EPA-903/9-8I-OOI
MAY 1981
Envlrenmental Fioteciion Agency
U.S. ENVIRONMENTAL PROTECTION AGENCY
CALCULATIONS FROM
COMPLIANCE EMISSIONS
OF LONG- AND SHORT-TERM
S02 CONCENTRATIONS IN THE
SOUTHWEST PENNSYLVANIA
AIR QUALITY CONTROL
REGION
Region III Library
owfrimnental Protection Agency
U ", ['r.vircnrcflntal Pretedton Ageaqf
ho , in III Im'oimation ftesoufc* --
Ct:iftr (CPW5?) x ;.»
841 1'hKinut Street ';'' :-|
PJhi- Je^liia, PA 19107 /_;. ^J
EPA Report Collection
Information Resource Center
US EPA Region 3
Philadelphia, PA 19107
MIDDLE ATLANTIC REGION-1
6th and Walnut Streets, Philadelphia, Pennsylvania 19106
-------
EPA - Online Library System Record
Page 1 of 3
(/.$. Environmental Projection Agency
Libraries
Contact Us | Print Version
EPA Home > Information Sources > Libraries > Online Library System > OLS Record
OLS Record
Q
Search
Swtiira
Serial Record
[ Display Records as Bibliography ] [ Item Status ]
RECORD NUMBER: 1 OF 1
Main Title
CORP
Author
Publisher
Year
Published
Report
Number
Stock
Number
OCLC
Number
Subjects
Subject
Added Ent
Collation
Holdings
Abstract
Calculations from compliance emissions of long- and short- term SO2 concentrations
in the Southwest Pennsylvania air quality control region.
Cramer (H E.) Co., Inc., Salt Lake City, UT.; Environmental Protection Agency,
Philadelphia, PA. Region III.
U.S. Environmental Protection Agency,
1981
EPA-903/9-81-001 ; 68-02-2547; TR-81 -136-01; EPA-68-02-2547
PB81 -226078
37287447
Air pollution; Sulfur dioxide; Mathematical models; Concentration (Composition);
Pennsylvania; Transport properties; Monitoring; Inventories; Air quality control region;
Numerical solution
Air-Pollution--Pennsylvania
1 v. (various pagings) : ill., maps ; 28 cm.
LIBRARY CALL NUMBER LOCATION
EJAD EPA 903/9-81-001 Region 3
Library/Philadelphia,
PA
NTIS PB81-226078 Most EPA libraries have a NTIS
fiche copy filed under the call
number shown. Check with
individual libraries about
paper copy.
This report describes the results of dispersion-model calculations of maximum annual, 24-
http://cave.epa.gov/cgi/nph-bwcgis/BASIS/ncat/lib/ncat/DDW?W%3D+%28TITLE+PH+... 8/17/2006
-------
EPA - Online Library System Record
Page 2 of 3
Notes
Corp Au
Added Ent
Place
Published
PUB Date
Free Form
NTIS
Prices
Holdings
Modified
Bib Level
OCLC Time
Stamp
Cataloging
Source
Language
OCLC/NTIS
Type
1 OCLC Rec
| Leader
hour and 3-hour average ground-level S02 concentrations for selected areas in the
Southwest Pennsylvania Air Quality Control Region (AQCR). The primary purpose of the
model calculations was to assist EPA Region III and the Pennsylvania Department of
Environmental Resources in determining the attainment or non-attainment of the National
Ambient Air Quality Standards (NAAQS) for SO2 in the Beaver Valley and Monongahela
Valley Air Basins exclusive of Allegheny County. All of the dispersion-model calculations
were made using the LONGZ and SHORTZ dispersion models with 1980 compliance
emissions inventories containing 492 major SO2 sources located within the Southwest
Pennsylvania AQCR and in Ohio and West Virginia near the western border of the AQCR.
The only calculated maximum that exceeds the NAAQS for SO2 is the maximum annual
average concentration at an isolated grid point located on high terrain about 1 kilometer
north of the Monessen Plant of Wheeling-Pittsburgh Steel. The model calculations also
indicate contributions of major SO2 sources located along the Ohio River in Ohio and
West Virginia to the air quality in the Southwest Pennsylvania AQCR.
Prepared for U.S. Environmental Protection Agency by H.E. Cramer Company, Inc. under
contract no. 68-02-2547, task order no. 2. "May 1981 " "EPA-903/9-81-001." Project
officers: Brian McClean and Alan Cimorelli
United States. Environmental Protection Agency.
{Washington, D.C.} :
1981.
PCA11/MFA01
LIBRARY Date Modified
EJA 19970718
m
19970714084123
OCLC/T
ENG
OCLC
MERGE
01004nam 2200253Ka 45020
Display Records as Bibliography j f Item Status
If there is an AREA footer, put it here.
EPA Home | Privacy and Security Notice | Contact Us
http://cave.epa.gov/cgi/nph-bwcgis/BASIS/ncat/lib/ncat/DDW?W%3D+%28TITLE+PH+..
8/17/2006
-------
EPA-903/9-81-001
May 1981
CALCULATIONS FROM COMPLIANCE EMISSIONS OF LONG-
AND SHORT-TERM S02 CONCENTRATIONS IN THE
SOUTHWEST PENNSYLVANIA AIR QUALITY
CONTROL REGION
Prepared for:
U. S. Environmental Protection Agency
Under
Contract No. 68-02-2547
Task Order No. 2
nwmatten
g/il riiesiir ui Sired V >
PhaafclpMi, PA Ml«7 fc^^^
H. G. Cramer company, inc.
UNIVERSITY OF UTAH RESEARCH PARK
POST OFFICE BOX 8049
SALT LAKE CITY, UTAH 84108
-------
DISCLAIMER
This report was furnished to the Environmental Protection Agency by
H. E. Cramer Company, Inc., University of Utah Research Park, P. 0. Box
8049, Salt Lake City, Utah 84108, in fulfillment of Contract No. 68-02-
2547, Task Order No. 2. The contents of this report are reproduced herein
as received from H. E. Cramer Company, Inc. The opinions, findings, and
conclusions expressed are those of the author and not necessarily those
of the Environmental Protection Agency.
ii
-------
ACKNOWLEDGEMENTS
We are especially indebted to Mr. Brian McClean and Mr. Alan
J. Cimorelli, our EPA Region III Project Officers, for their technical
assistance and guidance during this study. The emissions inventories
and most of the air quality data used in the study were principally
supplied by Mr. Wick Havens and Mr. Kenneth Bowman of the Pennsylvania
Department of Environmental Resources and by Mr. William Muer and
Dr. Roger Westman of the Allegheny County Bureau of Air Pollution Control,
We also received valuable assistance from the Southwest Pennsylvania
Planning Commission, from Mr. James Smallwood of the West Virginia
Air Pollution Control Commission and from the Ohio Environmental
Protection Agency.
111
-------
EXECUTIVE SUMMARY
This report describes the results of dispersion-model calcula-
tions of maximum annual, 24-hour and 3-hour average ground-level S0~
concentrations for selected areas in the Southwest Pennsylvania Air
Quality Control Region (AQCR). The primary purpose of these calculations
was to assist EPA Region III and the Pennsylvania Department of Environ-
mental Resources in determining the attainment or non-attainment of the
National Ambient Air Quality Standards (NAAQS) for S02 in the Beaver
Valley and Monongahela Valley Air Basins outside of Allegheny County.
Allegheny County was excluded from the dispersion-model calculations in
this report because similar calculations were made in a study performed
by TRC, Inc. for the Allegheny County Bureau of Air Pollution Control.
However, because the SO- sources in Allegheny County affect the S0_
concentration in the Beaver Valley and Monongahela Valley Air Basins,
the major SCL sources in Allegheny County were included in the emissions
inventories and dispersion-model calculations contained in this report.
All of the dispersion-model calculations were made using the LONGZ and
SHORTZ computer programs containing the long-term and short-term dispersion
models described in Appendix A. In the annual average S02 concentration
calculations for determining attainment or non-attainment of the NAAQS,
the annual 1980 compliance emissions inventory in Appendix C.I was used
with the meteorological inputs given in Appendix B.I which are based on
the 1965 hourly surface weather observations made at the Greater Pittsburgh
Airport and Greater Pittsburgh Airport mixing depth statistics for 1960
through 1964. The 1965 Greater Pittsburgh Airport data were selected as
typical of a worst-case year in which the meteorological conditions were
conducive to high ground-level S0~ concentrations in the Southwest
Pennsylvania AQCR. In the short-term concentration calculations for
determing attainment or non-attainment of the NAAQS, the short-term 1980
compliance emissions inventory in Appendix C.2 was used with the hourly
surface weather observations and twice-daily rawinsonde observations
made at the Greater Pittsburgh Airport on 28 August 1976. This day was
selected as the worst-case 24-hour period for high calculated SCL concentra-
iv
-------
tions in the Southwest Pennsylvania AQCR on the basis of a detailed analysis
of five years (1973 through 1977) of. meteorological observations from
the Greater Pittsburgh Airport. Specifically, the hourly wind observations
for these five years were analyzed to isolate those cases in which wind
directions from the southwest quadrant persisted for more than 12 hours
with mean wind speeds greater than 2 meters per second. Previous dispersion-
model calculations showed that these persistent wind conditions produced
the highest 24-hour SO concentrations in the Southwest Pennsylvania AQCR.
Of the 20 cases found in the persistence analysis which satisfied the
above criteria, the 28 August 1976 case resulted in the highest calculated
24-hour average SO- concentrations in the Southwest Pennsylvania AQCR.
The annual and short-term 1980 compliance emissions inventories
in Appendix C.I and Appendix C.2 contain 429 individual sources located
within the Southwest Pennsylvania AQCR and along the Ohio River Valley
near the western border of the Southwest Pennsylvania AQCR in eastern
Ohio and West Virginia. The 1980 compliance emissions inventories used
in the dispersion-model calculations were obtained from the following
agencies:
Allegheny County Bureau of Air Pollution Control -
Sources in Allegheny County
Pennsylvania Department of Environmental Resources -
Sources in the Beaver and Monongehela River Air
Basins and in other areas of Pennsylvania
EPA Region III, the Ohio Environmental Protection Agency,
and the West Virginia Air Pollution Control Commission -
Sources in Ohio and West Virginia
The S0~ emissions and other source parameters supplied by the above
agencies were carefully evaluated by the H. E. Cramer Company for accuracy
and completeness and were also reviewed and updated by the Pennsylvania
DER and EPA Region III.
v
-------
The results of the long-term and short-term dispersion-model
calculations made using the 1980 compliance emissions inventories are
presented in Table I which lists the maximum annual, 24-hour and 3-hour
average SCL ground-level concentrations calculated for the New Castle,
Beaver and Monessen grids. These grids, which are described in Section
5.1 and shown in Figure I, were selected as the areas within the Beaver
Valley and Monongahela Valley Air Basins with the highest SO,, ground-
level concentrations on the basis of dispersion-model calculations made
using the LONGZ computer program with the annual 1980 compliance emissions
inventory and the gross calculation grid shown by the dashed lines in
Figure I. The major S0« sources and SO- monitor sites shown in Figure I
are identified in Tables II and III. The only calculated maximum which
exceeds the NAAQS for SO,, is the maximum annual average concentration
for the Monessen grid which occurs at an isolated grid point located on
high terrain about 1 kilometer north of the Monessen Plant of Wheeling-
Pittsburgh Steel. The calculated concentrations at adjacent grid points,
which are at a distance of 1 kilometer, range from 40 to 61.5 micrograins
per cubic meter. The area in which the calculated annual average concen-
trations are above the NAAQS is therefore less than 0.25 square kilometers
in extent. Emissions from the Monessen Plant of Wheeling-Pittsburgh
Steel are responsible for approximately 58 percent of the maximum annual
average concentration calculated for the Monessen grid.
An important feature of both the long-term and short-term
model calculations for the New Castle and Beaver grids is the relatively
large contributions from major S0« sources located along the Ohio River
in eastern Ohio and West Virginia. For example, the combined emissions
from all eastern Ohio and West Virginia sources account for 39 percent
and 35 percent of the maximum annual average concentrations calculated
for the New Castle and Beaver grids, respectively. According to the
short-term model calculations, the combined emissions from all of the
eastern Ohio and West Virginia sources account for about 60 percent of
the maximum 24-hour and 3-hour average concentrations calculated for
each of these grids. The air quality impact of the major S0« sources in
vi
-------
M
W
PQ
2
V '
1
o
CO
CU
4-1
H
13
O
o
CJ
g
H
X
X
d
o
H
4J
ca
Vl
4-J
d
cu
CJ
d ^
O CO
6 oo
C5 *^X
t3
r]
rri
s
o
Calculati
Grid
d
o
H
4-1
ca
d
cu
u
o
o
cu
00
Vl
cu
Si
.
d
d
'*"N
0 O O
in co r*-*
cs o **
O 0 O
«-
oo cu
13 ca d
55 o> o
£5 PQ K^J
d
o
H
ca
g
o
d
o
o
0)
00
ca
Vl
0)
M
^
o
CM
^£> O O
co co' r-~"
co o i "-o
^-i m CM
r-l CM CM
CU
1
r^
4-1 d
w cu
ca vi co
o cu co
> 2
3 ca d
CU CU O
z PQ a
o
H
4J
c8
Vl
4J
d
cu
0
d
o
CJ
cu
00
Vi
cu
%
Vl
1
1
CO
VO o O
* *
CO CO r>
co o -
t-^. m co
co co O
co oo r»
0)
^ ca d
a) cu o
!Z P3 S
d
H
ca
Vl
Vl
CU
cu
ca
&
i i
CU
d
o
4-1
d
H
o
Cu
-H
Vl
00
0)
4-1
iH
O
co
H
ca
CO
Vi
3
0
a
0
6
I
alculated
CJ
K
vii
-------
FIGURE I. Map of the Southwest Pennsylvania AQCR and surrounding areas.
Dashed lines show the gross grid and shaded areas show the
five smaller grids used for dispersion-model calculations.
Major S02 sources and source complexes are indicated by the
numbers 1 through 31 and S02 monitor locations used for model
validation are indicated by the letters A through M.
Vlll
-------
TABLE II
LIST OF MAJOR S00 SOURCES IN FIGURE I
Source
Numb er
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
Source Name
Penn Power, West Pittsburgh
St. Joe Minerals
J & L Aliquippa
Duquesne Light, Phillips
Duquesne Light, Bruno t Is.
Duquesne Light, Cheswick
J & L Pittsburgh
USS Homestead
USS Edgar Thomson
USS Duquesne
USS National
USS Irvin
USS Clairton
Duquesne Light, Elrama
Penn Power, Mitchell
Hatfield Power
Armstrong Power
Keystone Power
Homer City Power
Seward Power
Conemaugh Power
Ohio Ediwon, Sammis
Ohio Power, Toronto
National Steel, Wierton Div.
Wheeling-Pittsburgh (North)
Wheeling-Pittsburgh (South)
Cardinal Power
Ohio Edison, Burger
Kammer Power
Mitchell Power
PPG
Penn Power, Mansfield
Wheeling-Pittsburgh, Monessen
UTM Coordinates (m)
X
553,105
556,065
564,400
565,260
580,680
602,330
589,000
592,100
597,260
598,324
597,000
593,180
595,500
592,000
587,340
591,570
628,979
640,141
652,756
666,882
664,582
531,700
533,500
534,300
532,500
534,600
530,000
520,500
515,320
515,800
512,600
549,049
594,170
Y
4,531,767
4,502,301
4,497,500
4,491,020
4,479,680
4,487,800
4,473,900
4,473,100
4,471,685
4,470,025
4,467,500
4,465,700
4,461,500
4,456,200
4,452,810
4,412,040
4,531,996
4,502,155
4,486,147
4,474,602
4,471,929
4,485,500
4,481,800
4,474,400
4,466,700
4,463,300
4,455,800
4,417,500
4,410,320
4,408,670
4,399,600
4,498,067
4,446,350
-------
TABLE III
LIST OF S02 MONITOR SITES IN FIGURE I
Site
Symbol
A
B
C
D
E
F
G
H
I
J
K
L
M
Site Name
(Operator)
New Castle (DER)
Beaver Falls (DER)
Fairview (Penn Power)
Route 68 (Penn Power)
West Beaver (Penn Power)
Midland (DER)
Baden (DER)
Logans Ferry (BAPC)
Downtown (BAPC)
Hazelwood (BAPC)
North Braddock (BAPC)
Liberty Boro (BAPC)
Glassport (BAPC)
UTM Coordinates (m)
X
554,830
557,750
544,820
550,720
548,950
546,330
565,090
605,154
585,150
589,762
596,680
596,210
594,190
Y
4,537,240
4,510,785
4,504,390
4,500,640
4,502,000
4,498,340
4,498,380
4,489,115
4,476,600
4,473,952
4,472,835
4,464,150
4,463,580
Site
Elevation
(m MSL)
257
220
390
238
375
249
230-
268
256
284
275
340
234
-------
eastern Ohio and West Virginia is greatest at or near the western border
of the Southwest Pennsylvania AQCR. To obtain preliminary estimates of
this impact, we included the two border grids 3 and 4 shown in Figure I
in both the annual and short-term model calculations using the 1980
compliance emissions inventories. The annual average calculations
showed a maximum concentration of 101.4 micrograms per cubic meter on
the Pennsylvania-West Virginia border about 10 kilometers northeast of
Weirton, West Virginia. Calculated concentrations are above the annual
NAAQS within a narrow border strip approximately 17 kilometers long and
4 kilometers wide extending north and south from the point of the calculated
maximum. Sources in eastern Ohio and West Virginia respectively account
for about 35 percent and 53 percent of the maximum annual concentration.
However, there are some points located within the areas where the calcu-
lated concentrations are above the annual S0? standard at which sources
located in Ohio account for more than 50 percent of the calculated
concentrations. The short-term model calculations show a maximum 24-
hour average concentration of 425.2 micrograms per cubic meter and a
maximum 3-hour average concentration of 2375 micrograms per cubic meter
at a point located about 4 kilometers east of the Pennsylvania-West
Virginia border and about 10 kilometers southeast of Weirton, West
Virginia. Sources in West Virginia account for 83 percent of the calcu-
lated 24-hour maximum and for virtually 100 percent of the calculated 3-
hour maximum. We also found several other 3-hour cases in 1976 in which
the calculated 3-hour maximums in this border area are greater than 2000
micrograms per cubic meter and are produced almost entirely by emissions
from sources in West Virginia.
We point out that the model calculations for the border areas
described above are preliminary values because they were made using
worst-case meteorological conditions selected for other grid areas and
therefore may not be representative of the worst-case situations for the
border areas. In addition, there are questions about the accuracy of the
emissions inventories for the West Virginia and Ohio sources used in the
XI
-------
model. Therefore, while we believe the results of our model calculations
for the border areas serve to point out potential SCL air quality problems,
an additional effort beyond the scope of the present study is required
to define the magnitude and areal extent of the annual and short-term
SCL concentrations along the western border of the Southwest Pennsylvania
AQCR.
For purposes of validating model calculations made by the
LONGZ and SHORTZ computer programs for the Southwest Pennsylvania AQCR,
we originally planned to use 1975 emissions inventory, meteorological
data and S0? monitor data. However, because of deficiencies in the
Pennsylvania DER and Allegheny County monitor data for 1975 and problems
with emissions inventories, it was necessary to use the 1976/1977 emissions
inventory in Appendix C.3 and meteorological data for 1976 and 1977 in
combination with SC- monitor data for the years 1976 through 1979 in the
model validation study. Table IV lists the observed annual average S0~
concentrations at various monitor sites and the calculated annual average
SCL concentrations at these monitor sites. The concentrations were calcu-
lated by using the LONGZ computer program with the 1976/1977 emissions
inventory in Appendix C.3 and the meteorological inputs in Appendixes
B.2 and B.3 developed from the 1976 and 1977 hourly observations at the
Greater Pittsburgh Airport. As shown in the table, the observed concen-
trations are less than half of the concentrations calculated at the Baden
monitor in 1976 and at the Hazelwood monitor in both 1976 and 1977. We
believe the comparison of observed and calculated concentrations at the
Baden monitor can probably be disregarded in view of the difference in
time between the observation period and the period for which the calcula-
tions were made, as well as the 45.7-percent valid data recovery rate
at the Baden monitor. We are unable to explain satisfactorily the large
model overpredictions at the Hazelwood monitor, except to point out that
a detailed analysis of the calculated concentrations in the vicinity of
this monitor showed a very large spatial gradient of concentration over
an area of about 1 square kilometer. Calculated concentrations equal to
the observed concentrations are found within a distance of about 500 meters
to the south and northwest of the monitor site.
xii
-------
w
Q W
U W
£
rH
VQ
O^
rH
T3
^i
0)
CO
0
01
4_1
cO
rH
~j
O
J
cO
a
d
o
13
CU 4
H
J
4J Cfl /~s
CO JH n
rH 4
3
a
j e
\) 6C
rH U 3-
Cd d **~^
U 0
u
t
cu
j>
i-t
CU
0)
0
0)
4-1
CO
rH
3
CJ
\ 1
d
0
13 -H
4-1 03 x-v
CO M M
( \ .
-j e
3 d -^
o
u oo
rH a 3-
rt
-J *""" ^
0 0
u
d
o
H
T3 4J
0) CO /
->
> »nm
M 4-)
cu d -
tn a)
jo o
O d v
0
CJ
d
o
i
n
4J 13
ca o
> -I-
co a
tn PL,
£>
O
^
o
4-1 111
H 4J
d -H
O CO
^
B
-^
bO
^
*^
*-x
J_J
o
4J
cd
M
OJ
a.
o
<-s
coo^OLor^ o -~~t »~H o^ *«o
~*d" i"H pH LO "JD
*
»f r**- o> co o** LO *~n " O> O^ O^ O^
CO CO OO H I » » i4 « 1 f-H ? 1
r^- r^- r*-*
*^ ** . *^^.
^ 0s* O
>-j
^ /-*V V«v O
/^^s p^ pij V-) n3 J-t MM
fVj fi^ QJ p^ QJ 0} 1) QJ O
WOrH N-' SMS SM 13
QV-/W So oooo d o
^^ en 13 cu p i d p i ^o PM ?**» S O
i-4 CO d "r^ &H 4-J x^» O ^~s S >~^
d d cud cud PO -ucj rncj
QJ ^> pi rH i-* d 4-1 d 4-1 d CU C1 1 P p i CU PH
13 CO S W 13 'H CU CO CU 3 CU n ^ 3
-------
13
01
3
H
4-1
a
o
o
£
rH
£
cn
rH
1 TJ
0)
}_l
01
to
o
0)
4->
cd
i i
^3
O
, |
cd
u
(3
0
13 vH
QJ 4-1
4-J
cd >.
u ' ^
j ro
rH 4J E
3 C
2 ^^
O Ol bC
t-l C
J 3-
ca d s--
u c
5
o
13
01
S-i
01
CO
o
13
0)
4-1
ca
rH
3
a
, |
ca
u
a
o
13 T
O! 4.
H
j
4-1 Cd /-v
cd I
-J CO
rH 4J E
3 C! ---
O 0) M
rH O 3
U 0
O
C
O
13 4-1
01 cd f
*> S-4O"!
-N
to 4J g
o> a ~~
-^
CO O> 00
,a u ^t.
o c --
0
u
c
o
rH
4-J 13
cd o
> -H
o> 01
tO p-i
f>
O
^
!
*S
-N
-1
M O
0 W
4-J 0) (
Tj
H 4-> to
C -H -,
13 to
T3 !-i
4-1 cd ai
to to P^i
o «
CXx-"s /^N 07 ^s
to o & u c u
CO PJ 4-1 Pn cd P-j
cd
-------
TABLE OF CONTENTS
Section Title Page
DISCLAIMER ii
ACKNOWLEDGEMENTS ±±±
EXECUTIVE SUMMARY iv
1 INTRODUCTION 1
2 S02 EMISSIONS INVENTORY 9
2.1 S0£ Industrial Emissions Inventory
for 1976/1977 10
2.2 S0£ Industrial Compliance Emissions
Inventory for 1980 11
2.3 SOo Area Source Emissions Inventory 12
3 METEOROLOGICAL INPUTS 23
3.1 Meteorological Inputs Required by
the Long-Term Model (LONGZ) 23
3.2 Meteorological Inputs Required by
the Short-Term Model (SHORTZ) 30
4 SO2 DECAY AND BACKGROUND
4.1 Adjustment of the Model Concentration
Calculations for S02 Decay 33
4.2 Estimates of S02 Background Concentration 36
5 LONG-TEEM CALCULATIONS 41
5.1 Calculation Procedures 41
5.2 Long-Term Calculation Results 46
6 SHORT-TERM MODEL CALCULATIONS 55
6.1 Calculation Procedures 55
6.2 Short-Term Calculation Results 58
7 MODEL VALIDATION 68
7.1 SO2 Monitor Data Used for Model
Validation 69
7.2 S0? Emissions Inventory Data Used for
Model Validation 70
xv
-------
TABLE OF CONTENTS (Continued)
Section
Title
7.3 Long-Term Validation Calculations
7.4 Short-Term Validation Calculations
73
75
PRELIMINARY CALCULATIONS OF S02 CONCENTRATIONS
ALONG THE PERIMETER OF THE SOUTHWEST
PENNSYLVANIA AQCR 79
8.1 Long-Term Calculations 79
8.2 Short-Term Calculations 81
REFERENCES
Appendix
A
MATHEMATICAL MODELS USED TO CALCULATE GROUND-
LEVEL CONCENTRATIONS A-l
A.I Introduction A-l
A.2 Plume-Rise Formulas A-7
A.3 Short-Term Concentration Model A-9
A.4 Long-Term Concentration Model A-16
A.5 Application of the Short-Term and
Long-Term Concentration Models in
Complex Terrain A-21
METEOROLOGICAL INPUTS USED IN THE LONG-TERM
MODEL CALCULATIONS (BASED ON METEOROLOGICAL
OBSERVATIONS MADE AT THE GREATER PITTSBURGH
AIRPORT) B-l
B.I 1965 Seasonal and Annual Inputs used
for the 1980 Compliance Calculations B-2
B.2 1976 Seasonal and Annual Meteorological
Inputs Used for Model Validation
Calculations B-9
B.3 1977 Seasonal and Annual Meteorological
Inputs used for Model Validation
Calculations B-l6
xv i
-------
TABLE OF CONTENTS (Continued)
Section Title Page
C S02 EMISSIONS INVENTORIES C-l
C.L Southwest Pennsylvania AQCR Annual
Compliance Emissions Inventory C-2
C.2 Southwest Pennsylvania AQCR Short-
Term Compliance Emissions Inventory C-29
C.3 Allegheny County 1976/1977 Emissions
Inventory for Validation Study C-57
C.4 Corrections and Changes to the
Emissions Inventory C-80
xvii
-------
SECTION 1
INTRODUCTION
This study was originated under EPA Contract No. 68-02-2522 to
provide dispersion-model calculations of the air quality impact of
future S0? emissions in the Southwest Pennsylvania Air Quality Control
Region (AQCR) for the years 1980, 1985 and 2000 for use in planning for
industrial growth and regulatory activities. During the course of the
study, in response to changes in the Clean Air Act, the Scope of Work
of the contract was modified to provide for dispersion-model calculations
to assist in evaluating changes in the Pennsylvania State Implementation
Plan and the attainment or non-attainment of the National Ambient Air
Quality Standards within the Beaver Valley and Monongahela Valley Air
Basins. This work was continued under EPA Contract No. 68-02-2547, Task
Order No. 2 which specifically addressed the requirement for dispersion-
model calculations of worst-case annual, 24-hour and 3-hour ground-level
SO- concentrations, using 1980 compliance emissions inventories, in the
Beaver Valley and Monangahela Valley Air Basins.
