Final Report
TRENDS IN SULFUR DIOXIDE EMISSIONS FROM
THE ELECTRIC UTILITY INDUSTRY AND AMBIENT
SULFUR DIOXIDE CONCENTRATIONS IN THE NORTH-
EASTERN UNITED STATES, 1975 TO 1982
SYSAPP 85/011
15 January 1985
-------
Final Report
TRENDS IN SULFUR DIOXIDE EMISSIONS FROM
THE ELECTRIC UTILITY INDUSTRY AND AMBIENT
SULFUR DIOXIDE CONCENTRATIONS IN THE NORTH-
EASTERN UNITED STATES, 1975 TO 1982
SYSAPP 85/011
15 January 1985
Prepared for
Terry L. Clark
U.S. Environmental Protection Agency
Meteorology and Assessment Division
Atmospheric Sciences Research Laboratory
Research Triangle Park, North Carolina 27711
Contract 68-01-6614
Prepared by
Alison K. Pollack
C. Shepherd Burton
Systems Applications, Inc.
101 Lucas Valley Road
San Rafael, California 94903
110600ef81*16'tR
-------
ACKNOWLEDGEMENTS
We are grateful to Mithra Moezzi and Gary Lundberg for their expert
computing assistance, Bill Oliver for assisting in the development of the
power plant emissions data base, Tony Thrall for technical guidance, and
Terry Clark for providing data and technical guidance. We are also
indebted to Jack Durham, who provided initial encouragement and sought and
secured financial support.
ii
-------
CONTENTS
Acknowl edgements 11
List of Illustrations iv
List of Tables vi
Preface vi i
1 INTRODUCTION 1
2 AMBIENT SULFUR DIOXIDE TRENDS 3
Ambient Sulfur Dioxide Data Base 3
Monthly Mean and Average Daily Maximum Concentrations 6
3 POWER PLANT SULFUR DIOXIDE EMISSION TRENDS 10
Data Sources 10
Emission Estimates 16
Summaries of Emission Estimates 19
Comparisons with Other Emission Estimates 19
Trends in Monthly State Total Power Plant
Sulfur Dioxide Emissions 24
4 CORRELATIONS BETWEEN MONTHLY POWER PLANT SULFUR
DIOXIDE EMISSIONS AND SULFUR DIOXIDE AMBIENT CONCENTRATIONS 25
Correlations at the State Level 25
Correlations at the Local Level 32
Correlations at the Regional Level 39
5 SUMMARY, FINDINGS, AND RECOMMENDATIONS
FOR FURTHER WORK 45
References R-l
81*164 i m
-------
ILLUSTRATIONS
1 S02 monitoring stations in the northeastern United
S ta tes 7
2 Electric power plants in the northeastern United
States and their 1975-1982 average annual S02 emissions 21
3 Seasonally adjusted total monthly power plant SC^
emissions, Illinois, 1975-1982 27
4 Seasonally adjusted monthly average daily maximum S02
concentration, Illinois, 1975-1982 28
5 Seasonally adjusted average SC^ concentration vs
seasonally adjusted power plant S0£ emissions, Illinois,
1975-1982, monthly data 29
6 Seasonally adjusted average daily maximum SC^
concentration vs seasonally adjusted power plant S02
emissions, Illinois, 1975-1982, monthly data 30
7 Monthly average S02 emissions and ambient concentration
Goudey Power Plant and Monitor 330480007F01; Binghamton,
New York 34
8 Monthly average S02 emissions and ambient concentration
Hick!ing Power Plant and Monitor 331880003F01; Elmira,
New York 35
9 Monthly average S02 emissions and ambient concentration
Chesterfield Power Plant and Monitor 481560004F02;
Richmond, Vi rgi nia 36
10 Monthly total power plant S02 emissions, Ohio River
Valley states, 1975-1982 40
11 Monthly average S02 concentrations, Ohio River
Valley states, 1975-1982 43
12 Monthly average daily maximum S02 concentrations, Ohio
River Valley states, 1975-1982 44
8-UbH 1 .
-------
A1-A22 S02 monitoring sites by state A-l
B1-B22 Monthly average S02 concentration and number of
reporting sites, 1975-1982, by state B-l
C1-C22 Monthly average daily maximum S02 concentration
and number of reporting sites, 1975-1982, by state C-l
D1-D22 Monthly total electric utility S02 emissions,
1975-1982, by state D-l
-------
TABLES
1 Number of monitors and amount of monitoring data available
in the SAROAD ambient sulfur dioxide data base 5
2a Weighted average sulfur content of coal delivered to
electric power plants in 22 northeastern states, 1975-1982 12
2b Weighted average sulfur content of oil delivered to
electric power plants in 22 northeastern states, 1975-1982 13
3 Electric power plants with flue gas desulfurization
systems operating between 1975 and 1982, 22 northeastern
s ta tes 14
4 Ash retention for fuels consumed by electric power plants 18
5 Estimated sulfur dioxide emissions from electric power
plants in 22 northeastern states, 1975-1982 20
6 E. H. Pechan & Associates estimates of sulfur dioxide
emissions from electric power plants in 22 northeastern
states, 1975-1982 22
7 Gschwandtner and Gschwandtner estimates of sulfur
dioxide emissions from electric power plants in 22
northeastern states, 1975-1982 23
8 Correlations between seasonally adjusted utility sulfur
dioxide emissions and seasonally adjusted ambient
sulfur dioxide concentrations 31
9 Correlations between monthly SC^ power plant emissions
and ambient S02 concentrations (monthly average and
monthly average daily maximum) 38
10 Ohio River Valley states annual power plant S02 emissions
and ambient S02 concentrations, 1975-1982 41
vi
-------
PREFACE
This report describes work performed under the first phase of what was
originally perceived to be a multi-year study, and as such should be
viewed as a report of work in progress. The intention of this phase was
to assemble and evaluate data bases for statistical analyses of relation-
ships among acid rain precursors, including ambient sulfur dioxide
concentrations and sulfur dioxide emissions, and to investigate the
evidence for trends in these precursors. We found that there were no
extant sulfur dioxide data bases, either ambient or emissions, with enough
time resolution for statistical analyses. We therefore developed the
necessary data base with the approval of the EPA project officer. This
major effort consumed most of the first-year project resources. Some
statistical analyses were also performed, which were of a survey and
exploratory nature, as we assumed that detailed statistical analyses would
follow in the second year. Trends in sulfur dioxide emissions and air
quality were found, and in some instances strong correlations between
sulfur dioxide emissions and ambient concentrations were seen. This
report should be viewed as an interim progress report in a longer-term
study; the preliminary findings suggest several hypotheses for further
study.
8»+16«t 1 Vll
-------
1 INTRODUCTION
Since the mid-1970s, there has been growing concern about the harmful
effects of the deposition of acidic substances. The primary acid-forming
substances of concern are sulfur and nitrogen oxides, which are associated
with both natural and anthropogenic sources. Acidic deposition is
postulated to adversely affect aquatic ecosystems, forests and crops, and
even building materials. Hundreds of lakes in the Adirondacks in upstate
New York have been declared "dead," devoid of many of their former species
of fish; these effects are postulated to be the result of high-level,
long-term emissions of sulfur dioxide ($02) from the heavily
industrialized upper Midwest states. Other plausible postulates involve
the "acidifying effects" of ground cover and soils in the region, changes
in fish-stocking practices, or some combination of causes. Uncertainties
abound, obscuring judgments about the effectiveness of mitigating actions,
or even the need for action.
Some advocate control or regulation of harmful emissions, and call for
reductions in emissions of sulfur dioxide principally from the electric
power industry, the dominant source of S0£ emissions. A key issue in the
debate over forced S02 emissions reductions is the extent to which acidic
deposition and precursors of acidic deposition, e.g., ambient S02, will be
reduced as a result of reductions in emissions. For example, will a 50
percent decrease in S02 emissions result in a 50 percent decrease in
ambient S02 and sulfate concentrations and a 50 percent decrease in acidic
deposition, or only a 25 percent decrease? Considerable effort is being
expended in sophisticated mathematical model development and in the design
of field measurement programs to attempt to answer this and related
questions.
-------
In this report we consider still another alternative to modeling and
design-specific field programs. As a result of the economic recession in
the late 1970s, actual sulfur dioxide emissions were reduced. Data from
the late 1970s and early 1980s, then, can be used to examine the results
of actual decreases in S02 emissions. The purpose of this study is to
examine recent trends in sulfur dioxide emissions and acid deposition
precursors, specifically ambient S02, and to assess the degree of
correlation between the two. This work is seen as the first step in a
two-step process, in which the second step would be to examine the
association between reductions in S02 emissions and/or ambient S02 and
sulfate concentrations and sulfate deposition.
In this report we address the first step, reporting on our examination of
trends in (1) sulfur dioxide emissions from electric power plants, and (2)
ambient S02 concentrations from 1975 to 1982. The region of study is the
heavily industrialized northeastern United States, extending to Wisconsin
and Illinois on the west and to Tennessee and North Carolina in the
south. Twenty-one states plus the District of Columbia (hereafter
referred to as 22 states for simplicity) are included in the study. In
Section 2 we discuss the development of the ambient sulfur dioxide data
base, and in Section 3 we discuss the development of the power plant
sulfur dioxide emissions data base. Correlations between these S02 emis-
sions and ambient concentrations at local, state, and regional levels are
examined in Section 4. Finally, Section 5 presents our conclusions and
recommendations for further study.
84161+ 2
-------
AMBIENT SULFUR DIOXIDE TRENDS
AMBIENT SULFUR DIOXIDE DATA BASE
The National Aerometric Data Branch of EPA maintains the National Aero-
metric Data Base (NADB) of ambient air quality monitoring data, known as
SAROAD (^Storage ^nd Retrieval _of Air Quali ty JJata). Data are submitted to
the NADB by federal, state, and local government agencies responsible for
criteria pollutant monitors.
Ambient S0£ is measured either by a continuous ("instantaneous") monitor-
ing instrument or by a 24-hour bubbler (integrated) device. Continuous
monitors record a value every hour for a possible total of 8760 hourly
measurements in a year; bubblers record one measurement per 24-hour period
(midnight to midnight), and sampling is performed once every six days.
Prior to 1978 many S0£ monitors were 24-hour bubblers. In 1978 the EPA
required that all S02 bubblers be modified with a temperature control
device to rectify a sampling problem: when temperatures reached a certain
degree, not all of the S0£ was collected; therefore, S02 levels tended to
be underestimated (Neligan, 1978). Subsequently, many S0£ bubblers were
retired and replaced with continuous monitors. Because the bubbler
instrument modification would complicate the interpretation of trends, the
bubbler data are not used in this study.
We received from NADB the hourly data for all S02 monitoring stations for
the 22 states, 1975 to 1982, for this study. Each monitoring site is
identified by a unique 12-character SAROAD code that indicates the state,
the area within the state (city or county), the site within the area, the
managing agency, and the project classification (e.g., point source
surveillance vs. background monitoring); these codes are defined in EPA's
-------
AEROS Manual of Codes (EPA, 1983). In many instances, the controlling
agency or the project classification for a site changed although the site
remained in the same physical location; therefore, for many monitoring
stations, there are multiple SAROAD codes. We define "site" as a given
physical location; thus one site may have two or more SAROAD codes. The
22-state NADB tapes contained raw data for 1783 SAROAD codes from 1117
actual monitoring sites.
From the hourly S02 monitoring data we calculated the following monthly
summary statistics for each site:
Number of hourly measurements
Arithmetic mean concentration
Number of valid days (in which there are at least 75 percent of the
hours between 9:00 a.m. and 9:00 p.m.)
Average daily maximum hourly concentration (calculated only from
valid days)
Information on the number of S0£ monitors and the amount of data available
at the monitors is given by state in Table 1. The number of monitors in
each states ranges from 6 (District of Columbia) and 7 (North Carolina and
Vermont) to 121 (Indiana); the median number of monitors per state is
40. Generally, those states with relatively high S02 emissions have a
greater number of ambient S02 monitors. The last five columns in Table 1
show that much data is missing at most of the monitors. Only a small
percentage of the mentions have substantial long-term data; the percentage
of monitors with greater than 50 percent of each year's hours for seven or
eight years ranges from 0 percent (District of Columbia, Maine, Maryland,
New Hampshire, North Carolina) to a high of only 37 percent in
Tennessee. The middle columns in Table 1 show that the number of monitors
is relatively constant for each state, from year to year, though some
states such as Maine and Indiana add monitors to their networks each year.
