US. ENVIRONMENTAL PROTECTION AGENCY
REGION VIII
AIR & HAZARDOUS MATERIALS DIVBION
DENVER . COLORADO 8C5295
-------
Publication No. EPA-908/1-76-009
PEDCo- ENVIRONMENTAL
NORTH DAKOTA AIR QUALITY
MAINTENANCE AREA ANALYSIS
Prepared by
PEDCo-ENVIRONMENTAL SPECIALISTS, INC.
Suite 13, Atkinson Square
Cincinnati, Ohio 45246
Contract No. 68-02-1375
EPA Project Officer: David Kircher
U.S. ENVIRONMENTAL PROTECTION AGENCY
Region VIIJ
Air Planning & Operations Section
Denver, Colorado 80203
SUITE 13 • ATKINSON SQUARE
CINCINNATI. OHIO 45246
513 / 771-4330
Task Order No, 19
Prepared for
June 1976
ipamxPD
25535-
BRANCH OFFICES
Suite 110, Crown Center
Kinui City, Mo. 64108
Suite 104-A, Profeulonal Village
Chapel Hill, N.C. 37614
-------
This report is issued by the Environmental Protection Agency to
report technical data of interest to a limited number of readers.
Copies are available free of charge - as supplies permit - from
the Air and Hazardous Materials Division, Region VIII, Environmental
Protection Agency, Denver, Colorado 80295, or may be obtained, for
a nominal cost, from the National Technical Information Service,
5285 Port Royal Road, Springfield, Virginia 22151.
This report was furnished to the Environmental Protection Agency
by PEDCo-Environmental Specialists, Inc., Suite 13, Atkinson Square,
Cincinnati, Ohio 45246, in fulfillment of Contract No. 68-02-1375.
The contents of this report are reproduced herein as received from
the contractor. The opinions, findings, and conclusions expressed
are those of the authors and not necessarily those of the Environmental
Protection Agency.
Publication No. EPA-908/1-76-009
-------
ACKNOWLEDGEMENT
This report was prepared for the U.S. Environmental
Protection Agency and the North Dakota State Department of
Health by PEDCo-Environmental Specialists, Inc. Mr. Kenneth
Axetell was the PEDCo project manager and principal author
of this report.
Mr. David Kircher was the project officer for the U.S.
Environmental Protection Agency. Messrs. Dana Mount and Jay
Crawford were the principal contacts with the North Dakota
Department of Health, Division of Environmental Engineering.
The author appreciates their contributions to this study.
ii
-------
CONTENTS
Pa9g
1. INTRODUCTION 1-1
2. SUMMARY 2-1
3. McLEAN-MERCER-OLIVER AQMA 3-1
4. CASS COUNTY AQMA 4-1
REFERENCES R-l
APPENDIX
A. NORTH DAKOTA AQMA POINT SOURCE SUMMARY A-l
B. AIR EMISSION SOURCES FROM A LURGI DRY-ASH B-l
GASIFICATION FACILITY USING LIGNITE COAL
iii
-------
FIGURES
No. Page
1.1 North Dakota AQMA's 1-2
3.1 McLean-Mercer-Oliver AQMA Analysis Area 3-4
3.2 Area Source Grids 3-9
3.3 Receptor Locations 3-23
3.4 1974 Geometric Mean Particulate 3-28
Concentrations
3.5 Predicted 1980 Geometric Mean Particulate 3-31
Concentrations
3.6 Predicted 1985 Geometric Mean Particulate 3-32
Concentrations
3.7 Predicted 1985 Annual Arithmetic Mean S02 3-35
Concentrations
4.1 Cass County AQMA Analysis Area 4-4
4.2 Area Source Grids for Caes County 4-7
4.3 Receptor Locations 4-11
4.4 1974 Geometric Mean Particulate 4-14
Concentrations in Cass County AQMA
iv
-------
TABLES
NO.
1.1 National Ambient Air Quality Standards
for Designated Pollutants
2.1 Results of AQMA Analyses
3.1 Locations of Point Sources
3.2 Comparison of Estimated Annual Lignite
Consumption from Coal Impact Study and
from Megawatt Capacity
3.3 Point Source Emissions in Analysis Area
3.4 Area Source Particulate Emissions for
McLean-Mercer-Oliver Counties
3.5 Area Source Emissions of SO2, Hydrocarbons, 3-15
and Oxides of Nitrogen for AQMA Counties
3.6 Apportioning Factors by Source Category 3-16
3.7 Particulate Emissions by Grid and Source 3-17
Category, 1974
3.8 Particulate Emissions by Grid and Source 3-18
Category, 1980
3.9 Particulate Emissions by Grid and Source 3-19
Category, 1985
3.10 Comparison of Model-Predicted and Measured 3-26
Pollutant Concentrations in AQMA Analysis
Area (Annual Averages)
3.11 Comparison of Model-Predicted and Measured 3-30
Pollutant Concentrations in AQMA Analysis
Area (Short-term Peak)
3.12 Source Contributions to Predicted 1985 3-34
Ambient Concentrations at Selected Receptors
Page
1-3
2-2
3-8
3-11
3-13
3-14
v
-------
No. Page
3.13 Oxidant Sampling Data—Selected Sites in EPA 3-38
Region VIII
3.14 Ozone Data for Ohio, Pennsylvania, and 3-40
Maryland
3.15 County-Wide Hydrocarbon Emission Densities 3-42
4.1 Particulate Emissions for Cass County 4-5
4.2 Cass County Emissions by Grid and Source 4-9
Category, 1974
vi
-------
1. INTRODUCTION
The North Dakota Department of Health, Environmental
Engineering Division has identified two areas that have the
potential of exceeding the National Ambient Air Quality
Standards (NAAQS) because of existing air quality and/or
projected growth over the next ten years (1975 to 1985) .
The U.S. Environmental Protection Agency has published these
areas as designated Air Quality Maintenance Areas (AQMA's).
The two AQMA's are McLean-Mercer-Oliver and Cass County.
McLean-Mercer-Oliver is designated for four air pollutants:
particulates, sulfur dioxide, oxidants, and oxides of nitro-
gen. Cass County is only designated for particulates.
Figure 1.1 shows the location of the two AQMA's.
Once designated as an AQMA, a detailed analysis of the
impact of projected growth on air quality is required. This
report presents such an analysis for the two areas.
If the AQMA analysis demonstrates that neither the
primary nor secondary NAAQS will be exceeded, no plan for
maintenance of standards is required and the AQMA can
possibly be de-designated. However, should the analysis
show a problem in attaining the NAAQS by 1975 and/or main-
taining the standards in the future, revisions to the North
Dakota State Implementation Plan (SIP) and development of an
air quality maintenance plan will be required.
The procedures used in the AQMA analysis are consistent
with the proposed regulations on maintenance of National
Ambient Air Quality Standards, 40 CFR, Part 51.^" Table 1.1
shows the primary and secondary NAAQS for the four desig-
nated pollutants. A ten year projection period has been
1-1
-------
I
ro
PEMBINA
TOWNER CAVALIER
BOTTINEAU
RENVILLE
Dl VIDE
BURKE
Mf HENRY
WILLIAMS
WALSH
RAMSEY
MOUNTRAIL
BENSON
6RAND FORKS
NELSON
EDDY
¦HERIDAN |WELLS
TRAILL
M SKENlfE
STUTSMAN
KIDDER
\URLEI6H
BILLINGS
60LDEN
VALLEY
BARNES
RICHLAND
RANSOM
LA MOURE
LOGAN
EMMONS
GRANT
HETTINGER
mvintosm
BOWMAN
ADAMS
SfOUX
Figure 1.1. North Dakota AQMA's.
-------
Table 1.1. NATIONAL AMBIENT AIR QUALITY STANDARDS
FOR DESIGNATED POLLUTANTS
Primary NAAQS, Secondary NAAQS,
Pollutant Frequency ug/m^ ug/m3
Suspended
particulate
Annual geometric mean
Maximum 24 hours
75
260
60
150
Sulfur
dioxide
Annual arithmetic mean
a
Maximum 24 hours
Maximum 3 hours3
80
365
1300
Nitrogen
dioxide
Annual arithmetic mean
100
100
Photochemical
oxidants
Maximum 1 houra
160
160
£
Not to be exceeded more than once per year.
1-3
-------
used in both AQMA's. In McLean-Mercer-Oliver, "208" water
planning activities for the Lewis & Clark planning area have
not yet developed demographic projections. The only other
federally-sponsored planning program, the Northern Great
Plains Resource Program, employed a ten year projection
period because reliable data on energy-related development
were not available beyond that time span. In Fargo, the
current regional land use plan has a projection year 2000,
longer than the normal analysis interval for AQMA planning.
Fargo is not part of a "208" planning area and no other
federally-sponsored planning programs were identified in the
AQMA. Therefore, the ten year period recommended in the
AQMA guideline documents was also used here.