All of the dispersion model calculations for the study were
performed using the LONGZ and SHORTZ computerized models described in
"Diffusion Model Calculations of Long-Term and Short-Term Ground-Level
Sulfur Dioxide Concentrations in Allegheny County, Pennsylvania"
(EPA-903/9-75-018). These models are fully documented in the "User's Instruc-
tions for the SHORTZ and LONGZ Computer Programs" (Bjorklund and Bowers,
1979). Figure 1-1 is a map of the Southwest Pennsylvania AQCR showing
the various calculation grids used in the study. The dashed line defines
the 100 x 140 kilometer gross grid which encloses most of the Southwest
Pennsylvania AQCR. This gross grid was used for LONGZ dispersion-model
calculations in the beginning of the study to isolate specific areas of
high annual average S0« concentration in which more detailed model
calculations were required to define the air quality impact of SO emissions
-------
WESTMORELAND CO
WASHINGTON CO.
50 5 10 15 20 KILOMETERS
FIGURE 1-1. Southwest Pennsylvania AQCR showing the Beaver grid (1),
New Castle grid (2), Northwest Border grid (3), South-
west Border grid (4) and the Monessen grid (5).
-------
from major sources and source complexes. A 5-kilometer spacing of grid
points was used for the central 60-by-100 kilometer area of the gross
grid and a 10-kilometer spacing was used for the remaining outer portions
of the grid. The five shaded areas in Figure 1-1 show the smaller grids
used for the detailed dispersion-model calculations with the 1980 compli-
ance emissions inventories in which a 1-kilometer spacing of grid points
was employed. The terrain contours for each of the five smaller grids
are presented in Figures 1-2 through 1-5. In areas of calculated high con-
centrations and steep gradients, the grid spacing was further reduced to
improve the definition of the concentration field. For example, a 100-
meter grid spacing was used in the Baden area and a 500-meter grid
spacing was used in the Monessen area (see Figures 5-2 through 5-4).
Emissions inventories for S0? sources located within Allegheny
County were supplied by the Allegheny County Bureau of Air Pollution
Control (BAPC). Emissions inventories for all other S0~ sources located
within the Southwest Pennsylvania AQCR were provided by the Pennsylvania
Department of Environmental Resources (DER). Emissions inventories for
sources located in eastern Ohio and West Virginia near the western
border of the Southwest Pennsylvania AQCR were obtained by EPA Region
III from the responsible state agencies in Ohio and West Virginia.
Section 2 of this report contains a description of the 1976/1977
and 1980 S0? emissions inventories used in the dispersion-model calculations,
Meteorological inputs used in the dispersion-model calculations are
described in Section 3. Adjustments made in the model calculations for
SO- decay and estimates of SO* background concentrations are discussed
in Section 4. Results of the long-term and short-term dispersion-model
calculations made using the 1980 compliance emissions inventories are
presented in Sections 5 and 6, respectively. Model validation calculations
and comparisons with SO- monitor data are discussed in Section 7. Section
8 contains a discussion of preliminary calculations of SO,., concentrations
along the western perimeter of the Southwest Pennsylvania AQCR. There are
three appendices to this report. Appendix A contains a description of
-------
4514 -
4512 -
4510 -
4508 -
4506 ^
4504 -
4502
4500 -
4498 -
4496 -
4494 -
4492 -
4490 -
555 557 559
563 565 567
FIGURE 1-2. Terrain contours in meters above Mean Sea Level for the
Beaver grid. The numbers 2, 3 and 4 respectively refer
to the St. Joe Minerals, J & L Aliquippa and Duquesne
Light, Phillips Plants.
-------
4539
4537
4535
4533
4531
4529
4527
4525
300
553
555
557
559
FIGURE 1-3. Terrain contours in meters above Mean Sea Level for the New
Castle grid. The number 1 refers to the Penn Power, West
Pittsburgh Plant.
-------
4454 -
4452 -
4450
4448 -
4446 -
4444 -
4442 -
4440 -
4438 -
4436 -
4434
4432 -
4430
585 587 589 591 593 595 597 599 601 603
FIGURE 1-4. Terrain contours in meters above Mean Sea Level for the
Monessen grid. The number 15 refers to the Penn Power,
Mitchell Plant.
-------
4535 -
4525 -
4515 -
4465
542
544
546
548
550
FIGURE 1-5. Terrain contours in meters above Mean Sea Level for the
Northwest (3) and Southwest (4) Border grids.
-------
the LONGZ and SHORTZ dispersion models used in the study. Tabular sum-
maries of the meteorological inputs used in the LONGZ model calculations
for the compliance cases and the validation study are contained in Appendix
B. The S0_ emissions inventories used in the study are given in Appendix C.
-------
SECTION 2
S02 EMISSIONS INVENTORIES
The 1976/1977 and 1980 (compliance) SO,, emissions inventories
used with the LONGZ and SHORTZ dispersion models were developed from in-
formation supplied by the following agencies:
Pennsylvania Department of Environmental Resources
Allegheny County Bureau of Air Pollution Control
West Virginia Air Pollution Control Commission
Ohio Environmental Protection Agency
EPA Region V
EPA Region III
Information pertaining to industrial point sources located in the Beaver
Valley Air Basin, the Monongahela Valley Air Basin and to other sources
in Pennsylvania located outside Allegheny County was obtained from the
automated emissions inventory files of the Pennsylvania Department of
Environmental Resources. Emissions data for S0? sources located within
Allegheny County were obtained from the Allegheny County Bureau of Air
Pollution Control. Data pertaining to S0~ sources located in the West
Virginia Panhandle were obtained from the West Virginia Air Pollution
Control Commission by EPA Region III. After reviewing the draft final
report, the Pennsylvania Department of Environmental Resources and the
West Virginia Air Pollution Control Commission made some revisions in the
1980 compliance emissions inventory. These revisions,which were received
after the dispersion-model calculations had been completed, are listed
in Appendix C.4 for reference. Emissions data for the S0_ sources in
Ohio along the Ohio River near the Pennsylvania border were obtained
principally by EPA Region III from the Ohio Environmental Protection
Agency and EPA Region V.
-------
2.1 S02 INDUSTRIAL EMISSIONS INVENTORY FOR 1976/1977
The 1976 and 1977 S0? emissions inventory data for the Southwest
Pennsylvania AQCR were reviewed by the H. E. Cramer Company and missing
data, questionable parameter values and other potential problems were
noted. The year 1976 was originally intended to be the base year for
the initial modeling of S0» emissions. However, principally because of
the lack of valid air quality measurements and to some extent the unavail-
ability of emissions data, the emissions data for the two years 1976
and 1977 were combined to develop the S0? emissions inventory used with
the initial executions of the models. The SO,, emissions inventory
parameters for each source that are required to execute the LONGZ and
SHORTZ dispersion models are:
Source Location (UTM coordinates)
Plant Grade Elevation
SO^ Emission Rate
Source Type (Stack)
Stack Exit Gas Temperature
Stack Exit Volume
Stack Exit Diameter
Stack Height
Source Type (Building or Area)
Source Length
Source Width
Source Height
The SO- emissions inventories obtained from the various agencies
were converted to standard metric units and forwarded with the list of
questions and missing data to the originating agencies for their review
as to accuracy and completeness. The short-term emissions inventory
required additional work in developing appropriate emissions rates from
the available data. This process took several months and required a
10
-------
number of iterations in order to determine whether plant operating
levels, maximum allowable emissions or rated boiler capacities should be
used in establishing short-term S0? emission rates.
2.2 SCL INDUSTRIAL COMPLIANCE EMISSIONS INVENTORY FOR 1980
Upon completion of the 1976/1977 emissions inventory, the
various agencies involved started to develop an emissions inventory for
the years 1980-1982 to be used in the dispersion modeling of future
years for determining compliance with air quality standards and for
developing control strategies. As this task was nearing completion, the
goals of the study were modified to reflect the changes in the environmental
regulations and the need for using dispersion-model calculations to define
the areas of attainment and non-attainment for S0~ within the Southwest
Pennsylvania AQCR. Because of these modifications in the goals of the
study, emphasis was placed on the development of an S0« emissions
inventory for the year 1980 and a specific review of the inventory for
1980 was conducted by the Pennsylvania Department of Environmental
Resources and the Allegheny County Bureau of Air Pollution Control. The
1980 emissions inventory for SO., sources located within Allegheny County
presented a special problem for the Bureau in that SO emissions are
heavily dependent upon the operations of several integrated steel mills
located within the County and the availability and mix of the fuels
consumed by these mills. For example, an individual source may use
fuels that vary from natural gas with a very low sulfur content to coal
with a sulfur content of a few percent which results in a very wide
range of possible S0? emissions. The regulations of the Allegheny
County Bureau of Air Pollution Control allow individual sources to burn
the mixed fuels available at any given time. For these reasons, considerable
judgement is required in arriving at a reasonable estimate of the fuel
mixture to be used in determining the SO- emission rates for both short-
term and annual model calculations. Therefore, the 1980 compliance SO-
11
-------
emissions inventory for sources located within Allegheny County reflects
the Allegheny County Bureau of Air Pollution Control's best estimate of
the mixture of available fuels which would be consumed during 1980.
The 1980 compliance emissions inventory for SO,, sources located
in West Virginia is based upon the state's Regulation X and estimates by
the West Virginia Air Pollution Control Commission of the actual operating
levels for 1980.
The 1980 compliance emissions inventory for S0~ sources located
in Ohio is unchanged from the 1976/1977 SO- emission inventory as neither
EPA Region III nor the H. E. Cramer Company, Inc. was able to obtain
estimates of the SO- emissions from the Ohio sources for 1980. The
1976/1977 inventory was reviewed by the Ohio EPA and updated to reflect
changes in operating levels and emissions.
2.3 S02 AREA SOURCE EMISSIONS INVENTORY
During the development of the industrial SO- emissions inventory,
estimates of the area source S09 emissions inventory were also developed
using the techniques described below. Area sources were divided into
three major categories:
Mobile sources
Space heating
Industrial and commercial sources excluded from the
industrial emissions inventory
2.3.1 Mobile Sources
The four types of mobile sources that were considered in develop-
ing the area source emissions data were:
12
-------
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 Commission
(RPC). Both sets of data are for calendar 1972. The EPA data are allo-
cated by type of vehicle. The RPC data are in the form of total daily
vehicle miles per traffic zone and traffic-zone area. Table 2-1 presents
the EPA data and total daily vehicle mileage estimates developed from
the RPC traffic zone data. The total daily mileage estimates in the
table for the Southwest Pennsylvania AQCR given by the two data sets
differ by less than 1 percent. Differences in the totals for the indi-
vidual areas range from about 16 to 25 percent.
Estimates of the S0~ emissions from the exhaust of motor
vehicles were obtained by using the vehicle mileage data by vehicle
types in Table 2-2 supplied by EPA Region III with the following emissions
factors published by EPA (AP-42, Supplement No. 5, December 1975):
0.18 grams per mile for light duty vehicles
0.36 grams per mile for heavy duty gasoline powered vehicles
2.80 grams per mile for heavy duty diesel powered vehicles
Estimates thus obtained of the SO,, emissions from motor vehicle exhaust
for Pittsburgh and the six county areas are shown in Table 2-2.
SO emissions from the operation of railroads and river ves-
sels 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 RPC. The reported annual fuel oil consumption is
13
-------
TABLE 2-1
1972 DAILY VEHICLE MILEAGE FOR THE SOUTHWEST
PENNSYLVANIA INTRASTATE AIR QUALITY
CONTROL REGION
Area
Pittsburgh
Allegheny
County
Butler
County
Armstrong
County
Westmoreland
County
Washington
County
Beaver
County
TOTAL
(All Areas)
EPA Region 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
All
Vehicles
3,110,004
13,351,328
2,187,730
936,865
5,438,732
3,263,532
2,347,737
27,498,927
14
-------
TABLE 2-2
ANNUAL S02 EMISSIONS FROM VEHICLES FOR
SELECTED AREAS OF THE SOUTHWEST
PENNSYLVANIA AQCR
(g/sec/km2)
Area
(km2)
Incorporated Pitts-
burgh (142)
Allegheny County
(1886)
Butler County
(2057)
Armstrong County
(1689)
Westmoreland County
(2652)
Washington County
(2219)
Beaver County
(1139)
Light Duty
Vehicles
0.051
0.015
0.002
0.001
0.004
0.002
0.003
Heavy Duty
Vehicles
0.005
0.001
0.000
0.000
0.000
0.000
0.000
Diesel
Vehicles
0.015
0.003
0.000
0.000
0.001
0.000
0.001
Total
0.071
0.019
0.002
0.002
0.005
0.003
0.005
15
-------
25x10 gallons for railroads and 7x10 gallons for river vessels. The
EPA emission factor for railroad locomotives is 57 pounds of SO,, per
1000 gallons of fuel; the corresponding factor for commercial steamships
is 27 pounds of SO- per 1000 gallons of fuel oil. These data yield esti-
mates of the total SO- emissions from railroads of 712.5 tons per year
and 94.5 tons per year from river vessels. Assuming that the total com-
bined railroad and river vessel SO- emissions of 807.0 tons per year occur
in Allegheny County and Beaver County within a 1-kilometer corridor along
the Ohio, Allegheny and Monongahela Rivers, the area-source SO- emissions
from railroad and river vessels are approximately 4 tons per year per
square kilometer.
2.3.2 Space Heating
SO- emissions from space heating were estimated using the emis-
^ /
sion factor of 0.6 pounds per 10 cubic feet of natural gas published by
EPA (AP-42, February, 1973) and the estimated quantities of fuels being
consumed within the Southwest Pennsylvania AQCR. According to the Brown
Directory (Natural Gas, Operating Gas Companies, 1973), there are seven
natural gas companies serving the region as shown in Table 2-3. Distrib-
uting the natural gas being used for residential and commercial heating
on a population basis results in the SO- emission rates by counties given
in Table 2-4. Heat for the large buildings in downtown Pittsburgh is
provided by central steam plants. This area therefore uses relatively
small amounts of natural gas for space heating.
Coal and oil are not considered to be primary fuels for resi-
dential or commercial space heating. According to the Minerals Industry
Report, 64.5x10 tons of bituminous and lignite coal were consumed in
1972 in the Commonwealth of Pennsylvania. The industrial/commercial S02
emissions inventory for the Southwest Pennsylvania AQCR accounts for
48.0x10 tons of coal per year which is approximately 75 percent of the
above 1972 annual coal consumption in the Commonwealth. The Minerals
Industry report also estimates that 64.1x10 tons of coal were consumed in
16
-------
TABLE 2-3
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
(106 Cubic Feet)
154,327
1,639
67,849
182,258
416,916
17
-------
TABLE 2-4
S0? 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
(106 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)
36.1
2.4
4.7
0.8
3.5
1.8
8.5
2.9
4.8
1.7
67.2
18
-------
1972 by industry and utilities in the Commonwealth of Pennsylvania.
The annual consumption of bituminous and lignite coal during 1972 by
other users in the Commonwealth is therefore 0.4x10 tons. We assume that
75 percent of this coal (0.3x10 tons) is used in the Southwest Pennsyl-
vania Region and that one-half of this amount (0.15x10 tons) is consumed
in hand-fired burners used for space heating. The area included in the
4
Southwest Pennsylvania Region is approximately 1.75x10 square kilometers.
The total annual consumption of coal used for space heating is therefore
about 9 tons per square kilometer . If the average sulfur content of
this coal is 5 percent, the corresponding annual S09 emissions in the
Southwest Pennsylvania Region due to the use of coal for space heating
are about 0.8 tons per square kilometer.
The Bureau of Statistics, Research and Planning of the Common-
3
wealth of Pennsylvania reports that 384,636x10 gallons of fuel oils
were distributed in the Southwest Pennsylvania AQCR during 1974. The
3
emissions inventory accounts for 254,019x10 gallons which leaves
130,344x10 gallons for consumption by residential and commercial users.
The natural gas industry reports that they serve 74,316 customers who do
not use gas for space heating. The assumption that all of these customers
are using fuel oil for space heating in private dwellings leads to an
estimated consumption of 68,464x10 gallons of fuel oil or approximately
half of the unaccounted fuel. It is likely that there are additional
consumers not included in the gas industries report and that some of the
listed and unlisted non-natural gas consumers are classified as commercial
users. Therefore, we have put all of the unaccounted fuel oil into
area-source space heating. Using the emission factor of 71 pounds per
1,000 gallons of fuel oil with 0.5-percent sulfur for commercial burners
published by EPA (AP-42 February, 1972) results in a total emission rate
of 4627 tons per year or 0.39 tons per year per square kilometer.
19
-------
2.3.3 Aircraft Emissions
S02 emissions were estimated for the Greater Pittsburgh Airport
and the Allegheny County Airport using the emissions factors published
by EPA (AP-42, April, 1973) and the aircraft operations by type reported
for FAA operated towers during 1975. Table 2-5 gives the reported
operation from each airport tower. Aircraft operations at the Greater
Pittsburgh Airport were assumed to be allocated as follows:
Air Carrier 10% Heavy Aircraft
60% Long Range Aircraft
30% Medium Range Aircraft
Air Taxi 5% Helicopter
47.5% General Aviation Turboprop
47.5% General Aviation Piston
General Aviation 20% Business Jet
20% General Aviation Turboprop
60% General Aviation Piston
Military 80% Military Transport
20% Military Jet
The estimated S02 emissions rate is 188.5 tons per year for aircraft
operations at the Greater Pittsburgh Airport.
Aircraft operations at the Allegheny County Airport were assumed
to be allocated as follows:
Air Carrier 100% Medium Range Aircraft
20
-------
TABLE 2-5
AIRCRAFT OPERATIONS REPORTED FOR 1975
(Takeoff/Landing Cycles)
Greater Pittsburgh Airport
Air Carrier
Air Taxi
General Aviation
Military
172,331
54,230
46,179
12,425
Allegheny County Airport
Air Carrier
Air Taxi
General Aviation
Military
5
566
168,378
1,497
21
-------
Air Taxi 50% General Aviation Turboprop
50% General Aviation Piston
General Aviation 80% General Aviation Single Engine
Piston
10% General Aviation Turboprop
10% General Aviation Dual Engine
Piston
Military 50% Military Transport
50% Helicopter
The estimated SO., emissions rate is 10.0 tons per year for aircraft
operations at the Allegheny County Airport.
22
-------
SECTION 3
METEOROLOGICAL INPUTS
3.1 METEOROLOGICAL INPUTS REQUIRED BY THE LONG-TERM MODEL (LONGZ)
Meteorological inputs required by the long-term dispersion
model LONGZ described in Appendix A were principally obtained from the
seasonal and annual distributions of wind speed and wind direction,
classified according to the Pasquill stability categories, for the
Greater Pittsburgh Airport. These distributions were developed from the
surface observations using the definitions of Pasquill stability categories
given by Turner (1964), which are based upon solar radiation (insolation)
and wind speed. 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
Figure 3-1 shows the 1965 annual frequency distribution of
wind direction at the Greater Pittsburgh Airport. Inspection of the
figure reveals that most frequent winds at the Greater Pittsburgh Airport
are from the west. The 1965 seasonal wind distributions for the Greater
Pittsburgh airport are presented in Appendix B.I.
In the dispersion 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
u{z} = u{z }- (3-D
R
23
-------
WNW
WSW
6 8 0 12
ENE
ESE
FIGURE 3-1. Annual frequency distribution of wind direction obtained from
the 1965 surface observations at the Greater Pittsburgh Airport,
Percent frequency scale is shown at right center.
24
-------
where
u{z} = wind speed at height z above the surface
} = wind speed at a refere
p = wind-profile exponent
u{z } = wind speed at a reference height z_. above the surface
K K
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 stabiliza-
tion height for the concentration calculations. In the case of low-level
emissions, which are generally treated as building sources, the wind-pro-
file 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. If the assigned source height is below
the reference height, the wind speed at the reference height is used
in the model calculations. 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 3-1. These exponent values
are based on the results obtained by DeMarrais (1959) and Cramer, et al.,
(1972).
Our vertical expansion (a ) curves, which include the effects
Z
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 3-2 lists the values of the standard deviation of the
wind elevation angle 0' corresponding to the Pasquill stability categories
Ij
for rural and urban areas. The rural a' values are based in part on the
L
measurements of Luna and Church (1972) and are consistent with the 0^ values
implicit in the vertical expansion curves presented by Pasquill (1961).
In order to adjust for the effects of surface roughness elements and heat
sources, the 0 ' values for the stability category one step more unstable
E
than the indicated stability category are used in the calculations for
urban areas. (The entire Southwest Pennsylvania AQCR was considered to
25
-------
TABLE 3-1
WIND-PROFILE EXPONENTS USED IN THE ANNUAL
AVERAGE CONCENTRATION CALCULATIONS
Pasquill
Stability
Category
A
B
C
D
E
Wind-Speed Category (m/sec)*
0.0-1.5
0.10
0.10
0.20
0.25
0.30
1.6-3.1
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
8.3-10.8
0.10
0.10
>10.8
0.10
*Measurement height is 6.1 meters above the ground surface.
26
-------
TABLE 3-2
VERTICAL TURBULENT INTENSITIES FOR
RURAL AND URBAN AREAS
Pasquill Stability
Category
A
B
C
D
E
F
Og (radians)
Rural
0.1745
0.1080
0.0735
0.0465
0.0350
0.0235
Urban
0.1745
0.1745
0.1080
0.0735
0.0465
27
-------
be an urban area in the dispersion model calculations described in this
report) 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 stability category F.
The height of the top of the surface mixing layer is defined
as the height at which the vertical intensity of turbulence becomes effec-
tively zero. This condition is fulfilled when the vertical turbulent inten-
sity is of the order of 0.01 radians or smaller. Because 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 repre-
sent the top of the surface mixing layer. However, even with a surface-
based inversion, a shallow mechanical mixing layer will always be present
due to 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 observa-
tions 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 annual and 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 stabil-
ity categories and the median early morning and afternoon mixing depths
were averaged and assigned to the D stability category. Table 3-3 gives
28
-------
TABLE 3-3
MIXING-LAYER DEPTHS IN METERS USED IN
THE ANNUAL CONCENTRATION
CALCULATIONS
Pasquill
Stability
Category
Wind-Speed Category (m/sec)
0.0-1.5
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
(d) Summer
A
B
C
D
E
1730
1730
1730
960
190
A
B
C
D
E
1230
1230
1230
685
140
1730
1730
1730
1025
320
__
1730
1730
1235
740
(d) Fall
1230
1230
1230
740
250
1230
1230
970
710
1730
1295
1230
1190
__
1730
1295
1230
1230
1295
1230
29
-------
the seasonal median mixing depths for the joint combinations of the wind-
speed and stability categories determined for the Pittsburgh area.
The Briggs (1971) plume-rise formulas given in Section A. 2 of
Appendix A require the ambient air temperature as an input. For the sea-
sonal concentration calculations, seasonal average afternoon temperatures
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 cate-
gories. Table 3-4 lists the ambient air temperatures used in the long-
term calculations.
The Briggs (1971) plume-rise formulas given in Section A.2 of
Appendix A also require the vertical potential temperature gradient as an
input. Table 3-5 lists the vertical potential temperature gradients used
in the long-term concentration calculations. The potential temperature
gradients in Table 3-5 were assigned on the basis of the Turner (1964) and
Pasquill (1961) definitions of the Pasquill stability categories, the mea-
surements of Luna and Church (1972), and our own previous experience.
3.2 METEOROLOGICAL INPUTS REQUIRED BY THE SHORT-TERM MODEL (SHORTZ)
Meteorological inputs required by the short-term dispersion model
SHORTZ were principally obtained from the hourly surface observations and
the twice-daily rawindsonde data for the Greater Pittsburgh Airport. The
Pasquill stability categories were determined using the Turner (1964) pro-
cedures. Table 3-6 lists the lateral turbulent intensities for rural and
urban areas by Pasquill stability category which were used in the short-
term model calculations. The actual hourly values of all meteorological
inputs used with the short-term model are described in Section 6.1.
30
-------
TABLE 3-4
AMBIENT AIR TEMPERATURES USED IN THE ANNUAL
AVERAGE CONCENTRATION CALCULATIONS
Pasquill Stability
Category
A
B
C
D
E
Ambient Air Temperature ( 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
298.5
298.5
286.3
282.4
TABLE 3-5
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 (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
0.003
31
-------
TABLE 3-6
LATERAL TURBULENT INTENSITIES FOR
RURAL AND URBAN AREAS
Pasquill
Stability
Category
A
B
C
D
E
F
a! (radians)
Rural
0.2495
0.1544
0.1051
0.0665
0.0501
0.0336
Urban
0.2495
0.2495
0.1544
0.1051
0.0665
32
-------
SECTION 4
S02 DECAY AND BACKGROUND
4.1 ADJUSTMENT OF THE MODEL CONCENTRATION CALCULATIONS FOR S02 DECAY
During atmospheric transport and dispersion, some of the S0~ init-
ially emitted from stacks and other sources is oxidized to form sulfates.
Because the oxidation process reduces the total amount of SO-, it is neces-
sary to make allowance for this depletion in the model calculations to
preclude overestimation of the ground-level S0_ concentrations. This is
accomplished through the use of an exponential decay term as a direct mul-
tiplier of the source strength used in the model calculations. The math-
ematical expression for SO- decay is of the form
Q{t} = Q{t=0) e~Xt (4-1)
where
Q{t} = adjusted source strength at time t
t = x/u
x = downwind distance from the source
u = mean wind speed
Q{t=0} = initial source strength
X = S0~ decay rate
33
-------
In the above expression, the decay or loss of SO- through oxidation is
assumed to be represented by a first order reaction (Alkezwenny and Powell,
1977). The decay of S02 is also expressed in terms of the S02 half-life
or the time required to reduce the amount of SC^ initially contained in a
plume by 50 percent. If the decay rate X is in units of percent per hour,
the S02 half-life is equal to (100 &n 2)/X hours.