-------
TABLE 1. Numbtr of monitor* and amount of monitoring data available in the
BAROAD ambient sulfur dioxide data ba«»
Total number
of monitor*
Connecticut 84
Delaware 19
District of Columbia 6
Illinois 82
Indiana 121
Kentucky 73
Main* 6O
Maryland 31
Massachusetts 84
Michigan 41
New Hampshire 21
New Jersey 38
New York 1OO
North Carolina 7
Ohio 86
Pennsylvania 78
Rhode Island 8
Tennessee 98
Vermont 7
Virginia 29
West Virginia 1O
Wisconsin 1O2
Number of monitors uiith anu data in each uear
1979 1976 1977 1978 1979 198O 1981 1982
No. of monitors with at least 90%
data for the aiven number of uears
21
13
4
41
19
40
6
19
24
30
9
23
70
1
34
49
3
40
4
13
4
23
16
12
9
47
39
4O
3
23
16
31
7
24
92
1
42
42
2
77
4
13
4
16
19
12
9
93
42
48
9
23
17
29
9
23
9O
1
41
39
2
84
4
14
4
33
19
8
3
48
49
43
9
19
19
31
9
24
49
2
44
42
4
97
3
17
4
90
12
10
3
46
69
94
16
13
49
30
8
26
47
3
47
42
4
99
4
14
4
94
11
9
3
37
71
49
28
10
21
29
7
31
92
9
46
91
3
94
9
19
6
31
9
8
3
4O
79
42
39
10
19
20
9
32
44
9
92
39
4
93
4
19
2
41
9
8
2
38
102
29
40
10
40
16
12
27
44
9
92
48
4
62
4
19
6
47
O
9
2
1
10
21
7
21
9
33
2
4
3
31
0
12
19
4
2O
2
2
1
36
1-2
7
4
1
27
36
2O
27
9
33
1O
9
12
16
4
3O
29
1
21
1
6
6
44
3-4
2
3
3
18
38
16
9
7
7
4
6
3
16
2
21
17
1
6
1
9
1
14
9-6
3
0
1
13
19
9
3
6
7
19
2
4
22
1
14
12
1
19
2
6
1
4
7-8
7
6
0
14
11
21
O
0
4
10
O
16
19
O
9
9
1
36
1
6
1
4
-------
All monitoring site locations are provided in the SAROAD site file, which
we received from NADB. From this file we extracted latitude and longitude
for all S02 monitoring stations in the 22 northeastern states. Figure 1
presents the 22-state S02 monitor locations. From the map one can see
that SC>2 monitors are located predominantly in heavily populated urban
areas; for example, there are heavy concentrations of monitors in and
around New York City and Chicago.
Figures Al through A22 show the location of, and amount of data from, S02
monitors in all 22 states, ordered alphabetically. The maps show latitude
and longitude and county borders. Each SC^ monitoring site is indicated
with a circle on the map; the diameter of the circle corresponds to the
amount of data available over the eight-year period under study,
specifically on the number of months of at least 50 percent of the
possible hours. For those sites marked with an "X" inside the circle, 50
percent of the available data for at least six out of the eight years are
available. Four percent of the S02 monitoring sites are not shown on the
maps because their latitudes and longitudes are not in the SAROAD site
file. The maps show that S02 monitoring sites are primarily located in
clusters in urban areas. In Tennessee (Figure A18) and Kentucky (Figure
A6), however, the Tennessee Valley Authority (TVA) operates clusters of
monitors in the vicinity of large power plants.
MONTHLY MEAN AND AVERAGE DAILY MAXIMUM CONCENTRATIONS
Trends in monthly mean and average daily maximum S02 concentration were
calculated for each of the 22 states. The monthly averages are based on
averages across all reporting sites for each state; because the monitoring
sites tend to be clustered in urban areas, monthly average values usually
reflect concentration levels of those urban monitors rather than those of
background or rural monitors. Although the absolute levels of the
concentrations within a state may reflect higher urban concentrations, the
-------
Figure 1. SO? Monitoring Stations in the Northeastern United States
-------
relative levels (i.e., the increases and decreases observed) may be more
representative of all areas in the state.
Figures Bl through B22 present trends in monthly average SC^ concentra-
tions for the 22 northeastern states from 1975 to 1982. Three methods
used to calculate the monthly averages are plotted. When monthly averages
are calculated by averaging across all individual site monthly averages,
each site is weighted equally (thin line). When the sites are weighted
according to the number of hours in the monthly site mean, each hourly
observation in the month is weighted equally (thick dashed line). The
thick black line indicates monthly averages calculated by averaging across
individual site monthly averages, but only for those sites with at least
40 percent of the total possible hours in the month (the cutoff value of
40 percent was chosen according to the results of the data completeness
study by Thrall et al., 1984). The bottom of each plot shows the number
of monitoring sites available for calculating each monthly mean. There
the thin line indicates the total number of sites with at least one hourly
observation in the month, while the thick line indicates the number of
sites with at least 40 percent of the data available for the month.
The seasonal nature of monthly average S0£ concentrations can be seen in
Figures Bl to B22: S02 concentrations for these urban-representative
averages are highest in the winter when (1) the S02 emissions from low-
elevation, sulfur-containing fuel heating sources are highest, (2) the air
is relatively stagnant, and (3) mixing volumes are small; and lowest in
the hot summer months when emissions from low-elevation sources are less
and there is greater mixing. Some of the plots also display a secondary
"peak" in the summer season, with an amplitude that is about 10 percent of
the winter peak (cf. Figures B4, B13, and B18). The plots reveal that the
three averaging methods result in remarkably similar trends; the few
exceptions, such as New Hampshire in 1980 (Figure Bll) or Rhode Island in
1976 (Figure B17) occur when there are only a few sites and one site has
an erratic pattern of monthly means resulting from a limited number of
hourly observations in a month. The three averaging methods result in
similar trends because the majority of monitors report data for at least
-------
40 percent of the total possible hours in each month, as can be seen at
the bottom of each plot. Finally, more than half of the plots show
evidence of a downward trend. Only one plot (for Maryland) indicates an
upward trend in monthly average S0£ concentrations over the period.
Trends in average daily maximum S02 concentration for each of the 22
northeastern states are shown in Figures Cl through C22. The three
averaging methods shown in Figures Bl to B22 are also used here, with the
same assignment of plotting line to averaging method, except that the 40
percent cutoff applies to the number of valid days rather than hours. A
day is considered valid if at least 75 percent of the hours between 9:00
a.m. and 9:00 p.m. have valid observations. The general patterns and
seasonal cycles for average daily maximums are similar to those observed
in monthly averages. Massachusetts (Figure C9) has an unusually large
average daily maximum in December of 1977, which most likely is caused by
one site with an extremely high (possibly invalid) value. This deviation
is possible because the raw SAROAD S0£ data were not subjected to large-
scale checking; rather, they were taken at face value.
Btibt 2
-------
3 POWER PLANT SULFUR DIOXIDE EMISSION TRENDS
DATA SOURCES
Power Plant Fuel Consumption
Data pertaining to monthly power plant consumption of fuels are available
from the Energy Information Administration (EIA) on Form EIA-759 (also
called by its original name, FPC Form 4), the Power Plant Report. The
monthly electricity generation, consumption of fossil fuels, and fuel
stocks data for each power plant in the United States have been computeri-
zed by the Department of Energy (DOE). We received a tape from DOE
containing one file for each year, 1975 to 1982. We merged the files and
sorted them by plant codes and by fuel source for further analysis. For
the 22 northeastern states, there are approximately 31,000 records in the
file.
Fuel Quality
Fuel quality data are available on federal Form 423, the Cost and Quality
of Fuels Report. Each record in the file contains information on one fuel
delivery, so there may be multiple records for a given fuel type at a
plant in a given month; in addition, many plants consume more than one
kind of fuel. Included in this computerized data base are Btu content,
percent sulfur content by weight, and percent ash content by weight. We
received from DOE a data tape with separate files for each year, 1975 to
1982. These files were sorted by plant code and merged into one file
containing approximately 70,000 records for coal and oil deliveries.
10
-------
Cost and Quality of Fuels Reports are required only for plants with a
generating capacity of at least 25 MW. All power plants, however, are
required to file the monthly Power Plant Reports. For those small plants
with less than 25 MW generating capacity, we substituted average sulfur
content of oil and coal for each state in each year; these values were
also used for larger plants for which data were missing from the Cost and
Quality of Fuels file. For a given state and year, average sulfur content
was calculated as the weighted average sulfur content of all deliveries
for that state in that year separately for coal and oil; the weights used
were the tons of coal or barrels of oil delivered. The weighted average
sulfur contents of delivered coal and oil for the 22 states in each year
are listed in Tables 2a and 2b, respectively. Qualitatively, one observes
a downward trend in median percent sulfur (by weight) in coal (from
approximately 2.03 percent in 1975 to approximately 1.7 percent in 1981
and 1982). The reverse trend is observed for the median weight percent
oil sulfur content (1.18 percent in 1975 to approximately 1.3 percent in
1982).
Power Plant Flue Gas Desulfurization Systems
Flue gas desulfurization (FSD) systems, also known as scrubbers, have been
installed at many large power plants in the northeast. A computer data
base containing information about all existing and planned FGD systems is
maintained on EPA's Univac computer by PEDCo Environmental, Inc. of
Cincinnati, Ohio (PEDCo, 1982). This data base, known as F6DIS (for FGD
Information System), provided a list of all scrubbers in operation in the
22 states between 1975 and 1982; the power plants with scrubbers are
listed in Table 3.
Power Plant Location
Power plant location information (latitude and longitude) was extracted
from the stack file of E. H. Pechan & Associates of Springfield, Virginia;
11
-------
TABLE 2a. Weighted average sulfur content of coal delivered to
electric power plants in 22 northeastern states, 1975 - 1982.
(Units are tons sulfur per 100 tons coal)
1975
1977 I27_fi
12SQ 12S1 12S2
Connecticut
Delaware
District of Columbia
Illinois
Indiana
Kentucky
Maine
Maryland
Massachusetts
Michigan
New Hampshire
New Jersey
New York
North Carolina
Ohio
Pennsylvania
Rhode Island
Tennessee
Vermont
Virginia
West Virginia
Wisconsin
1.94
2.07
0.70
2.29
2.71
3.23
1.94
1.55
0.80
2.39
2.39
1.82
1.87
1.05
2.95
2.03
1.94
2.86
0.97
0.82
2.04
2.26
2
2
2
2
2
3
2
1
2
2
2
1
1
1
2
2
2
2
2
0
1
2
.06
.02
.06
.28
.57
.19
.06
.47
.06
.13
.17
.59
.83
.01
.89
.13
.06
.78
.06
.84
.86
.14
2
1
2
2
2
2
2
1
2
2
2
1
1
1
2
2
2
3
1
0
1
2
.01
.96
.01
.10
.54
.89
.01
.47
.01
.05
.28
.71
.76
.03
.78
.11
.01
.01
.49
.95
.88
.22
1.91
1.75
1.91
1.98
2.52
2.61
1.91
1.48
1.91
1.79
2.12
1.64
1.68
1.06
2.63
2.06
1.91
2.41
1.91
0.95
1.81
2.12
1.92
1.80
1.92
1.86
2.64
2.74
1.92
1.41
2.57
1.58
2.51
1.68
1.78
0.92
2.58
2.03
1.92
2.24
1.51
0.84
1.87
2.01
1.
1.
1.
1.
2.
2.
1.
1.
1.
1.
2.
1.
1.
0.
2.
2.
1.
2.
1.
0.
1.
1.