A base year of 1974 was used in validating the AQMA
modeling. This was the most recent year for which complete
data could be obtained. Projection years of 1980 and 1985
were used since they marked the midpoint and end of the ten
year analysis period and fulfilled the requirement in the
guidelines of analyzing for five year intervals. Emission
projections were also developed for the year 1975, but this
year was dropped from the analysis because emissions were
not significantly different than in 1974.
The AQMA analysis focuses on projected annual average
concentrations. Short-term maximum concentrations were cal-
culated, but were not considered to be as important in the
analyses due to the increased uncertainty of predicting
short-term variations in future occurrences and of deter-
mining the joint probability of maximum predicted emission
rates coinciding with adverse meteorological conditions.
The analyses of annual concentrations were considered suffi-
cient for identifying attainment and maintenance problems.
This report is divided into three additional chapters:
Chapter 2 summarizes the findings of the two AQMA analyses;
1-4
-------
Chapter 3 includes a detailed analysis of the McLean-
Mercer-Oliver AQMA for all four pollutants; and Chapter 4
presents the detailed analysis of the Cass County AQMA for
particulates.
1-5
-------
2. SUMMARY
The analyses in both AQMA's utilized current, compre-
hensive emission inventory data and state-of-the-art regional
diffusion modeling to predict ambient pollutant concentra-
tions from emission estimates. In the McLean-Mercer-Oliver
AQMA, projections of future development were exceptionally
good because of several recent studies that identified
specific facilities to be located in the area and their
expected secondary impacts on population, traffic, construc-
tion, etc. In Cass County, almost all the emissions were
shown to be from area sources, and projections were avail-
able for each area source category. The AQMA analyses
followed procedures prescribed in proposed Federal regula-
tions (40 CFR Part 51, Subpart D) and in the series of AQMA
guideline documents prepared by EPA. Therefore, the analy-
ses are considered to be complete and to contain best
available data and procedures.
The analyses showed that annual standards will be
maintained in both AQMA's with existing air pollution con-
trol regulations. The findings are summarized in Table 2.1.
Standards were shown to be maintained by a wide margin in
McLean-Mercer-Oliver, where ambient levels are projected to
be only slightly above background even after anticipated
startup of a coal gasification plant, several power plants,
and new coal mines by 1985. Even though there will be
several large-scale developments, they will be spread over
many miles (much larger area than a city occupies) and do
not appear to increase ambient concentrations substantially
in combination because of the distances between sources.
2-1
-------
Table 2.1. RESULTS OF AQMA ANALYSES
Highest predicted Applicable annual
concentration in standard,
AQMA Pollutant AQMA, ug/m3 ug/m3
McLean-
Particulate
32
60
Mercer-
Oliver
S02
8
80
N°x
3
100
Cass Co.
Particulate
48
60
a Values shown are geometric mean; for other pollutants,
the values are arithmetic mean.
2-2
-------
The sources will have to meet applicable New Source Perform-
ance Standards; tall stacks on the proposed point sources
also act to disperse their ground-level impacts. Finally,
the AQMA has been designated a Class II area for prevention
of significant deterioration, so the allowable increments of
deterioration for particulates and SO2 would be violated
long before the air quality standards are exceeded. A
possible Class I designation for the Fort Berthold Indian
Reservation in western McLean County would result in an even
more restrictive air quality increment for new development
and provide additional assurance that standards would be
maintained.
Analysis for maximum 24-hour levels indicates that
these standards will also be maintained by a wide margin.
Therefore, the need for an air quality maintenance plan to
protect ambient standards in the McLean-Mercer-Oliver AQMA
from being exceeded is not anticipated.
No changes in particulate air quality are expected in
the Cass County AQMA during the next ten years. Presently,
concentrations at the level of the annual secondary standard
are being recorded at one sampling site in Fargo; two others
3
(one in Moorhead) have readings of about 50 ug/m geometric
mean. County-wide emission totals would seem to indicate
high concentrations in rural parts of the county (94 percent
of particulate emissions in the county are from unpaved
roads and agriculture), but modeling shows that concentra-
tions in the rural areas are only slightly above background
and that all high concentrations are in the Fargo-Moorhead
urban area. Point sources, mainly grain milling and storage
elevators, are shown to contribute about half of the concen-
trations (above background) at these receptors. In checking
this apparent contradiction, it was determined that the
highest emission densities do occur in the urban area and
that the rural emissions are very uniformly distributed
throughout the remainder of the county.
2-3
-------
The short-term particulate standards will probably
continue to be exceeded in Cass County as a result of uncon-
trollable natural sources. On infrequent occasions, less
than 10 percent of the sampling days, all sampling sites in
and near the AQMA record concentrations above either the
primary or secondary standard. Since the frequency of
violations is approximately the same at the background
sampling site as at the urban sites, it was concluded that
the contributing sources are probably either agriculture-
related or from long-range transport. In either case, an
attainment plan would be incapable of preventing the
violations.
3
The one-hour maximum standard of 0.08 ppm (160 ug/m )
for photochemical oxidants will probably continue to be
exceeded in the McLean-Mercer-Oliver AQMA solely as a result
of natural sources. However, it was concluded that any
increase in oxidant concentrations attributable to new
hydrocarbon sources in the AQMA should be minimized by means
of a plan to control hydrocarbon emissions from new station-
ary sources.
In summary, the analyses for the North Dakota AQMA's
indicate that the only pollutant/AQMA requiring a mainte-
nance plan is for photochemical oxidants (hydrocarbon con-
trol) in the McLean-Mercer-Oliver AQMA.
2-4
-------
3. Mc LEAN-MERCER-OLIVER AQMA
ANALYSIS AREA
The McLean-Mercer-Oliver AQMA was designated as a
Natural Resource Development Area under the concept that the
high intensity of coal mining and related energy projects
proposed for this area in combination might cause air qual-
ity standards to be exceeded in the future, even though the
demographic projections used to identify regions with poten-
tial for rapid growth do not show a substantial change from
the present rural, agricultural character of this area with-
in the next 10 to 20 years.
The three county AQMA is located in the west central
part of the state, to the northwest of Bismarck-Mandan. It
has a land area of 3827 square miles (70 mi east-west by 60
mi north-south) that is presently about 95 percent farmland.
Almost 27 percent of McLean County, about 550 square miles,
is the Fort Berthold Indian Reservation and not under the
direct jurisdiction of state regulations.
Terrain varies from flat to rolling, with the sharpest
terrain features along river valleys. The area is bisected
by the Missouri River and Garrison Reservoir, with McLean
County to the northeast of the river and Mercer and Oliver
Counties to the southwest. The only bridge across the river
between these counties is at Washburn.
The estimated 1974 population of the AQMA was 20,400.
The three largest towns—Garrison in McLean County and
Beulah and Hazen in Mercer County—have populations of 1,200
to 1,700 each. Population projections used in the AQMA Area
3-1
-------
2
Source Emission Inventory indicate a large percentage
increase in population to 35,500 in 1980 as a result of
manpower necessary to construct and operate the proposed
facilities. However, by 1985 the stable population of the
area is projected to return to about 25,000 with the out-
migration of the construction workers and their families.
If more energy-related facilities are located in this AQMA
in the following decade, population could remain near the
1980 peak.
Much of the land in the three county area contains
commercially recoverable deposits of lignite coal. Three
large power plants in the AQMA are already using this lig-
nite and three more power plants, expansion of existing
plants, and a coal gasification plant are in the design or
construction stages. New mines will be opened to serve
these additional plants.
After the boundaries of the AQMA were established,
plans to build another coal gasification plant in Dunn
County, about 20 miles west of the Mercer-Dunn County line,
were announced. This plant is projected for completion in
1982, and may be followed by additional coal gasification
plants in the same county. Therefore, emission data and
projections have also been assembled for Dunn County as part
of the AQMA analysis so that the air quality impacts of
energy development in that adjacent county can also be
determined.
In order to maintain the degree of resolution necessary
to accurately locate concentration gradients with the atmos-
pheric dispersion model, the entire three county AQMA plus
Dunn County cannot be efficiently modeled as a single unit.
To include an area of 70 mi by 60 mi, receptor sites in the
model would have to be spaced at unacceptably large inter-
vals or computer running time would be prohibitive if a
close receptor spacing were used. Therefore, a smaller
3-2
-------
analysis area which should encompass the points of highest
ambient concentration in the AQMA has been specified. This
analysis area is shown in Figure 3.1. All existing and
projected point sources except a few small grain elevators
in McLean County and the proposed coal gasification plant in
Dunn County are included in this area. Since it is esti-
mated that the Dunn County coal gasification plant will have
the same emissions as the one in the analysis area, it
should not contribute to as high ambient concentrations as
the one in Mercer County because it is not surrounded by
other large point sources.
