone Co.
Colona Div. Ampco-
Pitts. Corp.
Brighton Electric Steel
Casting
Republic Steel Corp.
Metropolitan Brie, Inc.
Mayer China Div. of
Interpace Corp.
Interstate Amiesite Corp.
Duquesne Light Phillips
Power Station
Hussey Metals
Number of
Individual
Sources
5
3
1
4
3
3
10
2
5
1
UTiM Coordinates
X
(m)
557,473
561,254
561,498
556, 392
557,458
547,938
558,039
549,465
565,226
565,881
Y
(m)
4,520,452
4,522, 148
4,503,987
4,514,801
4,510, 399
4,517,208
4,511,021
4,499, 118
4,491, 098
4,491,474
Total
Particulate
Emissions
(tons/year)
35.9
35. 4
31. 2
26.0
25.4
16. 6
11. 5
7. 7
10,500.0
11.4
TOTAL INDIVIDUAL SOURCES 268
2-14
-------
TABLE 2-6
SOURCES, LOCATIONS AND PARTICULATE EMISSION RATES USED
TO CALCULATE 1973 ANNUAL AVERAGE CONCENTRATIONS
IN THE PITTSBURGH AREA
Source Details
Source
Number
1
2
3
4
5
6
7
8
9
10
11
12
13
14
Source Name
USS - Clairton Works
Duquesne Light Phillips
Power Station
USS - Edgar Thomson
Jones & Laughlin
Steel Corp.
USS - Homestead
USS - National Works
Heppenstall
USS - Duquesne
USS - Irvin
H. J. Heinz
Universal Atlas
Cement (USS)
Fort Pitt Steel Casting
Mcsta Machine
Pittron
Number of
Individual
Sources
129
5
7
10
10
7
1
5
7
2
1
3
2
2
UTM Coordinates
X
(m)
595,380
565,226
596,990
589,150
591,880
597,400
588,120
598,120
593,230
586,000
602, 178
597, 650
591,157
593,850
Y
(m)
4,461,930
4,491,098
4,471,670
4,474,030
4,473,220
4,467,330
4,480,930
4,469,830
4,465,650
4,478,900
4,478, 561
4,465, 190
4,472,405
4,464, 500
Total
Particulate
Emissions
(tons/year)
18,415.7
10,500.0
9,303.2
7,212.0
4,200.1
3,901. 8
3,702. 6
1,723.9
1,349. 6
543. 1
176.8
96.0
165. 8
120.3
2-15
-------
TABLE 2-6 (Continued)
Source Details
Source
Number
15
16
17
18
19
20
21
22
23
24
25
26
Source Name
Duquesne Light Brunot
Island Power Station
McComvay Torley
Pittsburgh Brewing
B & O Railroad
WABCO (includes Union
Switch and Signal)
Allegheny County Steam
Heating (Stamvix Street)
Allegheny County Steam
Heating (12th Street)
Bellefield Boilers
Duquesne Light Sinter
Plant
Mercy Hospital
Duquesne Light Co.
Elrama Power Station
West Penn. Power Co.
Mitchell Station
Number of
Individual
Sources
3
1
1
1
18
1
1
2
1
1
4
4
UTM Coordinates
X
(m)
580, 680
588,110
587,550
589,610
594,400
584,380
585,200
589,190
591,930
586,019
592,000
587,340
Y
(m)
4,479, 720
4,481,200
4,479,280
4,472,680
4,475, 550
4,477,300
4,477, 600
4,477, 100
4,456, 300
4,476,415
4,456,200
4,452, 810
Total
Particulate
Emissions
(tons/year)
104.0
88.0
79.5
70.0
67. 5
49.9
49.9
60. 0
203.6
8.0
27,418.8
439. 8
TOTAL INDIVIDUAL SOURCES 228
2-16
-------
of the Pasquill stability categories given by Turner (1964), which
are based on solar radiation (insolation) and wind speed. Tables 2-7
and 2-8 list the parameters that define the various stability categories.
The wind speeds in Table 2-9 are in knots because airport surface wind
speeds are reported to the nearest knot by the National Weather Service
and Turner's classification is based on this convention. The thermal
stratifications represented by the various Pasquill stability categories
are:
• A - Very Unstable
• B - Unstable
• C - Slightly Unstable
• D - Neutral
• E - Slightly Stable
• F - Stable
Because no cloud-cover data were available for Allegheny County Airport,
hourly surface wind observations at the Allegheny County Airport were
used with concurrent 3-hourly cloud-cover observations at the Greater
Pittsburgh Airport to generate the seasonal wind distributions by stabil-
ity category for the Allegheny County Airport.
Figure 2-3 shows the 1973 annual frequency distributions of
wind direction at the Greater Pittsburgh Airport (dashed line) and
Allegheny County Airport (solid line). Inspection of the figure reveals
that, although the two distributions are generally similar, the most
frequent winds at the Greater Pittsburgh Airport are from the west,
while those at the Allegheny County Airport are from the south and west-
southwest. The Allegheny County Airport wind distributions were used in
the calculations for the Pittsburgh area and the Greater Pittsburgh Air-
port wind distributions were used in the calculations for the Beaver Valley
and New Castle areas. However, it should be noted that the surface winds
2-17
-------
TABLE 2-7
PASQUILL STABILITY CATEGORY AS A FUNCTION
OF INSOLATION AND WIND SPEED
Wind
Speed
(knots)
0,1
2,3
4,5
6
7
8,9
10
11
£= 12
Insolation Index
4
A
A
A
B
B
B
C
C
C
3
A
B
B
B
B
C
C
C
D
2
B
B
C
C
C
C
D
D
D
1
C
C
D
D
D
D
D
D
D
0
D
D
D
D
D
D
U
D
D
-1
F
F
E
E
D
D
D
D
D
-2
F
F
F
F
E
E
E
D
D
TABLE 2-8
INSOLATION CATEGORIES
Insolation Category
Insolation Index
Strong
Moderate
Slight
Weak
Overcast < 7,000 feet (day or night)
Cloud Cover > 4/10 (night)
Cloud Cover < 4/10 (night)
4
3
2
1
0
-1
-2
2-18
-------
TABLE 2-9
WIND-PROFILE EXPONENTS USED IN THE ANNUAL AVERAGE
CONCENTRATION CALCULATIONS
Pasquill
Stability
Category
A
B
C
D
E
Wind-Speed Category (m/scr)f
0.0 - 1. 5
0. 10
0. 10
0.20
0.25
0. 30
1. (5 - 3. 1
0. 10
0. 10
0. 15
0. 20
0.25
3.2 - 5. 1
--
0. 10
0. 10
0. 15
0.20
5.2 - 8.2
--
--
0. 10
0. 10
--
8. 3 - 10. S
--
--
0. 10
0. 10
--
> 10.8
--
--
--
0. 10
--
^Measurement height is (>. 1 meters above the ground surface.
2-19
-------
NW
WNW
WSW
•ALLEGHENY COUNTY AIRPORT 1973
GREATER PITTSBURGH AIRPORT 1973
NNW N
102019
NNE
ENE
ESE
FIGURE 2-3. Annual frequency distributions of wind direction during 1973
at Allegheny County Airport (solid line) and the Greater
Pittsburgh Airport (dashed line). Percent frequency scale
is shown at left center.
2-20
-------
at the two airports cannot be expected to give an accurate representa-
tion of the low-level wind circulation over the entire area of interest.
The seasonal and annual wind distributions for the two Pittsburgh airports
are presented in Appendix B.
In the diffusion models described in Appendix A, the variation
with height of the wind speed in the surface mixing layer is assumed to
follow a wind-profile exponent law of the form
"{z} = U{ZR} (— ) (2-1)
where
u{z} = wind speed at height z above the surface
u{z } = wind speed at a reference height z above the surface
K K
p = wind-profile exponent
In the case of discharges from tall stacks, as discussed in Sections A.3
and A.5 of Appendix A, the wind-profile exponent law is used to adjust
the mean wind speed from the reference (airport-measurement) height to
the stack height for the plume rise calculations, and to the plume stabili-
zation height for the concentration calculations. In the case of low-
level emissions, which are generally treated as building sources, the
wind-profile exponent law is similarly used to obtain the wind speed at
the assigned source height which depends on the vertical dimensions of
the buildings or other structures. Values for the wind-profile exponent
p assigned to the various combinations of wind speed and stability for
the long-term calculations are listed in Table 2-9. These exponent
values are based on the results obtained by DeMarrais (1959) and Cramer,
et al. (1972).
2-21
-------
Our vertical expansion (o ) curves, which include the effects
of the initial vertical plume or building dimension, relate the vertical
turbulent intensity directly to plume growth (see Equation (A-13) of
Appendix A). Table 2-10 lists the values of the standard deviation of
the wind-elevation angle al corresponding to the Pasquill stability
Ij
categories for rural and urban areas. The rural a' values are based in
part on the measurements of Luna and Church (1971) and are consistent
with the a' values implicit in the a curves presented by Pasquill
LJ Z
(1961). In order to adjust for the effects of surface roughness elements
and heat sources, the 0' values for the stability category one step
LJ
more unstable than the indicated stability category are used in the
calculations for urban areas. A procedure of this type is suggested by
Calder (1971), Bowne (1974) and others. The E and F stability categories
are combined in urban areas because we believe that the effects of
surface roughness elements and heat sources are incompatible with the
minimal turbulent mixing associated with the Pasquill F stability category.
The height of the top of the surface mixing layer is defined
as the height at which the vertical intensity of turbulence becomes
effectively zero. This condition is fulfilled when the vertical turbulent
intensity is of the order of 0.01 or smaller. Since direct measurements
of the intensity of turbulence are not routinely made, indirect indicators
such as discontinuities in the vertical wind and temperature profiles
must be used to estimate the depth of the surface mixing layer. In the
simplest case, the base of an elevated inversion layer is usually assumed
to repesent the top of the surface mixing layer. However, even with a
surface-based inversion, a shallow mechanical mixing layer will exist
due to the presence of surface roughness elements and, in urban areas,
surface heat sources.
Holzworth (1972) has developed a procedure for estimating
early morning and afternoon mixing depths for urban areas from rawinsonde
2-22
-------
TABLE 2-10
VERTICAL TURBULENT INTENSITIES FOR RURAL
AND URBAN AREAS
Pasquill Stability
Category
-------
observations and surface temperature measurements. Tabulations of daily
observations of the depth of the surface mixing layer, developed by using
the Holzworth (1972) procedures, are available for most rawinsonde stations
operated by the National Weather Service. For the seasonal concentration
calculations, we analyzed seasonal tabulations of daily observations of
mixing depth and average surface wind speed at the Greater Pittsburgh
Airport for the period 1960 through 1964 (Environmental Data Service,
1966) in order to determine seasonal median early morning and afternoon
mixing depths for each wind-speed category. The median afternoon mixing
depths were assigned to the A, B and C stability categories; the median
early-morning mixing depths were assigned to the combined E and F stability
categories; and the median early morning and afternoon mixing depths
were averaged and assigned to the D stability category. Table 2-11 gives
the seasonal median mixing depths for the joint combinations of the wind-
speed and stability categories determined for the Pittsburgh area.
The Briggs (1971; 1972) plume-rise formulas given in Section A.2 of
Appendix A require the ambient air temperature as an input. For the
seasonal concentration calculations, seasonal average afternoon tempera-
tures measured at the Greater Pittsburgh Airport during the period 1963
through 1972 were assigned to the A, B and C stability categories; average
morning and evening temperatures were assigned to the D stability category;
and average nighttime temperatures were assigned to the combined E and F
categories. Table 2-12 lists the ambient air temperatures used in the long-
term calculations.
The Briggs (1971; 1972) plume-rise formulas given in Section A.2 of
Appendix A also require the vertical potential temperature gradient as an
input. Table 2-13 lists the vertical potential temperature gradients used
in the long-term concentration calculations. The potential temperature
gradients in Table 2-13 were assigned on the basis of the Turner (1964) and
Pasquill (1961) definitions of the Pasquill stability categories, the
measurements of Luna and Church (1972), and our own previous experience.
2-24
-------
TABLE 2-11
MIXING-LAYER DEPTHS IN METERS USED IN THE
ANNUAL CONCENTRATION CALCULATIONS
Pasquill Stability
Category
Wind-Speed Category (m/sec)
0-1.5
1.6-3.1
3.2-5.1
5.2-8.2
8.3-10. 8
>10. 8
(a) Winter
A
B
C
D
E
500
500
500
320
140
650
650
650
470
290
710
710
670
630
—
—
710
710
--
—
—
710
710
--
—
--
—
710
—
(b) Spring
A
B
C
D
E
1530
1530
1530
825
120
1530
1530
1530
920
310
—
1530
1530
1030
530
—
—
1530
1415
—
—
__
1530
1530
--
—
—
1530
—
(c) Summer
A
B
C
D
E
1730
1730
1730
960
190
1730
1730
1730
1025
320
1730
1730
1235
740
—
--
1730
1295
—
__
1730
1295
—
—
—
—
1295
—
(d) Fall
A
73
C
D
E
1230
1230
1230
685
140
1230
1230
1230
740
250
—
1230
1230
970
710
—
—
1230
1190
—
—
--
1230
1230
--
—
—
__
1230
—
2-25
-------
TABLE 2-12
AMBIENT AIR TEMPERATURES USED IN THE ANNUAL
AVERAGE CONCENTRATION CALCULATIONS
Pasquill
Stability
Category
A
B
C
D
E
Ambient Air Tempo rain re (°K)
Winter
273.2
273.2
273.2
271.2
269. 7
Spring
287.0
287.0
287.0
283.7
280.3
Summer
298.3
298.3
298.3
294.4
290.7
Fall
289. 5
289.5
289.5
286.3
282.4
TABLE 2-13
VERTICAL POTENTIAL TEMPERATURE GRADIENTS IN DEGREES KELVIN
PER METER USED IN THE ANNUAL AVERAGE CONCENTRATION
CALCULATIONS
Pasquill
Stability
Category
A
B
C
D
E
Wind-Speed Category (in/sec)
0.0 - 1. 5
0.000
0.000
0.000
0.015
0. 030
1. 6 - 3. 1
0.000
0. 000
0.000
0.010
0.020
3.2-5.1
0.000
0.000
0.005
0.015
5.2-8.2
0. 000
0. 003
8. 3 - 10.8
0. 000
0. 003
> 10.8
:
0. 003
2-26
-------
2.2.3 Calculation Procedures and Results
The source data discussed in Section 2.2.1 and the meteorological
data discussed in Section 2.2.2 were used with the long-term concentration
model described in Section A.4 of Appendix A to calculate annual average
ground-level particulate concentrations for the New Castle, Beaver Valley
and Pittsburgh areas. In each case, the calculation grid was comprised of
grid points spaced at regular 1-kilometer intervals and discrete points
corresponding to the locations of the air quality monitors. The procedures
described in Section A.5 of Appendix A were used to account for the ef-
fects of variations in terrain height over the calculations grid. No back-
ground particulate concentrations were incorporated in the calculated
concentrations.
Figure 2-4 shows the calculated isopleths of annual average ground-
level particulate concentration for the New Castle area. There are four
areas within which the calculated concentrations exceed 25 micrograms per
cubic meter. These areas of high concentration are almost entirely due
to contributions from nearby sources. For example, Shenango China accounts
for the majority of the calculated concentrations in the Union area. Simi-
larly, Fenati Brick contributes the majority of the calculated concentra-
tions in the New Castle area and Medusa Portland Cement contributes the
majority of the calculated concentrations in the Chewton area.
Figure 2-5 shows the calculated isopleths of annual average ground-
level particulate concentration in the Beaver Valley area. There are
three areas within which the calculated concentrations exceed 60 micrograms
per cubic meter. Jones and Laughlin Steel contributes most of the calcu-
lated concentrations in the areas northeast of Aliquippa. In addition to
Jones and Laughlin, Armco Steel and Ambridge Municipal Authority contribute
importantly to the calculated high concentrations east of Aliquippa.
2-27
-------
104006
4540
-— 4535
4530
4525
555
560
FIGURE 2-4. Calculated isopleths of 1973 annual average particulate
concentration in micrograms per cubic meter for the New
Castle area.
2-28
-------
FIGURE 2-5. Calculated isopleths of annual average ground-level
particulate concentration in micrograms per cubic
meter for the Beaver Valley area.
2-29
-------
Figures 2-6 and 2-7 show the calculated isopleths of annual
average ground-level particulate concentration in the Pittsburgh areas
of Clairton-Liberty Borough and Hazelwood-Braddock, respectively. As
shown by Figure 2-6, the calculated concentrations exceed the annual
Primary Air Quality Standard of 75 micrograms per cubic meter in a large
area approximately centered on the Clairton Coke Works and in a small
area west of Elizabeth. Additionally, Figures 2-6 and 2-7 show that the
annual standard is exceeded in at least three areas along the Monongahela
River. Emissions from the Clairton Coke Works are principally responsible
for the high calculated concentrations in the Clairton-Liberty Borough
area. In the region west of Elizabeth, emissions from both the Clairton
Coke Works and the Elrama Power Plant contribute to the high calculated
concentrations. In the Braddock area (see Figures 2-6 and 2-7),
emissions from the Edgar Thomson Works are responsible for the majority
of the calculated concentrations above the annual standard. In the
region between Homestead and Swissvale (see Figure 2-7), emissions from
the U. S. Steel Homestead facility account for most of the calculated
concentrations. Finally, emissions from the Jones and Laughlin facility
are responsible for most of the high concentrations calculated in the
region between Mt. Oliver and Homestead.
2.3 EXISTING PARTICULATE AIR QUALITY
2.3.1 Particulate Sampling Networks, Observed 1973 Annual
Geometric Mean and Maximum 24-Hour Concentrations
Figure 2-8 shows the locations of 33 hi-vol monitoring sites
in the Southwest Pennsylvania Intrastate AQCR at which particulate con-
centrations were measured during 1973. Two additional monitoring sites
located in Lawrence County are also shox^n in the figure. At most of the
monitoring sites, the air quality measurements consist of 24-hour
samples taken once every six days. At some of the hi-vol monitoring
2-30
-------
FIGURE 2-6. Calculated isopleths of annual average ground-level particu-
late concentration in micrograms per cubic meter for the
Clairton-Liberty Borough area.
2-31
-------
I
o
o
•H
6
c
•H
a
o
•H
0)
o
c
o
o
u
•H
cO •
CL. n)
0)
^H S-J
0) CO
3 CO
O M
M cq
60 I
-d
0) O
60 O
CO 5
S-l .H
0) QJ
> N
cO CO
EC
rH
CO 0)
G 4J
cO >-i
O
o
en a)
a)
a)
6
o. o
O -H
W J2
n) &
^n
d w
o S
rH 03
CO VJ
U 60
I—
I
Cxi
W
2-32
-------
sLAWRBflCE CO
BUTLER CO
ARMSTRONG CO.
BEAVER CO
INDIANA CO
12 l> 22
ALLEGHENY C0.23\ s|,
WESTMORELAND CO.
WASHINGTON CO.
33®
34
FAYETTE CO
GREENE CO.
50 5 10 15 20 KILOMETERS
FIGURE 2-8. Map showing the location of hi-vol monitoring sites in the
Southwest Pennsylvania Intrastate AQCR. The names of the
sites, the responsible agency and observed 1973 annual geo-
metric mean concentrations are given in Table 2-14.
2-33
-------
sites in Allegheny County, 24-hour samples are taken once every three
days. Table 2-14 lists, for each of the individual monitoring sites
corresponding to the numbers in Figure 2-8, the site name, the responsi-
ble agency, the site classification, the 1973 annual geometric mean
concentration, and the maximum 24-hour concentration in micrograms per
cubic meter. In the case of the Allegheny County hi-vol sites, the
maximum 24-hour concentrations refer to the period from July 1973 through
June 1974.
The site classification in Table 2-14, although somewhat arbi-
trary, is intended to isolate the monitors that are within a few kilometers
of large industrial sources from monitors located in urban (high popula-
tion density) and suburban (low population density) areas not subject to
the maximum impact of particulate emissions from large industrial
sources. Of the 35 monitoring sites in Table 2-14, 22 are classified as
industrial sites, 3 as urban sites (all in downtown Pittsburgh) and 10
as suburban sites. As might be expected, the highest observed annual
particulate concentrations occur at the industrial and urban sites.
With one exception (Springdale), the observed annual geometric mean
concentrations at these sites are above the National Annual Primary
Standard of 75 micrograms per cubic meter. At five of the industrial
monitoring sites (Bessemer, Midland, Baden, Braddock and North Braddock),
the observed air quality is approximately 2 to 3 times larger than the
National Annual Primary Standard. The lowest observed annual geometric
mean particulate concentrations occur at the suburban monitoring sites
and range from 56 to 102 micrograms per cubic meter. Of the 10 suburban
sites, 2 (Beaver Falls and Allegheny County Airport) show annual concen-
trations above the National Primary Standard, while 3 (Brighton/Beaver,
Greater Pittsburgh Airport, and West Brownsville) show annual concentra-
tions approximately equal to the National Primary Standard. At the re-
maining 5 suburban sites (North Fayette, South Fayette, Rostraver TWP,
Lower and Elco), the observed annual concentrations are below the National
Primary Standard. The National Annual Primary Standard was exceeded
during 1973 at all three urban sites (Downtown, Courthouse and Central Lab)
2-34
-------
TABLE 2-14
HI-VOL SAMPLING SITES IN THE SOUTHWEST PENNSYLVANIA INTRASTATE
AIR QUALITY CONTROL REGION AND SOUTHERN LAWRENCE COUNTY
Key
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
Site Name
Bessemer
New Castle
Koppel
Beaver Falls
Rochester
Vanport
Brighton/Beaver
Midland
Baden
Ambridge
Greater Pittsburgh
Airport
North Fayette
South Fayette
Bellevue T
Downtown T
Courthouse
Central Lab
Springdale
Logans Ferry T
Agency
DER1
DER
DER
DER
DER
DER
DER
DER
DER
DER
AC2
AC
AC
AC
AC
AC
AC
AC
AC
Classification
Industrial
Industrial
Industrial
Suburban
Industrial
Industrial
Suburban
Industrial
Industrial
Industrial
Suburban
Suburban
Suburban
Industrial
Urban
Urban
Urban
Industrial
Industrial
1973 Annual
Concentration
(fig/m3)
223-
129
99
83
93
99
75
214
152
104
70
54
65
103
105
157
126
71
77
Maximum
24 -Hour
o
Concentration
(Mg/m3)
695
334
212
210
224
259
233
621
289
237
312
129
207
260
311
305
327
233
201
Pennsylvania Department of Environmental Resources,
>
JAllegheny County Bureau of Air Pollution Control.
During 1973 for DER sites and during the period July 1973 through June 1974 for
AC sites.
2-35
-------
TABLE 2-14 (Continued)
Key
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
Site Name
Hazelwood T
Brad dock
North Braddock
Duquesne 1
Springdale T
Allegheny County
Airport
G las sport T
Coursin Hollow
Liberty Boro T
Clairton
Courtney
Rostraver TWP
Monessen
Lover
Elco
West Brownsville
Agency
AC
AC
AC
AC
AC
AC
AC
AC
AC
AC
DER
DER
DER
DER
DER
DER
Classification
Industrial
Industrial
Industrial
Industrial
Industrial
Suburban
Industrial
Industrial
Industrial
Industrial
Industrial
Suburban
Industrial
Suburban
Suburban
Suburban
1973 Annual
Concentration
(jL(g/m3)
94
173
153
116
94
102
83
13G
112
120
132
01
107
59
56
77
Maximum
24-Hour
Concentration
(Mg/m3)
402
604
496
273
285
395
325
482
452
286
410
157
336
131
230
148
2-1 6
-------
Annual particulate concentration measurements for the area im-
mediately west of the Southwest Pennsylvania AQCR were obtained from the
West Virginia Air Pollution Control Commission and the Youngstown Air
Quality Control Region through EPA Region III. The Ohio hi-vol measure-
ments for 1973 show a range of annual particulate concentrations from
a low of about 60 micrograms per cubic meter at the Youngstown, Ohio
Airport to a high of 175 micrograms per cubic meter at the No. 5 Fire
Station in Youngstown. The West Virginia hi-vol monitoring sites, most
of which are located along the Ohio River from New Manchester to Moundsville,
show a minimum annual particulate concentration at New Manchester of
75 micrograms per cubic meter and a maximum annual particulate concentra-
tion of 163 micrograms per cubic meter at Weirton. Because the Ohio River
Valley and the Youngstown area are heavily industrialized, most of the West
Virginia and Ohio hi-vol monitoring sites mentioned above are located
near large stationary particulate sources.
Rubin and Bloom (1975) have summarized long-term particulate
air quality trends within the Allegheny County portion of the Southwest
Pennsylvania Intrastate AQCR. Average particulate concentrations at all
monitors within the County reached a minimum during the summer of 1972
and achieved a maximum in the spring of 1974; average particulate concen-
trations within the County have decreased slightly since the spring of
1974. Rubin and Bloom note that the variation in the long-term parti-
culate air quality in Allegheny County qualitatively follows the vari-
ation in the steel production index. They used the Air Quality Display
Model (AQDM) to calculate 1972 annual average particulate concentrations
produced by the known industrial sources. From a comparison of calcu-
lated and observed concentrations, the annual average particulate back-
ground in Allegheny County was estimated to be about 45 micrograms per
cubic meter. Their analysis of 24-hour average particulate concentra-
tions observed during the period 1971 through 1972 at the two Allegheny
County monitors which are most remote from industrial sources (North and
South Fayette) also indicated an average particulate background of 45 to
2-37
-------
50 micrograms per cubic meter. As stated in Section 2.4 below, we also
estimate that there is a contribution to the observed annual average
concentrations at monitors within the Southwest Pennsylvania Intrastate
AQCR on the order of 50 micrograms per cubic meter that cannot reasonably
be attributed to emissions from industrial sources and/or classical area
sources in the AQCR.
2.3.2 Existing Short-Term Particulate Air Quality
The maximum 24-hour particulate concentrations listed in the
extreme right-hand column in Table 2-14 indicate that the National
Primary 24-Hour Particulate Standard of 260 micrograms per cubic meter
is equalled or exceeded at 16 of 22 industrial sites, at all three urban
sites and at 2 of the 10 suburban sites (Greater Pittsburgh Airport and
Allegheny County Airport). The Annual Primary Standard is exceeded at
a larger number of hi-vol monitoring sites and by larger percentages
than the 24-Hour Primary Standard. However, because hi-vol samples are
generally obtained only once every six days rather than on a continuous
24-hour basis, the frequency of occurrence of particulate concentrations
equal to or larger than the National 24-Hour Standard cannot be directly
ascertained. For this reason, it is difficult to reach a definite con-
clusion as to whether the 24-Hour Primary Standard or the Annual Primary
Standard is the more restrictive in terms of achieving the National
Primary Ambient Air Quality Standards. Rubin and Bloom (1975), fol-
lowing procedures suggested by Larsen (1971) that assume log-normal
distributions and using 1972 hi-vol measurements from Allegheny County,
obtained mixed results depending on the individual hi-vol measurement
site. Probably the fairest statement that can be made on this matter is
that both the Annual and 24-Hour National Primary Standards for particu-
lates are of concern in the Southwest Pennsylvania Intrastate AQCR be-
cause current air quality data show both to be exceeded at a majority
of the industrial monitoring sites, at all three urban sites and at two
suburban sites (Greater Pittsburgh Airport and Allegheny County Airport).
2-38
-------
In attempting to gain insight into some of the factors controlling
the short-term particulate air quality in the Southwest Pennsylvania
Intrastate AQCR, we selected nineteen 24-hour hi-vol sampling dates during
1973 for which concurrent measurements were available at 10 or more
monitoring sites operated by Allegheny County. On each sampling date, the
observed 24-hour particulate concentration at each monitoring site was
normalized by dividing by the 1973 annual geometric mean concentration
for that site to yield a particulate concentration index I. The purpose
of the normalization procedure is to reduce the observed 24-hour concen-
trations at all the monitoring sites to a relative basis and thus to
remove the confounding effect of the large range in absolute concentra-
tion values typically present throughout the monitoring network. On each
selected date, values if I for all monitoring sites were summed and divided
by the total number of sites to obtain an average index value of I.
Table 2-15 lists the average particulate concentration indexes
I in order of increasing values and the sampling dates, 24-hour mean surface
wind speeds, and total precipitation amounts for both the sampling
period and the previous 24-hour period. The meteorological data were
obtained from measurements made at the Greater Pittsburgh Airport.
According to Table 2-15, low values of the average index (relatively
low particulate concentrations at all monitoring sites) occur most fre-
quently in combination with mean wind speeds greater than 5 meters per
second and with precipitation occurring during the sampling period and/or
during the previous 24-hour period. Similarly, high values of the average
index (relatively high particulate concentrations at all monitoring sites)
occur most frequencly in conjunction with mean wind speeds less than
5 meters per second and in the absence of precipitation during the
24-hour sampling period as well as during the previous 24-hour period.
To illustrate the behavior of the particulate concentration in-
dex at the three classifications of hi-vol monitoring sites shown in
Table 2-14, we assigned the monitors to three groups as indicated in
2-39
-------
TABLE 2-15
AVERAGE PARTICULATE CONCENTRATION INDEX FOR ALLEGHENY
COUNTY, MEAN WIND SPEED AND PRECIPITATION DATA
FOR NINETEEN SELECTED 1973 CASES
Sample Date
(1973)
3-17
12-6
10-31
1-4
7-15
5-28
11-24
11-12
12-18
4-22
2-9
12-24
9-25
9-13
10-19
10-7
6-9
9-1
8-8
Pnrticulate
Concentration
Index (I)
0.43
0.58
0.65
0.68
0.71
0.76
0.93
0.94
0.95
0.98 '
1.04
1.06
1.11
1.11
1.26
1.33
1.34
1.46
1.64
Mean Wind
Speed
(m/sec)
9.8
6.7
4.8
8.7
4.2
5.2
5.1
4.4
3.5
6.2
4.3
3.9
4.4
3. 5
4.3
2.6
5.1
2.4
3.1
Precipitation (in. )
Sample Day
1.05
0.01
0.48
0.00
0.50
0.10
0.40
0.00
0.03
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
Prior Day
0.44
0.00
2.50
0.30
0.00
1.20
0.00
0.00
0.50
0.00
0.36
0.00
0.00
0.00
0.40
0.00
0.00
0.00
0.00
2-40
-------
Table 2-16. For seven of the sampling dates in Table 2-15, we calcu-
lated the average index I for each of the three groups and recorded the
number of monitors within each group at which the index value I was less
than unity and greater than or equal to unity. Results for the seven
dates are given in Table 2-17. The entries in the table show that when
the average particulate concentration index for all monitoring sites is
low (sample dates 3-17, 1-4 and 7-15), the index values for individual
monitors and the average index value for each group of monitors are also
low. Similarly, when the average particulate concentration idex for all
monitoring sites is high (sample dates 6-9, 9-1 and 8-8), the index
values for individual monitors and the average index value for each
group of monitors are correspondingly high.
Although the number of cases of index behavior investigated
is small, we believe there is a strong tendency for closely similar
relative 24-hour particulate concentrations to occur simultaneously at
all Allegheny County hi-vol monitoring sites. The exceptions that we
have noted on a given sampling date typically involve one or two moni-
toring sites. In many cases, the monitoring site is close to a single
large stationary source and the impact of source emissions at the moni-
tor is highly dependent on the wind direction. The Logans Ferry T hi-
vol monitor, for example, registers high index values and high concentra-
tions with moderate to strong west-soutlwest winds which transport the
stack emissions from the West Penn Power Plant to the monitor. In other
cases, anomalous index values occur at individual urban or suburban hi-
vol monitors that cannot easily be explained on the basis of known
industrial stack or fugitive emissions. To the extent that the results
of the index study described above are valid, they highlight the nece-
ssity for a careful and detailed consideration of meteorological factors
(wind speed, wind direction, preciptation) as well as source and emis-
sions factors in the interpretation of hi-vol measurements.
It is interesting that the sample index cases in Tables 2-15
and 2-17 indicate that the highest relative particulate concentrations
2-41
-------
TABLE 2-16
GROUPING OF HT-VOL MONITOR SITES
Group A
(Suburban)
Group B
(Urban)
Group C
(Industrial)
Greater Pittsburgh Airport
North Fayette
South Fayette
Downtown T
Courthouse
Central Lab
Cl aii-ton
Liberty Boro T
Coursin Hollow
North Braddock
Allegheny County Airport*
Bellevue T
Braddock
Hazehvood T
Glassport T
Duquesne 1
* Although Allegheny County Airport is classified as a suburban monitor in Table
2-14, it is included here in Group C because of its proximity to the other moni-
tors in the group.
2-42
-------
I
CM
W
CO
P-I
o
PS
O
W
w
CO
w
H
X O
w &
O M
"Z, hJ
H P
-J
CJ) W
¥, CO
o
o z
o
H CO
<2 pS
J O
5 H
U M
H O
p [ ^.^
O
W >
O I
3
w
PS
o
w
w
CQ
I— 1
Al
f— I
4J
• i— 1
W
0
4->
• i— I
C
£
o
• i— i
CO
o
3
§
0
CD
a
j-
i ^
X
CD
C
i — i
C!
O
~oJ
JH
O
O
0
O
01
be
OJ
f-i
0
£>
<^
O
ft
0
O
PQ
ft
o
O
«<
a
o
O
O
ft
o
O
PQ
a
o
o
<<
ft
0
O
o
ft
^3
O
J-(
O
PQ
3
o
O
-------
tend to occur with wind speeds below 5 meters per second. On the other
hand, the highest short-term ground-level SCL concentrations in Allegheny
County occur in conjunction with moderate to strong wind speeds (Cramer,
et al., 1975 and others). The S00 emissions are principally from tall
~"~ L
stacks and from a relatively small number of large stationary sources or
source complexes. There is thus the implication that a significant
fraction of the observed high particulate concentrations is caused by
emissions from sources other than tall stacks, possibly fugitive emissions
from industrial and other low-level sources. This matter is discussed
in more detail in Section 2.4 where observed particulate air quality at
various monitoring sites within the Southwest Pennsylvania Intrastate
AQCR is compared with the air quality at these sites calculated by means
of diffusion-modeling techniques using the updated emissions inventory.
2.4 COMPARISON OF CALCULATED AND OBSERVED 1973 ANNUAL AVERAGE PAR-
TICULATE CONCENTRATIONS
2.4.1 Presentation of Results
The results of the long-term diffusion-model calculations de-
scribed in Section 2.2 of the 1973 annual average particulate concentra-
tions at various hi-vol monitor sites are shown in Tables 2-18 through
2-20 together with the corresponding observed annual geometric mean con-
centrations. The annual average concentrations calculated by the model
are expected to be from 10 to 20 percent larger than the corresponding
geometric mean concentrations.
Table 2-18 lists the calculated arid observed particulate
concentrations for the three hi-vol monitor sites in the New Castle area
which, as shown in Table 2-14, are all classified as industrial sites.
At the Koppel and New Castle monitors, the calculated concentrations are
less than 10 micrograms per cubic meter which is less than 10 percent of
2-44
-------
TABLE 2-18
CALCULATED AND OBSERVED ANNUAL PARTICULATE
CONCENTRATIONS FOR THE NEW CASTLE AREA
Monitor
Koppel
New Castle
Bessemer
UTM Coordinates
X (m)
557,050
554,935
543,240
Y (m)
4,520,370
4,538,401
4,535, 965
Calculated
Concentration
(/-'g/m3)
5
8
341
Observed
Concentration
(Mg/m3)
99
129
223
2-45
-------
the observed concentrations. Both the Koppel and New Castle monitors
are located in industrialized areas, but are at least several kilometers
away from any major industrial source of particulates. If the calculated
concentration of 341 micrograms per cubic meter at the Bessemer monitor
is reduced by 20 percent to obtain an approximate geometric mean, the
adjusted calculated concentration of 273 micrograms per cubic meter is
in reasonably close agreement with the observed concentration. The
Bessemer monitor is only 200 meters from a large cement plant and the
observed particulate concentrations thus principally reflect emissions
from the plant.
Table 2-19 presents the calculated and observed concentrations
at hi-vol monitor sites located in the Beaver Valley area. The Koppel
site shown in Table 2-18 is repeated in this table. The calculated con-
centrations at all 8 hi-vol sites are substantially lower than the ob-
served concentrations. At four of the sites (Rochester, Koppel, Vanport
and Beaver Falls), the calculated concentrations are 10 micrograms per
cubic meter or less and are approximately one-tenth or less of the
observed concentration. At the four remaining sites (Ambridge, Baden,
Brighton/Beaver and Midland), the calculated concentrations are approxi-
mately one-half to one-third as large as the observed concentrations.
Table 2-20 lists the calculated and observed concentrations
for hi-vol moitor sites in the Allegheny County-Pittsburgh area. At six
sites (Bellevue T, Downtown T, Courthouse, Springdale, Springdate T and
Logans Ferry T), the calculated values range from 9 to 13 micrograms per
dubic meter while the observed concentrations range from 71 to 157
micrograms per cubic meter. At three sites (Central Lab, Hazelwood T
and Allegheny County Airport), the calculated concentrations are approxi-
mately 30 micrograms per cubic meter while the corresponding observed
concentrations range from 94 to 126 micrograms per cubic meter. At six
sites (Braddock, North Braddock, Duquesne 1, Glassport I, Coursin Hollow
and Liberty Boro T), there is generally close agreement between the cal-
culated and observed concentrations. At the Clairton hi-vol site, the
2-46
-------
TABLE 2-19
CALCULATED AND OBSERVED ANNUAL PARTTCULATE
CONCENTRATIONS FOR THE BEAVER VALLEY AREA
Monitor
Ambridge
Rochester
Koppel
Baden
Brighton/
Beaver
Midland
Van port
Beaver Falls
UTM Coordinates
X (m)
565,130
561,020
557,050
565,090
556, 720
546,300
556, 610
557,280
Y (m)
4,493,600
4,504,650
4,520,370
5,598,400
4,502,200
4,498,400
4,504,400
4, 513,450
Calculated
Concentration
36
10
5
73
43
60
8
8
Observed
Concentration
104
93
99
152
75
214
99
83
2-47
-------
TABLE 2-20
CALCULATED AND OBSERVED ANNUAL PARTICULATE
CONCENTRATIONS FOR THE PITTSBURGH AREA
Monitor
Bellevue T
Downtown T
Courthouse
Central Lab
Springdale
Logans Ferry T
Hazelwood T
Braddock
North Braddock
Duquesne 1
Springdale T
Allegheny County Airport
Glassport T
Coursin Hollow
Liberty Boro T
Clairton
UTM Coordinates
X (m)
579,738
585,143
585,060
588,167
602,976
605,167
589,667
596,536
596,643
597,786
603,020
591,119
594,298
596,512
596,284
594,869
Y (m)
4,482,286
4,476,019
4,476,738
4,476, 833
4,489,036
4,489, 107
4,473, 357
4,472,452
4,472, 833
4,469,464
4,487,830
4,467,214
4,463,369
4,463,348
4,464,238
4,461,869
Calculated
Concentration
(Mg/m3)
6
13
13
36
10
9
32
147
130
58
10
29
95
162
104
736*
Observed
Concentration
(Mg/m3)
103
105
157
126
71
77
94
173
153
116
94
102
83
136,
112
120
*Not representative because the monitor is too close to the Clairton Coke Works
for accurate model calculations.
2-48
-------
calculated concentration is approximately six times larger than the ob-
served value. The Clairton hi-vol monitor is located within 300 meters
of the coke oven batteries of the Clairton Coke Works and is thus very
sensitive to the quantity of low-level fugitive emissions assumed to be
discharged from the coke oven batteries as well as the modeling techniques
employed. As shown in Figure 2-6, there is a very strong concentration
gradient at the location of the Clairton monitor such that a few hundred
meters farther west from the Clairton Coke Works, the calculated concen-
trations are lower by more than a factor of two. Because of the modeling
techniques employed, we were unable to obtain a representative calculated
concentration for the Clairton hi-vol monitor. At distances of one
kilometer or more from source complexes, the calculation techniques
employed provide representative concentration estimates. A representa-
tive concentration estimate could be obtained for the Clairton hi-vol
monitor by redoing the calculations using somewhat different source
parameters for the coke oven batteries (see Section 6).
2.4.2 Discussion of Results
The long-term diffusion model and meteorological inputs used
to calculate the 1973 annual average calculation at the various hi-vol
monitors in the New Castle, Beaver Valley and Allegheny County-Pittsburgh
areas were also used previously to calculate 1973 annual average S0~
ground-level concentrations in Allegheny County (Cramer, et al., 1975).
Although there are fewer S0« monitoring sites than hi-vol sites, very
good agreement was obtained at all SO- monitoring sites between calculated
and observed SO,, concentrations. For this reason, we believe the modeling
techniques and meteorological inputs used in this study to calculate the
annual average particulate concentration are inherently capable of pro-
viding a fairly accurate representation of the impact of particulate
emissions from major stationary sources, assuming that the particulate
emissions data used in the calculations are essentially correct. It is
apparent from the comparisons of calculated and observed annual particulate
2-49
-------
concentrations that, except for seven industrial hi-vol sites (Bessemer,
Baden, Braddock, North Braddock, Duquesne 1, Coursin Hollow, and Liberty
Boro T), the calculated concentrations are 3 to 10 times lower than the
observed concentrations. These differences are much too large, we
believe, to be explained by deficiencies in modeling techniques or
meteorological inputs. When the estimated particulate emissions for
motor vehicles and domestic heating given in Tables 2-2 and 2-3 are used
in our long-term area source model, the maximum calculated contribution
of the combined emissions from these area sources is of the order of 1
microgram per cubic meter. Use of the semi-empirical Gifford and Hanna
(1973) ATDL area-source model with the same emissions from motor vehicles
and domestic heating yields a maximum annual ground-level concentration
of 8 micrograms per cubic meter contributed by motor vehicles and a
similar concentration of 5 micrograms per cubic meter contributed by
domestic heating. As noted below, we suspect the empirically-determined
coefficient A in the Gifford and Hanna ATDL model implicitly enhances
the contributions from motor vehicles and domestic heating.
We therefore conclude that the stack emissions from major
stationary sources and classical area source emissions are responsible
for only a relatively small fraction of the observed annual particulate
concentrations at all the urban, most suburban and some of the industrial
hi-vol sites. It appears that there is a large residual annual mean
particulate concentration of the order of 50 micrograms per cubic meter
in the observed concentrations that cannot reasonably be ascribed to the
combined particulate emissions from industrial sources and from classical
area sources such as domestic heating and motor vehicle emissions.
Rubin and Bloom (1975), who have also estimated a background
annual particulate concentration of about 50 micrograms per cubic meter
for Allegheny County, hypothesize that this is principally caused by the
advection of particulate emissions from large stationary sources located
in Ohio and West Virginia. Previous experience in applying our long-
2-50
-------
term area-source model shows that a particulate emission rate of the
order of 10 grams per second per square kilometer will produce an
annual average ground-level concentration of 50 micrograras per cubic
meter. In the simple ATDL dispersion model developed by Gifford and
Hanna (1973), the average ground-level concentration x is given by
(2-2)
where Q is the area-source emission rate in micrograms per second per
square meter, u is the mean wind speed in meters per second, and A is
an empirical constant set equal to 225 for particulates. For a mean wind
speed of 5 meters per second, the area-source emission rate required to
produce an average concentration of 50 micrograms per cubic meter is
approximately 1 gram per second per square kilometer. We suspect that the
success of the ATDL model may be attributed to the fact that the coefficient
A implicitly includes the effects of unquantified emissions and that, if
these unquantified emissions had been included in the estimates of Q,
the value of A would be reduced one order of magnitude.
2.5 SUMMARY OF WORK ACCOMPLISHED IN PHASE I; RECOMMENDED PHASE II
WORK OBJECTIVES AND MAJOR ACTIVITIES
2.5.1 Evaluation of Existing Emissions Data
As indicated above, the present emissions data, which are
almost entirely related to industrial sources of particulates, are
incomplete with respect to estimates of particle-size distributions and
of industrial fugitive emissions. Information is also lacking in a
number of other categories. Estimates of particle-size distributions
can be obtained by using the size distributions developed by EPA for
various types of sources. The information required to complete the
2-51
-------
other categories that are presently incomplete may possibly be obtained
from existing information sources. There is additionally a continuing
requirement to refine, update and correct the existing emissions inventory.
In terms of the major contract objective of developing an
emissions inventory that can be used in conjunction with diffusion
modeling techniques to make accurate projections of future air quality,
the most serious deficiency in the existing emissions inventory appears
to be with respect to the unidentified sources responsible for the
background particulate concentrations. As noted above, the present
emissions inventory including area-source emission estimates for motor
vehicles and domestic heating — when used in diffusion models — yields
annual average concentrations at urban and suburban monitoring sites not
directly impacted by emissions from large industrial sources that account
for only about 10 percent of the observed annual concentrations which
range from 60 to 100 micrograms per cubic meter. This argument leads to
the major conclusion of the Phase I study which is stated as follows:
The accuracy of the existing emissions inventory of industrial
stack and fugitive emissions can and should be improved. However, achi-
evement of a more accurate inventory of industrial particulate emissions
cannot by itself lead to the requisite improvement in the accuracy of
projections of future air quality because we believe industrial stack
emissions are not responsible for the major fraction of the observed
particulate concentrations which violate the National Primary Air Quality
Standards. The sources and/or processes primarily responsible for the
observed concentrations are at present unidentified and therefore cannot
be included in the emissions inventory. It follows that the primary
objective of the Phase II work should be the identification and quantifi-
cation of the sources and/or processes that are responsible for the major
mass fraction of the observed particulate concentrations.
2-52
-------
2.5.2 Evaluation o.f Existing Air Quality Data
Because of the problems that have surfaced in the past few
years relative to the achievement of the National Ambient Air Quality
Standards for particulates, it is apparent that there are a number of serious
deficiencies in existing air quality data. Some of these deficiencies,
such as the routine hi-vol sampling frequency of one sample every six
days, have been mentioned in Section 2.3. There are also questions as
to the required number of hi-vol monitoring sites and the optimum siting
of hi-vol monitors. Rubin and Bloom (1975) recommend, for example, an
expanded program of( ambient background monitoring in Western Pennsylvania
because the existing sites operated by Allegheny County at North Fayette
and South Fayette are not well situated to assess the advection of
particulates from major industrial sources in Ohio and West Virginia.
Additionally, there are requirements for continuous sampling of particulates
so that information can be obtained on the time history of particulate
concentrations within a 24-hour period. It is also clear that much more
attention must be given to the meteorological conditions that are
present when hi-vol samples are collected because particulate concentra-
tions have been shown to be very sensitive to wind direction, wind
speed, atmospheric stability and precipitation (see Section 2.3.2).
This meteorological information is also needed assist in identifying the
probable origins of the particulates.
In the analysis of hi-vol filters, little attention has been
given in the past to determination of the particle sizes and the species
that account for the major mass fraction of the material on the filters.
In order to assess the background concentration, it is essential to
identify the generic types and sizes of material that comprise the bulk
of the background material. The current separations of hi-vol collections
into organic and inorganic fractions and the elemental metal analysis
techniques employed are inadequate for this purpose.
2-53
-------
It is also important, to have information on the occurrence of
activities in the vicinity of hi-vol monitors, when samples are being
taken, that may cause local emissions of particulates. Example activities
of this type are highway and building construction or repairs, land
excavations, and agricultural activities such as plowing or harvesting.
2.5.3 Results of a Recent Study Sponsored by EPA of the
Suspended Particulate Problem in the Duwamish Basin
The results of a recent EPA-sponsored study conducted by the
Boeing Company in the Seattle, Washington area (Olsen, Almassy and
Wingert, 1975) are of special interest because they highlight many of
the problems associated with the assessment of background particulate
concentrations pointed out above and are supportive of many of the
conclusions we have reached in our Phase I study. These results also
point out possible sources of background concentrations. Olsen, et al. ,
summarize their study as follows:
"The Duwamish Basin area is a heavily concentrated industrial
region of South Seattle, Washington. Air quality data accumulated since
1965 indicate the primary and secondary national air quality standards
have been exceeded. Computer modeling studies predict only borderline
compliance with 1975 secondary standards. The objective of this study
was to determine the nature of suspended particulate in the Duwamish
Basin and subsequently quantify the impact of specific suspended particu-
late sources. A network of six sites each equipped with two high volume
air samplers (Hi-Vol) for simultaneous collection of particulate on fiber-
glass and membrane filters was operated from July through November of
1974. Results from gravimetric, elemental, and compound analyses were
combined with meteorological data for correlation and analysis. The
analytical approach toward identifying particulate types is a positive
step from simply making mass measurements. This approach leads toward
identification of airborne particulate, giving environmental agencies
2-54
-------
an opportunity to make assessments of size distribution, chemical elements,
and compound effects on public health and well being. Additionally, through
positive identification of particulate, positive corrective action can be
taken."
The results of this study indicated that the suspended particu-
late is extremely complex. Source tests revealed a multitude of elements
and chemical compounds present in the emissions from each source. To
further complicate matters, the soil in the area had become contaminated
by the emissions from industrial and area sources. Semiquantative analysis
of particulate by sources showed that about 27 percent was contributed
by natural sources (pollen and spores, wind erosion, open burning and
biologicals), 39 percent was contributed by vehicles and road dust, and
only 34 percent was contributed by industrial sources. It was concluded
that strategies for controlling suspended particulate concentrations must
be more encompassing than simply controlling industrial and vehicle emis-
sions; consideration must also be given to cleaning up roads and parking
lots as well as seeding barren areas because about 35 percent of parti-
culate is directly related to these sources.
2.5.4 Phase II Objectives
As stated in Section 2.5.1, the primary objective of Phase II
should be the identification and quanitification of the particulate
sources and/or processes responsible for the major mass fraction of
background particulate concentrations collected by hi-vol monitors.
Achievement of this objective requires that a major effort be made to
identify, through an analytical approach using optical microscopy, scan-
ning electron microscopy and dispersive X-ray techniques, the generic
types and sizes of particulates on hi-vol filters that comprise the major
mass fraction of the sample. We stress the use of hi-vol monitors rather
than specialized collection devices and the major mass fraction because
2-55
-------
the National Ambient Air Quality Standards are defined in terms of the
total weight of material collected on hi-vol filters. In order to re-
late the results of this type of hi-vol filter analysis to the sources
and/or processes of origin, it will be necessary to collect and analyze
concurrent meteorological data and other data pertaining to the existence
and nature of fugitive particulate emissions in the vicinity of the
monitoring site as well as emissions data from known stationary sources
that may be impacting the monitor site.
A second objective of Phase II is to complete and update the
existing emissions inventory for the Southwest Pennsylvania Intrastate
AQCR. For cost-effectiveness reasons, this effort does not involve
extensive stack testing or extensive measurements of fugitive emissions
from industrial sources. Instead, the principal emphasis should be
placed on the use of existing information, including the use of EPA
emissions factors and particle-size distribution criteria. We anticipate
that the bulk of the emissions data for industrial sources to be used
for updating the inventory can be obtained from the files of the Pennsylvania
DER and Allegheny County Bureau of Air Pollution Control.
A third objective of Phase II involves the upgrading of existing
diffusion-modeling techniques, including provision for the effects of
gravitational settling of large particles, and the use of the upgraded
modeling techniques to aid in the interpretation of the results of the
hi-vol analysis.
2.5.5 Description of Phase II Activities
Figure 2-9 shows the major activities required to achieve the
primary Phase II objective and other objectives. The plan for accomplish-
ing the Phase II activities depends heavily upon the participation and
cooperation of DER and Allegheny County since the plan calls for the use
2-56
-------
of their hi-vol sampling networks, meteorological data and the results
of their routine analyses of hi-vol filters. The major effort of Phase
II is the detailed anlaysis of hi-vol filters obtained from a 15-station
hi-vol sampling network. As shown in Figure 2-9, the initial step is
the acquisition and analysis of the first set of hi-vol samples from the
network. It is intended that this step serve to point out problem
areas, prove the analysis procedures, and provide basic information on
the types and sizes of particulates that are found on hi-vol filters in
the area. After several samples from the 15-station network have been
analyzed, the requirements for additional and special sampling should be
more easily specified. A requirement is shown in the figure for special
sampling of particulates from specific sources to aid in their identifica-
tion during the analysis of the hi-vol filters. Because the requirements
for special sampling to determine the effects of advection, meteorological
conditions and vehicle and people activities have not been defined, the
flow paths for these activities are not shown.
Hi-Vol Sampling Program
The hi-vol sampling program will utilize the existing particu-
late monitoring networks operated by DER and Allegheny County in the
Southwest Pennsylvania Intrastate AQCR. The names, locations and clas-
sifications of the 35 monitoring sites are given in Table 2-14. It is
logistically and economically impossible to make a detailed analysis and
evaluation of all hi-vol filters collected at all sites. It will
therefore be necessary to select a smaller number of monitor sites for
the filter analysis program. The selection should be made on the basis
of the site classification, site location and availability of meteooro-
logical measurements.
The hi-vol sampling program will be conducted in conjunction
with the control agencies' on-going sampling program. Hi-vol filters
2-58
-------
for the sampling program will be obtained during the same periods that
samples are obtained as part of the -control agencies' programs. It is
proposed that the hi-vol sampling equipment currently installed at the
selected monitor sites be used in the sampling program. In order to
obtain two simultaneous 24-hour samples at each site, there must be a
minimum of two operational hi-vol samplers at each site. This require-
ment may necessitate the procurement of some additional hi-vol samplers;
however, all of the Allegheny County monitor sites are currently equipped
with several hi-vol units. Hi-vol filter samples will be obtained using
the Gelman high-purity filter because this filter material is of suffici-
ent purity to allow optimum use of the detection limits of the various
methods of trace element analysis.
The handling of hi-vol filters obtained during the sampling
programs is of special concern. Care must be taken to eliminate the
possible loss of material during filter handling and shipment. Each
sample filter will be put into an individual container immediately on
removal from the hi-vol sampler head and sealed. The sealed containers
may then be handled and transported without the danger of losing material,
During the sampling program, a detailed sampler log will be
maintained. This log will contain all of the standard information
normally recorded for hi-vol samplers (i.e., dates flow rates, running
times, filter numbers, etc.). Additional information that will be
recorded in the sampler log includes:
• The meteorological conditions that existed at the
sampling site during the measurement period
• An estimate of the vehicle traffic in the area of the
sampling site during the measurement period
2-59
-------
• Unusual activities in the_vicinity of the sampling
site that occurred during the sample period (i.e.,
contstruction, excavation, agricultural activities,
etc.)
• Unusual industrial-source activities revealed by
visual observation of emissions
The above record-keeping requirements will necessitate periodic visits
to the sampling station throughout the 24-hour sampling periods. The
specific site locations and number of personnel available will determine
the exact times that visits should be made.
The meteorological measurements required for the analysis and
evaluation of the hi-vol samples obtained during a sampling period will
be obtained from various sources. The Allegheny County Bureau of Air
Pollution Control operates a number of meteorological monitoring sites
as part of their total monitoring program. Surface meteorological
measurements are also made at the Greater Pittsburgh and Allegheny
County Airports. These data can be readily obtained.
Hi-Vol Filter Analysis Procedures
The success of Phase II is heavily dependent upon a complete
and accurate description and analysis of the material that is collected
on the hi-vol filters during the sampling program described above. As
previously noted, several studies designed to explain some aspects of
particulate air quality have been conducted using various analysis
techniques to identify particulate types and origins. In general, these
studies have been confined to one type of analysis and have been directed
toward identifying particulates of specific sizes that originate from
specific sources. The filter analyses proposed for Phase II are intended
2-60
-------
to define the major mass fractions of the filter samples by generic
types; the analyses will include particulates ranging in size from about
5 micronmeters to the largest particles found on the hi-vol filters.