79
60
79
78
56
45
79
60
13
26
49
62
80
94
44
07
79
21
79
86
75
81
1.67
1.33
1.67
1.71
2.43
2.44
1.67
1.69
1.24
1.20
2.46
1.44
1.82
0.91
2.46
2.02
1.67
2.27
0.57
0.89
1.77
1.48
1.66
1.23
1.66
1.93
2.36
2.48
1.66
1.60
1.24
1.34
2.35
1.43
1.80
0.89
2.47
2.02
1.66
2.00
0.58
0.87
1.86
1.52
12
-------
TABLE 2b. Weighted average sulfur content of oil delivered to
electric power plants in 22 northeastern states, 1975 - 1982.
(Units are tons sulfur per 1000 barrels oil)
Connecticut
Delaware
District of Columbia
Illinois
Indiana
Kentucky
Maine
Maryland
Massachusetts
Michigan
New Hampshire
New Jersey
New York
North Carolina
Ohio
Pennsylvania
Rhode Island
Tennessee
Vermont
Virginia
West Virginia
Wisconsin
1225
0.71
1.44
1.52
0.89
0.33
0.50
3.51
1.85
1.19
1.18
3.00
0.55
1.77
0.53
0.87
0.67
1.31
0.99
0.17
2.61
0.16
1.27
122f
0
1
1
0
0
0
3
2
1
1
3
0
1
0
0
0
1
1
0
2
0
0
.62
.55
.53
.82
.47
.64
.39
.11
.69
.14
.03
.61
.72
.30
.56
.71
.57
.02
.16
.53
.15
.60
1977
0.
1.
1.
0.
0.
0.
3.
2.
1.
1.
2.
0.
1.
0.
0.
0.
1.
1.
0.
2.
0.
0.
60
56
66
88
50
42
70
12
69
20
96
70
83
33
91
81
56
03
43
50
35
55
122fi
0.64
1.54
1.26
0.86
0.55
0.31
2.45
2.08
2.60
1.14
2.83
0.75
1.81
1.01
0.88
0.89
1.46
1.03
1.32
2.58
0.44
0.53
1222
0
1
1
0
0
0
2
2
2
1
3
0
2
0
1
1
1
0
0
2
0
0
.73
.67
.35
.98
.47
.58
.34
.13
.81
.08
.01
.78
.11
.34
.36
.08
.49
.85
.48
.35
.57
.73
12SQ
0.73
1.67
1.34
1.10
0.44
0.54
2.20
2.09
2.79
1.10
3.20
0.75
2.06
0.35
0.84
1.07
1.55
0.81
0.59
2.16
0.43
0.66
12fil
0.81
1.97
1.40
1.09
0.41
0.53
1.73
1.65
2.97
1.16
3.15
0.87
2.22
0.35
0.85
1.04
1.91
0.81
0.42
1.94
0.32
0.50
19.S2
1.32
1.94
1.30
1.12
0.58
0.50
1.98
1.85
3.00
1.17
3.12
0.96
2.10
0.33
0.74
1.07
1.76
0.73
0.37
1.79
0.44
0.55
13
-------
TRBLE 3. Electric power plants with flue gas desulfurization systems
operating between 1975 and 1982, 22 northeastern states.
Company
Plant Name
Central Illinois Light
Duck Creek
No.
1
Gross
MW
416
Net
MN
378
Esc
MN
416
Init.
Start
Date
7607
Cmrcl.
Start
Date
7808
Design
S02 Rem.
Eff. %
85.30
Central Illinois Public Serv
Newton
Commonwealth Edison
Powerton
Delmarva Power & Light
Delaware City
Delaware City
Delaware City
Detroit Edison
St. Clair
Duguesne Light
Elrama
Hoosier Energy
Merom
Indianapolis Power & Light
Petersburg
Kentucky Utilities
Green River
Louisville Gas & Electric
Cane Run
Cane Run
Cane Run
Mill Creek
Mill Creek
Mill Creek
Monongahela Power
Pleasants
Pleasants
1
51
1
2
3
6a
1-4
2
3
1-3
4
5
6
1
2
3
1
2
617
450
60
60
60
325
510
490
532
65
188
200
299
358
350
427
626
626
575
400
154
487
460
515
59
175
192
277
334
325
420
580
580
617
450
60
60
60
163
510
441
532
65
188
200
299
358
350
427
626
626
7909
8004
8005
8005
8005
7506
7510
8112
7712
7509
7608
7712
7904
8012
8112
7808
7812
8010
7910
8106
8005
8005
8005
7509
7510
8202
7712
7606
7708
7807
7904
8104
8204
7903
8012
8012
89.50
75.50
90.00
90.00
90.00
90.00
83.00
90.00
85.00
98.00
85.00
85.00
95.00
85.00
85.00
85.00
90.00
90.00
Continued
14
-------
TABLE 3. Concluded.
Init. Cmrcl. Design
Company Gross Net Esc Start Start S02 Fern.
Plant Name No. MW MW MW Date Date Eff. %
Northern Indiana Pub Service
Dean H. Mitchell 11 116 94 115 7607 7706 90.00
Pennsylvania Power
Bruce Mansfield 1 917 780 917 7512 7606 92.10
Bruce Mansfield 2 917 780 917 7707 7710 92.10
Bruce Mansfield 3 917 800 917 8006 8010 92.20
Philadelphia Electric
Eddystone la 120 120 120 7509 7509 90.00
Southern Illinois Power
Marion 4 184 161 184 7904 7906 89.40
Southern Indiana Gas & Elec
A.B. Brown 1 265 250 265 7903 7904 85.00
Springfield Water, Light & Pwr
Dallman 3 205 192 205 8010 8101 95.00
Source: PEDCo Environmental, Inc., Flue Gas Desulfurization Information
System Data Base
15
-------
the data base was developed with guidance from, and the participation of,
the Utility Air Regulatory Group, the Utility Data Institute, and the U.S.
EPA. The file contains information, including latitude and longitude, on
1951 stacks of 779 power plants across the United States. Latitude and
longitude are missing for about 10 percent of the plants in the file; we
manually located all plants for which location parameters were missing if
their annual average sulfur dioxide emissions were more than 10,000 tons.
EMISSION ESTIMATES
Monthly sulfur dioxide emissions were estimated for all electrical power
plants in the 22 northeastern states for the years 1975 to 1982. At each
plant, monthly emissions were estimated from fuel consumption data and
sulfur content of fuel data as follows:
S
QS02 = c x
where
Q_SO = tons of S02 emitted;
C = tons of fuel consumed to generate electricity;
S = sulfur content of the fuel, percent;
A = ash retention, percent; and
2 = multiplier, since 1 ton of sulfur burned produces 2 tons of S02.
For each plant, S02 emissions were calculated separately for each type of
fuel consumed (coal and/or oil) and then summed to estimate total plant
emissions for the month. The sulfur content of each fuel for each plant
was calculated not for each month, but for each year, since there are many
plant-months with no fuel deliveries. An annual weighted average sulfur
content is a reasonable approximation because there is little variability
in sulfur content from one delivery to another within a given plant
84164 2
16
-------
(Burton et al., 1982). This assumption should not lead to any bias in the
calculated emissions. Ash retention for each fuel type is given by EPA
(AP-42 Emission Factors, 1983); these values are listed in Table 4. We
assumed that all fuels were burned in the same year in which they were
delivered. Errors in emission estimates from lags in fuel consumption
should be small since the trends in annual median coal and oil sulfur
content were observed to be small (cf. Tables 2a and 2b).
Electric power plants with FGDs have reduced S02 emissions. For these
plants, the following additional calculations were made:
v TG - SG x E/100
x TS
where
Q$Q = reduced monthly SOg emissions;
TG = total plant generating capacity;
SG = generating capacity of the unit with a scrubber; and
E = scrubber designed efficiency, percent.
Total plant emissions were thus reduced by the percentage of emissions
scrubbed. Monthly emission estimates were reduced beginning with the
month in which the FGD system started commercial (as opposed to initial
testing) operations, as listed in Table 3. For these calculations we
assumed that the scrubber, once in operation, was always operating at its
designed efficiency. This assumption will, on the average, lead to an
underestimation of emissions because it is more likely that scrubber
efficiencies will be somewhat lower than design values; however, the
overall downward bias in emission estimates is believed to be small.
17
-------
TABLE 4. Ash retention for fuels consumed by electric power plants
Fuel *type Ash Retention (Percent)
Bituminous coal 2.3
Subbituminous coal 12.3
Anthracite coal 0.0
All fuel oils 0.0
Source: EPA AP-42 Emission Factors, Supplement 13, 1983.
18
-------
SUMMARIES OF EMISSION ESTIMATES
Total annual electric power plant S02 emissions for each year, 1975-1982,
are given for each state in Table 5. The state with the highest emission
levels in all years is Ohio, by a wide margin (30 to 48 percent for the
1975-1982 period); Illinois, Indiana, Kentucky, and Pennsylvania also have
consistently high emissions. Average annual 1975-1982 S02 emission totals
for each power plant are shown on the map of the 22 states in Figure 2.
The diameter of the circle marking each plant is proportional to average
annual plant emissions: the largest circles indicate plants with more
than 100,000 tons emitted each year, while the smallest circles indicate
small plants with less than 10,000 tons emitted each year. In the plot
one can readily see the high concentration of the largest emitters of S02
in the Ohio River Valley, and on or near other bodies of water, e.g.,
Lakes Michigan and Erie.
COMPARISONS WITH OTHER EMISSION ESTIMATES
In two recent studies annual sulfur dioxide emissions from power plants
were estimated. E. H. Pechan & Associates (EHPA, 1982) estimated annual
emissions from 1976 through 1980; their totals by state are given in
Table 6. Gschwandtner and Gschwandtner (1983) estimated total annual
emissions of S02 and nitrogen dioxide (N02) for every five years since
1900 and also for 1978; their estimates of S02 emissions by state from the
electric utility industry for those years which overlap our study are
given in Table 7.
The data files used as a basis for the EHPA emission estimates are the
same files used as the basis for the estimates in this study. Although
similar calculations were performed, the state totals in Table 5 are
generally about 1 to 5 percent higher than the EHPA state totals shown in
Table 6. There are two reasons why our estimates are higher than those of
EHPA. First, we estimated S02 emissions for all power plants reporting
fuel consumption on the monthly Power Plant Report, while EHPA omitted
19
-------
IN3
Figure 2. Electric Power Plants In the Northeastern United States
and Their 1975 - 1902 Average Annual SO? Emissions
-------
TABLE 5. Estimated sulfur dioxide emissions from electric power plants
in 22 northeastern states, 1975 - 1982.
Emissions in 1000 tons per year
1225 1976 1977 122fi 1979 19JJQ 1281 12S2.
Connecticut 32.0 25.0 23.3 25.8 27.3 31.4 29.7 46.0
Delaware 57.8 62.1 60.3 56.3 62.2 56.6 69.9 57.2
District of Columbia 7.0 7.0 12.4 9.8 6.3 4.2 2.4 0.7
Illinois 1502.7 1511.6 1451.0 1352.2 1218.7 1173.6 1029.2 1056.5
Indiana 1482.2 1485.3 1494.0 1384.4 1576.6 1584.9 1487.6 1345.1
Kentucky 1443.9 1549.9 1384.7 1244.1 1144.6 1072.9 1157.5 1042.8
Maine 20.0 12.9 9.8 8.7 10.8 16.1 13.2 12.6
Maryland 190.6 221.3 200.8 223.4 208.5 228.2 200.6 207.8
Massachusetts 109.8 158.7 159.6 257.1 261.5 275.4 264.6 263.2
Michigan 1028.6 891.4 907.4 820.7 761.2 569.0 604.6 587.7
New Hampshire 60.4 51.3 60.5 52.9 80.3 81.7 69.1 61.3
New Jersey 111.1 115.8 130.2 116.7 106.9 105.1 102.4 96.2
New York 539.0 515.6 552.0 522.8 511.3 479.4 509.4 474.1
North Carolina 383.9 423.1 439.2 408.2 391.0 446.8 456.8 420.2
Ohio 2794.0 2860.6 2833.3 2637.5 2713.2 2359.5 2383.0 2295.3
Pennsylvania 1499.4 1510.7 1451.6 1445.9 1591.9 1637.3 1509.8 1481.1
Rhode Island 4.1 3.0 3.6 3.4 2.8 5.2 5.0 3.1
Tennessee 1082.7 1263.1 1288.9 1062.5 922.3 960.6 900.3 646.7
Vermont 0.3 0.2 0.4 0.2 0.4 0.3 0.2 0.4
Virginia 210.5 226.4 238.5 224.2 204.0 165.1 142.7 125.6
West Virginia 1053.5 1034.5 1026.8 916.0 1010.4 1023.0 949.3 886.6
Wisconsin 442.9 475.5 535.1 482.8 501.7 472.9 412.6 375.4
20
-------
TABLE 6. E. H. Pechan & Associates estimates of sulfur dioxide emissions
from electric power plants in 22 northeastern states, 1976 - 1980.