Measurements of the rate at which S0? is transformed to sulfates
in plumes from coal-fired power plants show an extremely wide range of
values depending principally on the time of day, season of the year, geo-
graphical location and level of pollutant concentrations (Sidebottom, 1972;
Tesche, et al., 1976; Lusis and Wiebe, 1976; Forest and Newman, 1976;
Wilson, et^ _al. , 1977; Gillani, £t al., 1978). From the data presented in
the above references, we believe that the values of the decay coefficient
X most likely to apply to the Southwest Pennsylvania AQCR range from 1 to
6 percent per hour with an average value of about 3.5 percent per hour.
The SCL half-life values corresponding to the above range in X are 11 to
69 hours with an average value of about 20 hours. An average S0~ half-
life of 20 hours was used for all of the calculations of ground-level SO-
concentrations in this study. This half-life is consistent with the decay
rate used by TRC, Inc. in a study of the impact of S09 emissions in Allegheny
County, Pennsylvania recently conducted for the Allegheny County Bureau of
Air Pollution Control.
Table 4-1 shows the annual average ground-level SO- concentra-
tions calculated at selected monitor sites using S0~ half-life values
of 3, 12, 20 and 30 hours. These calculations indicate that a change
in the S09 half-life from 12 to 30 hours produces a change of about 10
percent in the annual average concentrations. The observed annual average
S0~ concentration for each of the monitoring sites is also given in
Table 4-1. With the exception of the Midland monitor site, the observed
concentrations tend to support the use of a relatively long SO- half-life.
34
-------
TABLE 4-1
THE EFFECTS OF S02 HALF-LIFE ON THE CALCULATED ANNUAL-AVERAGE
GROUND-LEVEL SO2 CONCENTRATION AT SELECTED
MONITOR LOCATIONS
Monitor
Location
Beaver Falls
New Castle
Midland
Hazelwood
Logans Ferry
Downtown
o
S0? Concentration (yg/m )
S0_ Half-Life (Hours)
£
3
24.5
13.1
87.4
124.2
36.9
48.2
12
39.5
25.3
107.1
146.7
53.5
70.5
20
43.1
29.1
112.5
149.7
57.5
75.5
30
45.2
31.4
114.2
154.6
60.1
78.3
Observed
Concentration
47.2
39.3
78.6
141.5
73.0
73.0
35
-------
4.2 ESTIMATES OF S02 BACKGROUND CONCENTRATION
The S0? background concentration is defined as the concentration
contributed by all SO,, sources not included in the emissions inventory
used for the dispersion-model calculations. In this study, many of the
difficulties associated with obtaining accurate background estimates
were eliminated by the inclusion of the emissions inventory of all S02
sources located within the Southwest Pennsylvania Air Quality Control
Region with annual emissions of 10 or more tons, as well as the major
SO sources in the West Virginia Panhandle and along the eastern border
of Ohio. To the best of our knowledge, the emissions inventory used in
the model calculations includes all the large S0« sources located within
30 kilometers of the three principal grid areas (New Castle, Beaver and
Monessen) for which model calculations were made.
The SOp sources not included in the emissions inventory, and
thus the sources most likely to contribute to the background concentra-
tion in the study area, are:
Emissions from classical area sources located within
the study area
Emissions from major point sources located outside the
area included in the emission inventory
The area-source emissions inventory discussed in Section 2 was used with
the long-term dispersion model for area sources in Appendix A to calculate
the impact of area sources on ambient S0~ air quality in the Beaver Valley.
Emissions from the following specific types of area sources were included
in the calculations:
36
-------
Vehicle traffic in the Beaver Valley area which was
represented by 18 area sources
Home-heating in the Baden area
Aircraft operations at the Greater Pittsburgh Airport
Rail and barge traffic in a 1-kilometer corridor along
the Ohio River and extending along the Beaver River to
Beaver Falls which was represented by 131 area sources
Figure 4-1 shows the S0~ concentration isopleths resulting from
the long-term area source calculations. The maximum calculated annual
average concentration for the combined effects of emissions from home
heating, rail traffic and barge traffic is 2.7 micrograms per cubic meter
which occurs in Baden near the J & L Aliquippa Plant. Figure 4-2 shows
the isopleth pattern of SO,, concentration calculated for the single home-
heating area source in Baden. The dimensions of the Baden area source
are 1.6-x 1.6-kilometers and the SQ~ emission rate for home heating is 0.1
grams per second which is equivalent to 1.36 tons of SO,., per square kilom-
eter per year. The air quality impact of all the above area source emis-
sions calculated for the Beaver Valley by means of the short-term disper-
sion model for area sources is very similar to that calculated by the
corresponding long-term dispersion model. Specifically, the calculated
short-term concentration contributions from area sources are approximately
equal to the calculated long-term concentration contributions. The
only significant difference is a displacement of the isopleth pattern
reflecting the specific wind-direction distributions used in the short-
term calculations. We conclude that the maximum contribution of area-
source emissions to the short-term and annual SO background concentrations
is less than 5 micrograms per cubic meter. The results discussed in the
following sections of this report do not include any estimates or calcula-
tions of the contribution of area sources to the total S0? concentrations.
37
-------
4516
4514 -
4512 -
4510 -
4508 -
4506 -
4504 -
4502 -
4500 -
4498 -
4496 -
4494 -
4492 -
4490 -
555 557 559 561
563
565
567
FIGURE 4-1. Isopleths of annual average SCU ground-level concentration
in micrograms per cubic meter calculated for the Beaver
Valley using the long-term area source model. The numbers
2, 3 and 4 respectively refer to the St. Joe Minerals,
J & L Aliquippa and Duquesne Light, Phillip Plants.
38
-------
4514
4512
4510
4508
4506
4504
4502
4500
4498
4496
4494
4492
4490
555
557
559
561
563
565
567
FIGURE 4-2.
Isopleths of annual average SC>2 ground-level concentration
in micrograms per cubic meter calculated for the single
home-heating area source in Baden. The numbers 2, 3 and 4
respectively refer to the St. Joe Minerals, J & L Aliquippa
and Duquesne Light, Phillip Plants.
39
-------
The potential contribution to the background SCL concentration
from the atmospheric transport or advection into the study area from major
point sources located beyond the area included in the emissions inventory
was estimated as follows. According to the results of the annual average
concentration calculations described in Section 5 below, the combined
contribution of the Ohio and West Virginia sources in the New Castle
area, which is approximately 50 kilometers from these sources, is about 15
micrograms per cubic meter. Similarly, the combined contribution of these
sources in the Beaver area, which is at a distance of approximately 20
kilometers, is about 27 micrograms per cubic meter. The contribution
of the Ohio and West Virginia sources is thus approximately inversely re-
lated to the distance from the sources to the receptors or grid points.
If we assume that a similar number of other major S0? sources are located
at a distance of 50 to 100 kilometers, it follows that the maximum
contribution of these sources to the annual average background concentra-
tion in the study area would be about 10 micrograms per cubic meter.
However, the number of major S0« sources not included in the emissions
inventory used in this study which are located within 100 kilometers of
the New Castle and Beaver areas is very small. Therefore, we believe
that the annual average S0~ background concentration contributed by dis-
tant major S0~ sources not included in the inventory is insiginificant
compared to the contributions of the sources contained in the inventory.
The same reasoning leads to the conclusion that the short-term background
concentration from distant sources not included in the emissions inventory
is also insignificant.
40
-------
SECTION 5
LONG-TERM CALCULATIONS
5.1 CALCULATION PROCEDURES
Figure 5-1 is a map of the Southwest Pennsylvania AQCR and
adjacent areas in Ohio and West Virignia showing the locations of the
various calculation grids as well as the large SO., sources and source
complexes used in the long-term model calculations. The SO,-, monitor
sites used in the long-term model validation calculations are also
shown. The dashed lines in the figures define the limits of the gross
calculation grid used in the early part of the study to determine the
areas within the Beaver Valley and Monongahela Valley Air Basins in
which high ground-level S07 concentrations would most likely occur. The
overall dimensions of the gross calculation grid are 100 x 140 kilometers.
A regular 5-kilometer spacing between grid points was used within a
central 60 x 100 kilometer section of the gross grid which included most
of the major S0~ sources in the Southwest Pennsylvania AQCR. A 10-
kilometer spacing of grid points was used in the portion of the gross
grid outside this central section.
The five shaded areas in Figure 5-1 were selected for detailed
dispersion-model calculations with a 1-kilometer spacing of grid points,
using the annual 1980 compliance emissions inventory. The selection of
the five smaller calculation grids was based on the results of model
calculations made with the LONGZ computer program using the annual 1980
compliance emissions inventory in Appendix C.I with the meteorological
inputs in Appendix B.I developed from the 1965 hourly observations made
at the Greater Pittsburgh Airport. The Beaver calculation grid (Number
1 in Figure 5-1) measures 17 x 27 kilometers and is approximately centered
on the confluence of the Beaver and Ohio Rivers. The New Castle calcula-
tion grid (Number 2 in Figure 5-1) measures 7 x 10 kilometers and encloses
41
-------
FIGURE 5-1. Map of the Southwest Pennsylvania AQCR and surrounding area.
Dashed lines show the gross grid and shaded areas show the
five smaller grids used for dispersion-model calculations.
Major SC>2 sources and source complexes are indicated by the
numbers 1 through 33 and SC>2 monitor locations used for model
validation are indicated by the letters A through M.
42
-------
the heavily populated area around the city of New Castle. The two
western border calculation grids (Number 3 and Number 4 in Figure 5-1)
are each 4 x 17 kilometers and are located along sections of the border
of the Southwest Pennsylvania AQCR likely to be heavily impacted by the
large SO sources across the border in eastern Ohio and in West Virginia.
The Monessen calculation grid (Number 5 in Figure 5-1) measures 21 x 27
kilometers and is approximately centered on the Monengahela River Valley
near Charleroi. The major SO,., sources and the S0~ monitor locations in
Figure 5-1 are identified in Tables 5-1 and 5-2.
The meteorological inputs for the long-term calculations of
ground-level S0? concentrations were developed from observations made
at the Greater Pittsburgh Airport (See Section 3). We used annual and
seasonal statistical summaries of the wind speed and direction observations
for the years 1964 and 1973 through 1977 in an attempt to define a worst-
case year Gi.-^., a year in which the meteorology would lead to the highest
calculated annual average SO,, concentrations). As might be expected, the
definition of a worst-case year is very complicated because the calculated
maximum S0~ ground-level concentrations are critically dependent on both
the source-receptor geometry and the meteorology. Consequently, the worst-
case meteorology varies with receptor location. Because of these difficulties,
we decided to use 1965 as the worst-case meteorological year on the basis
of a. study by Rubin (1974). In this study, Rubin defined 1965 as the worst-
case meteorological year for S0« concentrations in Allegheny County
from the results of annual S0~ model calculations made using a fixed set
of point-source emissions data for Allegheny County with various sets of
meteorological inputs developed from the annual wind distributions for
the Greater Pittsburgh Airport.
The annual 1980 compliance emissions Inventory given in Appendix
C.I was used with the LONGZ computer program containing the long-term
dispersion model described in Appendix A to calculate annual ground-level
S02 concentrations in the five, grid areas shown in Figure 5-1.
The results of these calculations are described below and in Section 8.
43
-------
TABLE 5-1
LIST OF MAJOR S02 SOURCES IN FIGURE 5-1
Source
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
Source Name
Penn Power, West Pittsburgh
St. Joe Minerals
J & L Aliquippa
Duquesne Light, Phillips
Duquesne Light, Bruno t Is.
Duquesne Light, Cheswick
J & L Pittsburgh
USS Homestead
USS Edgar Thomson
USS Duquesne
USS National
USS Irvin
USS Clairton
Duquesne Light, Elrama
Penn Power, Mitchell
Hatfield Power
Armstrong Power
Keystone Power
Homer City Power
Seward Power
Conemaugh Power
Ohio Ediwon, Sammis
Ohio Power, Toronto
National Steel, Wierton Div.
Wheeling-Pittsburgh (North)
Wheeling-Pittsburgh (South)
Cardinal Power
Ohio Edison, Burger
Rammer Power
Mitchell Power
PPG
Penn Power, Mansfield
Wheeling-Pittsburgh, Monessen
UTM Coordinates (m)
X
553,105
556,065
564,400
565,260
580,680
602,330
589,000
592,100
597,260
598,324
597,000
593,180
595,500
592,000
587,340
591,570
628,979
640,141
652,756
666,882
664,582
531,700
533,500
534,300
532,500
534,600
530,000
520,500
515,320
515,800
512,600
549,049
594,170
Y
4,531,767
4,502,301
4,497,500
4,491,020
4,479,680
4,487,800
4,473,900
4,473,100
4,471,685
4,470,025
4,467,500
4,465,700
4,461,500
4,456,200
4,452,810
4,412,040
4,531,996
4,502,155
4,486,147
4,474,602
4,471,929
4,485,500
4,481,800
4,474,400
4,466,700
4,463,300
4,455,800
4,417,500
4,410,320
4,408,670
4,399,600
4,498,067
4,446,350
44
-------
TABLE 5-2
LIST OF S02 MONITOR SITES IN FIGURE 5-1
Site
Symbol
A
B
C
D
E
F
G
H
I
J
K
L
M
Site Name
(Operator)
New Castle (DER)
Beaver Falls (DER)
Fairview (Perm Power)
Route 68 (Perm Power)
West Beaver (Penn Power)
Midland (DER)
Baden (DER)
Logans Ferry (BAPC)
Downtown (BAPC)
Hazelwood (BAPC)
North Braddock (BAPC)
Liberty Boro (BAPC)
Glassport (BAPC)
UTM Coordinates (m)
X
554,830
557,750
544,820
550,720
548,950
546,330
565,090
605,154
585,150
589,762
596,680
596,210
594,190
Y
4,537,240
4,510,785
4,504,390
4,500,640
4,502,000
4,498,340
4,498,380
4,489,115
4,476,600
4,473,952
4,472,835
4,464,150
4,463,580
Site
Elevation
(m MSL)
257
220
390
238
375
249
230
268
256
284
275
340
234
45
-------
5.2 LONG-TERM CALCULATION RESULTS
The maximum annual average SO,, concentration calculated for
the Beaver grid of 76.8 micrograms per cubic meter occurs at a grid point on
the western boundary of the Beaver grid with UTM X and Y coordinates of 554
and 4,503 kilometers respectively. As shown in Figure 5-2, which presents the
annual average S0~ isopleth pattern calculated for the Beaver grid, this
maximum is located approximately 2.5 kilometers west-northwest of the St.
Joe Minerals Plant, which is indicated by the number 2 in the figure. Addi-
tional calculations made at grid points located 1 kilometer southwest, west
and northwest of the point of maximum concentration on the western
boundary show annual average SO,, concentrations which are less than the
maximum calculated on the boundary.
The contributions of individual sources and source complexes in
the 1980 compliance emissions inventory to the maximum annual average
concentration calculated for the Beaver grid are given in Table 5-3.
Sources in the Beaver Valley contributed 38.5 micrograms per cubic meter
or about 50 percent of the calculated maximum concentration. Of the 38.5
micrograms per cubic meter contributed by the Beaver Valley sources, 27.7
micrograms per cubic is contributed by St. Joe Minerals. The combined
Eastern Ohio and West Virginia sources contribute 26.7 micrograms per
cubic meter or about 35 percent of the total maximum annual average S0~
concentration calculated for the Beaver grid. Sources located in Allegheny
County contribute 7.2 micrograms per cubic meter, other Pennsylvania
sources contribute 3.8 micrograms per cubic meter and sources located in
the Monongahela River Valley contribute 0.6 micrograms per cubic meter.
These source contributions account for the remaining 15 percent of the
total calculated maximum concentration. Additional calculations were made
using a 100-meter spacing of the grid points within the small Baden grid
shown in Figure 5-2. The maximum concentration shown by these calculations
of 72.2 micrograms per cubic meter occurs at the grid point located east of
the J & L Aliquippa Plant with UTM X and Y coordinates of 565.6 and 4,497.8
kilometers, respectively.
46
-------
4516
4514
4512
4510
4508
4506
4504
4502
4500
4498
4496
4494
4492
4490
555
557
559
561
563
565 567
FIGURE 5-2. Isopleths of annual ground-level S02 concentration in yg/m
calculated for the Beaver grid using the 1980 compliance
emissions inventory. The numbers 2, 3 and 4 respectively show
the locations of the St. Joe Minerals, J & L Aliquippa and
Duquesne Light, Phillips Sources. The asterisk symbol incicates
the point of maximum concentration and the dashed lines show
the small Baden calculation grid.
47
-------
TABLE 5-3
CONTRIBUTION OF INDIVIDUAL SOURCES AND SOURCE COMPLEXES TO THE MAXIMUM
ANNUAL AVERAGE SO2 CONCENTRATION CALCULATED FOR THE BEAVER
GRID USING THE 1980 COMPLIANCE EMISSIONS INVENTORY
Source
Beaver Valley Sources
J & L Aliquipa
Medusa Cement
ARCO
B & W Wallace Run
B & W Tubular Products
Shenango China
Ashland Oil
Westinghouse Electric
Penn Power ,
West Pittsburgh
Penn Power, Mansfield
St . Joe Minerals
Bessemer Cement
Total
Allegheny County
Sources
Monongahela Valley
Sources
Other Pennsylvania
Sources
Eastern Ohio Sources
West Virginia Sources
Total
Annual Average
S0_ Concentration*
(yg/m3)
1.5
0.1
4.7
0.1
0.1
**
0.1
**
0.3
4.1
27.4
0.1
38.5
7.2
0.6
3.8
16.7
10.0
76.8
* Occurs at a grid point with the UTM coordinates X=554 kilometers and
and Y=4,503 kilometers.
3
** Less than 0.05 yg/m .
48
-------
Figure 5-3 shows the annual average SCL isopleth pattern calcula-
ted for the New Castle grid using the 1980 compliance emissions inventory.
The calculated maximum annual average SO- concentration resulting from the
emissions of all sources combined is 37.3 micrograms per cubic meter and
occurs at a point near the southern boundary of the New Castle Grid with the
UTM X and Y coordinates of 558 and 4,525 kilometers, respectively. Contributions
of the individual sources and source complexes to the maximum annual average
SCL concentration calculated for the New Castle grid are given in Table
5-4. The calculated contribution of the Beaver Valley sources to the
maximum concentration is 14.0 micrograms per cubic meter or about 38 percent
of the total. Of the 14.0 micrograms per cubic meter contributed by the
Beaver Valley sources, the Medusa Cement Plant contributed 9.3 micrograms
per cubic meter. The combined sources in Eastern Ohio and West Virginia
contributed 14.4 micrograms per cubic meter to the maximum or about 39
percent of the total.
The contribution of the Penn Power, West Pittsburgh Plant to the
annual average S0~ concentrations in the New Castle area was calculated
by using an operating level of 75 percent of capacity and a GEP (Good
Engineering Practive) stack height of 145 meters in place of the actual
stack height of 229 meters. The maximum annual average SO,, concentration
produced by the West Pittsburgh Plant occurs at a grid point with the UTM X
and Y coordinates of 556 and 4,532 kilometers, respectively. The calculated
total annual average S0>2 concentration at this point produced by all sources
combined is 31.2 micrograms per cubic meter. Of this, 5.8 micrograms per
cubic meter or 18.6 percent is contributed by the West Pittsburgh Plant
emissions; 4.4 micrograms per cubic meter or 14.1 percent, is contributed
by the sources in the Beaver Valley; 7.7 micrograms per cubic meter or 24.7
percent is due to the eastern Ohio sources; 5.2 micrograms per cubic meter
or 16.7 percent is contributed by the West Virginia sources; 4.2 micrograms
per cubic meter or 13.5 percent is due to the other Pennsylvania sources;
and 3.5 micrograms per cubic meter or 11.2 percent is contributed by
sources located in Allegheny County. Additional concentration calculations
49
-------
4541
4539 -
4537 -
4535 J
4533 -
4531
4529 -
4527
4525 -
IWEST PITTSBURGH
IGRID
553
555
557
559
FIGURE 5-3. Isopleths of annual ground-level SC^ concentration in yg/m
calculated for the New Castle grid using the 1980 compliance
emissions inventory. The number 1 refers to the Penn Power,
West Pittsburgh Plant. The asterisk symbol indicates the
point of maximum concentration and the dashed lines show the
small West Pittsburgh calculation grid.
50
-------
TABLE 5-4
CONTRIBUTIONS OF INDIVIDUAL SOURCES AND SOURCE COMPLEXES TO THE
MAXIMUM ANNUAL AVERAGE S02 CONCENTRATION CALCULATED FOR THE NEW
CASTLE GRID USING THE 1980 COMPLIANCE EMISSIONS INVENTORY
Source
Beaver Valley Sources
J & L Aliquipa
Medusa Cement
ARCO
B & W Wallace Run
B & W Tubular Products
Shenango China
Ashland Oil
Westinghouse Electric
Penn Power,
West Pittsburgh
Penn Power,
St. Joe Minerals
Bessemer Cement
Total
Allegheny County
Sources
Monongahela Valley
Sources
Other Pennsylvania
Sources
Eastern Ohio Sources
West Virginia Sources
Total
Annual Average
SO Concentration*
1 (ug/m3)
0.6
9.3
0.3
0.1
0.1
A*
0.1
**
0.9
0.6
1.7
0.3
14.0
4.3
0.3
4.3
8.8
5.6
37.3
* Occurs at a grid point with the UTM coordinates X=558 kilometers and
Y=4,525 kilometers.
3
** Less than 0.05 Ug/m .
51
-------
were also made at grid points spaced at 250 meters within the small West
Pittsburgh grid shown in Figure 5-3. The location and magnitude of the
maximum concentration were unchanged from the values given above.
The maximum annual average S0~ concentration calculated for the
Monessen grid is 86.2 micrograms per cubic meter at a grid point with the
HTM X and Y coordinates of 594 and 4,447 kilometers, respectively. This- point
which is located approximately 1 kilometer north of the Wheeling Pittsburgh
Steel Plant at Monessen, is an isolated point of high concentration with the
calculated concentrations at the surrounding grid points ranging from
40.0 to 61.5 micrograms per cubic meter. Figure 5-4 shows the annual
average SCL isopleth pattern calculated for the Monessen grid. The
contributions from individual sources and source complexes to the maximum
annual average concentration calculated for the Monessen grid are given
in Table 5-5. Sources in the Monongahela Valley contribute 53.6 micro-
grams per cubic meter or approximately 62 percent of the maximum.
Allegheny County sources contribute 6.4 micrograms per cubic meter (7.6%).
Other Pennsylvania sources contribute 9.0 micrograms per cubic meter
(10.4%). Ohio sources 9.9 micrograms per cubic meter (11.5%) and West
Virginia Sources contribute 7.1 micrograms per cubic meter (8.2%).
Additional concentration calculations were made using a grid spacing of
500 meters within the detail Monessen grid shown in Figure 5-4. The
location and magnitude of the maximum concentration were unchanged for
the values given above.
The results of the long-term model calculations for the two
western border grids are presented in Section 8.1.
52
-------
4454 -
4452 -
4450
4448 -
4446 -
4444 -
4442 -
4440 -
4438 -
4436 -
4434 -
4432 -
4430 -
DETAIL MONESSEN |
GRID!
585
587 589
593 595 597 599 601 603
FIGURE 5-4. Isopleths of annual ground-level SC>2 concentrations in yg/m
calculated for the Monessen grid using the 1980 compliance
emissions inventory. The number 15 identifies the Penn Power,
Mitchell Plant and number 33 identifies the Wheeling-Pittsburgh,
Monessen Plant. The asterisk symbol indicates the point of
maximum concentration and the dashed lines show the detail
Monessen calculation grid.
53
-------
TABLE 5-5
CONTRIBUTIONS OF INDIVIDUAL SOURCES AND SOURCE COMPLES TO THE MAXIMUM
ANNUAL AVERAGE S02 CONCENTRATION CALCULATED FOR THE MONESSEN
GRID USING THE 1980 COMPLIANCE EMISSIONS INVENTORY
Source
Monongahela Valley Sources
ARMCO
Mitchell
Elrama
Wheeling Pittsburgh, Monessen
Wheeling Pittsburgh, Allenport
Allied Chemical
Total
Allegheny County Sources
Beaver Valley Sources
Other Pennsylvania
Sources
Eastern Ohio Sources
West Virginia Sources
Total
Annual Average
S02 Concentration*
(yg/m3)
**
1.0
1.9
50.3
0.2
0.2
53.8
6.4
0.8
8.2
9.9
7.1
86.2
Occurs at a grid point with the UTM coordinates X=594 kilometers and
Y=4,747 kilometers.
** Less than 0.05
m
54
-------
SECTION 6
SHORT-TERM MODEL CALCULATIONS
6.1 CALCULATION PROCEDURES
The short-term 1980 compliance emissions inventory given in
Appendix C.2 was used with the SHORTZ computer program containing the
short-term dispersion model described in Appendix A and worst-case 3-
hour and 24-hour meteorological inputs, developed from hourly meteoro-
logical observations made at the Greater Pittsburgh Airport, to calculate
short-term concentrations within the five calculation grids described in
Section 5.1. In addition to the regular 1-kilometer grid point spacing,
a 200-meter spacing was used for the New Castle grid in a small area
around the Penn Power, West Pittsburgh Plant and a 500-meter spacing was
used for the Monessen grid in a small area of high terrain north of the
Monessen Plant of Wheeling-Pittsburgh Steel. The locations of high
terrain elevations and S0« monitor sites were also included as discrete
points in the short-term calculations. To reduce the number of requisite
computer calculations, portions of the calculation grids which the
meteorological observations indicated would be clearly unaffected by the
transport and dispersion of relevant source emissions were excluded from
the executions of the SHORTZ computer program.