The analysis a!rea also encompasses the towns with the
3
greatest potential for growth in the AQMA. Because of the
limited access across the Missouri River, the population
influx is predicted to concentrate in Mercer and Oliver
Counties near the new employment centers.
METHODOLOGY
The AQMA analysis for suspended particulate, S02, and
NO is based on the use of regional dispersion modeling to
predict ambient concentrations throughout the analysis area
in the base year and two projection years for comparison
with the air quality standards. The methods used in the
AQMA analysis are consistent with those described in the EPA
guideline series for air quality maintenance planning and
analysis. Volume 12 of the guidelines, Applying Atmospheric
Simulation Models to Air Quality Maintenance Areas,^ was
used to review available models for application in the
McLean-Mercer-Oliver AQMA. The models selected and their
required input data are described below. A separate analy-
sis was performed for photochemical oxidants? it is pre-
sented at the end of this chapter.
A 208 planning agency has been designated for the same
boundaries as the AQMA, but the agency has not yet developed
3-3
-------
GJ
I
FOIT KRTHOLD
INDIAN
RTHOID j
irvt,_r~~v\i
\ ^Boiin
\ ANG
VAUfV
MUt AM
MDU existing
N> MDU Beulc H
itt«« '
VNfttlWOOO I
United Pwi
I
¦fiasco
ouvO co
1
•M \
^ |
?
CS^».
m( i I an CO
Miiic«a
Figure 3.1. McLean-Mercer-Oliver AQMA analysis area.
-------
projections for future development of the area. After the
208 projection data are released, they will be compared with
those used in the present analysis for consistency, and any
necessary changes in the analysis will be made at that time.
The only deviation from standard methodology in per-
forming the AQMA analysis was in projecting emissions for
point sources in future years. No growth factors were used.
Instead, existing sources were assumed to continue emitting
at their current rates (with the exception of one source
scheduled for shutdown in 1977) and all new sources and
expansion of existing sources were identified specifically,
with emission estimates being made from available informa-
tion. Data on the size, location, and startup date for
proposed sources obtained from the Northern Great Plains
5 . .
Resource Program, the State Division of Planning's Coal
Impact Study,6 and directly from the developers were all in
agreement. Although this projection method neglects any
unannounced new point sources that would locate in the AQMA
during the next 10 years, it is probably far more accurate
than analyses using growth factors to estimate future emis-
sions and land use planning data to allocate these increased
emissions to locations within the AQMA.
Modeling
Two multiple-source atmospheric dispersion models are
available with capabilities of predicting annual average
pollutant concentrations throughout a study area: Air
7
Quality Display Model (AQDM) and Climatological Dispersion
8
Model (CDM). Both of these models employ Gaussian plume
dispersion and frequency distributions of annual meteorolog-
ical conditions. The version of CDM on the Department of
Health's computer has the capability for calculating source
contributions at specified receptor sites (not in the ver-
sion distributed in UNIMAP). Since CDM is a second genera-
tion multiple-source model and includes some improvements
3-5
-------
over AQDM such as consideration of differences in daytime
and nighttime dispersion characteristics, CDM was selected
for use.
The CDM is applicable for flat or rolling terrain. It
predicts annual arithmetic mean concentrations at any number
of receptor sites for two pollutants--particulate and SO,, in
this case. The model adds a constant background concentra-
tion to values predicted at each receptor site by the
dispersion calculations. The predicted concentrations can
also be calibrated by comparison with concentrations mea-
sured at sites within the area being modeled. The slope of
the line of best fit between measured and predicted concen-
trations is used to correct for persistent over- or under-
prediction and the correlation coefficient provides an
estimate of the accuracy of the simulation model.
The following input data are required for CDM:
Description
Source locations
Source emission rates
Stack parameters
Meteorology
Receptor locations
Background concen-
tration
Format
UTM coordinates
Tons per year
Stack height and diameter,
exit temperature and flow rate
Joint frequency distribution
of 16 wind directions, 6 wind
speeds, and 6 stability classes
UTM coordinates
Annual arithmetic mean
Potential 24-hour maximum concentrations in the AQMA
were simulated with another model, the PEDCo Dispersion
Model (PDM). The model is analogous to CDM in many aspects,
but utilizes hourly meteorological and emissions input data
and predicts short-term concentrations for averaging times
from one to 24 hours. Meteorological input data include
wind speed, wind direction, stability class, mixing height,
and temperature.
3-6
-------
The following sections discuss the input data used in the
McLean-Mercer-Oliver analysis.
Source Locations
The locations of major existing and proposed point
sources in the AQMA are shown in Figure 3.1. Universal
Transverse Mercator (UTM) coordinates for the existing
sources were obtained from NEDS listings and checked by
locating the sites on USGS quadrangle maps. The UTM coor-
dinates for new sources were determined by locating proposed
sites on USGS maps based on information provided by the
developers. As indicated previously, the identification of
new sources was accomplished by comparing information on
energy development projects from several different studies
and information sources,^^'6'® all of which were in sub-
stantial agreement. The UTM coordinates for all point
sources in the analysis area are listed in Table 3.1.
The spatial distribution of area sources was specified
by means of a grid system. Area source emissions were
apportioned into 10 km square grids covering the entire
analysis area. The locations of the 30 area source grids
and their assigned numbers are shown in Figure 3.2.
Source Emission Rates and Stack Parameters
Pollutant emissions were divided into two categories
for purposes of modeling—point sources and area sources.
Point source emission data for the base year 1974 and stack
parameters were obtained from file data in NEDS format, as
summarized in Appendix A. Emission estimates for new power
plant point sources were based on announced megawatt capaci-
ties of the plants, assumed load factors of 85 percent, and
emission rates equal to the Federal New Source Performance
3-7
-------
Table 3.1. LOCATIONS OF POINT SOURCES
UTM coordinates, km
Source name X Y
1974
1.
Farmers Elevator &
Mercantile
339.0
5258.0
2.
Underwood Elevator
339.0
5258.0
3.
Basin I
324.3
5238.9
4.
MDU Boiler 1
290.0
5238.0
5.
MDU Boiler 2
290.0
5238.0
6.
MDU Boiler 3
290.0
5238.0
7.
MDU Boiler 4
290.0
5238.0
8.
MDU Boiler 5
290.0
5238.0
9.
United Power, Stanton
323.7
5239.5
10.
Beulah Farmers Elevator
290.0
5237.7
11.
Beulah PV Elevator
290.0
5237.7
12.
Hazen Farmers Elevator
302.0
5241.0
13.
Minnkota Power
332. 0
5213.7
New,
by 1980
14.
Basin II
324.3
5238.9
15.
Minnkota Unit 2
332.0
5213.7
16.
United Power, Underwood
338.0
5250.0
New,
by 1985
17.
MDU Beulah 1
287.0
5233.0
18.
MDU Beulah 2
287.0
5233.0
19.
Basin IIIa
324.3
5239.0
20.
Basin Beulah
286.1
5250.2
21.
ANG Coal Gasification
286.0
5250.2
22.
ANG coal dust
286.3
5250.3
a Location still tentative.
3-8
-------
FOOT BHTHOID
l/V/I
-------
Standards (NSPS) for fossil-fuel fired steam generators:
C.
Particulate = 0.1 lb/10 BTU input
S02 = 1.2 lb/106 BTU input
NOxa =0.6 lb/106 BTU input
This method may overestimate actual emissions from the new
sources because many will have emission rates less than the
NSPS. The estimates of operating times using the value of
85 percent of capacity (and the additional assumptions of
6500 BTU/lb for the lignite and 10,000 BTU/kwh conversion
efficiency) appear to agree quite well with estimates from
the Coal Impact Study report6 of coal consumption at the
proposed plants, as shown in Table 3.2. The following stack
parameters were used for new power plant units for which
specific information could not be obtained:
Stack height = 550 ft
Stack diameter = 20 ft
Exit velocity = 60 ft/sec
Temperature = 380° F
For the ANG coal gasification plant, emission estimates
and stack data were obtained from the environmental impact
g
report. These emission estimates were compared with inde-
pendent calculations (see Appendix B) based on data in the
literature and appeared to be reasonable. The same emission
data were assumed for the proposed Natural Gas Pipeline
plant in Dunn County, although this source was not included
in the model because it was considerably outside the anal-
ysis area.
The performance standard for nitrogen oxides from lignite
fired plants is only a proposed standard at the present
time.
3-10
-------
Table 3.2. COMPARISON OF ESTIMATED ANNUAL LIGNITE
CONSUMPTION FROM COAL IMPACT STUDY AND
FROM MEGAWATT CAPACITY
Plant
MW rated
capacity
Lignite usage,
ton/yra
Lignite consumption
from Coal Impact Rpt.
ton/yr
MDU Beulah
440
2,520,000
2,500,000
Basin I Stanton
212
1,214,000
1,200,000
Basin II Stanton
440
2,520,000
2,500,000
Basin Beulah
440
2,520,000
2,500,000
Minnkota
234
1,340,000
1,500,000
Minnkota II
440
2,520,000
2,800,000
United Stanton
172
985,000
850,000
United Underwood
950
5,442,000
5,500,000
a Based on BTU value of 6500 BTU/lb for lignite, conversion
efficiency of 10,000 BTU/kwh, and 85 percent load
factor.