Emissions in 1000 tons per year
127.6. 1977 1978 1222 12SQ
Connecticut 25.2 23.5 26.0 27.6 32.1
Delaware 60.9 59.2 55.6 61.1 52.5
District of Columbia 7.2 12.6 10.4 6.7 4.6
Illinois 1428.8 1367.0 1292.9 1167.7 1125.6
Indiana 1443.1 1457.6 1351.2 1536.9 1539.6
Kentucky 1512.3 1356.5 1210.0 1130.0 1007.5
Maine 13.0 9.9 8.7 10.9 16.3
Maryland 218.2 198.0 220.5 205.2 223.2
Massachusetts 159.6 160.4 258.9 264.5 275.5
Michigan 887.6 905.1 806.9 741.0 565.4
New Hampshire 50.5 59.4 52.3 78.9 80.5
New Jersey 113.2 128.4 115.3 105.1 110.2
New York 512.8 548.0 520.0 508.1 480.3
North Carolina 410.2 427.2 396.4 379.5 435.4
Ohio 2749.8 2686.1 2462.6 2514.5 2171.6
Pennsylvania 1432.0 1381.1 1322.7 1415.1 1466.1
Rhode Island 3.0 3.6 3.4 2.8 5.2
Tennessee 1228.3 1257.6 1033.1 893.3 933.7
Vermont 0.4 0.4 0.3 0.4 0.5
Virginia 224.9 238.0 223.9 203.2 163.7
West Virginia 1010.4 1001.4 895.5 955.9 944.2
Wisconsin 469.7 514.7 471.7 496.3 485.7
Source: E. H. Pechan & Associates, "Estimates of Sulfur Dioxide Emissions from
the Electric Utility Industry," 1982.
22
-------
EffiLE 7. Gschwandtner and Gschwandtner estimates of sulfur dioxide emissions
from electric power plants in 22 northeastern states; 1975, 1978, and 1980.
Emissions in 1000 tons per year
1225 1978 12SQ
Connecticut 31.4 25.9 28.9
Delaware 33.7 36.0 54.1
District of Columbia 13.2 4.7 1.9
Illinois 1778.8 1305.5 1190.5
Indiana 1600.2 1325.7 1644.3
Kentucky 1276.8 1243.6 1052.1
Maine 20.1 8.7 14.0
Maryland 204.2 211.8 245.0
Massachusetts 101.9 252.4 264.3
Michigan 1070.0 808.6 603.4
New Hampshire 62.7 50.7 82.2
New Jersey 117.4 101.8 107.0
New York 577.2 490.5 461.8
North Carolina 437.6 419.8 430.1
Ohio 2661.7 2265.6 2346.2
Pennsylvania 1439.9 1143.6 1415.1
Rhode Island 5.1 3.8 4.2
Tennessee 1351.0 896.4 1003.0
Vermont 0.5 0.0 0.0
Virginia 219.9 220.7 158.2
West Virginia 1300.4 1092.6 1059.7
Wisconsin 555.9 845.1 523.3
Source: Gschwtner and Gschwandtner, "Historic Emissions of Sulfur
and Nitrogen Oxides in the United States from 1900 to 1980," 1983,
23
-------
from their analyses all plants with less than one ton per year of S02
emissions. Second, EHPA assumed ash retention values of 5 percent for
bituminous coal and 15 percent for subbituminous coal, while we used the
EPA AP-42 recommendations of 2.3 and 12.3 percent, respectively.
Gschwandtner and Gschwandtner emission estimates are based on total annual
fuel consumption figures in various Department of Energy reports (for
complete details, see the Gschwandtner and Gschwandtner report) and EPA
AP-42 emission factors. Their calculations were not done plant by plant,
as were ours and those of EHPA. Differences between our state total S02
emission estimates (Table 5) and the Gschwandtner and Gschwandtner
estimates in Table 7 are relatively large. Relative differences are
especially large for those states with low S02 annual emissions, but the
absolute differences for these states are relatively small; however, even
in state-years with annual S02 emissions above 500,000 tons (by our
estimate), the two sets of annual estimates vary by as much as 25
percent. We have not examined (and do not wish to speculate about) why
our emission estimates differ from those of Gschwandtner and Gschwandtner.
TRENDS IN MONTHLY STATE TOTAL POWER PLANT
SULFUR DIOXIDE EMISSIONS
Monthly total S02 power plant emissions are plotted by state in Figures Dl
through D22. The power plant emissions for most states have a regular .
yearly pattern, with a summer peak in July and August and an even higher
winter peak in December and January. In those states which have high S02
emissions, such as Ohio (Figure D15), Illinois (Figure D4), and Kentucky
(Figure D6), significant decreases can be seen. In states with relatively
few power plants and low emissions, such as Connecticut (Figure Dl) and
Delaware (Figure D2), trends in emissions are difficult to detect apart
from the large seasonal variability.
Btibt 2
24
-------
CORRELATIONS BETWEEN MONTHLY POWER PLANT SULFUR DIOXIDE EMISSIONS
AND SULFUR DIOXIDE AMBIENT CONCENTRATIONS
CORRELATIONS AT THE STATE LEVEL
Monthly statewide ambient S02 mean and average daily maximum concentra-
tions are presented in Figures Bl to B22 and Figures Cl to C22, respec-
tively. Monthly statewide power plant S02 emissions are presented in
Figures Dl to D22. These plots reveal, in general, long-term reductions
in ambient S02 concentrations in addition to long-term reductions in emis-
sions; however, no clear pattern emerges with respect to short-term
correlations between changes in emissions and changes in ambient S02«
Short-term correlations are not clearly apparent because of the highly
seasonal nature of both emissions and ambient concentrations: ambient S02
concentrations peak during the winter, whereas emission levels have a
winter peak as well as a summer peak. To examine the difference between
actual monthly concentrations and "typical" monthly concentrations (i.e.,
averages for each month of the year), we applied the following statistical
model to monthly emissions and S02 concentrations:
A
Y. = a* 1* T ^29 "*" ^12 12 '
where
1 if the observation occurs in month j, 0 if otherwise, j = 1 to
12;
average emissions or S02 concentration for month j, j = 1 to 12;
and
25
-------
Y. = predicted monthly emissions or SC^ concentration, i = 1 to 96.
The difference between the actual values Y^ and the predicted values
Y., or the residuals, are the seasonally adjusted observations. As an
example, the seasonally adjusted power plant emissions and monthly average
daily maximum S02 concentrations for Illinois are shown in Figures 3 and
4, respectively; these values should be compared to actual emissions and
maximum SC^ concentrations shown in Figures D4 and C4, respectively.
The correlations of interest are between the seasonally adjusted emissions
and the seasonally adjusted SC^ concentrations. If the differences
between actual and typical seasonal emissions are correlated with the
differences between actual and typical seasonal SC^ concentrations, then
short-term changes in ambient SC^ levels are related to short-term changes
in emissions. Figures 5 and 6 show seasonally adjusted monthly power
plant emissions in Illinois plotted against seasonally adjusted monthly
mean and average daily maximum S02 concentrations, respectively. In both
plots the correlations are relatively strong; the correlation in Figure 5
is 0.759, and the correlation in Figure 6 is 0.728. These correlations
indicate that in Illinois monthly changes in emissions are reflected in
monthly changes in ambient S02 concentrations.
Correlations between seasonally adjusted emissions and seasonally adjusted
monthly mean S0£ concentrations, and between seasonally adjusted emissions
and monthly average daily maximum S0£ concentrations are shown for all
states in Table 8. These correlations are highest in those states with
greater power plant S02 emissions. The degree of correlation depends to
some extent on the number of S02 monitors and the amount of data at each
of the monitors (see Figures Al to A22); the more monitoring data there
are, the less variability there is in average monthly S02 concentrations
and the more likely it is that actual correlations between emissions and
ambient S02 concentrations will be seen. In addition, the degree of the
correlation depends on the locations of the S02 monitoring stations
relative to the locations of the power plants with sizeable S02
BHibH 3
26
-------
35.0
-30.0 -
-35.0
11111111111111111111111111 n 111111111111111111111111111111111111111111111111111111111111:111!! i.:
1975 1976 1977 1978 1979 1980 1981 1982
Figure 3. Seasonally fldjusted Total Monthly Power Plant
502 Emissions. Illinois, 1975 - 1982.
27
-------
0.010
0.009
0.008
0.007
0.006
0.005
£0.004
o 0.003
or)
0.002
2 0.001
>,
--0.000
03
C
O
£-0.001
Q>
CO
-0.002
-0.003
-0.004
-0.005
-0.006
-0.007
111111111111111111111111111111111111111111111111111 ii 1111111111111111111111111111111111111 i i 111
1975 1976 1977 1978 1979 1980 1981 1962
Figure 3. Seasonally fldjusted Monthly flverage Daily Maximum
S02 Concentration, Illinois. 1975 - 1982.
28
-------
0.0090+
_ *
*E 0.0060+ * *
S * *
"" 2
g _ * * * *
'43 - * * * *
2 0.0030+ * * 2 * ** ****
£ * *
g * * 2
§ * * *
0 _ * * * * *
0^-0.0000+ * 2 * ** *
00 _ ****** *
"S - *******
t) _ ***** *
.^, *
5 -0.0030+ * 2 * * *
>, _ * * *
£1 _ * * 2* *
c - *******
° _ * * *
S -0.0060+
-0.0090+
-35. -25. -15. -5. 5. 15. 25. 35.
Seasonally Adjusted S02 Emissions (1000 tons)
Figure 5. Seasonally Adjusted Average S02 Concentration vs
Seasonally Adjusted Power Plant S02 Emissions,
Illinois, 1975 - 1982, Monthly Data
29
-------
0.0240+
_ *
_ *
*
IE" 0.0160+ *
£ *
^ * *
Q_ *******
'£> - * * *
£ 0.0080+ * *2 * *
c - *******
o - *******
o ** *
0 _ * * *
o^ 0.0000+ * * *
_ * ** **2 * * *
~s * * *** * * **
CO * *
^> 2 * *
S -0.0080+ ** * * *
>, * * * *
;z _ * * *2 ** *
c _ * ** *
CO * * *
S -0.0160+ *
-0.0240+
-35. -25. -15. -5. 5. 15. 25- 35.
Seasonally Adjusted SO^ Emissions (1000 tons)
Figure 6. Seasonally Adjusted Average Daily Maximum S02 Concentration vs
Seasonally Adjusted Power Plant S02 Emissions,
Illinois, 1975 - 1982, Monthly Data
30
-------
TABLE 8. Correlations between seasonally adjusted utility sulfur dioxide
emissions and seasonally adjusted ambient sulfur dioxide concentrations
Ohio
Pennsylvania
Indiana
Illinois
Kentucky
Tennessee
West Virginia
Michigan
New York
Wisconsin
North Carolina
Massachusetts
Maryland
Virginia
New Jersey
New Hampshire
Delaware
Connecticut
Maine
District of Columbia
Rhode Island
Vermont
(1)
2609.6
1516.0
1480.0
1286.9
1255.1
1015.9
987.5
771.3
513.0
462.4
421.1
218.7
210.1
192.1
110.5
64.7
60.3
30.1
13.0
6.2
3.8
0.3
(2)
0.484
0.522
0.280
0.759
0.715
0.729
0.319
0.808
0.117
-0.198
0.080
-0.080
-0.129
0.771
0.428
0.445
0.071
0.215
-0.287
0.101
0.106
0.115
(3)
0.598
0.486
0.262
0.728
0.367
0.444
0.370
0.659
0.346
-0.017
-0.046
-0.044
-0.082
0.757
0.465
0.242
0.003
0.053
-0.206
0.206
-0.030
0.167
(4)
15.7
24.7
13.3
15.4
15.2
10.8
14.5
14.9
20.2
13.9
9.4
25.5
54.9
34.4
74.1
48.8
9.4
80.7
12.3
48.5
46.8
76.4
(5)
38.0
47.4
46.4
30.5
55.9
48.8
21.2
42.7
82.9
19.9
7.5
90.0
53.4
49.8
83.1
63.5
16.1
81.1
42.6
68.3
70.0
81.4
(6)
26.6
10.7
45.5
20.3
22.0
21.9
13.8
7.3
57.1
26.7
45.0
10.9
34.4
24.1
41.9
31.0
35.3
26.9
31.3
20.6
38.8
22.9
(1) Average annual S02 emissions from electric power plants, 1000 tons, 1975 - 1982.