During Phase I, we investigated some of the analysis and
microscopy techniques that have been used in other particulate studies
in order to become more familiar with the strengths and weaknesses of
the various techniques that may be applicable to Phase IT. We found
that many previous studies have been inconclusive. Additionally, we
found that particle sizing and the identification of particles on hi-vol
filters through the use of optical and scanning electron microscopy are
not easy tasks. The accuracy of the identification of generic types of
particles is dependent upon the skill, bias and experience of the technician.
As part of our investigation, four commercial laboratories were invited
to identify the generic types of particles and estimate their contribution
to the total loading of sample hi-vol filters. Two laboratories responded
with an analysis report; one was unable to perform the work at that
time, and one declined to participate because it was noncompetitive in
this area. The filter samples that were supplied to the laboratories
were obtained from the Allegheny County Bureau of Air Pollution Control.
Table 2-21 lists the filter numbers and sampling sites.
The analyses of the filters by the two participating laboratories
are instructive in that, although each laboratory was supplied a portion
from the same filter (No. 28662), the results differed considerably.
The major particulates identified by both laboratories were of the same
generic type. However, the reported contributions of the various generic
types to the total mass differed. The reported generic types that
accounted for more than 95 percent of the total mass are listed in Table
2-22.
After discussing the discrepancies in the two laboratory
reports with several persons who have had related experience or who are
2-61
-------
TABLE 2-21
HI-VOL FILTERS SELECTED FOR LABORATORY ANALYSTS
Filter Number
Sampling Site
28662
28721
28709
28773
Liberty Boro T
North Braddock
Allegheny County Airport
North Braddock
TABLE 2-22
PERCENT BY MASS OF VARIOUS PARTICULATE TYPES REPORTED BY
TWO LABORATORIES FOR THE SAME HI-VOL FILTER SAMPLE
Particulate Type
Laboratory A Report
Laboratory B Report
Fly Ash
Soot
Iron Oxide
Kish
Starch
Biological
75-80%
10-20%
5%
40%
6%
50%
2-62
-------
currently involved in similar work requiring particulate identification,
it was found that the use of optical and scanning electron microscopy
frequently does not yield a positive identification of the particulates.
The Research Institute of the University of Utah has been
involved in several studies to identify and size particulate material
found in the atmosphere of the Southwestern United States. They use a
technique which combines scanning electron microscopy with energy-dis-
persive X-ray detection and analysis to identify particulates by shape,
size and elemental constituents. The accuracy of the identification of
generic types of particulates is improved by this technique because
particulates are identified on the basis of both their geometric character-
istics and their elemental composition. The deficiencies of this technique
are that energy-dispersive X-ray analysis does not identify elements
with atomic weights of 23 or less, only a very small portion of the hi-
vol filter sample can be analyzed because of the restricted size of the
sample chamber of the scanning electron microscope, and color cannot be
used as an identifier.
The filter analysis procedures proposed for the Phase II acti-
vities combine a number of analysis techniques and provide for cross-
checks and redundant analyses of individual filter samples. The proposed
filter analysis techniques are:
• Optical microscopy
• Scanning electron microscopy
• Energy-dispersive X-ray analysis
• Chemical elemental analysis
• Gravimetric analysis
Figure 2-10 is a flow chart of the proposed hi-vol filter
analysis procedures. The hi-vol sampling, shown at the upper center of
the diagram, is combined with a sample log that will include activities
and unusual conditions in the vicinity of the sampling sites during the
2-63
-------
METEOROLOGICAL
OBSERVATIONS
Wind Speed
Wind Direction
Precipitation
Temperature
Rawin
HI-YOL SAMPLING
III-VOL I-Jl.TKRS
FIELD FOR REDUNDANT
ANALYSIS
FILTER DATA BASE
Emissions
Meteorology
Sample Log
ANALYSIS RESULTS
Gravimetric 1
Gravimetric 2
Elemental Chemical (1)
SEM-EDAX (2)
Identification of Major
Particles
Optical 1
Optical 2
SEM-EDAX
Size Distribution
Optical 1
Optical 2
SEM-EDAX ANALYSIS
3 Sections Each Filter
Identify Specific Particles
Identify Particle Elemental
Composition
Elemental Analysis (2)
Total Specimen
OPTICAL ANALYSIS
3 Sections Each Filter
Major Particles
Identification
Size Distribution
SAMPLE LOG
Site Acti\ities
I nusual Conditions
GRA\ IME FRIC
ANALYSIS (2)
OPTICAL MICROSCOPY
3 Sections Each Filter
Major Particles
Identifii at ion
Size Distribution
SEM ANALYSIS
3 Sections F.ach Filter
Identih Specific Particles
FIGURE 2-10. Hi-vol filter analysis flow chart.
2-64
-------
collection periods. The detailed sample log is one of the most important
elements of the Phase II program. Two concurrent samples will be obtained
at each selected monitor site. As shown by the flow chart, one of the
hi-vol filters will go to the responsible control agency and one will go
to independent laboratories. The standard gravimetric analysis will be
performed on each filter and the results will be recorded in the Filter
Data Base. Portions of the filters will be retained after gravimetric
analysis for redundant analysis to test the accuracy and reliability of
the filter analysis procedures.
Lpdating of the Particulate
Emissions Inventory
The objective of this task is to continue the updating of the
particulate emissions inventory for the industrial sources in the South-
west Pennsylvania AQCR and to further upgrade the emissions inventory by
including estimates of emissions from:
• Mobile Sources: automobiles, trucks, aircraft,
railroads and ships
• Area Sources: redidential heating, commercial
heating and agriculture
• Fugitive Sources: industrial fugitive sources and
storage piles
The responsible control agencies are currently updating the
particulate emissions inventory for industrial sources as part of their
normal activities. As nei-r particulate emissions data are obtained or
developed by the control agencies, these data will be included in the
emissions inventory for use in the Phase II work. Additionally, the
current emissions inventory will be updated to include, if possible,
particulate size distributions and other requisite data.
2-65
-------
The upgrading of the current particulate emissions inventory
to include the mobile and area source emissions will be accomplished
using basic data obtained from federal, state, county and local agencies
and from the established EPA emissions factors. As early as possible in
Phase II, the following requests for data will be made:
• Pennsylvania Department of Transportation - Vehicle
mileage data for each county and the metropolitan areas;
take-off and landing cycles for the Greater Pittsburgh
and Allegheny County Airports; vessel traffic data for
the rivers in the area
• Pennsylvania Department of Taxation - Gasoline and diesel
fuel consumption data for each county and metropolitan
area
• Bureau of Census - Population data for each county and
metropolitan area
• Pennsylvania Department of Public Service - Public
utilities fuel distribution data
In addition to the above agencies, there are several trade or-
ganizations, both independent and government sponsored, that report on
industrial and commercial activities by product or industry type. These
organizations include the National Coal Institute, Market Statistics,
Inc., Fortune and Noyes Development Corporation. The information
available from these sources will be used to cross-check the basic data
obtained from the main sources and will also serve as reference material
for developing a fuel balance for the area.
2-66
-------
Upgrading of the Diffusion-Modeling
Techniques for Particulates
The diffusion models described in Appendix A and used in the
annual average particulate concentration calculations discussed in this
report do not account for the effects of gravitational settling. In
general, the effects of gravitational settlign are negligible for par-
ticulates with diameters less than about 20 micrometerss. For larger par-
ticulates, gravitational settling results in tilted plumes with the
plume axis inclined to the horizontal at an angle given by arctan V /u
s
(V is the gravitational settling velocity and u is the mean wind speed).
o
Also, larger particulates that come in direct contact with the ground
surface by the processes of turbulent mixing and gravitational settling
may be retained at the surface. Based on recent experiments at Dugway
Proving Ground, Utah (Boyle, et_ al_. , 1975), it appears that the reflec-
tion coefficient varies from zero (complete retention at the ground
surface to unity (no retention at the ground surface) as the particle size
decreases from about 100 to 10 micrometers or less for most ground surfaces,
However, values of the reflection coefficient for various particulate sizes
and ground surfaces are not: well established.
During Phase II, we will modify the short-term and long-term
diffusion models described in Appendix A by substituting revised Vertical
Terms in the model equations. The revised model equations will account
for the effects of gravitational settling and partial reflection at the
surface. To handle varying particulate sizes, the particulate size
distributions will be divided into categories, with a settling velocity
and reflection coefficient assigned to each category. Particulate
concentrations will be calculated for each category and will then be
summed.
As an example of the planned model updating, the Vertical Term
given by Equation (A-6) of Appendix A will be replaced by the
2-67
-------
following expression for the n particle-size category:
Vertical Term
E
a=0
, , 2a H - H + V x/u{H)
1 I m sn
a+1
exp
, , 2a H + H - V x/u{H)
11 m sn l J
a=l
2-1
1 , 2a H + H - V x/u{H)
1 / m sn l J
(2-3)
where
a-1
, / 2a H - H + V x/u{H}
_!_ I m sn
2"
4) = mass fraction of particulates in the n particle-size
category
y = reflection coefficient for particulates in the n particle-
size category (y is equal to unity for complete reflection
and zero for complete retention)
sn
settling velocity of particulates in the n particle-size
category
For convenience, 0 is defined to be equal to unity in Equation (2-3). We
point out that a potential problem with the application of Equation (2-3) in
the Southwest Pennsylvania AQCR is that the equation cannot be used with the
terrain-adjustment procedures described in Section A.5 of Appendix A with-
out violating mass continuity (see Section 6).
2-68
-------
SECTION 3
DESCRIPTION OF THE PHASE II PARTICIPATE SAMPLING PROGRAM
The Phase II particulate sampling program was conducted during
August and September 1976 at 15 monitor sites in the Southwest Pennsylvania
AQCR. Following the procedures outlined in Section 2.5.5, the samples were
obtained concurrently with routine 24-hour samples taken at the same sites
by the Allegheny County BAPC and the Commonwealth of Pennsylvania DER,
using colocated standard hi-vols. Details of the Phase II particulate
sampling program are given below and in Appendix D.
3.1 DESCRIPTION OF THE SAMPLER LOCATIONS
The 15 hi-vol monitoring sites used in the particulate sampling
program are listed in Table 3-1 and the site locations are shown in Figure
3-1. The site selection was made jointly by the H. E. Cramer Company
and the EPA Project Officer after consultation with the Pennsylvania DER
and the Allegheny County BAPC. The criteria used in site selection included,
in addition to logistical factors, the need for obtaining particulate
samples representative of industrial, urban, suburban and rural areas. The
only stations considered were those currently operated by DER and BAPC as
part of their regular monitoring programs for which a minimum of several
years' records of particulate measurements are available. The following
logistical factors were considered in selecting the monitoring sites for
Phase II:
• Availability of electrical power required to operate two
hi-vols simultaneously (one hi-vol was required for the
regular DER or BAPC filter sample and a second hi-vol was
used for the special sampling program)
• The requirement that each site be accessible to operating
personnel and observers on a 24-hour basis
3-1
-------
TABLE 3-1
LIST OF HI-VOL MONITORING SITES FOR PHASE II FILTER SAMPLES
Site Number
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
Name
Radon (COPAMS)*-
Reaver Falls (COPAMS)'
Koppel
Brighton To urns hip
Midland
FJleo
Dovnilcnvn
Central Lab
Ihizchvood *"*'
North Braddock (TIM)***
Duqucsnc 11
Liberty Boro (TlM)**"*'
CJ air ton
Creator Pittsburgh Airport
South F aye tic
Agency
HER
HER
DKK
DKK
DER
DKR
P.AC'P
BAC^P
BACP
BACP
BACP
BAC^P
liACP
BACP
P-ACP
Cl assific'ation
Industrial
Sul)urban
Industrial
Suburban
Industrial
Rural
Urban
Urban
Industrial
Industrial
Industrial
Industrial
Industrial
Rural
Rural
*COPAMS - Commomveallh of Pennsylvania Air Monitoring System (UER telemetered
data site)
**Hazel\vood T, a TM siU\' is located apj^roximatelv 0.5 kilometers south of the selec-
ted Hazelwood hi-vol site.
***TM - Allegheny County Bureau of Air Pollution telemetererl site
3-2
-------
LAWRENCE CO.
BUTLER CO
ARMSTRONG CO.
INDIANA CO
BEAVER CO
ALLEGHENY CO.
015
WESTMORELAND CO.
WASHINGTON CO.
FAYETTE CO.
GREENE CO.
5 0 5 10 15 20 KILOMETERS
FIGURE 3-1: Locations of Phase II hi-vol monitoring sites. The' numbers refer
to site names and other details given in Table 3-1.
3-3
-------
• Availability of meteorological observations (the COPAMS and
TM monitoring sites in Table 3-1 have meteorological
measurement capabilities)
• The requirement that the sites be grouped geographically so
that all monitors could be serviced and periodic activity
observations could be made at each site during each 24-hour
sampling period
The site classifications given in Table 3-1 are admittedly qualitative and
are intended only to serve as a basis for grouping monitor locations accord-
ing to the proximity of major industrial sources, population density and
traffic volume.
Figures D-2 through D-5 in Appendix D present 2-kilometer by 2-
kilometer topographic maps showing the major roads, railroads and buildings
in the areas surrounding the various monitoring sites. A brief description
of each monitor site is given in Section D.I of Appendix D.
3.2 METEOROLOGICAL CONDITIONS ON THE SIX HI-VOL SAMPLE DAYS
Meteorological conditions on the six sample days (10, 16, 22 and
28 August 1976; 3 and 9 September 1976) are summarized below.
Tables D-2 through D-7 in Appendix D list the Greater Pittsburgh
Airport hourly observations of wind speed, wind direction and ambient air
temperature as well as the Pasquill stability category determined according
to the Turner (1964) definitions for these days. Table D-8 gives the 24-
hour average wind speeds and precipitation data for the sample days, and
Table D-9 lists the mixing depths and vertical potential temperature
gradients obtained from the twice-daily upper-air soundings made at the
Greater Pittsburgh Airport before, during and after each sample day. The
3-4
-------
mixing depths in Table D-9 were estimated using the procedures outlined in
Section 3.3 of Cramer, et^ a_l. (1975); the potential temperature gradients
are the average values for the surface mixing layer.
10 August 1976
A slow-moving surface high pressure system passed over the Pittsburgh
area during the period 9-11 August 1976. Early morning ground fog cleared
at the Greater Pittsburgh Airport by about 1200 EST on 10 August 1976, but
broken middle clouds remained until about 1700 EST. Light-to-moderate
northwest winds persisted throughout most of the day, with the wind becoming
calm after about 2100 EST. Rawinsonde data for the Greater Pittsburgh
Airport show relatively deep mixing depths at 0700 and 1900 EST, with a
slightly stable potential temperature gradient at 0700 EST and an adiabatic
potential temperature gradient at 1900 EST. However, the 0700 EST sounding
on 11 August 1976 shows that, with clear skies, a strong surface-based
inversion formed during the night of 10 - 11 August 1976.
16 August 1976
A weak fionLal system passed over the Pittsburgh area on the morning
of 15 August 1976, with 12.2 millimeters (0.48 inches) of precipitation
recorded at the Greater Pittsburgh Airport for the 24-hour period ending
at midnight on 15 August. A surface hLgh pressure system followed the
front and remained over Pittsburgh on 16 August. Fog or haze was observed
during the period 0500 to 0800 EST on 16 August, and broken middle clouds
were observed during the period 1100 to 1600 EST. Winds were light-to-
moderate and from the west-northwest through north. Upper-air soundings
at tVie Greater Pittsburgh Airport indicate that:
3-5
-------
A shallow mixing layer wJLh a moderate sur fare-based
inversion developed during the night of J5 - 16 August
A relatively deep mixing layer with an adiabatic lapse
rate developed during the daylight hours on 16 August
A strong surface-based inversion with a corresponding
shallow mixing depth formed during the night of 16 -
17 August
22 August 1976
A surface high pressure system was approximately centered over the
Pittsburgh area on 22 August 1976. With the exception of scattered to
broken high cirrus clouds after 0800 EST, skies were clear in the Pitts-
burth area on 22 August. At the Greater Pittsburgh Airport, haze and
smoke were observed during the period 0100 to 0900 EST and haze was reported
after 1700 EST. During the period 1200 to 2100 EST, light-to-moderate north-
west and north winds were observed. The wind was calm or light and varia-
ble during the remaining hours of 22 August. Upper-air data for the
Greater Pittsburgh Airport show that strong surface-based inversions devel-
oped during the nights of 21 - 22 August and 22 - 23 August. The 1900 EST
sounding on 23 August shows a relatively deep mixing layer with a slightly
stable lapse rate.
3-6
-------
28 August 1976
A trace of precipitation was reported at the ('.renter Pittsburgh Air-
port on 27 August 1976 as a short-wave trough at 500 millibars passed over
the Pittsburgh area. Similarly, a trace of precipitation was measured on
28 August as an advancing frontal system induced light rain showers at about
2100 EST. Broken to overcast low clouds covered the Pittsburgh area after
0500 EST on 28 August, and haze or ground fog was reported throughout the
day. Upper-air soundings made at the Greater Pittsburgh Airport on 28
August indicate that the mixing depth varied from about 350 meters in the
early morning hours to about 700 meters in the afternoon. The vertical
potential temperature gradient within the surface mixing layer was approx-
imately equal to the moist-adiabatLc value during the period. Light-to-
moderate south or southwest winds persisted irom 01UO HST until about mid-
night when the wind shifted to the northwest, probably indicating a frontal
passage.
3 September, 197J3
A surface high pressure system, centered over the Great Lakes on the
morning of 2 September 1976, moved across the Pittsburgh area on 3 Sept-
ember. Calm winds and dense fog (visibility at or near zero) were ob-
served at the Greater Pittsburgh Airport during the period 0100 to 0900
EST on 3 September. After the fog cleared, light-to-moderate surface wind
speeds were measured, with the wind direction ranging from east through
south-southwest. Rawinsonde measurenipnts at the (".reitor Pittsburgh Airport
show a shallow mixing layer with a slightly stable lapse rate at 1900
EST. A surface-based inversion of moderate strength developed during
the night of 3-4 September 1976.
3-7
-------
9 September 1976
A weak cold front passed over the Greater Pittsburgh Airport at
about midnight on 9 September 1976. The storm resulted in 7.1 milli-
meters (0.28 inches) of precipitation at the airport for the 24-hour
period ending at midnight on 9 September. This precipitation occurred
in the form of light rain showers during the period 1900 to 2400 EST.
Skies were clear on 9 September until clouds in advance of the front
moved into the Pittsburgh area at about 1600 EST. Haze was reported
from 0600 EST until 2300 EST, when fog replaced haze as the major re-
striction to visibility. Winds were calm until moderate southwest winds
developed at about 0900 EST. After the frontal passage at about 2400 EST,
the wind direction shifted from southwest to northwest and the wind speed
increased to over 10 meters per second. According to the 0700 EST upper-
air sounding at the Greater Pittsburgh Airport, a strong-surface based
inversion formed during the night of 8-9 September. However, a
relatively deep surface mixing layer with an adiabatic lapse rate developed
by 1900 EST on 9 September and continued throughout the night of 9-10
September.
3.3 SAMPLING PROCEDURES
Except for a minor difference in the type of hi-vol filters and
some differences in the preparation and handling of the filters, the
procedures used to obtain the particulate samples were identical with the
procedures used routinely by BAPC and DER. Specifically, the filter
samples were collected by means of a standard hi-vols, colocated (i.e.
within 1 or 2 meters) with a BAPC or DER hi-vol at each monitor site, and
operated for 24 hours (midnight to midnight) on regularly-scheduled
BAPC and DER sample days which were on a 6-day cycle.
3-8
-------
Spectral grade (Gelraan No. 64948) glass hi-vol filters were used
for all of the particulate samples. The filters were preconditioned, weighed,
sequentially numbered and packaged in plastic in sets of seventeen unfolded
pairs by the Coors Spectro-Chemical Laboratory in Golden, Colorado. A set
of these filter pairs was shipped to Pittsburgh, Pennsylvania by air approxi-
mately two days prior to each of the scheduled sample days. At each monitor
site, on the day preceding the sample day, one of a pair of filters was
placed in the hi-vol and the other filter was retained in the plastic
container. On the day immediately following the sample day, the exposed
filter was removed from the hi-vol and the unexposed filter was removed
from the plastic container and placed on top of the exposed filter. The
two joined filters were returned to the plastic container and the container
was sealed. The DER and BAPC filters were usually installed or retrieved
several days before or after a sampling day. Also, the DER and BAPC
filters were folded prior to installation and on retrieval.
The filter pairs in the sealed plastic containers were collected
and returned by air, usually on the second day following each scheduled
sample day, to the Coors Laboratory. On arrival at Coors, the individual
filter pairs were logged, removed from the plastic containers, conditioned
and weighted. The total weight of the particulates on each filter pair
was obtained by subtracting the previously-determined weight of the un-
exposed filter pair from the weight of the exposed pair. The subsequent
procedures used by Coors to prepare filter samples for analysis by
optical microscopy and scanning electron microscopy are described in
Section 3.5.
Following a recommendation made in the Phase II program plan,
observer's logs were maintained for each filter sample at each monitor
site to record the performance of the hi-vol equipment as well as
meteorological conditions and activities observed in the vicinity of each
monitor site that might be of assistance in evaluating the measured
particulate concentrations (see Tables 3-2 through 3-7).
3-9
-------
I
n
W
rJ
^^
Q
VQ
r- X
C!-i -rH
c
H OJ
P d,
O < 1
<: Q
Q
co CD
W H
HCTl
H
Pi
o 6
PM O
rJ
O M-l
0
(U
CO O
Pi T)
w o
r> r-l
Pi p.
W QJ
pq ^_^
O
CO
•^
>_,
d
E:
CJ
Cd
rn
_
I_J
1)
u
CO
o
T
d
iJ C i-"*
COS
O — 1 ~-^
'J 4_l tD
c: ,-i
CJ
^ f:
O '"
r-l CJ
Uj d
^
-C~! ^-*.
10 cr
•r-l Ei:
D * — '
^i
11
to VJ
'11 CL
-C CO
n
S~i T3
U 01
•H d
en
M
i)
•^
rn
CJ
> I 0
-—I U 4-J
U ^}
•-1 cj . H
wi o
3 tt O
4-J d — i
M
'if1 o eo
f j fy^ r^
•M H
•H a u
TJ O C!
C S-J -r-l
U) U-i t^
JZ CU
en
<1J a) >,
rH > d
•-I -H S-j
a. 4_i a,
•H C/j
CJ • 60
U r-H 3
O tr) U-i
n M
^3 O ^3 -H
•-]- ^ CQ 4_l
^ 1 OJ -1-
-J r-l (I,
-d d r-i a.
d TJ d o
23 CQ U-i 1-^
4J
en
0)
^ci d
jj f •
3 >-j
O O
en x
^
VD
5
ca H
1
r-l
U
4-J U
C f-
•H 3
1 — 1
-d a,
o
u
3 • r-l
T3 C eu
O 3 >
S-l r-J -H
p. V-i
V-i
60 3 to
a o c;
•H ,C O
(-1 rH
•H -, -a a)
4J aj a
rH 4-1 Ul
3 a. -H
d 3 rH
U-i V-i d
c
cr>
CN
5
--T
r^-
1 — 1
, — 1
CM
O
o
CXI
cr,
O
, — r
• — i
"O
C
rH
•a
sj
T3
1)
' — 1
rH
n)
4-J
cn
C!
•H
4-J
'U • — 1
^N d
r
4-J J-j
O O
^ . I
1 _'
_
o
1 —1
m
~~r
• — i
— <
1 0
r— i
o
. — (
1 0
. — 1
. — 1
r*
o
4-J
0 C
0 •:
-H O
uj a
s visi- .
[^
o
•r-l
en
'-0
•H
r;
a)
^
4J
VJ
d
a)
^
rj
1'
D.
O
j
^
•-)
4J
CO
3
T)
£
O
•H
4J
U
-J
j-i
4J
en
O
U •
V4
rH O
d -a
O 0
Q
rJ 0
•H
'-M
eu d
• — 1 i-l
-£J H
lining i
d
•4-1
en
o
r^
3
rH
a.
4J
C
d
, — H
a,
tu
.^
o
o
_i
1^3
r-l
north- i
E3
0
V-l
U-l
-a
c;
•H
1 "
•
en
's,
u
d
4-J
en
d
CD
C
•
en
-------
CJ
3
d
o
CJ
H
(/)
"yi
d
E
CJ
[V{
f-0
-
u
0)
)>
u
1)
en
43
O
1
d
V-i x — •
d 0 E=
CJ -,H •~-^.
CJ 4J to
d i
O '•^
x—
O 'if
•-I U
u
Tn ,?[,
•H t:
q.
Vj rj
O) C,
43 CO
E
3
M -a
r— 1 • — 1
•r-l d
pj LI
w
rH
D
-H
CO
•
(J
T)
JJ 71
•Jl u
1/1
M 3
rd 'TJ
D
0)
wi >
d oi
•r-l f — |
£
d " '^
d M o
>-M O rH
T) D.
o) o a) o
^- fj r^
30) W l/l
-H rH 01
D. 43 3 PH
d u* o
HO) 3 cq
U O Q
1 -H If;
'11 i-J CO O
CO O CO M
^ '"' ^ 'w
^r; , i
CO r^>
, ,
O 0
• •
CTI tn
C 1 O
r - - t
— i C 1
in r-~
CJ CM
cr. cr.
O O
~vT ^O
(M r-J
CTI a-.
0 O
. — 1 • — 1
.-'•i CJ
u d
o w
,d 13 n
u -a 3
M rd CJ*
O Vj 3 M
•/-, CQ OH
u
'Jl M
3 0)
-a >
D
W) rH
d
rH O
rH r-t
•H
J-4 d
T3 -H
>
4_l Vj
d rH
0) 1
C co
0) CO •
r> *") Tj
rd 3
a, o
. , — i
rH 0) CJ
o u
Q V_| JJ
4T 3 W
CJ O 3
co '.n "3
vD
^O
, i
0
•
ON)
^-f
f*l
r >
en
cr.
O
o
m
O
' '
, — 1
^
4J
\4
1) O
^ 0
t-J Pi
1
M O
C ^ 0
•H U -H
[/) d -IJ
H 0) O
OJ 3 ^',
D. C7"
•J)
X) d i-H
d a)
U) >
a) -u ai
6 W rH
3 3
-H Q T)
a, c:
3
d • o
o
0 0
VD O-
o o
CT. CTi
O 0
.1 .1
IJ ^J
M LJ
O 4". 4-J
Ci. 4-J 01
VJ 3 >,
•H O d
-3 CO tu
3-11
-------
CO
1
CO
w
rJ
3
H
i —
en
r-H
H
CO
p
O
H
-.n
.r-l £
-^
• "
01
en vj
u d
.Q tr)
3
r-i 13
11 U
PL, 4-1
CO
d
OJ
• r-l
CO
or
r;
•H
.C
rn
•H
d
•H
n
•H
n3
*.
d
•H
r)
^i
^_i
D •
ij ^
IT)
J-> 13
OJ
S on
d
- -r-l
^iP S-I
0 D
Cu -3
X
^o
Mfl
CJ
•^
.d
in
•H
fj
•H
f^
•r-l
13
••
c^
•r-l
"J
^J
P
01 •
i-J '-' .
(^
4-J "^3
0)
?: M
£
•* -r-|
tn s-j
O 3
rU 13
CO
C-J
-ct
on
d'
•H d
r! .r-l
01
•r— i tn
d D
•r-l r-H
C tj
•H \* t
13 U
v_l
- 0
d -U
—1 O
1$ X<
^_J
^J
01 -
J-J ^-,
TO
J-J 13
OJ
f 00
d
** T-i
cn M
O 3
Uj T3
vD
, — l
i.n
•
v^
a
r-H
a.
6
C^J
tn
!>
b
r-H
OJ
-Q
r^i
OJ
-H
, — |
rt
00
d
•H
rc3
tn
•H
^
•H
G
•r-l
13
*•
CJ
•H
c3
^
^i
0) •
J-> >1
(T)
4-J -T3
0)
2 00
d
r, -r-l
to '>-<
0 3
in
vD
O
m
•-H
^ n u
d 3
•H S-i
-d cj
t-H
•rH !/l
3 • 1J
rD Vj -^
D O
4J i — 1 G
d ex en
i> 6
u n)
rt tn co
• r— ) U
13 cn -n
cr) rt 3
13
^ i — I
d" D r-|
•r-l > D
'-M GJ >
o — < u
O -1
r-l 11
1 d S
U D 0
Pi CO -H
i'^i
CO
ro
•
'J
•r-J
!.__)
'i_j
rl
S^
ij
jj
_rri
'.n
•H
^
o
m
.
r-l
r]
G
u
0
raking a
CO
D
E-
3
r-l
P.
fcO
d
•r-l
n
u
d
a
3
C/r*
,_]
3
11
to
rJ
O
ex
CO
•H
13
o
-U
0)
G
•r-l
4-)
60
rJ
O
r-l
i furnace
'^/
u
rW
»•
CO
0)
'^*
o
E
M
d
•H
^
o
a
J
Li
QJ
l-l
O
13
O
•
4J
d
a
13
•r-(
^>
QJ
a
'^
o
e
o
d
i_i
3
O
J3
1
m
. — i
•
D
r-t
-D
rt
CJ
o
•r-<
4_)
o
o
O
O
o
in
r J
O
OJ
OJ
O~)
rli
o
. — 1
en
, — i
OJ
CO
cr*
O
, — i
d
QJ
-3
a)
ca
OJ
OJ
rj
CO
o^
o
• . — 1
. — 1
CO
a-.
0
1 — j
V-i
D m
;> , — i
Cfl 1— I
o) ca
CQ P' i
^
Ol
og
co
CTl
o
. — 1
OJ
CO
O-i
0
r— 1
• 1
0)
D,
a.
o
* /,
^0
OJ
OJ
co
cr.
O
, — i
in
OJ
CO
en
O
• — 1
d
o
4J
j^
"on
•^ D,
V-l ^5
CO E—l
CO
O J
C 1
CO
Q-\
0
, — 1
OJ
CO
m
0
r— 1
X)
d
m
i-H
13
•H
i 1
-t
,—I
0-1
O
o
r-H
cj
cr.
o
CO
G>
o
O)
co
O
oj
'O
o
OJ
co
m
O
o
•j
rt
VJ
4-1
d rO
O rJ
OJ
O
r-j
en
O
O
oj
co
-a
o
o
0)
N
3-12
-------
a
•H
4J
c
o
u
I
01
en
-«
d
G
O)
c£
en
*~
o
>-*
M
•j
en
O
" 1
d
1-1 ,—,
u c m
COG
U 4-1 C.O
C . 1
U
,~x
O ew
r-l CJ
tu d
r 0 c,0
•H n
T~
01
en j-.
u d
O) C
,0 'JO
3
M 13
0) O)
iJ , — (
r- 1 t-J
•H d
tj-l I-1
en
a
u
• H
t/}
,
4-1
c;
a
13
•H
0)
W
CU
^
^
^-1
cx
r-H
d .id
F: ej
>-* d
O 4-1
'' c/1
CT- in
r - --t
'-'
0 O
CO O^
-1- ^
i n • — '
O O
cN r-j
r*-} (-^i
o> en
O O
n in
0 O
t>j 04
co m
cr> cn
O O
• — i . — i
4d 0)
ej C
o en
-d -a o)
4J -a 3
>j d a*
O Sj d n
;•; M a M
-
Ol
el) E
ro cu
3 >
o d
42 a.
Ml
C --I
•H O
4J O
rn r;
0) o
H-: en
G d
0 O
S-i
'4-J ^.'d
i-V
ai o
o
e
en
4J >,
0) G 4-J
E^ d en
o T-I 3
c/) o, ~ej
i
^J
O
.
g-,
CO
CO
O
, — I
C^J
ro
CTl
O
CT>
O
rj
ro
a^
0
, — i
^•^
4_l
tJ
o o
43 M
•H O
hJ CQ
0
en
(II t-H -3
4-i d t2
•-I d
,c - en
S -H C Ol
i ai o ^
JZ > ••-> O
en ai en p
-H rH en en
3 -H
4D C CU 0 ^j ^]
T) O M r-l
C M c CD a o
d W) -H >, V-i (J
42 d 3
>-, vj w - 4J u
d d 3 0) en T.
u cu CM rH T-I -.-I
so c; 42 o o
d G G
- en • eu
4d a) e/i o
CJ g 4J -r-l J-l OJ CJ
d d en 4J o N N
> — i ' — i z3 o nJ d d
ca CL, -d c o ~ "
in i~- ^-C
o o o
a-> oo o
in ^t 'O
m a~ "Z.
C I ~~t
r-l O O
-H ri -vt
r-j cj rj
ro ^^~) O
en o~> cr»
O O O
>—< en en
—i r j rn
CM oi eg
ro co ^
CJi O"i CT^
O O O
t-M . — 1 • — t
c;
O 4-10)
4-J 1-J 4-1
VJ O ,G 4-1
•H n. J-J cu
d Vj 3 ;>-,
^H -H 0 d
U <; e/i Ci,
3-13
-------
p
.J
.Cl
r*** *^
O1 C-j
rH 0)
a.
H P-
CO <
o a
OJ
OJ -H
OJ I
a
(U
w ^
cq
O
— •
&
-Q
• — '
I
Vj ^-~-
4-J C ~~)
C O E
rj ,,-j ~^,
'J 4J ^0
CJ
^
^ C
O '-M
^H U
EX, a
tz
i/i u!
r-l d
QJ C
JS 00
3
Vj "O
01 OJ
I — 1 — J
•H d
f r | \ 1
'Jl
r;
-j
4-1
•H
00
•Jl
11
I*
3
-H
CX
tr
d
•H
d
a
Cj
11
3
tr
13
c:
d
wi
d
-H
r1^
0
u
, ]
c^3
H1
•-C
-^
CO
in
CO
OJ
~^j-
r-)
o~>
01
O
m
vO
OJ
CO
Ol
O
, — t
£
o:
•a
tc
co
• 4J
4_J _C
d NJ
0) JrJ -H
T3 O i-J
•H • 3
^> J^ i^J
GJ 4-J 4J .
3 XI
'/I O O (X,
i — I 'yi ^'j
rj d
> 0) -H
OJ JT-
rH 4-J TJ r-4
QJ OJ
L-: O !fl -H •
O 4J O N O
rH rH d 'H
M) U X U4
P< C '4-J
O o o) d
pq ,-1 v) • Vj
4J 3 U 4J
» 'Jl 0 -H
>' M 0) '4-i -H
ctj o r-i d d
QJ T3 d r-l (-I
,c o ; • 4J 4j
ii
1 — 1
CO
;^-
CO
, — ;
O-l
VD
Ol
rn
a-.
o
r-l
VD
Oi
n
CTl
O
. — i
i_j
(U 71
> rH
Cl* • — f
OJ PJ
cq U-,
O 0
d is
4-J
3
-O "
P-I
-U
d d
QJ -H
•o
• r-V i-J
> 0)
OJ -H
N
rH d U
rH £ -rH
OJ '4-J
G '*-!
'/i - d
01 M
4J 0) 4J
rH rH
d ja ^
^ -H U
D- 'fl 3
'Jl -H Vj
< > -U
01
, — 1
o
in
c
m
r^j
ro
ro
04
fl
a^
o
-T
o")
OJ
O")
CTt
o
, — 1
,™{
01
Oj
CL,
O
^
rH
d
4-J
•H
CX
'Jl
0
r-|
a
o
•_^
'H
QJ
-^
O
E
'Jl
•^
U
d
, — i
n
QJ
'Jl
d
QJ
Q
o~l
T~~i
in
m
m
CO
CN
t ^
ro
OA
O
, — i
. — i
VD
O")
01
O
. — i
• — i
• — i
d
o
4-1
r^
t~-C
• H
yj
M
-0
d i-J
o d
-- 1 OJ
Jl d
_o
H OJ
0) 0
0) E
—I r-| VI
r; rx o)
-< E > S
d "J o •
d m -H ,H -3
•H —id
O ;: QJ H
Vj -H o .'--J i;
O 4J rH
4-J d 4-J
d 4-J QJ • Ul
VJ 'Jl rH 4-1 0)
QJ .n d ^ rH
d >•: -rH -J d
•H d u TI r
U "3 ^' '— 1 • r-l
d -j > LU o
•H -j- u QJ en x
C CO
1 ~~
o
•H
4-J CTi O
d
4-J r | --J-
'ji -
OJ
QJ
•^
O
e
•d
rH QJ
d 1-1
E j
O i-3
^ -J
O CO
• — i r--
Ol — i
m o
en co
•x G--
vC ^
<3" ^^
^O O
Lj""! VD
04 OJ.
n n
CTv o\
O O
m 01
m m
OJ 0-1
on n
O^ Ol
o o
-H t — t
§ rH
o d
4-J r-l
C 4-1
3 d
O CJ
Q U
,
QJ
, — 1
o
d
CJ
o
•H
4-J
O
d
n
d
J
3-14
-------
, — *
01
c
•H
o
u
"i
on
W
f— 4
9
H
en
'—•
d
b
CJ
erf
eo
""
M
CJ
1-1
CJ
XI
o
i-i C on
C O E
CJ -H ---,
o 4-j •£
C ,-:.
O ' — '
5 e.
r
-— '
i! ^
f-O yi
•--< r
r~
'j
in i-i
M d
CJ O
X) LO
M "a
cj 'j
i_j —1
r— | r-J
•H T
U-. J-J
71
r:
i — t
u
j~>
•H
to
^•' V
o rH
d - — i x)
r-l CJ -H
Xl ^- pi '/)
01 li -r-1
-• — 1 O >
T-] ^_J
cj v xj en
1-1 o a)
r-i __« fi
en >, d -H
C > M p<
O d O
•^ aJ a)
'Jl X CO
L/0 * Cj
•H r-l d
d d M
QJ 'Jl CO
OJ • >-i |
>, ^ « o a
> O C! C ^
d E o O
a) uj — i HP
.d e/i u xj
>•, to i
t-J d -M 00 Tj
i-3 (-1 C; to C
•-J M QJ ^J d
c^ r-^
vc r~
01 C]
0 O
en Ln
r j OD
' ' d^
vc m
CO -T
m ^r
cj ri
on o-i
CTi O"^
o - o
in ^
C4 Cl
o") on
en en
o o
< — i . — i
O J-i
o a
U o
' — t Xj T3
(U iJ Til
N VJ d
d o i-j
i; >: P3
.j
--H CO
XJ C
- -H -H
cj co x:
iJ -rJ (.J
•H ^> CIi
C (1)
^2 OJ p
-if! CT
O
• n
r-H en co
d C
d 3 -H
MO .^
O ^H O
c; t— t o
OJ
a) x c to
£ 00
10 CJ 4-J O
0) M MM
3d -r-t 4-1
cr p. d en
a o en
1 r"l 1 M
to d to o
tO M t/) T3
LJ CO t3 O
VO C-)
~-t r^-
Cj r-l
O O
— H CTi
• O-»
O 0
^H , — i
0)
C "-,
en 4J
(D M
3 QJ O
cr xj M
3 M -HO
a M j pa
^
'j
—- 1
' j
rj
u
en 'n
d
O j-i
•H ^
(0
en o
•H • \
B IJJ
0) 'J-j
•-^
cj) u
•H
-^ "J
O r-l
U U
• H
,-i -H x: >,
d d cj d
fl R > -rj
MM d
O 0 0 d
x, ?: >", to
- i o i - -r
C v^3 1^
- — • — r — •
0 O 0
O r- en
vD -,
•H -H O d
CJ <-(, in UH
3-15
-------
LO
1
pa
rJ
3
H
r^
CTl
rH
H
CO
[ 1
o
00
OJ
><
W
to
pa
o
p
X
*iH
£
OJ
p
p
d
•H
P">
1
a
•—i
-D
n)
H
E
O
S-l
U-l
0
c.
•c
o
c
0)
Pi
V)
— •*
5-4
[-J
f:
QJ
«;
en
"~
5-i
QJ
K>
U
CJ
in
jo
O
1
i-> C ~~i
d O E
CJ j i *vfi
Q v, ,.
U
r— .
3 £
O LM
-H -J
[^ O
v~~'
•LJ
— - •" v
^ L tJj
•H t-
1) > — '
'U
i-j ro
"j a
3
a) a
U r-\
i-H rH
-H c*3
[X, u
(H
QJ
4_J
•H
CO
i^' T 1
o o d
rH LJ d
^
^ 01 to cj
O -u O 3
cq d Uj 5-i
S.< 4-1
Q)
• X) • 0
>, o o d
^> l*~' T-l
tfl " ' >1J
Q) u_< -a
X • d QJ
-i-is-i en QJ
bo d -U ON
C CJ rH d
•H -J d OK
,C -H -H
o > d QJ
d QJ u en •
u i-> 30
3m o -H
CT r-| ^ > -C '-M
CJ > Q) Q) U-i
j > d N >-i d
•-j QJ D d nj 5-1
^ -H j:: .Cl •_•-. 4-J
r- -t
O ^O
r i •—
OJ —t
•r) Ct) rH
d QJ d
cq ca [-u
aj
N
d
a;
*
o
•H
'H
'-H
d
5-1
AJ
V?
U
3
(-1
u
' >
^
f£)
Q
^
..--1
r^>
^_!
°i
I —
vD
in
r I
r-^
CO
rl
ro
cn
O
i~H
VD
CO
01
n
CTi
O
rH
QJ
n-
a.
o
^
0) T)
1 — ' M
d
0 0
rH -U
" ' •-)
tn ~-t
• O -LJ
a; e u
N 3 1)
d rH 5-1
IX! O-, -H
-a
QJ
^ fo d
Ol -r-l -H
o en QJ
1J 3 5-1 N
d u Q) d
d u ex ~
CO
vC
r i
cr.
3
^,\
i -
-T
c-~>
CO
OJ
CO
CTi
O
OJ
CO
OJ
f>
CTi
0
r-4
~~)
b
d
rH
D-. Tj
2 -H
H ^
;j
N
d
*T^
•
O
•H
'-H
'H
d
5-1
4-1
.H
'ej t^i
r' ^>
U d
O 1)
-
— i
_. ,--•,
--J
o o
r j r^
. ' ~~i
^ ,",
^-,
i~-~~ r~~*
0 Ci
•1 r-J
vD ro
O CTi
r-l O
vO vO
O O^
ro OJ
^£> n
O CTi
—i O
. — I , — 1
d
^1
o
4J
O d
eJ 2
rH 0
LJ Q
en
— i
QJ
^,
a
rH
^>
O
, — (
.
" -ri
en o
o o
o
^ X
en N
d
<7j
5-j d
to
t ]
"3
QJ -H
Dd bO
0
— '
. — i
O
CO
r-]
<^-
LO
CJ
ro
CTi
0
, — i
n
LA
OJ
n
en
0
. — 1
rH
d
5_j
U
d ri
Q) d
U J
>%
d
J_g
M
-TJ
d
d
C
3
O
V_j
X3
M
^
O
d
rH
r»
rJ
^•J
^
I
r---
O
c-J
• — i
LO
,
cn
O
n
\£)
O
. — i
CO
o
rn
MD
O
r— 1
"0
o
o
~2
r—j
CJ
N
d
X
QJ
N
d
"
.
•/)
, — i
QJ
^>
QJ
r>
O
rH
C
"3
o
J_J
23
.
cn
D
^
o
n
3-16
-------
•a
0)
3
c
•H
4-J
LA
I
fcj
I-J
E-
r-£
O
r
oi
co
OJ
•*
^
tn
>Q
O
" r
n
COS
oj -H "-^
u 4j tn
c -i
O ^—^
u
3 f-
O '-.-1
r-l U
U-1 • J
^
Cfl tf!
| 5
OJ
t/l U
^-* d
o a
,o en
F:
a) o
4-J • — t
r-< rH
•H d
p^ u
V)
C
r— I
a
u
H
CO
en
d
o
•H
CO CJ
I/I CO
•H ,~
E iJ
a)
t-H a)
OJ -rH
QJ d
rH Q)
[> •
O CX OJ
•-< O N
J2 d
"J in x
0)
o •
H 03 rH
uJ cd
1 e !:
CO O ^
m M o
P
m
.
\D
vi^
CO
a-v
OI
ro
Oi
O
CO
CJ"
CN
'CTv
O
'J
o
-C -r)
4J '3
rJ d
O )-i
•,'; cq
t
0)
— I
a.
e
d
tn
a
•rH
4-J
d
4J
tn
S>i
d
u
^i
u
•H
4J
4J
CO
CC
VD
. — I
O
vO
o
""^
o
0
n
VD
o
, — 1
, — 1
c
CO
a)
3
tr
3 t-
Q r-
1 0 .
-d N fjc
o d c
C 'J^ -iH
QJ CO
3 CO
or • -H
*-> e
13 c;
C OJ r4
d C a)
•H 4J
fcC E r-l
r; o -H
•H >-J 'J-l
^ D,
O '-M
!J X O
S-j
a ai vj
0 > 0)
•U C
rH tn rJ
•H l-i O
d o cj
rH U
vj 0 rH
1 rH
to f-.p d
was
LJ -H C/5
CD
"^
» — i
O
.
a-*
ro
c
rxi
0
l^Q
O
t — 1
1 — 1
o
n
vD
O
, — 1
, — )
*^-t
4J
V-j
a) o
rQ r4
H -HO
H _] 03
V-l
^ U
O ,d
r" 1 i-)
>! ill
> a) w
d 60 O
a) c: 4J .
rC d OJ
M 4-1 N
>i o c; d
-H i oj n:
i-H ,d 13
d to -H
3 -H > .
CO C 0) CO
3 :? oj
COTE
3 rJ C 3
ca o rH
c; -H a.
O CO
4J • rn ^
r4 W -H U
•^ C B d
d o 'U u
rH -H CO
fj (n rH
1 V) D rS
CO 'H > 4-J
00 r, oj -H
U .1) rH 5
1 — '
VD
O
C7i
in
^-f
in
O
OT
vO
O
— t
1—1
O
P~J
vo
O
, — l
, — t
C
o
J_J
^1
•iH
d
rH
t_>
u
a)
O
O
o
O
S-l
O
a.
i-J
J3 -U
-U OJ
3 >,
O n)
00 t4
3-17
-------
W
r-l
H
P-i
W
en
O
O
LO
—x
^
v>
3
•o
QJ
CL
4-
rH U
u, ^
i_)
"*0 00
•r-l [__
,° s~'
"J
LO lj
M ~
cj rx
_c e/5
C
3
VJ 13
CJ U
4J • — (
• — 1 > — 1
•rH rj
(ii 4J
in
1 — (
1)
iJ
•H
00
4-J "J 1
i n. >-i 7i
I-I CJ rJ --H
3 c/) d U
10 CO
"J f-) ^,-J-
^ ~~*~ • — i £2
GO — < QJ 6
in d > >-i
d 'rH •• CO '*-<
;>, M f- rH
3 U >,
4_J x! 3 3 d
O O O 3
-H
DO d CO d) l~*
r! .d _d to 'M •
•H p^ 4J S-l -H QJ
,-^i en d r-i N
r4 d O rH TJ d
d 4J n:
CX J^ 1— I 4J
•u in i-j 3
1J -H J-l 1-3 ,Q
d 3 QJ >-i
CJ 4-J - CO
•J -d QJ • en -H
d QJ £ d C a
•r-l O QJ O d
-d d o V4 -H d
,
~T3)
cd
QJ
u
m
vO
c-i
r— J
,
OJ
— '
^f
ro
vD
O
. — l
. — i
m
d
QJ
cq
4J
QJ
U
d o
1 -H
K^l '1 1
!/* ' | [
6 d
S-{
J^ 4-1
1 1
•iH Ai
3 a
3
t>0 S-i
C 4-J
•rH
4J ^-,
4J -rJ
3 d
O 1)
4J
CX CO
d
^
a •
in -d
O
£> o
en 4j
^t
CO
. — t
B
t
in
in
r-
i — i
CO
^-j-
CO
VD
O
. — 1
. — l
f-^.
^j-
o-i
vO
O
rH
, 1
OT QJ
r-l D.
rH D,
d o
u. ^
b :•-
-H i — i n
-H rH
CJ -H IM
>i H -H .
M
• — 4 n, 10 4-J
O -H d CJ
!^' ^ t — 1
CO -u IP Q,
— i in CO
•H V4
;-: iJ 4-j TJ
O Q -u
rH .d O ~J
a. -H
co n en u
to o d
J-J S-l M
t~) '4J S-i SI-1
rH CJ GO
CO I1 l-i to CJ
— i > d S-i
rH .n -rH O VJ 3
d -rH 4-1 'fl CO 4J
ti 3 t,n ~3 —l -H
o >-i 3 d -HO
;: u '-i d L, r?
^ CC --T
OD r i i^
in ^t in
• • •
CJ
rH
.,
O
,J
oo
0 1
O
, — i
in
0 ]
o
^j-
rn
vD
O
, — i
, — i
cn
m
m
•^rj
0
•— f
• — 1
C
^
O
4_)
d
|>
o
Q
•>
O
.
CJ
-H
n
•H
'-0
•H
^>
en
CJ
e
3
rH
CX
.
CJ QJ
OO N
d d
d ~
O rH
-------
0)
3
d
•H
4-1
in
^vj
P
-i
E
0)
rrf
en
P
CJ
P
0)
c/)
b
rj
cog
o -H ---
r J 4J * r
C :i
0
3 f:
O '4-4
—I U
'"-'
u
-I™ ' — •
to Ml
•H e
-
n ]
cn P
CJ Q
jo co
n
D
V.
1) O
*-J • — i
T— 1 ^H
' J
7)
C
i-J
00
• r*
-', '-0
r-H 4-J
,-H tn
rd 3
U T3
•H
J-J i — 4
P 0)
lU >
> QJ
r-l OJ
fin N
C ^ CT3
•H O ffi
C/J - — 1
•H
i-i CJ
C P
y) wo
0) QJ "d
[:' • 30
3 v-i cr
rH O 3 Vl
a. ^j • Q aJ
O CJ | IM
nJ N CO r-t
v) o a co 3
^ >; n^ ^j m
a^ o o
r ) --.T in
in m o
• • •
O O ro
in in \r
in CT' ^ \
vC C^ O
CM OJ O]
VD vD o]
r 1 —i 01
n ro co
vO vO vD
0 O 0
' ' ' ' ' '
,o ,n ,
r-l o-| P|
vO vO vO
0 O O
i 1 r— H 1—4
. — 1 — I i-H
O ^ OJ
o a c
o to
-i -r: "U a)
0) 4J XI 3
N P d cr
m OP 3 ,
K >-, m Q ^
.
i)
, — i
^
•H
7)
•H
J>
71
OJ
e
3
r~l
(H,
c^
o
4-J
^-1
•H
T)
! j
O
1
c/)
CO
I^J
r.^
[^ —
0
»
o
"*
^>
. — .
. — i
o
OO
O~)
•-D
O
' '
CD-,
ro
vO
O
, 1
r— (
X
4-J
}~l
D 0
•H O
j m
C 'rJ
o 3
-H ^j U 4
wi HO
in 13
•rH P • JZ!
G Cfl 0) 4-1
1) ^ — 1 3
0 ,fl O
• — 1 4-J C\j 01
OJ QJ
> 4-1 O 1)
O) t/l -rl r-l
^H OJ -LJ -H
3 o a
^J Cj
O OJ 00
f^ -C p n3
-u O i— 1
C T3 -0
o o o
4J 4-J O
•H on •
rt r» P w) P • •
>H -H a) G O —1 •-!
u J-J — i -H u d t^
1 U--I O. > '-I T; f^
00 -i-l f: •-! C P P
C/J P rf P O O O
^j xi w "d C ;' s'
c r ^ cr
cr cc r^
ui 10 o
• • •
VD CO —I
in -~t Ln
r^ — '
•H D, -LI a)
d P 3 >^
i-H — I O rt
u <; co LU
3-19
-------
I
01
PJ
H
p
•H
^o i:
cn Q.
rH D
D
w
PQ d
w
H LO
W 1
CO O
CT* Q,
PS rQ
0 cd
o e
0 0
CO
i:
Crf OJ
LiJ CJ
PS 13
W 0
CO r-l
cq p-
0 U
PH
'•/I
v«
^j
d
e
0)
erf
t/1
_
CJ
u
01
'Jl
o
d
u d 01
d O E
a -H \
d . i
O ' —
U
^— .
5 c
O v-i
rH U
r- , — ,
cn M
•H C:
• —
1)
r; -yj
I) O)
1 \ T— <
d
C/)
•n o
CJ l '
o
-H tn •
Cn rJ 0)
cn 01 N
•H u d
lb CiJ .Xj
W G
IH O «
O O rJ
CO r-l 01
rH
' — I tij? CX
01 d n
> -H cd
o) d w
' 1 rW
3 U-J
^? rQ O
O
rH • • .C _£
CJ O 'Jl iJ
i-J N N d >-"
oj cd d v-i o
W3 3^ ^ H d
1 - 0 | >£
01 O-J — '
p-^ cn ol
in O in
^f -
r^ ro r- 1
rj O CO
CN OJ CJ
C^ f-") f-,-)
O^> C7> O>
O O O
I — i 1 — 1 - — I
~-i a> f-^
r-~ \£> vD
CN CN] CM
ro ro r~>
O~i CTi O>
o o o
I — t 1 — 1 . — 1
Lj i — 1
d -II en o)
01 > — 1 0,
U Cd rH CX
d a) d o
CQ a in u;
•
i — i
d
^
S-l
o
CO
VD
MD
^
co
' — )
^
cn
OJ
cn
O
i — t
, — i
r-l
CO
OJ
cn
o
. — 1
. — 1
— ^
d
u
n
•H O.
Vj ' :
m H
*>
6 i
rH O '.1 1
-" r 1 (!
"T.I [J-, .-) rt
C CQ C./1 H '/I
cd d -.I
~ -H --J f=
u d o i-j
d -H rH IJ 'XJ
-i-J ^ X3 t
cn O 3 >,
o ra r-i rd
jJ d P. ^
— 1 ., -H cd
Q- w :•; -i
'" d d cr
^ 0 •- • CJ d
rH .H .,, V., .J .._, .
3 in d O r •/) CJ
fc C/) O —i J rJ N
•H >-i ex ,d -1) d
0) E J-J H U D, 3:
rH 01 cn d w
.Q '71 13 -H
•H 'H W 'J -r)
CJ 0) ij 13 -H U
r) > O Jj rH U 1)
>-i Ol -r) d — 1 3 — 1
u —) o £ --•; ,-Q a.
-------
n
e
CJ
01
>
u
T3
01
3
G
•H
4-1
G
O
U
td
rJ
CO
3
O
tx
E
.