3-11
-------
Point source emission data used in the CDM model for
the base year and projection years are summarized in Table
3.3.
Area source emissions and projections by source cate-
gory on a county-wide basis were obtained from the North
2
Dakota AQMA Area Source Emission Inventory. Tables 3.4 and
3.5 show summaries of these emissions. A detailed descrip-
tion of the procedures used to estimate and project the area
source emissions can be found in the inventory report. The
three area source categories with most particulate emis-
sions—unpaved roads, agriculture, and mining—are all
fugitive dust sources. Unpaved road emissions were esti-
mated from data on vehicle miles of travel on county roads
provided from EPA's Compilation of Air Pollutant Emission
Factors, Supplement 5. Agricultural emissions were based on
the number of acres planted in different crops and emission
rates derived from U.S. Department of Agriculture's wind
erosion equation. Fugitive dust emissions from lignite
surface mining were estimated by surveying the operations at
one mine in Mercer County and calculating emissions by
operation. These estimates were expressed in terms of
lb/ton of coal mined for use in determining emissions from
other mines in the study area.
In order to allocate the county-wide area source emis-
sions into the 30 grids in the analysis area, the methodol-
ogies described in Volume 13 of the AQMA guidelines,
Allocating Projected Emissions to Sub-County Areas,^ were
employed to develop apportioning factors for each source
category. The parameters used for apportioning the area
source emissions are summarized in Table 3.6; the resulting
area source emissions by grid for the three analysis years
are presented in Tables 3.7, 3.8, and 3.9.
For the projection years 1980 and 1985, it was assumed
that the percentage of total traffic in a county within each
3-12
-------
Table 3.3. POINT SOURCE EMISSIONS IN ANALYSIS AREA
Estimated emissions,
ton/yr
Source name
Partic
so2
N°x
1974
1.
Farmers Elevator &
10
0
0
Mercantile
2.
Underwood Elevator
4
0
0
3.
Basin I
46
10,499
3,962
4.
MDU Boiler la
2,05
173
54
5.
MDU Boiler 2a
205
173
54
6.
MDU Boiler 3a
205
173
54
7.
MDU Boiler 4a
218
346
108
8.
MDU Boiler 5a
218
346
108
9.
United Power, Stanton
3,822
8,448
2,682
10.
Beulah Farmers Elevator
4
0
0
11.
Beulah PV Elevator
10
0
0
12.
Hazen Farmers Elevator
17
0
0
13.
Minnkota Power
5,740
17,863
5,128
New,
by 1980
14.
Basin II
1,638
19,659
11,466
15.
Minnkota Unit 2
1,638
19,659
11,466
16.
United Power, Underwood
3,537
42,446
24,759
New,
by 1985
17.
MDU Beulah 1
1,638
19,659
11,466
18.
MDU Beulah 2
1,638
19,659
11,466
19.
Basin III
1,713
20,553
11,991
20.
Basin Beulah
3,276
39,318
22,932
21.
ANG Coal Gasification*3
1,117
12,426
8,191
22.
ANG coal dust
219
0
0
a Electrostatic precipitator installed in 1975.
b Hydrocarbon emissions = 1093 ton/yr.
3-13
-------
Table 3.4. AREA SOURCE PARTICULATE EMISSIONS
FOR McLEAN-MERCER-OLIVER COUNTIES
Particulate emissions, ton/yr
McLean County Mercer County Oliver County Dunn County
Source category 1974 1975 1980 1985 1974 1975 1980 1985 1974 1975 1980 1985 1974 1975 1980 1985
Fuel combustion:
1. Lignite coal
63
62
56
51
130
127
115
104
53
52
47
42
64
63
57
51
2. Distillate oil
9
10
16
11
5
5
9
6
2
2
3
2
3
3
5
4
3. Residual oil
7
7
8
9
4
4
5
5
1
1
1
1
3
3
3
4
4. Natural gas
3
3
5
4
0
0
0
0
0
0
0
0
0
0
0
0
5. LPG
1
1
2
1
2
2
3
2
neg
neg
neg
neg
1
1
2
1
Burning:
7. Open burning
5
5
5
5
neg
neg
neg
neg
5
5
5
5
11
11
11
11
Mobile sources:
10. Highway vehicles
72
72
102
65
27
27
38
24
14
14
20
13
27
27
38
24
11. Off-highway
67
67
67
67
33
33
33
33
23
23
23
23
34
34
34
34
13. Railroads
4
5
7
10
1
1
2
2
1
1
2
2
1
1
2
2
Fugitive dust:
16. Unpaved roads
39,610
39,610
39,610
39,610
16,180
16,180
16,180
16,180
10,240
10,240
10,240
10,240
19,770
19
,770
19,770
19,770
17. Agriculture
4,679
4,679
4,679
4,679
3,439
3,439
3,439
3,439
2,551
2,551
2,551
2,551
4,221
4
,221
4,221
4,221
18. Construction
39
171
0
0
96
0
835
96
96
160
0
0
0
0
643
0
19. Mining
0
0
0
1,725
1,587
1,587
4,904
5,635
437
437
1,256
1,256
0
0
0
2,300
20. Paved roads
267
284
597
406
96
102
220
149
45
49
110
73
86
92
180
139
Total
44,826
44,976
45,154
46,643
21,600
21,507
25,783
25,675
13,468
13,535
14,258
14,208
24,221
24
,226
24,966
26,561
-------
Table 3.5. AREA SOURCE EMISSIONS OF SO,,
HYDROCARBONS, AND OXIDES OF NITROGEN
FOR AQMA COUNTIES
Pollutant emissions, ton/yr
SO, HC NO
2 X
County/Source category 1974 1975 1980 1985 1974 1975 1980 1985 1974 1975 1980 1985
McLean County:
Lignite coal
39
38
34
31
3
3
3
2
16
16
14
13
Distillate oil
6
6
10
7
2
2
3
2
17
18
30
21
Residual oil
25
25
29
32
1
1
1
1
19
19
22
25
Natural gas
neg
neg
neg
neg
2
2
3
2
25
27
44
31
LPG
1
X
2
1
1
1
2
1
8
9
14
10
Open burning
neg
neg
neg
neg
9
9
9
9
2
2
2
2
Highway vehicles
32
33
59
43
1095
979
1312
563
736
770
1098
737
Off-highway vehicles
44
44
44
44
308
308
308
308
564
564
564
564
Railroads
10
12
18
24
17
20
31
41
66
78
119
160
Evaporative
0
0
0
0
96
101
162
105
0
0
0
0
Total
157
159
196
182
1534
1426
1834
1034
1453
1503
1907
1563
Mercer County:
Lignite coal
81
79
71
65
5
5
5
4
32
31
28
25
Distillate oil
3
3
5
4
1
1
2
1
12
13
21
15
Residual oil
14
14
16
18
1
1
1
1
11
11
13
14
LPG
1
1
2
1
1
1
2
1
11
12
19
14
Open burning
neg
neg
neg
neg
1
1
1
1
neg
neg
neg
neg
Highway vehicles
12
13
22
16
413
369
495
212
277
290
413
277
Off-highway vehicles
22
22
22
22
153
153
153
153
281
281
281
281
Railroads
2
.2
4
5
3
4
5
7
14
17
25
34
Evaporative
0
0
0
• 0
41
43
69
46
0
0
0
0
Total
135
134
142
131
619
578
733
426
638
655
800
660
Oliver County:
Lignite coal
33
32
29
26
2
2
2
2
13
13
12
11
Distillate oil
1
1
2
1
neg
neg
neg
neg
8
9
14
10
Residual oil
5
5
6
. 6
neg
neg
neg
neg
3
3
3
4
LPG
neg
neg
neg
neg
neg
neg
neg
neg
3
3
5
4
Open burning
neg
neg
neg
neg
10
10
10
10
2
2
2
2
Highway vehicles
6
6
11
8
218
195
261
112
147
154
219
147
Off-highway vehicles
15
15
15
15
108
108
108
108
196
196
196
196
Railroads
2
2
4
5
3
4
5
7
11
13
20
27
Evaporative
0
0
0
0
19
20
33
22
0
0
0
0
Total
62
61
67
61
360
339
419
261
383
393
471
401
Dunn County:
¦
Lignite coal
40
39
35
32
3
3
3
2
16
16
14
13
Distillate oil
2
2
3
2
1
1
1
1
3
3
5
4