(2) Correlation between seasonally adjusted SO. emissions and seasonally adjusted
monthly average SO. concentration.
(3) Correlation between seasonally adjusted SO. emissions and seasonally adjusted
monthly average daily maximum S02 concentration.
2
(4) B for seasonal adjustment regression model for SO. emissions.
2
(5) K for seasonal adjustment regression model for monthly average SO.
concentration.
(6) R for seasonal adjustment regression model for monthly average
daily maximum SO. concentration.
31
-------
emissions. In Michigan, for example, all of the SOg monitors except one
are located near major power plants.
o
The Rc values in the last three columns in Table 8 indicate the percentage
of variation in emissions and S02 concentrations that is explained by the
f\
seasonal adjustment model. Low R£ values indicate that variation in
emissions or SC^ concentrations from monitor to monitor cannot be
explained by the seasonal adjustment model, i.e., that no regular seasonal
pattern can be detected; high R2 values indicate that most of the
variation is explained by the seasonal adjustment model, i.e., that
emissions or SC^ concentrations follow a very regular seasonal pattern.
In general, the seasonal adjustment models fit better with states having
low emissions and few SC^ monitors. In states that have many power plants
and SC>2 monitors, greater levels of aggregation reduce effects of seasonal
patterns relative to long-term trends.
CORRELATIONS AT THE LOCAL LEVEL
Relationships between individual power plant emissions and monitored S02
concentrations were examined for a few selected sites. Such relation-
ships, however, are inherently difficult to analyze because of missing
data at most S02 monitoring stations. Very few stations were in existence
during the entire eight-year period under study, and even when monitors
are operating they rarely record measurements for all of the hours in a
given year. From the set of monitors with at least four years of at least
50 percent of the total possible hours each year, we selected three with
power plants nearby for further analysis; two of the monitor pairs are in
New York near the Pennsylvania border, and the third is in eastern
Virginia.
The first power plant-monitor pair chosen is located near Binghamton, New
York, near the Pennsylvania border. The Goudey power plant, owned by New
York State Electricity and Gas, is located in Johnson City, just east of
Binghamton. The generating capacity of the plant was 145.7 MW from 1975
a 32
-------
to 1978, was decreased to 103.7 MW in 1979 and 1980, and was increased to
118.7 MW in 1981. The S02 monitor (SAROAD identification 330480007F01) is
a population-oriented monitor located at a water treatment plant southeast
of the Goudey plant.
The second power plant-monitor pair is also located in upstate New York
near the Pennsylvania border but further east near Corning and Elmira.
The Hickling power plant is owned and operated by New York State Elec-
tricity and Gas, and is located in East Corning. The generating capacity
of the plant was 70 MW between 1975 and 1980, and was increased to 83 MW
in 1981. The S02 monitor (SAROAD ID 331880003F01), is located at a water
treatment plant southeast of the plant, and is population-oriented.
The third power plant chosen for analysis serves a much larger population
base, Richmond, Virginia. The Chesterfield plant is owned and operated by
Virginia Electric Power and is located in Chester, south of Richmond. The
plant had a generating capacity of 1484 MW until late 1981, when it was
decreased slightly to 1352 MW. The source-oriented S02 monitor (SAROAD
identification 481560004F02) is located in Hopewell, a few miles southeast
of the plant.
Of the three S02 monitors, only the Chester monitor has nearly complete
data. For that monitor sufficient S02 monitoring data are available to
compute average daily maximums for all but one month (there were some
values recorded, however, for that month, so a monthly mean could be
calculated). The two New York S02 monitors, though, are missing many
months of data; the monitor near Goudey is missing 14 consecutive months
in 1977 and 1978, and the monitor near Hickling begins in 1977 and is
missing a month at the end of 1978.
Figures 7, 8, and 9 show the estimated power plant emissions and monthly
mean and average daily maximum S02 concentrations for each of the three
sites chosen. The thicker line in the upper portion of the plots shows
power plant emissions; the scale for emissions is on the right-hand side
of the plots. The thinner lines in the bottom portion are monthly mean
3 33
-------
0.060
0.055
0.050
0.045
Q_
Q-
VW
C0.040
o
0.035
0.030
CM
O
CT>
Q>
S0-025
c_
Q)
a:
±0.020
.c
c
o
1600
0.015
0.010
0.005
B-BBIS 1975
111111111111111111111111111111111 111II11111111111111111111111111111111111111II1111111111
1976 1977 1978 1979 1980 1981 1982
-200
FIGURE 7. Monthly average S02 emissions and ambient concentration.
Goudey Power Plant and monitor 330480007F01; Binghamton, New York.
34
-------
0.060
0.055
0.050
0.045
0.
Q_
C0.040
O
-p
-------
0.120
0.110 -
7000
Q>
£0-050
11111111111111111111111 h 111111111111111111111111111 n 11111 11111 n M 1111 n 11111111 1111111 n 11
1975 1976 1977 1978 1979 1980 1981 1982
FIGURE 9. Monthly average S02 emissions and ambient concentration.
Chesterfield Power Plant and monitor 481560004F02; Richmond, Virginia
36
-------
S02 concentration (the lower of the two lines) and average daily maximum
S02 concentration; the scale for SC^ concentrations is on the left-hand
side of the plot. Statewide power plant emissions and SC^ concentration
patterns are relatively smooth because of the large number of data values
across each state; however, at these individual power plants and SC^
monitors, there are no regular seasonal patterns, and the seasonal
adjustment model fits relatively poorly.
Because of the irregular patterns in emissions and SC^ concentrations, and
because monthly mean concentrations are sometimes based on only a few data
values, correlations betwen emissions and S02 concentrations are not very
high, especially at the New York sites, as can be seen in Table 9. Table
9 also shows correlations between monthly emissions and SC^ concentrations
in the summer months only, April through October, inclusive. Because
there is only low-level mixing during the winter months, these months are
more likely to represent ground-level sources; for this reason, we
examined summer months separately. Restricting attention to just the
summer months improves the correlations slightly at the two New York
sites, even more so for the Virginia site. The second half of Table 9
presents correlations between yearly average SC^ concentrations and total
annual emissions at each site, both for all months in each year and for
just summer months. In general, because of the effect of smoothing the
data by averaging within each year, correlations between annual emissions
and annual SC^ levels are higher.
Correlations between emissions and average daily maximum SC^
concentrations are higher than correlations between emissions and monthly
mean SC^ concentrations, in general. The largest by far of the three
plants, Chesterfield, shows the highest correlations between SC^ emissions
and ambient SC^ concentrations, most likely because the large power plant
is the dominant SC^ source near the monitor. The two New York power
plants are much smaller and may not be the dominant SC^ source in the
vicinity of their respective associated SC^ monitors.
37
84161+ 3 °
-------
TABLE 9. Correlations between monthly SCL power plant emissions and
ambient S02 concentrations (monthly average and monthly average daily
maximum).
(1) Goudey power plant, SCL monitor 330480007F01; Binghamton, New York
Monthly data
All months Summer months
All months Summer months
Average
Maximum
.122
.266
-.165
.112
-.108
.118
-.111
-.115
(2) Hickling power plant, SCL monitor 331880003F01; Elmira, New York
Mnthl
All months Summer months
Yearly data
All months Summer months
Average
Maximum
.242
.286
.187
.355
-.062
.425
.332
.505
(3) Chesterfield power plant, SCL monitor 481560004F02; Richmond, Virginia
Monthly data
All months Summer months
Yearl
All months Summer months
Average
Maximum
.195
.292
.367
.461
.534
.687
.586
.725
38
-------
The correlations between emissions and S0£ concentrations in these local
sites can be summarized as follows:
Correlations are highest for point-source monitors near large power
plants;
Correlations between monthly emissions from power plants and monthly
average daily maximum SC^ concentrations are somewhat higher than
correlations between emissions and mean SC^ concentrations;
Correlations between power plant emissions and ambient SC^
concentrations are improved when the monthly data are aggregated to
yearly averages; and
Correlations between power plant emissions and ambient SC^
concentrations are higher when just the subset of summer months is
considered relative to all months. This is a period in which
emissions are high and mixing of emissions from elevated sources is
greatest.
CORRELATIONS AT THE REGIONAL LEVEL
We now examine the correlation between power plant SOg emissions and
ambient S02 concentrations for a large region of the northeastern United
States. This region, which includes and surrounds the Ohio River Valley,
consists of the six states Illinois, Indiana, Kentucky, Ohio, Pennsyl-
vania, and West Virginia. These states are among the seven states with
the highest annual S02 emissions from power plants.
Total monthly power plant emissions for these six states are presented in
Figure 10. As was seen for many state total emission plots, there is a
regular seasonal pattern of emissions peaking in both summer and winter,
with the winter peak higher than the summer peak. Total annual emissions
for the region are listed in Table 10; a consistent decrease in emissions
39
-------
950
CO
c
o
oo
c
o
00
00
LU
eg
o
en
900
850
800
750
700
650
600
550
1975 1976 1977 1978 1979 1980 1981 1982
Figure 10. Monthly Total Power Plant S02 Emissions
Ohio River Valley States, 1975 - 1982.
40
-------
TABLE 10. Ohio Piver Valley states annual power plant SCL emissions and
ambient SCL concentrations, 1975 - 1982.
(Includes Illinois, Indiana, Kentucky, Ohio, Pennsylvania, and West Virginia)
Annual power plant
S02 emissions, Annual average Annual average daily
Year 1000 tons S0_, ppm maximum S02* ppm
1975
1976
1977
1978
1979
1980
1981
1982
9775.7
9952.3
9599.7
8828.2
9057.6
8590.9
8088.2
7648.5
.0185
.0184
.0177
.0167
.0169
.0142
.0119
.0123
.0499
.0506
.0477
.0432
.0417
.0371
.0326
.0338
41
-------
took place beginning in 1976. Monthly average S02 and average daily
maximum S02 concentrations for these six states are presented in Figures
11 and 12, respectively. The S02 concentrations also show fairly regular
seasonal cycles, with peaks occurring during the stagnant winter months;
this pattern is more pronounced for mean S02 levels than for average daily
maximum S02 levels.
As can be seen in Table 10, regional power plant S02 emissions and ambient
S02 concentrations have decreased substantially in the region during the
study period. From 1975 to 1982, total six-state S02 emissions from power
plants decreased 22 percent, average ambient S02 concentrations decreased
33 percent, and average daily maximum S02 concentrations decreased 32
percent, indicating that emission reductions from sources other than power
plants have occurred. Indeed, if 71 percent of S02 emissions are from
power plants in the 1975 base year (as estimated by Gschwandtner and
Gschwandtner, 1983, p. 692) for the entire United States, then we estimate
that a 60 percent reduction in emissions occurred from 1975 to 1982 from
all sources in the area other than power plants (assuming a linear
relationship between S02 emissions and ambient concentrations).