The following rationale was used in selecting the meteorological
conditions most likely to cause the highest short-term ground-level con-
centrations in the New Castle, Beaver and Western Border calculation
grids. As shown in Figure 5-1, there are a very large number of major
S0_ sources located in the Ohio River Valley to the west and south-
southwest of these calculation grids. According to short-term 1980
compliance emissions inventory in Appendix C.2, approximately 51 percent
of the total emissions for all S0~ sources combined is from sources
located in Ohio and West Virginia. Emissions from sources located in the
Beaver Valley account for about 4 percent of the total emissions from all
55
-------
sources; emissions from sources in Allegheny County and the Monongahela
River Valley account for about 9 percent of the total emissions and the
remaining 36 percent of the total emissions are contributed by other
Pennsylvania sources. We analyzed the hourly surface observations from
the Greater Pittsburgh Airport for the five-year period from 1973 through
1977, using the PRSIST computer program, to determine the occurrence frequency
and time duration of persistent wind directions required to transport emis-
sions from the sources in Ohio, West Virginia, Allegheny County and the
Monongahela River Valley to the New Castle and Beaver calculation grids.
The results of the PRSIST analysis showed that there were a number of
cases during the five-year period of record of 24-hour or longer persistence
of wind directions within the 60-degree angular sector from 180 to 240
degrees. These are the wind directions required to transport emissions
from the major SCL sources in Ohio and West Virginia to the New Castle
and Beaver calculation grids, as well as the two Western Border grids
shown in Figure 5-1. On the other hand, the PRSIST analysis revealed
that wind directions within the sector required to transport emissions
from sources located in Allegheny County and the Monongahela River
Valley to the New Castle and Beaver calculation grids persisted for only
a few hours. We therefore concluded from the results of the PRSIST
analysis that the worst-case short-term meteorological conditions for
the New Castle, Beaver and Western Border calculation grids were most
likely to occur with persistent wind directions from the south-southwest
and southwest. Model calculations were made using the short-term 1980
compliance emissions inventory with the SHORTZ computer program and the
Greater Pittsburgh Airport hourly meteorological observations and mixing
heights for the cases of maximum persistence of these wind directions
obtained from the PRSIST analysis. These calculations showed maximum
hourly plume-centerline concentrations of approximately 800 micrograms
per cubic meter at the western edge of the New Castle grid when the
mixing heights were between 190 and 250 meters. However, according to
the results of the PRSIST analysis, these mixing heights did not persist
in any case for as long as 10 hours, which is the minimum time duration
required to yield calculated 24-hour average concentrations equal to or
56
-------
greater than 24-hour SCL NAAQS. The maximum hourly plume-centerline
concentrations calculated for the western border of the Beaver grid for
the selected wind-persistence cases were approximately 1000 micrograms
per cubic meter under the same meteorological conditions as those speci-
fied for the New Castle grid.
Because of the relatively long transport distances (30 to 60
kilometers) from some of the sources in Ohio and West Virginia to the
New Castle and Beaver calculation grids as well as possible terrain
channeling and trapping effects, we reviewed the assumption made in the
model calculations that meteorological observations made at the Greater
Pittsburgh Airport could be used to estimate plume trajectories,
wind speeds and mixing heights over these distances. Terrain
elevations in the Ohio River Valley are approximately 200 meters (650
feet) above mean sea level (MSL). The average elevation of the high
terrain at the sides of the valley and beyond is about 300 meters (1000
feet) MSL with a few isolated points that are 400 meters (1300 feet)
MSL. The stack heights associated with the major SO,., sources located in
the Ohio River Valley are generally greater than 150 meters. Assuming
that the plume rise is approximately equivalent to the stack height, the
plume stabilization heights are thus about 500 meters MSL which is well
above the highest terrain elevations. Therefore, we conclude it is
unlikely that the plumes would be trapped within the river valley during
periods with an established wind flow with minimum speeds of a few
meters per second. Under these conditions, the plumes would be well
above the local terrain features and would thus not be subject either to
terrain channeling or trapping. Emissions from low-level sources will
generally be channeled by the terrain when the mixing heights are at or
below the terrain heights. When the mixing heights exceed the terrain
heights, these emissions will be transported in the same general direction
as the tall stack emissions.
There is a question whether the hourly Greater Pittsburgh
Airport meteorological observations are representative of the concurrent
57
-------
meteorological conditions in the.Ohio River Valley and in the western
part of the Southwest Pennsylvania AQCR. It has been our experience
that the measurements at the Greater Pittsburgh Airport are most likely
to be representative of the meteorological conditions over this area
when the measured hourly wind directions persist within a 30- or 60-
degree sector for periods of 6 hours or longer. These persistent wind
directions occur only when the large scale circulation pattern is well
established with mean surface speeds greater than 3 meters per second.
Using the results of the PRSIST analysis of the 1973-1977 Greater Pittsburgh
Airport meteorological data in combination with dispersion-model calcula-
tions, we selected 28 August 1976 as the case most likely to produce
maximum short-term S0_ concentrations in the New Castle and Beaver
areas. Table ,6-1 lists the hourly meteorological input parameters
used with short-term dispersion model computer program SHORTZ to calculate
S0~ ground-level concentrations in the New Castle and Beaver grids pro-
duced by the short-term 1980 compliance emissions inventory (see Appendix
C.2). In these calculations, we assumed an S09 half-life of 20 hours
(see Section 4.1).
We performed dispersion-model calculations using the gross grid
and the hourly meteorological observations from the Greater Pittsburgh
Airport for 6 April 1975 and 16 August 1976 when persistent winds were
observed from the north and west, respectively. The purpose of these
calculations was to determine if these wind directions were likely to
result in maximum short-term concentrations that exceeded those calcu-
lated for 28 August 1976. The calculated maximum concentrations for both
of the cases were significantly lower than those calculated for 28
August 1976.
6.2 SHORT-TERM CALCULATION RESULTS
Figure 6-1 shows the calculated 24-hour average SO- isopleth
pattern for the Beaver grid. The maximum 24-hour average SO. concentration
58
-------
so
W
(J
H
tJ U
Pasquill
Stability
Category
^- \
.. -.H __.
O n)
}-i
^^
^\
- W TJ
t> cfl
M
s*^
PM
N 5
O t*i
<1 0
CD ^
<-i
S- -v
rt o
H ^
/ N
*63
T3 T} O
C cu a)
H 0) CO
>-? Q.->^
w 6
Wind
Direction
(deg)
^H
o !3
«d
w w w w
sO so so so
SO SO so SO
O O O O
m m in m
sO so so sO
o*\ CJN C^ CJ\
csl OJ CM CM
o o o m
. CM
m in so i i
CM CM CO -3"
r-- co m o^
CM CM CM CM
o m o o
CO O\ so sO
r~- r~ oo CXD
co co r^~ . r~- r^
o o o o
CO CO CO CO
i i i 1 r I r 1
O O 0 O
o o o o
oo so so m
o^ o> o^» o>
CM CM CM CM
r-x O> O>
CM CM CM CM
O O O O
O 0 0 O
r-~ r~ r>. r--
CM CM O O
i I i t so sO
;) vf co co
co >tf i i m
i i ,-H r-« o
CM CM CM ro
O i-H CM CO
CM CM CM CM
59
-------
4516
4514 -
4512 -
4510 -
4508 -
4506 -
4504 -
4502 -
4500 -
4498 -
4496 -
4494 -
4492 -
4490 -
555
557 559
563
565 567
FIGURE 6-1. Isopleths of 24-hour average SC>2 ground-level concentrations
in Ug/m^ calculated for the Beaver grid using the short-term
1980 Compliance Emissions Inventory. The numbers 2, 3 and
4 respectively refer to the St. Joe Minerals, J & L Aliquippa
and Duquesne, Phillips sources. The asterisk symbol indicates
the point of maximum concentration.
60
-------
calculated for the Beaver grid, using the 1980 short-term compliance
emissions inventory, is 251.6 micrograms per cubic meter, which is
approximately 70 percent of the 24-hour NAAOS of 365 micrograms per
cubic meter. This maximum occurs at the grid point on the western edge
of the Beaver grid with the UTM X and Y coordinates of 554 and 4,503
kilometers, respectively.
The contributions of individual sources and source complexes
to the maximum concentration calculated on the Beaver grid are given in
Table 6-2. The sources in the Beaver Valley contributed 104 micrograms
per cubic meter or about 41 percent of the calculated maximum concentra-
tion. Of the 104 micrograms per cubic meter contributed by the Beaver
Valley sources, 89 micrograms per cubic meter are due to the Penn Power,
Mansfield Plant and 15 micrograms per cubic meter are due to the ARCO
Plant. The Ohio sources and the West Virginia sources respectively
contributed 127.5 and 20.1 micrograms per cubic meter (51 and 8 percent)
to the maximum 24-hour average SO,., concentration calculated for the
Beaver grid.
The maximum 3-hour average SO,, concentration calculated for
the Beaver grid, using the short-term 1980 compliance emissions inventory,
is 835.5 micrograms per cubic meter, which is 64 percent of the NAAQS of
1300 micrograms per cubic meter. The Beaver Valley sources contributed
289 micrograms per cubic meter or about 35 percent of the calculated
maximum concentration. The eastern Ohio sources and the West Virginia
sources respectively contributed 468.5 and 78 micrograms per cubic meter
(56 and 9 percent) to the calculated maximum 3-hour average concentration
which occurred at the same grid point as the calculated maximum 24-hour
concentration.
The maximum 24-hour average SO,, concentration calculated for
the New Castle grid, using the short-term 1980 compliance emissions
61
-------
TABLE 6-2
CONTRIBUTIONS FROM INDIVIDUAL SOURCES AND SOURCE COMPLEXES TO THE
MAXIMUM 24-HOUR AVERAGE S0£ CONCENTRATION CALCULATED FOR THE
BEAVER GRID USING THE 1980 SHORT-TERM COMPLIANCE EMISSIONS
INVENTORY
Source
Beaver Valley Sources :
J & L Aliquipa
Medusa Cement
ARCO
B & W Wallace Run
B & W Tubular Products
Shenango China
Ashland Oil
Westinghouse Electric
Penn Power ,
West Pittsburgh
Penn Power, Mansfield
St. Joe Minerals
Bessemer Cement
Total
Eastern Ohio Sources
West Virginia Sources
Total
24-Hour Average S02 Concentration*
**
**
15.0
**
**
**
**
**
89.0
**
**
**
104.0
127.5
20.1
251.6
* Occurs at a grid point with the UTM coordinates X=554 kilometers and
Y=4,503 kilometers.
** Less than 0.05 micrograms per cubic meter.
62
-------
inventory, is 115.8 micrograms per cubic meter which is about 32 percent
of the NAAQS of 365 micrograms per cubic meter. This occurs at a grid
point at the northwest corner of the grid with the UTM X and Y coordinates
of 555.0 and 4,533.6 kilometers, respectively.
The contributions of individual sources and source complexes
to the maximum concentration calculated on the New Castle grid are given
in Table 6-3. The combined sources in West Virginia and eastern Ohio
contributed 69.4 micrograms per cubic meter or about 60 percent of the
calculated maximum concentration. With respect to the Beaver Valley
sources, the Penn Power, West Pittsburgh Plant contributed 46.3 micrograms
per cubic meter (approximately 40 percent of the total concentration)
and the other Beaver Valley sources contributed less than 0.01 micrograms
per cubic meter.
The maximum 3-hour SO.., concentration calculated for the New
Castle grid is 337.4 micrograms per cubic meter which occurs at the same
grid point as the calculated maximum 24-hour average concentration. The
Penn Power, West Pittsburgh Plant contributed 163.4 micrograms per cubic
meter or 48 percent of the total. Sources in Ohio and West Virginia re-
spectively contributed 109.4 and 64.6 micrograms per cubic meter (32 per-
cent and 19 percent). The contributions from all other sources are insig-
nificant.
The meteorological data for 28 August 1976 in Table 6-1 were used
with the short-term 1980 compliance emissions inventory in Appendix C.2
and the SHORTZ computer program to calculate 3-hour and 24-hour S0? concen-
trations for the Monessen grid. Because the wind directions on 28 August
1976 were from the south and southwest, the Monessen Plant of Wheeling-
Pittsburgh Steel is only major S0« source contributing significantly to
the ground-level concentrations calculated for the Monessen grid. The
calculated maximum 24-hour average S0« concentration for the Monessen grid
is 226.9 micrograms per cubic meter and the calculated maximum 3-hour
average SO- concentration is 703.9 micrograms per cubic meter. Both of
63
-------
TABLE 6-3
CONTRIBUTIONS FROM INDIVIDUAL SOURCES AND SOURCE COMPLEXES TO THE MAXIMUM
24-HOUR AVERAGE S02 CONCENTRATION CALCULATED FOR THE NEWCASTLE
GRID USING THE SHORT-TERM 1980 COMPLIANCE EMISSION INVENTORY
Source
Perm Power,
West Pittsburgh
Other Beaver Valley
Sources
West Virginia
Sources
Eastern Ohio
Sources
TOTAL
24-Hour Average
S0« Concentration*
(yg/ro3)
46.3
<0.1
23.8
45.6
115.8
* Occurs at a grid point with the UTM coordinates X=555 kilometers and
y=4,533.6 kilometers.
64
-------
these calculated maximums occur at the same grid point which is located
on high terrain about 100 meters above plant grade at a distance of about
1.5 kilometers from the plant. The UTM X and Y coordinates of this grid
point are 594.4 and Y=4,447.0 kilometers, respectively. Emissions from
the Monessen Plant of Wheeling-Pittsburgh Steel contribute all but a few
micrograms per cubic meter to the calculated 24-hour maximum concentration
and are entirely responsible for the calculated 3-hour maximum. Figures
6-2 and 6-3 show the isopleths of calculated 24-hour and 3-hour SCL concen-
trations for the Monessen grid.
65
-------
4460
4459
4458
4457
4456
4455
4454
4453
4452
4451
4450
4449
4448
4447
4446
444'
i i
200-
,50
100
MONESSEN
1 I
>90
591
592
593
594
595
596
FIGURE 6-2. Isopleths of 24-hour average SC>2 ground-level concentra-
trations in yh/nP calculated for the Monessen detail grid
using the short-term 1980 compliance emissions inventory.
The number 14 refers to the Duquesne Light, Elrama Plant.
The asterisk symbol indicates the point of maximum concen-
tration.
66
-------
4460
4459
4458
4457
4456
4455
4454
4453
4452
4451
4450
4449
4448
4447
4446
4445
590
i i I
i i i
20C
200-^
(GDj-SOP
MONESSEN
591
592
593
594
595
596
FIGURE 6-3. Isopleths of 3-hour average S0£ ground-level concentra-
tions in yg/m^ calculated for the Monessen detail grid
using the short-term 1980 compliance emissions inventory.
The number 14 refers to the Duquesne Light, Elrama Plant.
The asterisk symbol indicates the point of maximum concen-
tration.
67
-------
SECTION 7
MODEL VALIDATION
Previous model validation studies in which the ground-level
S0? concentrations calculated by the LONGZ and SHORTZ computer programs
have been directly compared with measured values of average annual and
short-term S00 concentrations are described in reports by Cramer, et al.
2.
(1975; 1976; 1977) and by Bjorklund and Bowers (1979). The results of
these previous validation studies show that, on the average, the annual
S0~ concentrations calculated using the LONGZ program are within about
10 percent of the corresponding measured values. Similarly, the 3-hour
and 24-hour concentrations calculated by the SHORTZ program are within
about 20 percent of the corresponding measured values when reasonable
allowance is made for the fact that the mean hourly wind direction
measurements used in the model to fix the exact position of hourly mean
plume centerlines with respect to SO monitor locations are reported
only to the nearest 10 degrees. It is important to recognize that the
maximum source-receptor distances in these previous model validation
studies are of the order of 10 kilometers and we have not been able to
test the accuracy of LONGZ and SHORTZ model calculations for source-
receptor distances greater than about 30 kilometers. We also point out
that the results of all model validation studies are critically depen-
dent on the accuracy and representativeness of both the emissions inven-
tory and air quality measurements as well as the accuracy and representa-
tiveness of the meteorological observations used in the studies. As
explained below, we originally planned to use 1975 emissions inventory
data, meteorological data and S02 monitor data in the validation study
of the LONGZ and SHORTZ model calculations for the Southwest Pennsylvania
AQCR. However, principally because of deficiencies in the SO,, monitor
data, it was necessary to use various combinations of emissions inventory,
meteorological and S09 monitor data for the years 1976 through 1979 in
the model validation study.
68
-------
7.1 S02 MONITOR DATA USED FOR MODEL VALIDATION
During 1975, the Pennsylvania Department of Environmental
Resources (DER) operated continuous SO- monitors within the Southwest
Pennsylvania AQCR at four COPAMS stations: New Castle, Beaver Falls,
Baden and Charleroi. Copies of the hourly SO- concentration measurements
for 1975 from these four stations were obtained from DER and from the
EPA air quality measurement retrieval system SAROAD. Both data sets
showed valid data retrieval rates ranging from 11 to about 45 percent
which are too low for use in validating long-term dispersion-model calcula-
tions. The Allegheny County Burea of Air Pollution Control (BAPC) operated
a network of seven continuous S0_ monitors in 1975. However, during this
period, there were serious problems with calibration procedures and
instrument operations at all of these monitoring sites. Because of the
resulting uncertainties in the measurements, the 1975 Allegheny County
BAPC SO- monitoring data are also unsuitable for model validation purposes.
The lack of adequate SO,, monitor data for 1975 thus forced the
use of SO measurements for other years in validating the dispersion
models. The Allegheny County BAPC provided us with annual average SO-
monitor data for 1976. The Pennsylvania DER replaced the SO- monitors
at the COPAMS stations in 1978. Since this replacement, the valid data
recovery rates have improved but are still below desirable levels. We
used the annual average SO- monitor data for 1978 and 1979 from the
following DER stations in the model validation study:
New Castle
Beaver Falls
Baden
Midland
The valid data retrieval rates at these stations during 1978 and 1979
ranged from 31 percent at New Castle to 58 percent at Midland. The
Pennsylvania Power Company has operated an SO- monitoring network in the
69
-------
vicinity of the Bruce Mansfield Power Station since 1974. Measurements
from three continuous monitors in this network made during 1977 were
used for both the long-term and short-term model validation. The names
of these Pennsylvania Power Company monitor stations are: Fairview,
Western Beaver and Route 68. Table 7-1 lists the SCL monitor sites and
periods of observation used in validating the LONGZ dispersion-model
calculations. Locations of the monitor sites are shown in Figure 5-1
and in Table 5-1. In validating the SHORTZ dispersion-model calculations,
we used only the Pennsylvania DER and Pennsylvania Power Company monitor
data.
7.2 S02 EMISSIONS INVENTORY DATA USED FOR MODEL VALIDATION
The emissions inventory data used for model validation were
obtained from four agencies:
Pennsylvania Department of Environmental Resources (DER)
Allegheny County Burea of Air Pollution Control (BAPC)
EPA Region III
West Virginia Air Pollution Control Commission
The Pennsylvania DER supplied the emissions inventory for sources located
within Pennsylvania that were outside of Allegheny County. This DER inven-
tory is based on product or fuel through-put for the years 1976 and 1977
and was reviewed and updated by DER personnel for the model validation study.
The emissions inventory for sources located in Allegheny County, which was
developed by the Allegheny County BAPC, is based on the actual operating
levels for 1976 and includes quarterly operating data for the U. S. Steel
sources. The 1976 S0~ emissions inventory for Allegheny County sources
used for model validation reflects updates and changes made jointly by
the Allegheny County BAPC and EPA Region III. The SO emissions inventory
for sources located in West Virginia was developed by the West Virginia
70
-------
w
PQ
H
CO
u w
0 0
r-.
ON
H
,0
ON
r-\
13
0)
[>
V)
<1>
to
0
0)
4-1
cfl
T-H
3
a
CB
C
13 t
0) 4
D
-I
j
4J CO -~x
cO !
t-l 4
-4 CO
-j e
3 C <-.
a c
-H c
i) so
j 3.
U 0
O
rr-t
5
J>
S-i
0)
w
O
0)
4-1
CO
r-l
3
CJ
.ml
r^
nj
c
I
13 T
D
-\
U ±1 ^.
4-1 CO ^
CO M ro
i-t 4
-» S
3 C -<
O 0) 6D
iH C
J rt
CO C ^
U 0
U
C
0
13 4J
0) CO /
s
> j-im
H W
0) C -
3
*%.
W 0) t>0
ja o ;
o c ^
o
u
C
o
H
4-J TJ
CO O
> -H
M M
01 OJ
w PM
O
a
^/
M O
0 4
j
4-1 Q) CO
H 4J 1
-i
C -H 0)
O C/3
^ C
^
^
D
-s
oo o *^D to r^> o -st1 -"d" *^~ ^o
co O i i vO to O c\i r^- o co
Oi li iOr-i iH H O r-iO
co-vfc^Lor^ cxj f-H i i o^ioo
*
CO P-* cO CO CT\ , LO »-H i I Q^ ^O
^0
co CD LO C^ T~H P*** >j" OO r^« C5
LO I~H co r**- *c^ oo ^H r*** o^ *sj*
OiHr-4O o\ o^
r^ r*^- r^*
***** **"-*. ***->.
CO CO CO
II 1 P-. r^ r^. r*-- r !****> i
C^ CJ*t GT\ O> CJ** CTi O^
CO OO CO t 1 t-H r ( i 1 *-4 ^-(^_|
r^ r^ r^
CJ\ CJ\ CTN
M
P^WOJ Q
-------
-a
H
4-1
c
O
O
I
F^»
W
r^
ON
rH
vO
r-.
rH
' 13
0)
t>
J^
0)
03
0
0)
4-1
cd
rH
3
CJ
cd
c
o
T3 -H
0) 4.
J
4-1 CT) /-s
CD >
iH 4.
3 C
J 00
j B
O CU 60
r~i C
C^J C
; 3-
U 0
o
cu
v-i
0)
CO
o
01
4-1
cd
r" !
3
O
cd
o
c
c
)
TJ -H
QJ 4-1
J ' CT
cd j-
j ^ ^
i CO
rH 4-1 B
3 C --.
CJ 0) 60
rH C.
) 3J
a z ^
u c
)
a
C
0
H
T3 4J
cu cd /-
> i-jco
H 4-1 £
cu C ~-
-\
03 CU 60
,£> 0 ;:
O d --
o
u
fi
o
4-1 T3
cd o
> -H
cu cu
CD p..
O
,**
>-
rJ C
1
s
%
i
)
O 4-1
4-1 HI cd
H 4-1 >-
C -H Q
^
)
X
S O
' ^
00 rH CT\
CT. 00 O
O O rH
r^ ^" ' oo
r^ CN \^
CTi ^H vD
r f
m csi r~^.
O CTv CN
rH O rH
vO O- u-i
rH 00 f~-
O^ O^ tr>
000
^O rH CO
0*1 ON r^
**O VO ^>O
r^- r^* r^
ON ON ON
i i i i i i
a
0 >>
TJ !-J
-a !H
4-) cd cu
M M f-^4
O pq
p,^%. ^-^ CO /^^
co u ^ u do
co PL, 4J pu cfl pu
cd ^ M <3 60 <|
rH pq O pq O pq
72
-------
Air Pollution Control Commission and is based on typical source operations.
The emissions inventory for sources located in Ohio was developed from
industrial source information supplied by the Ohio Environmental Protection
Agency to EPA Region III and from information on Ohio power plants contained
in reports published by the Federal Power Commission. The annual emission
rates for the Ohio power plants were set at 65 percent of the maximum
generating capacity. The 1976/1977 S0? emissions inventory used for the
model validation calculations is listed in Appendix C.3.
7.3 LONG-TEEM VALIDATION CALCULATIONS
Because the SO- emissions inventory supplied by the Pennsylvania
DER covers both 1976 and 1977, and the emissions inventory provided by
the Allegheny County BAPC is for 1976, we made average annual concentration
calculations with the long-term dispersion model LONGZ for both 1976 and
1977. The meteorological inputs used in these calculations were developed
from the 1976 and 1977 hourly observations made at the Greater Pittsburgh
Airport, following the procedures described in Section 3. The detailed
tabulations of seasonal and annual meteorological inputs for 1976 and
1977 are contained in Appendices B.2 and B.3.
The long-term dispersion model LONGZ was executed with the
1976/1977 emissions inventory listed in Appendix C.3 and the 1976, 1977
Greater Pittsburgh Airport meteorological inputs to calculate annual
average ground-level concentrations at the SO- monitor sites listed in
Table 7-1. The locations of these monitor sites are also shown in
Figure 5-1 and the UTM coordinates and site elevations are given in
Table 5-1. It is apparent from the discussion of monitor data for model
validation in Section 7.1 and from Table 7-1 that the observation periods
for the four Pennsylvania DER monitors (Baden, Beaver, New Castle and
Midland) are for 1978 and 1979, while the emissions inventory and meteoro-
logical data used in the model calculations are for 1976 and 1977.
73
-------
Consequently, there are serious reservations, for model validation
purposes, about the significance of any comparisons made between the
model calculations for these monitor sites and the observed concentrations.
The observed and calculated annual average S0~ concentrations
at all monitor sites are shown in Table 7-1 together with the ratios of
the observed concentration and the calculated 1976 and 1977 concentrations
at each monitor site. There are three ratios in the table that are less
than 0.5 which indicates that the model concentrations overpredict the
observed concentration by more than a factor of two. One of these low
ratios occurs at the Baden monitor when the annual average concentration
for 1977 calculated by the LONGZ computer program is used to form the
ratio. The other two low ratio values occur at the Hazelwood monitor
where the calculated annual average concentrations for both 1976 and 1977
are much larger than the observed 1976 concentration. We have not investi-
gated the possible reasons for the model overprediction at the Baden monitor
site because the observed concentration is for the period from September
1978 through August 1979 while the calculated concentrations were made on
the basis of the 1976/1977 emissions inventory and the 1976 and 1977
meteorological data from the Greater Pittsburgh Airport. However, the
large model overpredictions at the Hazelwood monitor are of serious concern
and, as described below, we have made a detailed analysis of the 1976
annual average concentrations calculated by the LONGZ computer program in
the immediate vicinity of the Hazelwood monitor.
If the observed and calculated values at all the monitor sites
in Table 7-1 are accepted at face value, the overall average ratio of
observed and calculated concentrations is 0.79 when the calculated 1976
concentrations are used; the value of the overall average ratio is 0.71
when the calculated 1977 concentrations are used. If the Hazelwood monitor
is excluded from consideration, the overall average ratio values are
0.92 and 0.84 for 1976 and 1977, respectively. If both the Baden and
74
-------
Hazelwood monitors are excluded, the values of the overall average ratio
are 0.99 for 1976 and 0.97 for 1977. We conclude that, except for the
Hazelwood and Baden monitors, the annual average SO. concentrations calcu-
lated by the LONGZ computer program are in good agreement with the
annual average concentrations measured at the SCL monitor sites.