Residual oil
9
9
10
12
neg
neg
neg
neg
7
7
8
9
LPG
1
1
2
1
1
1
2
1
9
10
16
11
Open burning
1
1
1
1
20
20
20
20
4
4
4
4
Highway vehicles
12
13
22
16
417
373
500
214
216
226
322
216
Off-highway vehicles
23
23
23
23
160
160
160
160
290
290
290
290
Railroads
1
1
2
2
2
2
4
5
9
11
16
22
Evaporative
0
0
0
0
36
38
57
40
0
0
0
0
Total
89
89
98
89
640
598
647
443
554
567
675
569
3-15
-------
Table 3.6. APPORTIONING FACTORS BY SOURCE CATEGORY
Source
category
Apportioning
factor
Source of
apportioning data
Fuel combustion
Open burning
Highway vehicles
Off-highway
Railroad
Unpaved roads
Agriculture
Construction
Mining
Paved roads
dwelling units
dwelling units
miles of road,
both paved and
unpaved
acres of cropland
miles of track
miles of unpaved
roads
acres of cropland
power plant sites
location of major
mining operations
and acres mined
miles of paved
roads
county highway maps
county highway maps
county highway maps
county ASCS offices
county highway maps
county highway maps
county ASCS offices
State Planning Division
Coal Impact Project
Report, and mining
companies
county highway maps
3-16
-------
Table 3.7. PARTICULATE EMISSIONS BY GRID AND
SOURCE CATEGORY, 1974
Source category, ton/yr
1 0)
>i
<0
•a
iH
3
CO
G
flj 1—1
0
to
O
¦P
G
•O
i—i
rH 5
G G
A -H
i &
rH
•H
CO
-H
a)
(0
0) £2
CD M
tr>
•H
ft «o
u
G
G
>
•p
P 0
ft P
-H a)
m -h
(0
G O
O
•H
as
o
Grid
0
O Si
EG >
O £3
K
D U
c
U
S
04
Eh
1
2
1
2
111
159
941
2
2
2
2
1066
257
neg
1329
3
3
1
3
904
317
1228
4
2
neg
1
2
686
176
867
5
neg
neg
1
1
764
91
neg
857
6
6
neg
2
1
neg
838
90
171
8
1116
7
2
1
2
neg
494
159
9
667
8
2
1
1
neg
635
151
10
800
9
33
1
3
neg
565
317
10
929
10
14
1
3
neg
649
317
13
997
11
1
neg
1
1
neg
623
53
679
12
1
neg
2
1
925
86
1015
13
34
1
1
neg
445
119
397
6
1003
14
5
neg
1
2
neg
562
237
99
3
909
15
3
neg
1
1
neg
749
139
893
16
2
neg
1
1
675
72
5
756
17
2
neg
1
1
neg
281
59
96
1090
12
1542
18
1
neg
1
1
neg
436
86
525
19
2
1
1
529
89
6
628
20
3
neg
1
1
707
153
865
21
3
neg
1
1
661
153
819
22
3
neg
1
1
621
153
5
784
23
15
1
1
1
519
131
5
673
24
1
neg
1
1
501
114
618
25
2
1
2
614
127
6
752
26
2
neg
1
1
729
153
886
27
2
neg
1
1
501
153
658
28
3
neg
1
1
527
153
6
691
29
6
1
1
1
593
134
437
5
1178
30
2
neg
1
1
427
111
96
8
646
3-17
-------
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
i—(
nJ
¦P
O
E-<
942
1330
1228
867
859
956
3621
814
941
1015
679
1015
1541
1203
892
764
1847
525
635
865
818
792
685
618
759
886
658
701
2003
561
Table 3.8. PARTICULATE EMISSIONS BY GRID AND
SOURCE CATEGORY, 1980
Source category, ton/yr
0)
o
CO
c
M
•H
o
Ul
3
•P
<0
•H
m
t)
-P
O
o
-p
CP
>i a)
(0
tj
rH
3
Cr>
u
Ul
G
nj i—t
a
0
to
o
-P
C
h n
C c
A -H
1 X5
1—1
•H
(0
a)
3 0
a 3
•h a>
H-l -H
nJ
c 0
Cn
o
•H
(0
fa o
o ,a
« >
O A
u
<
u
a
&
2
2
2
111
159
2
2
2
1066
257
1
2
2
3
904
317
2
neg
1
2
686
176
neg
neg
2
1
764
91
1
7
neg
2
1
neg
838
90
18
2
1
2
neg
494
159
643
2300
20
2
2
1
neg
635
151
23
31
2
3
neg
565
317
23
13
2
3
neg
649
317
31
1
neg
1
1
neg
623
53
1
neg
2
1
925
86
32
1
1
neg
445
119
96
834
13
5
neg
1
2
neg
562
237
390
6
2
neg
1
1
neg
749
139
2
neg
1
1
675
72
13
2
neg
1
1
neg
281
59
96
1380
27
1
neg
1
1
neg
436
86
2
1
1
529
89
13
3
neg
1
1
707
153
2
neg
1
1
661
153
3
neg
1
1
621
153
13
14
1
1
1
519
131
18
1
neg
1
1
501
114
2
2
1
614
127
13
2
neg
1
1
729
153
2
neg
1
1
501
153
3
neg
1
1
527
153
16
5
1
1
1
593
134
1256
12
2
neg
1
1
427
111
19
3-18
-------
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
i—i
-p
o
EH
941
1329
1227
867
858
2673
2970
805
929
1003
678
1014
1971
1492
892
759
1742
525
631
864
818
787
673
617
754
885
658
695
1999
554
Table 3.9. PARTICULATE EMISSIONS BY GRID AND
SOURCE CATEGORY, 1985
Source category, ton/yr
a
a)
0
ui
a
•H
¦O
0
(0
3
¦P
m
•H
m
4J
O
0
4-1
>i a>
>i
rH
3
(0
G
(0 rH
<0
0
a)
3
3
•H
> <0
O
+J
a
T3
H 5
c c
1 JC,
rH
•H
(0
•H
a)
o) ii
3 O
a. 3
•H a)
IW -H
Ifl
a o
0
•H
Pm o
O A
a >
O 43
K
D U
<
u
23
04
2
1
2
111
159
2
1
2
1066
257
1
2
1
3
904
317
2
neg
1
2
686
176
neg
neg
1
1
764
91
1
6
neg
1
1
neg
838
90
1725
12
1
1
1
neg
494
159
2300
14
2
1
1
neg
635
151
15
28
1
3
neg
565
317
15
12
1
3
neg
649
317
21
neg
neg
1
1
neg
623
53
1
neg
1
1
925
86
28
1
1
neg
445
119
96
1272
9
4
neg
neg
2
neg
562
237
683
4
2
neg
1
1
neg
749
139
2
neg
1
1
675
72
8
2
neg
1
1
neg
281
59
1380
18
1
neg
1
1
neg
436
86
2.
1
1
529
89
9
2
neg
1
1
707
153
2
neg
1
1
661
153
2
neg
1
621
153
9
12
1
1
1
519
131
8
1
neg
neg
1
501
114
2
1
1
614
127
9
1
neg
1
1
729
153
2
neg
1
1
501
153
3
neg
1
1
527
153
10
5
1
1
1
593
134
1256
8
2
neg
1
1
427
111
12
3-19
-------
grid would remain the same as in 1974, or that the addi-
tional traffic due to growth would not be concentrated in a
few grids. Data on new and upgraded roads were obtained
from the State Highway Department. Projections of new
housing units by grid (actually by town) were obtained from
the same reference as the population projections used in the
area source inventory, Social Impacts of Six Coal Conversion
3
Plants, prepared by the State Division of Planning. Loca-
tions of new coal mining sites were determined from the Coal
Impact Study and contacts with the mining companies.
Meteorology
The primary meteorological input data for CDM—the
joint frequency distribution of six wind speed classes, 16
wind directions, and six stability classes in the format of
the day-night STAR program—was available from the Bismarck
National Weather Service Station for the five year period
1967-1971 (eight observations/day). The Bismarck weather
station is about 25 miles south of the edge of the analysis
area. The environmental impact study for the ANG coal gasi-
fication plant included short-term meteorological data
collection from a site just north of Beulah; analysis of
g
this data indicated that wind speeds and directions at
Beulah and Bismarck were almost the same except for slightly
higher average wind speeds recorded at Beulah. No differ-
ences would be anticipated within these distances in an area
of the Great Plains lacking sharp terrain features.
Additional wind speed and direction data were collected
in or near the analysis area as part of the Northern Great
Plains Resource Program. Two of the NGP sites were located
at Stanton in Mercer County and McClusky in Sheridan County.