The seasonal adjustment model applied to statewide monthly emissions and
ambient S02 concentrations was applied to regional emissions and ambient
S02 concentrations. The residuals from these models were then correla-
ted. There is a correlation of 0.749 between changes in emissions from
the seasonal pattern and changes in monthly average S02 from seasonal
patterns; for monthly average daily maximum S02 concentrations the
correlation is 0.766. Simple regression analyses reveal that an emissions
decrease of 100,000 tons of S02 from power plants in the region in a given
month from what would normally be expected for that month of the year is
associated with a decrease of .031 ppb in monthly average S02 (from what
would normally be expected for monthly average S02 for that month) and a
decrease of .084 ppb in monthly average daily maximum S02 (from what would
normally be expected for the month).
42
-------
0.030
0.025
0.020
Q-
Q-
C
o
-P
c
CD
O
c
o
CM
O
tn
0.015
0.010
0.005
0. 000
1975 1976 1977 1978 1979 1980 1981 1982
Figure 11. Monthly Rverage 502 Concentrations.
Ohio River Valley States, 1975 - 1982.
43
-------
0.080
0.075
0.070
0.065
0.060
^0.055
c
o
-U
c
0)
O
C
CM
O
to
0.050
0.045
0.040
0.035
0.030
0.025
0.020
1975 1976 1977 1978 1979 1980 1981 1982
Figure 12. Monthly flverage Daily Maximum S02 Concentrations-
Ohio River Valley States, 1975 - 1982.
44
-------
SUMMARY, FINDINGS, AND RECOMMENDATIONS FOR FURTHER WORK
For this project we have constructed two large data bases that are of
interest to researchers in the problems of acidic deposition. The first
data base consists of estimates of sulfur dioxide emissions from the
electric power industry in the northeastern United States for the years
1975 to 1982. The emissions estimates for individual plants were calcula-
ted from (1) monthly reports of fossil fuel consumption, and (2) descrip-
tive annual reports of fuels delivered to each plant. We believe that
this monthly emissions data base for individual plants is unique. The
second data base consists of monthly patterns of ambient sulfur dioxide.
Using data from the SAROAD system of the National Aerometric Data Branch,
we calculated monthly mean and average daily maximum concentrations for
all S02-recording monitors in the 22 northeastern states.
Examination of trends in monthly power plant S02 emissions and ambient S02
concentrations revealed seasonal patterns in both. Emissions of S02 from
power plants peak during the summer cooling season and the winter heating
season, whereas ambient S02 peaks during the stagnant winter months, with
some evidence of a secondary summer peak. Substantial decreases in both
emissions and ambient S02 occurred during the 1975-to-1982 study period.
For example, annual power plant emissions from the heavily industrialized
six-state Ohio River Valley region decreased 22 percent from 1975 to 1982,
and annual average ambient sulfur dioxide in the region decreased 32
percent during the same period.
Correlations between power plant S02 emissions and ambient S02 were also
examined. Because emissions and ambient S02 exhibit regular seasonal
patterns, but not the same seasonal patterns, emissions and ambient S02
trends were seasonally adjusted before correlations were examined.
45
-------
Correlations were then calculated at the local, state, and regional
levels. In general, the higher is the level of aggregation, the higher
are the observed correlations. Our conclusions from examination of a
select set of individual power plant and S02 monitor pairs are as follows:
(1) Correlations are highest for point-source monitors near large
power plants;
(2) Correlations between monthly emissions from power plants and
monthly average daily maximum S02 concentrations are somewhat
higher than correlations between monthly emissions and monthly
mean S02 concentrations;
(3) Correlations between power plant emissions and ambient S02
concentrations are improved when the monthly data are aggregated
to yearly averages; and
(4) Correlations between power plant emissions and ambient S02
concentrations are higher when just the subset of summer months
is considered relative to all months. This is a period in which
emissions are high and mixing of emissions from elevated sources
is greatest.
At the state level, correlations between seasonally adjusted power plant
S02 emissions and ambient S02 concentrations vary. In general, higher
correlations are observed in those states with higher levels of emis-
sions. The calculated correlations are affected by the availability of
ambient S02 monitoring datae.g., in a given state the number of opera-
ting monitors can vary greatly across time; monitors are not evenly
distributed but rather are centered in urban industrialized areas; and
most monitors, even if operating during the entire study period, have many
periods of missing data.
Correlations are highest at the regional level, where the greatest amount
of aggregation was performed. In the six-state Ohio River Valley region,
5 46
-------
the correlation between seasonally adjusted monthly power plant emissions
and seasonally adjusted monthly average ambient S02 was 0.749. With a
simple regression analysis, we calculated a decrease of 0.031 ppb average
S02 concentration for a 100,000 ton decrease in S02 emissions in the
region.
Since most of our work effort consisted of constructing the heretofore
nonexistent data bases of monthly power plant emissions for individual
plants and monthly summary statistics for individual S02 monitoring
stations in the 22 northeastern states, data bases of sulfate
concentrations (or a surrogate measure such as visibility) and sulfate
emissions in the area need to be compiled, and available acid
precipitation measurements need to be acquired. We feel that we have only
begun to analyze the needed data bases, and therefore have recommendations
for further analysis of these data bases.
The ambient S02 data base is considerably complicated by the irregular
periods of data measured by the existing monitors, and by the different
times when the monitors began recording. One possible solution is to use
only long-term monitors with a minimum amount of data each month or each
year. Another possibility is to construct weighted averages, where the
weights are proportional to the amount of data available for a monitor in
a given month. Two possible methods can be used to account for the heavy
concentration of S02 monitors in industrialized urban areas. One
calculates a weighted average of monitors in a given geographical area
where the weights are evenly distributed among urban, suburban, and rural
areas; the other calculates separate averages for urban, suburban, and
rural areas. Spatial averaging techniques, such as Kriging, two-
dimensional moving averages, and two-dimensional splines, can also be
applied to the monitoring data to down-weight monitors that occur in
clusters.
The emissions data base consists of S02 emissions from power plants
only. Although the majority of S02 emissions occur from power plants, we
do not know the extent to which S02 emission trends in the power plant
B"4ibt b 47
-------
sector are the same as those from other sources. It would be useful to
estimate monthly SC^ emissions from other sources and then examine trends
in total S02 emissions. Many studies have estimated annual S02 emissions
from sources other than power plants (e.g., Gschwandtner and Gschwandtner,
1983); these annual estimates can be disaggregated into monthly estimates
and added to our monthly emissions for specified geographical areas. In
addition, because ambient SC^ concentrations in certain parts of the
northeastern United States are affected by Canadian sources, it would be
useful to include in the data base monthly estimates of Canadian emis-
sions, especially those from southeastern Ontario.
It is often postulated that shorter stacks are associated with local
impacts and that taller stacks, through long-range transport, are associa-
ted with regional impacts over larger areas downwind of the stacks. In
our emissions data base, we did not consider stack height because of
project resource constraints. It would be useful, then, to apportion
emissions from each power plant by stack height. One difficulty to
overcome in carrying out this analysis arises in associating fuel consump-
tion data reported by the electricity generating unit with the appropriate
stacks at each plant.
Many additional statistical analyses can be performed on the S02 emissions
and ambient S0£ data bases. For example, time series analyses can be used
not only to seasonally adjust trends but also to relate trends with
different seasonal patterns. Regression analyses can be used to relate
emissions from multiple sources to ambient S02 recorded at one monitoring
station. In addition, principal components analysis and canonical
correlation are two techniques that can be applied to sets of emissions
sources and S02 monitors to determine relationships among them.
A key question that we have not attempted to answer in our analyses is
that of data requirements for detecting relationships. For example, how
are the emissions data and ambient S02 data best utilized to detect
associations, and what improvements, if any, can or need to be made to the
data bases? Also, once a relationship, linear or nonlinear, has been
:> 48
-------
detected, how is variability in the estimated degree of strength of the
relationship best estimated, considering the many sources of variability
in the data bases?
Finally, one of the most obvious research efforts to follow is that of
studying trends in acid precipitation data and relating them to trends in
emissions and acid precipitation precursors. Many acid precipitation
monitoring networks are currently operating throughout the northeastern
United States, some have been in operation for most of the 1975-1982 study
period. Such monitoring networks are operated by the U.S. Geological
Survey, National Acidic Deposition Program, and the Utility Acid Precipi-
tation Study Program. Analysis of acid precipitation data, however, must
proceed carefully, for it is necessary to take into account not only
different sampling schedules (e.g., bulk monthly collection versus event-
only collection), but also different measurement methodologies and site
locations.
49
-------
REFERENCES
Burton, C. S., J. P. Nordin, and T. E. Stoeckenius. 1982. "Variability
(Uncertainty) in Sulfur Emissions: A Summary of Current Knowledge
and the Effect on Ambient Standard Attainment Demonstrations of
Adopting Some Simple Models of Sulfur Variability." Paper presented
at the AMS conference in Woods Hole, Massachusetts, September 1982.
E. H. Pechan & Associates, Inc. 1982. "Estimates of Sulfur Oxide
Emissions from the Electric Utility Industry. Volume I - Summary and
Analysis." E. H. Pechan & Associates, Inc., Springfield, Virginia
(EPA-600/7-82-061a).
EPA. 1982. "Compilation of Air Pollutant Emission Factors." 3rd ed.,
Supplement 13. U.S. Environmental Protection Agency, Research
Triangle Park, North Carolina (AP-42).
EPA. 1983. "AEROS Manual Series Volume V: AEROS Manual of Codes (Second
Edition)." U.S. Environmental Protection Agency, Office of Air
Quality Planning and Standards, Research Triangle Park, North
Carolina (EPA-450/2-76-005a).
Gschwandtner, G., and K. C. Gschwandtner. 1983. "Historic Emissions of
Sulfur and Nitrogen Oxides in the United States from 1900 to 1980."
To be published.
Neligan, R. E. 1978. Memorandum to directors of the Surveillance and
Analysis divisions and Air and Hazardous Materials Division, and the
regional quality control coordinators, EPA Regions I through IX, U.S.
Environmental Protection Agency, 25 July 1978.
Pechan, E., and S. Rothschild. 1983. "Newly Revised Stack File."
Memorandum from E. H. Pechan & Associates, Inc., Springfield,
Virginia.
PEDCo Environmental, Inc. 1982. "Flue Gas Desulfurization Information
System Data Base User's Manual." PEDCo Environmental, Inc.,
Cincinnati, Ohio (NTIS PB83-146209).
Thrall, A. D., J. L. Baptista, and C. S. Burton. 1984. "An Examination
of Air Quality Data Completeness Requirements." Systems
Applications, Inc., San Rafael, California (SYSAPP-83/185).
emeu e R-l
-------
73
41
Figure Al. Connecticut SOZ Monitoring Sites
-------
76
39
Figure A2. Delaware S02 Monitoring Sites
-------
o
Figure A3. Dist. of Columbia S02 Monitoring Sites
-------
87
Figure A4. Illinois SOZ Monitoring Sites
-------
B6
41
40
39
38
BB
Figure A5. Indiana S02 Monitoring Sites
-------
89
88
87
86 85
-i !H*-»-
84
83
37
36 4
36
83
82
Figure A6. Kentucky SOZ Monitoring Sites
-------
71
70
69
44
70 69
Figure A7. Maine S02 Monitoring Sites
67
-------
79
78
- 38
76
Figure A8. Maryland S02 Monitoring Sites
-------
43
42
70
Figure A9. Massachusetts S02 Monitoring Sites
-------
47
46
45
44
91 90 89
j j n\ i i M ; i ..:, i , = :.
: f r : '
1 :
88
85
84
83
82
1 , , , ' s i i ; ! i
..-,., .. i , i , . , . i_» j i ,
47
.._..
-{
i
]
!" i 1
i ! !
...L..-I. .
f]
!
i j .
' ' , i , ' ' ! ! i
4- , , f ; * 1 » t , U
i i 1 ' i i 1 ' i
4 I ,- i ] .. * .4 - i | li»4 ) i
i ' i i ' < ' ' > '
..1... 1 i .'..,. ..!...! i 1 i ,..!.... i ,1
f 1- 4 / , 1 I I 4 , J 1
/ ' / i i i ' '
^"? >/ I i * < < ' ' l {
U V/; ^ | | »t |r
. I : t | | 4 i , i i t . , i ,
, , , < \ 1 , , ! ! 1 r \ ,-, i
' , i ! ' l
4 11-4 1 '- . t ' ,
>4llf«ii'l~| L t i ', \
'*'',. j I >- .
t »
T.I I
rv. , , , - , , , \ - * , , ,
43
i
i
! ] . !