E
3
O)
-U
I—I
•F«(
PlH
<
• — 1
f^ •
p: a)
M N
o rt
>: :d
CO <~ I
-t -t
r 1 01
O O
r^ n
— 1
— 1 -H
.X 0)
CJ G
C CO
.G TJ OJ
4J TJ 3
LI co' cr
O L< 3 M
X CQ Q M
1
n
0
o
Li
OJ O
rH T3
,a o
•H
C/) M
•H C
> 0
0) -U
0 00
3
rH
a. •
rH
G -H
0 -H
4J JZ
\-l
-H M
rt a)
• — ) t>
U 0
1
CO bO
CO G
LJ -H
co
r-i
O
O
^
c
r--
CO
t"O
VD
O
. — i
00
CO
ro
O
i — r
ij
v^i
oi o
j3 i-i
-H O
rJ M
•
G
o
•H
in
en
0)
C/}
C
•H
rH
O
0
^
CJ
CO
•
0)
N
rt
M
n.
rH 3, '/)
rH l-i
n) TJ D
0) "3
en u ca
• -^ C Li
CO U 1) CJ
01 d -H
-G iJ !^-i •
en en o -LJ
3 en
a, « en CD
on a) 01
G C fi .G
0) -H 3 4-J
O) -G rH M
M a Cu O
6C C G
0)
G 3 • 0
o cr o) .u
4J >
rJ - -H rH
•H t-0 4-J rH
a! G 'J -H
rH -H Cd -G
O MJ
M >i S-i
CO rd M O
co f:
y o > o
r i
U~l
VD
i.n
G^
r-i
CO
n
v£)
O
, — 1
rH
CO
C~l
O
. — i
. — i
C
o
4.J
}-l
•H
CD
rH
U
rJ
OI • 4-1
l-l 5-1 CO
''J
'U N
G U-I rH 3
•H O _ ~
SO ,G O
G -U rH •
•H Ll £
-V O 'J-i O
Li C O Li
O >i-j
^ O en OJ
a.j -a (-1
tn 30
LI eo o G
CJ >-J rH 4-1
N O) O
O 4-J i-l
•3 01 'U rH
G on 3
13 Ll !/]
G O r) Oi
3 in H.J, vj
G
O
OJ
CT
3
O
O
O
G
O
•H
CO
CO
OJ
en
o
o
CJ
en
O
03
o
O
ri
v£)
O
-.1- O
CTl rH
CN rn
n vo
O-i O
O -H
(-1
O
a.
.G
J_l
3
o
CO
3-21
-------
3.4 SUMMARY OF FILTER SAMPLES, WEIGHTS AND PARTICULATE CONCENTRATIONS
Tables 3-2 through 3-7 which are reproduced from Tables D-10
through D-15 of Appendix D list the monitor sites, filter numbers, total
weights, 24-hour average particulate concentrations and observer's remarks
for each sample day. Of the total of 96 filter samples obtained during
the program, four samples were considered questionable because of equipment
malfunctions or related problems and one sample was considered questionable
because of the very samll weight of collected particulates. Excluding
these five questionable samples, the lowest 24-hour average particulate
concentration of 42 yg/m was measured at the Beaver Falls monitor on
10 August 1976. On each of the other five sample days, the lowest measured
concentration occurred at the Elco monitor site (no sample was obtained
at the Elco monitor on 10 August 1976). The highest concentration of 404
3
yg/m was measured at the Midland monitor site on 9 September 1976. The
highest concentration measured on each of the other sample days occurred
at a different monitor site. However, the highest concentration on a
given sample day, as might be expected, occurred at an industrial monitor
site.
3.5 SAMPLE PREPARATION
The hi-vol filters obtained during the particulate sampling program
were analyzed by both optical microscopy techniques and scanning electron
microscopy (SEM) techniques. The optical microscopy analysis was done
by the Coors Spectro-Chemical Laboratory and the SEM analysis was done
by the Environmental Studies Laboratory of the University of Utah Research
Institute. An optical analysis of six filters was performed by Walter C.
McCrone Associates, Inc. to provide an overall check on the accuracy of
the Coors analysis. The preparation of all filter samples as well as the
shipment of the samples for SEM analysis to the University of Utah Research
Institute was the responsibility of the Coors Spectro-Chemical Laboratory.
3-22
-------
H. E. Cramer Company personnel collected the filters from the
fifteen hi-vol monitor sites tin the day after the sample day. As described
in Section 3.3, a clean cover filter was,placed on top of each exposed
filter immediately on removal frori.the hi-vol to prevent any losses of
collected material and/or possible-'contamination during shipment. The
filter packets, each consisting of an exposed filter and a cover filter,
were then placed in heavy-duty, self-sealing plastic envelopes and shipped
by air to the Coors Laboratory. Each filter and each cover filter was
pre-assigned a seven-digit identifying number.
The first steps in the preparation of the samples by Coors involved
the conditioning and careful weighing of the filter packets to determine
the total mass of particulates collected at each monitor site. Next, five
3.8-cm diameter circular sections were cut from along the diagonal of each
filter packet. The locations of these five sections on the filter are
indicated in Figure 3-2. The circular specimens were cut while the cover
filter was still protecting the primary filter. This was done to avoid
dislodging any particles from the filter during the cutting process. A
0.64-cm diameter circular sample was then cut from the central 3.8-cm
diameter circle. The primary and cover filter sections of this smaller
sample were carefully separated and each mounted on a 0.64-cm diameter
SEM stub with a thin film of Duco cement. The samples were next coated
uniformly with gold to achieve an adequate ground for SEM scanning and
to protect the samples from loss of particles in shipment. Most of the
filter samples were shipped in this manner to the University of Utah
Research Institute. However, the final set was left intact as cover
plus primary filter and shipped between two glass slides. Both the cover
filter and primary filter of this final shipment were mounted at the SEM
laboratory. The 3.8-cm diameter circular sections minus the 0.64-cm
diameter centers were used for the optical microscopy analysis.
3-23
-------
1119032
FIGURE 3-2. Location of 3.8-cm Diameter Circular
Sections on Hi-Vol Filter Packet.
3-24
-------
3.6 OPTICAL MICROSCOPY ANALYSIS PROCEDURES
Both the optical microscopy and SEM techniques were used to deter-
mine the percent by mass and percent by number distributions of the particles
on the filters. The data from both techniques were also used in a detailed
analysis of the participates collected on the hi-vol filters on three
sample days: 16 August, 28 August and 9 September 1976. These three
sample days were selected from all the sample days for detailed analysis
principally on the basis of meteorological conditions. The other sample
days were undesirable for detailed analysis since the winds were generally
light and variable with calms reported for many of the hours. The optical
microscopy procedures used to analyze the fifteen filters on each of the
three sample days are described below.
The principal instrument used for the optical analysis of the
filters was the polarizing microscope. It was found by experimentation
on several filters that the best method of obtaining samples for optical
analysis was to peel off the top surface of the filter and examine the
particles found on it. Actual removal of the particles from the filter
xvas attempted using hot Aroclor. However, this technique was found to
be unsatisfactory. In the technique that was selected, the top layer of
the exposed filter was carefully peeled off and placed on a cover slide.
* >v
The glass fibers in the filter were made invisible with an immersion oil
with an index of refraction of 1.52. The glass slide was then examined
under the polarizing microscope and the particles were counted and sized.
The circular sample was also examined under the microscope to determine if
any particles remained on the second layer of filter material. If some
particles were present on the second layer of filter, this layer was care-
fully peeled off and prepared for observation in the same manner as the
first. The procedure was repeated until all the particles in a localized
2
area of 0.00968 cm were counted. Essentially all of the particles were
concentrated in the first two filter layers on most of the filters (Coors
Spectro-Chemical Laboratories, 1976a).
Refractive index oil made by Cargille.
3-25
-------
The above sample-preparation technique used in the optical
microscopy studies resulted in the counting and sizing of particles
embedded in the filter as well as those on the surface. In the SEM
analysis, only particles on the surface of the filter were examined. It
was noted that a few of the larger particles, which were identified as
CaCCL agglomerates, disintegrated in the immersion oil (Coors Spectro-
Chemical Laboratories, 1976b). Only a small percentage of the particles
appeared to be of this type and only a few of these actually broke into
smaller particles in the immersion oil.
Particles on the filter were classified according to size and
physical characteristics and the number in each category found on the
sample was counted. Particles were examined with both transmitted and
reflected light using various polarizations. This allowed observation
of the particles' morphology, transparency, color and other physical
properties which aided in determining generic types.
Coors Spectro-Chemical Laboratories separated the particulates
into the following six major categories (McCrone, et_ _al., 1967): flyash,
nonmagnetic iron oxide, magnetic iron oxide, quartz, biological particles,
and calcium carbonate. The magnetic iron oxide particles were
separated from the nonmagnetic iron oxide particles by the application
of a magnetic field.
The flyash category consists of many different types of particles
and, in this study, most of the particles in this category appeared to have
been formed by combustion processes. Observations of flyash (White, 1977)
show a wide range of sizes from about 0.1 ym to 300 urn. The particle shapes
are varied and include hollow translucent or transparent spheres, flakes,
irregular opaque particles, fused agglomerates of fine particles or small
spheres, and large, porous, partially burned gritty particles. The gritty
particles result from incomplete combustion of powdered coal and range
3-26
-------
in size from about 10 to 300 pm.
The number of particles in each category were separated into five
size ranges: 5-10 pm, 10-20 pm, 20-50 pm, 50-75 pm and > 75 pm. These
particle sizes represent the coarse mode of the traditional bimodal
distribution and thus provide no information on the fine particle mode.
The size of spherical particles was determined by measuring the particle
diameters with a calibrated occular on the optical microscope. The sizing
of the irregular particles was accomplished by using Martin's diameter
(McCrone and Delley, 1973).
The particle count data in each category and size range were
analyzed to determine a calculated filter weight and mass distribution
by size and category. The number of particles in each category and each
size range were used to calculate an average particle volume. For the
size range which has a minimum diameter x, and a maximum diameter x~, the
average particle volume V is (Herdan, 1960)
TT/6 X dx
Xl (3-1)
X2
dx
Xl
The volume mean diameter D was calculated from the expression
1/3
D
VM
3-27
-------
Calculated values of DTri, as well as the arithmetic mean diameter D.,, are
VM AM
listed in Table 3-8 for each size range. The volume of all the particles
in the i category and the j size range, V.., was calculated from the
expression
V. . =
N. .
D.
VM
(3-3)
where D refers to the j
. th
size range. These volumes were calculated for
.th
each size range within the i " category and summed to obtain the total volume
of particles in that category. The percent by mass within each size
range was then determined. Next, a density was assigned to each category
and the mass of particulates in the observed filter area was calculated.
The following densities were assigned to the categories used for the
optical microscopy analysis (DeNevers, 1977 and Kraus, et_ al_. , 1936):
3 3
1.5 g/cm for flyash, 4.9 g/cm for magnetic and nonmagnetic iron oxide,
33 3
2.5 g/cm for quartz, 1.0 g/cm for biological particles, and 2.7 g/cm for
calcium carbonate. By summing the masses in each category, we obtained
2
the total mass of particles on a standard filter area (0.00968 cm
for the optical microscopy examination) for each monitor site and sample
day. Assuming a uniform distribution on the filter (Coors Spectro-Chemical
Laboratory, 1976a), the total mass of particles on the filter was calculated
by dividing the mass on the standard filter area by the ratio of the
2
standard area to the total filter area of 423 cm . The percent
by mass and percent by number in each size range and category were also
determined for the total filter area.
The optical microscopy data for all fifteen monitor locations on
the three selected sample days were analyzed. The particle counts for the
standard area, the percent by number, overall percent by mass, the percent
by mass within each category, and the calculated total filter mass are
presented in Tables D-16 through D-74 in Appendix D of this report. Where
3-28
-------
TABLE 3-8
AVERAGE PARTICLE DIAMETERS FOR EACH SIZE RANGE
Size Range
(ym)
D
AM
(ym)
D,
VN
(ym)
5-10
10 - 20
20 - 50
50 - 75
>75
7.75
15
35
62.5
87.5
7.77
15.54
37.02
63.32
88.09
3-29
-------
available, the analysis of the cover filters is also included. Similar
results obtained from the combined analysis of the primary filter and the
corresponding cover filter for those cases in which the data are available
are presented in Tables D-75 through D-88 of Appendix D. Table 3-9, which is
reproduced from Table D-16 in Appendix D, is an example of the type of
information obtained from the optical analysis of primary filters.
3.7 SCANNING ELECTRON MICROSCOPY ANALYSIS PROCEDURES
The following procedures were used in the scanning electron
microscopy (SEM) analysis. Pairs of photomicrographs were taken of
selected areas on the sample stub at different magnifications. The
generic particle size distributions were determined from these photomicro-
graphs. Also, on some of the filters, certain particles were selected
and examined with energy dispersive X-ray analysis (EDAX) techniques
to determine their elemental composition. These particles were labeled
on the photomicrographs so that both the physical description (shape,
size and texture) of the particle and its elemental composition could be
utilized in identifying the particle. The particle size ranges used in the
SEM analysis were the same as those used in the optical analysis (5-10 urn,
10-20 ym, 20-50 urn, 50-75 urn and >75 ym). Particles were grouped in
three generic categories: spherical, irregular and agglomerate. The
agglomerate classification included lacy or frothy particles formed at
high temperatures as well as traditional agglomerates composed of accre-
tions of small particles. The generic size distributions were determined
by counting the particles on pairs of microphotographs made at various
magnifications which depended on the particle size range. A magnification
of 500X was used for the 5-10 ym range and 10-20 ym range; a magnification
of 200X was used for particles in the 20-50 ym size range; and a magnification
of 100X was used for the 50-75 ym and >75 ym sizes. All of the counts
were converted to a standard are;
seen at a magnification of 100X.
2
were converted to a standard area of 0.02053 cm which is twice the area
3-30
-------
TABLE 3-9
RESULTS OF OPTICAL PARTICLE COUNTS
Fi It.or Nuinlx-r: 101)3219 Total Massol Part icula le.s (mu;):
Coors: 361 Agency: 139
Monitor Location: Baden Calculated Mass (n;;): 2G1
Date of Sample: 8/16/76 Air Concent ni M' MI (Pf;/111' ):
LXX_TS: 80 Agency: 67
Category
Flva.sh
5-lOu
10-20y
20-50(1
50-75U
>75;j
Totals
Iron Oxide
5-10n
10-20y
20-50)i
50-75).!
>75y
Totals
Magnetic
Iron
5-10y
10-20)]
20-50(i
50-75)j
>75}J
Totals
Particle
Count
300
78
23
3
4
408
34
11
0
0
0
45
31
10
3
0
0
44
Percent By
Number
52.3
13.6
4.0
0.5
0.7
71.1
"
5.9
1.9
0
0
0
7.8
5.4
1.7
0.5
0
0
7.7
Percent By Vviss
Al 1
Cn te^' '!" i >'.s
1.8
3.8
15.3
10.0
35.9
6G . 9
0.7
1.8
0
0
0
2.4
0.6
KG
(3 . 5
0
0
8.8
Within
Cate^oiy
2.8
5.7
22.9
14.9
53.6
100
27.9
72.1
0
0
0
100
7.1
18.4
74.5
0
0
100
3-31
-------
TABLE 3-9 (Continued)
Fi 1 (or Number: 1093219 ((out iiin^l)
Catcgoiy
Quarts
5-l()u
10-20)1
20-50|j
50-75ji
>75u
Totals
Biological
5-lOu
10-20u
20-50)j
50-75u
>75tj
Totals
CaOO,
, i
5-10iJ
10-20u
20-50u
50-75y
>75)j
Totals
lUl'AI S
Count
15
1
9
1
0
26
11
10
3
0
0
2^1
18
6
3
0
0
27
574
(All Categories)
1
Percent By
Number
2.6
0.2
1.6
0.2
0
4.5
1.9
1.7
0.5
0
0
4.2
3.1
1.0
0.5
0
0
4.7
100
Pi Tcont Hv Mass
Al 1
Categories
0.2
0.08
10.0
5.6
0
1 5 . 8
0.01
0.3
1.3
0
0
1.7
0.2
0 . 5
3.6
0
0
4.3
100
V,i thin
Category
1.0
0.5
63.3
35.2
0
100
2.6
19.2
78.1
0
0
100
4.6
12.3
83.1
0
0
100
3-32
-------
The total filter weight and mass distribution by size and category
were calculated using the count data in each category and size range with
Equations (3-1) through (3-3). The following densities were assigned to
the three generic categories (Ursenbach, 1977a) : 2.4 g/ctn3 for spheres,
3 3
2.7 g/cm for irregular particles, and 1.5 g/cm for the agglomerates.
The total mass of particulates calculated'for the standard filter area
was divided by the ratio of the standard area to the total filter area
2
of 423 cm to obtain the total mass on the filter.
The SEM data for all fifteen monitor locations on the three
sample days were analyzed. The particle counts over twice the standard
area viewed at a magnification of 100X, the percent by number, overall
percent by mass, the percent by mass within each category, and the
calculated total filter mass are presented in Tables D-89 through D-156
in Appendix D. Some of the cover filters were also analyzed and the
result? of th<" combined primary and cover filter analysis for these cases
are presented in Tables D-157 to D-167 of Appendix D. Table 3-10,
reproduced from Table D-89 of Appendix D, shows the results of the SEM
analysis of the primary filter used for the optical analysis summarized
in Table 3-9. In the SEM-EDAX measurements, an impinging electron beam
causes atoms to be excited to a higher energy level and in the subsequent
decay, the atoms emit X-rays in which the energy depends in a characteristic
manner on the particular elements present. The elements and their
relative concentrations were obtained by integrating the individual
energy peaks of the spectrum. These peaks were measured for a set time
and count rate on each sample and an average standardized background
spectrum was subtracted from each set of data. As mentioned above, the
samples to be analyzed by EDAX were first uniformly coated with a film
of gold which served as an electrical ground for the electron beam. A
measured amount of gold was vaporized for each sample and the samples
were continuously rotated during coating so that an even film was deposited.
3-33
-------
TABLE 3-10
RESULTS OF SEM PARTICLE COUNTS
(Reproduced from Table D-89 in Appendix D)
Fj 1 tor Numlx.'r. 1003210
Monitor I/icnlion: Baden
Date of Sample: 8/IG/7G
Category
Sphere
5-lOp
10-20(j
20-50M
50-75u
>75u
Totals
Irregular
5-10u
10-20y
20-50y
50-75u
>75u
Totals
Agglomerate
5-lOu
10-20jj
20-50u
50-75p
>75u
Totals
TOTALS
Particle
Count
225
75
0
0
0
300
1075
300
60
0
0
1435
825
275
24
10
0
1125
2860
(All Categories)
Total Mass of Particular's (m^) :
Coor.s : ](]\ ,\^i
-------
Since there was no significant variation in gold thickness between particles
and between filters with this method of coating, the ratio of the integral
of the element's energy spectrum to the integral of the gold spectrum was
used to compare the composition of individual particles (Ursenbach, 1977b).
Use of this ratio to describe the particle removes any difference in count
rate or accumulation time which may have occurred between different
samples. This process could be performed for as many as seventy peaks
and, in most cases, was performed for twelve to twenty particles on a
filter segment. The ratios of the elemental peak with respect to the gold
peak for many of the filters were used to rank the elements which were
present. From the concurrent pairs of photomicrographs, the particle's
size and physical description were obtained. These physical parameters
were used along with the elemental composition of the particle to obtain
a tentative identification of the particle. The physical properties,
elemental composition reported according to descending size of the peak
to gold ratio, and possible identification are tabulated in Appendix D.
Example tabulations are shown in Tables 3-11 through 3-13 which are
reproduced from Tables D-194, D-195 and D-221 in Appendix D. The
photomicrographs are shown in Figures D-20 through D-86 of Appendix D.
As noted above, the EDAX analysis of the filters consisted of
both an examination of individual particles and an overall scan of the
major elements of the filter. The ratios of the elemental energy peak
with respect to the gold peak for the scans are presented in Tables D-168
to D-170 of Appendix D. Table 3-14, which is reproduced from Table D-170
in Appendix D, is an example of these ratios.
3-35
-------
TABLE 3-11
PARTICLE ANALYSIS OF SELECTED FILTERS
(Reproduced from Table D-194 in Appendix D)
Monitor Silo: No. B ruddock Dale o( Sample: 28 August 1976
Filter Number: 1093298 Photomicrograph Stub Number: 57
Particle Number
on Photomicrograph -:
1
2
3
4
5
6
7
8
9
10
11
12
Fl em cuts in Order of
Spectral Height
Si, A], Jv, Fe, Ca,
Ti, S
Si, Ca, Fe, Cl, S
Si, Al, Fc, K, Ca, Ti
Si, K, S, Ca, Al,
Fe, Cl
Si, Al, Fe, S, K
.Si, Ca, Al, Fe, S,
Mg, K, Cl
SI, Al, Ca, K, Fe,
S
Si. ( ;i, S, Fe, Al
Si, _A[, Fe, K, Ca,
Ti, S
Si, Al, Fe, S, Ca, K
Si, Ca, S, Al, Fe, K
Si, Ca, S, Fe, Al
( low level)
Dimension £
Description
1 0 microns
irregular
1 5 microns
irregular
25 microns
irregular
12 microns
spherical
30 microns
agglomerate
35 microns
irregular
1 0 microns
irregular
1 5 microns
spherical
5 microns
spherical
12 microns
irregular
30 microns
agglomerate
f>5 microns
irregular
Probable
Identification
industrial particle
industrial particle
industrial particle
ilyash
industrial particle
soil dust
industrial particle
fly ash
fly ash
industrial particle
industrial particle
organic - rubber
See Fieure
3-36
-------
TABLE 3-12
PARTICLE ANALYSIS OF SELECTED FILTERS
(Reproduced from Table D-195 in Appendix D)
Monitor Site: No. Braddock (Cover Filter) Date uf S.implc: 28Augl97G
Filtcjr Number: 1093299 Photonm:rogi aph Stub Number: 57C
Particle Number
on Photomicrograph"
1
2
3
4
5
G
7
8
9
10
11
12
13
L'lemonls in Order of
Spectral Height '
1'C, S, Ca
Ca, Si, Fe, St Al
Cl
Ca, JS, Si, Al
Ca, S, Si, Al
Sj., Al, Ca, S, Fe, K,
Mo
Si, Ca, Ti, Al, K
Mn, Fe, S, Cl, Zn
S^ Al, K, Fe, Ca
Sij 1'^Cj Ca, Al, K,
S, Ti
S (low level)
S, Si, Al, Fe, Ca
(low level)
Ca, S, Si, Fe, Al
Si, Al, Fe, S, Ca,
K
organic
Dimension cV
1 Jesc rip Li on
20 microns
spherical
20 microns
irregular
1 ,r) m icrons
irregular
50 microns
agglomerate
,'JO microns
irregular
15 microns
spherical
10 microns
irregular
20 microns
agglotneraLe
45 microns
i rregula r
20 microns
agglomerate
30 microns
agglomerate
40 microns
agglomerate
1 5 microns
spherical
Probable
Identification
fly ash (iron
oxide)
soil dust -
limestone
industrial particle
industrial particle
fly ash
fly ash
fry ash
industrial particle
organic - rubber
industrial particle
industrial particle
industrial particle
pollen
* See Figure
3-37
-------
TABLE 3-13
ANALYSIS OF SELECTED LARGE PARTICLES
(.Reproduced from Table D-221 in Appendix D)
Uuquesne II
111 0926
Monitor Si te:
Filter Number:
Date of. Sample: 10 August 1976
Particle Number
(see Figure)
(Fig.l>-37)
2 (Fig.D-58)
3a
3b
Elements in Order of
Spectral Height
Si , Al., Ca_, Fe, K,
Ti
Si, Al, K, Fe
Si, S + sma1 1 peaks,
may be organic
Si, Ca, Fe, S , K
spectra may come
from attached parti-
cles
Fe
Zn, Cl, Si, Na, Fe
Sj, Ca, Fe, + small
peaks
Si , Fe f sma 1 1 peaks
Si, Al, Fe, K
Dimens i OILS 6,
Descr i p tion
/t(Jx20 micron'
agg lomerate
with
particles on
surface
40 microns
agglomerate
25 mi crons
crump 1ed
40 microns
i rragular
wit h
particles
attached
25 microns
s p here
32 microns
agglomerate
80 microns
irregu lar
50 microns
i rregu lar
I OOx2()Micron:
a n g u 1 a r
30 microns
irregular
Probable
Ident i f ication
industrial
particle
industrial
particle
ash
ash
iron oxide
zinc chloride
industrial par-
tic! e (a!umi-
num oxide)
coal
coa L
soil dust
3-38
-------
TABLE 3-13 (Continued')
Monitor Site:
Filter Number:
Duquesne II Date of Sample: 10 August 1976
11 10920 (Continued)
Particle Number
(see Figure)
(FiK.D-65)
Elements in Order of
Spectral Height
Al, Mg, Si, Fe, Ca
Si_, Ca (particles 1,
2 and 6 may be joined
into a single parti-
cle)
SJ , Ca, Fe 4- smal 1
peaks
Si, S, Ca, Fe
Fe, Si
SL , Ca, Fe
Ca
Dinieiis ion-s&
De s c r i p t i o n
60 microns
irregular
55x30 micron
irregular
par tides
on surface
45 microns
agglomernte,
par tic 1es
on surface
20 mii runs
spherical
wi th par ti-
des on its
surface
6 microns
spheri cal
20 microns
particles on
surface
12x17 mi cron
triaugu1ar
with parti-
cles on the
surface
Probable
Identification
industrial
particle
industrial
particle
industrial
particle
i ndustrial
part icle
iron oxide
industrial
particle
limestone
3-39
-------
TABLE 3-14
SUMMARY OF EDAX ELEMENTAL ANALYSIS FOR 9 SEPTEMBER 1976 FILTER
SAMPLES. ENTRIES ARE RATIOS OF INTEGRAL ENERGY PEAKS
TO THE INTEGRAL ENERGY PEAK OF GOLD
(Reproduced from Table D-170 in Appendix D)
F. laments
Na
Mg
Al
Si
An
S
Cl
Cd
K
Ca
Ti
Cr
Mn
Fe
Co
Ki
Zn
Cu
I Ig
As
Pb
Important
Flements
In Rank
Order
Filter
Downtown
1106323
0. 04
0. 00
0. 32
5. GG
1. 00
0. 00
0. 00
0. 00
0. 02
0. (i2
0. 00
0. 00
0. 00
0. 00
0. 00
0. 00
0. 00
0. 00
0. 00
0.00
0. 00
Si, Ca, Al
Central Lab
i 106327
0. 11
0.08
0.41
3. 64
1. 00
0. 07
0. OG
0.09
0. 23
0. 87
0. 10
0.02
0. 08
0. 35
0. 04
0. 03
0. 01
0. 01
0. 02
0. 00
0. 00
JM, Ca, S,
Al, Fe, K,
Na, Ti, Cd,
Cu
Central Lab
1106328
Cover Filter
0. 13
0. 08
0.47
4.43
1. 00
0.27
0. 07
0. 05
0.22
1. 00
0. 07
0. 07
0. 07
0.24
0. 03
0. 05
0.03
0. 02
0. 00
0. 00
0. 00
_Si, Ca, Al,
S, Fe, K,
Na
Airport
1093294
0. 1C
0.04
0.35
4.36
1.00
0.28
0.00
0.21
I). 70
0.06
0.00
0.06
0.05
0. 17
0.01
0.00
0.0]
0.05
0.00
0.00
0.00
Si, K, Al,
S, Ctl, Fe,
Na, Cu
Airport
1093295
Cover Filter
0,21
0.08
0.44
5.41
1.00
0.16
0.13
0.00
0.26
0.90
0.11
0.06
0.05
0. 16
0.04
0.00
0.02
0.02
0.00
0.00
0.00
Si, Ca, Al,
K, Na, S, Fe
Cl, Ti
3-40
-------
TABLE 3-14 (Continued)
Elements
Na
Mg
Al
Si
Au
S
Cl
Cd
K
Ca
Ti
Gi-
ft In
Fe
Co
Ni
Zn
Cu
lift
As
I'b
Important
K lemon Is
In Hank
Order
Filler
So. F aye tie
1100310
(l. 31
0.-14
0. 83
2. 3G
1. 00
0. 07
0. HI
0. (} 1
0. 71
0. 5!)
0. 49
0. 31
0.25
0. •}]
0. 23
0. 11
o. ir>
0. 19
0. 01)
0. 07
0. 09
Si, Al, Cl_, K
S, Cd, Ca, T^
Mg, Fe, Na,
Cr, Mn, Co,
fiu. /n. Nir
Hg, Pb, .As
3-41
-------
3.8 STUDY RELIABILITY
Because of the complexity of the filter analysis procedures, it
is important to identify the major uncertainties and to ascertain the
potential sources of bias. As mentioned above, each filter was assigned
a randomly-generated seven-digit number which was the only identifying
information available to the analysts. There were also three different
laboratories involved in the filter analysis and the results obtained by
these laboratories were compared as part of the check for bias in the
results. Comparisons of the findings of the various laboratories,
the uniformity of filter loading and the uncertainties involved in
particle counting are discussed below.
3.8.1 Quality Control
To check on the quality of the optical filter analysis, six filters
were sent to Walter C. McCrone Associates, Inc. for an optical microscopy
analysis. The filters were identified only by a seven digit number. At
McCrone Associates, Inc., samples were prepared from each filter by
scraping material from randomly selected areas of the filter. Individual
particle species were identified. The percent by number and the size range
of each species determined by McCrone Associates are given in Tables D-231
through D-236 of Appendix D. Table 3-15 lists the results obtained by
Walter C. McCrone Associates, Inc. and the Coors Spectro-Chemical Labora-
tories for the six filters. The McCrone optical data for these six filters
are grouped in three categories: combustion products, minerals and
biological particulates. The combustion products category corresponds
to the flyash category reported by Coors and the mineral class category
includes the Coors iron oxide, quartz and calcium carbonate categories.
Only the results for percent by number could be used for this comparison.
The data in Table 3-15 show good agreement between the results from the
two laboratories. In some cases, the biological particulate fraction
found in Coors' analysis is larger than that found by McCrone Associates.
3-42
-------
TABLE 3-15
COMPARISON OF OPTICAL MICROSCOPY STUDIES PERFORMED
BY McCRONE ASSOCIATES AND COORS SPECTRO-CHEMICAL
LABORATORIES
(The entries are all in percent by number)
Filter Nuirbei & Location
] 093225 Brighton Twp 8/16/76:
McCrone Analysis
Coors Analysis
1093227 Midland 8/16/76:
McCrone Analysis
Coors Analysis
1093205 Duquesne 11 8/16/76:
McCrone Analysis
Coors Analysis
1093209 Liberty Boro 8/16/76:
McCrone Analysis
Coors Analysis
1093239 So. Fayette 8/16/76:
McCrone Analysis
Coors Analysis
1093284 Brighton Twp 8/28/76:
McCrone Analysis
Coors Analysis
Combustion
Products
83
63
70
73
67
68
70
90
62
70
89
61
Mineral s
16
25
28
20
27
27
26
8
37
25
7
35
Biological
Materials
0
12
<1
7
<3
5
<2
2
0
5
3
4
3-43
-------
3.8.2 Uniformity of Filter Loading
Our extrapolations of the filter analysis data for small filter
segments to the total filter area depends on the assumption that the
filter loading is uniform. To test the uniformity of filter loading,
Coors Spectro-Chemical Laboratories analyzed five filters, of which two
were cover filters, to determine the validity of this assumption. On
each of these filters, three areas given by the areas A, C and E shown
in Figure 3-2 were analyzed. Because of the difficulty in counting the
smallest particles and the sparse data for the very largest particles,
the particle counts were compressed into three size ranges: 5-20 ym, 20-50
ym and >50 ym. The hypothesis that the distribution of particles on the
filter was uniform was tested using a chi-square test for differences in
probability (Conover, 1971). The test statistic T has a chi-square
distribution and is given by the expression
- E..
0-4)
where 0.. is the observed value in the i row and j column. E.. is
the expected value if the hypothesis is valid in the same (i,j) cell and
is given by
R.C.
Eii = ~J~i (3-5)
where R. is the sum of the entries in the i row, C. is the sum of the
entries in the j column, and N is the total of all observations (the
total of all i row sums or the total of all j column sums). Because of
sample area was very small with respect to the entire filter area, a value
for a of 0.005, which corresponds to a chi-square value greater than 14.9
for a sample of four degrees of freedom, was selected as the rejection
criterion. Two of the five filters tested, one cover filter and one primary
3-44
-------
filter, were rejected at the 99.5% confidence level. The other three filters,
two primary filters and one cover filter, were not rejected at the 99.5%
confidence level. We interpret the results of this statistical test to be
supportive of the hypothesis of a uniform distribution of particles on
the filters.
3.8.3 Comparisons of Measured Filter Weights and Filter
Weights Calculated from Optical and SEM Data
The mass of particulate loading on each filter was determined
directly by weighing the filter before and after exposure in the hi-vol
measurement program. The total mass of particles on each filter was also
calculated from the optical microscopy and SEM data assuming a uniform
loading on the filter.
The total mass of particulates on each filter determined by
weighing and the total mass calculated from the optical microscopy data
are given in Tables D-16 through D-88 in Appendix D. The total particulate
mass on each filter determined by weighing and the total mass calculated
from the SEM data are given in Tables D-89 through D-167 of Appendix D.
Figure 3-5 is a plot of the measured mass versus the calculated filter mass
for the primary filters which were analyzed by optical microscopy methods.
The solid line in the figure represents a linear least-squares fit to the
data. A similar plot of the measured mass versus the calculated filter
mass for the combined primary and cover filters analyzed by optical
microscopy techniques is shown in Figure 3-6. The corresponding plots of
the measured mass versus the calculated mass for the filters analyzed by
the SEM technique are shown in Figures 3-7 and 3-8, respectively. Table
3-16 shows the linear least-squares equation for each plot as well as
the correlation coefficient and standard error.
Because the measured mass was obtained by weighing both the primary
and cover filters, it might be expected that the mass calculated for the
combined primary and cover filters should agree better with the measured
3-45
-------
TABLE 3-16
STATISTICAL COMPARISON OF MEASURED AND CALCULATED
FILTER WEIGHTS; Y = CALCULATED MASS
AND X = MEASURED MASS
Analysis Technique
Optical anaysis of
primary filters
Optical analysis of
combined primary
and cover filters
SEM analysis of
primary filters
SEM analysis of
combined primary
and cover filters
Equation of
Linear Fit
Y = 87 + 0.57X
Y = 27 + 1.3X
Y = 100 + 0.32X
Y = -70 + 1.8X
Sample Size
41
13
41
11
Correlation
Coefficient
0.54
0.82
0.46
0.78
Standard
Error (rag)
110
140
78
190
3-46
-------
1000
0
100
200
300 400 500
MEASURED MASS (mg)
600
700
800
FIGURE 3-5.
Plot of measured mass and calculated mass of primary
filters based on optical analysis. The solid line is
a linear Least-squares fit to the data and the dashed
lines are the 95-percent confidence limits.
3-47
-------
1000
900-
800-
(/)
700-
= 600-
Q.
O
£
o
- 500-
0>
00
o
LJ
I-
co
UJ
1 I ' I
400-
300-
200-
0
100
200
600
700
FIGURE 3-6.
300 400 500
MEASURED MASS (mg)
Plot of measured mass and calculated mass for combined
primary and cover filters based on optical analysis.
The solid line is a linear least-squares fit to the data
and the dashed lines are the 95-percent confidence limits.
800
1-48
-------
-------
CO
Ld
CO
E
o
o>
e
CO
CO
LJ
1000
900-
800
700
600
500
400
I 300
co
LJ
200
100
0.
0
100
200
300
400
500
600
700
800
FIGURE 3-7.
MEASURED MASS (mg)
Plot of measured mass and calculated mass of primary
filters based on SEM analysis. The solid line is a
linear least-squares fit to the data and the dashed
lines are the 95-percent confidence limits.
3-49
-------
1100
/ I
1000
900
/
/
/
/
/
/
/
800
700
co
c/>
UJ
£ 600
o>
_§
(S)
en
500
w
9
Q
UJ
UJ
400
/
f
300
200
100
Q
0
100
FIGURE 3-8.
200
300
400
500
600
700
800
MEASURED MASS (mg)
Plot of measured mass and calculated mass of combined primary
and cover filters based on SEM analysis. The solid line is
a linear least-squares fit to the data and the dashed lines
are the 95-percent confidence limits.
3-50
-------
mass than with the mass calculated for the primary filter alone. This
expectation tends to be borne out by the regression analysis, although
there are significantly fewer cases of calculated mass available for the
combined filters. The general result is that the masses estimated from
the primary filters alone are less than the measured masses while the
masses estimated from the combined primary and cover filter analyses are
slightly higher than the measured masses. Also, there appears to be little
difference in the accuracy of the estimated masses between the optical
and SEM data. The estimates are very sensitive to the number of large
particles found on the filter segments. This factor and the extremely
large multipliers used to extrapolate the results of the particle analysis
for filter segments to the total filter area principally account for the
large scatter in the data points plotted in the above figures.
3.8.4 Comparison of Agency's and Coors1 Measured Particulate
Concentra tions
As noted previously, two hi-vol measurements were made concurrently
at each monitor site, one by the responsible agency (DER or BAPC) and the
other by the H. E. Cramer Company. The H. E. Cramer Company's filter
samples were sent to the Coors Spectro-Chemical Laboratories for analysis.
The total weights of particulates on the filters at each monitor site are
listed in Tables D-16 through D-88 in Appendix D. Also, 24-hour average
particulate concentrations at each monitor site were determined independently
by Coors and the agencies from the filter weights and the hi-vol flow rates.
We tested the null hypothesis that the average difference observed between
the Coors and either DER or BAPC filters was zero by using a Student's t
test for the n pairs of data. The test statistic used was (Mendenhall
and Scheaffer, 1973):
d - y
(3-6)
where d is the average of the differences between the paired monitors,
S, is the standard deviation of the differences, and y is zero.
Q o
3-51
-------
The above test showed that the null hypothesis of zero difference
between the Coors and Agency concentration estimates could be rejected at
a confidence level greater than 95 percent. The difference D between the
two sets of concentration measurements at the co-located monitors is given by
"d ± ./2 = o.o ^ (3-7)
where V is the number of degrees of freedom (equal to n-1) and t ,„ _ ~
is the Student's t value at the 95-percent confidence level. The value
of D obtained from the above expression is
D = 11 ± 9 yg/m3 (3-8)
which shows that Coors1 24-hour average concentrations tend to be slightly
larger than the concentrations obtained by DER and BAPC. This result may
possibly be explained by the difference in filter handling procedures.
The filters analyzed by Coors were carefully removed from the monitors
immediately after the completion of the sample period, a clean cover filter
was placed on top of the exposed filter, and the combined filters were
then placed in a plastic packet. The agency filters were folded in half
so that the contaminated side was enclosed and then the filter was
placed in an envelope. It is possible that in the process of folding the
filters some of the particles were dislodged from the filter and hence
lost. It is also possible that other differences in the laboratory
procedures may have been responsible for this difference in the concentra-
tion measurements.
3-52
-------
SECTION 4
RESULTS OF THE OPTICAL AND SEM/EDAX ANALYSES
OF THE HI-VOL FILTER SAMPLES
The detailed results of the optical microscopy, scanning electron
microscopy (SEM) and energy-dispersive X-ray (EDAX) analyses of the hi-vol
filter samples are presented in Section D. 4 of Appendix D. Some of the
figures and tables contained in Appendix D are reproduced in this section,
which discusses the results of the hi-vol filter analyses. It is important
to note that the optical microscopy, SEM and EDAX analyses were principally
restricted to the coarse particle mode (particle diameters greater than 3
micrometers) of the bimodal mass distribution characteristic of urban
areas (Whitby, et jiJL. , 1972 and others). The small particles of the fine
particle mode (diameters less than about 1 micrometer) are not detectable
by optical techniques. Also, as discussed in Section 4.1 below, the
results of the optical and SEM filter analyses showed that the contributions
of particles with diameters less than about 5 micrometers to the total
filter loadings were small. The current National Ambient Air Quality
Standards (NAAQS) for suspended particulates consider only the total
weight collected on the hi-vol filters and do not address particle size.
For these reasons, we believe that the neglect of particles smaller than 5
micrometers in diameter did not significantly affect the results of this
study.
4.1 PARTICLE MASS DISTRIBUTIONS
The optical microscopy and SEM data given in Section D. 4 of
Appendix D were used with the parameters given in Table 4-1 to obtain the
percentage allocation of total particulate mass on the hi-vol filter
samples by generic category and by monitor site classification. Each
monitor site was classified as an industrial, urban or rural site following
the general scheme outlined in Table 3-1 in Section 3, but also taking into
4-1
-------
TABLE 4-1
PARAMETER VALUES USED FOR ANALYSIS
OF FILTER SAMPLES
Particle
Size
Categories
(um)
Particle
Diameter*
(ym)
Generic
Categories
Particle
Density
(g/cm )
Standard
Reference
Area
(cm )
(a) Optical Microscopy Analysis
_> 5 < 10
_> 10 < 20
> 20 < 50
^ 50 < 75
_> 75
7.77
15.54
37.02
63.32
88.09
Flyash
Nonmagnetic Iron Oxide
Magnetic Iron Oxide
Quartz
Calcium Carbonate
1.5
4.9
4.9
2.5
2.7
0.00968
(b) SEM Analysis
>_ 5 < 10
> 10 < 20
> 20 < 50
^ 50 < 75
_> 75
7.77
15.54
37.02
63.32
88.09
Spherical
Irregular
Agglomerate
2.4
2.7
1.5
0.02053
^Geometric mean.
4-2
-------
account the meteorological conditions (especially the mean wind speeds and
wind directions) on each of the sampling days. The following monitor sites
were classified as industrial sites:
• Baden, Midland, Hazelwood and Clairton on 16 August 1976
• Baden, North Braddock and Liberty Boro on 28 August and
9 September 1976
The Downtown and Central Laboratory monitor sites were classified as urban
sites on 16 August, 28 August and 9 September 1976. The rural sites on all
three sampling days were Elco, South Fayette, Greater Pittsburgh Airport
and Brighton Township.
Figures 4-1 through 4-4, which are reproduced from Figures D-12
through D-15 in Appendix D, show the cummulative mass distributions by par-
ticle diameter obtained from the optical microscopy analyses for all monitor
sites combined and for industrial, urban and rural monitor sites. The
corresponding mass median particle diameters and mass mean diameters are
listed in Table 4-2. The mass median diameter is the particle diameter for
50-percent cumulative mass. The mass mean diameter D is given by
n
D_
f. d. 3
where f. is the fraction of the particles in the size range represented by
the diameter d.. As shown by Table 4-2, the mass median diameter is about
72 micrometers at both industrial and urban sites and is 52 micrometers at
rural sites. The mass mean diameters at industrial, urban and rural sites
are 68, 60 and 52 micrometers, respectively. Thus, the results of the
4-3
-------
O
rH I
CO O
U -H
•H 4-1
4-1 3
p, T/~l
O -H
CU 4-J
£1 W
4-1 -H
6
o w
>-i Cfl
m co
g
T)
cu cu
C > •
•H -H ^-N
cfl 4-1 CN
CO
^2
O 3 O
en 3 01
0) O >-i
•U 3
•H cu ao
W ^2 -H
4—i r_T^
o w g
4-130
•H O M
o w
cu cu
H C O
H -H 3
CO rH T3
O
V-i T3 M
o cu ex
n 01
CO Pi
C cfl ^
o ^o
•H
4-1 CU •
M-l
rd H 0)
•H >
r-l O
4-J • g
en u) cu
•H -H r-l
T3 W
>! CO
W rH rH
W cfl cfl
CO C O
S cfl -H
to
cu ?>~, o
CX rH
O O
O -H
CO
r— I
O
M ^
O 4-1
I
~d-
w
4-4
-------
o
o
c\J
O O O O O
O (D 03 r- tO
(uirh
O
LT>
O
rO
O
c
O)
u
L_
o
o.
to
<
S
LJ
>
h-
,--r
_J
Z)
13
O
6 w
o in
"T3 <1) I
Q
C -H
•H 4-1 01
ca co 1-1
4-1 rH 3
^D 3 60
o e -H
tn o
0)
4-1 01
•H ,C
tn 4-1
t-t
o
4-J
•H
a
o
e
S o)
o o
en "a
o
O) M
C CX
rH -H 0)
R) rH P5
O
•4-1
C
O
•H
W
•H
T3
01
x; •
en •n
fl aj
-o >
o
0) 0
f, . 60
iH O
cfl H
C O
CO -H
CO
S
01
•H
4-1
cfl
I—I
3
CM
I
w
&
o
O 4J
O -H
U2 S
O
^ c
u o
•H -H
0 4-1
O !-i
•H 4-1
4J en
D,-rl
O T3
4-5
-------
o
o
CM
o o o o o o
O en oo f- co m
O
O
ro
O
CO
- c\i
O
ct)
o G
•H O
4-1 -H
p, | i
O 3
0) -H
4J 4-1
CO
E -H
O 13
m co
co
13 CO
o) B
R
•H 01
CO > •
4-1 -H -—•
O 03
cn
cu
p
4-J 3 0)
•HUM
CO 3
01 M
V< ,R -H
O 4J (i-i
4J
•H co E
R 3 O
O O t-i
6 r^ i j j
CO
R 13
CO CU 0)
^3 R O
M -H 3
3 M 13
O
M 13 >-i
O CU fX
4-4 rR CD
CO Pi
R CO *~*
O 13
•H
4-1 0)
P
13
H 0)
J-i O
4-1 . E
CO CO 0)
H -H >-l
T3 CO
>^ CO
CO rH iH
CO CO CO
CO R
U
E CO
bO
0) >> O
P> ft iH
•H O O
4-1 O -H
CO co fi
•H O
3 ^ ^3
S 0 4J
3 -H -H
I
-d-
W
«
p
o
310licdVd
4-6
-------
4-7
-------
TABLE 4-2
PARTICLE MASS MEDIAN AND MASS MEAN DIAMETERS AND STANDARD
DEVIATIONS OBTAINED FROM THE OPTICAL MICROSCOPY ANALYSIS
Monitor
Classification
Industrial
Urban
Rural
Composite
Mass Median
Diameter
(urn)
72
73
52
68
Std. Deviation
(ym)
5
5
5
5
Mass Mean
Diameter
(ym)
68
60
52
65
Std. Deviation
(ym)
7
10
12
7
4-f
-------
optical analysis indicate that the particulate concentrations measured by
the hi-vol samplers were primarily attributable to very large particles.
Some of these large particles were biologicals (spores, insect parts, etc.).
However, Figures 4-1 through 4-4 indicate that the exclusion of biologicals
does not significantly affect the conclusion that very large particles were
primarily responsible for the measured particulate concentrations. For
example, the composite mass median diameter is 68 micrograms per cubic meter
if biologicals are included and 57 micrometers if biologicals are excluded.
Figure 4-5 compares the composite cumulative mass distribution
of hi-vol filter particles developed from the optical microscopy data with
the composite cumulative mass distribution developed from the SEM data.
Although the composite mass median diameter for the SEM data is 27 micro-
meters and the composite mass median diameter for the optical microscopy
data is 68 micrometers (all particles) or 57 micrometers (biologicals
excluded), the SEM data also support the conclusion that very large
particles primarily accounted for the mass loadings on the hi-vol filters.
Table 4-3 gives the mass median and mass mean particle diameters obtained
from the SEM data for industrial, urban and rural monitor sites and for all
monitor sites combined. In contrast to the results of the optical analysis
(see Table 4-2), the SEM mass median and mass mean diameters do not show any
significant difference between rural sites and industrial or urban sites.
Possible reasons for the differences between the optical and SEM cumu-
lative mass distributions include:
• The SEM analysts encountered considerable difficulties
in counting particles because the fibers of the hi-vol
filters could not be made transparent as was done with
an index oil in the optical analysis
• The SEM analysis considered only the particles on the
surface layer of the filters, while the optical analysis
4-9
-------
C\J
CU
CX
o
rH
CU
cu
co
u
•H
60
O
rH
cu o
J2 -H
4-J , f*
U S 4J
•H O -H
>-l CO
cfl CO
CX r i i i
o • o
CO r<
w 4-) MH
C co
O 'O T3
•H CU
4-J S C
3 pq -H
fi co ca
•H 4-J
4-J C O
W CO
•H fi
TIJ ^>-, O
CX -H
CO O 4-J
to O 3
CO CO rD
B O -H
tj ^
CU U 4-1
> -H tfl
•H B -H
4J 13
CO rH
rH co en
3 o en
E -H cfl
3 4-i E
0)
O 0)
cu
•H
4J T3
CO 4J Cd CU
O rH >
cx B 3- o
E O E E
O In 3 CU
O <4-l U M
w
o
oo o o
O CD 00 N
8 ?
O
K)
(uirt)
4-10
-------
TABLE 4-3
PARTICLE MASS MEDIAN AND MASS MEAN DIAMETERS AND
STANDARD DEVIATIONS OBTAINED FROM THE SEM ANALYSIS
Monitor
Classification
Industrial
Urban
Rural
Composite
Mass Median
Diameter
(ym)
28
27
26
27
Std. Deviation
(ym)
3
3
4
3
Mass Mean
Diameter
(ym)
36
35
32
35
Std. Deviation
(ym)
6
6
7
6
4-11
-------
considered particles on and beneath the surface layer of
the filters
• The particles were viewed at an abitrary orientation in
the SEM analysis, while the use of an index oil in the
optical analysis tended to present the largest side of
non-spherical particles for view
The discovery of the presence of very large particles which
accounted for most of the particulate mass collected on the hi-vol filters
was a completely unexpected result because these large particles are not
contained in the particulate emissions inventory for Southwest Pennsylvania.
Excluding biologicals, the large particles appear to be of industrial origin
as evidenced by a lacy or frothy structure, an agglomerate form or the
appearance that the particle had solidified from another state. Examples
of a few of these large particles are shown in Figures D-68, D-69, D-72,
D-73, D-75, D-76, D-77, D-78, D-85 and D-86 in Appendix D. Although the
large, non-biological particles appear to be of industrial origin, it is
possible that these particles were deposited on the ground prior to the
sampling days and were then resuspended on the sampling days by vehicle
traffic or the wind. Five of the large particles were selected for SEM-
EDAX analysis. These particles are shown in Figures D-87 through D-91 in
Appendix D. Table 4-4 gives the results of the SEM-EDAX analysis for the
five particles, all of which appear to be of industrial origin if coal
particles are assumed to be fugitive emissions from industrial sources
(storage piles, open railroad cars transporting coal, etc.).
We point out that relatively few previous studies have identified
large particles as the principal contributors to the mass loading on hi-vol
filters in urban areas. However, many of these studies were primarily con-
cerned with the respirable (small) particulates which have the greatest
potential for adverse health effects, and large particles either were not
considered or the experimental procedures were biased against their collec-
tion. Although it is possible that large particles actually were present
4-12
-------
TABLE 4-4
PARTICLE ANALYSIS OF FIVE SELECTED LARGE PARTICLES
Particle Number
(See Figure)
Elements in Order of
Spectral Height
Dimension &
Description
Probable
Identification
1 (Fig. D-87)
2 (Fig. D-88)
3 (Fig. D-89)
4 (Fig. D-90)
5 (Fig. D-91)
Fe, Si, Ca, Si
Fe, Si, (low level
peaks)
Fe (low level peaks)
Fe, S, (low level
peaks)
Fe (low level peaks)
100 ym lacy,
agglomerate
110 ym flat,
irregular
160 ym flat,
irregular
140 ym agglo-
merate
150 ym
angular,
irregular
industrial
particle
industrial
particle
organic ash(?)
industrial
particle
(soot)?
organic, coal
4-13
-------
in these previous studies, it is also possible that the presence of large
particles is restricted to certain areas such as the Southwest Pennsylvania
region.
The uncertainty in the calculated filter mass loadings due to
uncertainties in the observed particle size distributions is of concern
because of the importance of the large particles discussed above. Conse-
quently, we used the same data that we used in Section 3 to test the
uniformity of the filter loadings (regions A, C and E on five of the
filters) to determine the uncertainty in the particle counts and hence in
the calculated filter mass loadings. The 20- to 50- micrometer, 50- to 75-
micrometer and greater than 75- micrometer particle size categories were
combined in a single category with a geometric mean diameter of 67.8
micrometers. Because the particle counts for the three regions of the
filters were not assigned to generic categories (flyash, quartz, etc.), it
was necessary to estimate an average particle density for each filter in
order to calculate the mass loading of the filter. We used the percentages
by mass in the various generic categories previously determined for the
filters (see Tables D-16 through D-74 in Appendix D) to calculate the
weighted-average particle density for each filter. The only size category
considered as a source of uncertainty in the calculated filter mass load-
ings was the category for particles with diameter above 20 micrometers
because this category accounted for over 50 percent of the mass on each of
the filters. For each of the five filters, we used the particle counts for
the three filter areas to calculate the mean and standard deviation of the
number of large particles. The minimum and maximum number of large parti-
cles on each filter were then estimated using the t distribution with two
degrees of freedom (Mendenhall and Scheaffer, 1973)
'a/a (S//T) (4'2)
where X is the mean, S is the standard deviation, n is equal to three and
4-14
-------
t ,„ is the t distribution value at the (1-0.) confidence level. For the
a/2
90-percent confidence level, t .„ is 2.92. (The 90-percent confidence
level rather than the 95-percent confidence level was used because the
three regions of each filter had very small areas in comparison with the
area of the entire filter.) The range of mass loading on each filter was
then calculated using the range in the number of large particles obtained
from Equation (4-2) in combination with the average particle density and
the geometric mean diameter of 67.8 micrometers. The resulting uncer-
tainty in the calculated filter mass loadings arising from the uncertainty
in the number of large particles was found to be about 60 percent, although
the uncertainty for a specific filter was slightly dependent on the total
mass loading of the filter.
4.2 ELEMENTAL ANALYSES
Particulate emissions arising from some industrial production
processes may have unique elemental compositions which can be used to
identify the origin of the particulates. Consequently, SEM-EDAX techniques
were used to examine individual particles as well as entire filter segments.
The detailed results of the EDAX analyses are given in Section D.4 of
Appendix D. The following subsections discuss the EDAX analyses of indivi-
dual particles, entire filter segments, trace elements which may present
a health hazard, and various types of slag found in the Southwest
Pennsylvania area and on the hi-vol filters. We point out that the data
are at times sparse and only semiquantitative in nature.
4.2.1 Elemental Analysis of Individual Particles
The particles containing each element were categorized according to
particle diameter (0 to 10, 10 to 20, 20 to 40 and greater than 40
micrometers) and monitor site classification (industrial, urban and rural).
The particle elemental compositions were then examined for nonuniformities
between particle size ranges or monitor site,classifications. The height
4-15
-------
of each element's integral energy peak recorded during an EDAX scan was divi-
ded by the height of the integral energy peak of gold to obtain a ratio
which was essentially independent of any discrepancies between filters in
factors such as counting rates and beam intensities. The elements were then
ranked according to descending magnitude of the energy ratio and used in the
Krutjkal -Wa 11 i s test statistic, which is given by (Conover, 1971) as
12 V^ - 3(N+1) (4-3)
_
N(N+1) L-J n.2
1=1 1
where N is the total number of observations, 11. is the sum of the ranks
asbiguud to the i size-range or site-classification category, k is the
number of categories and n. is the number of observations in the i cate-
gory. This statistic, which has an approximate chi-square distribution for
(k-1) degrees of freedom, was used to test the following hypotheses: (1)
the k categories had identical elemental distributions, and (2) the
elemental distribution of at least one of the k categories differed from
that of the other categories.
The results of the analyses described above did not identify the
specific sources of particles examined, but did indicate that some
elements had nonuniform distributions. At the 97.5-percent confidence
level or above, the relative concentration of the following elements were
highest in the smallest particle-size categories:
• Silicon found in the agglomerate particles from rural
monitor sites
• Potassium and magnesium found in the agglomerate particles
from industrial monitor sites
• Titanium and manganese found in the combined spherical,
4-16
-------
irregular and agglomerate particles from industrial
monitor sites
With the above exceptions, it was not possible to reject, at that 97.5-per-
cent confidence level, the hypothesis that all other elements tested had
uniform distributions between particle size categories and between monitor
site classifications.
The SEM-EDAX analyses also revealed some site-to-site differences
in the types of particles found on the hi-vol filters. For example, the
results of the analyses of the filters from the Baden and Clairton monitors
on 16 August 1976 are shown in Tables 4-5 and 4-6, respectively. Although
only a limited number of particles from each filter were selected for
analysis, the Baden and Clairton filters appear to have contained more coal
particles than most of the other filters which were analyzed. Both the
Baden and the Clairton monitor sites are located near potential sources of
airborne coal particles. Open coal storage piles in the vicinity of the
Clairton monitor are potential sources of fugitive emissions. Similarly,
the Baden monitor is in close proximity to railroad tracks over which coal
is transported in open railroad cars.