At the time that the AQMA analysis began, only a few months
of data were available from these sites for comparison with
3-20
-------
the Bismarck readings. Linear regression analysis of three-
hour average wind speed and direction readings for each of
the NGP sites and the Bismarck station showed only fair
correlations:
Correlation coefficient,
HGP site comparison with Bismarck data
Wind speed Wind direction
Stanton 0.67 0.76
McClusky 0.52 0.67
However, many problems were experienced initially with the
meteorological equipment at these NGP sites, and the correla-
tion is attributed to instrumentation problems rather than
to different wind patterns at these nearby locations.
After reviewing all available wind speed and direction
data, it was determined that the data from Bismarck in the
day-night STAR format would be representative of the analy-
sis area.
The other meteorological data required for input to
CDM—mean annual nocturnal and afternoon mixing heights—
were also obtained from the Bismarck weather station, where
the only routine upper air measurements in the state are
made. The mean nocturnal mixing height is 359 m; the mean
afternoon mixing height is 1448 m.
For the short-term model, conditions simulating the
most stable atmospheric conditions and a low mixing height
(low level inversion) were specified. The other two atmos-
pheric conditions often associated with short-term peak con-
centrations—plume looping (unstable atmosphere) and plume
trapping aloft followed by breakup of stable conditions—
are not appropirate for time periods of 24 hours. There-
fore, the following meteorological input data were used:
3-21
-------
Wind direction = 240°
Wind speed = 2.0 m/sec daytime
1.0 m/sec nighttime
Stability = D for daytime
E for nighttime
Mixing height = 800 m maximum
500 m minimum
Temperatures = diurnal variation in
40's and 50's
Receptor Locations
Receptor locations for prediction of ground-level
concentrations in the CDM model can be specified at any
points in the analysis area. In order to assure complete
coverage of the area, a rectangular array of receptors at
the corners and center of every area source grid was speci-
fied, as shown in Figure 3.3. In addition, four other
receptors were located at the UTM coordinates of the four
sampling sites in the area so that predicted and measured
concentrations could be compared. This design resulted in a
total of 76 receptors spaced throughout the 50 by 60 km
area.
Background Air Quality
Background concentration is that portion of measured
ambient pollutant levels that cannot be reduced by con-
trolling emissions from local man-made sources. For SC>2 and
NO , background is usually considered to be negligible. For
particulate, the non-reducible fraction may originate from
long-range transport, natural sources such as windblown dust
and biological debris, and atmospheric aerosol formation.
The method specified in the Federal regulations (Section
51.13c) for estimating particulate background is to use the
concentration measured at a nonurban site in or near the
3-22
-------
FOUT BERTHOtD
U>
I
to
W
INOIAV^Vi m 7
UIIN
¦dCiLC|
ouvft co
OiiVIt CO
IEGCNO
0 I 2 3 4 s imxs
Figure 3.3. Receptor locations.
-------
study area that is unaffected by nearby emission sources.
All four of the high volume sampler sites in the AQMA would
be characterized as nonurban. The one with the lowest
3
annual geometric mean, the Beulah WET site at 20 ug/m in
1974, also probably has the lowest surrounding source impact
of the four sites. The site has the lowest concentration
3
recorded in the state, and the values of 24 ug/m arithmetic
3
mean and 20 ug/m geometric mean compare well with back-
ground concentrations determined for other areas in the
northern Great Plains states. Therefore, the readings from
this site were used for particulate background concentra-
tions in the modeling.
BASE YEAR AIR QUALITY AND MODEL VERIFICATION
Even though the AQMA is sparsely populated, there is a
relatively good network of sampling stations to provide
measurements of baseline air quality: four particulate and
S0o sites and two NO sites. The stations in the NGP study
were installed during 1974, so annual averages for that
calendar year are not available. However, a continuous year
of data was generated (September 1974 through August 1975)
which has been considered as base year in this analysis.
All the sampling data for the three pollutants show concen-
trations to be just above background in the base year, and
none of the measurements approach either the annual or
short-term averages for primary or secondary air quality
standards.
The CDM model was first run for the base year 1974 so
that the modeling results could be verified by comparison
with measured concentrations. The model predicts annual
arithmetic concentrations at all specified receptor sites.
The pollutant NO was not analyzed directly with the model,
which only handles two pollutants per run. Instead, NOx
3-24
-------
concentrations were assumed to be proportional to predicted
S02 concentrations at receptor sites since the point sources
are the same and the two pollutants are emitted in a con-
stant ratio by the point sources. Model-predicted values
are compared with measured values in Table 3.10.
For particulate, the only pollutant for which quanti-
tative data are sufficient to calibrate the model/ linear
regression between predicted and measured values yields a
slope of 0.088 and a y-intercept of 30.0 for the line of
best fit, and a correlation of 0.40. Ideally, the slope and
correlation should approach 1.0 and the y-intercept should
be near zero when background has been included in the pre-
dicted concentrations. However, a regression analysis for
such closely spaced points is subject to large errors and is
not really indicative of the close agreement between pre-
dicted and measured concentrations. Review of the data
pairs shows that the model was within 8 ug/m of the mea-
sured value in all four cases and that the ratio of pre-
dicted to measured values was 1.02.
There are some limitations on the accuracy of any model
that would be inherent in the input data for this analysis
area:
0 The concentrations due to source emissions are
much less than background, which was assumed
to be constant throughout the analysis area.
Small variations in actual background in dif-
ferent parts of the AQMA would have a strong
influence on the total measured concentrations
and on the resulting correlation between pre-
dicted and measured concentration.
0 The majority of particulate emissions are from
fugitive dust sources which, by their nature,
are difficult to quantify or to locate
accurately.
3-25
-------
Table 3.10. COMPARISON OF MODEL-PREDICTED
AND MEASURED POLLUTANT CONCENTRATIONS IN AQMA ANALYSIS AREA
(ANNUAL AVERAGE)
Predicted ann.
Measured ann.
arith. mean,
arith. mean,
Pollutant
Receptor site
ug/m^
ug/m3
Partic
Beulah WET
32
24
Beulah WP
35
35
Stanton
33
39
Washburn
32
35
SO-
Beulah WET
1
neg
Beulah WP
4
neg
Stanton
4
bubbler = 14
contin. = neg
Washburn
2
bubbler = 15
contin. = neg
NO
V
Stanton
1.3
bubbler = 3.5
A
contin. = neg
Washburn
0.7
bubbler = 3.4
3-26
-------
In view of the reasonable values obtained with the
model even with the limitations cited above, the CDM model
has been applied with a calibration factor of 1.0 to predict
future particulate concentrations in the analysis area. For
S00 and NO , the measured and predicted values are both so
low that no assessment of the predictive capabilities of the
model can be made. Therefore, the model will also be
applied with no calibration correction for these two
pollutants.
Prior to plotting the predicted particulate concentra-
tions on a map of the analysis area, the values were con-
verted from annual arithmetic to annual geometric means so
that they could be compared directly with the applicable
primary and secondary standards of 75 and 60 ug/m"^, respec-
tively. This was accomplished by use of the following
equation relating the two measures:
2
arith. mean _ el/2 In SGD
geom. mean ~ (eq.l)
where SGD = standard geometric deviation
Standard geometric deviations of the four sampling data
sets varied from 1.65 to 2.55. Since background is such a
large portion of the predicted concentrations at all the
receptors, the SGD of 1.86 from the site used as background
was assumed to be representative for all locations. This
represents a conservative assumption, as a higher SGD would
produce lower geometric mean values. By solving equation 1
above for an SGD of 1.86, it can be determined that the
geometric mean is 0.825 of the predicted arithmetic mean at
a receptor.
The predicted particulate concentrations are shown in
Figure 3.4. Concentrations are fairly uniform throughout
the analysis area and, as indicated by the sampling data,
3-27
-------
FORT SEKTHCHD
INDIAN
25
25
30
U)
I
to
00
•25 ug/m'
LEGEND
Figure 3.4. 1974 geometric mean particulate concentrations.
-------
not much above background. No isopleth maps were prepared
for S0„ and NO concentrations because predicted concentra-
*L X
tions were not substantially above zero for either pollutant
in any part of the area. The highest predicted S09 level
3 3
was four ug/m compared to the standard 80 ug/m ; the high-
3
est NO level was slightly above one ug/m compared to the
x 3
standard of 100 ug/m .
The short-term PDM model also predicted peak concen-
trations well above the standards throughout the analysis
area. Model-predicted values are compared with measured
maximum concentrations at sampling sites in Table 3.11. For
particulates, the average ratio between predicted and mea-
sured 24-hour maximum values was 0.58; the ratio between
predicted and second high measured values was 0.73. The
highest predicted 24-hour concentration (uncorrected) at
3
any receptor in the analysis area was 99 ug/m , including an
3
estimated background of 24 ug/m . Thus, even considering
the tendency of the model to underpredict, it does not
appear that short-term particulate concentrations above the
secondary standard are occurring. All predicted 24-hour S00
o 2
and NO concentrations were less than 50 and 20 ug/m ,
respectively.