--if
-4 4 : i-
L ,. i i i , i i t ,
i i i ' * i I I
' ' i i ' i i i , ,
i t . , , ! , (- , , i j i j. ; « .1 , j , j , i , i i , i
i i i t t . i i
i i i i , i i i , , . .
i;,;; [;i; i ;r;;j; | r;;;,; i ;;.
»4;,,:.: i'
4-
1 r ' * ' " j :<*'!' i - *
! . . . . I ! ! . ! 1 . . .
42
rtt
L
4
i
'"! !""
i
...]..
i1
91
42
90
89
88
Figure A10. Michigan S02 Monitoring Sites
-------
73
71
Figure All. New Hampshire S02 Monitoring Sites
-------
74
39
- 39
74
Figure A12. New Jersey SOZ Monitoring Sites
-------
45
44
43
42
41
41
Figure A13. New York S02 Monitoring Sites
-------
84
79
78
77
76
34
33
83
82
81
80
79
78
77
76
Figure AH. North Carolina S02 Monitoring Sites
-------
82
81
80
41
40
41
39
39
Figure A15. Ohio S02 Monitoring Sites
-------
80
79
40
40
76
75
Figure A16. Pennsylvania SOZ Monitoring Sites
-------
42
42
71
Figure A17. Rhode Island S02 Monitoring Sites
-------
90 89
H4 K-i t-
i ! i ! ! U
88
87
86
34
90
82
Figure A18. Tennessee SOZ Monitoring Sites
-------
45
Figure A19. Vermont S02 Monitoring Sites
-------
83
82
81
444-
80
37
36
36
77
76
Figure A20. Virginia S02 Monitoring Sites
-------
78
40
39
38
40
39
38
82
81
80
79
78
Figure A21. West Virginia S02 Monitoring Sites
-------
94
43
Figure A22. Wisconsin S02 Monitoring Sites
-------
0.021
0.021
D_
a.
-0.018
-P
CD
-P
C
S0.015
c
o
LJ
£0.012
01
O)
o
"0.009
(X
^0.006
o
0.003
0.000
V)
a<
tn
10
00 i-
o 5
_L
1975 1976 1977 1978 1979 1980 1981 1982
Figure Bl. Monthly flverage S02 Concentration and
Number of Reporting Sites. Connecticut. 1975 - 1982.
-------
0.030 -
0.027r-
£0.024 -
c
^0.02lh
00
0.018
o
c.
o
CJ
csj 0.015
o
en
en
^0.012
Q)
>
cr
0.009
°0.006
0.003
0.000
-------
0.036
0.032
CO-
OL
c
o
-p
CO
c
CD
o
50.
CM
O
cn
0) 0,
O)
CD
t_
Q)
0.028 -
024 -
020 -
016 -
1.008 -
0.004 -
0.000
1975 1976 1977 1978 1979 1980 1981 1982
Figure 83. Monthly Overage S02 Concentration and
Number of Reporting Sites, District of Columbia. 1975 - 1982.
-------
1975 1976 1977 1978 1979 1980 1981 1982
Figure EH. Monthly Rverage S02 Concentration and
Number of Reporting Sites. Illinois. 1975 - 1982.
-------
0.030 -
1975 1976 1977 1978 1979 1980 1981 1982
Figure B5. Monthly Rverage S02 Concentration and
Number of Reporting Sites. Indiana, 1975 - 1982.
-------
0.027 -
1975 1976 1977 1978 1979 1980 1981 1982
Figure B6. Monthly Rverage S02 Concentration and
Number of Reporting Sites. Kentucky, 1975 - 1982.
-------
0.045 -
1975 1976 1977 1978 1979 1980 1981 1982
Figure 87. Monthly flverage S02 Concentration and
Number of Reporting Sites, Maine. 1975 - 1982.
-------
i o
0.030
1975 1976 1977 1978 1979 1980 1981 1982
Figure B8. Monthly flverage S02 Concentration and
Number of Reporting Sites, Maryland. 1975 - 1982.
-------
Q.
Q-
0.022
0.020
0.018
c
o
^0.016
CD
u
C
o
I.0H
1.012
CM
O
in
o> 0.010
CD
l_
> 0.008
cr
>^
JE 0.006
c
o
0.004
0.002
0.000
m
32
I
I
I
I
I
1975 1976 1977 1978 1979 1980 1981 1982
Figure B9. Monthly flverage S02 Concentration and
Number of Reporting Sites, Massachusetts, 1975 - 1982.
-------
0.020 -
0.018
a-0.016
Q.
c
- 0.014
CO
-P
o0.012
o
c
o
CM
O
CD
0)
CO
CD
d
0.010
1.008
0.006
-P
c
0.004
0.002
0.000
CO
Q}
m 20
<4-
o
2 10
I
I
I
I
1975 1976 1977 1978 1979 1980 1981 1982
Figure B10. Monthly Rverage S02 Concentration and
Number of Reporting Sites. Michigan, 1975 - 1982.
-------
0.050
_ 0.045
Q_
Q_
C
o
-p
CO
c_
-p
C
Q)
0.040
0.035
1.030
O)
CO
C_
Q)
>
(X
>,
0.020
^0.015
si.
-P
C
o
0.010
0.005
0.000
CO
a>
-p
CO
0)
JD
_L
JL
1975 1976 1977 1978 1979 1980 1981 1982
Figure Bll. Monthly flverage S02 Concentration and
Number of Reporting Sites, New Hampshire. 1975 - 1982.
-------
0.030
0.027
c
o
-£0.021
c_
-P
C
S0.018
c
o
CJ
£0.015
en
CD
£0.012
CD
cr
>.0.009
JT
^J
C
£0.006
0.003
0.000
CO
(^20
Q-
O
510
I
I
I
I
I
1975 1976 1977 1978 1979 1980 1981 1982
Figure B12. Monthly flverage S02 Concentration and
Number of Reporting Sites. New Jerseyt 1975 - 1982.
-------
0.033 -
1975 1976 1977 1978 1979 1980 1981 1982
Figure B13. Monthly flverage S02 Concentration and
Number of Reporting Sites. New York, 1975 - 1982.
-------
0.032 -
5.0.028
a.
c
o
OD
t_
-P
C.
|0.020
o
CJ
CM
° 0.016
Q)
O)
ro
<_
>0.012
CE
£0.008
o
0.004
0.000
CD
to
O
t_
CO
/₯V
i
i
J_
I
I
1975 1976 1977 1978 1979 1980 1981 1982
Figure BH. Monthly flverage S02 Concentration and
Number of Reporting Sites. North Carolina, 1975 - 1982.
-------
0.030
0.027
Q_
DL
~0.024
c
o
ro 0.021
c
CD
O
LJ
CNJ
CD
O)
CD
1.018
1.015
0.012
CD
CE
^0.009
-p
c
o
0.006
0.003
0.000
Q)
-P
cn32
O
t_
CD 4 r*
_Q 16
I
I
1975 1976 , 1977 1978 1979 1980 1981 1982
Figure 815. Monthly flverage S02 Concentration and
Number of Reporting Sites, Ohio, 1975 - 1982.
-------
1975 1976 1977 1978 1979 1980 1981 1982
Figure B16. Monthly Overage S02 Concentration and
Number of Reporting Sites, PennayI vania, 1975 - 1982,
-------
0.045
0.040
5-0.035
c
o
"ro 0.030
c
CD
£0.025
CM
O
cn
CD
CD
CO
£_
CO
0.020
0.015
.c
"c 0.010
o
0.005
0.000
00
CD
-p
CD
U-
O
0)
FDCnFT
I
1975 1976 1977 1978 1979 1980 1981 1982
Figure B17. Monthly flverage S02 Concentration and
Number of Reporting Sites. Rhode Island. 1975 - 1982.
-------
0.018 -
0.016 -
5-0.014
c
o
-p
C
CD
|0.010
0.008
CM
o
en
03
O)
(0
>0.006
"c 0.004
o
0.002
0.000
CD
-P
o
C-
0>
1975 1976 1977 1978 1979 1980 1981 1982
Figure B18. Monthly Rverage 302 Concentration and
Number of Reporting Sites. Tennessee. 1975 - 1982.
-------
0.027
~ 0.021
£
DL
o_
c 0.021
o
ro
0.018
CD
O
C
0.015
CM
O
CO
2,0.012
01 0.009
i0.006
0.003
0.000
-------
0.024 -
1975 1976 1977 1978 1979 1980 1981 1982
Figure B20. Monthly flverage S02 Concentration and
Number of Reporting Sites. Virginia. 1975 - 1982.
-------
0.040 -
0.036 -
Q.
D.
0.032 -
c
o
-P
c
Q)
CVJ
0 0
tn " '
028
024
020 -
03
O)
CD
0.016 -
-P
c
o
0.004 -
0.000
1975 1976 1977 1978 1979 1980 1981 1982
Figure B21. Monthly flverage S02 Concentration and
Number of Reporting Sites. West Virginia, 1975 - 1982.
-------
0.022 -
1975 1976 1977 1978 1979 1980 1981 1982
Figure B22. Monthly Rverage S02 Concentration and
Number of Reporting Sites. Wisconsin. 1975 - 1982.
-------
1975 1976 1977 1978 1979 1980 1981 1982
Figure Cl. Monthly flverage Daily Maximum S02 Concentration
and Number of Reporting Sites. Connecticut. 1975 - 1982.
-------
0.054
_0.048
E
Q.
CL
C0.
O
-P
CD
£0,
c
CD
(J
c
£ 0,
C\J
O
en
to
c_
0}
>
ec0.
042
036
030
024
018
00-012
0.006
0.000
CO
CD
o
c_
CD
1111111111111
_L
J\
_L
_L
vv vv
J_
_L
1975 1976 1977 1978 1979 1980 1981 1982
Figure C2. Monthly flverage Daily Maximum S02 Concentration
end Number of Reporting Sites. Delaware. 1975 - 1982.
-------
0.063
0.056
E
a.
DL
0.049 -
0.042 -
c
0)
(J
50.
(M
O
CD
0> 0,
O)
o
c_
0)
>>
c.
035 -
028 -
021 -
I. 014 -
0.007 -
0.000
1975 1976 1977 1978 1979 1980 1981 1982
Figure C3. Monthly Rverege Daily Maximum S02 Concentration
and Number of Reporting Sites. District of Columbia. 1975 - 1982.
-------
0.063
0.056
Q.
Q-
0.039
(D
£0.042
c
0)
u
c
20.035
(M
O
cn
o>0.028
O)
ID
L_
OD
^0.021
0.007
0.000
n
(9
-P
11 111 I 11 11 11 I
I
I
I
I
I
I
I
1975 1976 1977 1978 1979 1980 1981 1982
Figure C4. Monthly flverege Daily Maximum S02 Concentration
and Number of Reporting Site3t Illinois. 1975 - 1982.
-------
0.072
0.064
CL
CL
0.056
c
o
-p
£0.048
c
CD
O
C
o a
cj "
(M
o
en
0,0.032
O)
o
CD
£0.024
1.016
0.006
0.000
1975 1976 1977 1978 1979 1980 1981 1982
Figure C5. Monthly flverage Daily Maximum S02 Concentration
and Number of Reporting Sites. Indiana. 1975 - 1982.
-------
0.090 -
1975 1976 1977 1978 1979 1980 1981 1982
Figure C6. Monthly flverage Daily Maximum S02 Concentration
end Number of Reporting Sites. Kentucky. 1975 - 1982.
-------
0.126 -
1975 1976 1977 1978 1979 1980 1981 1982
Figure C7. Monthly flverage Daily Maximum S02 Concentration
and Number of Reporting Sites. Maine. 1975 - 1982.
-------
0.060
Q.
5-0.048
c
o
£0.042
c
S0.036
o
LJ
S0.030
en
o>
O)
£0.024
Q)
>
cr
>.0.018
x:
-p
c
i0.012
0.006
0.000
-------
0.240F
0.216 -
O-
Q.