4.2.2 Elemental Analysis of Entire Filter Segments
An SEM-EDAX elemental analysis was performed on segments from a
total of 33 hi-vol filters. The results of the elemental analysis of the
filter segments for 16 August, 28 August and 9 September 1976 are given
in Tables 4-7, 4-8 and 4-9, respectively. (Tables 4-7 through 4-9 are
reproduced from Tables D-168 through D-170 in Appendix D.) Of particular
interest are the differences in the elemental compositions of the particles
from the rural South Fayette site on 28 August and 9 September 1976.
Although both of these days were days of south to southwest winds, Table
4-8 shows that the dominant elements at the South Fayette site on 28 August
1976 were aluminocilicates and other soil-related species. Table 4-9 shows
that elements such as lead, mercury, arsenic, zinc, copper, cadmium,
4-17
-------
TABLE 4-5
PARTICLE ANALYSIS OF SELECTED FILTERS
Monitor Site1: Baden Da te o[ Sample: 16 August 1976
Filter Number: 1093219 Photomicrograph Stub Number: 45
Particle Number
on Photomicrograph*
1
2
3
4
5
6
7
8
9
10
1 1
12
KlemenLs in Order of
Spectral Height
Ca, Fe , S, Mn
(low level)
Si, S, Ca
(lew level)
Si, Ca
(low level)
Fe, Ca, Si
Si, Ca, Al, S
S, Si, Cl, Ca
(low level)
Ca, S
(low level)
Si, Ca, S
(low level)
Si, Ca
(luw level)
S, Si, Ca
(low level)
Si, Zn, Ca, S
(low level)
Si, Fe, Ca, Al , S
(low level )
Dimension R
Description
40 ym
irregu lar
12 ym
i rregular
1 5 ym
ir regul ar
10 ym
i rregular
10 ym
irregular
1 5 ym
irregular
50 Um
i rregular
2 0 um
i rrcgu la r
1 2 nm
i rregu lar
7 Um
irregu lar
10 Mm
irrt'gu Lar
10 Pm
i rregu lar
Probable
Identification
organic - coal
organic - coal
organic - coal
organic - coal
organic - coal
organic - coal
organic - rubbei
mi nera 1
organic
organic
organic - coal
organic - coal
4-18
-------
TABLE 4-6
PARTICLE ANALYSIS OF SELECTED FILTERS
Monitor Site: Clairton 1 )ute <>l Sain pie: 1 G August 197(5
Filler Number: 1093211 Photomicrograph Stub Number: 43
Particle Number
on Photomicrograph
1
2
3
4
5
(j
7
8
9
10
11
12
Flcmenls in Order of
Speetrnl Height
Si, Fe, Ca, S (low
level)
Si, Fe, Ca (low
level)
Si, Ca, Fe
Fe, Si, Ca, K, A],
Pb
Si, Fe, Ca (low
level)
Si, Fe, Ca (low
level)
Si, Mg, Ca (low-
level)
S, Si, Cl, Ca, Al,
Na, K, Zn
Al, Si, Ca
Al, Si, S, Ca (low
level)
Al, Ca (low level)
Si
Dimension ^
Description
15 ym
irregular (long)
12 ym
irregular
20 ym
irregular
20 Mm
irregular
25 ,Jm
agglomerate
2 0 VJm
irregular
1 5 ljm
irregular
1 5 Um
irregular
20 ym
irregular (long)
If) ym
irregular
25 ym
irregular
8 Urn
i )' regular
Probable
Identi lieation
organic
organic-coal
mineral
iron oxide
industrial particle
organic-coal
organic -coal
industrial particle
(sulfate?)
industrial particle
organic-coal
organic-coal
organic-coal
t-19
-------
TABLE 4-7
SUMMARY OF EDAX ELEMENTAL ANALYSIS FOR 16 AUGUST 1976 FILTER
SAMPLES. ENTRIES ARE RATIOS OF INTEGRAL ENERGY PEAKS
TO THE INTEGRAL ENERGY PEAK OF GOLD
Elements
Na
MS
Al
hi
Au
S
Cl
Cd
K
C;i
Ti
Cr
Mn
Fe
Co
Ni
Zn
Cu
Hg
As
PI)
Tm port: in I
Elements
fn Hank
Order
Filler
Baden
109 32 19
o. on
o. on
n. oo
1.48
1 . 00
0. 00
0. 00
0. 00
0. 00
0. 00
0. 00
0. 01)
0. 00
0. 02
0. 00
0. 00
o. oo
0. 00
0. 00
0. 00
0. 00
Si
Beaver Falls
1093221
0. 00
0. 00
0. 00
1. 75
1 . 00
0. Hi
0. 00
0. 00
0.00
0.00
0. 00
0. 00
0. 00
0. 00
0. 00
0. 00
0. 00
0. 00
0.00
0. 00
0. 00
Si.S
Koppel
1093223
0. 00
0. 00
0. 02
2. 13
1. 00
0. 14
0. 00
0. 00
0. 00
0. 00
0. 00
0. 00
0. 00
0.00
0. 00
0. 00
0. 00
0. 00
0. 00
0. 00
0. 00
Si, S
Brighton T\vp
1093225
0. 14
0. 00
0. 4(i
G. 76
1 . 00
0. 00 -
0. 00
0. 00
0. 10
0. 89
0. 00
0. 00
0. 00
0. 02
0. 00
0.00
0. 00
0. 00
0. 00
0. 00
0. 00
Si, Ca, Al,
Na, K
Midland
1093227
0.25
0.10
0.45
4.04
1.00
0.23
0. 03
0.02
0. 18
0.54
0.03
0.03
0.01
0.43
0.00
0.00
0.00
0.00
0.00
0.00
0.00
JM, Ca, Al,
Fe, Na, S,
K, Mg
4-20
-------
TABLE 4-7 (Cont.)
SUMMARY OF EDAX ELEMENTAL ANALYSIS FOR 16 AUGUST 1976 FILTER
SAMPLES. ENTRIES ARE RATIOS OF INTEGRAL ENERGY PEAKS
TO THE INTEGRAL ENERGY PEAK OF GOLD
Flemenls
Na
Mg
Al
Si
Au
S
Cl
Cd
K
Ca
Ti
Cr
Mn
Fe
Co
Ni
Zn
Cu
IIR
As
1'h
ImporL'inl
Flements
In Hank
Order
Midland
1093228
Cover Filler
0. 33
0. 11
0.47
5.23
1.00
0. 15
0. 07
0. 03
0. 22
0. G7
0. 05
0. 01
0. 01
0.45
0. 01
0. 00
0. 00
0. 03
0. 00
0. 00
0. 00
Si, Ca, Al,
Fe, Na, K,
S, MU, Cu
Filler
Flco
1093213
0. 00
0.00
0. 00
1.43
1.00
0. 00
0. 00
0. 00
0. 00
0. 00
0. 00
0. 00
0. 00
0. 00
0. 00
0.00
0. 00
0. 00
0. 00
0. 00
0. 00
Si
Down low n
1093215
0. 00
0. 00
0. 00
1 . (54
1. 00
0.01
0. 00
0. 00
0. 00
0. 00
0. 00
0. 00
0. 00
0.01
0.00
0. 00
0. 00
0. 00
0.00
0. 00
0. 00
Si
Cenlral Lab
1093211
0. 11
0. 00
0. 37
5. 9(5
1.00
0. 22
0. 00
0. 00
0. 11
0. 75
0. 00
0. 00
0. 00
0. 01
0. 00
0. 00
0. 00
0. 00
0. 00
0.00
0. 00
Si_, Ca, Al,
S, Na, K
Ilazelwood
1093201
0.00
0.00
0.08
1.50
1.00
0.08
0.00
0.00
0.04
0.22
0.00
0.00
0.00
0.05
0.00
0.00
0.00
0.01
0.00
0.00
0.00
Si, Ca
4-21
-------
TABLE 4-7 (Cont.)
SUMMARY OF EDAX ELEMENTAL ANALYSIS FOR 16 AUGUST 1976 FILTER
SAMPLES. ENTRIES ARE RATIOS OF INTEGRAL ENERGY PEAKS
TO THE INTEGRAL ENERGY PEAK OF GOLD
Elements
Na
MB
Al
Si
Au
S
Cl
Ccl
K
Ca
Ti
Cr
Mn
Fe
Co
Ni
Zn
Cu
ll£
As
Pb
Important
Klcments
In Hank
Order
No. Braddock
1093203
0. 11
0. 09
0. 24
1. 70
1 . 00
0.24
0. 09
0. 0*
o. 1:3
o. 31
0. 07
0. OS
o. on
0. 10-
0. 0(1
0. Of)
0. 01
0. ()fi
0.01
0. 0,3
0. 04
Si, Ca, Al,
S, K, Na,
Fe, Cu, Mg,
Ph, As, Cl,
Cd^Cr, Ti
Duquesne II
1093205
0. 00
0.00
0. 07
1. 17
1.00
0. 12
0.00
0.00
0,04
0.20
0. 00
0. 00
0. 00
0. 13
0. 00
0. 00
0. 00
0.01
0.00
0.00
0. 00
Si, Ca, Fo,
S
Filter
Liberty Horn
1093209
0.00
0.00
0.07
1. Gl
1. 00
0. 01)
0. 00
0.00
0.05
0. 28
0. 00
0. 00
0. 00
0. 05
0. 00
0.00
0. 00
0. 00
0. 00
0. 00
0.00
Si, Ca
Clairlon
1093211
0.00
0.00
0.00
0.32
1 . 00
0.00
0.00
0.00
0.00
0.04
0. 00
0. 00
0.00
0. 19
0.00
0.00
0.00
0. 00
0.00
0.00
0.00
Si, Fe
Airport
1093229
0.24
0.03
0.32
3.56
1.00
0.2G
O.Ofi
0.05
0.17
0.51
O.OG
0.02
0.01
0.45
0.01
o.oo
0.00
0.04
0.00
0.00
0.00
Si, Ca, Fe,
Al, S, Na,
K, Cu
4-22
-------
TABLE 4-7 (Cont.)
SUMMARY OF EDAX ELEMENTAL ANALYSIS FOR 16 AUGUST 1976 FILTER
SAMPLES. ENTRIES ARE RATIOS OF INTEGRAL ENERGY PEAKS
TO THE INTEGRAL ENERGY PEAK OF GOLD
Elements
Na
Mg
Al
Si
Au
s
Cl
Cd
K
Ca
Ti
Cr
Mn
Fe
Co
Ni
Zn
Cu
Htf
As
Ph
Important
Flements
In Hank
Order
Kilter
Airport
1093230
Cover Filter
0. 14
0.07
0.4G
3. 59
1 . 00
0. 15
0. 01
0. 01
0. 11
0. 55
0. 04
0. 05
0. 02
0.27
0.00
0. 00
0. 00
0. 00
0. 00
0. 00
0. 00
Si, Ca, Al.Fe
S, Na, K
So. Fayette
1093239
0. 00
0. 00
0. 13
3. 79
1.00
0. 03
0. 00
0. 00
0.00
0.48
0. 00
0. 00
0. 00
0. 00
0. 00
0. 00
0. 00
0.00
0. 00
0. 00
0. 00
Si, Ca, Al
4-23
-------
TABLE 4-8
SUMMARY OF EDAX ELEMENTAL ANALYSIS FOR 28 AUGUST 1976 FILTER
SAMPLES. ENTRIES ARE RATIOS OF INTEGRAL ENERGY PEAKS
TO THE INTEGRAL ENERGY PEAK OF GOLD
Elements
Na
Mg
Al
Si
Au
S
Cl
Ccl
K
Ca
Ti
Cr
Mn
Fe
Co
Ni
Zn
Cu
"8
As
Pb
Important
Elements
In Hank
Order
Brighton Twp
1093284
0.08
0.00
0. 3G
5. 36
1 . OU
0. 37
0. 00
0.00
0.06
0.58
0. 00
0. 00
0.00
0.02
0.00
0.00
0. 00
0. 00
0. 00
0.00
0.00
Si, Ca, S,
Al
Hazelwood
1106308
0.22
0. 10
0.39
3. 58
1. 00
0.45
0.02
0.01
0. 15
0.49
0. 04
0.02
0.00
0.08
0.00
0.00
0. 00
0. 01
0. 00
0. 00
0.00
Si, Ca, S.A1,
Na, K, Mg
Filter
Ilazelwood
110G309
Cover 1- ilter
0.24
0. 10
0,45
4. 71
1.00
0. 15
0. 02
0.02
0. 16
0. 64
0. 04
0.01
0.01
0. 12
0. 00
0.00
0.00
0.01
0.00
0.00
0.00
Si, Ca, Al,
Na, K, S, Fe,
Mg
No. Brad dock
1093298
0.20
0. 06
0. 38
3. 32
1. 00
0. 54
0.03
0.07
0. 17
0.77
0.01
0.02
0. 03
0. 35
0. 02
0.01
0.01
0.01
0.00
0.00
0.00
_Si, Ca, S, Al,
Fe, Na, K
No. Braddock
1093299
Cover Filter
0.18
0.09
0.45
3.83
1.00
0. 15
0.03
0.02
0.16
0.66
0.05
0.02
0.01
0.20
0.00
0,00
0.00
0.02
0.00
0.00
0.00
Si, Ca, Al,
Fe, Na, K, S
4-24
-------
TABLE 4-8 (Cont.)
SUMMARY OF EDAX ELEMENTAL ANALYSIS FOR 28 AUGUST 1976 FILTER
SAMPLES. ENTRIES ARE -RATIOS OF INTEGRAL ENERGY PEAKS
TO THE INTEGRAL ENERGY PEAK OF GOLD
Elements
Na
Mg
Al
Si
Au
S
Cl
Cd
K
Ca
Ti
Cr
Mn
Fe
Co
Ni
Zn
Cu
Ilg
As
Pb
Important
hi em en ts
In Hunk
Order
Filter
Duquesne II
1106300
0.27
0.15
0.58
6. 35
1.00
0.12
0.10
0.08
0.30
0.87
0.07
0.06
0.05
0.08
0.03
0.02
0.01
0.03
0.01
0.00
0. 00
Sij Ca, Al,
K, Na, Mg,
S, Cl, Cd, Ti,
Cu
Uuquesne II
1106301
Cover Filter
0. 24
0. 10
0.48
5. 19
1. 00
0. 08
0.04
0. 00
0.20
0.60
0. 04
0.03
0.01
0. 14
0. 00
0.00
0. 00
0.01
0. 00
0.00
0. 00
Si, Ca, Al,
Na, K-, Fe,
Mg
Liberty Boro
1106302
0. 10
0. 04
0.44
3. 65
1.00
0.45
0.02
0.03
0.19
0. 53
0. 06
0. 04
0. 03
0. 14
0.01
0.00
0.00
0.03
0. 00
0.00
0.00
Si, Ca, S.A1,
K, Fe, Na,
Cu
Liberty Boro
1106303
Cover Filter
0.25
0. 06
0.41
4.48
1.00
0.13
0.05
0.03
0. 18
0. 59
0. 05
0.02
0.02
0.43
0. 02
0.00
0. 01
0.01
0. 00
0. 00
0. 00
Si, Ca, Fe,
Al, Na, K,
S
So. Fayette
1093278
0. 06
0. 02
0.37
2.55
1.00
0.43
0.01
0. 00
0. 14
C. 34
0. 03
0.01
0.01
0. 19
0. 00
0. 00
0. 00
0.00
0.00
0.00
0. 00
Si, S, Al,
Ca, Fe, K
4-25
-------
TABLE 4-9
SUMMARY OF EDAX ELEMENTAL ANALYSIS FOR 9 SEPTEMBER 1976 FILTER
SAMPLES. ENTRIES ARE RATIOS OF INTEGRAL ENERGY PEAKS
TO THE INTEGRAL ENERGY PEAK OF GOLD
Elements
Na
iMg
Al
Si
Au
S
Cl
Cd
K
Ca
Ti
Cr
Mn
Fe
Co
Ni
Zn
Cu
II g
As
PI)
Important
Elements
In Rank
Order
Filter
Downtown
1 106323
0. 04
0. 00
0.32
• 5. 66
1. 00
0. 00
0. 00
o. oo
0. 02
0. 02
0. 00
o. oo
0. 00
0.00
0. 00
0. 00
0. 00
0. 00
o. oo
0. 00
0. OH
Sij Ca, Al
Central Lab
110(1327
0.11
0.08
0.41
3. G4
1 . 00
0. 07
0. 0(5
0. 09
0.23
0. 87
0. 10
0. 02
0. 08
0. 35
0. 04
0. 03
0. 01
0. 04
0. 02
0. 00
0. 00
_Hi, Ca, S,
Al, Fe, K,
Na, Ti, Cd,
Cu
Central Lab
110032H
Cover Filter
0. 13
0. 08
0.47
4.43
1 . 00
0.27
0. 07
0. 05
0. 22
1. 00
0. 07
0. 07
0. 07
0.24
0. 04
0. 05
0. 03
0.02
0. 00
0. 00
0. 00
Si, Ca, Al,
S, Fe, K,
Na
Ai rport
1093291
0.16
0.01
0.35
1.36
1. 00
0.28
0.00
0.21
0. 70
0.06
0.00
0. 06
0 . 0 5
0.17
0.01
0. 00
0.01
0.05
0.00
0.00
0.00
Si, K, Al,
{-', Cd, Fe,
Na, Cu
Airport
10 932!) 5
Cover Filter
0.21
0.08
0.44
5.41
1.00
0. 16
0. 13
0.00
0.26
0.90
0.11
0.06
0.05
0.16
0.04
0.00
0.02
0.02
0. 00
0.00
0.00
Pi, Ca, Al,
K, Na, !-', Fe
Cl, Ti
-------
TABLE 4-9 (Cont.)
SUMMARY OF EDAX ELEMENTAL ANALYSIS FOR 9 SEPTEMBER 1976 FILTER
SAMPLES. ENTRIES ARE RATIOS OF INTEGRAL ENERGY PEAKS
TO THE INTEGRAL ENERGY PEAK OF GOLD
Klements
Na
Mg
Al
Si
Au
S
Cl
Cd
K
Ca
Ti
Cr
AIn
Fo
Co
Ni
Zn
Cu
n«
As
1'b
ImporLinL
Klenu'iils
In Hank
Ordrr
Filter
So. Fay e tie
110(5310
o. :n
0. 44
0. 85
2. 30
1. 00
0. (i7
0. SI
0. (H
0. 71
o. 59
0. 49
0. 31
0,25
0. 4 1
0. 2,']
0. 11
0. 15
0. 19
0. 09
0. 07
0. 09
Si, Al, Cl, K,
S, Ctl, Ca", Ti
i\[g, I-'e, Na,
("r, Mn, Co,
Cu. /n. Ni,
4-27
-------
titanium, sulfer, chromium, nickel and chlorine were present on the filter
for 9 September 1976. The South Fayette monitor is located on the roof of
the South Fayette Township High School, and vehicle traffic was very light
on all sampling days except 9 September 1976, which coincided with the
opening of school. Thus, the presence at South Fayette on 9 September 1976
of particles containing elements such as lead, mercury and arsenic probably
can be attributed to vehicle emissions.
4.2.3 Elemental Analysis of Trace Elements
Metals such as cadmium, chromium, cobalt, nickel, mercury, copper,
zinc, arsenic and lead appeared only rarely and in trace quantities on the
hi-vol filters. Although the total contributions of these elements to the
filter mass loadings were small, these elements are of concern because of
their potential for adverse health effects. The SEM-EDAX elemental analysis
of individual particles (see Section 4.2.1 above) showed that only a very
few of the particles with diameters larger than about 5 micrometers contained
any of the trace elements. However, the results of the SEM-EDAX elemental
analysis of entire filter segments (see Section 4.2.2 above) revealed the
presence of trace elements on some of the filters. Because the analysis
of entire filter segments included both small (diameters less than 5
micrometers) and large particles, we conclude that trace elements generally
are concentrated in small particles.
Tables 4-7 through 4-9 identify most of the hi-vol filters containing
trace elements. It should be noted that, if the ratio of the integral
energy peak of a trace element to the integral energy peak of gold was less
than 0.03, Tables 4-7 through 4-9 do not indicate the presence of the
element even if it was actually present. The trace elements found on the
filters are summarized as follows:
• Cadmium, chromium and cobalt were present in trace
amounts at the Midland site on 16 August 1976, at
4-28
-------
the Airport site on both 16 August and 9 September
1976, and at the Liberty Boro site on 28 August 1976;
an elemental analysis is not available for any of
these sites on any of the other sampling days except
for the Liberty Boro site on 16 August 1976 when no
trace elements were found
• Cadmium, chromium, cobalt, nickel and mercury were present
in trace amounts at the Central Laboratory site on 9
September 1976 and at the Duquesne II site on 28 August
1976; however, none of these metals was present at the
Central Laboratory site on 28 August 1976 or at the
Duquesne II site on 16 August 1976 (the filters for
these sites on the remaining sampling days were not
analyzed)
• Cadmium, chromium, cobalt and nickel were present in
trace amounts at the North Braddock site on 16 and 28
August 1976, and mercury, arsenic and lead were also
present at the North Braddock site on 28 August 1976
(the North Braddock filter for 9 September 1976 was
not analyzed)
• Cadmium, chromium, cobalt, nickel, mercury, copper, zinc,
arsenic and lead were present in trace amounts at the
South Fayette site on 9 September 1976 (see Section 4.2.2.
above)
Zinc, which is not actually considered to be a trace element, was
present primarily in the large particles. (Of the individual particles
that were analyzed and found to contain zinc, the largest particle diameter
was 230 micrometers.) In general, only about 2 percent of the particles
were found to contain a non-negligible amount of zinc. However, approximately
4-29
-------
40 percent of the particles selected for analysis from the hi-vol filters
at the Brighton Township site contained zinc. With the exception of a single
particle from the Brighton Township filter for 16 August 1976, all of the
particles containing zinc at the Brighton Township site were found on the
filters for 28 August and 9 September 1976. The prevailing winds on 16
August 1976 were from the north, while the prevailing winds on 28 August
and 9 September 1976 were from the south and southwest. We therefore
conclude that a source of particles containing zinc is south or southwest
of the Brighton Township site. The zinc smelter at St. Joe Minerals,
which is about 2.8 kilometers southwest of the Brighton Township site,
possibly is the source of the particles which contained zinc.
4.2.4. Correlation of Elemental Compositions of
Atmospheric and Slag Particles
Slag produced by the steel industry is found throughout Southwest
Pennsylvania in slag dumps and has been extensively used in road and
building construction. Thus, slag particles are likely to be a primary
constituent of non-traditional fugitive emissions. We therefore performed
an analysis of samples of the different types of slag found in the area to
determine the correlations between the elemental compositions of these slag
samples and the compositions of some of the particles found on the hi-vol
filters. The results of the SEM-EDAX analysis of slag samples from electric
air furnaces, blast furnaces and basic oxygen furnaces are given in detail
in Tables D-227, D-228 and D-229 in Appendix D.
To test for correlations between the elemental compositions of
the slag samples and the elemental compositions of particles found on the
hi-vol filters, the slag samples for the three types of furnaces were
divided into four categories: (1) large slag particles, (2) small parti-
cles adhering to the surface of large slag particles, (3) small slag
particles, and (4) small particles adhering to the surface of small slag
particles. The ratios of the integral energy peaks of ten elements
(silicon, calcium, aluminium, potassium, sulfur, magnesium, iron, titanium,
4-30
-------
chlorine and manganese) to the integral energy peak of gold were used to
characterize the elemental compositions of the various particles. That is,
each particle was characterized by the rank of the peak-to-gold energy
ratios for the ten elements. The Spearman's rank statistic (rank correla-
tion coefficient) was then used to compare the elemental compositions of
particles found on the hi-vol filters with the elemental compositions of
the slag samples. This statistic is given by Mendenhall and Scheaffer (1973)
as
6 T
N(N2-1)
N
Y^ 6 2 (4"5)
where N is the number of elements (ten) and <5. is the difference between
the paired ranks for the i element. Because of the semi-quantitative
nature of the data, p was required to be greater than or equal to the value
given by Conover (1971) for the 99.5-percent comnfidence level before a
positive correlation was assumed to exist. Additionally, only spherical
and agglomerate particles were considered because these particles were
considered most likely to be of industrial origin.
As noted in Section 4.1, each of the various monitor sites was
classified an industrial, urban or rural site. Because only 12 to 20
particles per filter were selected for the SEM-EDAX analysis of elemental
composition, we combined the particles selected from five or six filters
within each site category in order to increase the sample size. We then
used Equations (4-4) and (4-5) to determine which of the industrial
(sphere or agglomerate) particles found on the filters had elemental
compositions which were correlated (at the 99.5-percent confidence level)
with the elemental compositions of the slag samples. The diameters of these
4-31
-------
particles, as determined from the SEM photomicrographs, and the particle
densities for the two types of industrial particles were also used to cal-
culate the fractions of the particulate mass loadings on the filters that
correlated with the slag samples. The corresponding ambient particulate
concentrations which correlated with the slag samples were estimated by
multiplying the total air concentrations by the calculated fractions of
filter mass loadings.
Tables 4-10, 4-11 and 4-12 give the estimated contributions of the
industrial particulates which were correlated with slag samples to the
filter mass loadings and ambient particulate concentrations at industrial,
urban and rural monitor sites. We point out that Tables 4-10 through 4-12
do not include the filter samples which did not show a correlation with
any of the vaious types of slag particles. The filter samples which did
not show any correlation with slag were all found at rural sites (Brighton
Township and South Fayette on both 28 August and 9 September 1976 and Elco
on 28 August 1976). The prevailing winds on 28 August and 9 September 1976
were from the south and southwest. On 16 August 1976, a day with prevail-
ing north winds, the filter samples from the Brighton Township, South
Fayette and Elco sites were correlated with at least some of the slag
samples. Thus, the correlation of filter samples with slag samples at
these rural sites appear to be dependent on the prevailing wind direction.
The effects of wind direction on the filter samples are discussed in
greater detail in Section 4.3.
4.3 WIND - DIRECTION EFFECTS ON HI-VOL FILTER MASS LOADINGS
Meteorological conditions, especially the mean wind directions
during the sampling days, influence the constituents of the hi-vol filter
samples. For example, as discussed in Section 4.2.3, the presence of
relatively high zinc concentrations at the Brighton Township site appears
to be restricted to days with the south winds required to transport parti-
cles containing zinc to the site from a zinc smelter immediately to the south,
4-32
-------
o
1 — 1
1
"*
W
rJ
<£
H
CO
W
H
M
CO
Pi
O
H
I— i
^r
a
1^
r-l
M
Pi
H
CO
p
O
fs
M
£-«
<
Q
£3
o
["T j
CO
W
H
<3^
j— 3
t)
U
H
H
Pi
<^
P-t
! "I
I-H
H
CO
p
Q
53
I-H
CO
W
h-J
CJ
M
H
Pi
<£
P-J
O
<£
t-J
W
PC
H
M
1 — "
o
w
H
<^
, 1
W
Pi
ei
o
u
w
tf
<^
pr{
u
M
ns
>^-
rH
,0
CO
H
to
•H
4J
C
•H
TD
0)
4-J
^
O
a
0)
Pi
CO
CO
CO
£7j
^j
cu
4J
•H
PM
O
CO
4J
rj
CU
CJ
J^
0)
PH
T3
C
CO
to
C
0
•H
4-1
CO
4-1
C
a)
0
0
o
CJ
^
•H
<
rH
rH
<3
•H
4-J
C
CO
3
o-
m
CTv
•
0
0)
i-Cn
4-1
4-1
Cfl
Cfl
CU
rH
P.
g
OS
CO
00
cfl
r~H
CO
0)
4-J
X
4-1
•H
^
T3
cu
4J
tO
rH
CU
)H
^
O
o
cu
P
CO
O
•r-l
ft
^
a)
CO
o
4-1
cu
S-!
CO
i
•H 1 rH
W 0) T-H M
M !H H
cu cu
CJ 4-1 W
!H rH CO
CU <4H -H CO
P-I o p^ s
Air Con-
centration
(ys/m3)
j Location
i
i
j
r~~
i— i in r--
CM rsi
-4-
rH rH CM
CM (M
00
o <3- m
CM CM
l~-
O O rH
CM CM
r^.
rH LO r~-
CM CM
sfr
rH r-< CM
CM CN
[-«
i— i in i--
CM Cx|
vt
rH •—! CM
CM CM
Sphere
Agglomerate
Total
^D
p^
^0
r-H
00
1 Midland
O VO vO
O r> r~
o o o
1 — 1 I— 1
o r^ r^~
o o o
r— 1 I— 1
Sphere
Agglomerate
Total
VD
r^~
^o
i — i
CO
Hazlewood
,
000
CM CM
O CO CO
m u~i
o o o
CM CM
O CO co
in in
o o o
CN CM
O CO CO
m in
o r-- r^.
o o o
CM CM
Sphere
Agglomerate
Total
vO
r»
00
CM
CO
Midland
m
i — I O^ t — l
i-H CM
(N O CM
CO ro
in
<—< O~> rH
r-t CM
CM O CM
CO CO
m
• — i £>
r^.
CO
CM
00
Liberty
Boro
4-33
-------
I
H
<
P
P^
^S
O
Pn
CO
W
H
<
(J
5
u
HH
H
Pi
PH
rj
<
M
Pi
H
CO
^>
Q
Z
M
CO
W
K-l
u
M
H
Pi
<:
p^
o
<
nJ
CO
PC
H
M
3
o
W
H
<3
J
W
Pi
Pi
o
CJ
w
Pi
^**
-i S-J -, 0
Xrrl
o c
J-l
O 3
•H PL,
W
TO
M
01
U
CD
r-t
M
3
PM
4J
cn
03
i—l
PQ
O
Electri
Arc
Particulate
OJ
4-1
TO
Q
Monitor
4-1
a M
0) 0)
O 4J W
V-l rH M
O> 4-j .H TO
PH O Pn S
Air Con-
centration
(yg/m )
4-1
C r-H
0) 01
O 4~i W
r-l i— 1 CO
0) 4-1 -H TO
p* o pti S
Air Con-
centration
(yg/m3)
4-1
c n
CD 0)
O 4-1 W
r-i T-H Cfl
0) 4-( -H TO
PL, O Pn 2
Air Con-
centration
(yg/m3)
4-1
C M
0) 0)
O 4J W
r-l rH W
Ol 4-1 -H TO
PH O fin S
Air Con-
centration
(yg/m )
Location
O CM CN
co m
O CT\ 01
CN CM
O CM CM
CO co
o 01 01
CM CM
O CM CM
CO co
O Ol 01
CM CM
O CM CM
co co
o 01 01
CM CM
Sphere
Agglomerate
Total
8/16/76
rH
0)
&
o.
0
^
CN O CN
CM CM
co r~> o
CM ro
O O O
CM CN
O r^ r^-
CM CM
000
CM CN
o r^ r->
CM CM
CM O CN
CO O CO
Sphere
Agglomerate
Total
vO
r^
VD
rH
CO
Downtown
O CO CO
CM CM
O rH rH
CO ro
O CO CO
CN CN
O < — t i — 1
CO CO
O CN CM
CM CM
o m LO
CM CM
O vD ^D
1 — 1 1 — 1
O CO CO
•— t rH
Sphere
Agglomerate
Total
vO
r^-
vD
i — i
co
Central
Lab
o o- -*
CN CM
O < — I « — I
CO co
o — I
static
CM CM
O •— 1 •— 1
static
CN CN
O •— I •— i
static
CN CM
O •-! i— 1
static
Sphere
Agglomerate
Total
vD
r^
CT\
c^
Downtown
m
O O r-H
<± -j
r-l
•— ( O> O
OO CT.
LO
0 0 rH
-------
Csl
rH
|
-^f
W
nJ
<£
H
CO
W
H
M
CO
Pi CO
O W
H rJ
M CJ
r^H 1 *
O H
JleC P^
<£
f-J P^
0? 0
Pi rJ
CO
H
<; re
H
Q M
2 13
JH)
0 0
PH Ld
H
CO i
H a 1-1 H
Basic Oxygen
Furnace
Blast Furnace
Electric
Arc
Particulate
4-J
a M
0) 01
o 4J en
M i-i tn
a) LM -H n)
P-l O PH S
Air Con-
centration
(yg/m )
—
4J
C 1-1
01 (U
O 4J W
S-l r-l W
0) M-I -H 03
PL, 0 f^ S
Air Con-
centration
(yg/m )
4-J
a J-i
QJ 0)
O 4-1 t/)
M i-l M
J
^O O VD
O ^O
i — i
00
Brighton
Township
CO
.-1 vo 00
en en
CM
•-t -<)- m
CM CN
OO
^H VO OO
cn en
CM
I-H sf m
CM CM
CO
r-H VO 00
cn cn
CM
r-n -d- m
CN CM
oo
•
r-l v£) OO
en en
CM
r-i en
CM en
o> O cr>
t— 1 rH
r> r^ -J-
i — i
Sphere
Agglomerate
Total
^D
r^
^o
i — i
CO
Airport
O vfl ^O
CM CM
O 00 CO
I— 1 1— 1
o en cn
CM CM
O ^D v£>
1 — 1 1 — 1
o -
Sphere
Agglomerate
Total
v£>
r-s
vr>
i— i
00
South
Fayette
o in m
en en
O CT\ Oi
en on
o in in
on on
O O ON
cn on
o m m
on on
o o\ ON
en m
o m m
on en
O CT\ O\
m en
Sphere
Agglomerate
Total
vO
r^
oo
CM
00
Airport
oo
o i — t
m >xj
vO
o oo o,
en en
en
rH CT. O
r-> oo
Sphere
Agglomerate
Total
kO
r^
CT\
CTv
Airport
4-35
-------
Consequently, to examine the effects of wind direction on the composition
of the filter mass loadings, we used the results of the optical and SEM
filter analyses to estimate the percentage contributions of the various
particle generic categories to the filter mass loading at each of seven
monitor sites on each of the three sampling days for which optical and SEM
analyses were available. Table 4-13 gives the results of the optical
analysis and Table 4-14 gives the results of the SEM analysis. (The FeO
category in Table 4-13 consists of both magnetic and non-magnetic iron
oxide.) The prevailing wind directions on the three sampling days are also
shown in the two tables.
Tables 4-13 and 4-14 indicate that, at some of the monitor sites,
the constituents of the hi-vol filter samples during days with north winds
differ significantly from the constituents on the days with south winds.
For example, the contribution of flyash to the filter mass loading at the
rural Brighton Township site is significantly higher on the sampling day
with north winds than on each of the two sampling days with south winds,
while the contribution of quartz is significantly higher on each of the
days with south winds than on the day with north winds. The flyash
particles may possibly be attributed to emissions from sources in the
industrialized Beaver Valley; the quartz particles may have originated from
the field south of the site, from construction activities south of the site
or from dust resuspended by vehicle traffic south of the site. The rural
Elco site shows trends with respect to flyash and quartz particles that are
similar to, but less distinctive than, those of the Brighton Township site.
The industrialized Pittsburgh area is north of the Elco site and Washington
County Route 88 is south of the site. The contribution of iron oxide to the
filter mass loading at the Duquesne II site is significantly higher on the
day with north winds than on each of the two days with south winds. This
iron oxide, which was almost entirely magnetic iron oxide, might be
attributed to emissions from a steel mill located north or northwest of the
site. The Edgar Thompson, J & L (Pittsburgh) and U. S. Steel Homestead
Works are candidate sources of iron oxide particles at the Duquesne II
site. The flyash contribution at the Liberty Boro site does not show a
4-36
-------
CO
w
rJ
PO
01
PH
w
H
h— 1
Pn
Q
W
^
i-j
2
r4
U
M
H
P-i
O
Q
W
H
CJ
W
W
CO
O
CO
CO
^
K^
pq
H
U
PS
w
P-I
w
UJ
Cp
•^
C
•H
T3
cO
o
rJ
0)
I J
rH
•H
MH
O
&-S
4-1
si
60
•H
0)
£2
M
0)
4-1
rH
•H
W
n
*
rH
CO
U
60
O
rH
O
•H
pq
N
t-l
CO
3
0-
0
O>
[> !
rC
Uj
cO
1 1
W
x— N cO
60 01
J= 2
•
a
,—v rH
60 cd
6 U
V '
C
o
2 'H
H a
S 0)
S-j
•H
p
01
4-J
CO
P
M
O
1 J
•H
C
O
^r|
r^ Csl 00 f~°^ f — " ON
i— I O *O rH i — I CO - r-r r-^ t-~
<-* ^ ^ 1^-4 ^- ^*
1 CO CO I CO CO
S 1 1 "2. \ 1
53 CO CO Z CO CO
^O ^O vvO vO
r^« r^ vo r*^ r^ ^o
— ^- ^ — . r^ ^^^ ~~-^ r^
^O OO ~^ ^D OO "^-
i— 1 CM ON rH CM ON
CO CO ON CO CO CTN
CO, CU CU
H H H
C d C
O O O
4-J 4-1 4-1
C ti C rC ,C £.
0) Ol 0) 60 60 60
T3 T3 T3 *H *H -H
cfl CO cO V-i M S-J
pp pq pQ pp pq pq
rH 00
CO CO
CM CM
00 CO
CM r-
CM
rH O
rH O
O 00
CO OO
^O * ON ^o
r~ •
OO 00 C^
Td r^j ""O
o o o
o o o
5 ^ 5
rH rH rH
0) QJ OJ
N N N
cO CO cd
K K E
CM
m r-i o
^D in o
>D CO ~-.
rH CM ON
CO CO ON
A! ^ X
a u u
o o o
T3 T3 T3
TJ 13) ^3
CO CO CO
M M M
pq pq pq
• • •
O O O
^ y^ S2
o
rH
rH
MD
^J-
CM
00
CM
OO
m
m
rH
CO
CM
r-
m
rH
Z;
1
*-?•
vD
r^-
"^^
vO
rH
CO
M
H
Q)
C
W
QJ
p
cr
;3
P
CJN
vD CO
VO vD
CO CO
v£) CM
--
CM ON
OO ON
0 O
O O
pq pq
i>> ^
4-1 4-J
!H i-i
0) 0)
o n
•r-l -r-l
rJ W
4-37
-------
•
W
co
Q
H
U
W
W
CO
CO
CO
pq
H
S3
W
w
CM
CO
w
cO
^
>>
60
d
•H
'O
CO
O
^
OJ
4-1
rH
•H
MH
O
4-1
t~;
60
•H
Q)
(-1
0)
4-1
, — (
•H
1
•
(1)
4-1
CO"
S-i
g
O
, — |
60
60
S-i
CO
t — (
3
J-.A
ou
0)
£
M
QJ
ni
UJ
ex
CO
w
-^ cfl
60 QJ
-5 s
t
O
x— "S r— • f
60 tO
6 U
c
o
3 -H
q 4-*
H a
s CD
•H
Q
0)
4-1
CO
O
}-i
0
4-1
•H
c
o
s
CN to r^*- cr\ oo *^D
• cO
CN LO - •vj* ^*D CN cO CO
O P"-« LO f*0 ' — i *vl"
r*- co LO i — i
LO LO ^-O LO •"-* CNJ
1 — i
- f^« CO VD l/~) r-H
i — i o CN i— < T-H
i — i
^ 13 k2 p^; ^g ^g
I en co i co oo
s i i s i i
S en co 5 co co
^^
r>. i^*. r^> r*> r~-> r^-
"**^ '*^* "^^ *-••». *^^. ~^*^
\O CO CT> ^O CO CT»
rH CN **^ i-H CN -^
^^_ ~- — O^ "^-x ^--. O>
co oo co co
J5 ^ J5
fn ^-i ^H
C (3 C
O 0 0
4J 4-J 4-1
Cr^ r~* r^ r~* i"1
i— t 1— ( A- 1 i-i—i |J— <
a) aJ a) 60 60 oo
^ TJ 13 *H -H »H
cO cO CO S-i >-i !-i
pq pq pq pq pq pq
CO *O CN
^O CD ^O
LO CO --3*
ON CN CN
00 CN ON
CO vo ^cj"
r^ *~~i r^
-O- ON 00
i — I i — i
\£) CO CN
CN r-^. co
F— 1 1— 1 !— I
f-^ h-, l->
^33
1 CO CO
S 1 1
Z CO CO
^^
r^* r — r^>
— ^ ^--^ **~^
vo CO ON
r-t CN -^
-~ ^^^ Q^
OO CO
O O 0
o u u
i-H r-l i-H
WWW
r^» r^-
•^^ •^^^ -*
vO CO O"*\
i — ( CN "*--
"^-^ ""^— C^
CO CO
T3 T3 "^
0 0 O
000
r-H i-H i-H
a) a) a)
N N N
CO CO cO
ffi W ffi
O OO ON
co to co
CN -<]" LO
—1 >0 l~s
CN OO CO
r^ ^- co
ON VO CO
LO ON > t^ r^
-^^ ~-^. ' —
VD CO ON
*-< CN ~^
^^ ^ 0\
CO OO
^ ^ ^
o a u
0 0 O
13 13 TJ
tO cO cO
J_l J_( ^_j
pq pq pq
• • •
O O O
CN LO CO
\J^ vO
r^ p^ p^
~~ — **-^ ^ —
O OO ON
r-H CN \
-~~, -^ ON
OO OO
1— 1 M H
M M M
0) Q) 0)
a c c
CO W W
CL) CU C1J
333
cr tr cr
fS Q Q
OO ON co
OO -^r ON
*d" Si" ^^D
O CN O
r-- o ?*» ^\
4-1 4-1 4J
r-4 M l~<
ClJ <1J CL)
X J3 ,Q
•rl -H -H
4-38
-------
strong dependence on the wind direction during the sampling days, a result
that is consistant with the fact that the site is downwind of one or more
industrial source complexes for almost all wind directions.
4.4 DIFFERENCES IN FILTER SAMPLES BETWEEN
MONITOR SITE CLASSIFICATIONS
Table 4-15 shows the average particulate concentrations by moni-
tor site classification and generic particle category- This composite
summary was developed from the optical microscopy data presented in
Tables D-16 through D-88 in Appendix D. As might be expected, the highest
ambient particulate concentrations are found at industrial sites and the
lowest ambient particulate concentrations are found at rural sites. On the
average, the contributions of combustion products (flyash) to ambient
particulate concentrations at industrial, urban and rural sites are 68, 50
and 36 percent, respectively. Similarly, the average contributions of iron
oxide particles to ambient particulate concentrations at industrial, urban
and rural sites are 7, 4 and 2 percent, respectively. Thus, particles of
industrial origin (combustion products and iron oxide) significantly affect
ambient particulate concentrations at all types of sites, but are of
greatest importance at industrial sites. Non-industrial particles (bio-
logicals, quartz and calcium carbonate) account for over 50 percent of
ambient particulate concentrations at rural sites and for 25 to 30 percent
of ambient particulate concentrations at industrial and urban sites.
4.5 SUMMARY OF THE RESULTS OF THE FILTER ANALYSES
The principal result of the filter analyses described in this
section is that particles with diameters above 25 micrometers account for
over 50 percent of the mass loading on hi-vol filters in the Southwest
Pennsylvania area. In general, these large particles appear to be of indus-
trial origin. The estimated average contributions of industrial particles
4-39
-------
TABLE 4-15
COMPOSITE SUMMARY OF AVERAGE PARTICULATE CONCENTRATIONS
BY SITE CLASSIFICATION AND GENERIC PARTICLE CATEGORIES
Parameter
Number of Filters
Number of Sites
Average Filter Weight* (mg)
3
Average Air Concentration* (yg/m
Combustion Products
Iron Oxide
Biologicals
Quartz and CaCO
TOTAL
Parameter Value
Industrial
Sites
12
6
350 ± 170
130 ± 70
14 ± 19
30 ± 40
18 ± 16
190 ± 100
Urban
Sites
5
2
160 ± 80
70 ± 20
5+4
12 + 11
30 ± 30
140 ± 50
Rural
Sites
12
4
190 ± 70
40 ± 20
2+3
30 ± 40
30 ± 30
110 + 50
* The average filter weight or ambient air concentration plus or minus
one standard deviation.
4-40
-------
to the total ambient particulate concentrations at industrial, urban and
rural monitor sites are 75 percent, 55 percent and 40 percent, respectively.
The elemental compositions of many of these large particles are correlated
at the 97.5-percent confidence level with the elemental compositions of
the slag which is produced by the steel industry and is found throughout
Southwest Pennsylvania in slag dumps. This slag also has been extensively
used in road and building construction. Non-industrial particles (biologi-
cals, quartz and calcium carbonate) have estimated average contributions
to the total ambient particulate concentrations at both industrial and
urban monitor sites and at rural monitor sites of 25 to 30 percent and 55
percent, respectively. Trace elements are almost entirely restricted to
particles with diameters less than 5 micrometers. At many monitor sites,
the presence in the filter samples of certain elements or generic particle
categories appears to be determined in part by the mean wind direction
during the sampling period, a result that suggests that specific sources
are responsible for the presence of these elements or generic particle
categories. However, the data generally are far too limited to define any
specific source - receptor relationships.
4-41
-------
SECTION 5
REVISION OF THE PHASE I
EMISSIONS INVENTORY
5.1 EMISSIONS FROM INDUSTRIAL SOURCES AND SOURCE COMPLEXES
Our review of the Phase I particulate emissions inventory for
the Southwest Pennsylvania AQCR revealed several deficiencies that could
contribute to errors in diffusion-model calculations of particulate air
quality. For example, critical parameters such as stack heights were
missing for some sources, and some fugitive sources were represented as
surface-based point sources. With the assistance of the Pennsylvania
Department of Enivronmental Resources (DER) and the Allegheny County
Bureau of Air Pollution Control (BAPC), we obtained most of the missing
parameters. Also, the fugitive sources were converted for modeling
purposes from point sources to building sources. Inspection of the
locations given for the various sources showed that all of the individual
sources within several of the lage steel complexes were assumed to be at
the same point, whereas the actual separation of the sources was as much
as several kilometers. We acquired plant layouts for these source com-
plexes from the DER or BAPC and estimated the actual locations of the in-
dividual sources. Additionally, as explained in Section 6.1, some modifi-
cations in source parameters were made for the prupose of simulating the
behavior of emissions from some of the source types. For example,
new procedures were developed to calculate plume rise for the coke oven
emissions.
The particulate emission rates for coke ovens and quench towers
are major differences between the Phase I and Phase II emissions inventories.
The emission factors assumed for the coke ovens at the Clairton Coke Works
were:
5-1
-------
• Charging - 0.3 pounds per ton of coal
0 Pushing - 1.2 pounds per ton of coal
• Leaking doors - 0.6 pounds per ton of coal
The emission factors assumed for the coke ovens at all other steel facili-
ties were
• Charging - 0.8 pounds per ton of coal
« Pushing - 1.2 pounds per ton of coal
• Leaking doors - 0.6 pounds per ton of coal
The emission factor for quench tower emissions included both the particulates
associated with the quenching process and the suspended solids contained in
the quench water. Based on the water quality of the Monongahela River,
the total emission factor for quench towers was estimated to be 3.5 pounds
per ton of coal. The above emission factors for coke oven and quench tower
emissions were provided by Bloom (1977).
5.2 EMISSIONS FROM CLASSICAL AREA SOURCES
The area sources used in this study were divided into four major
categories:
• Mobile sources
• Vehicle reentrainment
• Space heating
• Industrial and commercial sources excluded from the
industrial/commercial emissions inventory
5-2
-------
5.2.1 Mobile Sources
The four types of mobile sources that were considered in
developing the area source emissions data are:
• Motor vehicles
• Railroads
• River vessels
• Aircraft
Daily motor vehicle mileage data for Pittsburgh and the six
counties comprising the Southwest Pennsylvania AQCR were obtained from
EPA Region III and the Southwest Pennsylvania Regional Planning and
Commission (RFC). Both sets of data are for calendar 1972. The EPA
data are allocated by type of vehicle. The RFC data are in the form of
total daily vehicle miles per traffic zone and traffic-zone area. Table
5-1 presents the EPA data and total daily vehicle mileage estimates
developed from the RFC traffic zone data. The total dialy mileage
estimates given in the table for the two data sets differ by less than 1
percent. Differences in the totals for the individual areas range from
about 25 to 16 percent.
Estimates of the partciulate emissions from the exhaust and
tire wear of motor vehicles were obtained by using the EPA vehicle
mileage data by vehicle types in Table 5-1 with the following emissions
factors published by EPA (AP-42, Supplement No. 5, December 1975):
• 0.54 grams per mile for light duty vehicles
• 1.21 grams per mile for heavy duty gasoline powered
vehicles (assuming 6 tires per vehicle)
5-3
-------
TABLE 5-1
1972 DAILY VEHICLE MILEAGE FOR THE SOUTHEWEST PENNSYLVANIA
INTRASTATE AIR QUALITY CONTROL REGION
Area
Pittsburgh
Allegheny
County
Butler
County
Armstrong
County
Westmoreland
County
Washington
County
Beaver
County
TOTAL
(All Areas)
EPA Retion III Data
Light Duty
Vehicles
3,458,169
13,578,293
2,086,405
1,029,916
4,931,992
3,015,601
1,905,955
26,548,162
Heavy Duty
Vehicles
177 ,722
450,671
34,133
44,779
138,280
62,049
80,675
810,587
Diesel
Vehicles
66,646
171,627
12,800
17,475
51,215
24,820
30,253
308,190
Total
All Vehicles
3,702,537
14,200,591
2,133,338
1,092,170
5,121,487
3,102,470
2,016,883
27,666,939
SWPRC Data
*A11
Vehicles
3,110,004
13,351,328
2,187,730
936,865
5,438,732
3,263,532
2,347,737
27,498,927
5-4
-------
• 1.70 grams per mile for heavy duty diesel powered
vehicles (assuming 8 tires per vehicle)
Particulate emissions estimates thus obtained for motor vehicles exhaust
and tire wear for Pittsburgh and the six county areas are shown in Table
5-2.
Particulate emissions from the operation of railroads and river
vessels were obtained by using the emission factors published by EPA (AP-42),
Supplement No. 4, December 1975) and fuel consumption data obtained from
the Southwest Pennsylvania RFC. The reported annual fuel oil consumption
is 25 million gallons for railroads and 7 million gallons for river vessels.
The EPA emission factor for railroad locomotives is 25 pounds of particu-
lates per 1000 gallons of fuel; the corresponding factor for commercial
steamships is 20 pounds of paticualtes per 100 gallons of fuel oil. These
data yield estimates of the total particulate emissions from railroads
and river vessels of 312.5 and 70 tons per year, respectively. Assuming
that the total combined railroad and river vessel particulate emissions
of 382.5 tons per year occur in Allegheny County and Beaver County within
1-kilometer corridors along the Ohio, Allegheny and Monongahela Rivers,
the area-source particulate emissions from railroad and river vessels in
these corridors are approximately 2 tons per square kilometer per year.
5.2.2 Space Heating
Particulate emissions from space heating by natural gas were
estimated using the emission factor given by EPA (AP-42, February 1973)
of 19 pounds per million cubic feet. Table 5-3 summarizes the natural gas
sold in the Southwest Pennsylvania AQCR during 1973 by the seven natural
gas companies serving the region (Brown Directory, 1973) and Table 5-4
gives the particulate emissions by county obtained by distributing the
natural gas used for residential and commercial heating on the basis of
population. Because heat for the large buildings in downtown Pittsburgh
5-5
-------
TABLE 5-2
ANNUAL PARTICULATE EMISSIONS FROM VEHICLES FOR
SELECTED AREAS OF THE SOUTHWEST
PENNSYLVANIA AQCR (tons/km2)
Area
(km2)
Pittsburgh
(142)
Allegheny County
(1886)
Butler County
(2057)
Armstrong County
(1689)
Westmoreland County
(2652)
Washington County
(2219)
Beaver County
(1139)
Light Duty
Vehicles
5.22
1.56
0.22
0.11
0.40
0.29
0.36
Heavy Duty
Vehicles
0.61
0.12
0.01
0.01
0.03
0.01
0.03
Diesel
Vehicles
0.32
0.06
0.01
0.01
0.01
0.01
0.02
Total
6.15
1.74
0.24
0.13
0.44
0.31
0.41
5-6
-------
TABLE 5-3
SUMMARY OF NATURAL GAS SOLD IN THE SOUTHWEST
PENNSYLVANIA AQCR IN 1973
Type of Customer
Residences with Heat
Residences without Heat
Commercial
Industrial
TOTAL
Number of Customers
876,182
74,316
68,057
1,031
1,019,586
Quantity of Gas
(10 cubic feet)
154,327
1,639
67,849
182,258
416,916
5-7
-------
TABLE 5-4
PARTICULATE EMISSIONS FROM NATURAL GAS
SPACE HEATING BY COUNTY
County
Allegheny
Lawrence
Beaver
Greene
Fayette
Indiana
Westmoreland
Butler
Washington
Armstrong
TOTALS
Population
1,605,016
107,374
208,418
36,090
154,667
79,451
376,935
127,941
210,876
75,590
2,982,358
Percent of Total
Population
53.8
3.6
7.0
1.2
5.2
2.7
12.6
4.3
7.1
2.5
100.0
Natural Gas
Consumed
(10 cubic feet)
120,416
8,057
15,667
2,686
11,638
6,043
28,200
9,624
15,891
5,595
223,812
Emissions
(tons/yr)
1,144
77
149
26
111
57
268
91
151
53
2,126
5-f
-------
is provided by central steam plants, this area uses relatively small
amounts of natural gas for space heating.
The Mineral Industry of Pennsylvania reports that 64.5 million
tons of bituminous and lignite coal and 4.3 million tons of anthracite
coal were consumed in the Commonwealth of Pennsylvania during 1972. Of
this total of 68.8 million tons of coal, 64.1 million tons were consumed
by industry and utilities. Thus, 4.7 million tons were used by other con-
sumers. The industrial/commercial particulate emissions inventory for
the Southwest Pennsylvania AQCR accounts for 48.0 million tons of coal
per year, or 69.8 percent of the total annual coal consumption in the
Commonwealth. We therefore assumed that 69.8 percent of the 4.7 million
tons of coal used by non-industrial and non-utility sources in the
Commonwealth was used in the Southwest Pennsylvania AQCR in hand-fired
burners. Also, on the basis of coal usage in the Commonwealth, we
assume that 93.75 percent of this 4.7 million tons was bituminous and
lignite, while the remainder was anthracite. The particulate emission
factors published by EPA (AP-42, February 1972) for bituminous and
anthracite coal are 20 and 10 pounds per ton of coal, respectively.
Thus, we estimate that coal usage in hand-fired burners in the Southwest
Pennsylvania AQCR accounted for particulate emissions of 0.03 million
tons per year, which is equivalent to 2.7 tons per square kilometer per
year.
The Bureau of Statistics, Research and Planning of the Common-
wealth of Pennsylvania reports that 384,636 thousand gallons of fuel oil
were distributed in the Southwest Pennsylvania AQCR during 1974. The
industrial emissions inventory accounts for 254,019 thousand gallons
for consumption by residential and commercial users. For the emission
factor published by EPA (AP-42, February 1972) of 15 pounds per thousand
gallons of fuel oil consumed by commercial burners, the total emission
rate is 980 tons per year or 0.08 tons per year per square kilometer.
5-9
-------
5.2.3 Aircraft Emissions
Particulate emissions were estimated for the Greater Pittsburgh
Airport and the Allegheny County Airport using the emission factors published
by EPA (AP-42, April 1973) and the aircraft operations by type reported
for FAA operated towers during 1975. Table 5-5 gives the aircraft operations
reported for each airport. Aircraft operations at the Greater Pittsburgh
Airport were assumed to be allocated as follows:
• Air Carrier —
10 percent Heavy Aircraft
60 percent Long-Range Aircraft
30 percent Medium-Range Aircraft
• Air Taxi —
5 percent Helicopter
47.5 percent Turboprop
47.5 percent Piston
• General Aviation —
20 percent Jet
20 percent Turboprop
60 percent Piston
• Military —
80 percent Transport
20 percent Jet
The estimated emission rate is 147.3 tons per year for aircraft operations
at the Greater Pittsburgh Airport.