Figure 7-1 shows the isopleths of the 1976 annual average SO
ground-level concentration within a distance of about 1 kilometer from the
Hazelwood monitor calculated by the LONGZ computer program. The location
of the monitor is indicated by the filled triangle near the center of the
figure. The asterisks show the locations of individual sources within
the J & L Pittsburgh source complex which is designated by the number 7.
The important feature in Figure 7-1 is the very large spatial concentration
gradient within an area of about 1 square kilometer centered on the monitor
site. Calculated values equal to the annual average concentration of 141.5
micrograms per cubic meter observed at the Hazelwood monitor in 1976 occur
at a distance of about 0.5 kilometers south and northwest of the monitor.
Although we are not able to explain the differences between the model
calculations and the observed concentration at the monitor, the calculated
isopleth concentration pattern in Figure 7-1 illustrates the need for
an increase in the density of the monitoring network in the vicinity of
major S0? sources to define accurately the ambient air quality.
7.4 SHORT-TERM VALIDATION CALCULATIONS
The SO,, monitor data available for validating the short-term
dispersion-model calculations for the five grid areas of the Southwest
Pennsylvania AQCR are very limited. The Pennsylvania DER provided 3-hour
and 24-hour maximums measured at the New Castle and Beaver monitor sites
for the 12-month period from September 1978 through August 1979. We also
obtained summaries of the S0~ concentration measurements, made during
1977 by the Pennsylvania Power Company in the vicinity of the Bruce
75
-------
4775.0
4474.5 -
4474.0 -
4473.5 -
589.0
589.5
590.0
590.5
591.0
FIGURE 7-1. Isopleths of calculated 1976 annual average SC>2 ground-
level concentration in micrograms per cubic meter for the
vicinity of the Hazelwood monitor. The monitor location
is indicated by the filled triangle. The asterisks show
the location of individual sources within the J & L Pittsburgh
source complex which is designated by the number 7.
76
-------
Mansfield Plant, which were above the 3-hour and 24-hour NAAQS for S09.
No SO- monitor data are available for the Monessen grid. Because of these
severe limitations in the short-term SO- monitor data, as well as the uncer-
tainties involved in applying the 1976/1977 annual emissions inventory
and hourly meteorological observations from the Greater Pittsburgh Airport
to the specific 3-hour and 24-hour time periods for which SO- monitor data
are available, direct comparisons of short-term model calculations with
SO- monitor data are of questionable value. For these reasons, we
decided to limit the short-term model validation study to comparisons of
SHORTZ model calculations, made using the 1976/1977 emissions inventory
in Appendix C.3 and the hourly meteorological inputs for 28 August 1976
in Table 6-1, with the highest short-term SO- concentrations measured at
the monitor sites mentioned above.
The maximum 3-hour and 24-hour average SO- ground-level concentra-
tions calculated by the SHORTZ computer program for the New Castle grid,
using the 1976/1977 emissions inventory in Appendix C.3 with the hourly
meteorological data for 28 August 1976 in Table 6-1, are 469 and 108
micrograms per cubic meter, respectively. During the 12-month period from
September 1978 through August 1979, the maximum 3-hour average SO-
concentration measured at the DER New Castle monitor site was 550
micrograms per cubic meter as compared with the calculated concentration
of 469 micrograms. The maximum 24-hour average SO- concentration measured
during this 12-month period at the New Castle monitor was 128 micrograms
per cubic meter as compared with the calculated maximum of 108 micrograms
per cubic meter.
The maximum 3-hour and 24-hour SO- concentrations calculated by
the SHORTZ computer program for 28 August 1976 at grid points near the
West Penn Fairview monitor, which is located at the eastern edge of the
Northwest Border calculation grid'(see Figure 5-1), are 1190 and 333 micro-
grams per cubic meter, respectively. The maximum 3-hour average SO-
77
-------
concentration measured at the Fairview monitor during 1977 was 1387.7
micrograms per cubic meter and the maximum 24-hour average SCL concentration
measured during 1977 at the Fairview monitor was 528.2 micrograms per
cubic meter. Of the 11 cases during 1977 when the 24-hour average SO,.,
concentrations measured at the Fairview monitor exceeded the 24-hour NAAQS,
the prevailing wind directions reported by the Pennsylvania Power Company
were between south and west-southwest for 8 cases which include the case
of the maximum observed 24-hour average. The reported average wind
directions for the other 3 cases were between east-northeast and east-
southeast. For the two cases during 1977 in which the 3-hour average S0~
concentrations were above the 3-hour NAAQS, the reported average wind
directions were from the southeast and south-southeast. Concurrent hourly
observations from the Greater Pittsburgh Airport show low wind speeds and
variable wind directions during the periods for these three cases.
We conclude that the 24-hour average concentrations calculated
by the SHORTZ computer program using the 28 August 1976 meteorological
data are supported by the limited S02 monitor data which are available for
the New Castle, Beaver and Western Border grids. The maximum 3-hour
average concentrations calculated by the SHORTZ program for these grids,
using the 28 August 1976 meteorological data, are lower than the maximum
3-hour SO- concentrations measured at the monitor sites. The prevailing
wind directions reported by the Pennsylvania Power Company for the Bruce
Mansfield Plant, for the period in which the maximum 3-hour S0? concen-
trations were measured at the Fairview monitor during 1977, indicate
that the worst-case 3-hour meteorology may be different from that repre-
sented by the 28 August 1976 observation from the Greater Pittsburgh
Airport used in the SHORTZ model validation calculations. However, a
detailed short-term model validation study which would provide the
requisite objective comparison of SO- monitor data with model predictions
is beyond the scope of this study and is probably not possible because
of the limitations in the existing emissions, meteorological and SO,-,
monitor data.
78
-------
SECTION 8
PRELIMINARY CALCULATIONS OF S02 CONCENTRATIONS ALONG
THE PERIMETER OF THE SOUTHWEST PENNSYLVANIA AQCR
8.1 LONG-TERM CALCULATIONS
The results of the long-term model calculations of S0~ ground-
level concentrations made for the gross calculation grid (See Figure 5-1),
using the annual 1980 compliance emissions inventory showed calculated
annual average concentrations above the annual SO,, NAAQS of 80 micrograms
per cubic meter along the western border of Pennsylvania with Ohio and
West Virginia. For this reason, the two Western Border grids 3 and 4
in Figure 5-1, which are outside of the Beaver and Monongahela Valley Air
Basins, were included in the detailed model calculations. In the calcula-
tions made using the annual 1980 compliance emissions inventory, concen-
trations above the annual standard of 80 micrograms per cubic meter occur
in a narrow strip approximately 17 kilometers long and 3 kilometers wide
located principally in the Southwest Border grid. Figure 8-1 shows the
calculated S09 isopleths for the Western Border area. The calculated
maximum annual average S0? concentration of 101.4 micrograms per cubic
meter occurs in the Southwest Border grid (4) at a grid point with the
UTM X and Y coordinates of 541.0 and 4472.5 kilometers, respectively. At
this point, Ohio sources contributed 35.8 micrograms per cubic meter (35
percent), West Virginia sources contributed 54.0 micrograms per cubic
meter (53 percent) and all of the Pennsylvania sources contribute 11.5
micrograms per cubic meter (11 percent) to the total. The contributions
of individual major SO,, sources and source complexes in Ohio and West
Virginia to the calculated maximum ranges from about 5 micrograms per
cubic meter to 20 micrograms per cubic meter. It thus appears that no
single source or source complex is responsible for producing the high
calculated ground-level concentrations.
There are points within the area of the calculated exceedance of
the annual standard where Ohio sources contribute more than 50 percent of
79
-------
4535 -
4525 -
4515 -
4505 -
4465
4495 i
4485 -i
4475 -
542
544
546
548
550
FIGURE 8-1. Isopleths of annual average S0~ concentration in yg/m
calculated for the Western Border area of the Southwest
Pennsylvania AQCR using the LONGZ computer program with
the annual 1980 compliance emissions inventory. The
numbers 3 and 4 refer to the Northwest Border and South-
west Border calculation grids, respectively.
80
-------
the calculated concentration. At the point with UTM coordinates X=541.0
kilometers and Y=4482.5 kilometers, the concentration is 83.9 micrograms
per cubic meter. Ohio sources contribute 46.4 micrograms per cubic meter,
West Virginia sources contribute 26.3 micrograms per cubic meter, and the
remainder of 11.2 micrograms per cubic meter is contributed by sources in
Pennsylvania. Thus, the sources in Ohio and West Virginia are jointly
responsible, according to these calculations, for the high concentrations
in the border areas.
8.2 SHORT-TERM CALCULATIONS
The two Western Border grids were also included in the model
calculations made using the short-term 1980 compliance emissions inventory.
Ground-level concentrations above the 24-hour NAAQS were calculated
within a very small area of approximately 3 square kilometers in the
Southwest Border grid. The calculated maximum concentration of 425.2
micrograms per cubic meter occurs at the grid point with UTM coordinates
X=544.0 kilometers and Y=4,466.0 kilometers. At this point, Ohio sources
contributed 60.6 micrograms per cubic meter, West Virginia sources
contributed 352.1 micrograms per cubic meter and Pennsylvania sources
contributed 12.4 micrograms per cubic meter to the total concentration.
A maximum 3-hour S0~ concentration of 2,375 micrograms per
cubic meter was calculated at the same grid point at which the calculated
24-hour maximum occurred. Practically all of this maximum (2,368 micrograms
per cubic meter) is contributed by sources located in West Virginia. We
found several other cases in which the calculated 3-hour maximum concentra-
tions were above 2,000 micrograms per cubic meter. These concentrations
are also almost entirely due to sources in West Virginia. However, cal-
culations necessary to define the maximum 3-hour and 24-hour concentrations
due to sources located in Ohio and/or West Virginia were not made.
81
-------
Additionally, there is some indication from the short-term
calculations that there may be an area of high ground-level concentration
near the Hatfield Power station. We calculated a 24-hour SO- concentra-
tion of approximately 200 micrograms per cubic meter produced by Hatfield
emissions at the extreme southern edge of the Monnessen grid on 28
August 1976 in the presence of southerly winds. The calculated maximum
3-hour concentration produced by emissions from the Hatfield Power
Station at the southern boundary of the Monnessen grid was approximately
650 micrograms per cubic meter.
We point out that the model calculations for the border areas
described above were made using worst-case meteorological conditions
selected for other areas and may not be representative of the worst-case
situations for the border areas. In addition, there are some questions
about the validity of the emissions inventory data for the West Virginia
and Ohio sources. While the results of our model calculations serve
to point out potential air quality problems in these border areas, an
additional effort beyond the scope of the present study is required to
establish the magnitude and areal extent of high SO concentrations
along the western and southern perimeter of the Southwest Pennsylvania
AQCR.
82
-------
REFERENCES
Alkezweeny, A. J. anrl D. C. Powell, 1977:
rate of SO
Environ.,
Estimation of transformation
to SO, from atmospheric concentration data, Atmos.
, 179-182.
Bjorklund, J. R. and J. F. Bowers, 1979: User's instructions for the
SHORTZ and LONGZ computer programs. Technical Report TR-79-
131-01, H. E. Cramer Company, Inc., Salt Lake City, Utah.
Bowne, N. E., 1974: Diffusion Rates. Journal of the Air Pollution Control
Association, 24/9), 832-835.
Briggs, G. A., 1971: Some recent analyses of plume rise observations. In
Proceedings of the Second International Clean Air Congress.
Academic Press, New York.
Calder, K. L., 1971: A climatological model for multiple source urban air
pollution. Proc. 2nd Meeting of the Expert Panel on Air Pollution
Modeling, NATO Committee on the Challenges of Modern Society, Paris,
France, July 1971, 33.
Cramer, H. E., et^ aL_., 1972: Development of dosage models and concepts.
Final Report under Contract DAAD09-67-C-0020(R) with the U. S.
Army, Deseret Test Center Report DTC-TR-72-609, Fort Douglas,
Utah.
Cramer, H. E., H. V. Geary and J. F. Bowers, 1975: Diffusion-model calcula-
tions of long-term and short-term ground-level SO 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, Pennsylvania. EPA Report 903/
9-75-018. NTIS Accession No. PB-245262/AS.
Cramer, H. E., J. F. Bowers and H. V. Geary, 1976: Assessment of the air
quality impact of S0~ emissions from the ASARCO-Tacoma smelter.
EPA Report No. EPA 910/9-76-028. U. S. Environmental Protection
Agency, Region X, Seattle, Washington.
Cramer, H. E., J. F. Bowers and H. V. Geary, 1977: Comparison of calculated
and observed hourly ground-level SO,., concentrations for the
ASARCO-Tacoma Copper Smelter, APCA Paper No. 77-58.3, presented
at the 70th Annual Meeting of the Air Pollution Control Associ-
ation, Toronto, Ontario, Canada, June 20-24, 1977.
DeMarrais, G. A., 1959: Wind speed profiles at Brookhaven National Labora-
tory. J. Met., 16, 181-190.
83
-------
REFERENCES (Continued)
Forest, J. and L. Newman, 1976: Oxidation of sulfur dioxide in power
plant plumes, BNL-21698, work done for ERDA by Department of
Applied Science, Brookhaven National Laboratory, Upton, New
York.
Gillani, N. V., R. B. Husar, J. D. Husar, D. E. Patterson and W. E. Wilson,
Jr., 1978: Project MISTT: Kinetics of particulate sulfur forma-
tion in a power plant plume out of 300 kn., Atmos. Environ., 12,
589-598.
Holzworth, G. C., 1972: Mixing heights, wind speeds and potential for urban
air pollution throughout the contiguous United States, USEPA, OAP,
Research Triangle Park, N. C., Publication No. AP-101.
Luna, E. E. and H. W. Church, 1972: A comparison of turbulence intensity
and stability ratio measurements to Pasquill stability classes.
J. Appl. Met., 11(4), 663-669.
Lusis, M. A. and H. A. Wiebe, 1976: The rate of oxidation of sulfur dioxide
in the plume of a nickel smelter stack, Atmos. Environ., 10,
793-798.
Pasquill, F., 1961: The estimation of the dispersion of windborne material.
Met. Mag., 90, 33-49.
Rubin, E. S., 1974: The influence of annual meteorological variations on
regional air pollution modeling: A case study of Allegheny
County, Pennsylvania. Journal of the Air Pollution Control
Association, 24(4), 349-356.
Sidebottom, H. W., 1972: Photooxidation of Sulfur Dioxides, Environ. Sci.
and Tech. , b_, 72-79.
Tesche, T. W., G. Z. Whitten, M. A. Yocke and M. K. Liu, 1976: Theoretical,
numerical and physical techniques for characterizing power plant
plumes, EPRI EG-144, Electric Power Research Institute, Palo
Also, Calfironia.
Turner, D. B., 1964: A diffusion model for an urban area. J. Appl. Meteor.,
3.(1), 83-91.
Wilson, W. W., R. J. Charlson, R. B. Husar, K. T. Whitby and D. Blumenthal,
1977: Sulfates in the Atmosphere: A progress report on Project
MISTT, EPA-600/7-77-021, U. S. Environmental Protection Agency,
Research Triangle Park, North Carolina.
84
-------
APPENDIX A
MATHEMATICAL MODELS USED TO CALCULATE
GROUND-LEVEL CONCENTRATIONS
A.1 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
z
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 a* and a' may be directly specified or may be assigned on
A Ci
the basis of the Pasquill stability category (see Section 3 of Cramer,
^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
u
R
H
m
li
3z
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
f (Table)
pfcfl (Table)
aE;i,k (Table)
(Table)
(W.\
U),
Hm;i,k,£ (Table)
u {zR}i (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 ic wind-speed cate-
gory
Standard deviation of the wind-elevation
angle in radians for the ifc wind-speed
category and kt*1 stability or time-of-day
category
Ambient air temperature for the k stabil-
ity or time-of-day category and &tn season
(OK)
Vertical potential temperature gradient for
the i^h wind-speed category and ktn stability
or time-of-day category (°K/m)
Median surface mixing depth for the i'" wind-
speed category, k stability or time-of-day
category and H*-". 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)
3
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
{h} \ 2Y
/ 3F
\ 2Y,2 /
(10h)
2/3
(A-l)
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)
Y- = the adiabatic entrainment coefficient ~0.6 (Briggs, 1972)
/ o
F = The initial buoyancy flux (m /sec )
(A-2)
3
V = The volumetric emission rate of the stack (m /sec)
TT 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)
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
-------
f
3w - 3u{h> \
u {h} < w/1.5
w/1.5 < u {h} < w
w
; u {h} >
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
6F
u{hh
2 ^
1/3
3F
1 - cos
10S1/2h
u{h}
-1 /2
;TT u{h} S ' <10h
1/3
-1 /?
;IT u{h} S ' >10h
(A-4)
where
'2
S
the stable entrainment coefficient~0.66 (Briggs, 1972)
= vertical potential temperature gradient ( K/m)
The entrainment coefficients Y-I an
-------
the calculated stable rise Ah to exceed the adiabatic rise Ah,T as
s N
the atmosphere approaches a neutral stratification (36/3z 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
KQ
TV u{H}a a
y z
{Vertical Term} {Lateral Term} {Decay Term} (A-5)
where
K
Q
u{H}
a ,a
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
2
+ exp
n=l
exp
. /2n H + H\2
1 I m
, 2n H -
1 / m
(A-6)
A-9
-------
where H is the depth of the surface mixing layer. The exponential terms
m
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
.1/2? 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 ,
m
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 - -r- .
(A-8)
4
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
{Decay Term} = exp [ - ty x/u{H} ] (A-9)
where
enging
= the washout coefficient A (sec ) for precipitation scav-
A-10
-------
0.692
T
1/2
, where T , is the pollutant half life in seconds
for physical or chemical removal
= 0 for no depletion (^ is automatically set to zero by
the computer program unless otherwise specified)
In the model calculations, the observed mean wind speed u is
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}
(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, et^ aj^. (1972),
,rd deviation c
given by the expression
the standard deviation of the lateral concentration distribution a is
a {x} = a! x
y A ry
x + x - x (1-a)
Y ryv
ax
ry
(A-ll)
x =
y
ax
y
ry x a
\ ry A
XR
x (1-a)
ry
< x
~
ry
(A-12)
A-ll
-------
where
ry
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)
= the lateral diffusion coefficient (~0.9)
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
on the basis of the Pasquill stability category .
Following the derivation of Cramer, e± al. (1972) and setting
the vertical diffusion coefficient 3 equal to unity, the standard devi-
ation of the vertical concentration distribution
sion
O is given by the expres-
z
az{x} =
(A-13)
x
zR
R
(A-14)
where
al =
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 a' may also be obtained from direct
b
measurements or may be assigned according to the Pasquill stability cat
egories. When a' values corresponding to the Pasquill stability cate
rs
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
2
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
0.5 Ah
2.15
(A-15)
The downwind distance to stabilization
^
K
is given by
lOh
X
R
TT u{h} S
1/2
; > 0 and TT u{h} S
1/2
< lOh
(A-16)
lOh
; - > 0 and TT u{h} S
1/7
'
> lOh
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 a is defined by the building crosswind dimension y divided
yo j- & JQ
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 a 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
£\ ill
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
K 0
X{x, y} = - {Vertical Term}
/2? u{h} 0 {x} y
(A-17)
{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
E o
x < 3 x
In
a'(x+x
x > 3 x
o
(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
L^i exp
n=l
. /2n H \ 2 /
- l I ? I i(
2 ^^"Kry ' 2 \
6H
m
/2? a {x}
z
2H
m
6H
m
10
10
(A-19)
The Lateral Term is given by the expression
{Lateral Term} =
-------
and
a ,{x} = o! (x-hc _/2)
(A-21)
The Decay Term is given by Equation (A-9) above.
The concentration at points interior to the area source is
given by
2 K Q
/2iT u{h} x y a'
o o Jj
In
a (x'+D+h
{Vertical Term} (A-22)
where
x' = distance downwind from the upwind edge of the area source (m)
A. 4
LONG-TERM CONCENTRATION MODEL
A.A.I
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, 0) is
given by
A-16
-------
/hTr A0T ~~.
exp [- * r/u. Hi>k)
(A-23)
i,k, £
= exp
JL
2
+ exp
2n H
n=l
. .
m;i,k,£
z;i,k,£
2n H . , -H.
(A-24)
where
A9f
S{0}
frequency of occurrence of the i wind-speed category,
jth wind-direction category and kfch stability or time-
of-day category for the &th season
the sector width in radians
a smoothing function
S{G} =
A8'
0
-9'> 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,8)
A-17
-------
As with the short-term model, the Vertical Term given by Equation
(A-24) is changed to the form
Vi k £,
^ ) .L ) IS.) A/ ,. «/-»
(A~26)
2H .
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,6) is calculated from
the seasonal concentrations using the expression
(A-27)
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 cr is defined as the
zo
building height divided by 2.15. Separate vertical turbulent intensities
<3* may be defined for the low-level sources to account for the effects of
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
2 K Q
/2? R A9'
u.{h} a
S{9} V,
z;i,k
(A-28)
exp -
where
R = radial distance from the virtual point source to the receptor
a/2
r =
distance from source center to receptor, measured along the
sector centerline (m)
r = effective source radius (m)
A-19
-------
y =
xy
lateral distance from the sector centerline to the receptor (m)
lateral virtual distance (m)
r cot
o 2
(A-29)
z;i,k
In
"GE
L°E
;i
;i
,k(
.k(
r
r
'+r
'-r
oy
o >
+
+
h
h
r < r
o
< 6r
E;i,k
r' + h
6r
(A-30)
1+2 7 exp
n=l
2n H
. , .
m;i,k,£
a
z;i,k
H . .
m;i,k,.
> 10
. l/6Hm;l.k.t
' 2\ ".il.k
< 10
(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
x y
O 0
u.{h} a' .
i E;i,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 TERRAIN
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
in s
A-21
-------
where
m
the depth of the surface mixing layer measured at a point
with elevation z above mean sea level
a
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
TH 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-
s
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
1 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
H;{Z} =
H
m
H + z
m a
- z ; z < z
(A-35)
A-22
-------
C1J
JS
4-1
c
H
ft
4-1
H
&
T3
CD
a
o
a
en
H
CU
T3
cu
N
H
t-l
H
J2
cd
4-1
tn
Q)
0)
iJ
Q)
0)
c
Q)
4J
0)
td
"0
a)
en
SJ
r-»i 0)
cn>-,
N cfl
--r-- rH
KB
SC M
c
.£ -H
JJ CO
PH -H
0)
M CJ
a td
rl H-4
X M
H 3
S CO
g
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
m
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
K a
level is adjusted to the stack height for the plume-rise calculations
by the relationship
u(h) =
; 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} =
; H < z + z_,
' o a R
(A-37)
A-24
-------
en
a
a
3
a
.H
rt
a
o
c
a)
o
a
o
o
o
PH
T3
H
(-1
M
O
4J
N
W-*
- 6
ex
a)
id
bD
C
H
X
H
J-J
a
a)
CM
W
O
M
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
METEOROLOGICAL INPUTS USED IN THE LONG-TERM AND MODEL CALCULATIONS
(BASED ON METEOROLOGICAL OBSERVATIONS MADE AT THE
GREATER PITTSBURGH AIRPORT)
B-l
-------
M
X
M
y
w
p-l
^
w
CJ
y.
^
M
&
o
a
00
o>
rH
1
1
O
w
CO
P
CO
H
P-t
J2
M
i-I
S5
O
en
" uj in
3: i-
S5~
Ci in to
LU U.
LU >- E:
a. o; o
en o o
CJ O
CJ UJ tO
= *- '
2* O
O if <£>
LU a.
LU > E:
a. a: o
1S> O O
C3 O
»* a ui c»
Z H-
0£ =S CJ
O -s
CD ^
»-< LU c* M to
i LU a.
2 CC LU > H:
o o a. etc o
w ' to o o
- o o
LU t O UJ f*>
03 -. 3 h-
>-. 30
H- Ct IM lO
to uj a.
ui >- r:
a. a; o
V) O O
cj uj in
3H 1-
*** ^C CXi
3 0
O i to
LU a.
LU 3- SC
a. a: o
y* o *>
o in
id LU N-
t_d /T
O 0 O O 0 O O- O O- -.C- 0 O -!> C- O O
'O O1 *^> <^ C^ "C> d> '^^ "C* "O "O* -C> '-^ O* "^s
o o o o o o o o o -o o o o o- o o
O O JI> O O O O O -O O O O -C- -C- O O
o o o o o o o o o o o o o o o o
o o o o o o o- o o o o o o o o o
o o o o o o o o o o o o o o o o
O O O O O O O O O O O -O -0 O O O
O O O O O O O1 -O -O O O O O O O- O
ooooooo-ooo-oooooo
c- c.- o c- o o -c- o o- o o o o o o o
O O O O O O O O O O O -O 'i> O O O
o c- o o -o o o o o- o o o o o o o
oooooooooooooo o o
C-" O O O O1 O O O1 O "C" O O O O O O
o o o o -o -c- o o o -c> o o -o o o o
oooooooooooooooo
oooooc-oooooooooo
0 0 0 0 O 0 0 O 0 -0 0 0 -0 O 0 0
oooooooooooooooo
c.- o- o o o o o o o -o o o o o o o
oooooooooooo oooo
O O O O O O O O £> O O O O «> O >
oooooooooooooooo
oooooooooooooooo
oooooooooooooooo
oooooooo-oooooooo
oooooooo-o-ooooooo
ir?**ooo*^ro^*'co^oo^ocrt-^-»rori-rN-rw
r-i'd-roty5i'iomro'*-'O'*vopoiMr>-oo
^^ oj T * co to in *" c\j o& c*j cr* co in c*u ^^ "*~*
r*j ^^ **4 *o c>j ^r ^^ *^* ^f ^r* fN* *.o 0*1 cvj (j^ (js
^^ «~4 O ^^ <^^ ^^^ ^r f~4 -f 4 »-4 «^ ^^ »^ rt^ <^ -O
oooooooooooooooo^
O" ^^ "*^ "O" *C^ ^Ci* "O* *C5* "O1 ^1? "C-^ "O1 ^^ ^^" 'O" '^^
00- 00 0000000 00000
<9-f»iT-«com'j>cO'«'-^<<^>inrs--<»' in co *-« CM o *- r*! 'JD cxi c»J o ro »3>« -»-^ r^
OOOO-OOOO'O-OOOOOOO
oooooooo-oo-oooooo
oooooooooooooooo
ooo|jao^cr»M5OfN*-' »"Tf-r^-O'jac\i*-«^^
_ *foJo-^ior-.^-<'jc>»ncM-^«vOcr>r*1>rr'
co tJ** csj oz> \o ii^ *>jo vc* *^" Oxi Ovi c\i ro ro ui U"5
r^> "\f* f->m (j ) ij^ tj^i ^* -c*> ix^ oj '-O o"1 ^o r*"* r*^ ^-j"
O O O O O O O *^ O O O O -7- O O *>
O O O O O O O £> O O -O O O O *£!> O
OO<><>-OO-C>OO'<>-C>O'-C><><>O
01
3Z. Lu -;> O O -O <> O ^:
O Ct; O O O --i O C: O -O C> O- -C> O -O C1 O O
*-« '-3 o u") o- in o in o ir> o in o u) c: in -o ui
o
-.- CJ IT-
CM TT
B-2
------- OCR error (C:\Conversion\JobRoot\00000AGL\tiff\2000VIIX.tif): Saving image to "C:\Conversion\JobRoot\00000AGL\tiff\2000VIIX.T$F.T$F" failed.