PROJECTED AIR QUALITY
The models were reapplied to the area using projected
emissions for 1980 and 1985, with other input data remaining
the same. Thus, the model calculated expected arithmetic
mean and maximum concentrations at receptor sites for 1980
and 1985. Particulate concentrations were converted to
geometric mean values using the procedure described in the
preceding section.
Figures 3.5 and 3.6 show isopleths of the predicted
1980 and 1985 particulate concentrations in the analysis
3-29
-------
Table 3.11. COMPARISON OF MODEL-PREDICTED
AND MEASURED POLLUTANT CONCENTRATIONS IN AQMA ANALYSIS AREA
(SHORT-TERM PEAK)
Predicted max Measured max Second high
Pollutant
Receptor site
24-hour,
ug/m^
24-hour,
ug/m3
24-hour
ug/m3
Beulah WET
45a
47
39
Beulah WP
55
161
105
Stanton
81
185
158
Washburn
75
128
101
Beulah WET
5
n.d.
n.d.
Beulah WP
12
n.d.
n.d.
Stanton
26
88
66
Washburn
14
72
61
Stanton
9
16
12
Washburn
5
34
13
Partic
SO,
NO
x
Predicted particulate concentrations include a background
of 24 ug/m3.
n.d. = not detectable.
3-30
-------
Ijj
I
u>
I—*
FOKT BERTHOLD
JNDIAN
tE iEtVATION
30 ug/m^
§m§ jr ^"N
Ouvft co
CtMftl
*
OovCt CO J
¦Ml On CO
25 og/m^
•C il*« co
Mi(<6* CE
Figure 3.5.
Predicted 1980 geometric mean particulate concentrations.
-------
FORT BERTHOLD
MCVCCCCO
OCJVfl CO
L SPEND
-------
area. These concentrations in both years are slightly
higher than for the base year, but still do not approach the
levels of the secondary standard. The highest expected
annual average concentrations in 1980 and 1985 were 32
ug/m3.
The contributions of different source categories to
ambient concentrations, as indicated by the CDM source con-
tribution tables, are shown in Table 3.12. Obviously,
undefined background sources 'are the major contributors at
all receptors. Most of the remaining source impact is from
area sources which contribute an even higher percentage than
their portion of the emissions in the area because they are
emitted at ground level and are uniformly distributed (near
all receptors). By reviewing the area source emission data
in Tables 3.7 through 3.9, it becomes apparent that most of
the area source contribution is from unpaved roads and
agriculture, but that most of the increase in particulate
concentrations between the base year and the projection
years is due to emissions from mining operations (unpaved
road and agriculture emissions are estimated to remain
constant in the projection years).
For S02 and NOx, ground-level concentrations in the
projection years do not increase nearly in proportion to the
increase in emissions shown in Table 3.3. Maximum concen-
trations of S02 in the area are projected at 7 ug/m in
1980 and 8 ug/m3 in 1985, still only 10 percent of the
annual standard. The 1985 distribution of ambient S02
concentrations throughout the analysis area is shown in
Figure 3.7. Levels of NO are projected to increase to 2.3
o x3
ug/m in 1980 and 2.7 ug/m in 1985. Virtually all of the
ambient S02 and N0X concentrations are due to the major
point sources.
Projected maximum concentrations follow trends similar
to the annual averages. The maximum 24-hour particulate
3-33
-------
Table 3.12. SOURCE CONTRIBUTIONS TO PREDICTED
19 85 AMBIENT CONCENTRATIONS AT SELECTED RECEPTORS
Percent of total ambient concentration
contributed by each source
Source Near Beulah Near Stanton Near Center
Background
61
69
67
Point sources
3
2
5
Unpaved roads
11
15
15
Agriculture
3
3
4
Mining
21
10
8
All other area
sources
1
1
1
Total
100
100
100
3-34
-------
Figure 3.7. Predicted 1985 annual arithmetic mean S02 concentrations.
-------
concentrations in the analysis area (uncorrected) are pre-
3 3
dieted to increase from 99 ug/m to 103 ug/m in 1980 and
123 ug/m3 in 1985. If these values are increased to com-
pensate for underprediction by the model, second highest 24-
3
hour concentrations would be approximately 141 ug/m in
3
1980 and 168 ug/m in 1985. Projected maximum 24-hour
3
concentrations of SO~ are still less than 100 ug/m , and
3
projected maximum NO concentrations are less than 30 ug/m .
From the modeling results, it can be concluded that
proposed developments during the next 10 years will not
cause air quality standards for any of the three pollutants
to be exceeded. Therefore, no maintenance plan should be
required. Also, existing regulations for prevention of
significant deterioration would be violated long before
ambient levels of the particulate or SC>2 air quality stan-
dards were exceeded in this area and thus act as a mainte-
nance measure in protecting air quality.
PHOTOCHEMICAL OXIDANT ANALYSIS
Existing concentrations of photochemical oxidant in the
McLean-Mercer-Oliver AQMA have been measured at only one
location, the proposed site of the ANG coal gasification
plant north of Beulah. In a sampling period from July 2
through July 30, 1974, the peak one-hour measured concentra-
tion was 0.117 ppm and the second highest reading was 0.105
ppm, compared to the NAAQS of 0.08 ppm. Oxidant sampling
was also performed at the proposed site for Natural Gas
Pipeline's proposed coal gasification plant in Dunn County
from September 1 through October 31, 1975, with the highest
one-hour level being 0.035 ppm."^ Since oxidant concentra-
tions are closely related to solar intensity and tempera-
ture, peak concentrations usually occur during the summer
months with longer daylight hours and higher temperatures.
3-36
-------
Therefore, sampling during summer may have shown comparable
levels to Beulah. The limited sampling data available
indicates that the national standard is already being
exceeded in the AQMA, even though very little of the poten-
tial development of the area has yet occurred.
Concentrations at or exceeding the standard have also
been observed at many other rural and low-population density
locations in the Rocky Mountain states (EPA Region VIII), as
shown by the data summarized in Table 3.13. No extended
sampling for oxidants has shown levels substantially below
the standard, so it might be concluded from this data that
naturally occurring background concentrations throughout
this part of the country approximate the standard of 0.08
ppm.
Oxidants are not emitted directly into the atmosphere
but are produced by a series of chemical reactions between
organic compounds (including hydrocarbons) and nitrogen
oxides in the presence of sunlight. Research has shown that
the rate of oxidant formation is affected by the specific
organic compounds present, the ratio of organic compounds to
nitrogen oxides, and the meteorological conditions such as
solar intensity, temperature, and atmospheric stability.
Generally, peak concentrations are measured within a
few hours of noon, although the emissions of precursor
compounds (organics and nitrogen oxides) contributing to
these peak concentrations may occur several hours earlier or
remain from the previous day. Since the atmospheric photo-
chemical reactions usually take several hours, the measured
oxidant concentrations may occur many miles downwind from
the points of emission origin. This transport phenomenon
has been demonstrated quite clearly by sampling data col-
lected at urban and rural locations in Ohio and surrounding
12
states in 1974. The rural sites, all located within 75
miles of major metropolitan areas, had significantly higher
3-37
-------
Table 3.13. OXIDANT SAMPLING DATA—SELECTED SITES IN EPA REGION VIII
Sampling site
Source of data
Peak one-hour
Highest Next highest
Date occurred
Highest Next highest
Beulan, ND
Dunn Center, ND
Colstrip, MT
Billings, MT
Douglas, WY
Fort Collins, CO
Oil Shale Area, CO
Tract A
Tract B
ANG Gasification
Plant EIS
117
Natural Gas Pipeline .035c
Interim Report
Colstrip Power .080
Plant EIS
Yellowstone County .156
Air Pollution
Control Agency
Panhandle Eastern .076
Gasification
Plant EIS
PEDCo .130
Area Oil Shale .089
Supervisor f
Dept. of Interior
Area Oil Shale .080
Supervisor,
Dept. of Interior
105
032'
. 152
075
127
07/20/74
09/14/75
06/74
07/04/75
06/21/74
10/18/75
Summer, 75
06/26/75
07/10/74
09/08/75
07/11/75
06/21/74
09/09/75
No sampling performed during summer months.
-------
concentrations than the urban locations. Data from this
study are shown in Table 3.14. Identifiable impacts on
oxidant air quality have been shown to extend 50 to 75 miles
12
from urban areas. However, the only urban area near the
AQMA is Bismarck-Mandan; its size and distance from the
sampling sites would indicate only occasional impacts,
certainly not with the frequency that the standard was
violated at the Beulah site (18 out of 27 sampling days).
The probable sources of oxidant in this AQMA and other
rural locations in Region VIII are: (1) downward transport
from the ozone-rich layers in the stratosphere, due to
strong vertical mixing, and (2) photochemical generation
from organics emitted by vegetation. Ozone transport from
the stratosphere may produce ground level concentrations as
high as 0.03 to 0.05 ppm over extended periods and can cause
even higher readings for one-hour peak periods under certain
meteorological conditions.