5-10
-------
TABLE 5-5
AIRCRAFT OPERATIONS REPORTED BY THE FAA FOR 1975
Type of Aircraft
Operations
Air Carrier
Air Taxi
General Aviation
Military
Number of Aircraft Operations
Greater Pittsburgh
Airport
172,331
54,230
46,179
12,425
Allegheny County
Airport
5
566
168,378
1,496
5-11
-------
Aircraft operations at the Allegheny County Airport were assumed
to be allocated as follows:
• Air Carrier —
100 percent Medium-Range Aircraft
• Air Taxi —
50 percent Turboprop
50 percent Piston
• General Aviation —
80 percent Single-Engine Piston
10 percent Turboprop
10 percent Twin-Engine Piston
• Military —
50 percent Transport
50 percent Helicopter
The estimated emission rate is 2.8 tons per year for aircraft operations
at the Allegheny County Airport.
5.3 REENTRAINMENT AND RESUSPENSION OF ROAD DUSTS BY MOTOR VEHICLES
Estimating the particulate emissions due to reentrainment and
resuspension of road dust by motor vehicles is a difficult task since it
requires a knowledge of the type of road, the dust loading of the road,
the silt content of the dust and vehicle travel rates. Cowherd and Man
(1976) relate dust loading on paved roadways to particulate emissions by
the empirical relationship where e is the emission factor (pounds per
vehicle mile), K is a constant which depends on the particle size, L is
the surface dust loading in pounds per curb mile and s is the fractional
silt content of the road dust. Sarter and Boyde (1972) report that dust
5-12
-------
loading on roadways varies with respect to land use patterns in the
surrounding locale. Table 5-6 summaries the results of field measurements
of dust loading for roads in 12 cities.
Within the Southwest Pennsylvania AQCR, dust loading of road-
ways varies from extremely heavy in the vicinity of industrial source
complexes to relatively light in commercial and residential areas; all
three types of areas may be located within a few kilometers of each other.
Traffic data of sufficient detail to assign vehicle mileage to each road-
way or type of roadway do not exist and could not be developed within the
scope of this study. Vehicle mileage data were obtained from the
Southwest Pennsylvania Regional Planning Commission for the entire
Southwest District. These data consist of vehicle mileage traveled
within each of 958 traffic zones and the size of each traffic zone. The
traffic zones vary in size from several hundred square meters in the
downtown Pittsburg area to several square kilometers in the outlying
areas of the district. For those areas of interest, the traffic zones
were combined into areas of approximately equal traffic density to reduce
the number of sources that were to be considered and to provide geograph-
ical representations of the areas for use in the diffusion model
calculations. The vehicle mileage data and size of each traffic zone
were used to produce a traffic density (i.e., vehicle miles per square
kilometer) for each traffic zone. An average traffic density for each
of the defined areas to be used in the diffusion modeling was calculated
from the traffic density data for each traffic zone within the area. For
each small area of heavy traffic density, a second area was defined and
superimposed on the larger area of average traffic density, and the
difference between the average and heavy traffic densities was assigned
to the second area. Table 5-7 presents the defined areas, their dimen-
sions and the associated traffic densities.
To evaluate the effects on particulate air quality of the
above emission factors for reentrained dust from paved roads, we used the
5-13
-------
TABLE 5-6
DUST LOADING ON PAVED ROADS AS A FUNCTION OF LAND USE
(SARTER AND BOYDE, 1972)
LAND USE
Residential*
Low/old/single
Low/old/multi
Med/ new /single
Med/old/single
Med/old/multi
Industrial
Light
Medium
Heavy
Commercial
Central Business District
Shopping Center
Dust
Loading
(Ib/curb
mile)
Weighted
Minimum
120
31
180
260
140
260
280
240
60
63
Maximum
1,900
1,300
1,200
1,900
6,900
12,000
1,300
12,000
1,200
640
Mean
850
890
430
1,200
1,400
2,600
890
3,500
290
290
Mean
1,200
2,800
290
*By average income level/neighborhood age/type of dwelling
5-14
-------
TABLE 5-7
UTM COORDINATES, DIMENSIONS AND TRAFFIC DENSITIES
FOR AREA SOURCES USED IN ESTIMATING VEHICLE
REENTRAINMENT OF ROAD DUSTS
UTM Coordinates at Center
of Area Source (ra)
X
555000
556500
556000
556500
557500
559000
561000
557500
565000
560000
559500
559500
560500
560500
556000
558500
559500
555000
557500
559500
555000
562000
565500
562000
561500
Y
Dimensions (km)
x y
Beaver Valley Grid
4515000
4514500
4512000
4515500
4515500
4515000
4515000
4514500
4513000
4512000
4508500
4506500
4506500
4505500
4508000
4506500
4505500
4505000
4504500
4504500
4503000
4509000
4507500
4507000
4505500
2.0
1.0
4.0
1.0
1.0
2.0
2.0
1.0
6.0
4.0
2.0
1.0
4.0
1.0
1.0
2.0
2.0
1.0
6.0
4.0
3.0 j 3.0
1.0
1.0
1.0
4.0
1.0
1.0
2.0
3.0
1.0
2.0
2.0
5.0
2.0
1.0
1.0
1.0
1.0
4.0
1.0
1.0
2.0
3.0
1.0
2.0
2.0
5.0
2.0
1.0
Traffic Density
(Daily Vehicle
Miles per km )
9565
9565
9565
3941
3941
3941
3941
3941
3941
3941
15953
15953
15953
15953
13418
13418
13418
13418
13418
13418
13418
3496
3496
3496
3496
5-15
-------
TABLE 5-7
(continued)
UTM Coordinates at Center
of Area Source (m)
X
562500
565500
560500
561500
562500
561000
557500
555000
557000
555500
558500
555000
557000
559000
562500
564000
561500
564000
561500
564000
566500
566500
566500
567500
566000
567500
Y
Dimensions (km)
x y
Beaver Valley Grid
4505500
4502500
4504500
4504500
4504500
4502000
4501500
4501000
4497000
4492500
4492500
4490000
4490000
4490000
4497500
4494000
4493500
4492000
4490500
4490000
4498500
4495500
4492500
4490500
4490000
4489500
1.0
5.0
1.0
1.0
1.0
4.0
1.0
5.0
1.0
1.0
1.0
4.0
3.0 3.0
2.0
6.0
3.0
3.0
2.0
2.0
2.0
5.0
2.0
3.0
2.0
3.0
2.0
3.0
3.0
3.0
1.0
2.0
1.0
2.0
6.0
3.0
3.0
2.0
2.0
2.0
5.0
2.0
3.0
2.0
3.0
2.0
3.0
3.0
3.0
1.0
2.0
1.0
Traffic Density
(Daily Vehicle
Miles per km )
3496
3496
8248
8248
8248
8248
8248
8248
3709
3709
3709
3709
3709
3709
8527
8527
8527
8527
8527
8527
8666
8666
8666
8666
8666
8666
5-16
-------
TABLE 5-7
(continued)
UTM Coordinates at Center
of Area Source (m)
X
571000
571000
577500
577500
577500
582500
597500
591750
595250
591750
595250
591750
595250
587750
589250
582500
582000
584500
584500
586250
588750
587000
584500
589500
589500
588500
586000
586000
Y
P:
4484500
4477000
4485500
4480500
4475500
4485500
4485500
4485500
4485500
4480500
4480500
4475500
4475500
4482500
4482500
4480500
4475500
4473750
4475250
4474500
4474500
4477500
4477250
4478250
4476750
4480500
4482000
4480000
Dimensions (km)
x y
-ttsburg Grid
7.0
8.0
5.0
5.0
5.0
5.0
5.0
3.5
3.5
3.5
3.5
3.5
3.5
1.5
1.5
5.0
4.0
1.5
1.5
2.5
2.5
3.0
1.5
1.5
1.5
3.0
2.0
2.0
8.0
8.0
5.0
5.0
5.0
5.0
5.0
5.0
5.0
5.0
5.0
5.0
5.0
1.0
1.0
5.0
5.0
1.0
1.0
3.0
3.0
4.0
1.0
1.0
1.0
3.0
2.0
2.0
Traffic Density
(Daily Vehicle
Miles per km )
7761
7761
9311
15268
15268
9879
9879
6383
6383
18829
18829
295Q9
29509
19993
19993
19993
24969
24969
24969
24969
24969
65899
108456
25184
25184
25184
19993
19993
5-17
-------
TABLE 5-7
(continued)
UTM Coordinates at Center
of Area Source (m)
X
589500
586000
588000
589500
587500
587500
587500
587500
591000
593000
595000
597000
599000
595000
601000
1
603000
605000
606500
606500
603500
601500
604500
606500
606500
606500
595000
603500
601500
604500
606500
606500
606500
Y
Dimensions (km)
x y
Clairton Grid
4471500
4471000
4471000
4470500
4467500
4462500
4457500
4452500
4471000
4471000
4471000
4471000
4471000
4465000
4471000
4471000
4471000
4471500
4470500
4466500
4461500
4461500
4462500
4461500
4460500
4455000
4456500
4451500
4451500
4452500
4451500
4450500
1.0
2.0
2.0
1.0
5.0
5.0
5.0
5.0
2.0
2.0
2.0
2.0
2.0
10.0
2.0
2.0
2.0
1.0
1.0
7.0
3.0
3.0
1.0
1.0
1.0
10.0
7.0
3.0
3.0
1.0
1.0
1.0
1.0
2.0
2.0
1.0
5.0
5.0
5.0
5.0
2.0
2.0
2.0
2.0
2.0
10.0
2.0
2.0
2.0
1.0
1.0
7.0
3.0
3.0
1.0
1.0
1.0
10.0
7.0
3.0
3.0
1.0
1.0
1.0
Traffic Density
(Daily Vehicle
Miles per km )
15347
15347
15347
15347
10578
10578
1698
1698
10690
10690
10690
10690
10690
10825
15696
15696
15696
15696
15696
4193
4193
4193
4193
4193
4193
4769
4769
4769
4769
4769
4769
4769
5-18
-------
diffusion modeling techniques described in Section 6.2 to calculate 1975
annual average particulate concentration for the area along the Ohio River
from Sewickley through the City of Pittsburgh and along the Monongahela
River to Swissvale. Table 5-8 gives the source inputs for the area sources
used in the model calculations. All of the vehicle mileage was assigned
to a commercial classification with a dust loading factor L of 290 pounds
per curb mile. The silt content s was assumed to be 10 percent and the
-4
constant k was set equal to 1.5 X 10 , yielding an annual emission rate
-4
of 7.9388 X 10 pounds per vehicle mile.
Figure 5-1 shows the calculated isopleths of annual average ground-
level particulate concentration attributable to emissions from resuspended
roadway dust. Annual concentrations as high as 30 micrograrns per cubic
meter are calculated in downtown Pittsburgh, with concentrations above
10 micrograms per cubic meter calculated in residential areas. If the
average dust loading for residential areas rather than the average dust
loading for commercial areas had been used in the model calculations, the
concentrations in Figure 5-1 would be increased by a factor of 4.1 (see
Table 5-6). Similarly, if the average dust loading for industrial areas
rather than the average dust loading for commercial areas had been used
in the model calculations, the concentrations in Figure 5-1 would be
increased by a factor of 9.7. Thus, resuspended roadway dust may have
a significant impact on particulate air quality in the Southwest
Pennsylvania ACQR, especially in downtown areas with heavy traffic.
5-19
-------
TABLE 5-8
ESTIMATES OF REENTRAINMENT EMISSIONS FROM MOTOR
VEHICLES BY AREA SOURCE FOR PITTSBURGH GRID
Source
No.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
UTM Coordinator at Center
of Area Source (m)
X
571000
571000
577500
577600
577500
582500
587500
591750
595250
591750
595250
591750
595250
587750
589250
582500
582000
584500
584500
586250
588750
587000
584500
589500
589500
588500
586000
586000
Y
4484500
4477000
4485500
4480500
4475500
4485500
4485500
4485500
4485500
4480500
4480500
4475500
4475500
4482500
4482500
4480500
4475500
4474750
4475250
4474500
4474500
4477500
4477250
4478250
4476750
4480500
4482000
4480000
Dimensions (km)
X
7.0
8.0
5.0
5.0
5.0
5.0
5.0
3.5
3.5
3.5
3.5
3.5
3.5
1.5
1.5
5.0
4.0
1.5
1.5
2.5
2.5
3.0
1.5
1.5
1.5
3.0
2.0
2.0
y
8.0
8.0
5.0
5.0
5.0
5.0
5.0
5.0
5.0
5.0
5.0
5.0
5.0
1.0
1.0
5.0
4.0
1.0
1.0
3.0
3.0
4.0
1.0
1.0
1.0
3.0
2.0
2.0
Emissions
(Tons/Year)
345
394
185
303
303
196
196
89
89
262
262
410
410
24
24
397
396
30
30
149
149
628
129
30
30
180
63
63
5-20
-------
c
•rl
C
O
•H
4-J
d 13
QJ a
o o
p) M
O
O T3
QJ
QJ >
4-) d
nj CX
r-l
o o
•H 5-J
nj 4-)
d, W
i—I T3
0)
QJ QJ
r-l -d
I C
13 a
*-i ^ _
3 M
O 3
H W
tO 0)
S-i
QJ
SO O
tfl AJ
QJ CU
> r-l
rrj jD
nj
rH -U
CTJ 3
3 ,0
C -H
rt -u
M-t 0)
o e
W CJ
J2 -H
QJ 3
O }-i
W QJ
•H p.
CJ g
j-J nj
rt VJ
r-f W
3 O
a M
r-l O
rt -H
u 6
w
oi
o
M
5-21
-------
SECTION 6
RESULTS OF THE PHASE II DIFFUSION MODEL CALCULATION
FOR PARTICULATE EMISSIONS FROM INDUSTRIAL SOURCES
6.1 INTRODUCTION
Diffusion model calculations of ambient particulate concentra-
tions for large geographic areas are complicated by many factors.
First, it is extremely difficult to obtain an accurate emissions inventory
because of the many fugitive and non-traditional sources of particulates
such as storage piles, roadways and construction sites. Second, there
are large uncertainties in the emission rates that apply to the various
types of sources. Even for the major industrial sources, the rates are
frequently based on emission factors which may not yield accurate estimates.
Also, these emission factors generally do not consider particles with
diameters greater than 30 micrometers because industrial emission control
systems are assumed to be very effective in eliminating large particles.
Additionally, there are some uncertainties about the collection efficiency
of hi-vol samplers for both large and small particles which in turn
raise questions about the accuracy of particulate air quality data.
The diffusion model calculations described in Sections 6.1 and
6.2 below were restricted to industrial particulate emissions. Although
attempts were made to assess the air quality impact of fugitive emissions
caused by entrainment of dust from paved roads (see Section 5), the
available emission factors are not sufficiently well established to
provide more than order of magnitude estimates of this impact. The
contributions of particulate sources outside of the Southwest Pennsylvania
AQCR to ambient particulate concentrations within the AQCR have been
partly accounted for by including emissions from some large electrical
power plants in Ohio and West Virginia in the diffusion model calculations.
6-1
-------
The effects of gravitational settling are not included in the
calculations described in this report for two reasons. First, controlled
industrial particulate emissions are comprised of small particles
(diameters less than 20 micrometers) which, as discussed in the following
paragraph, have negligible settling velocities. Second, the models that
we have developed and tested to include the effects of gravitational
settling and the partial reflection of particulates at the surface
(Dumbauld, et_ al_. , 1976) can strictly be applied only in flat terrain.
For the industrial particulate sources in the emissions inventory given
in Appendix C, variations in terrain height in the Southwest Pennsylvania
AQCR have a significant effect on calculated ground-level particulate
concentrations, while the calculated effects of gravitational settling
and partial reflection are minimal. Given these mutually exclusive
model options, the significant terrain effects were included in the
model calculations rather than the minimal settling velocity effects.
The dispersion of large particles differs from that of gaseous
pollutants in two ways. First, large particles are brought to the surface
through the combined processes of atmospheric turbulence and gravitational
settling. Second, large particles that reach the ground surface may be
partially or completely retained at the surface. The largest particles
in the industrial emissions inventory have diameters on the order of 20
micrometers, and the maximum density for these particulates is about 3
grams per cubic centimeter. The resulting terminal fall velocity is about
0.07 meters per second (McDonald, 1960). For an effective emission height
of 100 meters and the mean wind speed on 16 August 1976 (one of the sampling
days) of 3.5 meters per second, gravitational settling alone would bring
these particles to the surface at a downwind distance of 8,750 meters.
However, using the mean vertical turbulent intensity on 16 August 1976 of
0.0724 radians, turbulent mixing alone would bring these particles to the
surface within 1,000 meters. Thus, the effect of gravitational settling
in bringing the largest particles in the emissions inventory to the surface
6-2
-------
is almost an order of magnitude less than the effect of turbulent mixing.
For a settling velocity of 0.04 meters per second, Dumbauld, ej^ al_. (1976)
estimate that about 35 percent of the particulates reaching the surface
could be retained at the surface. Thus, the model calculations described
in this report could overestimate ground-level particulate concentrations
produced by industrial emissions at long downwind distances. However, the
relationship between settling velocity and fraction of material retained
at the surface hypothesized by Dumbauld, £t_ aj^. (1976) is speculative.
Also, the fraction of material assumed by Dumbauld, et_ al_. to be deposited
decreases rapidly as the settling velocity decreases. Consequently, the
neglect of gravitational settling and deposition in the model calculations
appears warranted for the small particles in the industrial emissions inven-
tory.
Table 6-1 lists by name the 35 hi-vol monitor sites operated by
the Pennsylvania DER and the Allegheny County BAPC that were used for com-
parisons of concurrent calculated and observed particulate concentrations.
Additionally, Table 6-1 gives both the Universal Transverse Mercator (UTM)
coordinates and the height above mean sea level (MSL) of each hi-vol site.
The monitor sites numbered from 1 through 15 in Table 6-1 were used in the
hi-vol particulate sampling program described in Section 3.
6.2 RESULTS OF THE ANNUAL AVERAGE CONCENTRATION CALCULATIONS FOR 1975
6.2.1 Source and Emissions Data
The development of the particulate emissions inventory used in
this study is discussed in Section 5, and the source and emissions data
for all sources included in the inventory are given in Appendix C in the
form required for input to the diffusion models described in Appendix A.
In general, the source and emissions data in Appendix C were developed
using conventional techniques and reflect actual source parameters.
However, some modifications were made for the purpose of simulating the
6-3
-------
TABLE 6-1
LIST OF MONITOR SITES
Monitor
Numbe r
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
Monitor Name
Batlen
Beaver Falls
Koppcl
Brighton Township
Midland
Elco
Downtown
Central Lab
Hazehvood
North Braddock
Duci ues no II
Liberty Borough T
Clairton
Greater Pittsburgh Airport
South Fayette
Springe! ale
Court House
North Fayette
Braddock
Duquesne I
Allegheny County Airport
Glassport T
Coursin Hollow
Henry Kaufmann
Brownsville
Lover (Charleroi)
Belle Vernon (Hostraver
Township)
Courtney
Amb ridge
Rochester
UTM X
(m)
565,090
557,750
557, OGO
556,790
5 16, 330
594,625
585,150
588,280
589,610
596,680
597,980
596,210
594, 890
566,100
570,330
602, 976
585,060
564,420
596,536
597,786
591,119
594, 190
596,512
587,125
594, 740
590,315
601,580
589,780
565, 130
561,030
UTM Y
(m)
4,498,380
4,510,785
4,520,375
4,506,085
4,498,340
4,437,300
4,476,600
4,476,970
4,473,400
4,472,835
4,469,720
4,464,150
4,461,870
4,483,000
4,469,750
4,489,036
4,476,738
4,476, 800
4,472,452
4,469,464
4,467,214
4,463,580
4,463,488
4,471,985
4,430,405
4,440,820
4,447,660
4,451,300
4,493,600
1,505,640
Elevation
(m MSL)
230
220
287
335
249
312
256
292
248
275
242
340
238
362
390
320
237
349
244
277
389
234
360
342
239
333
309
226
217
271
6-4
-------
TABLE 6-1 (Cont.)
Monitor
Number
31
32
33
34
35
Monitor Name
Vanporl
New Castle
Bessemer
Monessen
Swissvale
UTAI X
(m)
556,900
554, 830
543, 195
595, 205
594, 654
LfTM Y
(m)
-1,503,550
4,537,240
4,535,9CiO
4,445,995
4,474,424
Elevation
(m MSL)
217
257
340
237
280
6-5
-------
behavior of emissions from some of the source types. For example,
virtual temperatures were used to calculate plume rise for the quench
towers. The virtual temperature, which is the temperature of dry air at
the same pressure and with the same density as the effluent, accounts for
the high moisture content of quench tower emissions. As discussed below,
new procedures were also developed to calculate plume rise for the coke oven
emissions.
There are no generally accepted procedures for modeling the air
quality impact of particulate emissions from coke ovens. The coke ovens
form a large heat source with a surface temperature at the top of the ovens
of about 66 degrees Celsius (150 degrees Fahrenheit). We arbitrarily
assumed that the air in contact with the tops of the ovens is heated to
60 degrees Celsius (140 degrees Fahrenheit) and that the air rises from the
tops of the ovens at about 1 meter per second. The product of the horizon-
tal area of the coke ovens and the assumed vertical velocity of 1 meter
per second yields an effective volumetric emission rate which is used with
the assumed exit temperature of 60 degrees Celsius to calculate the heat
flux required for the plume-rise calculations. Although the actual heat
flux of the air rising from the coke ovens may differ by a factor of two
or so from our estimated value, the calculated plume rise varies according
to the cube root of the heat content and thus is relatively insensitive
to the assumed heat flux. In the case of the coke ovens at the Clairton
Coke Works, the calculated stabilization height for a mean wind speed of
5 meters per second is about 100 meters. This compares favorably with
visual observations of emissions from coking operations in the Pittsburgh
area reported by Rubin and Bloom (1974). They report that the mean
stabilization height appears to be about 135 meters and that this effec-
tive stack height for coke ovens yielded the best results in their
calculations made using the Air Quality Display Model (AQDM). Although
we calculated plume rise for each group of coke ovens as a whole, multiple
point sources were used to simulate the coke ovens in order to preserve
the horizontal geometry of these sources.
6-6
-------
6.2.2 Meteorological Data
Meteorological inputs required by the long-term diffusion model
described in Appendix A were principally obtained from the 1975 annual
frequency distribution of wind-speed and wind-direction categories at the
Greater Pittsburgh Airport, classified according to the Pasquill stability
categories. The annual wind statistics were developed from hourly surface
weather observations using the definitions of the Pasquill stability
categories given by Turner (1964), which are based on solar radiation (in-
solation) and wind speed (see Section 2.2.2). Tables B-ll through B-15
in Appendix B give the 1975 seasonal and annual joint frequencies of
occurrence at the Greater Pittsburgh Airport of wind speed and wind
direction, classified according to the Pasquill stability categories.
The E and F stability categories are combined in the tables because we
believe that the effects of surface roughness elements and heat sources
in urban and industrialized areas are incompatible with the minimal
turbulent mixing associated with the Pasquill F stability category. The
Greater Pittsburgh Airport wind data were used in the annual concentration
calculations for the Southwest Pennsylvania AQCR because we believe
that, of the wind data available in sufficient detail for modeling, the
Greater Pittsburgh Airport wind data are most likely to be representative
of the wind flow over the entire region. However, we recognize that the
Greater Pittsburgh Airport surface winds cannot be expected to give an
accurate representation of the low-level wind circulation in some areas.
As explained in Section 2.2.2, the long-term diffusion model
described in Section A.4 of Appendix A requires the following meteoroligical
inputs in addition to seasonal and/or annual wind summaries: (1) wind-
profile exponents by wind-speed and Pasquill stability categories, (2)
vertical turbulent intensities by Pasquill stability category, (3) mixing
depths by wind-speed and Pasquill stability categories, (4) vertical
potential temperature gradients by wind-speed and Pasquill stability
categories, and (5) ambient air temperatures by Pasquill stability category.
The wind-profile exponents and urban vertical turbulent intensities used
6-7
-------
TABLE 6-2
MIXING LAYER DEPTHS IN METERS USED IN THE 1975
ANNUAL CONCENTRATION CALCULATIONS
Pasquill
Stability
Category
A
B
C
D
E
Wind-Speed Category (m/sec)
0-1.5
1248
1248
1248
698
148
1 .6-3. 1
1285
1285
1285
789
293
3.2-5.1
-
1300
1300
976
653
5.2-8.2
-
-
1300
1153
-
8.3-10.8
-
-
1300
1191
-
>10.8
-
-
1300
1191
-
6-8
-------
TABLE 6-3
VERTICAL POTENTIAL TEMPERATURE GRADIENTS AND AMBIENT
AIR TEMPERATURES USED IN THE 1975 ANNUAL AVERAGE
CONCENTRATION CALCULATIONS
Pasquill
Stability
Category
A
B
C
D
E
Potential Temperature Gradient ( K/m)
Wind-Speed Category (m/sec)
0.0-1.5
0.000
0.000
0.000
0.015
0.030
1.6-3.1
0.000
0.000
0.000
0.010
0.020
3.2-5.1
-
0.000
0.000
0.005
0.015
5.2-8.2
-
-
0.000
0.003
-
8.3-10.8
-
-
0.000
0.003
-
>10.8
-
-
0.000
0.003
-
Ambient
Air
Temperature
(°K)
287
287
287
284
281
6-9
-------
to calculate 1975 annual average particulate concentrations are given in
Tables 2-4 and 2-10, respectively. We averaged the seasonal median mixing
depths given in Table 2-11 for use in the 1975 annual particulate concen-
tration calculations. Table 6-2 lists the resulting mixing depths. Table
6-3 gives the vertical potential temperature gradients and ambient air
temperatures used in the 1975 annual average concentration calculations.
The potential temperature gradients given in Table 6-3 are identical to
the potential temperature gradients given in Table 2-13. The average
afternoon temperature measured at the Greater Pittsburgh Airport during
1975 is assigned to the A, B and C stability categories in Table 6-3;
the average of the morning and evening temperatures is assigned to the D
stability category; and the average nighttime temperature is assigned to
the combined E and F categories.
6.2.3 Calculation Procedures and Results
Calculation Procedures
The emissions data for all sources given in Appendix C and the
meteorological data discussed in Section 6.2.2 were used with the long-
term diffusion model described in Section A.4 of Appendix A to calculate
annual average ground-level particulate concentrations at 5-kilometer
intervals on a gross 65-kilometer by 90-kilometer grid and at the locations
of the 35 hi-vol samplers identified in Table 6-1. The procedures
described in Section A. 5 of Appendix A x^ere used to account for the
effects of variations in terrain height over the calculation grid.
Annual concentrations were also calculated at 1-kilometer intervals for
four small grids: New Castle, Beaver Valley, Pittsburgh and Clairton.
The same emissions inventory was used for the gross grid and the small
grids.
The particulate emission rates assigned to the quench towers
at the various steel facilities included both the particulates attributed
6-10
-------
to the quenching process and all particulates and suspended solids con-
tained in the quench water. Because hi-vol samplers may not capture or
retain all of the types of particulates contained in the quench water,
concentrations were calculated both with and without -the effects of
quenching emissions.
Results of the Calculations
The isopleths of 1975 annual average ground-level particulate
concentration calculated without the effects of the quenching process for
the New Castle, Beaver Valley, Pittsburgh and Clairton grids are shown
in Figures 6-1 through 6-4. The calculated concentration isopleth patterns
with the effects of quenching included are very similar to those shown in
Figures 6-1 through 6-4, but indicate slightly higher annual average con-
centrations, especially in the vicinity of large steel works. Table 6-4
gives the 1975 calculated and observed annual particulate concentrations
for the 35 hi-vol monitor sites listed in Table 6-1. We point out that
the observed particulate concentrations in Table 6-4 are annual geometric
mean concentrations, whereas the calculated concentrations are annual
arithmetic mean concentrations. In general, the geometric mean concentra-
tions are expected to be about 10 to 20 percent smaller than the corre-
sponding arithmetic mean concentrations.
It is important to note that the calculated concentrations in
Figures 6-1 through 6-4 and Table 6-4 consider only the effects of emissions
from sources contained in the industrial particulate emissions inventory.
As discussed in Section 5, we also calculated the contributions to ambient
particulate concentrations of emissions from area sources. These emissions
were almost entirely due to the entrainment of dust from paved roads by
motor vehicle traffic. Figure 5-1 shows that these emissions may cause
localized "hot spots" in downtown areas with heavy traffic. However,
current entrainment emission factors depend on the type of roadway and
other assumptions and can vary by a factor of ten or more for the same
traffic volume. Because the calculated concentrations attributable to
6-11
-------
104006
4540
4535
4530
4525
555
5GO
FIGURE 6-J.
Calculated isopleths of 19/5 annual average ground-
level particulate concentration in micrograms per
cubic meter for the New Castle grid without the
effects of quenching.
6-12
-------
FIGURE 6-2. Calculated isopleths of 1975 annual average
ground-level particulate concentration in micro-
grams per cubic meter for the Beaver Valley grid
without the effects of quenching.
6-13
-------
c
•H •
00
a c
O -H
•H ,£
4J O
rt c
a cr
a)
O 4-1
a o
O
O CO
4->
a) o
4-1 0)
rt u-t
iH <4-l
3 4->
a) -H
rH ?
T3 13
C -H
3 !-J
O 00
,
60
QJ !-i
60 3
tfl ^
M W
0) 4-J
> 4J
cfl -H
P-.
H
03 0)
CO J-4
o
in u-i
r^-
cr\ j-i
^) 0)
O 6
en o
,C -H
•H ft
13 03
a) e
4J Cd
cfl )-i
iH 60
3 O
a n
<-t a
0 -rl
o g
6-14
-------
FIGURE 6-4. Calculated isopleths of 1975 annual average ground-level
particulate concentration in micrograms per cubic meter
for the Clairton grid without the effects of quenching.
6-15
-------
TABLE 6-4
OBSERVED AND CALCULATED 1975 ANNUAL AVERAGE GROUND-LEVEL
PARTICULATE CONCENTRATIONS AT THE DER AND BAPC
MONITORS IN THE SOUTHWEST PENNSYLVANIA AQCR
Monitor
Number
1
2
3
4
5
6
1
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
Monitor Name
Baden
Beaver Falls
Koppel
Brighton Township
Midland
Elco
Downtown (County Off.Bldg)
Central Lab
Hazelwood
North Braddock
Duquesne II
Liberty Borough T
Clairton
Greater Pittsburgh Airport
South Fayette
Springdale
Court House
North Fayette
Braddock
Duquesne I
Allegheny County Airport
Glassport T
Coursin Hollow
Henry Kaufmann
Broxvnsville
Lover
Rost raver Township
Courtney (Nex^ Eagle)
Ambridge
Rochester
Vanport
New Castle
Bessemer
Monessen
Swiss va le
3
Concentration 0£/m )
Observed*
132
78
103
•k-k-k
139
74
85
99
97
121
139
107
105
79
56
59
159
61
165
109
73
94
93
77
86
60
50
119
102
88
90
123
200
121
155
Calculated**
W/Quench
83
20
23
25
132
35
33
50
89
55
88
87
77
30
27
47
31
26
43
73
54
69
131
45
12
18
19
20
30
22
24
22
87
31
62
W/0 Quench
58
18
22
22
115
34
27
42
68
49
85
66
69
27
25
45
26
24
38
69
45
54
97
36
10
17
17
18
27
20
21
21
86
30
48
*Geomet;ric mean.
**Arithmetic mean.
***Not available due to excessively high readings caused by building con-
struction and dirt access road.
6-16
-------
area source emissions provide at best an order of magnitude estimate of
the effects of entrainment from paved roads, these calculated concentra-
tions are not included in Figures 6-1 through 6-4 or in Table 6-4.
6.3 RESULTS OF THE 24-HOUR CONCENTRATION CALCULATIONS FOR
SELECTED SAMPLING DAYS
6.3.1 Source and Emissions Data
Of the three sampling days for which detailed hi-vol filter
analyses are available, we selected 16 and 28 August 1976 for the 24-
hour concentration calculations becuase the prevailing winds on the two
days were from different directions. The prevailing winds on 16 August
1976 were from the nortlwest through north, while the prevailing winds
on 28 August 1976 x^ere from the south through southwest. We were not
able to obtain the actual emissions from the major particulate sources
on 16 and 28 August 1976 in order to improve the accuracy of the short-
term diffusion model calculations. In the absence of the actual emissions
data for the two sampling days, we used the average emissions data given
in Appendix C for the industrial sources in the 24-hour concentration
calculations.
6.3.2 Meteorological Data
Tables 6-5 and 6-6 list the hourly meteorological inputs used
in the short-term diffusion model calculations for 16 and 28 August 1976.
These inputs, which are based on Greater Pittsburgh Airport surface weather
observations and upper-air soundings, were developed following the tech-
niques outlined in Section 3 of the report by Cramer, e_t_ a_l_. (1975).
The observed wind directions, wind speeds and ambient air temperatures were
used as direct model inputs. For each hour, the wind-speed and cloud-cover
observations were used to determine the Pasquill stability category
following the Turner (1964) procedures (see Section 2.2.2). The vertical
turbulent intensity was assigned to each hour on the basis of the Pasquill
6-17
-------
in
w
ON
rH
H
g
PH
C/3
H
o
M
O
O
nJ
O
Pi
o
w
H
W
2
O
ffi
TJ
Cfl
4-1 J-l
rH C ^
ca QJ
r-l rH >>
D d 4-1
4-1 r£) 'H
r4 3 C
H 0)
4-1
c
M
TJ
cd
rH 4-1 !-l
0) C — '
o ,
4-1 3 4-1
0) M Cfl
> 3 a
H 0)
4-1
C
M
(I)
rH
•H 4-1
MH C
O Q)
r-l C
1 ft
-a x
a w
•H
£2
mtial
irature
it (°K/m)
4_l £X CJ
0 g -H
p ( nl *ri
H cfl
t-i
O
Ambient Air
Temperature
(°K)
00
c .c
•H 4-1 /-v
X ft g
•H <1) ^
g 0
TJ O
T3 QJ 0)
d a) w
Wind
Direction
(deg)
s-i s~*
•3 H
O W
Ed W
m in
^O VD
v*O ^Q
O O
0 0
m m
st st
o o
0 0
st st
rH rH
0 0
st st
CM CM
O O
o o
00 OO
CM CM
m in
rH rH
0 O
CM CM
o o
rH rH
0 0
0 0
o o
m m
vO *«O
^o ^o
0 0
o o
in in
o o
0 0
rH rH
O O
st st
CM CM
O O
0 0
in in
00 00
CM CM
m m
CM CM
rH rH
P*** vO
m o
CM CM
o o
CM m
0 0
0 0
CO st
o o
in rH
vo m
0 0
0 0
0 0
st st
CM CM
CM vD
st m
O rH
rH rH
r^ CM
rH H
VD st
o o
st st
co CO
o o
o o
in *-o
rH rH
rH rH
0 0
rH rH
0 0
m in
CO CO
r^- f^
0 0
0 O
st st
rH rH
0 O
st CM
o o
o o
o o
st co
ON ON
CM CM
rH in
r~- oo
CM co
rH rH
st CM
rH H
in st
0 O
st st
CO fO
0 0
0 O
r-x oo
rH rH
m m
"*O ^*Q
^O \^£i
o o
0 0
m m
st st
o o
0 0
st st
rH rH
0 0
O CM
O rH
o o
0 0
CM O
ON ON
CM CM
o m
O CM
m CM
rH rH
CM st
H rH
st in
0 0
st st
CO co
0 0
0 0
ON O
rH CM
in m
vD v£5
o o
0 0
in "~i
vD *»D
st st
o o
0 0
st st
rH rH
0 0
co m
CM CO
o o
o o
ON r~»
OO 00
CM CM
o m
in r~-
ON tsD
CM ^D
rH O
st CM
o o
o in
CO CO
0 0
o o
rH CM
CM CM
m m
vO *-O
vO **Q
o o
o o
in m
st st
o o
o o
st st
rH rH
0 0
^O 00
st m
0 0
O 0
CO OO
CM CM
o m
O CM
St rH
st r^-
m in
rH CM
0 0
H m
O CO
0 0
0 0
CO st
CM CM
6-18
-------
W
x£>
rH
H
C/J
p
OO
CN
O
PH
CO
H
)GICAL ]
ILY METEOROLC
B
o
re
cd
4-1 !-l
rH d ^
C3 CU
r-l i — 1 >~>
0) 3 4-1
Cfl r-l M
r-3 3 d
H 0)
4-J
d
M
cd
rH 4-1 V-1
CJ 0)
•H rH >-.
4-J 3 4-1
0) V-l W
> 3 d
H 0)
4-J
d
M
0)
rH
•H 4-1
MH d
O CD
n d
1 P.
C W
jiH
;ntial
^rature
it (°K/m)
4-1 P. CD
0 6 -H
PJ QJ T3
H 03
r-l
O
Ambient Air
Temperature
(°K)
60
d xi
•H 4-J ^-x
« P. E
•H 0) ^
S n
T3 O
13 0) 0)
d cu cfl
ts ui 6
Wind
Direction
(deg)
3 H
O CO
Pd w
LniniOinrHrHrHrHrHrHrHrHrHrHrHrHrHrHrHrHrHrHrHrH
xOxOxOxOLOininininininLomininininininminminm
xDvDxDxOOOOOOOOOOOOOOOOOOOOO
OOOOrHrHrHrHrHrHrHrHrHrHrHrHrHrHrHrHrHrHrHrH
OOOOOOOOOOOOOOOOOOOOOOOO
LoioioioiOLoioinioiOLOLoioioioioioinioioioioioio
xDxDvDiocOcocOcOcOcocococOcocococOcocOCOcOcocOCO
OOOOOOOOOOOOOOOOOOOOOOOO
OOOOOOOOOOOOOOOOOOOOOOOO
cocococococococococococococococococococococococo
rHrHrHrHrHrHrHrHrHrHrHrHrHrHrHrHrHrHrHrHrHrHrHrH
OOOOOOOOOOOOOOOOOOOOOOOO
xDxDxOxOxDxOxDiniO<)- O^* O^ C5> CTt O> CT\ O^ CT\ O> O> O^ CT\ CT> C^ O> CT^ CT* O^ O>
CNCNCNCNCNCNCNCNCNCNCNCNCNCNCNCNCNCNCNCNCNCNCNCN
ocNCNCNcNCNOinoiooinoinoor~.^oooooo
-^•Q>C^C^C^O»^}-Or^COOxDCOCT\xOxDOLOOOOOOO
co CN CN CN CN CN co ~-3~ *J" in xD xO r^ r^« oo oo co r^ ^, r^ r^- r*> r^> i — *
r^r^xOOcNcoo>r^x^-xDcNx4-cocxjcNOO
ininoinOLnoinininxDrHxDxDrHrHxDrHHxDrHrHxOxD
CNCNCNrHCNCNCOCNCNCNCOxj-^j-xDvDininxcfin-xT-xtxd-COCO
OOOOOOOOOOOOOOOOOOOOOOOO
OcocNr^xD-o-ooor^Or^ocNrHcocNcncNcOrHrHr^O
CNCNCNrHrHCNCNCNCNrHCNrHCNCNCNCNCNrHCNCNCNCNCNCO
OOOOOOOrHOOOOOOOOOOOOOOOO
OOOOOOOOOOOOOOOOOOOOOOOO
rHCNCO-O-LnvOI — COCT\CJ5rHCNCO
-------
stability category using the values given in Table 2-10 for urban areas.
The lateral and vertical turbulent intensities were assumed to be equivalent
for a 10-minute averaging period, and the t law of Osipov (1972) and
others was used to obtain the hourly lateral turbulent intensity. That
is, the hourly lateral turbulent intensity was obtained by multiplying
the corresponding vertical turbulent intensity by 1.43 (6 ). The
wind-speed measurements at the standard heights from the 0700 EST and
1900 EST upper-air soundings were averaged to obtain the vertical profile
of average wind speed within the surface mixing layer on each day. A
logarithmic least-squares regression technique (see Equation (3-4) in
Cramer, et al., 1975) was then used to calculate an average wind-profile
exponent for the day.
The 0700 EST and 1900 EST vertical temperature profiles for
15, 16 and 17 August 1976 and for 27, 28 and 29 August 1976 were plotted
on thermodynatnic diagrams. In the absence of a surface-based inversion,
the mixing depth was set equal to the height above the surface of the first
elevated stable layer. If a surface-based inversion existed, the mixing
depth was set equal to the minimum value of 125 meters determined by Cramer,
et_ ail_. (1975) on the basis of an analysis of Pittsburgh mixing depth
statistics. The vertical potential temperature gradients at 0700 EST
and 1900 EST were set equal to the average gradients through the depth
of the surface mixing layer. The 0700 EST mixing depths and vertical
potential temperature gradients were assumed to apply from midnight until
0700 EST on both 16 and 28 August 1976.
On 16 August 1976, the dry adiabat (line of constant potential
temperature) passing through the surface pressure and maximum afternoon
surface temperature intersected the 0700 EST temperature profile at a
height below the 1900 EST mixing depth. Consequently, mixing depths and
potential temperature gradients for the hours between 0700 and 1900 EST
on 16 August 1976 were obtained by linear interpolation. The mixing
depth and vertical potential temperature gradient at 0700 EST on 17
6-20
-------
August 1976 were assumed to apply at midnight, and mixing depths and
vertical potential temperature gradients for the hours following 1900
EST on 16 August 1976 were also obtained by linear interpolation.
On 28 August 1976, the maximum surface temperature of 299
degrees Kelvin occurred during the period 1500 through 1600 EST. The
intersection of the dry adiabat passing through the surface pressure and
this temperature with the 0700 EST temperature profile indicated a
mixing layer depth slightly larger than the mixing depth shown by the
1900 EST sounding. This larger mixing depth was assumed to apply at
1500 EST and 1600 EST, and the vertical potential temperature gradient
during this period was set equal to the near-adiabatic value of 0.001
degrees Kelvin per meter measured at 1900 EST. Mixing depths and potential
temperature gradients for the hours between 0700 EST and 1500 EST on 28
August 1976 were obtained by linear interpolation. Inspection of the
vertical temperature profile at 0700 EST on 29 August showed little
change in mixing depth and potential temperature gradient between 1900
EST on 28 August and 0700 EST on 29 August. Consequently, the mixing
depth and vertical potential temperature gradient were held constant
after 1900 EST on 28 August 1976.
6.3.3 Calculation Procedures and Results
Calculation Procedures
The emissions data for all sources given in Appendix C and
the hourly meteorological inputs listed in Tables 6-5 and 6-6 were used
with the short-term diffusion model described in Section A.3 of Appendix
A, including the terrain adjustment procedures outlined in Section
A.5, to calculate 24-hour average ground-level particulate concentra-
tions for 16 and 28 August 1976. Concentrations were calculated at the
locations of the 35 hi-vol samplers identified in Table 6-1 and at 1-kilo-
meter intervals for the four small grids described in Section 6.2.3.
Concentrations were calculated both with and without the effects of
6-21
-------
emissions from the quenching process because hi-vol samplers may not
capture or retain all of the types of particulates contained in the
quench water.
Results of the Calculations
The 24-hour average particulate concentrations calculated at the
locations of 34 of the 35 hi-vol samplers identified in Table 6-1 on 16 and
28 August 1976 are listed in Tables 6-7 and 6-8, respectively. Because of
the uncertainties about the actual particulate emissions from the major
industrial sources on the two days as well as the uncertainties about the
effects of emissions from non-traditional sources such as resuspended
roadway dust (see Section 5), no 24-hour concentration isopleth maps are
presented in this section. However, computer listings containing the
results of all of the short-term particulate concentration calculations
have been provided to EPA Region III.
6-22
-------
TABLE 6-7
OBSERVED AND CALCULATED 24-HOUR AVERAGE GROUND-LEVEL PARTICULATE
CONCENTRATIONS AT THE DER AND BAPC MONITORS IN THE
SOUTHWEST PENNSYLVANIA AQCR ON 28 AUGUST 1976
Monitor
Number
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
Monitor
Name
Baden
Beaver Falls
Koppel
Brighton Township
Midland
Elco
Downtown
Central Lab
Hazelwood
North Braddock
Duquesne II
Liberty Borough T
Clairton
Greater Pittsburgh
Airport
South Fayette
Springdale
Court House
North Fayette
Braddock
Duquesne I
Allegheny County
Airport
Glassport T
Cousin Hollow
Henry Kaufmann
Brownsville
Lover
Rostraver Township
Courtney (New Eagle
Ambridge
Rochester
Vanport
New Castle
Bessemer
Monessen
Concentration (yg/m^)
Observed
67
61
86
87
80
46
110
97
105
63
107
—
76
72
41
39
185
45
86
94
65
86
127
71
—
—
—
) 66
44
74
171
63
—
180
Calculated
w/ Quench
19
38
23
47
120
77
31
36
129
42
42
76
135
76
72
42
30
40
52
103
73
51
94
35
34
58
89
15
102
7
22
6
0
50
w/o Quench
19
38
23
47
120
67
27
31
112
30
39
71
121
72
65
39
25
39
33
99
57
48
89
30
29
52
83
13
75
7
22
6
0
41
6-23
-------
TABLE 6-8
OBSERVED AND CALCULATED 24-HOUR AVERAGE GROUND-LEVEL PARTICULATE
CONCENTRATIONS AT THE DER AND BAPC MONITORS IN THE
SOUTHWEST PENNSYLVANIA AQCR ON 28 AUGUST 1976
Monitor
Number
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
Monitor
Name
Baden
Beaver Falls
Koppel
Brighton Township
Midland
Elco
Downtown
Central Lab
Hazelwood
North Braddock
Duquesne II
Liberty Borough T
Clairton
Greater Pittsburgh
Airport
South Fayette
Springdale
Court House
North Fayette
Braddock
Duquesne I
Allegheny County
Airport
Glassport T
Cousin Hollow
Henry Kaufmann
Brownsville
Lover
Rostraver Township
Courtney
Arabridge
Rochester
Vanport
New Castle
Bessemer
Monessen
3
Concentration (yg/m )
Observed
183
137
—
151
248
40
70
—
76
108
—
136
115
97
59
80
125
76
196
87
53
—
125
60
—
—
—
56
148
143
150
343
—
88
Calculated
w/ Quench
248
25
37
22
161
18
9
64
29
64
95
325
14
13
6
41
9
9
34
97
12
23
382
11
2
2
6
3
16
33
36
104
65
16
w/o Quench
163
25
36
22
94
18
9
49
19
63
79
222
14
13
6
35
9
9
32
80
11
22
278
10
2
2
6
3
16
32
36
103
65
16
6-24
-------
SECTION 7
COMPARISON OF CALCULATED AND OBSERVED
PARTICULATE CONCENTRATIONS
7.1 ANNUAL CONCENTRATIONS FOR 1975
Table 7-1 compares the 1975 annual geometric mean particulate
concentrations observed at the 35 hi-vol samplers whose locations are given
in Table 6-1 with the corresponding annual arithmetic mean particulate
concentrations calculated both with and without the effects of emissions
from the quenching process. With the exception of the annual concentration
calculated at Coursin Hollow (either with or without the effects of quenching
included included), the calculated concentrations are all below the observed
concentrations. This result is not surprising if it is remembered that the
calculated concentrations consider only the effects of the quantified
industrial particulate emission and do not include the effects of fugitive
sources (natural and anthropogenic) or the possible effects of advection
from urban and industrial areas outside of the Southwest Pennsylvania AQCR.
The Coursin Hollow site is frequently affected by emissions from the Clair-
ton Coke Works. If the effects of quenching are excluded, there is a very
close correspondence between the calculated and observed annual concentra-
tions at Coursin Hollow. However, because the effects of advection and
emissions from unquantified sources should, at least to some extent, be
present at the Coursin Hollow site, the results of the annual concentration
calculations for Coursin Hollow suggest that either the hi-vol sampler is
not capturing or retaining the particulates associated with the quenching
process or that quenching emissions have been overestimated.
Table 7-1 shows that the correspondence between calculated and
observed annual particulate concentrations generally is best at sampling
sites located near major industrial sources (for example, Baden, Midland
and Coursin Hollow) and worst in downtown areas with heavy traffic (for
7-1
-------
TABLE 7-1
OBSERVED AND CALCULATED 1975 ANNUAL AVERAGE GROUND-LEVEL
PARTICULATE CONCENTRATIONS AT THE DER AND BAPC
MONITORS IN THE SOUTHWEST PENNSYLVANIA AQCR
Monitor
Number
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
Monitor
Name
Baden
Beaver Falls
Koppel
Brighton Township
Midland
Elco
Downtown (County Off. Bldg)
Central Lab
Hazelwood
North Braddock
Duquesne II
Liberty Borough T
Clairton
Greater Pittsburgh Airport
South Fayette
Springdale
Court House
North Fayette
Braddock
Duquesne I
Allegheny County Airport
Glassport T
Coursin Hollow
Henry Kaufmann
Brownsville
Lover
Rostraver Township
Courtney (New Eagle)
Ambridge
Rochester
Vanport
New Castle
Bessemer
Monessen
Swissvale
3
Concentration (yg/m )
Observed*
132
78
103
sV*A
139
74
85
99
97
121
139
107
105
79
56
59
159
61
165
109
73
94
93
77
86
60
50
119
102
88
90
123
200
121
155
Calculated**
w/Quench
83
20
23
25
132
35
33
50
89
55
88
87
77
30
27
47
31
26
43
73
54
69
131
45
12
18
19
20
30
22
24
22
87
31
62
w/o Quench
58
18
22
22
115
34
27
42
68
49
85
66
69
27
25
45
26
24
38
69
45
54
97
36
10
17
17
18
27
20
21
21
86
30
48
Observed Minus
Calculated
49-74
58-60
80-81
7-24
39-40
52-58
49-57
8-29
66-72
51-54
20-41
28-36
48-52
29-31
12-14
128-133
35-37
122-127
36-40
19-28
25-40
(38-4)
32-41
74-76
42-43
31-33
99-101
72-75
66-68
66-69
101-102
113-114
90-91
93-107
Geometric mean.
** Arithmetic mean.
*** Not available due to excessively high readings caused by building construction
and dirt access road.
7-2
-------
example, Downtown and Central Laboratory) and in rural areas (for example,
South Fayette). As explained in Section 5, emission factors for dust
entrained from paved streets by motor vehicle traffic depend on the assumed
type of street and can vary by about a factor of ten for the same amount
of traffic. Although there are considerable uncertainties in these emission
factors, the results presented in Section 5 indicate that dust entrained
from paved streets in downtown areas with heavy traffic may account for
significant fractions of the ambient particulate concentrations that can-
not be accounted for by the quantified industrial emissions. Additionally,
the observed concentrations in rural areas may include significant contri-
butions by natural sources (for example, pollen and wind-blown dust) that
are not included in the industrial emission inventory. The overall corre-
spondence between calculated and observed concentrations tends to be poorest
in the northwest corner of the calculation grid (the New Castle-Beaver
Valley areas). Because this area is relatively close to heavily industrial-
ized areas in Ohio (for example, Youngstown), the effects of advection
from outside the Southwest Pennsylvania AQCR may be greatest in this area.
The hi-vol sampler sites in Table 7-1 that may be classified
as rural or suburban sites are the Brighton Township, Elco, Greater
Pittsburgh Airport, South Fayette, North Fayette, Allegheny County Airport,
Brownsville, Lover and Rostraver Township sites. As shown by the column
at the far right of Table 7-1, the differences between observed and calcu-
lated annual concentrations at these sites generally are very similar.
On the average, the difference between observed and calculated annual con-
centrations at the rural and suburban sites are 40 micrograms per cubic
meter if the effects of quenching are included and 43 micrograms per cubic
meter if the effects of quenching are excluded.
In summary, the results of the annual diffusion-model calculations
provide circumstantial evidence that:
7-3
-------
• Unknown sources not included in the industrial particuate
emissions inventory significantly affect some monitoring
locations
• The entrainment of dust from paved roads by motor vehicle
traffic is probably one of the most significant of the
unknown sources
• Advection of particulates from urban and industrial areas
outside of the Southwest Pennsylvania AQCR may affect
ambient particulate concentrations within the AQCR.,
especially along the western edge of the AQCR
• Either the particulate emission rates for the quenching
process have been overestimated, or the hi-vol samplers
do not capture and/or retain a significant fraction of
these particulates
7.2 TWENTY-FOUR AVERAGE CONCENTRATIONS FOR 16 and 28 AUGUST 1976
It is of interest to consider the uncertainties in the observed
24-hour average particulate concentrations before comparing the calculated
and observed 24-hour average concentrations. Table 7-2 compares the 24-
hour average particulate concentrations measured by the Pennsylvania DER
or the Allegheny County BAPC hi-vol sampelrs on 16 and 28 August 1976 with
the 24-hour average particulate concentrations determined by Coors Spectro-
Chemical Laboratory using the hi-vol filters from the colocated samplers.
The differences in 24-hour average concentrations at colocated samplers
range from 3 to 54 micrograms per cubic meter. An uncertainty in a 24-hour
average particulate concentration of as much as 50 micrograms per cubic
meter significantly affects any attempt to compare concurrent calculated
and observed particulate concentrations. Additionally, an uncertainty
7-4
-------
TABLE 7-2
COMPARISON OF 24-HOUR AVERAGE PARTICULATE CONCENTRATIONS
MEASURED ON 16 AND 28 AUGUST 1976 BY THE
COLOCATED HI-VOL SAMPLERS
Monitor
3
Concentration (jlg/m )
DER or
BAPC
Coors
(a) 16 August 1976
Baden
Beaver Falls
Koppel
Brighton Township
Midland
Elco
Downtown
Central Laboratory
Hazelwood
North Braddock
Duquesne II
Liberty Boro
Clairton
Greater Pittsburgh Airport
South Fayette
67
61
86
87
80
46
110
97
105
63
107
—
76
72
41
86
86
90
65
83
66
139
110
131
79
145
114
52
76
69
Concentration
Difference
19
25
4
22
3
20
29
13
26
16
38
—
24
4
28
(b) 28 August 1976
Baden
Beaver Falls
Koppel
Brighton Township
Midland
Elco
Downtown
Central Laboratory
Hazelwood
North Braddock
Duquesne II
Liberty Boro
Clairton
Greater Pittsburgh Airport
South Fayette
183
137
—
151
248
40
70
—
76
108
—
136
115
97
59
203
164
133
149
268
43
101
119
71
153
—
130
61
110
91
20
27
—
2
20
3
31
—
5
45
—
6
54
13
32
7-5
-------
of the magnitude may significantly affect determinations of compliance
with the 24-hour primary and secondary National Ambient Air Quality Standards
(NAAQS) for particulates of 260 and 150 micrograms per cubic meter, re-
spectively. For example, whether or not the 24-hour primary NAAQS is
exceeded at the Midland site on 28 August 1976 depends on which of the
measured concentrations is accepted as valid.
In general, the 24-hour average concentrations measured by Coors
tend to be higher than the 24-hour average concentrations measured by the
DER or BAPC. Differences in the handling of the hi-vol filters (i.e.,
placing cover filters over the primary filters used by Coors) may account
for this result (see Section 3.8.4). However, the DER or BAPC concentrations
are, at times, higher than the concentrations estimated by Coors (for
example, see the Clairton monitor on 28 August 1976). The filters analyzed
by Coors were removed at the end of the sampling day, while the DER and
BAPC filters were removed at the start of the next sampling day. Thus,
although the DER and BAPC hi-vol samplers were static between sampling
days, the action of the wind alone may have caused the filters on these
samplers to accumulate additional particulates during the static period.