-------
UI 0.
UJ >- S
u_ lii O
CO O O
O O
ci ui in
oooooo-oooooooooo
o o- o ro o o r- o- o ro o ro oo « rv.
-O O O »- O -C- O-J O O T-t -C* "«t" U*> ^1* «.'«J
O O S>»tOOOIOO«-iO%r)vOCMMCM
OOO-C'O-OO-O-OO-OMrO- a:
o. fy f.>
to o o
o o
o ui u->
Cxi o oo u-. co r- uo
in o « IN- ^ ir-j -^< oj ro in
OOOOOOOO'OOOO'r-iOOO
oooooooooooooooo
W
Pi
§
W
PH
1/1
Pn
O
W
oe.
o
« UI
o
to
UI
- s:
Q- a: o
u> o o
t5 O
d UJ CO
UI 0.
UI i- El
a.QiO
to o o
C3 O
O UI ^
3 O
d OJ Ul
ui a.
ui 5- T:
cu a: o
to o o
o o
o UJ in
"OOOOOOOO « CM OJ «M »i «
inif)«oror<>-J-^- "-4 tO
ui a.
ui >- 3:
IX. lit O
.<> '.-j o
o «»
O'OO-OO'OOO-OOOO-OOOO
oooooooooooooooo
oooooooooooooooo
00:00 o-
' (JJ O- UT O
r UI
UJ '"^J rJ"
-> > > » o o o
o o o o c> -z> -o-
!j") O- b7 O U J C> t*">
rj tn r- o t\j iri N-
B-4
-------
^~*
H
M
M
H
CO
Q
53
<<
g.
O
M
H
O
fd
M
n
«\
w
w
Pn
CO
o
M
1-5
i^*
s
u
M
s
£3
o
o
o
Pn
0
w
^
o-
w
£
UJ
UJ
ii-
CO
^
UJ
LU
a.
CO
(Z*
=
s
0
UJ
UJ
a.
LT.'
^^ *^**
z
« s
o
*-i UJ di
> UJ
z
UJ t O
co . n:
1 | <
_i_J *"^
CO
0? UJ
UJ
a.
CO
O
z.
ra
o
UJ
UJ
a.
ty-i
o
3:
>
-
o
C3
UJ
h-
OH
CJ
in
CK:
o
0
UJ
I
CJ
^>
>-
CsC
0
CJ
UJ
1
XT'
CJ
ro
D~
C£
O
CJ
UJ
1
OC
0
O«J
cc
o
o
UJ
t
CJ
_
>-
lii
i:"*
C3
UJ
I
C3
CO
a_
o
in
..-s
u.
o
0
in
^
co
a.
S;
O
O-
f>
CO
X
co
a.
i:
o
o
o
ro
.
.^"*
cu
o
o
o
in
~
CO
a_
:c
o
.^*
in
IS.
^s
CO
UJ
:z uj
O (X
.--! CJ
»~ »_lj
'-.'> C^l
UJ
0- 0
o o
o o
o o
'C21 'C?
O 0
'-- O
o o
o o
o -o
0 0
o -o
0 -0
o o
0 0
o o
0 0
0 0
0 0
o <»
\J» OD
0 «
CO *.O
If) ,>
O *2>
^ lf>
O C.J
1-J
O i>
0 O
O C'-
O <^
o o
o -o
o -o
o o
-c* <>
o o
o o
C^ *O*
o o
o o
o o
o o
O 0
0 0
o o
0 0
o o
0 0
O 0
C- 0
OS Tf
s^l CO
rs- ro
rvj -f
csj ^1
O £>
0 O
* in
"rf* r^^
«-^ C*J
ro vo
00 (X-
o o
r> O
^u> ^>
C5^ "^"
* -rr
* ' m
U"t "**
vr -w
o o
o o
<- <'
0- O
s> in
U"< fv.
TT -.I/
<^> *I>
c> -o-
c- <:-
o o-
o o
o- o
o o-
O C--
o o
O1 O
o o
<!:. O
O 0
O 0
o o
o- o
o o-
o o
2- O
o o
0 -O
O 0
o o
o o-
1 CSS
J5 IT'
^ ro
ro *
«- 'J-
o o
'> O
-^* CO
co in
in ea
* -o
T^- -^J-
-«-l -^^
0 -0
f^> ^^
JJ CO
in in
CM O
C>j f-4
m r~-
"^ *^r
O C>
> .>
-o ;>
C> UT>
O O-J
C'S -v-"
o -o
^ o
C^ -C*
0 0
o o
o o
o o
o <>
o o
o o
-o o
o o
o o
O* *^>
o o
o o
0 O
o o
o o
O "?Z>
o o
o o
o o
o- cr»
ro «J3
vi> fs.
^-1 l%l
csj csj
o o
0 O
in o
ro -.0
Cfi vO
ro is-
JJ 'JO
»- o-
o ->
i o-
^^> |S_
-« ro
CO T
i^« tn
f^ in
o -o-
O 0-
O :'>
-'> O
^.> UT
U"5 F1^-
i--i ir.-
0 O
^* -C^
<1> o
<"^> -^
o o
o o
o o
c- o
o o
o o
o -c-
o o
0 O
o o
o o
o -o
o o
o -c>
o o
0 0
o o
o o
0 &
ro r-
in -o
i-o rs.
cr> co
< ro
o o
-o -=>
cr. in
-- ts,
in .M
^2^ *kO
Mi IS.
~ -O
O O
< o-
U- ISS
rr ro
».'>J -C>
r«> t-«
< xj-
^-i O
o o
o "O
-> >
^. in
> -M
IX- :>
o o
-o -o
o -c>
0 0
o ->
o o
0 0
O* O1
o
o o
o -o
O 0
O 0
o o
o -o
o o
o o
& o
o o
o o
O 0
0 0
« o
in in
IS. IS.
CSJ <*
.o ^-<
rs. >*Q
o o
o -o
csj in
>» ro
in ro
^» iy~t
O *C!*
0 0
-o o
*t- ro
**F (N-
U"> Is-
,j) js^
ir> rj-
o o
C* 'O"
7^ ->
o "'
O U"3
u"> r^
C-j -^r
O- O
C> -C^
o &
--> o
o o
C*1 "O1
0 0
^ ^
o -r>
o o
o- -a-
o- o
o o
o o
o o
<*5- <"^>
o o
o -o
..-. o
o- o
C- O
-o o
0 0
o- o-
T ro
T*- Ul
«r ro
or^ cr^
^> t-i
o o
O" '-I?"
fv Cf\
»-< CO
ro o
ro co
~-« 0
o o
o o
»-i CO
O '^)
C"-, -r^
r^ ^-*
r- in
o -:>
c> o
»> O
'> ^
^ IjJ
C- I'M
r»- w.
O O
o o
C^ *^
0 0
o o
0 0
o o
0 0
o o
o o
o o
o o
o o
o o
o o
o o
o o
O 0
o -o
o o
0 0
O 0
o o
0 0
o o
o o-
o o
0 Tf
ro co
* os
.o o
ro «»
0 0
O 0
ro -w
0 CO
r*» r**>
o »
p* 0
0 0
0 0
as o
cr» *-
««t O
rs, ro
O CSJ
o o
0 O
0 O
o o
^3* iO
in rs,
< ro
B-5
-------
CO
W
u
o
CO
W
s
|i!D
1-1
o
«
0
H
3
M
O
PH
Q
W
H
<]
>
W
,_5
W
pel
O /-x
pt| CO
W CO
W -~-
O
Ul
Ul
a.
'S>
d
2
i*
3
A
Ul
Ul
a.
CO
d
X:
^>«
31
Q
Ul
Ul
a.
CO
G>
2
»wi
3
0
Ul
Ul
a.
CO
o
z
l»l
3>
0
>
£li
O
CJ
Ul
I
ct
u
in
>-
ai
0
o
Ul
1
3-
Oi
0
(3
Ul
H-
a;
0
o
c>
+
Ul
o
0
in
*-
N.
*-H
o
o
+
Ul
o
o
in
o
o
uo
o
o
0
Ul
ro
r^
«>«
o
1
UI
o
o
o
in
ro
r»-
«*i
o
i
Ul
o
0
o
ID
n
fs.
-^
0
1
Ul
^
o
o
in
CO
*
* «
o
1
Ul
o
o
o
10
J>
*
«Mt
0
1
Ul
0
o
o
in
SO
V
«^
0
1
Ul
o
o
o
in
v£>
Tf
W
W
Ul O O O O
Ul >-*+ I I
o_ or LU LU ui LU
CO O O O O O
o o o o o
O Ul 111 -O O O
z t <* eo in w
»-> -*» I I
OL OC LU LU LU Ul
'.O O Cs -C- O O
O O O- O O-
o LU in o- f> o
z t * TO in in
>-«
O IT" O O « «
Ul O O O O
UJ J- + + I I
0- LE: UJ LU LU LU
co o o o o o
o o o o o
o ui in o o o
Z t * U3 IO 111
>-« - o M >^
3 O «- f>- *
o o
O O
Ul >- + + I I
o_ o£ ui ui ui ui
CO O O O O O
<_3 O O O O
a LU in o o o
z » * co in if)
t <
£> ui in o o o
2: t oj o o -^
ui o o o o
ui s- + + i i
0. Ct: LU UJ LU LU
CO O O O -C- O
o o o o o
o ua m o o o
3T i -«r co in in
>
3* tj »-« »" t>- «*
O
M
H
s~* ~+ £5 £> *-* *-»
Ul O f> O O
Ul 3- + 4- I I
O. dt: LU LU LU LU
Ul O O O- O O
O Ul IT- O O O
ae H- <* «j ""> m
i-< '- "- fN- *
Ci ^ O O -H -<
LU O O O O
LU >- -t- -t- I I
O. Cki U4 LU LU LU
in o o o o o
d ui
in o o o
sf co in 10
r-. -o rt vo
n
H
->->- 3-
(~ t- 'r- f
I--4 '-I » 1 »*
r.*j L«- tu cu
a; i -c -r.
( f_ i - >-.
co 07 to
B-6
x x
* f
co co
-------
£i '.O O O O
ijj > -t- "* I 1
0. Oi UJ UJ Ul UJ
-.0 O O O :> O
O O -C- O O
Cs UJ O O O !>
rz: t o o -o o
ii <3C O O O O1
3s i_> o o r*» r»7
o o -o o o
UJ C-- :> O O
Ul i- »» + *
O. CC LU LU Ul LU
U> O O- O O O
C3 O O O O
O Ul O O O O
Z H- O O O O
cc o o o o
3 <_j -" - *-i iM
o in o o co cij
UJ O O O O
UJ >- + * t I
a. ii: uj uj UJ ui
y> d O O -O *">
t_3 O O -O O
O UJ O O O O
Z H- O O O O
«
-< CO
H
&
Cd
H
5
O
W
PS
p
1
w
is
w
H
,J
<:
H
H
&
W
H
O
PM
3
O
l-l
H
tf
W
>
y> S^
z
M
>
hJ
w
w
w
w
w
erf
o
w
Q
SI
Q
PL,
P
UJ
Ul
a.
CO
d
3E
I-H
S
a
LU
UJ
LL.
CO
O
z
II
3
£>
Ul
UJ
a.
to
cj
H-4
3»
>-
or
o
(J
UJ
i~
-
a;
0
rjl
LU
1
o
o
+
UJ
o
Cv
o
0
0
o
o
o
+
LU
o
o
o
o
c-
o
o
o
1
LU
O
o
o
o
o
ro
CO
o
1
LU
o
o
o
o
o
in
o
I
Ul
o
o
o
o
o
v-4
o
1
Ul
o
o
o
o
o
ro
*-4
^
1
UJ
o
o
o
o
m
f>4
^^
O
1
Ul
c-
o
o
o
o
CVJ
Ed
§
Pi
W
w
1-J
§
o
a
M
&
O T>- O C- O O
Ul O O O O
UJ >- + -t- -» +
O. Ct: LU LU LU LU
to o o o o -o
o o o o o
O UJ O O O O
a: i <» o o- o
»-< O O
o o o o o
<=> UJ O O O O
z \- o o o o
>-< <£ O O U^> O
3 O '-«'-« CO
O CO O O O O
U) O O O O
Ul >-» **»
O. Qi Ul LU UJ UJ
o o o o
-i -+» I I
a. a: LU uj LU LU
10 O O > O O
i.n .' ". J> '.7>
O UJ O ^ O- O
Z I O O -O O
i
O UJ "> O O* O
X t O O O O
i-t
-------
_J O
> CJ
IB LU
- o
t > CO
»-4 a; r~-
_i o
-« rJ o
co ui co
- o
_j o
>-4 o ro
CO Ul CO
- o
I 3- CF-
_J O
*-1 O SO
CO LU CO
-a- -if TT ro
uj o o o o
LU >- 4- 4- 4-
a. a; uj ui uj uj
i/) o o o m o #
o o -c> CM o i
cj ui o o in *
Z I O O C1 CM
- in 3-
3 O >- >-! >-« so H-
LU O O O O CD
LU >- 4- 4- 4- 4- (j _ _ _, kO
PM
W
W H
>* W
O ffi
& ^
l-l
X
LU O O
LU 3- 4- +
n rs 111 l L i
CO O O O
sJ O O
O LU O O
Z I O O
HH <^c ro ro
O M «*
Z LU O -O
O LU 3- 4- 4"
CO O. O£ LU LU
3 O - >-
O
"*^ ro **^
o o t
4- 4- O
LU Ul LU
o o a:
111 O >-
CM in o
in CM -
>- in o
< SO LU
. LU
a.
ro ro o?
o o
4- 4- O
LU LU Z
o o -«
in o 3
CM in
SO CM U.
IS. Ul O
C7^ SO
. . LU
CJ
LU
o o o o
a.
co
l^ Ul LU
o o o
o o o
o LU m ui
z h- co ca
i" <£ CM CJ
rs cj f-1 -»^*
LU
4- 4- OC.
LU LU a:
O O =5
in o cj
!""" U^ *_?
CO CM O
CO C*"»
N. CM U.
C3
^^ 4 *& tjr ro r^ 3~*
LU ^^ *^> ^> O CJ
Ul I- 4- 4- + 4- Z
IX C£ LU LU LU LU LU
lO O O O O O CD
t.3 Ul U) «-* -> C?
«c^ LU r^- IN- m i/> tu
i "<
til iaj flij C&
.: a ' 4:
! - \~ I - t~
'Jl l/J s>> O">
-------
Ct SO «>
ui a.
ui j- ac
ex. oc o
to o o
cs o
o ui 10
a: i-
- V>
ui a.
ui >- c
a. ac o
to o o
ta o
a ui 4r>
X H-
» - ac
O. Q£ O
u> o o
o o
o ui o»
O 0
O 0
00000
00000
O 0
0 0
o o
o o
0 0
o o
0 0
0 0
o
o
00000
o
o
o
00000
0
o
00000
0
o
o
00000
o
o
o
00000
o
O 0
o o
o o
O 0
o o
0 O
o o
o
0
00000
o
0
o
o
o
o
o
o
o
0
00000
0
o o
o o
00000
00000
o o
> «-
oooooooo o o o o o o o o
oooooooooooooooo
oooooooooooooooo
oooooooooooooooo
oooooooooooooooo
oooooooooooooooo
oooooooooooooooo
oooooooooooooooo
CM
X!
P-.
o
w
w
PH
CO
-
UI t-
o> «-«
o n u>
UI 0.
ui > x
a. ac o
o» o o
u o
O UI fO
z *-
«* (M M
in a.
UI > JC
a. ex o
(O O O
o o
O UI V)
oooooooooooooooo
oooooooooooooooo
oooooooooooooooo
o
u
o1
fe
e> i to
ui a.
ui >- jc
a. oe o
woo
-« «A
ooooooooo
vOO
oooooooooooooooo
oooooooooooooooo
oooooooooooooooo
co
UI
3CUIOOOOOOOOOOOO
OOiOOOOOOOOOOOO
t- ui
-« ac
O Q.
oo
00
oo
00
B-9
-------
A <» W
Ul 0.
ui >» ac
a. oc o
woo
o o
* K
0. OC. O
woo
o o
O Ul CO
X t-
- ic
a. ae o
woo
C£ O
o ui ro
oooooooooooooooo
oooooooooooooooo
oooooooooooooooo
w
u
&
w
^
Pi
u
o
o>
o 1C
a. ae o
(O O O
o o
-------
H
h-1
H
C/3
2;
o
H
H
O
fd
PJ
C/3
O
M
55
O
u
o
o
53
w
!=>
o-
w
P3
fn
O vA x: ooowooN-ooroo wca- oooooooooooooooo
Z H- OOOOOOOOOOOOOOOO
31 O M OOO OO OO
o ui to oooooooooooo*«ooo
z » oooooooooooooooo
M
30 O
z - oooooooooooooooo
ft: ai o
O v
C9 'M%
UI t- O UI rO OIXOOMM^lMdlMOI^-ira-MMM
-------
e
M
M
FQ
cn
P
o
M
H
U
w
CkC
o
ui a.
ui >- x:
a. oe. o
to o o
o o
o ui n
z i
«*
ui a.
ui > z
o. at o
(0 O O
ta o
a ui m
z »-
- « *
31 U
c* * v>
ui a.
ui >- c
0.0:0
to o o
(9 O
O UI CO
X t-
M x:
a. oe. o
«o o o
C9 O
A UI W
«- - ac
a. ac o
(O O O
0 O
o ui n
oooooooooo«*ooooo
oooooooooooooooo
oooooooooooooooo
oooooooooooooooo
B
o « co
§
Pi
ui >- K
a. O£ o
co o o
C9 V>
00000000*10000000
oooooooooooooooo
CO
UI
ZUIOOOOOOOOOOOOOOOO
oaeoooooooooooooooo
B-12
-------
W
o
en
O vD O O *4 «
UJ O O O O
111 >- * + I I
Q. at ui uj uj ui
O) O O O O O
o o o o o
a ui m o o o
a: i «- CD r> r>
*- -*+ I I
a. o o o o o
o o o o o
o ui to o o o
ac »- * to r> r>
- r<- »
A « o o «« «»
UI O O O O
UI >«« I I
a. CK ui ut ui ui
O O O
Z »~ « 00 Irt 10
-« « N. o w %e
31 O M «. fw ^
a « o o « -i
UI 0000
UI >- + + I I
a. oc ui ui ui ui
co o o o o o
o o o o o
O UI tf> O O O
z i- * » ID r>
>-> 4- * I I
a. ac ui ui ui ui
to o o o o o
o o o o o
o ui m o o o
x i- «- oo m v>
-
at o «« « ts. *
o w o o -< -.
UI O O O O
UI >- -f * I I
a. oe ui ui ui ui
(O O O O O O
o o o o o
O UI « O O O
3C t- * 00 10 10
-« - o n so
a O « « fs. »
UI O O O O
UI >+.« I I
a. ae ui ui ui ui
«
- - o PO vo
3 O «-> <-4 fw «
O - « + I I
a. oe ui ui ui ui
- + * I I
o. a; ui ui ui ui
(O O O O O O
o o o o o
A ui in o o o
35 »- « O W> «
i-« - + + I I
ft. OC UI UI UI UI
co o o o o o
o o <> o o
c> ui r> o. o o
z i- « «& n n
- O O «i ^<
UI O O O O
UI 3- + + I I
a. of ui ui ui ui
o o o o o
C3 O O O O
o ui «r> o o o
a: t « o» 10 «
-«
*-" CM
oc oe a: ot
o o o o
ts o o o
ui ui ui uj
»- K- »- i-
o cj o
>->>- s-
oc as o; «
o o o o
C3 t3 O O
UJ UI UI UI
» I » »-
-------
O SO O O + + I I
O. Of Ul Ul Ul UJ
-*> + + +
Q. « Ul Ul UJ Ul
(A O O O O O
u o o o o
O Ul O O O O
Z »- O O O O
M * r*
O tf» O O »» I I
a. a: ui ui ui ui
Ul
Ul
a. oe ui ui ui ui
»J
Fd
Pd Pd
H Pd
»J O
»+ I I
o. oe ui ui ui ui
o> o o o o o
til O O O O
O Ul O O O O
X *- O O O O
«- « o o o o
3 O O O W f«»
O W O O CM -«
Ul O O O O
Ul >+« I I
a. ue ui ui ui ui
(O O O O O O
til O O O O
a ui o o o o
z t- o o o o
«- « o o o r>
3 o o o n -
e» +»*«
a. oe ui ui ui ui
(O O O O O O
19 O O O O
a ui o o o o
Z I- O O O O
- * -t- «*
a. ae ui ui ui ui
» O « -* (M OJ
O »* O O »* «M
Ul O O O O
Ul >-*« I I
a. o: ui ui ui ui
* «* »
a. oc ui ui ui ui
>-> D-
CO CO CO O)
-------
ui o o o o
ui >** + *
a. at ui ut iu ui
to o o o r> o
C9 O O Ol O
o ui o o <- r>
z i o o »» cw
-« « ro to » *»
3 o « -» »« »*
w
P3
£_; x~s
H E3
28 £
W Jj
&j W
§ w
H c/3
W
ei w
H Pi
- o
H- >- CP
-< oe i\-
_j o
«- C3 O
CD Ul CO
u> o
W> O
> o
*-><>
-> oe «>
_i o
«- o ro
- o
t-i QC
-I O
CQ UJ CO
I I- OJ
^ «
(O O
o
0>
3C V- O O ON f»l
« « » » »< «
9 O <-.- »
A * » <* * n
UI O O O O
UI >«*«*
a. os ui ui ui ui
CO O O O O O
u o o ir> o
o ui o o CM m
Z H> O O IO (M
i 4C r<> r<>
;» O ~ * -« v0
a ro « ro
Z UI O O O O
O UI >»-»» +
(A O. OC UI UI UI UI
o
<0 o ui o o
Z >- O O vO N
« + + o
o ui r> ir> rw r>
z i co co a» d
i . ro
ui o o o o
ui > +* + »
a. a: ui ui ui ui
(O O O O O O
t» r> r> o o
A ui P» r- io »
z i- < * f>- r*
-> « c< c* er> »
a» o -*->->-
oc oc at at
o o o o
C3
UI UI UI UI
CC CE
O O
CC
O
CD CD CD CD
-------
co
§-
B
I
to
ui a.
ui >. ac
ouoeo
to o o
19O
A UI IO
«-< a o*
oooooooooooooooo
oooooooooooooooo
oo-ooooo-ooooooooo
oooooooooooooooo
OOOOOOOOOOOOOOOO
OOOOOOOOOOOOOOOO
OOOOOOOOOOOOOOOO
«-« OOOOOOOOOOOOOOOO
ro
PQ
X
51
a
p
i
g
to
p
o
M
O
o
s
o
w
H
B
M
H
O
Q
W
W
PM
CO
M
W
O
a
w
o
o
o
ui a.
UI S- JC
euaeo
woo
O O
»* OW09
ae
O
oo
oo
OO
oo
oo
OO
OO
OO
ooo
ooo
OOO
ooo
ooo
OOO
OOO
OOO
oo
oo
OO
oo
ooo
ooo
OOO
ooo
OO
OO
OO
OOO
OOO
OOO
oo
oo
OO
oo
oo
OO'
OO
O
oooo
oooo
OOOO
oooo
oooo
OOOO
OOOO
O O
ui
to
a.
o
ui a.
ui>x:
a. ae o
woo
o o
o ui n
oooooooooooo.oooo
OOOOOOOOOOOOOOOO
eN9»9»«**^^-is-n oo
wN.t>-f>> roroK>a>^r^OM»««or-
OOOOOOOCK O-«OO«0
ooooooooooo o~o o o o
OOOOOOOOOOOOOOOO
OOOOOOOOOOO O-O O O O
'«-«* -
oooooooooooooooo
OOOOOOOOOOOOOOOO
UI
ZUIOOOOOOOOOOOOOOOO
ooeoooooooooooooooo
Muononoiooinonoinomoirt
»- ui
B-16
-------
£SS2S£SSSo
w
sa
H
Q
O
M
H
U
o r> to
UI Q.
uj > x:
to o o
C9 0
z >-
9 0
O O O 0 O 0 O
SooSooS
OOOOMCMOOCM
O * CO OOOOOOOOOOOOOOOO
ui a. tnoooooooocMfotocf»»«r4r<)
ui > s: NOOOOoooo^rocf>o>to«o
O. Of O ^-OOOOOOOO «M|S-»«^ o o
ui i o ui ro
to *- z t-
_l >-<«*
»- 90
CD w
to
£t - JC
a. oe o
to o o
o o
O UI «
Z H-
-i - 3C
O. (K O
o> o o
oooooooooooooooo
oooooooooooooooo
oooooooooooooooo
oooooooooooooooo
oooooooooooooooo
oooooooooooooooo
oooooooooooooooo
OUIN. OOOOOOOOOOOOOOOO
3C I OOOOOOOOOOOOOOOO
to
uu
ZUIOOOOOOOOOOOOOOOO
ooeoooooooooooooooo
-oononoinoinoiooioonoin
»- ui
B-17
-------
e
H
tJ
M
PQ
H
O
H
H
W
H
n
w
P-I
w
u
I
o-
w
O SO f> OOOOOOOOOOOOOOOO
uj > jc oooooocoo'«'Osem-HO
o ui in oooooooooooooooo
ZH- OOOOOOOOOOOOOOOO
: O ** OOOOOO O O O
**
onto oooooooooooooooo
o uj m oooooooooooo»*ooo
Z I- OOOO .OOOOOOOOOOOO
a o o o o o o
o - K Nso«oosorocNr^so««inrainr<>«nm
ovoto cumcy^-sMtrers-mocNosomirecNO
tO C3 ^^ ^X ^IT Ol IO CO CD ^^ CO IO *^ Oi CK ^4 ^ta ^^ ^^
o o v> o o o ** tM m d so r^ o «* c^ m so ^
to o ui co o o o o o o o o o o 01 w w ^* *** o
z t- oooooooooooooooo
>- >1
-------
a <«
ui
CA
a.
a. ae o
co o o
(0 O
o ui tr>
o «
o o o o S o o :
H
M
nJ
M
H
en
xE" Soooooooo<
o
H
H
u
w
pi
H
Q
Q
w
w
p-l
>-
ae
o
o o
(O
K
a. ae o
(O O O
o o
O UI CO
at t-
i-* JE
o. at o
tO O O
d O
o ui ro
oo
oo
oo
oo
oo
oo
oo
oo
oo
oo
oo
oo
oo
oo
oo
oo
oo
oo
oo
oo
oo
oo
oo
oo
ooo
ooo
ooo
ooo
ooo
ooo
ooo
ooo
oo
oo
oo
oo
oo
oo
oo
oo
oo
oo
oo
oo
oo
oo
oo
oo
ooo
ooo
ooo
ooo
ooo
ooo
ooo
ooo
oooooooooooooooo
oooooooooooooooo
oooooooooooooooo
oooooooooooooooo
co
ui a.