-p
to
o
c
o
0.192
168
CM
O
cn
CD
0. 120
096
CD
>
cc
0.072
-P
o 0,
018
0.024
0.000
32
n
CO
-P
in
<*_
o
SIB
E
3
I I I 11
_L
_L
_L
1975 1976 1977 1978 1979 1980 1981 1982
Figure C9. Monthly flverage Daily Maximum S02 Concentration
and Number of Reporting Sites. Massachusetts. 1975 - 1982.
-------
0.050
0.045
0.0.040
CL
c
--0.035
CD
00.030
o
c.
o
cj 0.025
o
tn
OJ
o
i_
0}
CE
1.020
0.015
JC
-P
c
1.010
0.005
0.000
n
CD
-P
10
I
I
I
I
I
I
I
1975 1976 1977 1978 1979 1980 1981 1982
Figure C10. Monthly flverage Daily Maximum S02 Concentration
and Number of Reporting Sites. Michigan, 1975 - 1982.
-------
1975 1976 1977 1978 1979 1980 1981 1982
Figure Cll. Monthly flverage Daily Maximum S02 Concentration
end Number of Reporting Sites. New Hampshire, 1975 - 1982.
-------
0.053
0.038
e
D.
c
o
|0.036
-P
c
CD
U
00.030
CJ
CM
O
in
CD
OJ
CD
0.023
CD
£ 0.018
C0.012
o
0.006
0.000
V)
CD
-P
E
3
I
I
1975 1976 1977 1978 1979 1980 1981 1982
Figure C12. Monthly Overage Daily Maximum S02 Concentration
and Number of Reporting Sites. New Jersey. 1975 - 1982.
-------
0.180
E0.160
D.
o_
§0.140
C0.120
oj
u
c
o
U0.100
CM
O
cn
00.080
to
c_
"-0.060
>.
r:
00.040
0.020
0.000
o>
I
I
I
I
I
1975 1976 1977 1978 1979 1980 1981 1982
Figure C13. Monthly Overage Daily Maximum 502 Concentration
and Number of Reporting Sites. New York. 1975 - 1982.
-------
0.090 -
0.080 -
Q-
D.
C
o
0.070 -
0.060 -
CD
c_
-p
C
o>
o
°0.050
eg
o
in
a> 0.040
O)
CD
l_
CD
er 0.030
|0.020
0.010 -
0.000
-------
0.150
0.135
5.0.120
Q-
.20.105
-4J
£0.090
o
c
o
CM
O
in
CD
0.075
.060
0)
a:
>»
.c
0.045
o 0.030
0.015
0.000
D
CD
-P
16
_L
_L
1975 1976 1977 1978 1979 1980 1981 1982
Figure CIS. Monthly flveroge Daily Maximum 302 Concentration
and Number of Reporting Sites. Ohio. 1975 - 1982.
-------
0.081
-0.072
D_
Q.
-p
CD
c_
£0.054
0>
CJ
c
o
CM
O
cn
§,0.036
Q>
>
cr
0.027
-P
00.018
0.009
0.000
o>
-p
^32
u-
o
16
J_
_L
tin
11111111111111
1975 1976 1977 1978 1979 1980 1981 1982
Figure C16. Monthly Rverage Daily Maximum S02 Concentration
and Number of Reporting Sites. Pennsylvania. 1975 - 1982.
-------
0.060
0.054
D.
c
o
"0.042
-4J
C
0)
O
(_>
CM
1.036
1.030
CD
en
to
c.
0)
0.024
c
o
0.012
0.006
0.000
V)
0)
en
<4-
°2
c.
CD
XI
I I 11 I I I I I I I I I I I I I I I I I I I I I I I I I I I I I II I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I
TlLJlFT
I
1975 1976 1977 1978 1979 1980 1981 1982
Figure C17. Monthly flverage Daily Maximum S02 Concentration
and Number of Reporting Sites. Rhode Island. 1975 - 1982.
-------
0.070
0.063
E
£0.056
c
o
CD
i_
-P
C
S0.042
c
o
LJ
£0.035
en
05
O)
£0.028
0)
>0.021
c
0. 0H
0.007
0.000
l
I
I
l
1975 1976 1977 1978 1979 1980 1981 1982
Figure CIS. Monthly flverage Daily Maximum 502 Concentration
and Number of Reporting Sites, Tennessee, 1975 - 1982.
-------
0.050 -
1975 1976 1977 1978 1979 1980 1981 1982
Figure C19. Monthly Rverage Daily Maximum 502 Concentration
and Number of Reporting Sites. Vermont. 1975 - 1982.
-------
0.040
0.036 -
0.032
Q_
0_
C
.50.028
ID
t_
-P
O
C
o
1.024
tM
O
tn
0.020
g»0.016
£_
0>
1.012
o 0.008
0.004
0.000
V)
0}
-p
10
03
XI
E
3
I
I
1975 1976 1977 1978 1979 1980 1981 1982
Figure C20. Monthly flverage Daily Maximum S02 Concentration
and Number of Reporting Sites. Virginia. 1975 - 1982.
-------
0.090
~0.080
E
CL
O.
C0.070
o
CD
*? 0.060
CD
o
c.
"0.050
CM
O
tn
03
CC0..030
§0.020
0.010
0.000
en
o>
in
««-
o
o>
xi
1975 1976 1977 1978 1979 1980 1981 1982
Figure C21. Monthly flverage Daily Maximum 502 Concentration
and Number of Reporting Sites. West Virginia, 1975 - 1982.
-------
0.048
D.
D-
0.042
c
o
-p
0.036
c
CD
O
1.030
CM
O
cn
0,0.024
a>
CD
cr 0.018
1.012
0.006
0.000
on
CD
o
c_
CD
E
3
I 11 I I I I I I I I I I I I II 111 I 1 111 11 I I I 11 I I I I I I I 11 I I I I I I 11 I 11 III I I 11 I I I II 11 I I I I I 11 I I I I I I I 11 I I I I I I I I I I I I I
I
I
1975 1976 1977 1978 1979 1980 1981 1982
Figure C22. Monthly Overage Daily Maximum S02 Concentration
and Number of Reporting Sites. Wisconsin. 1975 - 1982.
-------
V)
c 4
o "
r>
I 3
«M
o
«0
0
I
I I I I I I
1975
Figure 01.
1976 1977
1978
1979
1980
1981
1982
Monthly Total Electric Utility 502 Emissions.
Connecticut. 1975 - 1982.
-------
8
- 7
n
c
o
+3
V)
i s
-------
2.50
2.00
c
o
1.50
ID
C
O
E
UJ
tM
Si.00
.*>
o
0.50
0.00
1975
1976
1977
1978
1979
1980
1981
1982
Figure D3. Monthly Total Electric Utility 502 Emissions.
District of Columbia. 1975 - 1982.
-------
200
150
c
o
-p
c
-100
n
r>
CM
o
Vt
-p
o
50
e
1975
1976
1977
1978
1979
1980
1981
1982
Figure D4. Monthly Total Electric Utility 502 Emissions.
IIlinois. 1975 - 1982.
-------
180
150
i 120
r»
I"
V)
CM
O
V)
.**
O
60
30
0
I I I
i i
I I
i I
1975
1976
1977
1978
1979
1980
1981
1982
Figure D5. Monthly Total Electric Utility S02 Emissions.
Indiana. 1975 - 1982.
-------
150
120
c
o
oa
69
ca
«n
c
o
f)
t)
UJ
CM
O
(O
JJ
o
1975
Figure D6.
1976
1977
1978
1979
1980
1981
1982
Monthly Total Electric Utility S02 Emissions.
Kentucky. 1975 - 1982.
-------
3.00
1975
1976
1977
1978
1979
1980
1981
1982
Figure D7.
Monthly Total Electric Utility S02 Emissions.
Maine. 1975 - 1982.
-------
25
20
c
o
15
c
o
f)
UJ
lB
o
-P
o
I I
1975
1976
1977
1978
1979
1980
1981
1982
Figure D8. Monthly Total Electric Utility S02 Emissions.
Maryland. 1975 - 1982.
-------
25
20
eg
69
VI
2 is
r
V)
LU
CM
o
5 IB
o
I i i
i i 1
i I
i I t i i I
1975
1976 1977
1978
1979
1980
1981
1982
Figure D9. Monthly Total Electric Utility S02 Emissions.
Massachusetts. 1975 - 1982.
-------
100
1975 1976
Figure D10.
1977
1978
1979
1980
1981
1982
Monthly Total Electric Utility 502 Enissiona.
Michigan. 1975 - 1982.
-------
8
c
o
r»
c
o
UJ
CM
O
0
-P
o
0
I I I I
1 I
I I I I
I I
I I I I I I
1975
1976
1977
1978
1979
1980
1981
1982
Figure Oil.
Monthly Total Electric Utility S02 Emissions.
New Hampshire. 1975 - 1982.
-------
15
12
r>
c
o
n
c
o
r>
10
at
Csl
o
V)
-p
o
I I 1
i I i
i I
i i I
i I i
I i
1975
1976
1977
1978
1979
1980
1981
1982
Figure D12.
Monthly Total Electrifc Utility 502 Emissions.
New Jersey. 1975 - 1982.
-------
V)
eg
eg
r»
r>
n
UJ
cvj
o
V)
"20
o
10
0
I i
. «
1975
1976
1977
1978
1979
1980
1981
1982
Figure D13.
Monthly Total Electric Utility S02 Emissions.
New York. 1975 - 1982.
-------
50
c
o
§30
C
O
UJ
S20
(D
O
O
10
i I
i I
i l i
I
i I i i i I i i i I
1975
1976
1977
1978 1979
1980
1981
1982
Figure DH.
Monthly Total Electric Utility 302 Emissions,
North Carolina. 1975 - 1982.
-------
300
250
200
n
c
- 150
n
n
UJ
o
to
5 100
o
50
0
i i
1975 1976 1977 1978
1979
1980
1981 1982
Figure D15. Monthly Total Electric Utility 502 Emissions.
Ohio. 1975 - 1982.
-------
200
150
o
c
o
t»
c.
-100
UJ
cvj
o
V)
-P
o
50
0
I i i
I i
i i
J I
1975
1976
1977
1978
1979
1980
1981
1982
Figure D16.
Monthly Total Electric Utility 502 Emissions,
Pennsylvania. 1975 - 1982.
-------
1.00
*
0.90
0.80
-0.70
n
c
o
-p
0.60
c
-0.50
CO
0.30
0.20
0.10
0.00
1975 1976 1977 1978 1979 1980
1981
1982
Figure D17. Monthly Total Electric Utility S02 Emissions.
Rhode Island. 1975 - 1982.
-------
150
120
c
o
C9
CO
n
c
o
90
60
to
c
-p
o
30
I i i
1975
1976
1977
1978
1979
1980
1981
1982
Figure D18. Monthly Total Electric Utility S02 Emissions.
Tennessee. 1975 - 1982.
-------
0.150
0.H0
0.130
0.120
0.110
§0.100
-P
£0.090
£0.080
.!! 0.070
£
UJ
*0.050
o
0.030
0.020
0.010
0.0001 '.' ' ' '.
I i i i I i i iN i i
1975
1976
1977
1978
1979
1980
1981
1982
Figure D19. Monthly Total Electric Utility S02 Emissions.
Vermont. 1975 - 1982.
-------
30
25
C20
o
.**
n
c
2 is
n
n
UJ
CSJ
o
V)
10
0
j I
i I
I i i
I i
i I
1975
1976 1977
1978
1979
1980
1981
1982
Figure D20. Monthly Total Electric Utility 502 Emissions.
Virginia. 1975 - 1982.
-------
120
100
80
n
I 60
n
E
LU
CM
O
VJ
5 40
o
20
0
I I
i i
I I
1975
1976
1977
1978
1979
1980
1981
1982
Figure D21.
Monthly Total Electric Utility S02 Emissions.
Nest Virginia. 1975 - 1982.
-------
n
o
r>
i 30
r>
n
CM
O
en
520
o
10
0
i i
I
i i I i
i I i i i I
1975 1976
Figure D22.
1977 1978
1979
1980
1981
1982
Monthly Total Electric Utility S02 Emissions.
Wisconsin. 1975 - 1982.
------- |