Table 7-3 compares the calculated and observed 24-hour average
particulate concentrations on 16 and 28 August 1976 for the fifteen hi-vol
sampler locations used in the filter analyses described in Section 4. The
prevailing winds on 16 August 1976 were from the northwest through north,
while the prevailing winds on 28 August 1976 were from the south through
southwest. The Greater Pittsburgh Airport recorded 1.22 centimeters
(0.48 inches) of precipitation during the 24-hour period ending at
midnight on 15 August 1976; no precipitation was measured at the Airport
during the 24-hour period ending at midnight on 16 August 1976. Because
no precipitation was measured at the Airport on 27 August 1976 and only
a trace of precipitation was measured on 28 August 1976, the effects on
particulate air quality of non-traditional fugitive sources probably were
more significant on 28 August 1976 than on 16 August 1976.
7-6
-------
w
PM PM
O
O
O
O
^~N
t3 fO
d e
cfl --^
60
cy —
S-i en
o) d
en o
X> -H
O 4J
rt
d rJ
4H CJ
•H rH
Q cfl
u
d
o
•H
4J /-N
cfl co
M ^
4-1 -•-.
d 60
cu i^-
o ^
d
o
u
en
o
o
u
a
p i
<3^
PQ
O
Crf
W
Q
-d
0)
4-1
cd
rH
d
o
rH
cfl
"d
QJ
£
Q)
W
o
O
co
d
x;
o o
0)
d
cr
60
d
•rH
f~;
O
d
0)
d
O'
60
d
•rH
r"!
O O
z d
0)
d
cy
60
d
•H
[~l
o
d
CU
d
cy
60
d
-H
O O
*Z> d
0)
d
cy
60
d
•H
X!
O
d
a)
;3
o-
en
M
o
o
c_>
S-i
O O
P4
W PQ
Q
r<
O
4J
•H
d
o
S
^O
O
T 1
4-)
en
60
d
vO
rH
/ — \
CT3
r-t»r--oor-rHrsic^aNON^Droc^-o ON C) r— H ON in ^j~ CM vo ON
OO OO ON ^o OO ^O CO rH CO I^~^ *^" rH in ( — ^ vO
rH rH rH rH rH
r^rH^Di^-o^DOr^-incor-~ I \ocMrH
rH rH rH
M
O
P.
•H
r**^i ^C
& J-l 60
•HO i-t
x; 4-1 d
en cd ^ xi
d j-i o en V
en 3 o O O 4-1 4-i
rH O X3rdHVJ4J4J
rHH nj-dMO-HO)
cfl i-J T3 cfl qq P-, >~,
HrH4-id on]^ cn4-)4-icu
T3 C^ OH *r^ "^f CJ ^ C^ N M O"* O C$ flj ^3
nj 1) O V^ "fH ^~^ O CJ o3 O ^3 *rH i~H M O
pT) pQ |s^ p^ ^' PT^ [^ QJ prj J2j C~) i 1 C_5 C»D C/D
7-7
-------
TJ
0)
d
•H
4->
d
o
w
1-1
^— v
T-)CO
d e
cd ~-~
M
T3 P-
Qj ^— '
>
M W
o) d
co o
JO -H
0 4J
0)
d rJ
0) 4-1
d) d
IS 0)
4J O
a) d
43 O
o
0)
a 13
d a)
a) 4-1
M rd
0) rH
4H 3
>4H U
•H rH
Q n3
u
d
O
•H
4J /~v
rt co
rJ B
4-1 ---.
C M
n) ~~ 5
(XI ,~J.
O •— '
d
o
u
to
M
o
u
u
pj
«3
m
r4
0
erf
w
p
T3
0)
4-1
n)
rH
3
o
1— 1
n3
u
13
0)
}_!
0)
CD
43
o
60
d
•H
jC
o a
52 d
0)
3
o-
60
d
•H
42
U
d
a)
3
o-
60
d
•H
42
O O
ss d
OJ
3
cy
bO
d
•H
t_f"^
U
C
0)
3
cr
60
d
•H
42
0 0
& d
0)
3
o-
60
d
•H
42
O
d
0)
3
co-
co
rJ
O
0
u
rJ
O O
P4
erf -5
W M
Q
M
o
4J
•H
d
o
a
VD
f^
C^
rH
4J
CO
3
60
3
<
oo
CM
^ — ^
43
O^r^-r-^cMOcMa-iLn~D l CM ^3" CO vO
^OCMCOCMCT>rH ^J-rHv£)p-~-C-JrHrH
rH CM
coinr^cMrHooo'.Ki-cTixj-Lninvj-co^'O
ooco i m^ 42
ex M 60
•HO s-i
42 4J 3
CO CTJ ^ 43
d MO wa)
en 3 o o o 4J 4J
rH O 43l3rH>-4 4J4J
rHH nJI^MO-HO)
n) i-J "d rt CO PLI >>
fn d d oi-ia) d d
O T3 3rHOCQd>.OS-ilii
MrH4-lql OOJS COU4JCU
!2(D(l)42cd 4JJ-irH42d)rJj-i4-)42
Q) >CX60rH O d4-10!4-l 30) -H Cd4J
' O rd O-< *rH T U O ^ C W w CT1 r^ fu 1) ^
tdO)OrJ'HrHOa)a3O3'HrHMO
CQpQt«ipqSWPUtf2Qh-lOOW
7-?
-------
Inspection of Table 7-3 that the inclusion of the effects of
emissions from the quenching process in the model calculations does not
significantly affect the 24-hour average concentrations calculated at
the fifteen hi-vol sampler locations on 16 August 1976. The significant
overpredictions occur at samplers near major source complexes (the Midland,
Hazelwood and Clairton sites) and at the samplers in rural areas (the Elco
and South Fayette sites) . Neglect of the effects of depletion by dry
deposition is a possible explanation for the overpredictions at the Elco
and South Fayette sites which are south of industrialized areas. Excluding
the sites at which the observed concentrations are overpredicted, the average
difference between observed and calculated concentrations is about 50 micro-
grams per cubic meter for both the Agency (DER or BAPC) and Coors data,
both with and without the effects of quenching.
Table 7-3 shows that the effects of emissions from the quenching
process on the 24-hour average particulate concentrations calculated for
28 August 1976 are significant at the hi-vol samplers which are immediately
downwind of quench towers during periods of southwest winds (the Baden,
Midland and Liberty Boro sites). If the effects of quenching are excluded,
the only site at which the observed concentration is overpredicted is the
Liberty Boro site which is northeast of the Clairton Coke Works. Excluding
the monitors at which the observed concentrations are overpredicted, the
average difference between the observed and calculated concentrations for
the Agency samplers is about 75 micrograms per cubic meter, both with and
without the effects of quenching. Similarly, excluding the monitors at which
the observed concentrations are overpredicted, the average difference be-
tween the observed and calculated concentrations for the Coors samplers
is about 85 micrograms per cubic meter, both with and without the effects
of quenching. (The average difference between the particulate concentrations
measured by the colocated Agency and Coors samplers on 28 August 1976 is
10 micrograms per cubic meter.) Thus, the results of the model calculations
suggest that the contributions to observed particulate concentrations of
natural and anthropogenic sources not included in the industrial particulate
emissions inventory were larger on 28 August 1976 than on 16 August 1976.
7-9
-------
As noted above, this result is not unexpected because of the occurrence of
significant precipitation on 15 August 1976.
The contributions of individual sources to the 24-hour average
particulate concentrations calculated for the Baden site on 16 and 28
August 1976 are given in Tables 7-4 and 7-5, respectively. The corresponding
calculated source contributions for the fourteen remaining hi-vol sampler
sites are shown in Tables 7-6 through 7-33. It is important to note that
the calculated concentration given at the bottom of each table includes
the effects of emissions from the quenching process. However, if emissions
from the quenching significantly affect the total concentration given in
a table for a combination of sampler location and day, the contribution
of quenching emissions is isolated in the table.
7-10
-------
TABLE 7-4
CONTRIBUTION OF INDIVIDUAL SOURCES TO THE 24-
HOUR AVERAGE PARTICULATE CONCENTRATION
CALCULATED AT THE BADEN MONITOR ON
16 AUGUST 1976
Source
ARCO Polymers
Sechan Limestone
Duquesne Slag
Heckett Engineering Slag
Pennsylvania Power Co. (W. Pgh.)
Fenati Brick
Other Sources
All Sources Combined
Calculated Concentration
(ug/m3)
1.13
0.28
0.33
10.77
5.14
0.63
0.87
19.15
Percent
Contribution
5.9
1.5
1.7
56.4
26.9
3.3
4.3
100.0
Observed concentration (jJg/m ):
Coors: 86
Agency: 67
7-11
-------
-------
-------
TABU- 7-5
CONTRIBUTION OF INDIVIDUAL SOURCES TO THE 24-HOUR
AVERAGE PARTICULATE CONCENTRATION CALCULATED
AT THE BADEN MONITOR ON
28 AUGUST 1976
Source
J & L Aliquippa Works
J & L Aliquippa Quench .
Ohio Sources
W. Va. Sources
Other Sources
All Sources Combined
Calculated Concentration
(yg/m3)
, 141.11
85.48
12.10
3.69
6.04
248.42
Percent
Contribution
56.8
34.3
4.9
1.5
2.4
100.0
Observed concentration (yg/m )
Coors: 203
Agency: 183
7-12
-------
-------
TABLE 7-6
CONTRIBUTION OF INDIVIDUAL SOURCES TO THE 24-HOUR
AVERAGE PARTICULATE CONCENTRATION CALCULATED
AT THE BEAVER FALLS MONITOR ON
16 AUGUST 1976
Source
Armstrong Cork
Pennsylvania Power Co. (W. Pgh.)
Other Sources
All Sources Combined
— - -- j • —
Calculated Concentration
(Vlg/m3)
10.49
26.13
1.51
38.13
Percent
Contribution
27.5
68.6
3.9
100.0
Observed concentration (pg/m ):
Coors: 86
Agency: 61
7-13
-------
-------
-------
TABLE 7-7
CONTRIBUTION OF INDIVIDUAL SOURCES TO THE 24-HOUR
AVERAGE PARTICULATE CONCENTRATION CALCULATED
AT THE BEAVER FALLS MONITOR ON
.28 AUGUST 1976
1
Source
J & L Aliquippa Works
Crucible Steel
Crucible Quench
St. Joe Minerals
Ohio Sources
W. Va. Sources
Republic Steel
Duquesne Power & Light (Phillips)
Other Sources
All Sources Combined
Calculated Concentration
(Ug/m3)
1.53
1.04
.62
.31
12.24
2.72
4.29
2.24
.42
25.41
Percent
Contribution
6.0
4. 1
2.4
1.2
48.2
10-7
16.9
8.8
1.7
100.0
Observed concentration (yg/m ):
Coors: 164
Agency: 137
7-14
-------
TALBE 7-8
CONTRIBUTION OF INDIVIDUAL SOURCES TO THE 24-HOUR
AVERAGE PARTICULATE CONCENTRATION CALCULATED
AT THE KOPPEL MONITOR ON
16 AUGUST 1976
Source
Shenango China
Pennsylvania Power Co. (W. Pgh.)
Medusa Cement
Bessemer Cement
Other Sources
All Sources Combined
Calculated Concentration
(Ug/ro3)
.81
19.99
.77
.41
.70
22.68
Percent
Contribution
3.6
88.1
3.4
1.8
3.1
100.0
Observed concentration (yg/m ):
Coors: 90
Agency: 86
7-15
-------
TABLE 7-9
CONTRIBUTION OF INDIVIDUAL SOURCES TO THE 24-HOUR
AVERAGE PARTICULATE CONCENTRATION CALCULATED
AT THE KOPPEL MONITOR ON
28 AUGUST 1976
Source
J & L Aliquippa Works
J & L Aliquippa Quench
Crucible Steel
Ohio Sources
W. Va. Sources
Babcock & Wilcox
Duquesne Power & Light (Phillips)
Other Sources
All Sources Combined
Calculated Concentration
(lig/m3)
2.33
.71
.73
22.03
6.24
1.43
1.51
1.67
36.65
Percent
Contribution
6.4
1.9
2.0
60.1
17.0
3.9
4.1
4.6
100.0
_ — -. . . , — . . —
3
Observed concentration (yg/m ):
Coors: 133
Agency: Missing
7-16
-------
TABLE 7-10
CONTRIBUTION OF INDIVIDUAL SOURCES TO THE 24-HOUR
AVERAGE PARTICULATE CONCENTRATION CALCULATED
AT THE BRIGHTON MONITOR ON
16 AUGUST 1976
Source
Babcock & Wilcox (WR)
Pennsylvania Power Co. (W. Pgh.)
Other Sources
All Sources Combined
Calculated Concentration
(]Jg/m3)
1.84
44.08
1.68
47.60
Percent
Contribution
3.9
92.6
3.5
100.0
Observed concentration (yg/m )
Coors: 65
Agency: 87
7-17
-------
TABLE 7-11
CONTRIBUTION OF INDIVIDUAL SOURCES TO THE 2/4-HOUR
AVERAGE PARTICULATE CONCENTRATION CALCULATED
AT THE BRIGHTON MONITOR ON
28 AUGUST 1976
Source
Crucible Steel
Crucible Quench
St. Joe Minerals
Ohio Sources
W. Va. Sources
ARCO Polymers
Other Sources
All Sources Combined
Calculated Concentration
(yg/m3)
1.33
.79
1.09
12.97
3.30
2.18
.82
22.48
Percent
Contribution
5.9
3.5
4.8
57.7
14.7
9.7
3.7
100.0
Observed concentration (yg/m ):
Coors: 149
Agency: 151
1-U
-------
TABLE 7-12
CONTRIBUTION OF INDIVIDUAL SOURCES TO THE 24-HOUR
AVERAGE PARTICULATE CONCENTRATION CALCULATED
AT THE MIDLAND MONITOR ON
16 AUGUST 1976
Source
Crucible Steel
Pennsylvania Power Co.(W. Pgh. )
Other Sources
All Sources Combined
Calculated Concentration
(yg/m3)
108.31
10.87
0.55
119.73
Percent
Contribution
90.5
9.1
0.4
100.0
Observed concentration (yg/m ):
Coors: 83
Agency: 80
7-19
-------
TABLE 7-13
CONTRIBUTION OF INDIVIDUAL SOURCES TO THE 24-HOUR
AVERAGE PARTICULATE CONCENTRATION CALCULATED
AT THE MIDLAND MONITOR ON
28 AUGUST 1976
Source
Crucible Steel
Crucible Quench
Ohio Sources
W. Va. Sources
Other Sources
All Sources Combined
Calculated Concentration
(VS/m3)
46.28
67.06
36.14
11.69
0.09
161.26
Percent
Contribution
28.7
41.6
22.7
7.2
0.1
100.0
Observed concentration (jJg/m ) :
Coors: 268
Agency: 248
7-20
-------
TABLE 7-14
CONTRIBUTION OF INDIVIDUAL SOURCES TO THE 24-HOUR
AVERAGE PARTICULATE CONCENTRATION CALCULATED
AT THE ELCO MONITOR ON
16 AUGUST 1976
Source
USS Clairton Works
USS Clairton Quench
USS National Works
Duquesne Power (Elrama)
J & L Pittsburgh Works
J & L Pittsburgh Quench
Crucible Steel
Wheeling Pittsburgh Steel (Moness
Wheeling Pittsburgh Quench (Mones
Duquesne Slag
Ohio Sources
Springdale Power
Other Sources
All Sources Combined
Calculated Concentration
(yg/m3)
5.11
2.48
1.25
22. 10
8.20
6.77
0.81
;n) 2.38
sen) 0-84
1.20
14.11
3.58
8.51
77.34
Percent
Contribution
6.6
3.2
1.6
28.6
10.6
8.8
1.0
3.1
1.1
1.6
18.2
4.6
11.0
100.0
Observed concentration (yg/m ):
Coors: 66
Agency: 46
7-21
-------
TABLE 7-15
CONTRIBUTION OF INDIVIDUAL SOURCES TO THE 24-HOUR
AVERAGE PARTICULATE CONCENTRATION CALCULATED
AT THE ELCO MONITOR ON
28 AUGUST 1976
Source
Ohio Sources
Allied Chemical
Other Sources
All Sources Combined
Calculated Concentration
(yg/m3)
1.39
15.64
.87
17.90
Percent
Contribution
7.8
87.4
4.8
100.0
Observed concentration (yg/m ):
Coors: 43
Agency: 40
7-22
-------
TABLE 7-16
CONTRIBUTION OF INDIVIDUAL SOURCES TO THE 24-HOUR
AVERAGE PARTICIPATE CONCENTRATION CALCULATED
AT THE CENTRAL LAB MONITOR ON
16 AUGUST 1976
Source
J & L Aliquippa Works
J & L Aliquippa Quench
St. Joe Minerals
USS Saxonburg
H. J. Heinz
Pittsburgh Brewing
Heckett Engineering
Other Sources
Other Sources Combined
Calculated Concentration
(lig/m3)
12.09
4.30
1.26
1.01
12.00
1.66
.61
3.03
35.96
Percent
Contribution
33.6
12.0
3.5
2.8
33.4
4.6
1.7
8.4
100.0
Observed concentration (yg/m ):
Coors: 110
Agency: 97
7-23
-------
TABU', 7-17
CONTRIBUTION OF INDIVIDUAL SOURCES TO THE 24-HOUR
AVERAGE PARTICULATE CONCENTRATION CALCULATED
AT THE CENTRAL LAB MONITOR ON
28 AUGUST 1976
Source
USS Clairton Works
USS Clairton Quench
J & L Pittsburgh Works
J & L Pittsburgh Quench
Wheeling Pittsburgh Steel
Ohio Sources
Duquesne Power & Light (Elrama)
Other Sources
All Sources Combined
Calculated Concentration
(Iig/m1)
2.74
1.09
28.92
13.17
1.25
4.50
9.36
3.39
64.42
Percent
Contribution
4.3
1.7
44.9
20.4
1.9
7.0
14.5
5.3
100.0
Observed concentration (yg/m ):
Coors: 119
Agency: Missing
7-24
-------
TABLE 7-18
CONTRIBUTION OF INDIVIDUAL SOURCES TO THE 24-HOUR
AVERAGE PARTICULATE CONCENTRATION CALCULATED
AT THE DOWNTOWN MONITOR ON
16 AUGUST 1976
Source
Dixmont Hospital
H. J. Heinz
Marquette Cement
Calgon (Drying)
Calgon (Baking //I)
Calgon (Baking //2)
Shenago Inc. Boilers #7 & 8
Shenago Inc. Coke Ovens
Vulcan Materials
J & L Aliquippa Works
St. Joe Minerals
ARCO Polymers
J & L Aliquippa Quench
Shenago Quench
Shenago Quench
Other Sources
All Sources Combined
Calculated Concentration
(yg/m3)
1.68
.60
1.72
.41
1.15
2.47
.73
3.95
.41
8.98
1.17
.58
.81
1.96
1.94
2.77
31.33
Percent
Contribution
5.4
1.9
5.5
1.3
3.7
7.9
2.3
12.6
1.3
28.7
3.7
1.9
2.6
6.3
6.2
8.7
100.0
Observed concentration (yg/m )
Coors: 139
Agency: 110
7-25
-------
TABLE 7-19
CONTRIBUTION OF INDIVIDUAL SOURCES TO THE 24-HOUR
AVERAGE PARTICULATE CONCENTRATION CALCULATED
AT THE DOWNTOWN MONITOR ON
28 AUGUST 1976
Source
Crucible Steel
Wheeling Pittsburgh Steel
Wheeling Pittsburgh Quench
Ohio Sources
W. Va. Sources
Duquesne Power & Light (Phillips)
Duquesne Power & Light (Elrama)
Hatfield Power
Other Sources
All Sources Combined
Calculated Concentration
(lig/m3)
.13
.49
.19
4.43
.12
.15
2.87
.61
.43
9.42
Percent
Contribution
1.4
5.2
2.0
47.0
1.3
1.6
30.5
6.5
4.5
100.0
Observed concentration (yg/m ):
Coors: 101
Agency: 70
7-26
-------
TABLE 7-20
CONTRIBUTION OF INDIVIDUAL SOURCES TO THE 24-HOUR
AVERAGE PARTICIPATE CONCENTRATION CALCULATED
AT THE HAZELWOOD MONITOR ON
16 AUGUST 1976
Source
J & L Pittsburgh Works
J & L Pittsburgh Quench
Shenango Works
Shenango Quench
J & L Aliquippa Works
USS Saxonburg
Other Sources
All Sources Combined
Calculated Concentration
(US/m3-)
86.79
11.88
4.46
3.72
8.26
2.64
11.09
128.84
Percent
Contribution
67.4
9.2
3.5
2.9
6.4
2.0
8.6
100.0
Observed concentration (yg/m ):
Coors: 131
Agency: 105
7-27
-------
TABLE 7-21
CONTRIBUTION OF INDTVIDUAL SOURCES TO THE 24-HOUR
AVERAGE INARTICULATE CONCENTRATION CALCULATED
AT THE HAZELWOOD MONITOR ON
28 AUGUST 1976
Source
USS Clairton Works
USS Clairton Quench
USS Irvin
J & L Pittsburgh Works
J & L Pittsburgh Quench
Wheeling Pittsburgh Steel
Wheeling Pittsburgh Quench
Ohio Sources
Duquesne Power & Light (Elrama)
Other Sources
All Sources Combined
Calculated Concentration
(Ug/m3)
1.94
.75
.37
2.72
8.45
1.66
.64
4.15
7.43
1.13
29.24
Percent
Contribution
6.6
2.6
1.3
9.3
28.9
5.7
2.2
14.2
24.4
4.8
100.0
Observed concentration (yg/m ):
Coors: 71
Agency: 76
7-28
-------
TABLE 7-22
CONTRIBUTION OF INDIVIDUAL SOURCES TO THE 24-HOUR
AVERAGE PARTICULATE CONCENTRATJON CALCULATED
AT THE NORTH BRADDOCK MONITOR ON
16 AUGUST 1976
Source
USS Homestead Works
J & L Pittsburgh Works
J & L Pittsburgh Quench
Wabco
J & L Aliquippa Works
J & L Aliquippa Quench
USS Saxonburg
Heckett Engineering
Pennsylvania Power (W. Pgh.)
Other Sources
All Sources Combined
Calculated Concentration
(yg/m3)
3.35
13.39
10.12
0.49
5.38
2.15
1.15
0.64
1.57
3.82
42.06
Percent
Contribution
8.0
31.8
24.1
1.2
12.8
5.1
2.7
1.5
3.7
9.1
100.0
Observed concentration (yg/m ):
Coors: 79
Agency: 63
7-29
-------
TABLE 7-23
CONTRIBUTION OF TNDFVIDUAL SOURCES TO THE 74-HOUR
AVERAGE PARTICULATE CONCENTRATION CALCULATED
AT THE NORTH BRADDOCK MONITOR ON
28 AUGUST 1976
Source
USS Clairton Works
USS Clairton Quench
USS Irvin
USS National
USS Edgar Thomson
USS Duquesne
USS Homestead
Ohio Sources
Duquesne Power & Light (Elrama)
Other Sources
All Sources Combined
Calculated Concentration
(US/m3)
2.90
1.12
1.74
12.73
30.06
1.51
1.54
3.66
6.16
2.97
64.39
Percent
Contribution
4.5
1.7
1
2.7
19.8
46.7
2.3
2.4
5.7
9.6
4.6
100.0
Observed concentration (yg/m )
Coors: 153
Agency: 108
7-30
-------
TABLE 7-24
CONTRIBUTION OF INDIVIDUAL SOURCES TO THE 24-HOUR
AVERAGE PARTICULATE CONCENTRATION CALCULATED
AT THE DUQUESNE II MONITOR ON
16 AUGUST 1976
Source
USS Edgar Thomson
USS Homestead
J & L Aliquippa Works
J & L Aliquippa Quench
St. Joe Minerals
USS Saxonburg
H. J. Heinz
Heckett Engineering
Pennsylvania Power (W. Pgh.)
Ohio Sources
Springdale Power
Mercer Lime & Stone
Other Sources
All Sources Combined
Calculated Concentration
(Ug/m3)
15.29
5.41
7.81
2.83
0.70
0.94
1.23
0.55
0.80
1.61
0.92
0.70
3.81
42.60
Percent
Contribution
35.9
12.7
18.3
6.6
1.6
2.2
2.9
1.3
1.9
3.8
2.2
1.6
9.0
100.0
Observed concentration (jJg/m ) :
Coors: 145
Agency: 107
7-31
-------
TABLE 7-25
CONTRIBUTION OF INDIVIDUAL SOURCES TO THE 2/4-HOUR
AVERAGE PARTICIPATE CONCENTRATION CALCULATED
AT THE DUQUESNE II MONITOR ON
28 AUGUST 1976
Source
USS Clairton Works
USS Clairton Quench
USS Irvin
USS National
Ohio Sources
Duquesne Power & Light (Elrama)
Other Sources
All Sources Combined
Calculated Concentration
(yg/m3)
42.33
15.42
2.63
20.56
2.88
8.38
2.98
95.18
Percent
Contribution
44.5
16.2
2.8
21.6
3.0
8.8
3.1
100.0
3
Observed concentration (jJg/m ):
Coors: Missing
Agency: Missing
7-32
-------
TABLE 7-26
CONTRIBUTION OF INDIVIDUAL SOURCES TO THE 24-HOUR
AVERAGE PARTICULATE CONCENTRATION CALCULATED
AT THE LIBERTY BOROUGH MONITOR ON
16 AUGUST 1976
Source
USS National Works
USS Edgar Thomson
USS Duquesne Works
USS Homestead
J & L Pittsburgh Works
J & L Pittsburgh Quench
Shenango Works
J & L Aliquippa Works •
J & L Aliquippa Quench
USS Saxonburg
Ohio Sources
Pittron
Other Sources
All Sources Combined
Calculated Concentration
(Ug/m3)
2.75
6.16
.97
10.32
3.93
3.76
.86
4.66
.80
1.05
29.78
2.16
8.85
76.05
Percent
Contribution
3.6
8.1
1.3
13.6
5.2
4.9
1.1
6.1
1.1
1.4
39.2
2.8
11.6
100.0
Observed concentration (yg/m ):
Coors: 114
Agency: Missing
7-33
-------
TABLE 7-27
CONTRIBUTION OF INDIVIDUAL SOURCES TO THE 24-HOUR
AVERAGE PARTICULATE CONCENTRATION CALCULATED
AT THE LIBERTY BOROUGH MONITOR ON
28 AUGUST 1976
Source
USS Clairton Works
USS Clairton Quench
Ohio Sources
Other Sources
All Sources Combined
Calculated Concentration
(yg/m3)
209.28
103.64
2.80
9.86
325.58
Percent
Contribution
64.3
31.8
1.0
2.9
100.0
Observed concentration (yg/m ):
Coors: 130
Agency: 136
7-34
-------
TABLE 7-28
CONTRIBUTION OF INDIVIDUAL SOURCES TO THE 24-HOUR
AVERAGE PARTICIPATE CONCENTRATION CALCULATED
AT THE CLAIRTON MONITOR ON
16 AUGUST 1976
Source
USS C lair ton Works
USS Clairton Quench
USS Irvin
USS Edgar Thomson
USS Homestead
J & L Aliquippa Works
Ohio Sources
Other Sources
All Sources Combined
Calculated Concentration
(lJg/m3)
80.40
10.88
5.60
5.45
1.92
3.36
15.00
12.50
135.11
Percent
Contribution
59.5
8.1
4.1
4.0
1.4
2.5
11.1
9.3
100.0
Observed concentration (yg/m ):
Coors: 52
Agency: 76
7-35
-------
TABLE 7-29
CONTRIBUTION OF INDIVIDUAL SOURCES TO THE 24-HOUR
AVERAGE PART1CULATE CONCENTRATION CALCULATED
AT THE CLAIRTON MONITOR ON
28 AUGUST 1976
Source
Ohio Sources
W. Va. Sources
Duquesne Power & Light (Elrama)
Other Sources
All Sources Combined
Calculated Concentration
(Ug/ro3)
2.04
.23
11.90
.42
14.59
Percent
Contribution
14.0
1.6
81.5
2.9
100.0
Observed concentration (yg/m ):
Coors: 61
Agency: 115
7-36
-------
TABLE 7-30
CONTRIBUTION OF INDIVIDUAL SOURCES TO THE 24-HOUR
AVERAGE PARTICULATE CONCENTRATION CALCULATED
AT THE GREATER PITTSBURGH AIRPORT MONITOR
ON 16 AUGUST 1976
Source
Duquesne Power & Light (Phillips)
J & L Aliquippa Works
Crucible Steel
Heckett Engineering Slag
Pennsylvania Power Co. (W. Pgh.)
J & L Aliquippa Quench
Crucible Quench
Other Sources
All Sources Combined
Calculated Concentration
(Ug/m3)
19.41
39.00
4.42
1.22
4.01
3.56
1.16
3.80
76.58
Percent
Contribution
25.3
50.9
5.8
1.6
5.2
4.6
.1.5
5.1
100.0
Observed concentration (Ug/m ):
Coors: 76
Agency: 72
7-37
-------
TABLE 7-31
CONTRIBUTION OF INDIVIDUAL SOURCES TO THE 24-HOUR
AVERAGE PART1CULATE CONCENTRATION CALCULATED
AT THE GREATER PITTSBURGH AIRPORT MONITOR
ON 28 AUGUST 1976
Source
Ohio Sources
W. Va. Sources
Other Sources
All Sources Combined
Calculated Concentration
(Ug/ra3)
9.99
2.47
0.15
12.61
Percent
Contribution
79.2
19.6
1.2
100.0
Observed concentration (l-ig/m )
Coors: 110
Agency: 97
7-38
-------
TABLK 7-32
CONTRIBUTION OF INDIVIDUAL SOURCES TO THE 24-HOUR
AVERAGE PARTICULATE CONCENTRATION CALCULATED
AT THE SOUTH FAYETTE MONITOR ON
16 AUGUST 1976
Source
Duquesne Power & Light (Phillips)
J & L Aliquippa Works
Crucible Steel
Heckett Engineering Slag
Pennsylvania Power Co. (W. Pgh.)
Fenati Brick
J & L Aliquippa Quench
Crucible Quench
Ohio Sources
Other Sources
All Sources Combined
Calculated Concentration
(yg/m3)
5.82
15.61
2.61
1.02
2. 17
.74
5.48
1.03
34.59
2.96
72.03
Percent
Contribution
8.1
21.7
3.6
1.4
3.0
1.0
7.6
1.4
48.0
4.2
100.0
Observed concentration (yg/m ):
Coors: 69
Agency: 41
7-39
-------
TABLE 7-33
CONTRIBUTION OF.INDIVIDUAL SOURCES TO THE 24-HOUR
AVERAGE PARTICULATE CONCENTRATION CALCULATED
AT THE SOUTH FAYETTE MONITOR ON
28 AUGUST 1976
Source
Ohio Sources
W. Va. Sources
Other Sources
All Sources Combined
Calculated Concentration
(Ug/m3)
5.86
0.15
0.08
6.10
Percent
Contribution
96.1
2.5
1.4
100.0
Observed concentration (yg/m )
Coors: 91
Agency: 59
7-40
-------
REFERENCES
Bloom, B,, 1977: Private communication.
Bowne, N. E., 1974: Diffusion rates. J. Air Poll. Control Assoc., _24(9)
832-835.
Boyle, D. G., et. al., 1975: DC-7B aircraft spray system for large-area
insect control. Dugway Proving Ground Document No. DPG-DR-C090A,
U. S. Army Dugway Proving Ground, Dugway, Utah 84022.
Briggs, G. A., 1971: Some recent analyses of plume rise observations. In
Proceedings of the Second International Clean Air Congress,
Academic Press, New York.
Briggs, G. A., 1972: Chimney plumes in neutral and stable surroundings.
Atm. JEnv., .6(7), 507-510,
Calder, K. L., 1971: A climatological model for multiple source urban air
pollution. Proceedings ofthe Second Meeting of the Expert Panel
on Air Pollution Modeling, NATO Committee on the Challenges of
Modern Society, Paris, France, July 1971, 33.
Conover, W. J., 1971: Practical Nonparametric_Statistics, John Wiley and
Sons, Inc., New York, 462.
Coors Spectro-Chemical Laboratory, 1976a: Observation of Atmospheric
Particles Collected in Pittsburgh Area, progress report under
customer order number 415104025, 10 November 1976.
Coors Spectro-Chemical Laboratory, 1976b: Observation of Atmospheric
Particles Collected in Pittsburgh Area, progress report under
customer order number 415104025, 2 December 1976.
Cowherd, C. and C. 0. Mann, 1976: Quantification of dust entrainment from
paved roads. APCA Paper No. 76-5.4, presented at the 69th
Annual Meeting of the Air Pollution Control Association, Portland,
Oregon, June 27 - July 1, 1976.
Cramer, H. E., et al., 1972: Development of dosage models and concepts.
GCA Corporation Final Report under Contract DAAD09-67-C-0020(R)
with the U. S. Army, Deseret Test Center, Fort Douglas, Utah.
PTC Report No. DTC-TR-72-609.
Cramer, H. E., H. V. Geary and J. F. Bowers, 1975: Diffusion-model calcu-
lations of long-term and short-term ground-level S02 concentrations
in Allegheny County, Pennsylvania. H. E. Cramer Company Technical
Report TR-75-102-01 prepared for the U. S. Environmental Protection
Agency, Region III, Philadelphia, PA. EPA Report 903/9-75-018.
NTIS Accession No. PB-245262/AS.
3-1
-------
REFERENCES (Continued)
DeMarrais, G. A., 1959: Wind speed profiles at Brookhaven National Labora-
tory. J. Meteor., 16, 181-190.
DeNevers, N., 1977: Private Communication.
Dumbauld, R. K., J. E. Rafferty and H. E. Cramer, 1976: Dispersion-deposi-
tion from aerial spray releases. Preprint Volume for the Third
Symposium on Atmospheric Turbulence, Diffusion and Air Quality,
American Meteorological Society, Boston, Mass.
Environmental Data Service, 1966: Tabulation III, daily mixing depths and
average wind speeds—Pittsburgh, Pa. Job No. 6234, National
Climatic Center, Federal Building, Asheville, N. C.
Environmental Protection Agency, 1969: Air Quality Display Model. Pre-
pared by TRW Systems Group, Washington, D. C., available as
PB-189-194 from the National Technical Information Service,
Springfield, Virginia.
Environmental Protection Agency, 1972: Complication of air pollutant
emission factors (revised) . Office of Air and Water Programs^
Publication No. AP-42, U. S. Environmental Protection Agency,
Research Triangle Park, N. C.
Gifford, F. A. and S. R. Hanna, 1973: Modeling urban air pollution.
Atm. Env., ^, 131-136.
Herdan, G., 1960: Small Particle Statistics, 2nd Ed., Butterworths,
London, 33.
Holzworth, G. C., 1972: Mixing heights, wind speeds and potential for
urban air pollution throughout the contiguous United States.
Publication No. AP-101, U. S. Environmental Protection Agency,
Research Triangle Park, North Carolina.
Huber, A. H. and W. H. Snyder, 1976: Building wake effects on short stack
effluents. Preprint Volume for the Third Symposium on Atmos-
pheric Turbulence, Diffusion and Air Quality, American Meteoro-
logical Society, Boston, Massachusetts.
Kraus, E. H., W. F. Hunt and L. S. Ramsdall, 1936: Mineralogy, an
Introduction to the Study of Minerals and Crystals, McGraw-Hill,
New York, 465.
Larsen, R. I., 1971: A mathematical model for relating air quality measure-
ments to air quality standards. U. S. Environmental Protection
Agency Publication No. AP-89, Environmental Protection Agency,
Research Triangle Park, N. C.
-------
REFERENCES (Continued)
Luna, R. E. and H. W. Church, 1972: A comparison of turbulence intensity
and stability ratio measurements to Pasquill stability classes.
J. Appl. Met., JJ,(4), 663-669.
McCrone, W. C., R. G. Draftz, J. G. Delley, 1967: Particle Atlas. Ann
Arbor Science Publishers, Inc., Ann Arbor, Michigan.
McCrone, W. C. and J. G. Delley, 1973: The Particle Atlas Edition Two,
Ann Arbor Science Publishers, Inc., Ann Arbor, Michigan, 265.
McDonald, J. E., 1960: An aid to computation of terminal fall velocities
of spheres. J. Me t. , J/7, 463.
Mendenhall, W. and R. L. Scheaffer, 1973: Mathematical Statistics with
Applications, Droxbury Press, Massachussetts, 561.
Olsen, R. H., M. Y. Almassy and A. L. Wingert, 1975: A study of the sus-
pended particulate problem in the Duwamish Basin. EPA Report No.
910/9-75-010. U. S. Environmental Protection Agency, Region X,
Seattle, Washington.
Osipov, Y. S., 1972: Diffusion from a point source of finite time of action.
In AICE Survey of USSR Air PollutionLiterature -Volume XII,
distributed by National Technical Information Service, Springfield,
Virginia.
Pasquill, F., 1961: The estimation of the dispersions of windborne material.
Met. Mag., _9_0, 33-49.
Pasquill, F., 1962: Atmospheric Diffusion. D. Van Nostrand Co., Ltd. -
London, 297.
Rubin, E. S. and H. T. Bloom, 1974: Evaluation of air pollution control
strategies for attaining secondary air quality standards in
Allegheny County, Pennsylvania. Project No. 196, Pennsylvania
Science and Engineering Foundation, Harrisburg, Pennsylvania.
Rubin, E. S. and H. T. Bloom, 1975: Maintenance of ambient particulate
standards in an industrialized region. Paper No. 75-06.1 presen-
ted at the 68th Annual Meeting of the Air Pollution Control
Association, Boston, Mass., June 15-20, 1975.
Sartos, J. D. and G. B. Boyd, 1972: Water pollution aspects of street
surfaces contaminants. EPA Report No. EPA-R2-72-081, U. S.
Environmental Protection Agency, Research Triangle Park,
North Carolina.
-------
REFERENCES (Continued)
Turner, D. B., 1964: A diffusion model for an urban area. J. Appl, Meteor.,
_3(1), 83-91.
Uhrsenbach, W., 1977a: Private communication.
Uhrsenbach, W., 1977b: Analysis of Hi-vol Filters by Scanning Election
Microscopy and Energy Dispersive X-ray Procedures, Progress
Report No. 3 by Environmental Studies Laboratory of the Univer-
sity of Utah Research Institute, 1 June 1977.
Whitby, K. T. R. B. Husar and 3. Y. H. Lui, 1972: The aerosal Distribu-
tion of Los Angeles smog. In Aerosals and Atmospheric Chemistry,
G. H. Hidy, Ed., Academic Press, New York, 348.
White, H. J., 1977: "Electrostatic Precipitation of Fly Ash," J. Air
Poll. Control Assoc., 27, 114-120.
8-4
-------
APPENDIX A
MATHEMATICAL MODELS USED TO CALCULATE
GROUND-LEVEL CONCENTRATIONS
A.I INTRODUCTION
The computerized diffusion models described in this appendix
fall into two general categories: (1) Short-term models for calculating
time-averaged ground-level concentrations for averaging times of 1, 3,
8, and 24 hours; (2) Long-term models for calculating seasonal and
annual ground-level concentrations. Both the short-term and long-term
concentration models are modified versions of the Gaussian plume model
for continuous sources described by Pasquill (1962) . In the short-term
model, the plume is assumed to have Gaussian vertical and lateral con-
centration distributions. The long-term model is a sector model similar
in form to the Environmental Protection Agency's Climatological Dis-
persion Model (Calder, 1971) in which the vertical concentration dis-
tribution is assumed to be Gaussian and the lateral concentration dis-
tribution within a sector is rectangular (a smoothing function is used
to eliminate sharp discontinuities at the sector boundaries). Vertical
plume growth (a ) in the short-term and long-term models and lateral
plume growth (a ) in the short-term model are calculated by using tur-
bulent intensities in simple power-law expressions that include the ef-
fects of initial source dimensions. In both the short-term and long-
term models, buoyant plume rise is calculated by means of the Briggs
(1971; 1972) plume-rise formulas, modified to include the effects of
downwash in the lee of the stack during periods when the wind speed at
stack height equals or exceeds the stack exit velocity. An exponent law
is used to adjust the surface wind speed to the source height for plume-
rise calculations and to the plume stabilization height for the concen-
tration calculations. Both the short-term and the long-term models
contain provisions to account for the effects of complex terrain.
Table A-l lists the hourly meteorological inputs required by
the short-term concentration model. Lateral and vertical turbulent
A-l
-------
intensities o' and a' may be directly specified or may be assigned on
A EJ
the basis of the Pasquill stability category (see Section 3 of Cramer,
j2_t al_., 1975). The Pasquill stability cateogry is determined from
surface weather observations using the Turner (1964) wind-speed and
solar-index values. Mixing depths may be obtained from rawinsonde or
pibal measurements, or they may be assigned on the basis of tabulations
of the frequency of occurrence of wind speed and mixing depth (available
from the National Climatic Center for synoptic rawinsonde stations).
Potential temperature gradients may be obtained from measurements or
assigned on the basis of climatology.
Table A-2 lists the meteorological inputs required by the
long-term concentration model. Joint-frequency distributions of wind-
speed and wind-direction categories, classified according to the Pasquill
stability categories, are available from the National Climatic Center.
Alternately, surface wind observations may be analyzed to generate wind-
frequency distributions by time-of-day categories (night, morning,
afternoon and evening). Vertical turbulent intensities may be deter- "
mined from a climatology of actual measurements or may be assigned on
the basis of the Pasquill stability categories. Median mixing depths
may be determined from the seasonal tabulations of the frequency of
occurrence of wind-speed and mixing depth prepared by the National
Climatic Center. Vertical potential temperature gradients may be as-
signed to the combinations of wind-speed and stability or time-of-day
categories on the basis of climatology.
Table A-3 lists the source input parameters required by the
short-term and long-term diffusion models. As shown by the table, the
computerized short-term and long-term models calculate ground-level
concentrations produced by emissions from stacks, building vents and
roof monitors, and from area sources. Both the short-term and long-term
models also use a Cartesian coordinate system (usually the Universal
Transverse Mercator system) with the positive X axis directed toward the
east and the positive Y axis directed toward the north.
A-2
-------
TABLE A-l
HOURLY METEOROLOGICAL INPUTS REQUIRED BY THE
SHORT-TERM CONCENTRATION MODEL
Parameter
Definition
R
H
m
li
9z
Mean wind speed at height z (m/sec)
K
Mean wind direction at height z (deg)
K
Wind-profile exponent
Wind azimuth-angle standard deviation in radians
Wind elevation-angle standard deviation in radians
Ambient air temperature ( K)
Depth of surface mixing layer (m)
Vertical potential temperature gradient (°K/m)
A-3
-------
TABLE A-2
METEOROLOGICAL INPUTS REQUIRED BY THE
LONG-TERM CONCENTRATION MODEL
Parameter
Definition
p. (Table)
aE;i,k (Table)
Ta;k,£ (Table)
(Table)
H . (Table)
m;i,k,£
(Table)
Frequency distribution of wind-speed and
wind-direction categories by stability or
time-of-day categories for the &tn season
Wind-profile exponent for each stability or
time-of-day category and i*- wind-speed cate-
gory
Standard deviation of the wind-elevation
angle in radians for the I1-" wind-speed
category and k'-'1 stability or time-of-day
category
Ambient air temperature for the k stabil-
ity or time-of-day category and £*-" season
Vertical potential temperature gradient for
the if"1 wind-speed category and ktn stability
or time-of-day category (°K/m)
Median surface mixing depth for the i^"1 wind-
speed category, k stability or time-of-day
category and £t'1 season (m)
Mean wind speed at height z for the i wind-
speed category (m/sec)
A-4
-------
TABLE A-3
SOURCE INPUTS REQUIRED BY THE SHORT-TERM
AND LONG-TERM CONCENTRATION MODELS
Parameter
Definition
Stacks
Q
X, Y
z
s
h
V
T
Building Sources
Q
X, Y
z
h
L
W
6
Area Sources
Q
X, Y
Pollutant emission rate (mass per unit time)
X and Y coordinates of the stack (m)
Elevation above mean sea level of the base of the
stack (m)
Stack height (m)
Actual volumetric emission rate (m /sec)
Stack exit temperature ( K)
Stack inner radius (m)
Pollutant emission rate (mass per unit time)
X and Y coordinates of the center of the building (m)
Elevation above mean sea level of the base of the
building (m)
Building height (m)
Building length (m)
Building width (m)
Angle measured clockwise between north and the
long side of the building (deg)
Pollutant emission rate (mass per unit time)
X and Y coordinates of the center of the area
source (m)
Elevation above mean sea level of the area source (m)
A-5
-------
TABLE A-3 (Continued)
Parameter
Definition
Area Sources
(Continued)
h
L
W
6
Characteristic vertical dimension of the area
source (m)
Length of the area source (m)
Width of the area source (m)
Angle measured clockwise between north and the
long side of the area source (deg)
A-6
-------
A.2 PLUME-RISE FORMULAS
The effective stack height H of a buoyant plume is given by the
sum of the physical stack height h and the buoyant rise Ah. For an adiabatic
or unstable atmosphere, the buoyant rise Ah is given by
AhN
/ \l/3 ,
' —^- (10h)Z/J I f (A-l)
_ u {h}
where the expression in the brackets is from Briggs (1971; 1972) and
u{h} = the mean wind speed at the stack height h (m/sec)
Y1 = the adiabatic entrainment coefficient ~0.6 (Briggs, 1972)
/ o
F = The initial buoyancy flux (m /sec )
;V ' ' a ' (A-2)
TT \ T
s
3
V = The volumetric emission rate of the stack (m /sec)
= IT r w
r = inner radius of stack (m)
w = stack exit velocity (m/sec)
2
g = the acceleration due to gravity (m/sec )
T = the ambient air temperature ( K)
3.
T = the stack exit temperature ( K)
S
The factor f, which limits the plume rise as the mean wind speed at stack
height approaches or exceeds the stack exit velocity, is defined by
A-7
-------
; u {h} < w/1.5
(3w-3"(h}) ; w/1.5 w
(A-3)
The empirical correction factor f is generally not applied to stacks with
Froude numbers less than about unity. The corresponding Briggs (1971)
rise formula for a stable atmosphere (potential temperature gradient
greater than zero) is
Ah
3F
6F
1/3
.u{hh2 s
1 - cos
/10S1/2h
u{h)
-1/2
;TT u{h} S ' < lOh
1/3
;TT u{h} S 1/2>10h
(A-4)
where
'2
S
the stable entrainment coefficient~0.66 (Briggs, 1972)
99
3z
vertical potential temperature gradient ( K/m)
The entrainment coefficients y1 and y? are based on the suggestions of
Briggs (1972). It should be noted that Equation (A-4) does not permit
A-8
-------
the calculated stable rise Ah to exceed the adiabatic rise Ah,... as
s N
the atmosphere approaches a neutral stratification (30/3z approaches 0)
A procedure of this type is recommended by Briggs (1972).
A. 3
SHORT-TERM CONCENTRATION MODEL
A.3.1
Elevated Sources
The atmospheric dispersion model used to calculate hourly
average ground-level concentrations downwind from an elevated continuous
source is given by
IT u{H}o a
y
{Vertical Term} {Lateral Term} {Decay Term} (A-5)
where
K
Q
u{H}
0 ,o
y z
scaling coefficient to convert input parameters to
dimensionally consistent units
source emission rate (mass per unit time)
mean wind speed at the plume stabilization height H (m/sec)
standard deviations of the lateral and vertical con-
centration distributions at downwind distance x (m)
The Vertical Term refers to the plume expansion in the vertical
or z direction and includes a multiple reflection term that limits
cloud growth to the surface mixing layer.
{Vertical Term}
,exp
+ exp
n=l
exp
- /2n H
1 [ m
1 /2n H - H
1 / m
H\2
(A-6)
A-9
-------
where H is the depth of the surface mixing layer. The exponential terms
in the infinite series in Equation (A-6) rapidly approach zero near the
source. At the downwind distance where the exponential terms exceed exp(-lO)
for n equal 3, the plume has become approximately uniformly mixed within
the surface mixing layer. In order to shorten computer computation time,
Equation (A-6) is changed to the form
/2rF a
{Vertical Term} = —^—- (A-7)
/n
m
beyond this point. Equation (A-7) changes the form of the vertical concen
tration distribution from Gaussian to rectangular. If H exceeds H ,
the Vertical Term is set equal to zero which results in a zero value for
the ground-level concentration.
The Lateral Term refers to the crosswind expansion of the plume
and is given by the expression
{Lateral Term} = exp
(A-8)
where y is the crosswind distance from the plume centerline to the point
at which concentration is calculated.
The Decay Term, which accounts for the possibility of pollutant
removal by physical or chemical processes, is of the form
where
{Decay Term} = exp [ - ^ x/u{H} ] (A-9)
= the washout coefficient A (sec ) for precipitation scav-
enging
A-10
-------
0^92
T '
Ll/2
where TI . Is the pollutant half life in seconds
for physical or chemical removal
= 0 for no depletion (\p is automatically set to zero by
the computer program unless-otherwise specified)
In the model calculations, the observed mean wind speed u is
K
adjusted from the measurement height z to the source height h for
K
plume-rise calculations and to the stabilization height H for the con-
centration calculations by a wind-profile exponent law
u{z} = U{ZR}
R
(A-10)
The exponent p, which is assigned on the basis of atmospheric stability,
ranges from about 0.1 for very unstable conditions to about 0.4 for very
stable conditions.
According to the derivation in the report by Cramer, ejt a_l. (1972),
the standard deviation of the lateral concentration distribution a is
y
given by the expression
ay{x}
a» x
A ry
x + x - x
Y
(1-a)
v
ax
ry
(A-ll)
x =-<
y
ax
ry
x a:
V ry A/
~~ ^1-
- XR+Xry(1'a) '
ry
ry
(A-12)
A-ll
-------
where
the standard deviation of the wind-azimuth angle in
radians
x = distance over which rectilinear plume expansion occurs
downwind from an ideal point source (~50 meters)
= the standard deviation of the lateral concentration
distribution at downwind distance x (m)
K
= the lateral diffusion coefficient (~0.9)
ry
7yR
a
The virtual distance x is not permitted to be less than zero. The lat-
eral turbulent intensity a' may be specified directly or may be assigned
A
on the basis of the Pasquill stability category.
Following the derivation of Cramer, et al. (1972) and setting
the vertical diffusion coefficient 3 equal to unity, the standard devi-
ation of the vertical concentration distribution
sion
a is given by the expres-
az{x} -
£ (x
(A-13)
X
> XT
(A-14)
where
a
zR
standard deviation of the wind-elevation angle in
radians
the standard deviation of the vertical concentration
distribution at downwind distance x (m)
A-12
-------
The vertical turbulent intensity o' may also be obtained from direct
ij
measurements or may be assigned according to the Pasquill stability cat
egories. When 0' values corresponding to the Pasquill stability cate
£j
gories are entered in Equation (A-13), the resulting curves will differ
from the corresponding Pasquill-Gifford curves in that Equation (A-13)
assumes rectilinear expansion at all downwind distances. Thus, a
values obtained from Equation (A-13) will be smaller than the values
obtained from the Pasquill-Gifford A and B curves and larger than the
values obtained from the D, E and F curves at long downwind distances.
However, the multiple reflection term in Equation (A-6), which confines
the plume to the surface mixing layer, accounts for the behavior of the
D, E and F curves (decrease in the expansion rate with distance) in
a manner that may be related to the meteorology of the area.
Following the recommendations of Briggs (1972), the lateral
and vertical standard deviations of a stabilized buoyant plume are
defined by
V = °zR
0.5 Ah
2.15
(A-15)
The downwind distance to stabilization x^. is given by
K
X
R
lOh
: If <-
u(h} S 1/2 ; || > 0 and TT u{h} S 1/2 < lOh
lOh ; — > 0 and IT u{h} S l'2 > lOh
a Z
(A-16)
A-13
-------
A.3.2 Application of the Short-Term Model to Low-level
Emissions
The short-term diffusion model in Section A. 3.1 may be used to
calculate ground-level concentrations resulting from low-level emissions
such as losses through building vents. These emissions are rapidly dis-
tributed by the cavity circulation of the building wake and quickly
assume the dimensions of the building. Ground-level concentrations are
calculated by setting the buoyancy parameter F equal to zero. The
standard deviation of the lateral concentration distribution at the
source CJ is defined by the building crosswind dimension y divided
yo j t> j Q
by 4.3. The standard deviation of the vertical concentration distribution
at the source is obtained by dividing the building height by 2.15. The
initial dimensions 0 and 0 are assumed to be applicable at the
yo zo
downwind edge of the building. These procedures are in good agreement
with the results of recent wind-tunnel experiments reported by Huber and
Snyder (1976). It should be noted that separate turbulent intensities
0' and 0' may be defined for the low-level sources to account for the
A hi
effects of surface roughness elements and heat sources.
A. 3. 3 Short-Term Concentration Model for Area Sources
The atmospheric dispersion model used to calculate ground-
level concentrations at downwind distance x from the downwind edge of
an area source is given by the expression
X(x, y} = - - — ^- - (Vertical Term}
(A-17)
2? u{h} 0 {x} y
z o
{Lateral Term} {Decay Term}
where
Q = area source strength in units of mass per unit time
y = crosswind source dimension (m)
A-14
-------
a {x} = -<
z
0' X
E o
x < 3 x
In
al(x+x /2)+h
Li O
x > 3 x
(A-18)
x = alongwind dimension of the area source (m)
h = the characteristic height of the area source (m)
The Vertical Term for an area source is given by
{Vertical Term} =
1+2 J exp
n=l
i / 2n H
II m
2 \a {xj
X
2ir a {x}
z
2H
m
i / 6H
1 / m
> 10
i / 6H
1 / m
(A-19)
The Lateral Term is given by the expression
{Lateral Term} =
-------
and
a {x} = o! (x+x /2) (A-21)
y A o
The Decay Term is given by Equation (A-9) above.
The concentration at points interior to the area source is
given by
? K n F°F (x'+1)+hl )
x{x'} = _ ^ In — V {Vertical Term} (A-22)
/2rT u{h} x y a' L a'+hJ\
O O Ji b )
\ '
where
x' = distance downwind from the upwind edge of the area source (m)
A. 4 LONG-TERM CONCENTRATION MODEL
A. 4.1 Elevated Sources
The atmospheric dispersion model for elevated point and volume
sources is similar in form to the Air Quality Display Model (Environmental
Protection Agency, 1969) and the Climatological Dispersion Model (Calder,
1971). In the model, the area surrounding a continuous source of pollu-
tants is divided into sectors of equal angular width corresponding to the
class intervals of the seasonal and annual frequency distributions of wind
direction. The emission rate during a season or year is partitioned
according to the relative wind-direction frequencies. Ground-level con-
centration fields for each source are translated to a common reference
coordinate grid system and summed to obtain the total due to all emissions.
For a single source, the mean seasonal concentration at a point (r, 6) is
given by
A-16
-------
2 K Q
TT r A9' .'4-f
S{9} V
1,K,
exp
[- i|; r/u. H.jk>
(A-23)
V. , = exp
"IK.!?