UI >- JC
a. at o
CO O O
til O
M
O UI N-
X H-
-
-------
w
O VD O O v* »*
tu o o o o
Ul > * « I I
a. oe ui ui LU ui
(O O O O O O
(9 O O O O
a ui n o o o>
z t- * oo to in
»-» « N. o n «
a» o »- tw *
O ^ O O
UI O O O O
Ul >-*>* I I
a. oc ut ui uj ui
- « co in in
« +« I I
a. oe ui ui ui ui
(O O O O O O
C9 O O O O
o ui n o o o
z i- * o» 10 to
M »+ I I
a. ae ui ui ui ui
to o o o o o
C9 O O O O
o ui n o o o
z i- * «o n in
w « rv. o w *
a> o « ^ rw «
a * o o »- -<
ui o o o o
Ul >*> I I
a. « ui ui u ui
(O O O O O O
«J O O O O
o ui n o o o
z »- « a> r> m
«- «* I I
a. ae ui ui ui ui
(O O O O O O
t9 O O O O
o ui in o o o
ac t- * a» « in
« « r>. o n SB
o w o o ** *
Ul O O O O
Ul >«+ I I
a. oc ui ui ui ui
- «
A r*> » o -« -»
Ul O O O O
Ul >»» I I
a. oc ui ui ui ui
o »
o ui in o o o
z *- * « « in
<-i « N- o r*> %P
3» CJ »- « N. *
O »+ I I
a. tt ui ut ui ui
(O O O O O O
(9 O O O O
o ui n o o o
z »- « a> in in
»-« « N. o ro «o
«S C* O O -!-«
Ul O O O O
Ul 5- + + I I
o. ae ui ui ui ui
«o o o o o o
C9 O O O O
o ui in o o o
z »- «r a> « in
t * + I I
a. ae ui ui ui ui
o> o o o o o
cji o o o o
A ui in o o o
z »- * co m r>
1-1 «*
a» o « »i r>- "><** I I
a. ae ui ui ui ui
to o o o o o
C9 O O O O
a ui in *> o o
x t- * as » in
M - o ro so
« w v>
B-20
CD GO CD tO
-------
Ck « O O CM CM
Ul O O O O
Ul >- » * I I
O. Of. Ul Ul Ul Ul
to o o o o o
19 O O O O
O Ul O O O O
z t- o o o o
- ****
a. oc ut ui ui ui
«3 O O O O O
o o o o o
O Ul O O O O
z t- o o o o
»- ** I I
a. oc ui ui ui ui
to o o o o o
C3 O O O O
O Ul O O O O
X I O O O O
«c o o o o
3» o o o n w
r>
o o o o
o o o o
ui
Ul
O. Of Ul Ul Ul Ul
<0 O O O O O
13 O O O O
O Ul O O O O
z i- o o o o
«-
w /-s
pi s
H >
$$
w &rf
w w
H W
Di
hJ O
<| W
H P
W Q
H PL,
O P
PH ^
Ck « O O CM CM
Ul O O O O
Ul >»* I I
a. ae ui ui ui ui
CO O O O O O
o o o o o
O Ul O O O O
a: i- o o o o
>-i « O O O O
3 O O O l*> W
o ro o o CM -<
Ul O O O O
Ul >»+ I I
a. ot ui ui ui ui
«* I I
a. of ui ui ui ui
-+«* +
a. ot ui ui ut ui
w o o o o o
o o o o o
O IU O O O O
z t- o o o o
M o
O CM O O O O
Ul O O O O
Ul > + + * +
CL Of Ul Ul Ul Ul
(O O O O O O
(9 O O O O
O Ul O O O O
z »- o o o o
«- o r>
9 O «i ^ CM CM
W
O -< O O -H
Ul O O O O
Ul >-** I I
a. oe ui ui ui ui
to o o o o o
o o o o o
O UJ O /> O O
ac H- o o o o
-« a: o o m o
3 O O O -« tO
o o o o o
Ul O O O O
Ul >-«*» 4-
O. OS Ul Ul Ul Ui
(O O O O O O
ta o o o >
d Ul O O O O
Z I- O O O O
HI »-
CO CO CD CO
CO CO
B-21
-------
t- >
O O O
O V0
UI
UI >-
a. ac ui tu ui ui
10 O O O V) O
O O O ft! O
a ui o o ~ r>
3E I O O «rv CX
* « ro w -« to
«- CJ»
at ui
ac i-
CO
w
Pi
& /-N
H *Z
2§S
W i-4
CL, pq
§ t4
H W
W
Pi W
M Pi
< O
W
H O
W <3
M H
to
MI O
a> ui
- O
I- >- 00
M CK IS.
_l O
-» « O
at ui ca
ac i- CM
>- o
I >- O*
1-1 at 9*
Ui O
»-. O o
_l O
-i C9 VO
at ui a>
SB
H
PH
W /-v
O W
Pi
Pi W
W H
a
2!
M
x
a ir> * * * w
ui o o o o
ui >* + « -i-
o. ae ui ui ui ui
CO O O O ID O
O O O N O
O UI O O »- «
z i- o o 9> e*
- « n n « n
31 O « ^ v«
UI O O O O
UI 9- » + « *
a. a: ui ui ui ui
co o o o o o
u o o r> «
o ui o o o* n
5C t- O O » N
Z UI O O O O
O Ul >--+ *< +
co a. oe ui ui ui ui
O
co o ui o o CM n
Z t- O O NO CM
o M "« ^r ro r>>
UI O O O O
UI > + + * +
a. oc ui ui ui ui
CO O O O O O
O O O «f> O
Ck ui 10 n N. n
Z »- CD CO CO C4
-<
» O »- - tv. «M
a »« * to ro
ui o o o o
UI > + ««».
a. oe ui ui ui ui
(O O O O O O
ca *> ir> o o
o ui i^. N. *> r>
z K- « * N. N.
-> «»
3» O ~* -« \O «
*- CM W <#
O
CO
oe at oc. at
o o o o
O O O > 3- >-
CD CD CD fitt
-------
APPENDIX C
S02 MISSIONS INVENTORIES
C-l
-------
OC 3C
O %.
Ul
o
SO
-
ON
SO
<-
oo
oo
HP
CM SO
sO rs.
*-* V
HP 00
ro «
ro rs.
ro ro
tf) SO
o
CM
CM
m
in
sO
CM
.
o
sO
H- m
HP 00
CM -«
in
_
cr.
HP
to
o
-
o
SO
0
«
00
00
n
Ut
sO
CM
CM
O O
m in
CM CM
~ rs. rs.
CM CM CM
CM CM CM
OS
CM
CM
OTk
CM
CM
O CM IM
CM CM CM
CM CM CM
CM
CM
CM
CM
CM
CM
a.
a.
CO
Z
o
04
to
04
Ul
o
a.
oe.
OB UJ
00
_i a.
a. ui
a. w
CO
in
sO sO
OO 40
CM CM
o o o
CO O O
CM «~ -4
(^ Q^ Q^
r>- so so
O 00
in oo
m H- HP
rs. H" H-
eo rs. r*.
n to tn
m
ao
ro i*»
o
o
CM
in
CO O O
f> sO v»
CM
CM
O
O
00
N.
00
o
ro
ro
CM
o
SO
O
o
o
o
CD
oo
o
o
CM
H-
O
so
O CM
eo o «r»
O
CM
in
oo
o o
OO 00
SO vO
er> er%
rs. N.
o
o
SO
o o
oo oo
so sO
«
O t-«
3 z
o o
>- Ul
I- Of
O O 04
-
o
ro
CM
CO
UJ
oe.
UJ
o
x. a
Ul Ul
C9 i->
Ul Ul
Ul
<= Of
ce
OQ o
w*
X -
C9 «
a:
=>
CO
CO
O
o a.
Cvl
z _j z z »
oooojao ooooujo* o
CM
UJ
CM Ui
oe
co
UJ
a.
o
Z
-« o- CM ro H- co
o« O
CM
oo souJtnui
*->«iuio>-> uj
z ro -« i
w o FN. CM
»^in «r o
- o
ui z
S> «
C0 Z
o-^Ojccviujro
-»»«o:«z
a:
-------
in
o
o
to
o o
in m
o co CM
CM ro in
CM
CM
o o -* »- in in in
»o so so so in m m
o
so
CM
so
CO
IN.
CO
ro
CM
in
CM *
m o CM
in ~<
rs.
CM
SO
ro ro
in in
in m
is- ro o
m in
CM
ro in T
ro in «
m CM »«
in *
ro o
CM *
ro
o
CM
ro
esi
CM
ro
so oo
m oo
CM *-
NT
row
o oo
00 SO
roro
r*. <«
inin
o a*
CM C* in in
rororo
00 O
o *-
o>
ro
o
^
0
v-t
CM
o
*4
ON
so
CM
z m
0 sO
-« in
SO
n
CM
O
SO
CM
00
0
SO
00
fs.
in
n CM
O CM
ro -«
o o
0 is.
ro ro
00 00
o o
O ON
oo N.
rs. N.
in in
o o
CM -
m in
ro ro ro
CM ^ oo so r- - CM * ui
o cs <-< o a: ac uiso K x
C9 C3 O =0 ZZCM Ui O
x x x x to . ui
m
O
m
OO
CM o>
o o
OS -«
CM
CO
a o
x:
o
at
X >-
rsi oc
oo i ON ae
N. -J rs. ui
oa a.
C-3
-------
o
in
o
in
o
in
o
m
o
in
o
in
o o
fw in
in
in
in o
CO CJN
o
so
in
o
in
ro
o
in
oo
-* ro
« CM
in
ro
CO O
CM
SO
0ft ^&
o
ro «
o
o
o
o
CM
00
o
00
o
00
o
CO
o
00
o
00 00
o o
00
o
in
ro
o
o
o
o
o
o
o
in
ro
o
CO
CM
in
ro
ro
oo
so
CM
(M
CM
CM
f 00
N. ro
CM CM
oo
ro
CM
is. r>-
CM CM
ON
CM
CM
CM
ro
CM
CM
CM
CM
CM
CM
CM
CM
CM
CM
CM
CM
CM
CM
CM
CM
ro
CO
o
ro
CM
00
o
o
ro
00
in
ro
00
o
CM
X
0
O
n
00
in
ro
00
o
o
in
5844004
ro
00
o
«*
00
5864004
ro
00
o
m
in
5874004
ro
00
o
N-
00
O
o
SO
«
00
in
* ro
N. 00
0 0
so OS
so «O
SO SO
5990004
5965004
o
o
o
*»
O
00
o
o
so
«*
00
m
CM CM
ro e>
O 0
CM CM
» »
ON ON
6044004
6044004
CM
r>
49230<
6036004
CM
CM
O
ON
oo
N.
0
O
«*
CM
00
m
0
0
r*.
ON
CM
00
o
CM
(M
00
N.
tf)
r>
ro
ro
CO
ft.
CM
00
5783904
o
o
CM
^4
ro
00
5780304
o
0
SO
v-4
ro
00
o
00
ff»
r*.
r«-
in
o
o
«*
CM
ro
00
o
in
ON
r-.
r*.
in
oc
o
Ul
o o
ON -«
O
o
Of.
at
at
o ui M
oo > co
o
co
o o
SO t- CM
_l Ul ON
>-« oc -r-.
co
o
so
o
a: o
ui
co
oc
ui
JC
ae
o
o.
o
3:
o co
^->
. 3:
a.
ui
3 CM co ro
O CO CO
CO CM
OOOCCOOOCOOOI
x -i ac co
o cc cc cc
_J tU Ul Ul
ui at ae ac
«-« - 3- >- _l
oe ox oc oc _J
i I ui ui ui ui ui
ac co _i i- t- »- o
OO OC9OCCO»-«*«K-
-------
CM
CM
O
O
o
o
CM
CM
CM
CM
CM
CM
o *
ro
ro
CO
CM
00
o o
so so
cis o
IS. CO
IS. IS.
in in
CM
CM
^.
IS.
ro
CO
in
CM
CM
OS
in
m
in
vo
in
OS
CO
m
os
in
CM
CM
^.
rv.
ro
CO
in
CM
CM
OS
r>
so
in
so
o
CO
CO
in
os
in
CM
CM
f^.
ro
CO
n
CM
CM
OS
in
«*
rs.
$
in
&
N.
in
0s
in
CM
VO
in
ro
in
in
CM
CM
in
o
0s
r>
SO
o
ro
OS
in
OS
in
rs,
CM
VO
in
ro
in
in
CM
CM
in
o
OS
m
S
o
ro
OS
m
os
in
CM
VO
in
ro
in
m
CM
CM
in
o
OS
*>
SO
O
ro
Os
m
OS
in
o
in
os
in
*
in
m
CM
CM
«"«
in
CM
OS
SO
O
SB
ro
in
os
in
o
in
0s
in
*
in
in
CM
CM
-
in
ro
OS
SO
£
in
ro
in
os
in
o
in
os
in
T
in
in
CM
CM
-
o
«-
SO
in
V4
CM
m
OS
in
*
0
m
IS.
m
CM
CM
SO
m
IS.
CO
SO
«*
in
CM
CM
in
OS
m
*
o
m
rs.
in
CM
CM
SO
in
OS
CO
so
o
M
CM
in
0s
in
*
in
,»,
in
in
in
CM
CM
so
o
+
SO
0
M
v*
m
OS
in
*
in
_
m
m
in
(M
(M
sO
o
sO
*4
sO
in
0s
0s
^
0s
in
SO
o
CO
*
in
CM
CM
SO
in
0S
SO
SO
m
V*
IS.
in
OS
in
<£
-J
O.
CM«A»Z <*«~»*-i>«<»-i«*CMCMCMO
-------
ooooooinmm ro so » * CM
0000
Ol Ol CM CM
01 01 O
SO S0 ON
oooooooooooo r<- *
ro co
^* _ ,* _ o (M (s| 00
»* ** ** Qft 00 SO SO ^f1
ON Oft ON flft
ro n ro ro
Ol CM CM Ol
o in n oo
o ...
» CM CM CO
«4v««*v4*v4 v« O^
O ~4 »H CO
-. oo o> m
*'*'*-*'*'*sfisOsai SO SO SO so N.
CMrMCM(S|«MCMCM(MCM CM CM (M Ol Ol
CMCMCS1CMCMCMCMCMCM CM Ol Ol Ol Ol
ON O» ON «N
01 Ol 01 Of
Ol Ol 01 Ol
Ol
Ol
OJ CM CM
01 01 Ol
oooooooonmin
o
o
o
o
o
OJ
Oft ON
Ol
o
o
o o o
O3 09 r»
ro
ooooioirorooiocN
ooooo o
SOS0S0S0S0SOSOV0VO
01
so
tn
oi
SO
CM
01
SO
01
o
01
SO
o
o
01
SO
is. » n oo
IS. V) Ol Oft
O O O Oft
O O O Oft
is. rs. is. so
h> 0s 01 01
n so so so
ro «4 .oi
r^soooNoovorooioi
oo co o oft oft oft oo co oo
mnsonininininin
oo ro
in ro
ON ON
o
so
oft
oo
rs.
o
o
oo
in
in
in
in
is, o 01 «
-------
o o
Of* O»
oo ao
oo oo
in
ro
ro
to
ro
ON o^ CO 40
CO CO N* t»»
m to -«r *
ON ff* O^ O^
CM CM CM ft
CM CM CM CM
O O O O
|S« fN» 0% ON
ro ro 'r ^r
ro ON o CM
* CM ro so
M ~+ O O
O O ON ON
ro « CM
IN. CO s»
oo » m
is. IN. CO
in in
in
ro
m
oo
m
^f CMCMCMCMCMCMCMCMCMCM oo so so co oo oo mm
ro ^r *& *o* *o* ^ ^i* in m m m oo ^> ^j* r*» rs. rs. ** so
ON ro ro
ro *^«^«-«^«w^»*»^»^«-«»^ oo roro roro ro ro r*.
«*
CM CMCMCMCMCMCMCMCMCMCM CM CMCM CMCMCM CM CM
CM CMCMCMCMCMCMCMCMCMCM CM CMCM CMCMCM CMCM
» OOOOOOOOOO O OO 000 OO
is. is.is»r«>r^r^r^r«>r^rs>N. 09 co oo oo oo oo ^ ^
N CMCMCMCMCMCMCMCMCMCM ro roro rororo < <
«* *f~4<~«<~uicM
CM oc u. rsj
>
ac
-> co to
m in o m o m co
*- >- Q3 v* CO £3
ro * in oc so rs.
rs. r*» r*» o rs. rs.
C-7
-------
00
ro
ro
ro
PO
o o o o o
ro ro ro ro in
CM
in
CM
m
r-
CM
SO
in
o
o
in
tn
is. r*. r- is. is.
SO SO SO SO SO
in
CM
o -
o
CO
CM
^4^4^4M«4 CM CM SO
CMCMCMCMCM ro fO CM
CD
ro
CO
in
m
o
o
SO
is.
SO
in
ON
in
IS.
so
o o
CO CO
in in
F>fc f^
so
so
in
ON
in
so
o
ro
ro
Pw
so
o co PO o in o in CM
co CM ON * CM m «o ~
sosouois.p^ P- r>> P>-
inininmm in to in
sOsO«OsOsO so SO so
o
o
SO
o
IS.
SO
0%
m
o
CM
in
in in in
M «^ CM
CM CM CM
tw fs. IS.
in in tn
in
CM
CM
in
o
o
«*
IS.
ON
in
inotnis-so ro cr> o
* in in in in -« <« -«
-«-<-<-^ CM CM CM
ro ro ro ro ro ro ro ro
ON ON ON ON ON Oft ON OS
tnininintn in in in
o
CO
V0
in
ffv
UJ
O
«e
acxtuso
OCIS.UCM
=> «
o
cr>
o
ON
* SO
000000*0*0
zoocoo9coco->'cr>
-i ..... ro-in-
O
co
cc
Ul
cc
oc
Ul
oe
to
co
V>
oe
ui
CO
oe
oe
co -i -J i
-«-<>->cM
.jcococococoocisocisoas
X
o
-
ox
t-ro
-------
IS.
CM
ro
IS.
-i ro
o os
o
00
o
ro
ON
ro
n ro
ro
10 in
ro ro ro
ro ro fo
in in in
CM
so
CM
CM
CM
(M
O
CM
CM
CM CM
CM CM
CM CM
CO
O
ro
so so
n in
CM
ON
CM
CM
O ««
ro r*.
CM ro
ON en ON
CM CM (M
CM CM CM
ro
CO
o
o
-
ox
UI
ro
CM
O>
oe
UJ
* »-t in
a os o T«
00 -TH 00 t
co
UJ
O. -<
X:\OCB
uj as => o
»-<-
o«iN«-i
«KOZ
COCM3
eo
-------
v« <~4 w4
o o totoioinooooooooooooiocacM moo o 4v*«H*>«<*>«v*v*tvH«»f«t>OO POIO PO OO
po PO PO PO PO PO PO PO PO PO PO PO PO PO PO PO PO PO ^tr 10 10 CM PO CM PO
oo mianoooooooooooooocacM po to o o
^W CO 4^ ^^ 4^ 4^ ^^ ^^ ^^ X^ *** **H ** *"4 9^ W4 «>4 V4 «M ^^ ^v P^ fw ^^ ^J
PO PO PO PO PO tO tO tO tO tO tO tO tO fO fO tO tO to tO tO tO tO tO *" to
* *r ^«r'*'*'«-'*ir'«'<«-<«''*-i»*''»i<*-'*'<- T^r « *
<«* *>«''*i*' ON
m m
OOC0POPOPOPOOOOOOOOOOOOO*«POO CO * SO CM
(O
PO O CM
CM CM
- CM -< I- « _J VO Ul SCO
* _J >- -J CM pg
CA
:*
O
_l CMCMCMrMCMCMCMCMCMCMCMCMCMCMCMCMCMCMCMPOCMCMCOCM«4CM ZCMO.CM
C-10
-------
CM
in
CO CO
SO SO
<"* CO CO « OS CO ON
os ro ro os o^ ro os
CO CO CO CO
SO VO SO NO
CO
SO
in
r>
in m in
rs.ro so ro CM m ON v* o
o is. rw
O r»
r» so
CM
is. is- rw is. is. is. rw is. is. rw rwrwrwrw rw co co
rw rw is, is.is.is.is.rwis.rw rwrwrwrw rw co co
« <»«» ««*»«»<**« * **<«< * in r»
ro so so
ro -« -«
in so so
is. fv N- ON ON ON ON os o\ ON os ON es CN
« so in-«-'«fooinsoujososrs.so co o o
« so rwoooeor-cMosocosooso -< ON ro
......... oc .......
ro-« ro-« »- MBCOe
O Ul XUIUI
U. > _l _l
CM ro *invofwoooNoo£-^cMro'«-» m >->*>-« rw
-«-'«>cM<«-««- tn
oc rw of
ui ro ui
J -J
.4 is. i»«
O CM O
oa eo
9- 3-
Ul UI
_l O _l
** SO H*
Of CM ac
C-ll
-------
ON
so
so
CO
ON
CM
ca
so
o
* o o in
in in
ON ON
o
SO
o
CM
so
ON
so
**
SO
ro
ON
in
o
CM
SO
SO
in
in
in
CM
~ ON
SO 00
* o
r>- r- ON
CO
CO
o
* * o
in in in
+ * ro
ro ro ro
us so «
N. o
co rs.
« o
*-> o
«o rs.
in
CM
sO
in
SO
CM
in
in
CM
CM
in
ro
CM
in
ro
CM
in
0
^
SO
O
0
ON
ro
rs.
o
SO
»4
ON
00
m
o
ON
SO
O
SO
is.
ro
IS.
O
SO
CM
ON
09
in
O
SO
*
O
CM
n
in
IS,
O
00
r-
|S-
09
n
o
SO
"*
o
CM
m
in
rs.
o
CO
rs.
rs.
00
m
o
in
o
o
in
N.
m
CM
ON
in
so
IS-
so
CM
r-
CM
rs.
ON
in
to
ae
ac
C3
ac
co
ui
ro
i
CO
oe
ui
uiinoeinoo rs. so»->oooocMin _i rs. u> ON m ON « oNoecMCMsoior- ui *- to co
CMt-OOOOInI IK- UI
I- CO >- _J _J oc ac x: - zsOOsOsOIS.OCOaO (/>
-------
o
ro
in
tr>
in
in
m
in
in
ro
o
SO
in
o
SO
in
o
SO
m
so
*.
so
IS.
sO
is.
sO
r*.
in
ro
CM
in
ro
CM
in
ro
(M
in
ro
OS
CM
01
OS
CM
CM
os
CM
CM
OS
CM
CM
CM
CM
OS
CM
CM
OS
CM
CM
OS
CM
CM
OS
CM
CM
OS
CM
CM
CM
CM
in
O O
ro «
o
o
o
OS
o
«s
o
OS
o
OS
o
OS
o
OS
IS.
in
o
o
ro
CO
oo
in
o
o
o
00
v*4
OS
00
in
o
«*
»*
in
00
oo
in
o
OS
fO
O
o
V*
OS
CO
in
0
V*
O
OS
SO
O
in
in
to
os
m
o
.*
o
OS
SO
o
is.
m
CO
os
in
o
CM
O
OS
SO
O
in
in
00
o>
in
o
CM
O
OS
SO
O
SO
m
00
Os
in
0
ro
o
OS
SO
O
^f
m
00
OS
in
OS
ro
o
OS
so
o
so
in
00
OS
in
o
ig-
o
OS
SO
O
in
in
00
cr>
in
0
^r
o
OS
so
O
SO
m
00
Os
in
o
in
o
OS
SO
o
(f
m
00
OS
m
o
SO
o
Os
SO
o
5*
V)
00
os
in
Ul
o
CO
U. SO
*- o
CM t
o a:
CM =>
CM
CO O
U. Ul
O X O
o - oo
CM Ul CM JC CM
CO
Ul
=>
o
CO
CO
« -CM
ro
so
Ul
Ul
Ul
Ul
Ul
Ul
Ul
Ul
Ul
Ul
CO Os CO OS CO OS CO OS CO OS CO OS CO OS CO OS CO OS CO OS
CslCMCMCMCMCMCMCMCMrM
C-13
-------
« vo rororoin ««cM«ooflococococo CMCM CM o
OP* 0^ *"** 4^*4 **** ^^ SO 0^ CVJ CVj O^ O1* C^ O^ flfr CP* C^ m CNI IO
ro CM CM CM
O O O O O *
ON ro ro ro CM co
«"4 <^>4 w-4 «"« »« «-4 * «* O *^ «**
CM N. rororocrv rorv.r-vocooororororo
vo vo ^..^.^^ roo^^rorororo
co ^t ^ ^4 « ^
oo ro VOVOVOCM ovoooroinnmninin
CM
00 '
in
r>
ro cr> CM CM CM CM
ro 01 po r^ ro fo
CM CM CM CM CM CM
oooooooooo to in eo « ro
ro ro r*) PO ro f*> f^ p*> ro fo fo ro ro to PO
CMCMCMCMCMCMCMCMCMCM CM CM CM CM (M
O O O O
5M»<<-««^«-«« OO
o
CO
tr>
CM -<
in in in o
nnconnnnnnn «-
to
ui
o
oe
=>
o
to
ui
m o oooo OOVON.OOOOOO CM CM
ort in mmmm ^'^^^'^^'^'v^f- roro
oooo « *
ro o
oo o
CM o ro ro ro ro
m CM r*^ r*« r^* r^«
vO ON OO OO QO 00
in 10 *n in ir> in
V O O O O IO O CM CM
F^ is» 0rs C4 ^i* so 0^ v) (rt
l N* f^
inminnminininnini-inin
ce
o
in
ro
in
ro
to
CM
in
in
CM
V0
oo
ro
in
00
in
in
vo
ro
ON
in
ui
ac
t-
-4=9CMUICMCMroro
(O
CO
UI
z
o
Ul
Ul
CO
X
C3
=>
CO
CO
I
I
a.
ui
ui
ui
ui
i-
-------
o
PO
-
co
CM
in
*"
ro
ro
m
OS
M
CM
SO
SO
»
10
ro
f;
o
ro
ro
in
OS
<^
CM
CO
CM
CM
CM
n
00
"*
CM
ro
CM
CM
SO
IS.
so
in
OS
*
«-l
ro
*f±
Mr
m
00
^0
-
OS
Cf»
rs.
n
ro
r>-
*
CM
ro
CM
o
«*
so
so
in
rs.
£
o
M
ro
*
n
m
^f
~
^
'-
CM
*
r*.
N.
m
CM
ro
CM
so
CO
so
OS
OS
O
OS
m
in
_*,
^r
«
^
r»-
CM
*
r*.
rs.
**
CM
ro
CM
so
40
so
OS
IS.
OS
O
o
SO
f
n
to
VO
~
SO
SO
OS
Is.
fv.
in
CM
ro
CM
CM
SO
r*.
OS
£
0
e»
in
*&
W
r>
VO
10
CM
rs.
CM
00
r>
ro
CM
ro
CM
ro
*
SO
OS
£
O
OS
r>
in
so
CM
N-
01
CO
r>
ro
CM
ro
CM
ro
*
»
OS
IS.
5
O
OS
in
m
SO
CM
OS
ro
CM
CM
OS
CM
ro
CM
ro
ro
in
m
OS
o
CM
£
m
so
CM
OS
ro
CM
CM
OS
CM
ro
CM
ro
ro
in
00
in
5
o
CM
2
in
SO
CM
OS
ro
CM
CM
OS
CM
ro
CM
ro
ro
m
in
OS
o
CM
?
in
so
CM
OS
ro
CM
CM
os
CM
ro
CM
ro
ro
in
SO
m
Cf>
o
CM
?
in
so
CM
OS
ro
CM
CM
or*
CM
ro
CM
ro
ro
in
in
in
IS.
OS
O
CM
5
in
SO
CM
OS
ro
CM
CM
OS
CM
ro
CM
ro
ro
in
in
fs-
CJS
O
CM
*
in
o
co
ut
Ul
CD
CM
ro
in
SO
oo
-i
ro so -H ui ro << v oc zc a_ o-«a.--ici.»-.a.-~a.«a.
«» in _i CM »- H- ui ui
** CO CO D£ 0£ C9 O C9 C9 C9 CJ O
oco t- zzzzxzz
Ul CO 'CO . _ M 1-1 ._. l-i i-l «-i
O O O O Ul O I O I O I O I O I OOOO<-«O«*O~4O-«O~4O
coocaozo
-------
CO
IN.
oo
to
rs.
oo
IN.
oo
r- o
oo ro
o
ro
-+ in
ON CM
03
N.
CO
IN.
09
NO
CM C
ON 0
ro r1
»* V
M CM O
N ON ^
o ro IN.
4 9^ v^
o o o
* ^»- FN.
IN. IN. in
~* ~*
Tf ^.
in in
^ ^
ro ro
^c
in
ro
CM
00
CM
i-l
CM
00
CM
r-l
CM
OO
(M
w
00
CM
«>>
CM
CM
ro
CM
CM
ro
CM
CM
ro
CM
CM
ro
CM
CM
ro
CM
CM
ro
CM
CM
ro
CM
CM
ro
CM
CM
ro
CM
CM
ro
CM
CM
ro
CM
CM
ro
CM
CM
ro
CM
CM
ro
CM
CM
ro
CM
n»
ro
m
ro
m
9>
o
CM
<«
SO
in
i*»
ro
m
CM
in
ON
o
CM
f
NO
m
r»»
to
n
^
m
ON
o
CM
ro
NO
in
n
ro
in
at
*
ON
O
CM
ro
Sft
in
n
ro
m
OK
»
ON
O
CM
ro
NO
in
r«?
ro
in
00
««
er»
o
CM
ro
NO
in
*>
,4
NO
V*
«<
ON
O
m
*
SO
in
«?
M
SO
CM
»
Oh
O
in
*
so
in
oo
ro
«
NO
in
ON
O
CM
ro
NO
in
oo
ro
o
o
00
o
r-
*
NO
in
00
ro
<*4
in
w
o
so
CM
VO
in
OB
ro
ON
*
o
NO
CM
NO
m
00
ro
NO
«
0
NO
CM
NO
in
00
ro
«
in
o
CM
ro
NO
in
oo
ro
CM
in
o
CM
ro
NO
in
09
ro
n
in
ON
o
CM
ro
NO
in
v
N.
CM
CM
If)
ON
0
r-
CM
VO
m
« CM
ON ^
-* CM
UJ
o
oe
CM
O
-«ro>-IN.»OO»ONaO -
CM I CM<4-CMOCMOCMOCM<«-CM'«'CM'*>rO.jrO_l
o
-------
SO
IV.
in
rs.
ro
« ro
as in
-* CM
o so
oo
o
CO SO
in m
o
ro
ro
o
ro
ro
o
ro
ro
os
ON
ro
m
rs.
in
o
in
so is-
ro is.
m in
CM
ro
CM
CM
ro
rs.
- is.
CM CM
in in
CM CM
in m
rs. r-
oo eo
so so
in m
m in
o
CM
sO
O
0
m
^
o
ro
so
^f
m
m
o
CM
SO
O
O
n
^*
o
ro
SO
^f
m
m
o
CM
SO
O
O
«
^f
0
ro
SO
^
m
m
o
CM
sO
0
O
m
*^
0
ro
SO
^f
m
m
n
m
CM
m
_
r>
^j*
SO
CM
in
in
m
m
m
ro
m
OS
»
^*
< ro
so oo
* CM
* m
ro so
so m
CM CM
CM CM
ro ro
ro ro
* «4
n tn
^M, §
^^ ^T
00 09
OS OS
^ *
sO so
m m
m m
m
IS.
OS
OS
ro
«te
^r
o
rv.
ro
m
m
o
00
m
in
*
m
ro
CM
so
m
O9
CM
OS
00
Ul
o
ui
111
o
oo
SO Z O
M Ul SO
CM
o
CM
o
CM
o
zoo
=» oo or
cc o.
CM *-UIO SO O
roo
ooe
m
CM UJ -*
co -« cj m
OE
to
ro-ocroujcooo
-ot oacroro
in in in in
o
O
o
CD
CD
x:
*
-------
so
CM
CM
CM
o
ro
CM
CM
ro
in
oo
CM
CM
r>.
r«-
so «« so
CM IS- *»
o
o
in
o
O
OS
o
OS
o
SO
* CM
N.
ro
in
in
o
CM
OS
in
o
CM
in
o
CM
O
OS
O -O
ff» OS
CM CM
SO SO
ro ro
ro
r>. m ro
is, <« ro
» in in
oo
so o
SO «
ro *
os so so o o
SO IS. OS TT *
. ^4 CM fw IS.
so ro r» « i
o CM «* » *
# <« * ro ro
CO
so
o
CM
O
00
CM
in
ro
CM
CM
CM r>.
CM *
« ro
os
esj
CM
OS
CM
CM
Os
CM
CM
OS
CM
CM
ro
CM
CM
ro
CM
CO
ro
CM
CO
ro
CM
CO
ro
CM
oo
ro
CM
OO OS OS
ro « ««
CM CM CM
CM
SO
ro
CM
OO
00
in
oo
in
in
n
in
in
in
m
o ro
OS OO
esj -*
oo
so
oo
to
*
in
ro
CM
so
in
o
00
in
o
in
*
n
o o *
oo oo os
in in M
*« so
O O CO
» « T
ro
CM
in
in
oo «-
ro ro ro in ro
CM CM CM r<- r-
so so so to os
in m in m in
is. is. is. is. r- rs. is.
vfi ^O *St ^O *ff ^0 vO
is. is. is. fs. is. O O
4 »i »-« _i «i QO OO
ro ro ro ro ro os os
in m in in in ««
* * * * «* » *
in in in in m os os
o o o o o *««
«H o^ ~+ »-l < O O
ro ro ro ro ro os os
m in m in m t «*
m in in m in mm
«-* oo
oo CM
-* CM
o o
m in
** «*
ro o
CM so
CM M
in so
m m
m in
to
tu
i- a:
at
=>
m
tU
x
so z o
«- -
*kroui 3 -H o CM o
z oe. z co
CO UI C9 UI
Of K Ot Z -J
CO
ui
at
ui
O
O.
a:
z
ui
o
so
SO
ro
o
o
o
o
ui
i i
u.
CO
z
-------
IS.
CM
r-
ro
*+
CM
00
so
CO
ON
tf>
o o
ro ro
CM CM
O O
o in
SO CO
* CM
o
SO
XT
00
ro
«*
SO
SO
ON
O O
in CM
O is,
r.. SO
O CM
CM «-
V)
«*
ro
o
in
ro
o o
ro in
-« oo
CM oo
CM
ro
ro
in
o
in
in os
o ro
CM
ON
ON
CO
o
in
o
CO
o
<« o *+
CM
ro
V0
so
ro
o
ro
ro
CM CM
ro ro
ro ro
o
CM
CM
ro ro
ro ro
CM
CM
ro
CM
o
in
*
in
so
o
SO
in
in
Is.
SO
O
in
oo
in
ro
in
*
o
so
00
CM
*
in
is.
so
o
in
oo
in
ro
m
*
o
so
00
CM
*
10
QC
o
(O
O 0
0 0
fs. fs.
SO SO
ON ON
ON O*
«4 «*
ro ro
to in
ON ON
r- r--
ON CN
CD CD
CM CM
so SO
o
in
CD
CO
ON
o
ro
in
r-
ON
o>
CM
ro
so
o
^!
^
CM
r»
n
4
CM
O
in
*
*r
«*
o
it
SO
O
*!
^
CM
r-
»»
_l
SO
00
*
SO
in
is.
CM
in
so
0
in
0
ro
ON
CM
ON
4
IS.
*
CM
00
in
*
sO
SO
O O
9> ON
SO SO
CM CM
0 0
sO SO
CM CM
00 CO
CO 00
sO sO
SO SO
SO so
*
ro
«*
CM
O
M-
O
CM
^4
*
O
r>
in
^
an
m
o o
m in
ro »^
ro ON
ON CO
ro ro
w f
* *
ON *
00 <«
CM in
ON O
-4 CM
SO sO
(A
3C
Of
Ul
Ul
ro
i£ ON
U O
ec CM
on
Ul
m ro
CM ro
CM CM
m m
o o
o ui
eo »- i-
o « «n M oo i-i ON
O -* U4 Z -< 3C -4
if in at =3 in o in
in mm
ui
a*
o
o.
C9
x
o
oc
CO
JC
Of
o * =3 *
o - Z -4 Z -4
ui « o in o so
ic o z o cj o
or
ui
or o o ai in
Ul O
2» «# O OL CM
C3 m o*
0. 00 O N.
-j in
O U
or ***
H CM I >4
Ul is. N. ON ON _J
-------
o
os
o
CO
in
(M
oo
in in
01
m n rs.
o o r».
in m TT-
in
.-i os
in M
o in so os
CM o is- o
o
ro
oo
inin
>-« «- c>i
to
CM
so »»
so in
ro CM
C3
Z -.
co
oo os os
O OS OS
so
inin
into
in
so
in
(M
ro ^ oooo oo r>>is.
oo o »«^^« o> oo
CM (M CMCMCMCM - (MCM
O
O.
o
ro -<
in so
ro
is,
m
in
ON
so
m
o
CM
o
o
CM
r>
o
in so o
CM r». CM
co rs. «
rs. o «
in so so
tf>
UI
O
oc.
x.
o
Q£
UI
»-
(O
^ 00 l-l N.
ro z n o CM
Q£
o >
eo ui
ui -i
z z
-I Z ~4 UJ
O 3 -< 3C CM
^ CO - CO -H
so VO VO
CO
CO
- CO
ui
OQ x
JC O
UI >-i
3- C9
O UI
Z O£
m ««
o
o
« o
o at
z -
O U CO
-< UI Z
C3 OS O
UI M
oe >- t-
to
JC >- Z
o :> o
or ui o
u. o: ^
so
SO
O
o
ON
00
so
o
o
CM
ro
ro
in
eo + f os r> os
os in in in o in
** i ^4 CM ro CM
in in
os
ors-ors-oooo
O-S»ON.OOOO
romconnminn
O 00
O "» O O
in z r- rs.
^ ^s. *^ \ in m m m rs. 9 m in
OOv^CD*^- uimco-H^cij coooio
tt 2»CDi~iSOsOCMOO»-iOSi-i
ui
C9
CO
o ro o
o o
UI
UI
O^OOOOOOOOOOOOO
00 00 CO OOOQOOCD 09 OO OO
C-20
-------
m sososo in inininin ooooinininmin inoooooooooo
*- ro ro so ^> co co oo oo r*- CM CM CM r*. o ^> o is. ro ^> ^> ^> ^> o o o o os os
»« <**« so rorororo voosr^cois.~*-«»'«o is-mmininininoDoocooo
os oooo is. M _i »« .« soroso-a-sdit-'q-^-tn CMCM«MCMCMCMCM«OOO<«>'<«>
-in ....
nf IS. IS. CM SO SOSOSOSO IS-CM«-iCMCMSOSOsOro sOCMCMCMCMCMCMCMCM»-t*"<
ro os os o ** Tt" ro fo fo so ** *-* i* -1-4 9-t ** »* i-t
CM CO OO 'T-
IS- 00 oo * in * * <* * f^'O'"*''*''*'1*'*''* *<»«'«'<»»»'***»'*»
O OSOSO OOOO OOOOOOOOO OOOOOOOOOOO
CM -MMCM CM CMCMCMCM CMCMCMCMCMCMCMCMCM CMCMtMCMCMCMCMCMCMCMCM
m CM CM ^t* os f*. r» rs. is. r^ so so so so so ^* so so *^ os os os os os os os os ro ro
is. m in r^« *^ ^^ ^r1 *^ ^h r*. ro ro ro ^^ *^ *^ ^^ ^r os ^^ ^^ ^^ ^^ ^ ^ *^ *t* ro ro
^» ^> t^f fo ^o fo ro fo fo fo fo fo fo fo
<««'«>»* « «*«'« ^t-'*'T<*'>r'«r'»>'<>'* '*i'«'*'*<«><*'«'^r<*^r^-
* v1*1* * *»<'« »»***>«*»» »«<'»**<»«»>'»
O£>OOO O OOOO OOOOOOOOO OOOOOOOOOOO
OUIOOO O OOOO OOOOOOOOO OOOOOOOOOOO
os co o o o is. r«.r>.N.is. intnininininmmin vosoi^sososcsosovosovo
OSi-iOOO fO CMCMCMCM CMCMCMCMCMCMCMCMCM **<'*'«««'*'«<«**
CM > ro ro ro ro ro ro ro ro ro ro ro ro ro ro ro ro ro ro ro ro ro ro ro ro fo ro ro ro
inuiintnin in m in in in ininininininininin minminininininininin
S-'
CO
-------
ooo r»- is- * m om
o
o
o
ro
o
o
o
in
co co co co CM o> o is- in in
«*'»'<» o> so roo «-« »* «
o
CM
*-4 «-l ** SO V*
CM <*
CM O CO vO SO
-« * ro CM CM
o
in
so
o
o
in
o
r-
*»'«* O O *-« SO ^H «-«»-
««-<« inin «-i i so so so
so so so «« mm in in in
«<* ^f in in CM CM ***
O O O O O OO ~t ~4 ~«
CM CM CM CM CM CM CM ro ro ro
ro
*
^>
o
0
l\-
*
CM
in
^%
z
31
O
Ct
1
3
z
co
s^»
O
O
ro
o
CM
in
<
o
o
o>
in
ro
in
o
o
ro
o
CM
m
«r
o
o
o>
in
ro
in
o
o
ro
o
CM
in
*
o
o
ON
in
ro
in
£0
O
UI
OC
CO
UI z
o o -«
z z o
N. o r- o o
V. O >. O O
in o in o o
x, is. > r«- r»
o o o o
UI O UI O
to r~ o> r-
>_i SO >- vO
> -< » «
ui m ui m
O
o
so
tn
O
o
CO
CO
in
o
o
ro
ro
ro
r>
UJ
_j
_i
»-«
=»
o
cc
co
co
at
x
Z I-
l~ O *+
UI C9
a ui z
O£ O
03 > I-
CO -«
to
oc
o
o>
on
UI
on
111
CO
oe
ui
cocoooivomiSOCM ooo
i « m i CM oo
C0 C* M
z z o
UI UI
CO 09 OO 00 CO OO O3 O9 OO CO
co o » a:
UI C£ UI O
an a. of o
o o H
GO aa u. _i
o a _i o
ui ui
-------
00
in
r-
m
o
SO
ro
ro
IO
ro
CM
CO tO
ro so
CM »i
in
in
ON
XT
ON
in
m
IS-
O
CM
ON
««
(S.
O
m
o
SO
CM
v>4
N.
O
in
0
0>
*
«
N.
o
in
o
ON
«*
<-<
IS.
O
in
o
CM
v-4
VH
ro
ro
so
o
«
**
CM
ro
ro
m
o
**
J
CM
ro
ro
in
in
«*
|sj
ro
oo
in
so
'*
ro
so
«
in
00
so
SO
r>.
CM
ro
o o
-* ON
ro in
«- 0
-
O ON ON
CM -«-«
O
CM
CM
ro
CM
ro
CM
CM CM
m
CM
CM
CM M
CM
ro ro
oo oo
SO
ro
o
o
ON
ro
(M
in
oo
ON
0
0
m
«<4
m
00
o
o
in
*
m
00
OS
0
to
V*
w*)
v>
oo
0
10
**
«-^
10
ro
5308001
10
o
o
ON
ro
in
in
ON
0
o
ON
ro
in
in
ON
o
0
ON
«*
ro
in
in
ON
0
O
ON
ro
r>
o
0
0
m
0 0
+* ~*
0 0
CM ro
ro ro
m in
in m
«N ON ON
s. in v. in N.
«- SO »- SO
O O A O O
Ul O Ul O Ul
o> « tn * o>
i i ro i i ro >-«
> ro » ro >
ui m ui in ui
oc. a: oc
ac
ui
o
CD
oe
-i > »^^-^-i^ »<
a.
o
o
Of ro 3 Ul CM « to O O
l~ ^* Ul O£ O£ O O
a oo oe ui ui
minvcM K a. CM ro
SO sO sO SO |S.
-------
00
SO
o
CM
in
CM
oo oo in ro
ro o in o
»* in
00 o O
in so is.
o
CM in *-
ro *
CM
os *» is. 4
CO * N. *
00
00
co
CM
CM
in
in
is-
ro
ro
in
o
r-
os
is.
o
in
ID
o
so
ro
CM
so
ro
00
ro
so
ro
CM
OK
o
is.
ro
0"
ON
o o
IS. ~4
Is.
ro
«-« «~i »« CM
so so va CM
m in in in
CO
o
CM
r>
o
CM
O
CM
* * O
O O «
CM CM CM
ON ON o o is. s. r- N. r^ is.
OS M OOO O O O O O
* CM CMCMCM CMCMCMCMCM
m -i
in
os -«
SO 00
* ro
oo
o r». o N- o o o
o "s, o N. o o is,
m in o in o o so
in N. in *s. m in oo
V0 ** so <- SO so o
OS
is. o
N.O
incM
O
ro
ON Oft
rs-ors-o
N.ON.O
in «J> in vo
eo
rs. o rs. o r»- o o o
S. O N. O "s, 0 O O
m -> ro ro in »-i-*
ro>ro3»ro ro t > -<
in ui in ui in in in ui in
oe. of. oc
o oooo o oooooo o o
O Ul O Ul O O Ul O Ul O Ul O O O
oo co so to so is. to tn co ON to ON os os
ro MCMMCM CM o »- o »-t o o o
ftl 3** ^^ !^ «^i *^ 3^ P*> 1^ r*> «^ PO fO fJ
m ui in ui in in ui m ui w ui in in m
oc oc ac at oc
*^ s^ ********
to
on
oo ui ON ac.
CM os ro ~>
**
- i-iO
UI
Of
UI
ui o ui ui >
ac N. cc oc o
o _J so -< o «- o o o
>- 4C U. U. Ul
in ui o a: < in =)
u.
ui
a:
UI
Q«:
O
oo
UI
oQc
Ou »- -^ »
O
CM
-«
in
»^
OS
ro
in
»^
OS
in
*"
OS
in
«"«
OS
oOi:
-Ul
v0»-
-------
ro
in
is.
in
in o
in «-
SO
o
CO
ro
CM
in
o
o
in
fv.
SO
o
CO
CM
ro
o o
(M
O
SO
o
o
CO
* in
in
ro
in
OP.
O SO
O SO
IS. CO
ro
ro
so
so
00
in
(M
0*
ro
cr»
ro
or.
ro
ro
ro
in
SO 4^
so o
oo *
is. is-
o o
(M
in
CM
oo
5f
CM
00
^
CM
CM
ro
CM
CM
ro
CM
CM
ro
CM
CO
ro
is.
so
CM
00
ro
CM
CM
ro
CM
SO
CM
CM
SO
CM
SO
IS.
CM
m
CM
o
o
CM
so
OS
o
o
CM
SO
- co
- -
9%
0001
«l
QTS
osco
o
CM
CM
ON
CM _l
-
ON
CM
M W CM
CM CM
CM CM
ON OS
ui
ui oe
Vi? tift
O U,
CJ O£
Ul
e> <=>
^g ^>* ^g
-«
1-4 Or
Ul
V> I
Z l-
9
_i oe
u, uj
-------
o
CXI
tn
rs.
ro
o
CM
tn
in
in
to o
so ro
OS
CM f
CM -*
in
in
in
in
in
in
o
o
o
o
OS
CM
CM
O
o
N.
in
CM
CM
CM
SO
CM
ro
ro
so
so
so
o
o
o
in
o
ro
o
ro
in
ro
ro
to in
or» «
in rs.
CM ro
v » V. O
so tn so in so
in N. in N* in
SO »- SO " SO
O
OS
in
so
o
o
CO
ro
ro
in
o
o
CO
ro
ro
m
o
o
co
ro
ro
co
ro
ro
o
o
co
ro
ro
in
o
o
ro
*
ro
in
o
o
*
ro
CM
in
o
o
OS
CM
CM
in
o
o
in
ro
ro
in
o a o a o
O UJ O Ul O
in to so to so
ro 1-4 ro -> ro
ro > ro > ro
in ui in ui in
CK os.
o
o
in
ro
ro
in
»-
fie CC
Ul Ul
-J
00 I- 3 OC
H. o co - _i
SC O. H- « « OS -
ootooini OOSOCOUIOG
OQ -oui >-i o =3 t ca
oszoe>- CO *
- Z I Z - tO
3 »_! 1-1 -< Ul 00 «*
o
u.
C9
Of
Ul
oc
ae
ui
Of
oc
Ul
en
cm
o'-o
o
co
i
so
I
U)
ac ac
ui in ut
t- »
»- CM »
-------
VO
CM
CM ~
O
o
CO
in
vo
o
o
PS.
in
vo
o o
o o
VO CM
in in
vo vo
0S 0S OS.
o N. o fs- o rs. o
O -V. O N. O V. O
« in * in * in *
<* V. * N. * N. *
O
O
o
00
ro
0S
o N. o
o N. o
o in o
ro >s. ro
o
o
CM
rs.
o
o
CM
IV.
O
O
ro
rs.
o
o
o
ro
N.
o
o
in
ro
ro
in
o
o
in
ro
ro
in
o
o
in
ro
ro
in
o
o
in
ro
ro
in
O O O C» O O O O O O C, O O O O O
OUIOUIOUIO O O O UI O O O O O
rocorotorocoro ro oo o co o o « o o
,*._*,-. v -* * * in^in r> in n in
ro>-ro>-ro>-ro ro ro ro»ro ro ro ro ro
inuimuiinujin to in nuin in in in in
O£ OJf O£ O£
at
>-.
u.
co
ui
o
oc
o
CO
A CO O
X X
3
u.
ui
o.
CO
ui ca
a.inz
CO -
CM »- _J
Cf> CM
CM «k
< * ro
CM
UI
Z
z
CM
CT>
CM
ro iv. ~t
Oft OS O
CM CM ro
ro
ro
0S
m
ro
OS
vo
ro
0S
# * *
- CM
r> r>
ro ro
0S 0V
ro
in
ro
0s
m
ro
OS
in
in
ro
0S
vo
in
ro
rs.
in
ro
0S
oo
m
ro
0S
C-27
-------
10
O
fO
CO
O
CM
o
CM
n
o
r>
so
so
ro
o
CM
o
CM
CM
ro
CM
o
ro
ro
10
CM
o
o
o
ro
o
o
o
ro
o
o
CM
o
to
to
in
o
o
o
IO
ro
o
o
-«
ID
ro
v>
on
IO
ro
SO
ro
ro
cr*
C-28
-------
Of X
ro
_i co
O N.
> X
O
SO
CM sO
so r>-
o ^ m
CM * 00
os »<
so * co
--< ro *-4
CM
r»
CM -H
r> to.
O
CM
SO
SO SO ^
I-- O Or>
O
OS
tn in CM t-
so co os CM
CSJ * »l O
O
O
CO
CM -«
SO N.
-« 00 fO
CM -i
O
O
X
CO
o
cv
-i ca >
> JE o
Z Ul O
Ul I
v. a. r>
o ui o
a. . -- Z
13 Z O
CO Ui *-
co x ui
> Ul Ul
t QC OC
z
=>
o
CJ
>
z
Ul
C9
Ul
CM
CO
to
o o
m 10
(SJ CM
so so
00 00
CM
10
CO
If)
o o
o o
in tf>
CM
CM
CM
00
o
o
SO
IS.
N.
o
o
CM
r>
CO
r>
CM csj
CM CSJ
r«> ro
so so
o o
* *
^ (s-
CM N.
CO O
o o
O Ul
ro co
00 "
Is. »
in ui
on
ui
. o<
Ul 31 I- O K-H-UI
Of CO O Z _l Z Z H-
COO N. O OO.OOUO_ICO OOOOUIO*-
os co o i- CM ntn>-«o. os «-<** x co |