/H \2
1 1 i k £ \
2 la . , J
\ z;i,k, «,/ _
CO
^— >
n=l
exp
, /2n H . . -H. . \2
1 / m;i,k,£ i,k, a\
2 \ az;i k £ /
exp
2n H . , +H.
z;i,k,£
(A-24)
where
A9'
S{8}
frequency of occurrence of the i wind-speed category,
jth wind-direction category and kth stability or time-
of-day category for the £th season
the sector width in radians
a smoothing function
S{6} =
A6'
>!-e' I < AO'
ii-e'1 > A0'
(A-25)
the angle measured in radians from north to the center-
line of the jth wind-direction sector
the angle measured in radians from north to the point
(r,9)
A-17
-------
As with the short-term model, the Vertical Term given by Equation
(A-24) is changed to the form
/2? a „
2H z':\; (A-26)
m;i,k,£
when the exponential terms in Equation (A-24) exceed exp(-lO) for n equal
3. The remaining terms in Equations (A-23) are identical to those previously
defined in Section A.3.1 for the short-term model, except that the turbulent
intensities and potential temperature gradients may be separately assigned
to each wind-speed and/or stability (or time-of-day) category; the ambient
air temperatures may be separately assigned to each stability (or time-of-
day) category for each season; and the surface mixing depths may be separately
assigned to each wind-speed and/or stability (or time-of-day) category for
each season.
As shown by Equation (A-25), the rectangular concentration distrib-
ution within a given angular sector is modified by the function S{9) which
smoothes discontinuities in the concentration at the boundaries of adjacent
sectors. The centerline concentration in each sector is unaffected by con-
tributions from adjacent sectors. At points off the sector centerline, the
concentration is weighted function of the concentration at the centerline of
the sector in which the calculation is being made and the concentration at
the centerline of the nearest adjoining sector.
The mean annual concentration at the point (r,0) is calculated from
the seasonal concentrations using the expression
4
l V^
Xa{r,0} = ~ > . Xnfr.e} (A-27)
£=1
A-18
-------
A.4.2 Application of the Long-Term Model to Low-Level Emis-
sions
Long-term ground-level concentrations produced by low-level emis-
sions are calculated from Equation (A-23) by setting the buoyancy parameter
F equal to zero. The standard deviation of the vertical concentration dis-
tribution at the downwind edge of the building a is defined as the
z o
building height divided by 2.15. Separate vertical turbulent intensities
a' may be defined for the low-level sources to account for the effects of
tj
surface heat sources and roughness elements. A virtual point source is used
to account for the initial lateral dimension of the source in a manner iden-
tical to that described below for area sources.
A.4.3 Long-Term Concentration Model for Area Sources
The mean seasonal concentration at downwind distance r with
respect to the center of an area source is given by the expression
ro}
2 K Q
2? R A9'
u.{h) a
S{8} V
z;i,k
i,k,£
(A-28)
exp -
r' - r )/u.{h}l
o' i J
where
R = radial distance from the virtual point source to the receptor
vl/2
r =
distance from source center to receptor, measured along the
sector centerline (m)
r = effective source radius (m)
A-19
-------
y =
X =
y
lateral distance from the sector centerline to the receptor (m)
lateral virtual distance (m)
= r
(A-29)
z;i,k
In
' . . (r'+r ) +
E;i,k\ o/
1 . . (T ' -r \ + h
E;i,k\ o/
E;i,k
r1 + h
r <
o
6r
6r
(A-30)
3
1+2 N ^ exp
Z—/
n=l
, / 2n H
!_ m;i,k,
a . ,
z;i,k
TT
m;i,k,£
. I /6Hm;i,k,£ V
iY\ aZ;i,k /
' 2 \ a
> 10
< 10
z;i,k
(A-31)
and the remaining parameters are identical to those previously defined.
For points interior to the area source, the seasonal average
concentration is given by the expression:
2 K Q V^
/2if x y .
o o i, j ,k
-1- 5 J 3 "- J . J i
111
^i{h} 0E-' k
a' . (r"+l) + h
°E-i k + h
vi,k,^
(A-32)
A-20
-------
where
r" = the downwind distance, measured along the sector centerline
from the upwind edge of the area source (m)
A.5 APPLICATION OF THE SHORT-TERM AND LONG-TERM CONCENTRATION MODELS
IN COMPLEX TERPvAIN
The short-term and long-term concentration models described in
Sections A.3 and A.4 are strictly applicable only for flat terrain where
the base of the stack (or the building source) and the ground surface down-
wind from the source are at the same elevation. However, both models
may also be applied to complex terrain by defining effective stabilization
heights and mixing depths. The following assumptions are made in the model
calculations for complex terrain:
• The top of the surface mixing layer extends over the
calculation grid at a constant height above mean sea
level
• Ground-level concentrations at all grid points above
the top of the surface mixing layer are zero
• Plumes that stabilize above the top of the surface
mixing layer do not contribute to ground-level con-
centrations at any grid point (this assumption also
applies to flat terrain)
In order to determine whether the stabilized plume is contained
within the surface mixing layer, it is necessary to calculate the mixing
depth H*{z } at the source from the relationship
m s
H*{z } = (H + z - z } (A-33)
m v s' \ m a s/
A-21
-------
where
H = the depth of the surface mixing layer measured at a point
with elevation z above mean sea level
a
m
z = the'height above mean sea level of the source
Equation (A-33) is represented schematically in Figure A-l. As shown by
the figure, the actual top of the surface mixing layer is assumed to
remain at a constant elevation above mean sea level. If the height H of
the stabilized plume above the base of the stack is less than or equal
to H*{z }, the plume is defined to be contained within the surface mixing
Til S
layer.
The height H of the stabilized plume above mean sea level is
given by the sum of the height H of the stabilized plume above the base
of the stack and the elevation z of the base of the stack. At any eleva-
tion z above mean sea level, the effective height H'{z} of the plume cen-
terline above the terrain is then given by
H'{z} =
H -z; H -z>0
o o —
0 ; H - z < 0
o
(A-34)
The effective mixing depth H'{z} above a point at elevation z
above mean sea level is defined by
m
; z > z
H +z - z ; z < z
ma a
(A-35)
A-22
-------
0)
.c
4J
4-1
•H
01
C
•H
n3
4-1
d
o
u
0)
•d
a)
N
•H
cfl
4J
W
0)
,c
4-1
M
0)
X!
4J
QJ
•H
>-i
0)
4J
Q)
O
4J
3 •
V4
r-^-i OJ
cn>~.
N Cfl
PS 60
(3
JS -H
4J CO
PL, -H
QJ e
T3
a)
60 U
a rt
•H 4H
X M
•H d
W
Pi
a
M
Pn
A-23
-------
Figure A-2 illustrates the assumptions implicit in Equation (A-35). For
grid points at elevations below the airport elevation, the effective mix-
ing depth H'{z} is allowed to increase in a manner consistent with Figure
A-l. However, in order to prevent a physically unrealistic compression
of plumes as they pass over elevated terrain, the effective mixing depth
is not permitted to be less than the mixing depth measured at the airport.
It should be noted that the concentration is set equal to zero for grid
points above the actual top of the mixing layer (see Figure A-l).
The terrain adjustment procedures also assume that the mean wind
speed at any given height above sea level is constant. Thus, the wind
speed u above the surface at a point with elevation z above mean sea
-tv a
level is adjusted to the stack height for the plume-rise calculations
by the relationship
; h < z + z
o a R
(A-36)
where h is the height above mean sea level of the top of the stack. Sim-
ilarly, the wind speed u{H} used in the concentration calculations is given
by
u{H}
L [— -I ; H > z + ZD
R \ z^ / o — a R
R
; H < z + ZD
o a R
(A-37)
A-24
-------
S g
S "i
2. a
en
C
o
•H
cd
o
c
o
•H
C
0)
u
c
o
o
o
M-l
•H
o
ex
•H
i-j
60
O
t!
0)
C
60
•H
in
CO
- S
53
bO
C
•H
0)
•H
4-1
O
OJ
W
CNJ
W
Pi
o
A-25
-------
It should be noted that the terrain-adjustment procedures
outlined above provide a very simple representation of complex plume-
terrain interactions that are not yet well understood. Because the
model assumptions are generally conservative, it is possible that concen-
trations calculated for elevated terrain, especially elevated terrain
near a source, exceed the concentrations that actually occur. It should
also be noted that the procedures described above differ from previous
"terrain-intersection" models in that terrain intersection is only
permitted for a plume contained within a mixing layer. That is, terrain
intersection is permitted for all stability categories, but only for a
plume contained within the surface mixing layer.
A-26
-------
APPENDIX B
SEASONAL AND ANNUAL FREQUENCY DISTRIBUTIONS
OF WIND SPEED AND WIND DIRECTION
Tables B-l through B-5 list the 1973 seasonal and annual sum- "
maries of the joint frequency of occurrence of wind-speed and wind-direction
categories, classified according to the Pasquill stability categories,
at the Greater Pittsburgh Airport. These wind summaries were developed
using the Turner (1964) procedures which use surface wind-speed and cloud-
cover observations to estimate the Pasquill stability category during each
hour. The corresponding 1973 seasonal and annual wind summaries for the
Allegheny County Airport are given in Tables B-6 through B-10. In the
absence of cloud-cover observations for Allegheny County Airport, the
Allegheny County Airport wind-speed data were merged with concurrent
Greater Pittsburgh Airport cloud-cover data to determine the stability
categories. The 1975 seasonal and annual wind summaries for the Greater
Pittsburgh Airport are presented in Tables B-10 through B-15.
B-l
-------
ooooooooooo
H
1
m
TABLE ]
2:
0
H H
U M
W i-J O
erf M g
M pq M
O <3 erf
1 H £3
P CO Q
M ,_] H
srig
!§•&
->
•— OOOOOOOOOOOOOOOOO
CO
- ooooooooooooooooo
o m ooooooooooooooooo
>- U 40
at ui
o v>
« 1 OOOOOOOOOOODOOOOO
Ul ...
•- fl-
CC >-<| OOOOOOOOOOOOOOOON
i— a*« ooooooooooooooooo
to
0
o - ...
0
>- u —
(J O
Ul
>- Ul 0
dD X ......
(/)tO tOU)W 3CXX4
UJ lu xuJUitntoa^XK-
M en o
B-2
-------
I
M
W
iZ
O
H H
O M
W hJ O
Pi M 13
M PQ M
Q <£] p^
I H P
P CO O
M _!
13 O" I
<£ CO
Q P-i
W
W W
PM X
co H
P O
E2 H
M
& O
Pn M
O P
W O
a o
13 U
W
CO iH
H
H fn
M O
PM
_, O
Pi !Z
W M
H Pi
H w
o w
W H
H
H
O
W CO
& W
O- M
W Pi
pi O
Pn O
W
H H
M U
'O
Pi
O
u
OOOOOOOOOOOOOOOOO
— oooooooooooooooo —
I OOOOOOOOOOOOOOOOO
M OOOOOOOOOOOOOOOOO
OOOOOOOOOOOOOOOOO
>- o -^
o o
OOOOOOOOOOOOOOOOO
OOOOOOOOOOOOOOOOO
t~ ^ o
-^ oooooooooooooooor*
B-3
-------
« m — od—-*-^«sfs_
O 00000000
o- o o o o o —
O O O O O
^
Sx
H H
a H
W iJ O
erf H 2;
M pa M
P <} erf
I H D
P CO P
M _J H
13 KJ &
H O
O P Pj
!3 cr erf
•33 CO H
<3 <3
O Pu
w td
W W O
PL, tr! Crf ro
C/D H t3 r^
i m a\
P O CO rH
ro g: H H
I M H Pn
« ^ O M O
^: cu ,
w PH M erf
(_j o Q erf w
^ fd O H S
H U U <| 5
J3 O W W
W <3 Crf
erf o w
C» P 23
tD W W H
CJ 1-1 ffi
O fu H
O H
en H
PH co <:
O i CJ W
u w
a •> M
Cd oo Crf
C3 W O
O- H O
Ed erf W
erf o H
PM o <
w o
H H
a CO
v- a.
— Crt
s.r> o-oooo-ooooo — OOOOOT
UJ
t OOOOOOOOOOOOOOOO —
f*> •^e ac
o
o ooooooooooooooooo
•^ oooooooooooooooooo
V<-I-»< «OOOOOOOOOOOOOOOOO
_ UJ O
X - • • -
a: >" -«
*- A
(A
OC UJ O OOOOOOOOOOOOOOOOO t OOOOOOOO— OOOOOOOO
UJ C -<
t— ^/ O
Ull OOOOOOOOOOOOOOOOO -• OOOOOOOOOOOOOOOO-t
—I Z ^
>-, o o at
CD jr »r> ooooooooooor>oooor> «- o
3C 3C X
(/>
B~4
-------
—> o o o o o •
OOOOOOOOOO
o
H H
O H-1
U J O
(M H E3
M CO M
Q J
pi w H
Pd O H fd w
U H EC
O PM EH
O M
CO H
PIH en <|
O ^
O
CO
W
M
co pi
Pd O
• M O
Pi W
O H
O <
H H
55 be u
O OOOOOOOOOOOOOOOOO *-• t-
—• OOOOOOOOOOOOOOOOO Cf VI
I OOOOOOOOOOOOOOOOO 1-1
1*1 OOOOOOOOOOOOOOOOO O
jc ooooooooooooooooo . (j
t/> on m .-« — .— 4^-^^M— • •-«ooj*r*>ro — o f>
lu
VO OOOOOOOOOOOOOOOOO t- 3k
I OOOOOOOOOOOOOOOOO
OOOOOOOOOOOOOOOO —
o
•— OOOOOOOOOOOOOOOON
O OOOOOOOOOOOOOOOOO
(_) -* « OOOOOOOOOOOOOOOOO
uj ooooooooooor>oooof^ -^
N. I OOOOOOOOOOOOOOOOO
«r -^ ooooooooooooooooo to
UJ >-(_>! OOOOOOOOOOOOOOOOr4
UJO (MLUro OOOOOOOOOOOOOOOOO
>— '~* tJc«coMOO-"-^'*>'*>'— 's^^-rv>^i*^^-'i**1's-
r>^ooor>r»ooooooooooo »-a.
^ OOOOOOOOOOOOOOOOO _J
I OOOOOOOOOOOOOOOOO I-H Ck~«
f^>
^ OOOOOOOOOOO^OOOOOO O
s-- ft ooooooooooooooooo oor^MMorg^r^rgrorv-ro^tr-^or)
UK- ooooooooooooooooo ro rou»*rm«s»mr^N-»**r»»>h- — — o
UtO OOOOOOOOOOOOOOOOO I OO(>O<*
V> t" • ** OOOOOOOOOOOOOOOOO
\ ...
•r —
^, o
ooooooooooooooooo m <>oooomno*o»-«ovsij»into^-
OrO OOOOOOOOOOOOOOOOO O -* O - tftJOO— OOO«-«OOOOO
UJ t OOOOOOOOOOOOOOOOO *•• OOOOOOOOOOOOOOOO— i
a. o
3C /^>
a o ot
X tO OOOOOOOOOOOOOOOOO i— i O
31-* OOOOOOOOOOOOOOOOO U>O XXX (rttrtC''1 l/)tOl/l JCXXCC
I OOOOOOOOOOOOOOOOO UJ UJ 3C UJ UJ T >f ~X Zt XI—
o ac to o
c»
I --
o at
•-> o
£_>(_) XX3C (OtOU) C->t/> 3C 3C 3 a
tU UJ X UJUJ(/l{n 3 3» 7. t-
B-5
-------
13
O
o o o o o
OOOOOOOOOOOOOOOOC*
OOOOOOOOOOOOOOOOO
ooooooooooooooooo
OOOOOOOOOOOOOOOO C*
2;
o
M
H
o
w
- O O O O <
oooooooooo-ipjoo—-ori
OOOOOOOOOOOOOOOOO
ooooooooooooooooo
O-CXOOOOOOOOOOOOOOO
Q >-i
I H
Q H
53 H4
M H
Q H (6
*Z CO O
•5 P-I
,J Prf
Q iJ M
W H <3
W
O"
O
— OOOOOOOOOOOOOOOO-
LO
I
PQ
M W CO
fe W H
H H
(^ M
00^
W 0^
u o w
S3 2 H
W i— i <3
s o o
u o
CJ O W
o -l fn
CO H "
S5 tn to
w co w
P < M
o- i-J oJ
U CJ O
K O
PH - W
CO f-l
H W tO fl 10 CO (/>
<_) (_J
QC
-------
c c>oc-ooooooo
ooooooooooo
ooooooooooo
• OOOOOOOOOOOOOOOO
Z •— u- —•
O C> O O
O C O —
OOOOOOOOOOO OOOOO
ooc ooooooooooooo
o o c oooooooo-ooooo
O O O OOOOOOOOOOOOO
H H
O M
W i-J
on
N -SJ
E
O CM
UJ
UJ in
ooooooooooooooooo
OOOOOOOOOOOOOOOOO
OOOOOOOOOOOOOOOOO
•OOOOOOOOOOOOOOOOO
UJ t£ OOOOOOOOOOOOOOOOO
3_ ~ — 0~0 — oo — 000—0:0 —
•-*— OOOOOOOOOOOOOOOO—'
3 i OOOOOOOOOOOOOOOOO
I
PQ
w
1-J
H &
CO M
f""l l~-1 E—I
Z O* Pi
< CO O
< p-t
p P-I Pi
W M
w w <;
P O 55 iH
J3 H P
H O PLJ
J3 O U O
^
fn H >H pd
o Q is PJ
Pi W H
w o w a
O O O H
53 O W !3
W '
Pi
O M H
O fe -i o Pi
G o
55 " O
W CO W
p W H
cf M <;
[d Pi U
Pi O
Pn O
w
5S -(_>( OOOOOOOOOOOOOOOOO
>- LU ra ooooooooooooooooo
t— a.
—• (/* tn
— ft. l«— OOOOOOO—'O-^OOOOO*^
I— 3» —' OOOOOOOOOOOOOOOOO OOOO — OOOOOOOfSJOoOO
\j IOOOOOOOOOOOOOOOOO
zzzcc
B-7
-------
OOC'CC
• o- o o o
o <: ooooooooooooooo
oooo-c-ooooo-c oooooo
OOOOOOOOOOOOOOOOO
oo 00000© ooo-oooooo
I*. QC OOOOO-OOOOOO — «
PQ
Ed
5!
H
53
O
M
H
U
fd
Pi
M
P
1
P
M
P
Ed
Ed
P i
CO
P
M
En
M
U
Ed
U
U
O
En
U
^
Ed
C3
O1
gj
H
53
M
0
>-i
—I
-1
_}
-<
-
OQ
-
-J
m
«
>-
CK
CATEGO
~j
:
a
UJ
O-
Q
3
Z
O
o
DC
OOOOOOOOOOOOOOOOO ->
ro ooooooooooooooooo c.
c
CO C,
•>, ^
1 OOOOOOOOOOOOOOOOO >
fM OOOOOOOOOOOOOOOOO t-
trt
o
in rofj— '«— f*J(\jr«j»ijO'j)rorM(o-HOfNj«j» y
I OOOOOOOOOOOOC>OOO*T
M
O L
*.*» ooooooooooooooooo
oooc-oooooooooo-^c- —
o
V
— oooooooooooooooorw
c. ooooooooooooooooo
t—
« s?s?ss"S«::sssss^
o
s« OOOOOOOOOOOOOOOOO
—
o - • •
<
o
1 OOOOOOOOOOOOOOOOO
-
ooooooooooooooooo
— ooooooooooooooooo
0
oe
o
o 2:xz u>tnto VJtntn xacaco:
C^ 0
i_) r— OOOOOOO^OOOOOOOO'-O
[. UJO OOOOOOOC-OOOOOOOOO
3 trt r-
3 X
J X!
v^ 0
Uj' OOOOOOOOOOOOOOOOL-)
UJc>OOOOOOOOOO
Q- r^
1/1
o
-7 ooSSSSSo-oSSSoooS
o
O OOOOOOOOOOOOOOOOr-
^-
OOOOOOOOOOO — OOOO^
O OOOOOOOOOOOOOOOOO
C9
1C WW OOOOOOOOOOOOOOOOO
J X O5
* — N
- d-
-• (/> m
_J
f->
o
^ OOOOOOOOOOOO^OOO— i
z "•
0 OC
fcrf O
cjo z:2:z uitoco o^coto 3:3ca:a.
ac i/> o
B
"~
-------
c o o o o o
OOOOOOOOOOOOOOOQ
-oooooooooooooooo
OOOOOOOOOOOOOOOQ
x- a: (_> m
•OOO-T^OOC*^ —,«--..
00
I
s
o
M ^
H H
U M
W hJ
PJ M
M pq
fi < O
I H ^
Q CO M
IS £i
M (-J &
^2 *1 P
P P
Z O-
^ C/3
« PM
W
W W
PH Cd
W H
I
O O
K H
M
ts o
K
fn M
° 9
W O
O O
13 O
W <
s
. p
e> w
o H
O pn
O H
W
PM I/i
O <3
S^
& «
W co
C3 W
0-M
pij pc;
S O
Pn O
W
H H
!3 <3
M 0
O
!~3
n
r^
ON
O Pt<
CJ O
>-< PiJ
2 W
W S
*T< ^
H-4 1^4
o t>
W CO
I-J
^1
s
CO
W
M
PC!
o
o
W
H
O f, OOOCfOOOOOOC-OC/OOOO C< OJC OOOOOOOSOOOOOOOO'
o v» h-
>- Cj CO - UJ in -J
K- a.
_j *—
t— 3---I OOOOOOOOOOOOOOOOO
(/)
-.O OOOOOOOOO^OOOOOOO O 1f>
U X
— UJ JC —•
o •-* in
C3 OOOOOOOOOOOOO'OOOO
OB
OO OOOOOOOOOOOOOOOOO
-J I OOOOOOOOO-^OOOOOOtfl
•H OS* OOO-OOOOOOOOOOOOOO 00
CD Z
t— :» '^-'ooooooooooooooooo
oooo — ooooooooooor> OCA
I OOOOOOOOOOOOOOOOO UJ «:
O • ' >- v- M
>- UJCX OOOOOOOOO-OOOOOOOO
M- a.
•— irt tn
>- UV— OOOOOOOOOOOOOOOO-— CO X O —'-»'*r*T(v>'-<'v>f*»—' ^ CO OB ao ro ff* \*
UJ *r
t— ^ O f^>
i—a. loooooooor*) — — o — oooro
co z m
1-^ OoOOOOOOOOOOOOOOO
C-OO— 'i
B-9
-------
OOOOOOOOOOOOOOOO
ooooooooooooooooo
o r*i
>• O CP
>- i_>i~ —'Oooj—> o o —• rv "•--• — ^- o o o f^-
OC U.O OOOOOOOOOOOOOOOO-"
o
H H
CJ M
W i-3
Pi H
M fQ
Q <; O
i H a
O co H
ft t3 H
a o-pi
<3 CO O
ED
P-l 05
H
w w -l CJ Pi
CJ O
Pd co pa
P W H
ex M <;
Pd pi U
Pi O
Pn O
Pd
JO
o
*- ^ to oooooooooooooooo-*
o
-.0 OOOOOOOOOOOOOOOOO O 10
*T7 O O
o ooooooooooooooooo
1 OOOOOOOOOOOOOOOOO
OOOOOOOOO-OOOOOOOO
« _1
OOOOOOOOOOOOOOOO —
OOOOOOOOOOOOOOOOO
— m oo co -^i •*-
— t/> —
QO x n
OOOOOOOOOOOOOOOO--
CfOOOOCiOOOOOOO-OOOO
OOOOOOOOOOOOOOOOO
x z a:
-------
OOOOO
OC-OOO
OOOOOC-OOOOOOOOOOO
OOOOOOOOOOOOOOOOO
OOOOQOOOOOOOOOOOO
ooooooooooooooooo
r- co «r> \r> r-- \,->
OOOOOOOOOOOOOO
I
pq
H
o
W
erf
S£
I M
Q i-l
2; M
H eg
*~g •<:
H
P CO
H M W
U en H
2; co erf
H <^ O
S ij o
O* U W
Fn CO U
W
H M
H O
o o
f-3 W
OOOOOOOOO^OO-nOOOO'T
OOOOOOOOOOOOOOOOO
OOOOOOOOOOOOOOOOO
>• UJ 10
«— din moo'-ooryeooo^ooroi-o^toror-jco—•
ooooooooooooooooio
o o o o ,o
ooooooooo
I— OOOOOOOOOOOOOOOOO
ooooooooooooooooro
! OOOOOOOOOOOOOOOOO
eo ooooooooooor-oooor^-
OOOOOOOOOOOOOOOOO
O OOOOOOOOOOOOOOOOO
—• OOOOOOOOOOOOOOOOO
(t>
XXX (rtCOV)
•B-ll
-------
*-• ooooooooooooooooo
o o o o o
ooooooooooooooooo
I
cfl
w
55
o !>•<
M H
H M
U _}
W M O
Pi PQ 2
H •< H
P H g
P P
H ,J H
£2 H Pi
P O
P O* PL,
P-i
co p r~>
I O PQ cr»
P H co r-H
M O H (t(
^S 25 M O
M pu,
O S Pi W
O W H
W O H S
o o <; H-i
PJ S
pi p o w
pi W ffi
P H W H
O fe p~t
U M H
O CO
co H
O ,-4
U «
>-• CO
o « w
53 CO M
W W Pi
P M O
O1 Pi O
WOW
Pi O H
fl-l fT1 - LJ —•
I OOOOOOOOOOOOOOOOO
>- ID O
ooooooooooooooooo
O (J
UJ U
DC (/>
: z eft c/> trt cfl to w>
LU UJ C/> C8 31
B-12
-------
O O O O O i
O O O O O
Pd
w
pa
JOINT FREQUENCY OF OCCURRENCE OF WIND-SPEED AND WIND-DIRECTION
CATEGORIES, CLASSIFIED ACCORDING TO THE PASQUILL STABILITY
CATEGORIES, AT THE GREATER PITTSBURGH AIRPORT DURING
THE SPRING OF 1975
0 — t-
— ' at u »
a. ar ot «
O OOOOOOOOOOOOOOOOO (A i— h-
— OOOOOOOOOOOOOOOOO O tn
1 OOOOOOOOOOOOOOOOO •-•
u) r*> ooooooooooooooooo tb
U CD
V)
u x: <^oooooooo«Nor>ooor?a) - o
O- DC Ul •-•
CO O C/>
-• UJ O
»- (L
O -J 1
ui in
o
to
^ o
«•* «
UJ >- t_J 1
_J 1 OOOOOOOOOOO-OOOOO<~> UJ K
« X Vt Vt V> W IA XXX
-------
^~*| A .
W ffl
fd W O
PL| EC Pi LT>
w H p r^
n P O C/3 r-l
iH 2; H H
I M H Pn
pq tS O H O
oo-oooooooooooooo.-
CO OOOOOOOOOC»OOOOOOO
OOOOOOOOOOOOOOOOO
o ooooooooooooooooo
-< ooooooooooooooooo
o ooooooooinooooooor)
.-• OOOOOOOOOOOOOOOOO
j OOOOOOOOOOOOOOOOo
ro ooooooooooooooooo
o
M >l
H H
U M
W »J O
Ptf M IS
M PQ H
P <; Pi
I H P
P en p
M ,-J H
;s .-! Pi
M O
P P PL,
z o- PS
ai ^- ooooooooooooooooo
\0 OOOOOOOOOOOOOOOOO
u on
O P Pi W
prf W g
W O H S
U U H O en
U W
•Z, •> M
W uo pj
P W O
O- M O
W Pi W
Prf O H
FH o <;
W O
H H
2: <;
— oooooooooooooooory
« n — i
OtJ
at co
B-14
-------
o o o •
~* ooooooooooooooooo
s^
H H
O H
O
I H p
Q co p
M H-] H
s 1-J pi
M O
Q P P-i
z o- pi
-l O CO
u w
^ »\ I—I
W CO pi
p W O
cr M u
f£J pj fjj
Pi O H
PL, O O
•— I OOOOOOOOOOOOOOOOV
:» o ooooooooooooooooo
1U
z
3
I OOOOOOOOOOOOOOOOO
o
*~ OOOOOOOOOOOOOOOOf)
>• o —« « ooooooooooooooooo
o o? to oooooooooooooooon
3- UJ o aeum ooooooooooooooooo
_J k OOOOOOOOOOOOOOOOO LU C
M O •* OOOOOOOOOOOOOOOOO K vN
CD Z • flC ^*^-«*M^^^^OOOO^|OOr*O<^*
U) o OOOOOOOOOOOOOOOOO
QC UJO OOOOOOOOOOOOOOOOO I OOOOOOoO— OO-HOOOO(M
t— ^> O
o an ooooooooooooooooo ooo<->Nr*>»->h)ir)Nrtf-Mvtooor^
Ul> OOOOOOOOOOOOOOOOO —• OOOOOOOOOOOOOOOOO
>- UJ •* OOOOOOOOOOOOOOOOO I OOOOOOOOOOOOOOOOO
i— a. o • -
M 6O *-
_j ac ^
» O OK
CO ZIO ^-.—.« — —.*-! — -.-.«.«-*«—.-.». MO
»~ a — ooooooooooooooooo LJ *_» z x >: tot^to w> y> w z z z a
B-15
-------
GO OOOOOOOOOOO-^OOOO"-*
OOOOOOOOOOOOOOOOO
O OOOOOOOOOOOOOOOOO
<-< OOOOOOOOOOOOOOOOO
53
O
H
H
O
W
Pi
O H
I M
P i-l
55 M
M PQ
H
P cn
o oooooooo*-ooooooo —
—• OOOOOOOOOOOOOOOOO
I OOOOOOOOOOOOOOOOO
ro ooooooooooooooooo
X OJ
V CO OOOOOOOOOOOOOOOOtM
O <•« OOOOOOOOOOOOOOOOO
t- OOOOOOOOOOOOOOOOM
>• UJ IT)
i-J M
Q H-1 -) W
, _]
-------
APPENDIX C
INDUSTRIAL PARTICULATE EMISSIONS INVENTORY
This appendix contains the industrial particulate emissions
inventory for the Southwest Pennsylvania Intrastate Air Quality Control
Region (AQCR) in the format required for input to the computerized dif-
fusion models described in Appendix A. With the exception of the Source
Type, the emissions inventory is self-explanatory. Source Type 0 is a
stack source, Source Type 1 is a building source and Source Type 2 is an
area source. Sources 8001 through 8015 are 2-kilometer by 2-kilometer
area sources centered on the locations of the 15 hi-vol samplers used in
the special monitoring program described in Section 3. These sources are
assigned unit emission rates (1 ton per year) so that the concentrations
calculated for the special monitoring sites attributable to localized
emissions of resuspended roadway dust can be scaled to the various emis-
sions estimates. A more detailed discussion of the emissions inventory
is given in Section 5.
C-l
-------
Pi
o
C!
;S W
O t>
M 3
CD M
W
Pi CO
r-1 0
CD I I
pi co
£—1 CO
Z M
O S
o w
PH W
H H
M <<
Pi
iJ W
% 1
2: o
z: cj
W
P-. rj
&H i — i
CO Pi
&\ H
£5 CO
IE P
I"1 Q
O IH
CO
r--
en
i — i
a
•H
4-J
QJ
E
^
, — |
0
v!
U
ca
4J
CO
QJ
4-J
CO
1
3
CJ
•H
4J
S-l
CO
p-t
1
ft i--J i—
^0 -J fz;
4-J -r-l ^—-
co n3
(0
Pi
C ''""•^
o -x u
•H -X QJ
WO) C/}
03 4-J \.
•H ct3 CO
B Pi B
w ^
• -K
4-J p , ^-N
•nex;
X OJO
W H — •
a
O '-N
^ -H r-3
O -U CO
CO CO 2
4J >
co aJ B
f— { ^— x
I-M
jj
cj &0 /— s
cd -H g
4-1 QJ s— • '
cn ;x;
>^
£J ^
p ^ — '
P
X
'->
6 ^
P
Ot.
M
•H a) >•>
CO 4-J *~~--
co rt C
•H PJ C
B 4-i
w • — '
Q)
a
EH
"O
C
CO
*v
QJ S-i C
B QJ O
CO J3 -H
Z B 4-J
3 CX
QJ Z -H
cj S-i
S-l QJ 0
3 U CO
O S-l QJ
CO 3 Q
0
CO
o
4-J
C
QJ m
*rH ^-3"
rC) VO
m m
in - — i
^£) OO
m oj
01
r-
on
0 0
m
-*
OJJ
C
•H —I
Cl ^-"^
QJ
QJ S-l
S-l rH
cj rn
CO
QJ ""
vj i^>j
O
CJ pq
un ~
01
01
o
m
o
i —
m
a^
o
o
on
i — i
O
OJ
=;^
^
pq
W
CJ
in
£N
O CT^ CT\
C3^ CT\ C^
r~ o o
CT\ in in
•— i 04 ol
*^D O O
r-H O O
VD in in
f3~\ CTv (y\
OJ Ol CN
CN Ol Ol
-3- 01 oj
oj m m
000
O O .-H
oo
r-H CN OO
CN =
rH
pq S-i r^
1 — 1 1 — 1
QJ pq pq
C
N QJ Q)
C iH rH
QJ -H -H C
pq Pi pi -H
J>
S-i
\£> r^ 00 H
o-l 01 01
CO
co
P
01 O O O O
CTi "*O ^O ^O O
i — 1 i — 1 i — 1
CN i— i CK oo on
co co in in on
-3-01 on CT\
m o ^o on on
in ^o ^D ~3~ on
<3~ ^o on vo in
on on co on on
oo oo co co oo
Ol OJ CN OJ Ol
r-~ oj a> in r— i
CN in rH r^ 01
O O O O o
•-H o O in O
f~- ^D m VD ^O
in in in in in
vO vO ^O **O *sO
^3" ^vj" -CT ~3" ^3"
cTi cr^ cr\ o>
m m m m m
oo o o CN en
m r^- oi in in
co i — i o o^
•— i
on
CO
m
on
00
ol
, — t
01
O
O
r^
m
^15
01
o
^-
1
r-H
^ii=
[1
CO
QJ
PS
j j
,n co
cd c
.H 0
CO -H
4J
cO
m z
CO
CO
CO
0
o o o
CO CO CO
i — 1 rH i — 1
a\ en o
01 r-H r-1
01 01 01
O ol m
co en co
^0 vX5 ^O
0 O 0
co on on
OJ OJ CN
vO **O ^'D
^3" ~3" ^}*
O O O
on on co
on on on
r-~- r-~ r-~
*-O ^D ^O
-3- O^ O^
CN OJ 01
OJ 01 og
O r^ o
rH 04 CO
O O o
O 0 0
in in in
r- r^ r-.
^o ^o ^-O
-3- CO CO
<)- 01 co
I-H on
-------
a
*rH
4-1
0)
g
3
r— 1
O
^
O
cd
4-1
en
0)
cd
r*H
3
U
•H
4-1
^1
cd
PH
^
O
cd
CO
o
•H
to
CO
•H
E
W
4-J
•H
X
w
_
o
cd
4-1
CO
o
cd
4-1
CO
>-*
g
EH
^>
X
V-H
H
^D
c
Q
•H
CO
CO
•H
E
w
0)
p.
>^
H
"
0)
g
cd
Z
O)
a
3
o
CO
-JC
CO x-s
3 E
•H •— '
TJ
cd
Pi
-X O
•K 01
0) CO
4J — -~
cd ro
P- E
*~*
• -5i
a. x^
E ^
0)C
H ^
C
O x-v
•rH i — 3
JJ 01
03 21
£>
01 ^
PJ
r-1
•H £
CLt v- — '
&
*&
^ — s
-^
0) t*.
4-J •— »
cd c
ci o
4-1
- — '
r;
rd
!-i C
0) O
-9 'H
6 4J
3 P.
z; -H
S-i
d) CJ
CJ CO
!w 01
3 Q
O
CO
OOOO O r~^ in
01
4-1
C
v^oloiOO ' — ^OO rorocococoOciJO
^o 01 01 oo in in in o-i ro c^ ro m ^ *r-i in
^f-roroin in inininininin^
4
— inino oo>a> c^a^OAaNt^oia^c^'
olc-ororo rooioi oiojojoioioioioi
ojolojol olojoi oiojojoloiolojoi
C) o*\ Q>I - - •<}" co cr\ o c\j i — * oo O
cri-^ruoco ^LOCO cococo^j-srcococo
^r--r^r^ cooooO csitNCNiCvicN^r1— i-— '
cr^ CTI CPI o^ o^ o^ o^ o~\ CT^ o^ cr\ cr\ &\ cr* &\
LO LT~1 UO UO UO LO LO L/O LO LJ™1 *-O I-O ^O UO k-O
^
a en
c co -£: ~fc ~**^ J-( J*J C^ f A
r-H0303>^ LO'-dPHp-t -M J-jCO-M -U
!-i -HUCJX 0) LO LO LO
CM CN r— t OJ CO CO
CM CN CM! ON CM CM
i — * r^**" CNI iQ vo |>O
\£) **~H ^r co
00 O O O O O
\D O O 00 ro OJ
r- m >
- QJ O O
•-I 4-1 Cd X
P-l M ^J 4-1 OJ Oj
CO -rl 3 ^ ^
Lr, CJ i-H O O O
pq co w co CJ U
O t— i ol co
-------
CJ
•H
S-i
4J
0)
B
3
r-H
O
>
.M
O
cd
4_1
in
a)
4—1
cd
r-l
3
0
•H
4J
r-l
ra
P-,
^
o
cd
4-1
CO
C
o
•H
en
en
•H
B
W
4J
•H
X
W
^
U
m
4J
CO
^!
CJ
ed
4->
CO
>>
f^-4
H
P
X!
y-t
H
£3
C
O
•H
CO
cn
•H
e
w
0)
D-
>-,
H
CJ
E
cd
2
OJ
CJ
r-l
3
o
CO
-X
-X
•X
cn ,-s
3 e
•H ^
TJ
nl
ty
PS
^^,
•X U
-X QJ
QJ cn
4-1 ~~-~
n3 c-"l
f* ^
a A
B U!
QJO
H v-x
C
0 /-^
•H J
4-) CO
n) S
>
OJ g
i— 1 -^x
W
4J
x:
oc.^
•H e
QJ ^-^ ^
K
'B'
^_,
LJ
OJ >,
4-) ~—
cd c
oi o
4J
1 — ^
-a
c
cd
!-i C
QJ O
X! -H
B 4J
3 G-
Z -H
t-l
QJ 0
U CO
r-l CD
3 0
O
CO
o o o
o o o
• • •
o o o
• • I
o o o
000
00 CO CO
r\ f r.
CTs Cn
CO CO CO
LO LO LO
CO 00 CO
• • •
, — / 1 — 1 , — I
r-H i — 1 i — I
r-- r~ r~-
o o o
cn w to
x: p; a c
M QJ QJ QJ
r-l > > >
a '^ o o o
XI T3
en aj a) cu QJ
4-1 3 ^ ^ ,M
4-1 C O O O
•H -rt O O U
PL, 4J
c
,-J O v£> r^ CO
U O O O
u3 -^-^ , — i i — I i — I
>-)
O
O
.
CM
O
.
-cr
uo
, — i
CN
CO
LO
CM
CN
CN
CO
CO
O
LO
r~
CO
r^
•JCNCNCNCNCNCNCNCNCNCNCNCNCN
ooooooooooooooooooooo
cccccccccaccccccccccc
ajiDaiajajcuiiaiaJcuttJQJCuajQjQjajQJajcua)
>>>>>>>>>>>>>>>>>>>p>>
ooooooooooooooooooooo
ooooooooooooooooooooo
UOUOUOUUOOUOCJUOUOUUOU
OOOOOOOOOO"— If-Hr-lr-li-l.— Ir-l
C-4
-------
V
a
nj
4->
CO
O
•rH C
*-< O
4-J -H
CL> W
e w
3 -H
rH B
O W
O 4-1
cd -H
4-J X
CO W
a
nj
CO
a
CD
4-1
CO
£
tJ
X
s
P
CJ
cd C
r-H O
P -H
O U)
•H CO
iJ -H
CO LJ
01
CX
>»
H
0)
g
ra
^;
0)
0
}_l
3
O
CO
-K
-K
cn ^-v
3 e
•H N S
-d
5
^-^
•K O
•x cu
0) W
-u -\
n3 ro
^ p
^
4 ^
CX s~^
B W
aio
H -^
c
O /-N
*(H t— 4
4J CO
cd X
tu E
i — i • — '
w
li^
•H 'g
QJ '--^
PC
Is
9"
^^
l-l
QJ K^
4-1 ~~-
cd c
Pd 0
l)
-a
ca
t-i C
cu o
r-Q *H
g 4J
p p-
J_J
(1) CJ
a en
-------
-X
•X
^i -K
U CO ^-s
ca D a
CO T3
TO
pi
o
•H C *-
O -K C
4-i -H * a
a) en 01 a
6 en 4J --
•H n] c*")
e pi E
O W v-
o 4J cx /-
d -H 6 U
X 010
CO U H -^-
C
O .".
^ -H J
O 4-1 CO
to n) g
4-1 >
co a) g
i—i ^-^
pa
nj -H E
co rn
X
01
4-1
rH O 'ZT
3 -H a> ^
o w 4-1 ^,
•H en (U c
•u -H pi o
C
cd
ai o
rJ^ 'H
3 a.
0)
o
(-4 oi a
DOW
O M
ocTiOOvOr^oo>
OO
>>>>>>>>>>>>>>>>>
OOOOOOOOOOOOOOOOO
>>>>>>>>>>>
OOOOOOOOOOO
OOOOOOOOOOOO
UOOCJUOCJUUCJUU
OOOOOOOOOOOOOOOOOO
UCJUUUUUOOOOUOUCJUUU
C-6
-------
y
U
CO
CO
O
•H C
M O
4-> -H
0) M
B w
3 -H
<-< E
O W
>
^
CJ 4J
CO 'H
4-1 X
co w
V?
o
co
4-1
CO
a
to
^j
CO
>>
1
X
;=>
0)
4-1
r-t O
3 -H
o tn
•H W
4-J -H
cd w
eu
0)
a
H
0)
6
m
p^
0)
o
M
13
O
CO
•X
-K
-X
Cfl y-^
3 B
13
/^
-X U
•K 0)
ai c/i
4-) ~^.
rt <•<->
Pi E
•- — '
• -X
p , r — N
e^>
HM
01 0
H "-^
C
O '—-
•H H4
4-1 CO
to S
>
ai e
rH •-— '
J-J
fcO ,-N
•H B
ai -— -
W
y" N
^
y— V
E
t-t
CU ^t
4-J ~-^
cO C
Pi 0
il.
-a
c
m
M C
0) O
,0 -H
g 4-1
3 a.
Z -H
V-i
QJ O
CJ W
)-i 0)
3 Q
0
CJ1
W
^
J_J
0
t-5 / — v
TD
C 01
0 3
4J C
M -H
•H 4-1
CO C
r-H 0
O U
CO
CO
^3
oooooooooooooooooooooooooooo
oooooooooooooooooooooooooooo
oooooooooooooooooooooooooooo
•>
OO
OO
rr-.
oooooooooooooooooooooooooooo
UOUOUOOOUUOOOOOOCJUUOCJUUOUUUO
O- — lcN^o^^lJO^Dr^cX)O^O• — <(NrO
-------
y^
O
rt
4J
CO
o
•H c
rJ O
4-1 -iH
o) cn
E co
3 -H
•H e
o w
^.
O 4J
rt -H
4-1 X
u
rt
CO
o
rt
4J
i
X
*£-!
H
^
o3 £
-H 0
3 -H
a w
•H cn
4-1 -H
rt w
PH
01
a
>\
H
0)
E
rt
^
cu
o
1-1
3
O
CO
•X
_!^
^
CO /— ,
3 E
•r-l ^-"
13
rt
s~~.
•X U
-X 0>
01 Uj
4-1 ----
rt co
pc; g
^^
_,,
a. ^-s
e ^
010
H ^
C
O ^N
"rH (— J
4J CO
rt ^
0) E
W
4-J
cc ^-v
•H E
^^
~1=
'E
l-i
Ol >-.
4-1 --,
rt c
Pi 0
4— '
-a
C
rt
l-i C
Oi O
,£) -H
e w
3 P.
£2 *H
1-1
o) a
u cn
^J Ol
3 Q
O
CO
cn
,y
1^
p
2^ -• — s
13
C 01
O 3
4-1 C
1-J -H
•H 4-1
rt c
rH O
U O
CO
CO
ZO
oooooooooooooo
oooooooooooooo
oooooooooooooo
O
CM
O
CM
OO
i—i
04
ON ON ON
OOOOCDO-— !
O O
CM
CM
CM
CO
O
o
vO
O
CN
CM
OO
ON
o
o
CO
OOOO-H--iOOOOOOOO
oooooooooooooo
o
o
CM
LO
OO
ON
ON
o
oo
co
CO
ON
ON
wcnwcnwcnwcntocncncncnt/}
oo
ooo
O O O O
OOOOOOOOOOOO
oouuouucjuoacj
OO
uu
OOONO—i
OO1—!<—I
E
rt
Ol
CO
£>,
4-1
C
3
O
r^i
C
0)
"M
01
rH
rH
<;
4J
01
0)
l-l
4-1
CO
.n
4-1
CM
r-l
4J
cd
QJ
PC
1-1
rH
pq
O
vD
o
1 — I
e
rt
0)
CO
C
3
O
o
K*N
£
01
00
01
, i
rH
^
^
r*S
•H
^
c
rt
4J
CO
4J
rt
0)
PC
1-1
rH
PQ
O
r--
O
f — (
-------
^ •£
O <•/] ^^\
03 3 E
4-1 -H x_^
CO T3
OS
O
•r-t C •—-•
1-J O * O
4J -H -K 0)
O w 0) CO
g m 4-1 -~-
3 -H W ro
r-l E Pi 6
O W ^
r-^ • ':*
U 4-1 P. s~*
a -H g u;
•U X CUO
c/) [i] H x- '
C
o /-^
Ai -H J
a 4-1 c/i
CT3 ftj ^rj
4~> >
en a) g
1 x^/
w
4J
a CJD ^
TO -H g
4J GJ *— '
^ ,-N
|S
I--*-!
"-1 Is
§ ^
0)
rtj p ,—.
rH O M
3 -H OJ >•,
O CO 4-1 ^^
•H cn ra C
4J -H Di O
>-j e 4-j
ct! M v^
P-I
cu
P,
H
-a
c
tfl
"
cy M C
g CJ 0
cd f-i -H
^30,
CD PS -H
0 (-1
(-4 O 0
3 O W
O M OJ
00 3 Q
o
LO
o oo o ooo o
1 — ' r — -o^ co r^o^o CN
CO OJ-^f r— t , 1 , — IrH . — I
4J 4-1
c c
co cooo o oaicu u-\
CO 00 00 r-^ VJD -H -H O
in tn m in co o o - tn
II
co *o ^D in r~^ ^ r~-- r— i
co in in ^o ^ \o ^o p*-
CN O4 CN CN CNCNCN CN]
co co (J~\ o>o-\ r--
m in >n in mmm m
o oo o cooooo co
* * • • • . •
O OO c^ r-~^oO I-H
t^- oj --j " ai
0) -J,u cp_co(/i os-i
•r^ t~i i — (i — ti — 1 |4_1, — | O3CCCT3 .XIrH
JJ0^ oipapq cu-Hpq -Ht_),2^: pq
0- -r-l -H CO 4J 4-J
'4-1 bC H M c
OO CU O .— I QJCUO 3i— i vD r- OO
c^ MOO C>CN ,n^r-3- p
C-9
-------
-X
-X
v? -x
O CO ^-,
CO 3 g
4-1 -H --^
CO TD
a
•H C s~~.
S-i O -i< CJ
4J -H -X CJ
a) co a> co
g CO 4-1 ~~^
3 -H cfl co
M g Cd £
O W ^
CJ 4-1 CX A
cfl 'H E ^
4-1 X dJO
CO U H ^— -
C
O <~s
y *H | 1
O 4-1 CO
cfl cfl ^
4J ^*
co a) g
a
4-1
0 60 ^s
Cfl -H £
4-1 QJ ^^
CO prj
>•*
s 'e'
EH • — '
X
X— N
S P
r j ^_^
£j
rj
Cfl C -—.
rH O M
3 -H CJ >,
CJ C/] 4-1 *•-—
•H M cfl C
4-1 -H Pi O
(-1 £ 4-1
Cfl W ^ — ••
P-,
0)
P-
Jx.
H
TD
C
cfl
^
1) rJ C
E 0) o
Cfl ,Q -H
« £ 4-1
3 D,
01 ^ -H
CJ rJ
l-l CJ U
3 O W
O S-i QJ
CO 3 Q
O
CO
O O O O O v£>
O CM O CM CO CO
• » • • • 1 — 1
•O- CO <—t CM
i — i cr» co i — i ' — i o
• • • • • in
co in cr\ in ' — i
r^ ~d" O^ OO CO OO CM
CM CM r- < •—< I— I CO
CM CM CM CM CM CM
i • o^ o CT\ ^o in
O CM •— ( r— |
—I CM
O m o O o O
r~^ •— i O o O o
i — l OO ' — 1 i — I i — I CO
' — 1 t^- CT\ C^ CTi vO
o^ co i^- ^ r^^ in
^3" ^3° *3" ^3" *-d~ ^3"
O CO 00
PM co 1-1 4-1 CM i— 1 4-1 -H
O QJ T3 -r( 4^ )-l
3 3 CacOMtOd)
O CO OCO Cfl 34J4-J -H4J
Pi x-x >-i OH (-1 rH -H-rH i-J fi
01 rH rH COOCC -rl
CJCXpq ^pq \-t r- 1 fi, H-, ojco
C -rl CJ d C
CO rH -H 4-1 O CO
QJ rH r- 1 |5 O O'rHOrHCM OJO
3-H^O COr-- C4JOOOOOO 3O^
cr-C—i OJ •— i 3 cfl •—< -— H r-n c^t— i
J3Pn' — 1 ^ ^— t $-14-)' — 1, — 1, — 1 -j, — i
Q N— ' U pq oo Q
o o
r^ CM
• .
CO CTi
. ,
CT> i — 1
CM CM
O^ CO
-3" CO
co m
r-, f-^
CM CM
CM CM
CM CO
rH CO
^3" *"3"
m m
-* vD
, .
v£> O
O O
t-t
O
•H
4J
cfl
S-i
O
P.
O
O CO
^1
-------
-X
•K
^ -K
U CO x~>
ca 3 E
4-J -H ^—-
CO
os
u
rJ O -K 0
J-l -H -K 0)
cu en cu co
E en j-j \
a -H ro co
r-H i= os E
_p W -^
X . -K
u 4-> a. X-N
R -H E x;
4-J X OO
OO [d H ^-x
C
O /-^
O 4-J en
cd ca S
4-J £>
e/j cu E
4-J
o "to ^
n) -H g
4-J CU ^-^
C/3 K
^ x-s
S E
§-4 ^>
^
X
/• — \
S E
E— i ^^
£D
CU
CO C /-N
rH O M
3 -H CU >,
O W 4-) -~~.
•H CO CO (2
4-J -H PS O
V-4 £ 4-J
CO W -^
P-I
CU
a.
>>
H
T3
C
CO
•*
O) rJ C
E cu o
Cd ^3 -H
Z E 4-1
3 a,
CU Z -H
O (j
s-i cu o
3 O CO
O rJ 01
e/: 3 p
o
M
^D O O O O O O
CM CO LO r^ LO - O vO
• — * -—<
O to LO to O O CM
O> r-H — 1 r-J O CO CO
1 — ' cr> ON o~\ CT* o\ ^o
LO LO to LO CO O CN
vO CO (3D CO r-~ CO CO
^ »v r, r\ *. r, r,
o to LO LO o c oo oo oo oo oo r-v
LO LO LO LO LO tO LO
O OO Ol O~! rH ^O CT^
* * • • . . .
^ Q^ ^j- ^j- cO CM i i
O> CM ' — I ' — ' C CO i — 1 PL, CU CO
CU3CO^CO
-------
•X
^ *
cj en ,— *
4-1 -H -— '
co i3
CD
Pi
0
•H C i— s
^ o -x o
•I— ) *rH -*i Ql
01 co CD en
g en 4-1 \
D -H CD CO
0 a ^
v| • -!<
o 4-1 CL /-N
CD -H E tid
4-1 tt 0)O
CO W H ^
C
O '~>
^ -(H t-J
a 4-i co
•w >
co oi E
| •• s
-P
U fcO y-x
m, i c;
•M £n
4-1 CJ ^-~-
CO IE
p-l
s 'E'
t-H ^- '
3
X
— ' 'a'
& ^
' )
01
ij
. j /—) ",
r— i i_j tM
D -H a >,
u en 4-1 ---
•H en CD c
4-1 -rl pi O
^ E 4-1
CD [i] ^^
>^
01
>;
H
•a
c
cD
^
0) 1-1 C
E u o
(D r£l 'H
2 E 4-1
3 P-
01 Z -H
CJ (-1
t-i oi a
d o w
OS-JO)
00 3 Q
0 •
CO
OLO OOOOOO O
vO r-i oicocor-coro r-~
. — ^O LO OO ^— H LO CO O ' — I
CO ' — 1 • — 1 CO ' — 1 < — 1 ' — f
OACT\OAOAO^ r-~
CM 01 04 o-i 04 oi oj r-s) 04
O4O COOOOOCOLOLO -3-
OJ
OO OrHojr-iOCO O
COO LOLOLOLOLO^t O
^D r^ LO LO LO LO LO LO LO
COCO OJOJOIOJOJOJ O-
COCO CX3COOOCOCOCO O
H CJ]
O 0) 0) C
t-H4-l — ( CM 4J 4-1 4-1 T-t
CD Cfl =fti =S= Cd -H -rH 3
CJ 0) J-l S-l S-l QJ
•H ,G OCMoOQJCllO) V-i
E ^ SCPCrHrH pq
0) W OJ -rH -H -H -H 4-) 4J
-C )-i C. >-.r^^iOO)QJ ^c
C_J i — i p SH CD CD C rC .c^ TJ co
P3CO OpQfflrH;23 O >-<
U C 3
rH C 0 ^
rH OOI O O — < OJ CO <}" LO M O W
•H ^DO CJOCOOOCOCOCOOO 4-1 CTi 4J
> coco rH cocococoroco 4-1 co 4-J
Z U P-, PL,
O LO
CN •— I
I — 1
o o
r^
i — i
CO LO
CO 04
-d-
•o 01
OJ O4
04 CM
<± 0
VO r-l
O CN
CO OO
04 LO
01 CO
r-~ oo
sr ~3-
O r~^
LO r~-^
LO OJ
r-T
-------
4J -H
CO t>
O
•H C
o -x
•H -K
OJ 0) 01
CO 4-1
•H CO
E erf
o w
a 4J a,
cO -H 6
X
ai
O 60
cO -H
4-1 a)
co s
X
CU
4-J
O
-H
tj
•rl
4J
Cd
4J -H Pi O
ME 4-1
cfl W --^
Pu
a.
>,
H
T3
q
a
t-i C
0) O
,fl -H
QJ
CJ
a
-H
O !-) QJ
CO 3 Q
o
CO
ooooooooooo
^LOOOOOOOOOO
OOrooOOOOOOOO
(N CNJ r\l
O
OOCMu-iu-iOOOOOO
O3COOOO30OCOOOOOCOOOOO
oooo—1000000
ooooooooooo
=:fcOOOOOO
>
O
OO
fHr- IOOOOOOOOO
fflfQCJUCJUUUOUO
W
Cfl
QJ
•r-l
C
O
O
O
o
CM
O
O
o
CJ
C
•H
o
ca
•H
S-J
(1)
C
u
1—I
D
o
LO
01
I—
o
LO
LO
CM
CM
CM
00
CC
t*
-a-
o
o
CM
eq
o
00
C-13
-------
,-ti
u
d
4-1
co
CJ
•H C
M O
J 1 ... |
4-J M
QJ CO
6 w
3 -H
rH 6
0 W
>
r^
O 4-1
ro -H
4-1 X
CO W
^
CJ
d
4-J
CO
X
u
d
4-J
CO
>*
g
H
p
X
^^
£-t
H
p
01
4-J
co C
' — 1 O
3 -H
O CO
•H CO
4-J -H
)~< E
rH riJ
CiJ r1-!
PH
0)
ex
:>,
H
r.
CJ
,-H
vU
£2
0)
a
M
3
O
CO
.;;
-X
W ^.
3 G
•H ^-'
TJ
d
C4
,^^
-X 0
jf n i
"is UJ
a) co
4-J -\
d co
& e
^
• -X
PH/-N
6 &
GO
C
O '"••
•H rJ
4-J CO
d g
>
o e
rH •— '
W
4-1
.C
tO^-N
•H g
O ^-'
r^l
'e
i.
M
CJ >,
4J ~-»
d c
« o
4-1
T3
C
d
M C
CJ O
rP 'H
e w
D CX
rS -H
SH
CJ CJ
0 en
r-l 0)
a Q
0
CO
O O CO
r- CM ^H
co CO
\O ' — I i — I
CO
0 0 —1
CO CO
C C
0 0
•H -H
4-1 4J
d d
r-l r-l
CJ CU
P. P.
Of~ *
*• — i
OJD 60
C C
•H -H
4J 4J
r-l rH rH
rH QJ QJ
W g g
O O ro -3-
u cr> o> cr\
rP -3- 1 ' 1
O
CN
r^
r~~
i — i
CM
in
CO
CO CO
CO CO
-
m m
CO CM
•— 1 CM
o o
CO
C
O
•H
4-J
d
lH 00
QJ C
CX -H
O 4-)
CO
to d
C u
•H
4-1 C
rH O
OJ r-l
g W
CO O
CJ> C'
-3- m
1 — 1 1 — '
-a
C
d
CO
r>
TJ
rH
J3
g
to
C
•H
C
d
QJ
rH
U
CM
O
in
. — i
o
-3-
r-~
~d-
O
CM
CO
-D O
vo co
o o o
lH
QJ
0
CX,
d
-a
oo
c
•H
lH
fX
CO CO CO
1-1 1H rJ
pq
o oo
-------
•K
-X
_ s> "X
CJ CO ^
4-1 -H ^
CO TD
Cfl
Pd
o
•H q /-v
4-1 -H -X (U
o co a) cn
E cn 4-1 —
3 -H ca co
rH p pi 6
o w ^
rM! • -X
CJ 4-J CX x — s
CO -H 6 W
4-J X DO
CO Ld H v-^
C
O x-x
r*5 -H rJ
O 4-1 CO
cd cd ^
4-J ^>
CO GJ 6
w
4-J
0 W)x~s
cd -H £
4-1 QJ v^
oo pa
p-1
2u S
H ^— ^
p
X
— 'c'
r^. t-
r v ,.
| — i
CJ
4-J
cd C x-s
i~H O M
3 -H 01 >-,
o cn 4-1 -^
•H cn ca c
4-1 -H Pi O
we 4-1
OJ
>1
H
-a
rt
CJ M C
6 OJ 0
CO rQ -H
« e 4-1
D a.
OJ ^ -H
CJ M
1-1 ai a
3 o cn
O >-i D
CO 3 P
o
CO
LO o o
o to o
CO
rH rO
r,
^-O
i~ H co rO
vD CO CO
LO LO CO
vO O \O
CN| csl r^J
c\l oj c-4
*
CO r-H
.H vD ^D
(Xl V0 ^D
•iH =ffc -
CN
CN
co
<^-
ON
i — 1
r^
r^.
ON
o-
~*
CM
i — i
•-*
LO
O
ON
1 — 1
CN
O
CN
=s=
Oi
c
•H
r-l
0)
4-1
C!
CO
OO
o
T=
PH
PQ
-d-
O
"^
O
CJ\
^o
1 — 1
o
co
^0
LO
^o
CM
CM
LO
LO
^
CO
r^
r--
ON
^D
CO
CO
VD
CM
CM
VO
VD
ON
vD
r^
ON
O
a)
^
o
o
ON
0
^
CM
r-
O
O
LO
vD
CM
CM
0
, — 1
o
LO
r-
r-
ON
^
r-l
Q
0
O
N
o*
^
o
r-
_j~j
CO
•H
rH
O
-H
O
~"T
1 — 1 , — I
CO CT>
-------
S>
O
cO
4-J
CO
a
•H q
4-1 'H
CJ CO
e w
D -H
rH E
O PJ
y
U 4-1
CO -H
4-1 X
co pa
u
cO
4-1
CO
a
CO
4-1
CO
>H
g
§
k_>
--N
^ ,
6
1 1
QJ
CO C
rH O
3 *H
U CO
•H CO
4-1 -H
!H B
co pa
P-,
QJ
ex
^1
H
»v
QJ
g
CO
2
QJ
O
M
3
O
CO
*
CO /g.
•H *—s
T3
CO
PS
'— >
•£ QJ
a) w
4-1 ^~~
CO oo
PH B
'~-/
• -X
PL, /"""N
B X
QJO
H ^
C
0 s~*
4J CO
CO IS
£>
QJ E
W
00 /-N
•H B
OJ ^
33
E
^,
!H
4-1
CO c
pi o
4J
s — •
T3
(3
CO
SH C
QJ O
ft -H
E J~>
3 CX
SH
CD CJ
o in
r-i a)
o n
0
CO
(• —
o
rH
CM
CO
^j-
CO
co
LO
to
CO
0-1
o
CO
o
^Q
o
oo
ON
-3-
O
vD
T-_-
QJ
4J Cfl
CO r-l
rH
QJ M
rH
i n
•H r-H
a
1
LO
CO
CM
CM
i — (
ry\
WJ
LO
r^
cr»
• — 1
LO
CNJ
LO
~^-
LO
vO
(^
r-H
O
iO
-3-
i — i
1
CM
ON
C
O 'H
TD
0) C
N^ *H
O lH
U O
tO O
LO sO
0 0
~3" <3"
o
, — 1
sD
OO
CM
4-1
C
QJ
•H
rD
4
IO
CO
CN
i — i
OO
o
ON
, — 1
OO
ON
0
ON!
ro
HJ
LO
^3-
LO
o
o
1 — 1
1 — 1
o
o
r~~
i — i
-TT=
>1
rrj
3
O
P-l
U
QJ
rH
PJ
rs]
sD
O
-3-
O
tO
CN
O
i-H
, — ,
, — 1
OO
to
oo
CM
CN
O
cr\
i-H
00
as
0
C^J
n~\
<~^J
LO
^3"
to
sD
1^,
OO
O
r— 1
I — -
1 1
=fe
CO CO
Cl rH
•H CO
Cfl QJ
•H C
C -H
•H ^_,
QJ
O
OO >->
vO
o
1 ~rt
^»T=
CO CJ QJ
* r-l 4-1
-3- PH C
CO -H
~^= r-'N CO
SH
CO CO O
SH 4-J C
rH O -H
PQ PS M
• — i CN CO
p-^ p-^ p-^
O O O
~-3" <3" -
0 0
-------
•X
r-^ '1C
U CO ,—,
CT) 3 E
4-) -H -— '
CO T3
cd
o
•H C ^-v
!-i O -K O
4J -H -j,' cy
01 co QJ cn
e co 4-1 -^
3 -H ca co
u ^
rij- • -X
CJ 4-1 CX ,-N
cd -H e $6
4-> X OJO
co U H v— '
c
O ,-N
4->
g 7
H v^
3
X
„_ o
H -^
£ZJ
0)
a c ,-N.
i— 1 O !-J
CJ CO 4-1 .
•H CO CO 0
4-j -H ee o
u e 4-1
cd W v^
0)
a.
H1
-a
c
cO
"
CJ L-I C
g U 0
^ E 4-1
D a.
cj ;s -r-i
U l-l
t-j a) a
DOW
O >-< 0)
CO D Q
o
CO
^O ^3 C> • — 1 O 1 — CN vi
~3" r^^ co cr\ o CN r~
• • • • • •
p — t • — i
r~-Or^cNO —id r-
. . . .0 . .
<3" * -U
£ a
0)r-i(UcNO r^r^- u-
•H i — t *H LO o r^» r^- f^-
PjQrOrjOrO'-H -xf<3~ LT
<£, <$
r— ^ r— ( i— J i-^ r— i Q^ Q^\ ^
•v^" -j" -vj* v j" •N^~ f**^ cx] r~*
OJCMCMCMCN CN^Nj O
C\l LO O*\ ' — 1 CO i — 1 LO i—
CO CN < — i CO ^O LO 0s
OOOO^H OO LT
t— I I-H ,—H i — ir--- CNCN u"
CN OJ CN CN OO \vD ^D C^
CNJCNCNCNCN OO if
ooooo oo ^
LO LO LO LO uO LO LO if
^• 'vD VD VO VO 'vD CN
^OCN^tf-H. — 1 i— 1 , — 1 LT
r-H LO t—
O O O O •— t OO C
r,
o^
hvO
1—1 \^O
1 r~<
r- -H w
vD CTN -TI- d)
•— 1 r^ r~( r— | r^ Qr—
rt rjojJ^c/iOQJ COLO o c*"1
^ 03 r*^ GJ co 4-J C*0 t/j =?*~ ro j .— ^ —-•_
C'~- -coat-i ojcn ^23cr
DC 'o ^ ^ a ^
O -H PL, o CO
•— 14-J O ^CNCO-a- r-H CNJ OrH r-
d oo cooocooo o cj^a% o 1-1 c
• O O OOOO O OO 42CO r-
4-icj) vD
j CM CN c^
1 LO LO CO LO -i Po c
w w p-i a) 4-j M
00 -H
CN CO -H O — I
O O >-i .C -H
1 — ' < — i n 4_j . — i
^T
-------
X
cj
cO
4-J
OO
u
•H C
SH O
-U -H
eu cn
G co
3 -H
T-H E
jD W
v>
^
CJ 4-1
4-J X
00 W
i v!
U
cfl
4-1
00
O
CO
4-J
00
>>
•£
H
X
s
H
01
CO C
rH 0
3 -iH
CJ CO
•H CO
4-1 -H
l-i E
cO W
0)
ex
H
0)
E
CO
2^
Ol
u
S-i
;3
O
CO
-X
'en s- -.
•3 e
•H ^
T3
CO
f — ^
* CJ
-X 0)
o) en
4J \
cO co
[V^ g
•*-'
• -X
Q-Q>
oio
H ^—
O '—-
•H rJ
4J CO
cO ^C
>
oi E
i — l **— '
4-1
OC x~ >
•H E
0) ^
K
'E
g
^x
^
S-i
eu ^
4J ~\
CO C
K! o
4-1
^
T)
C
^ c
0) O
o *H
E 4-1
3 CX
Z *H
S-i
0) O
cj cn
S-i 0)
3 P
o
in
T-^i
SH
O
u
Ml
C
O
S_i
4-1
CO
E
<^
O O O LO O
v^ H
P-i
iO
LO
^
i — i
CO
cn
Ol
a
o
S-i
PH
^O
LO
^
i-H
CO
cn
01
CJ
o
M
P-i
r~-
LO
^.j-
OO
CO
CNCOLOCO
C
O^O^roooOcoo'") >o ^o
1 — 1 • — 1
4-1 C C
o 5 o
03
pj
H
•^
CJN
f- —
! 1
^
C-18
-------
^ -K
o cn ^^
03 D g
4J -H ^-s
en 13
03
erf
cj
.1 —i ^_^
l-i O -X CJ
4-1 -rl -K 0)
6 cn 4-J --~
D -^ 03 co
rH E Crf E
0 W ^--
_y • -!'
O 4-J O, <— v
03 -H E v^
4-J X CUO
en W H • — •
C
O ^N
^ -rl rJ
a 4-J en
03 o) J2
4-1 >
en 0) g
W
^S^
4J 0) v_x
K"
§ J,
t— i ^— '
X
IS G
H -^
| i
01
4-J
CO a ^ — \
rH O V-l
CJ CO 4J --J
•H W 03 G
4-J -H Crf O
l-i6 J->
03 W *— •
01
ex
•»
H
T3
C
cfl
QJ (j f
'-y r-* l—t
6 01 o
CtJ rQ *H
ss 6 4J
D a
0) Z -rl
O VJ
!-i 01 0
3 a en
o 1-1 CD
en 3 Q
o
en
v£> ^D O O
~j-
-------
1
O CO /-^
co a B
_^J ,p_j y_x
CO TJ
(0
Pi
a
•rl C .--*
JH O -V. U
4J -H -X CU
a) co cu co
g CO 4J -^
3 -ri m cn
rH 6 Pi g
o w *-^
*^ • -X
a 4-1 px ,-s
cd -H e jsi
•w x aio
CO W H "-x
C
O x— .
^ -rH i-J
O 4-J CO
CO CO g
-J-I P>
cn cu 6
W
1j Oo^
CO -H E
4-J CU ^-^
co ;E
>H
S '9'
E~H •• — '
( — J
X
X — v
a e
H — '
D
cu
4-J
ec C >-s
i— 1 O J-i
3 -H 0) Ps
O W 4-1 \
•H CO CO C
4-J -i-i pi O
JH £ 4J
m w v-x
PH
CU
a
^
H
13
C
CO
r.
CU JH C
E CU 0
CO n *rl
2: E 4-J
3 P.
CD g -H
O JH
JH CU O
D U M
O JH CU
CO 3 Q
O
CO
cn cn o vo • — i O OvOOino
oo r~^ r~^ cn O vD en ' — i ^0 O CN O *"O O
LnCn. CNCN CN r-HrHrHrH
CN 1—
r~ — r^- in oo oo *~H CT> Q°\ cn (^\ ^\
VO^O'-O r— 1 i — 1 LTl OOOOO
oom cncn en enenenenen
ooooco enen m ^H — i rH rH rH
CTi C^ Cn r — I ' — ) O> -3" ^3" *3" ^3" ~J"
cni — i r, r* r, * r,
cnc^c^ ^o\£> in ooooo
-J CN rH
C rH CN
D-. /~N »v *
E CO rH O
O rH r, rH
CJ JH O> JH i — i JH I-H
S CU XVn X QJOOcJrHrHCJrH
O rH OCU O CCnOO— lCn=ferHCO^:
Pt|x— , -H U > — I CN O •— I O O=!fe -H
T3 PL, rHcoenen rH en 4-1 •< <—t '-psS"Ps
CO 1 1 'H CU ^^ ^ 'H =^ CO -ClrCXlCU rC^Cl)
•HCU cotOr-f 3PQ 3 cu OTrHww>rHcn>
C-H Jninrd JHJH ^v J-i e DO-)DCrH3C
efl<4H rHi-HO 1-3^ rH 1— 1 t^JCU rH -H jHi-HjnjHOCOJ-IO
>co pqpQU JH pqpq M W rJ CJ^UCJOpqCJO
' — 1 C .^- CO ,y ^
X CO U rH O -H C
co" r-ncNen OP rnesi OJ-i rH o -H c\ien E m JH m inm inm
CU PQ -vf ^" ^5" ed E-H -
-------
o
*rH
S-i
-LJ
OJ
E
3
r-i
O
^x>
a
eg
4-1
CO
-<
|
X
^
^)
C
o
'w
w
•H
E
W
OJ
ex
>x,
H
r.
1)
P
5
|2j
01
o
M
3
O
CO
^f
•K
0) ^-x.
•r-l xE
TJ
eg
pi
-X 'ej
"K CLJ
CU CO
4-1 --x.
eg co
pi ^
• -K
CX '-N
S ^
OJO
H ^
C
O '-x
•H r-4
4-1 CO
eg s
£>
-i C
CD o
rf~i 'H
E 4J
3 ex
Z -H
r-l
CU U
O 01
Jj CU
3 a
o
CO
o o o m
CO CO r-l
CO O CN] i — i
• O • •
4-1
rj
o^ o oo cu
CO O CO -H
co ,-< co rQ
|
^j- , — i , — i , — i
x^ in m in
CN CO CO CO
• — i o in co
CN •— I
a-. r- r- r-.
CN CN CN CN
CM OO 00 CO
r-x, co co co
O co co co
O O
V^D -H r^ a>
ro LO •^ i — 1
p>-> r^ -^
QJ CU O !-i
C > be --I cu cn
o c eg !>-, ^_t
1 j o CQ ^ ^J CU
>-J 01 O -^r Q N
r-J CU - O -H
T3 eg E » c --i >, >-i
C M -H rC a) I f-i a)
3 O »-J 03 CU ' — ' eg £>
O CX 3 M O 4J r-{
P1-* 3 CU r-i CJ ' — 1 O 3
O rH C_) CO =fe A P-l
4-1 4-1
u w
cu t — i eg . — i CN co
s-i co cj vO 1 —
CN CTN
OOO
OOO
vD vD vD
CO CO CO
OOO
CM CN CM
m m m
r- r-- r-
m in in
m m m
OOO
•—i co O
CM
OOO
m
CM
r.
f-^ fj^ , — (
1— i r— 1 CN
• — 1 i — 1 I — 1
=J£- r£j~ =^-
E E E
CU C1J CU
4-14-14-1
01 t/1 OD
r**i r^ r^i r*~>
r-l C/l CO CO
-o
C *"U T3 ID
3 e s £
Pq co cn co
X
QJ • — ' CM CO
4-i in in in
C in m m
cu ^d~ --d" -4
4J
(^
QJ
•H
n
•f
CO
O
CO
CM
i — i
O
O
xD
CO
o
CM
in
r^
LO
LO
t_O
^o
vD
CO
O
CN
m
r^.
in
LO
0
OO
o
CM
O
1 — I
=££;
to
C
0)
Q
0£
£
•H
,M
cd
CQ
•vO
LO
10
*^O t — I LO ^O
4J 4-J 4-J 4-1
CJ CH C d
QJ CU CU CU
•H vH -H -H
^Q O rO O
E J E E
00 OO OO 00
0 O 0 O
co co co co
i— i r^ vO cr^
• — 1 ^H
0 O 0 0
o o o o
xD vD vD \O
CO CO co CO
O O O O
CN CM CN CN
m m in m
r-^ r- r^ r^
in in in m
LO LO LO LO
co ^ D
m c: c c cd
T— 1 *r— t "rH *rH C4
O C C C r~1
4-J -H -H -H CJ
O V-l V-l r-4
Di CJ U O O
Oi]
£
OO CTl O r-H CTj
in in xD xD cj
in in in in QJ
^j~ ^ <]-
-------
-X
"a 'to ^~>
CO 3 g
4-1 -H ^
CO T3
Pi
O
•H C ^
SH O -i< CJ
4-1 -H -K CU
CU CO 0) CO
g CO 4-1 \
3 -H CO C-0
rH g Pi g
O W ^
^ . -x
a 4-i ex ^--N
rt -H g ^
4-J X CUO
CO pj H ^~"
G
o ^-N
rM -H |J
O 4^ CO
CO CO S
4-1 >
CO O! g
W
4— '
O 00 s-,
CO -H g
4-J CU ^^
CO pU
^-\
S g
H ^
cu
CO C '~-
.— 1 O S-l
3 -H QJ >•>
O CO 4-1 \
•H CO cfl C
4J -H Pi O
S-l g 4-1
CO W v^
PJ
01
a,
H*
-a
co
01 M C
g oj o
CO ,£3 -H
z; g 4-i
3 0,
01 3 -H
O SH
M 0) 0
3 O to
O S-l 0)
CO 3 Q
O
CO
CO
C
•H
t-C ^ — \
CJ -O
cu
0 3
M (±
C -H
CO 4-J
C C
CU O
rC 0
CO v_x
O
O
.—< CN CN
• • •
CN CN CN
CNCNCNCMCNcNCNncNt— ICNCNCN
OOOOOOOc-OOr~-OOO
CMtN
OO
CNtNCNCNIOJCNCNCNCslCNCNCNCN
OOOOOOOOOOOOO OO
o CN
o CM
o
CM
CNI o
CN CN
O
OOOOOOOOOOOOO
CN
oo
o o
i — i
o
i — 1 o~)
cn =s=
cu
to cn
to co
cu cu
a u
o o
PH O-i
<-
=fe
to
to
0)
o
o
S-i
PH
m
=s=
CO
to
0)
a
o
S-i
PH
VO
=fe
to
to
cu
o
o
S-i
PU
r-^ oo
^t ^
co to
to to
0) 0)
a o
0 0
S-i S-i
PH PH
i — i
CN
* — I
cn
, — i
=ft-
cn
to
cu
a
o
P-,
0
CN CN
, — 1
to ro
to =s=
cu
a to
O SJ
S-l r-l
PH rn
0
=s=
to
to
cu
u
o
S-l
PH
O
=S=
to
to
CU
CJ
o
S-i
PH
ro
O
=S=
to
CO
0)
O
O
S-i
PH
CN
O
i — i
.-H
O
=te
to
to
cu
o
o
•H
_c
, — 1
O
, — i
^TT=
C
CU
£>
o
0)
y
CO
p~]
i — i
cn
in
•
o
0)
^
CO
m
CN
cn
in
•^
m
O
, — i
=^
^H
01
TD
C
•H
S-i
O
O
^
C
cfl
H
00
C
•H
r|
a
4_1
CO
pa
co
cn
in
^
\£)
0
, — I
=fe
^
0)
T3
c
0)
T— 1
P3
^j-
cn
in
^
-------
u
o
4-1
CO
o
•H C
S-i O
4-1 -H
0) en
E CO
3 -H
H E
0 W
^
a 4-1
CO -H
4-1 X
CO fd
^
CJ
cfl
CO
^
CJ
cfl
4J
CO
>•
1
X
s
H
0)
4—1
cO C
CO co
^"^
. -K
p ^-^
E ^fj
0)O
H ^
C
O /— s
•H i "1
4J CO
cfl S
>
QJ E
i— H ^— ^
W
4-1
x:
OC ,-V
•H E
OJ -^
3
'E
Is
^
&_<
0) >~,
4-1 ->.
ci3 C
Pi O
4-1
^^
-d
d
cfl
J_, (-
QJ O
^Q -H
E i *
3 ex
Z -H
M
CJ O
o en
i-i 0)
3 Q
O
CO
0
o
o
1 — 1
o
0
1 — 1
^J-
LO
ro
o
1 — 1
CN
00
-3-
VO
°^,
o^
*^o
o
o
LO
CO
CN
o
1
1 — I
o
0) W
C QJ
O c/]
4-J CO
W 0)
ai cj
E 0
•H !-i
t-j eM
rjj
CO r-l
CJ O
aj r-.
O r^- co
^ ^ *\
CTv LO CN
cr» cj\
CO
ro
LO O
ro o
a-* o o o
. . . o
r^- 10 co i—i
CN O 00
LO r—I 1—I O
CN •—I r—I O
CO CO CO r-H
oooo
CN CN CN CN
oooo
r—i CO CO ro
CN CN CN CN
LO LO LO LO
w\ r* r\ r.
OOOO
CN CN CN CN
LO LO LO
LO LO LO
CN
co
a->
o
O O O
O CO CO CO
O*\ *^D r-H *vD
CO r—I
ro O
en
en o
•—t r—
CM
r~-
<•
CO CO CO OO
co co co ro
CN CN CM CM
vO
co co co ro
LO LO LO LO
LO LO LO LO
oooo
ro co co CO
LO LO LO LO
LO LO LO LO
oooo
CO O O O
10 ro ro co
CM r^ -d- oo
oooo
M
rj;
•H
j_i
00 60 60 QJ
Cfl
rH
CO
OJ
c
CO
0)
3
CT
3
Q
C C C QJ
•H -H -H C
60 00 60 -H
60 60 60 60
cO cd cfl £
rH lH H W
CO CO CO
4-J
4J
rH CM ro 0)
r^* i ~ r^-* ^ s^
*O v£) '-O cj
•-3" ^1" •<]" QJ
rXl
CM CM CN
r— 1 r— 1 , 1 ^
-::= ' 1 a
X
60 O
a o
•H rH
60 -H
CO S
Cfl
rH KJ
CO
^
O
-H O
O u
r^ , o
^3" Cfl
PO
co 10 en co
60 CM CN QJ CX
C r-l r-l > E
•H =S= =fc -H O
LH 4-i cj
s-i w en -H
cfl fin £n 60 S-i
-------
^
ej
CO
4-1
a
•H C
l-t O
4-1 -H
cu en
E en
P -H
,-1 E
O PJ
A;
O 4-1
CO -H
4-1 X
C/3 W
y
O
CO
jj
C/3
Ai
O
CO
4-1
C/5
3
X
g;
H
cu
j_j
co e
iH O
D -H
cj en
•rH en
4-1 -H
M E
CO W
CU
a.
H
QJ
E
CO
K
QJ
U
l-i
P
O
C/}
-K
•X
en s-*.
P E
•H v — '
cO
s-*^
•K O
•K cu
cu en
4-1 "^-~
CO CO
PH E
^°~*'
„
O , ,-— ^
E W
cuo
H "-^
d
O X-N
•H nJ
4-J C/)
co 2
cu E
) ^_/
w
4-1
Ofi^-N
•H E
QJ ^-^
33
3
^6
^
1-4
CU >1
4J \
CO C
K 0
4-1
'— '
C
CO
l-i G
QJ O
rd -H
E 4J
P a.
IZ -H
^-i
CU O
ej en
J-4 CU
P O
O
e/3
j-i
C
QJ
g
cu
U
cO
en
P
T3
QJ
S
O
O")
LO o
- — I CO
LO o o
•—t CO CO
O O O CO rn o
LO LO LO CO CO LO
ro ro n ro ro ro
LO LO LO LO LO LO
CN CV| CN CN CN CN
00
o o o o o o
CM CM CM CN CM CM
C
CU
LOLOLOLOLOLOLOLOLOLOLOLOLOLOLOLOLOLOLO
CXICNCNCNCNCNCNCNCNCNCNCNCNCNCNCNCNCNCN
O^ <-< CM O 00
CN O^i •—I -
i — i
ft ssaoo
(— H
ft ssaoo
CO
1 — 1
ft ssaoo
, — i
en
en
01
o
o
o
CM
ft ssaoo
i — i
CM
ft ssaoo
CM
CM
// ssaoo
CO
CM
ft ssaoo
CM
, — l
1
-------
a o) /--,
to 3 B
4-J -H --^
CO T3
cfl
Pi
U
C >—<•
l-l O -X U
•H -K CU
CU 0) CU CO
0) 4-1 \
3 -rl co co
e pi e
o w ^—
a 4-j ex s-
co -H 6 ^
X QJO
co pa H --
o ^
-^ -H t_)
CJ 4-1 CO
co crj ^
4J >
in cu E
j-i
u oo/->
CO -H g
4-1 a) v—'
co pc
£
X
CU
cu
4-1
CO
3 -H
CJ W
•H 03 CO C
4-J T-I Pi O
M g 4-1
CO W v--
PH
cu
a.
c
CO
CU l-i C
E CD O
CO J3 T-I
Z 6 4J
3 O-
0)
a
cu u
o en
a
3
o
LO 3 Q
o
CO
o o
on on
o o o o
on
00
OOOi—(Oinininm
COOOCOvOOO'-H'—I i—I i—I
00
s±
Ambient
CN
CO
on
o
on
CO
m
CO
^
o
r-H
on
in
on
on
on
Ambient
m
on
CO
CN
O
on
CO
m
on
m
-*
o
CT<
CN
m
on
i — i
00
rC
CO
on
m
m
on
CM
m
0
m
CO
m
in
-d-
o
oo
CM
in
O^
^sO
vQ
cM-HCOcnocO'-icoocMcMCOLncOr-Hon^ocMcjn
OO i-H O CM •— I CM
4-14-14-1 4-14-1 4-14-1
(^ {H £H C CH CH C
CM r~^ on in CM cu CD CD cr» o** o^ o^ CD cu • — i • — i ' — i cu cu
CM r~~ CO O £N -rl T-l -H CO CO CO CO -H -H •— 1 i— 1 i— l -H -rl
coononcococococococococococococococoro
CM ^o CM o co CM CM CM CN CM
-------
o
•H
}_i
4-J
QJ
6
;3
i— 1
0
_^
O
cd
4-1
CO
QJ
Cfl
t— j
3
a
•H
4-1
S-i
cfl
PH
Tk^
U
03
4-1
CO
C
O
•rl
to
cfl
•H
g
J_J
•H
X
w
^
CJ
cfl
4J
CO
O
cfl
4-J
CO
>->
£
p
X!
s
£— i
1 i
c
O
•H
CO
Cfl
•H
E
w
OJ
a.
p>-,
H
*•
QJ
e
cfl
z
Q)
O
}-i
D
O
CO
•X
•K
-X
Cfl x~x
3 e
•H >—
T3
Cfl
^-^
-x a
-:< QJ
QJ CO
4-1 ^
cfl ro
pi e
• -X
CX x-N
QJO
H v^
C
O '~"N
•H nJ
4-1 CO
cfl g
OJ 0
W
4J
"Sc ^
•H g
0) s-'
X
3
,, — s
6
• — •
^
j_j
QJ >i
4->
03 C
Pi o
4-1
*• — '
-a
03
>-i C
QJ O
rQ -H
E 4-1
3 P.
Z -H
>-l
QJ O
CJ TO
M QJ
3 Q
O
OO
o o
v^-
.
o oo
o
1 — 1
o o
o in
i — I St
0^1 in
CM CO
CO CO
o in
(-H
CM CO
CM v£>
CO CO
m in
O sf
r-t O
^ r,
st OO
m m
m m
st st
• •
O 00
^o r^*
CN
01 O
vO "
- O OJ
C --i O
QJ 1 —I
Q) -H =fe
!-< O
CJ •— 1 CO
CO =ft= QJ
Cfl vD
0) " ?~, Cfl O
C ,£ QJ ^ 0) •— I
o to > a o -
W 3 C vH 0 m
co ^j o ^i MO
u o pq p^ ^
4J
p 'H
O •— < 4J r-H
D- OO 03 O
CO C -H
cfl in QJ m
£> f^-i
^j-
o
1 — 1
~ii=
Cfl
QJ
cn
en
QJ
u
O
P-4
CN
O
T-H
in
vn
^j-
.
00
,
^o
st
0
^,^J-
in
CO
CO
vO
0^
^
CO
in
CM
o
».
00
in
in
CM
.
^j-
i — I
CM
O
,— (
1 — 1
1 — i
1
r^
O
T— H
O O
O -H
p*^
M
TU
p;
^j
O
^
0)
r-t
4-1
Cfl
Cfl
U
£j
Q)
o
Cfl ^H
C ^
QJ
> CO
O cfl
r-H
Q) O
y fX
O 3
CJ U
^ CM
vD ^O
r-H I — 1
m m
st
r~^.
CM
co
OO
o
in
st
CM
CO
st
CO
OO
CM
cfl
O
O
o
4-1
1-1
0)
4-J
C
QJ
P.
t-i
cfl
O
CO
C
OJ
O
o
o
0)
0)
m
C-26
-------
4C
•K
^ J<
CJ to ,--,
CO 3 E
4-J -H ^
CO tl
CO
U
•H d ^— N
M O -X CJ
4-J -H -X CD
01 M 01 CO
E CO 4J \
3 -H CO CO
"H g Pi E
,? W ^
^ . *
CJ 4-J ft /-— s
CO -H E S4
4-1 X 01 0
W Ld H ^
c
O '->
CJ 4J CO
CO CO S
4-J ^-
co CD E
4-1
cj "ob /-s
cO -H E
4J CD *• — '
CO X
>1
g Is
p
X
-^
P o
p
0)
CO d /-s
rH O r-l
3 -H CD >,
CJ W 4-1 — —
•H CO CO C
4-1 -H PC O
>-l {2 4-1
co W ^
CD
O.
>^
H
T3
d
CO
^
CD r-l d
e CD o
cd ^o *H
,Z £ 4-J
3 ft
CD 23 -H
CJ M
r-l 0) O
3 O CO
O r-l 0)
CO 3 P
o
CO
o o oooo m ooo o
CO **Q 1 • • — 1 <3" ~3" CN
• • • • • • •
CN r-H
CO —1 CN <—* CM CN|
ri
1 — ^
oo oo o in **o ^f in
-d" ' — ( C7>
• • •
>D -a- o vo
• • • •
:}•> CN O LO
-H m CN
4J
c
•o CD O m
CM o Loin- — 10s! o co'Hin o
-^ -D
n xi LO
^
^H i— 1 r— 1 CM
LO ' — 1 ^O vO vsD ^D ^O CQ CO 00 <~~l
CM CO CO CO CO CO CO
CO co CO
o^ in co co co co ^d
^O vD ^«O CO
ip vO ^O ^o
CJN - — i ^t" CT\ C}
'^ r^ r~^- LO
co o oooo d" ^vl" <3~ ^3" ^3" ''d" *^1~
CTi •— 1 CO CO CO CO VO
^d" ^Cf ~d" ^f
CM CN CN LO
r- in Lnminin r~- CNCMCM oo
co ^^ r^^ p"*- r~^> r^- r~^>
m in LO CT>
CNJ <^~ o^ cr* o> o> CD oo oo co r"--
\£) ^-O LO LO LT^ LO '^O
co o *^ -lr-l CrH -rl
CW Dffl O CU CM CL, PH MpQ f_J
o d X o
J-l O CO ,-O r-l
4J- — 1 4J' — 1 CO- — ICNCO^t CD' — 1 CD
cncN coco -, O WOOOO do r-i
^-l v^D CD ^-D CO vD **O ^O ^) O *^O CD
o en
>^ r-l C Pi r-l
r-l CJ O rH
Q co cj £>-, pq
^
0)
rH CM CO ft 1— 1
-------
u
•H
VJ
4J
QJ
e
--j
rH
O
*t
CJ
d
4-J
to
0
4-J
rH
3
U
•H
4J
Vj
(S
^
u
d
4-J
CO
C
o
•H
en
en
•H
E
W
4J
•H
X
W
v;
U
d
4-J
CO
O
d
4-J
K_
g;
^
X
5-;
H
3
C
O
•H
W
en
•H
S
W
QJ
>N
H
QJ
£
d
2;
QJ
CJ
VJ
3
O
in
•K
-X
j*
en ^-s
3 E
•rH ^-^
T3
d
OS
^-^
-X CJ
J.< QJ
QJ en
4-> -^
d co
^ 3
• -K
p, ,^-s
E N^
QJO
H ^
C
O X-N
•H HH
4-1 tO
d ^
>
QJ e
W
j_j
•H 1=
QJ *~s
£
*G
^^
VJ
QJ >*
4-J -~-,
d a
oS o
4J
'
C
d
M d
ey o
,£j -H
E 4-J
3 a
rS 'H
VJ
QJ U
o en
VJ OJ
3 a
o
w
0 0
r-H r~~
. t
£>
CO \£>
CO CO
LO 0-
o CO
vO ON
CM CO
*^ rv
^0 CO
0 0
CM CM
CO CO
" Vj *v
VJ r-H CIJ r-H
QJ CO S CO
[5 ^rii O ::>i-
O PH
PH to en
VJ >N SH
j: rH 4-1 rH
6C pq -H PQ
3 U
d
E r-H VJ r-H
OJ CO QJ ON
CO E O
O ^O O v£>
O O O vD
CO OO O
•H
rf^
OJ
QJ
pa
i — i
r-H
vD
en
C
QJ
O
QJ
V?
O
U
QJ
>
•H
f
QJ
QJ
pa
CN
vO
CO
o
r-H
1
. — 1
o
i-H
^T=
en
rH
rH
O
PS
rH
d
C
o
4-J
d
en
QJ
o
o
VJ
PH
i — i
CM
v£>
^3-
O
rH
c^-
en
en
QJ
a
o
VJ
PH
CM
CN
CD
LO
o
i — i
~rfc
CO
en
QJ
U
O
Vj
PH
CO
CN
3
O28
-------
•K
•K
^ -K
O 0) x—*
CO 3 B
4-1 'H •—'
CO T3
cd
o
•H C
S-i O -X
4J -H -X
O> 01 QJ
E en
3 -H
•H E
O W
ro
CJ 4-1 p. ^-
cd *H £ ^
4-J X Q)O
co W H >-
C
O '•^
^ -H rJ
a 4J co
co ro a
4-J >
co o) E
w
CJ CUO r—N
a) -H e
4-J
CO
01
X
H
QJ
4-1
cd C
rH O
3 -H
CJ 03
•H 03 Cd
4-J -H Pi
r-< 6
co a
PH
QJ
ex
cu >,
c
cd
cu in c
E 01 o
CO ,0 -H
3
ex
•H
CJ O
3 CJ 03
O S-l QJ
CO
3 O
O
CO
O
CO
CN
O
o
CN
r>
LO
o
o
CN
O-
QJ
C
QJ
-C
U
CO
ro
I
o o o o o o o
•—I CN O O O CN CO
o
CT"! O CD **O <3"
r-n o o o --D CO CN CO O l^~ CO
-3- CT\ O> CT> CT»
LO LO- LO LO LO LO LO
O
on
co
-3-
CO
CTi
f,
-^ °
QJ
fl^ ^
O O
pq U
01 ^O ^-O
^j.
o
^
T~
0)
C
0)
o
01
^
o
u
LO
•
-------
-X
•X
kyj _V
o tn ^-\
CO 3 6
4-1 -H v_x
co ID
rS
a
•H [2 <-^
»i o -x a
4-J -H -X 0)
ai to o) cn
E tn 4-J \
3 -H cfl co
<-t e & e
0 W ^
^ . -x
O 4-J CX-^
cfl -H B t4
4-1 X 0)O
C
0 ^
AS -H ,_)
cj 4-J cn
CO CO ^
4-> >
LO QJ 0
1 i ,
f^ N— ^
W
CJ 60 /~\
CO -rl g
4J QJ ^— '
CO [J-4
X
s "E
^
X
s 9"
H -^
^ >
OJ
CO C x~.
r-t O S-l
3 -H QJ >-,
CJ LO 4-* — -.
•H cn co c
4-J 'H Pi O
^ e 4J
to pJ ^
PH
01
O.
^x,
H
-a
c
CO
r.
0) M C
6 OJ O
CO ,Q -H
z e 4J
3 CX
QJ 13 -H
O M
^ OJ u
3 CJ W
O t-i 0)
W 3 Q
O
CO CO CO CO
O^ CTv CD^ O>
• • • •
i — l i — 1 i — l i — l
O"\ O> OA
00 OO CO OO
r^* r^- r^^
^c, „ ^ ~^,
f*>- r^« r^** p"^-
C^ O^> C^ CT>
LO LO LO LO
o o o o
r-4 ,-H r—^ i 1
'— < ^— ) r— H |-H
O O O O
cn tn w w
tn tn cn tn
01 0) OJ OJ
CJ U O O
o o o o
^-1 ^4 J-4 i^-J
PH PH P-l PJ
CO Cfl CO CO
O -H -H -H -H
M 0000
I I 6 B
co <) <2 <3 o
CO CO OO
o
O
LO
o
o
LO
CN CN
CO CO
CO CO
CN CN
O O
vO **O
CN
CO
O
CN CN
CO CO
CO CO
CN CN CM
CN CN Cs|
CO CO CO
i—I <—l i—t
CN CN CN
000
OOO
" »v •*
CN CN CM
OOO
LO LO LO
r- O
O O
CO CO CO
OOO
0 0 O
CO
CO
O O O
i—I '—I CO
CN CN CN
CO CN ON
CN r~- O\
O\ '—I ^£>
i-l CN CM
O O vT>
CO CO O
CN CN CN
CN CM CN
CO CO CT\
CO CO 00
O O O
000
CM CN CN
LO LO
o o o
ooo
o o o
CM CM CM
O1* CTv (Ti
to LO LO
CO CM O-
o r- CM
o co a*>
f. ^ ».
CT\ r- CO
ooo
CM
CO
oo
ro
WW
O O
CM CM
•-H CM CO
CO co CO
^ ^ (-1
i-H i-H .-1
m m m
--I CN CO
i—I i—I .—i
CN CN CN
CM
CO
4-J
rC
M
•H
hJ
01
C
tn
QJ
3
cr
3
Q
. — i
CO
^>fc
W
J-l
^-^
CO
e
cO
^
iH
W
v-^
1 — 1
pq
. — i
CM
0
r^
CO
CO
-Tfe
1-1
rH
m
CO
CN
O
r--
-------
o
•H
J_j
4J
0)
e
3
O
^
0
4-1
CO
01
cd
rH
3
u
•H
4-1
M
cd
PH
•K
y .£
U 10 ^N
cd 3 p
4-1 -H ^-<
CO TD
pi
C ' — •-
O -K O
•H -X 0)
to a) w
CO 4-1 ^^
•H cd CO
e pi e
w ^--
• -X
4-1 p i , — s
•H e ^
X 0)O
W H ^/
C
O /-->
A; -H ,j
a 4-i co
cd cd S
4-1 >
co a) g
W
4-1
^ f.
O 00-— -
Cd -rl 6
4-1 CU ^-^
CO 3:
p-1
£5
p
X
'-?
H •"•
3
C x-x
0 r4
•H a) x
W 4J --~
W nj C
•H Pi O
e 4-1
w ^^^
0)
p.
£>^
H
C
rd
0) i-4 C
e oj o
Cd rQ *H
Z E 4-1
3 P.
0) Z -H
rH O) CJ
3 U W
O i-i CU
CO 3 O
o
CO
o o
—i O
. •
CM CO
CM CM
• •
o oo
o o
>— i r^*
00 00
m in
v£> O
• •
cd
01
m
4-1 4-13
•H -H O
d c H
o o
sa s c
0
rH 4-1
CN (U CO f~^
O p. O 00
o a o -H
00 O 00 Li
y; pq
M
<
;_,
O
4J
•H
C
O
^
^j-
O
o
oo
o
o
o
CM
O
o
o
CM
OO
ro
CN
O
-------
•X
•X
44 *
a cn r*.
cd 3 e
4-1 -H -^
CO T3
cd
Pi
0
•H C /-N
S-l O -X O
4J -r-l -X O>
cu cn a) cn
E w 4-1 ^~
3 -H cd o->
.-1 E K 6
o w ^^
p>
4*! • -X
a 4-i ex ^
cd -H E W
4J X QJO
CO W H --•
C
o -^
44 -H ,-J
O 4-1 CO
cd cd s;
4-1 >
CO Q) E
J N^^
W
4<5 42
CJ 00^-s
cd -H E
4-1 OJ >— "
cn 33
>|
g Is
^>
X
V-H e;
g v5
^D
CU
4-J
ca C ^
^-l O t-i
3 -H 01 >,
U W 4-1 ^~,
•H cn cd c
4-1 -H erf o
S-i E 4-i
cd w -^
PH
0)
ex
X
H
-a
c
cd
QJ M C!
E oi o
Cd 43 -H
2 E 4J
3 ex
01 IZ -H
0 S-i
S-i CU O
3 O W
O S-i 0)
CO 3 Q
0
CO
o
0
o
o
CM
o
o
o
CN
i-H
1 — I
CO
CN
o
o
p->
r~
CO
CU i— I
VJ -H
< 3
m
S-i
o a)
4-J 0
•H -H
C! M-i
O M-l
S 0
>i
•~D 4-J
o o c
O O 3
iH 00 O
H O
O
o
o
o
CN
O
o
o
CN
CN
o-i
CN
CN
O
O
0
r>
vD
r~~*-
CN
o
o
o
CT\
OO
0
o
o
o
CN
O
O
o
CM
vD
CN
m
CO
co
r\
CM
O
CO
o
o
o
CXI
O
O
o
CN
o
CM
O
CO
0)
0
S-i
3
0
CO
cd
QJ
t-i
<;
S-i
0
4-1
43 -H
CO C
iJ 0
s
i-H
ctj
S-i OO
4J O
C 0
QJ 00
o
0)
u
S-i
3
o
CO
cd
0)
S-l
•H
C
O
s
o
r— 1
O
CO
QJ
a
t-i
3
O
CO
cd
0)
S-i
<
S-i
o
4J
M -H
>-l C
o
a) S
C
w
QJ •— 1
3 1-1
o- o
3 00
Q
C-32
-------
o
-H
^-4
4_)
01
e
D
rH
O
CJ
CO
4-1
CO
0)
4-J
CO
rH
3
a
•H
4-1
)~l
CO
PH
V*
O
CO
4-1
CO
C
o
•H
CO
CO
•H
e
H
4-1
•H
X
W
s^
O
CO
4-1
CO
^
a
cfl
4J
CO
S
H
|
C
O
•H
CO
CO
•H
e
W
cu
ex
P^,
EH
r.
CU
CO
z:
CU
o
^
3
o
CO
•K
-X
•i;
en /— .
3 6
•H s — ^
TJ
cfl
Pi
, — s
-x a
-X 01
0) CO
4-J \
CO ro
Pi £
i~
£ s/
aio
H -~s
a
0 ^
•H i-J
4J CO
co 2
a> e
W
rC
OC.-N
•H e
01 -— -
K
'e'
^_^
^e
^-s
^_l
0) >>
4J ^\
d C
f* 0
4-J
s — '
TD
C
cO
r-l G
0) O
,0 -H
e 4-1
3 a
^-i
cu a
CJ CO
r-< QJ
3 O
O
CO
0
O
o
0
vO
O
CN
O
.
i — i
CN
QJ
O
14
3
O
CO
CO
CU
<^
r^~|
CxD J-i
3 0
O 4J
M -H
0 C
PQ O
^
^
4-1
r-l CN
0) rH
^3 O
•H OO
r-l
oooooooo
o
o
o
-------
#
J"
^ *
O CO /-N
co 3 e
4J -H v_^
CO T3
«
fy
l-M
o
•H C ^-
1-1 O * C.
J-i -H -X CL
CU CO CU CT
6 W 4-> -•-.
3 -H rd co
^H 6 PS E
O W *-
p>
^ • *
O 4J PL,/--.
03 -H E >4
4-1 X CUO
co w H v_/
q
O /~s
-M -H ,J
O 4-1 CO
03 cd g
4-J >
CO CU g
r~H '••— ^
w
4J
^ X!
O Mx-s
03 -H g
4-1 CU s-'
CO 33
>H
*£. '?
H ^
S
X
X 6
H 0-
3
CU
J-J
rd C ^
i-( O M
3 -H 0) to
a w 4-1 ---^
•H CO rd C
•M -H Pi O
MS 4-J
cd w N-/
A^
CU
a
H
TD
3
03
CU M C
B cu o
ca xi -H
2 E 4J
3 P,
d) g -H
CJ t-l
s-i cu a
DOM
OS-id)
CO 3 Q
o
CO
ooo o oo o oo
ooo o •* ^O >-£) x^- ^j- vo >^£) ^
ro ro. ro ro LO LO ro ro ro
uj ^U ^ij LO LO LO LO CN CN
CNCNCN c^ roro ro CNCN
CNCNCN CN CNCN CN CNCN
OOOCO OO CO t— 1 LO LOLO
— • i CO i— i ,— i -nro ro roro
r~~4 t — 1 r-H
OO^D o oo o o CT* CTN o^ r^ r^^ ^f oo co
<*' ^f\ t\ n^.
o> -v0
v£>r-~C_^ CO r— l !— l ^H \D ^D
ooror^ o roro f— H ^Ovo
CM CN ^
OOO O OO O OO
x;
M
S-l
XI 3
TO M ^
D- S-l W .
C. XI Xl XI XI DXIXl 4-) X U^XI
•HCJCIO CJ X1CJCJ 4-1 CJ CCJU
3CIC-1C3 (3 C0(i!c- *H (II I — 'rJC
CTCUCUO) OJ 4JCUQJ Pn^-sCU QJCU
•H333 3 4-133 1 fj 3 -33
r-n o- o* cr a) o* -HO-O- occuo" ocro-
< i-H FM C co 00
X) -H CO d
fji^-oor- -H-H ,-j,— ICN rHcuo cd^-(04
I-H i— i CN U^D oo cuc^r c3- O — i CU^-ir-i
CTi O"-\ CT> S-l CJ^ O^i O> XI ^ CTi _£j CT^ O>
^ U I-) % ^ CO
C-34
-------
o
•rH
(H
QJ
6
3
i— I
O
U
cd
4-1
CO
0)
cd
rH
O
•H
4J
^1
cd
PH
y
O
4J
in
C
0
to
w
•H
B
w
4-)
•H
X
W
^
o
cd
4-1
CO
o
cd
4-1
CO
X
g
3
X
,,_,
EH
J 1
c
0
co
w
•H
B
W
QJ
£
^
OJ
B
cd
3
QJ
U
>n
3
O
CO
•X
•X
-x
en ,-N
•H - —
T3
cd
Pi
,, — ,,
**C QJ
'cu en
4-> --~
cd co
Pi £5
~
ft^
E ?v^
QJO
H ^
C
O '"•N
•rH h4
4J CO
cd ^
>
^
UJ r^i
4J -^.
cd c
Pi 0
•M
s — '
T3
C
c^
tj C
QJ O
f> -H
e 4-1
2 ("Vi
:z -H
jn
QJ a
o en
!-l QJ
3 Q
O
CO
LO LO O O O O
LO o r~^ O cri LO
i — i i — i i — i co
a> LO
O OO LO CN < — 1 1 —
ro I-H CN r-i LO
co
o>oM-uo- ro ^a~
--a* — i ^r oo oo vo co
v£) * n
' — 1 I— t
CW rHQJ JS M ' 4-1 -rH
M (_i V-l in -H E O i-J OQO
i — ! i — 1 i — t 4J -H p | , — 1
-pqpqpqcoi-! pq t-HC
w v-i cd o
i-i O C 4J
QJ • — ^CNiro
i — t i — i i — i LO r^
rH —1 CN
LO CN LO CN LO
O CN O CN O
LO en
CO CO CO
CO CO CO
LO LO LO
CO O CO
-------
o
•H
M
4-1
0)
B
3
i— i
O
>
A!
O
cd
4-J
CO
0)
4-1
cd
tH
3
O
•H
4-1
M
cd
PH
Ai
O
m
4-J
CO
c
o
•H
cn
cn
•H
e
W
4-1
•H
X
W
A!
O
cd
4J
CO
ij
^
a
cd
4J
r/3
><
g
t3
X
£1
H
to
C
O
•H
W
cn
•H
c
W
01
a.
>.
H
OJ
e
cd
^
cu
o
1-1
3
o
CO
•X
-X
cn ^-s
3 e
•H -^
t)
cd
&
,-s.
-X 0
•x at
0) CD
4-1 - —
cd ro
* sE
t _v
(2L, /— N
e «
(DO
H v^
C
O '~^
•H ^4
4J CO
TO g
>
%
4-1 ~^
cd c
Pi O
4-1
TJ
rj
^ C
01 O
,0 -H
e 4-1
3 CX
2 -H
^-t
0) O
a cn
M 0)
3 Q
O
CO
0 0
O O
m m
o o
m in
st st
o o
CN CM
OO CO
1 — t , — 1
0 0
o o
ro ro
m in
sD vO
sr st
si" sr
o o
o o
m in
•-, ^
ro ro
ro ro
m m
CO st
so r-s
r- ^H
i — i i — i
rCl >l >1 >~>
C5JD >-( s-i . >-i
>-i oi 01 o)
3 4-> 4-1 4-1
,JO 4-1 cn 4-1 4-1
co cd j£ cd j2 cd
4J pq cfl M CD M
4J 0) 3
•H 01 iJ Ol PH 0)
PM A! A! A!
1 O CO O CO O
M c_j r^ c_j ^^ r_3
C "~
•H
rH sr in ^D
0) CO ro CO
0) O O O
,c o o o
12 •— i — i — i
o
ro
CO
, — f
m
m
st
o
CN
ro
CN
O
O
ro
m
vD
sr
st
O
0
m
**
ro
ro
m
in
.
o
CTi
sO
, — 1
>,
^
01
si 4-1
CJ 4-1
d fd
0) M
3
cr o)
Ai
CO O
~fe O
r^
ro
O
O
i — i
O
o
m
o
in
st
o
CM
CO
O
o
ro
in
vO
st
sj-
O
O
m
n
ro
ro
m
sO
o
ro
i-H
bO
C
•H
rH
T3
c
Cfl
E
CO
~r,~.
CO CO
ro >£>
r-~ i-s.
in r-s
r-< O
sD ro
OJ CN
•st -st
-J -si-
ro ro
co co
O O
ro ro
ro ro
O O
o o
rsj ro
ro ro
r* ^
m m
•—i i—i
m m
CN
co
cn
ro
O O
O
m
o o
O CO
O
m
CN O
-st m
o o
1—I I—I
CN ex)
sD
ro
O O
o o
co CO
O O
-st T=
1-4
tH
ffl
CN
St
O
0
1 — 1
a
o
•H
4J
cd
4-1
CO
!-j
0)
5
o
PH
T— 1
01
,fl
O
4-1
•H
^
CN
n
1 — 1
~£fe
tn
M
r- 1
FQ
1 — 1
m
o
o
, — i
T3
1-1
cd
>>
,e
o
4-1
•H
$
CO
CN
in
0
0
t — t
CN Si-
r—I
cd
a
.^
g
0)
U
TJ
0)
•rl
r—l
H
-------
u
•H
^j
JJ
0)
e
3
O
^
O
to
4-J
CO
CU
03
rH
3
u
•H
4-1
J_l
to
PM
v^
a
cfl
4J
CO
e
o
•H
CO
CO
•H
e
W
4-J
•H
fxj
w
^
u
03
4-J
CO
a
03
4J
CO
^
g
X
s
H
C
O
•H
CO
CO
•H
e
W
01
ft
^
H
CO
g
03
^
0)
o
f-l
3
O
CO
-X
•K
-X
CO ^
3 e
•H s — '
TD
03
, — s
-X O
•K a)
CU CO
4J — -
o3 on
PH S
"
. -X
ft ^
6 !^
cuo
H •*-*
C
o /— •
•H ,-4
4J tO
03 S.
cu ^e
w
4—1
00 /-N
•H e
cu *->
K
,-v
^
"9"
"^
^^
^-i
cu >^
4J — ~.
03 C
PH O
4-1
'^ — '
C
03
r~l 3
CU O
^ -H
e 4-1
3 o 1
a -H
^_i
0) U
o in
M CU
3 O
O
CO
o o
LO -d-
• .
CM > — l
LO O
• «
< — i on
-3- on
CM •-<
o o
LO on
<]-
c^ c^
LO LO
LO
j=
w
, — i
pq
^«
CM
O
, — 1
i — i
X
C
0)
ft
e
o
u
£-fl
cu
[5
O
O
•H
^
O
CM
i — i
-ifc
^ ^
4J
c
rH
P-l
TD
TD
•H
H
CO
L^
^
pq
, — i
on
o
i — i
i — i
on
j_i
rH
pq
CM
on
O
, — i
. — i
C-37
-------
a
•H
^
4-1
OJ
^
3
rH
O
CJ
cd
4-J
CO
CU
4— i
CO
rH
3
a
•H
4-1
M
cd
PM
v1
U
cfl
4-J
CO
rj
O
•H
CD
CD
•H
E
W
4J
•H
X
w
^
o
ca
4-1
CO
CJ
cfl
4-J
CO
;x
2
g
X
^
p
£3
C
O
cn
cn
•rH
e
W
OJ
p.
£>>
H
OJ
e
r?
0)
u
M
3
O
CO
•i;
-X
cn /~,
3 6
•H >— •
-a
cd
Pi
^-^
•iC CJ
-K 01
CD CD
4J -~~>
cd m
pi e
CL /-N
E bi
QJO
H ^
C
O s-*.
•H ,J
4-1 C/3
cd s
I>
QJ E
W
4J
60 /-N
•H E
a) *-^
K
'g'
^
^-s
e
v^^
^— V
M
QJ !>i
4J ^~
cfl C
Pi O
4-1
^— ^
Td
C
cd
j_, pi
0) O
n -H
E 4-1
3 a
t-<
Q) CJ
a cn
*-) a)
3 Q
0
CO
v£> ^D
01 vO
• .
CO CO
in o
• •
o ^o
O CN
O"\ O
».
! 1
l~~ O
vD LO
LO LO
-------
|