Other studies have shown that organic compounds emitted
by vegetation may increase oxidants by as much as 0.02 to
13
0.05 ppm. Normally, atmospheric conditions which would
produce high concentrations from downward mixing from the
upper atmosphere are not conducive to high oxidant genera-
tion rates from vegetation, and vice versa. However, the
additive effect from these two sources gives values that
support the hypothesis that measured concentrations of 0.08
ppm and higher in rural areas in the Great Plains are almost
entirely due to natural sources of oxidant.
The Guidelines for Air Quality Maintenance Planning and
14 1 ~
Analysis indicate that analyses for oxidants should be
performed on a regional scale or AQMA-wide basis. There-
fore, hydrocarbon emission densities from man-made sources
for subcounty areas such as the area source grids or anal-
ysis area used in the CDM model were not estimated. County-
wide emissions were assumed to represent a regional scale;
3-39
-------
Table 3.14. OZONE DATA FOR OHIO, PENNSYLVANIA,
AND MARYLAND
City
cl
Maximum one-hour concentration, ppm
Urban
Cincinnati, OH
Dayton, OH
0.18
0.13
0.15
0.14
0.14
0.15
Columbus, OH
Canton, OH
Cleveland, OH
Pittsbugh, PA
Rural
Wilmington, OH
McConnelsville, OH
0.18
0.16
0.17
0.17
0. 20
Wooster, OH
McHenry, MD
Dubois, PA
a Occurred during June 14 to August 31, 1974.
Source: Control of Photochemical Oxidants—Technical Basis
and Implications of Recent Findings. U.S. Environmental
Protection Agency, Research Triangle Park, North
Carolina. Publication Number EPA-450/2-75-005. July
1975.
3-40
-------
emission densities were estimated for each of the four
counties for the base year and projection years, as shown in
Table 3.15.
The resulting emission densities are significantly
lower than average urban hydrocarbon emission densities (100
2
to 1000 ton/mi /yr) reported to have an impact on measured
12
peak oxidant concentrations. Even though regional emis-
sion density in the AQMA is projected to increase approxi-
mately 230 percent by 1985, this density will still remain
quite small as compared to an urban density.
From available air quality data and related research,
it can be concluded that the NAAQS for oxidant will probably
continue to be exceeded as a result of natural contributions
alone. The increase in concentrations over background levels,
that are the result of current and projected emissions of
organic compounds from man-made sources in the AQMA, cannot
be determined from available atmospheric dispersion models
or the estimating procedure described in Appendix J to 40
CFR 51. An air quality maintenance plan would be unable to
assure strict maintenance of the NAAQS for oxidants because
of the impact of natural sources. However, it would still
be prudent to minimize the effect of projected growth by
requiring, through the mechanism of a maintenance plan,
control of new hydrocarbon emission sources.
3-41
-------
Table 3.15. COUNTY-WIDE HYDROCARBON EMISSION DENSITIES
Hydrocarbon emissions, ton/yr
McLean Mercer Oliver Dunn
Source category 1974 1980 1985 1974 1980 1985 1974 1980 1985 1974 1980 1985
Point sources
0
2721
2721
1170
2367
8500
851
2111
2111
0
0
1093
Fuel combustion &
open burning
18
21
17
9
11
8
12
12
12
25
26
24
Highway vehicles
1095
1312
563
413
495
212
218
261
112
417
500
214
Off-highway
vehicles
308
308
308
153
153
153
108
108
108
160
160
160
Railroads
17
31
41
3
5
7
3
5
7
2
4
5
Evaporative
96
162
105
41
69
46
19
33
22
36
57
40
Total
1534
4555
3755
1789
3100
8926
1211
2530
2372
640
747
1536
Area of2County,
mi
2065
1042
721
1992
Emission density, 0.7 2.2 1.8 1.7 3.0 8.6 1.7 3.5 3.3 0.3 0.4 0.8
ton/mi2/yr
-------
4. CASS COUNTY AQMA
ANALYSIS AREA
Fargo, the largest city in North Dakota with a popula-
15
tion of 56,000, is located in Cass County. Together with
West Fargo (pop. 6,500), the two cities account for almost
80 percent of the county's population. The urban area also
includes Moorhead, Minnesota, located immediately adjacent
to Fargo. Moorhead has a population of about 30,000. Cass
County and Clay County (containing Moorhead) are designated
as a Standard Metropolitan Statistical Area (SMSA). The
two-county bistate area also comprises the planning bound-
aries for the Fargo-Moorhead Metropolitan Council of Govern-
ments. However, the Minnesota portion of the urban area was
not designated as part of the AQMA. Limitation of the AQMA
to Cass County makes determination of an analysis are$ much
more difficult.
More than 98 percent of the 17 50 sq mi area of Cass
County is in farmland, either as row crop or pasture. In
recent years, the amount of land in crops has been increas-
ing at the expense of pasture land and summer fallow.
Although urban sprawl has been identified as a potential
lfi
problem by the Council of Governments, it certainly would
not change the percentage of land area in farmland signifi-
cantly within the next 10 or 20 years. The particulate
emission sources in Cass County are primarily agriculture
related, such as grain storage, milling, drying, bulk ferti-
lizer handling, wind erosion of cropland, and traffic on
unpaved farm roads.
4-1
-------
Existing (1974) particulate concentrations at all
sampling sites in the two counties are less than the annual
secondary standard, with the exception of the Fargo indus-
3
trial site which had a geometric mean of 62 ug/m in 1974.
Over the past several years, the three sites in Cass County
have measured approximately the same concentrations each
year—the rural site near background level, the commercial
3
site at approximately 50 ug/m , and the industrial site in
the sixties. A very slight long-term trend toward improved
air quality is noted from the historical data. However, the
data from all sampling locations have relatively high stan-
dard geometric deviations, and the 24-hour primary standard
3
(260 ug/m maximum) has been exceeded in most years at all
three sites in Cass County and at sites in Clay County,
Minnesota.
The terrain in the area is extremely flat. The Red
River of the North forms the state boundary dividing Fargo
and Moorhead, but there is no discernible river valley or
flood plain. The river, which flows north toward Canada,
has a slope of only about one ft per mi.
JThe Fargo area also has very high wind speeds. National
Weather Bureau statistics for the past 10 years show Fargo
to have the highest average wind speed of any U.S. city,
17
14.4 mph. The winds combined with the flat terrain of the
Red River valley cause severe wind erosion problems, which
probably are the cause of the maximum particulate readings
exceeding the short-term air quality standards.
The AQMA is an area of moderate population and economic
growth. The estimated 1980 and 1985 populations of Fargo
18
are 58,400 and 61,000, respectively. Almost all the popu-
lation growth in Cass County is expected to be in Fargo and
16
West Fargo.
For the same reasons as those discussed in Chapter 3,
the analysis area cannot include all of Cass County.
4-2
-------
Therefore, a smaller analysis area of approximately 30 km by
50 km has been specified, as shown in Figure 4.1. This area
includes almost all of the county's population and point
sources, plus the areas with potential for growth. Although
Clay County is not part of the AQMA, point sources from the
Moorhead area have also been added in the analysis because
of their possible impact on ambient concentrations in Fargo.
The uniform distribution of area sources in Cass County
indicates that ambient concentrations in the portion of the
county excluded from the analysis area should not be sub-
stantially different than concentrations in the rural por-
tions of Cass County included in the modeling area.
METHODOLOGY
The AQMA analysis for suspended particulate in Cass
County is based on the use of regional dispersion modeling
to predict ambient concentrations throughout the analysis
area. The same model as that employed in the McLean-Mercer-
Oliver AQMA, the Climatological Dispersion Model, was used
in this analysis, for essentially the same reasons as des-
cribed in Chapter 3. However, only base year conditions
were simulated in this AQMA because neither the total emis-
sions nor percentage contributions by source category were
shown to change significantly from the base year to the
projection years, as shown by the emission inventory summary
in Table 4.1. Modeling results would not be able to show
any distinction between base year and projection year air
quality.
Even though particulate concentrations are not expected
to change in Cass County, the area was still modeled to
determine whether sampling sites are located at points where
maximum concentrations in the county would be recorded, or
whether other areas might have higher concentrations that
4-3
-------
Argutvillt
head
1-94
\ Kindred
CASS CO.
mCHlAND CO.
Figure 4.1. Cass County AQMA analysis area.
4-4
-------
Table 4.1. PARTICULATE EMISSIONS FOR CASS COUNTY
Particulate emissions, ton/yr
Source category 1974 1975 1980 1985
Fuel combustion:
Lignite coal
neg
neg
neg
neg
Distillate oil
109
110
114
119
Residual oil
92
92
113
131
Natural gas
11
11
12
14
LPG
3
3
3
3
Wood
2
2
2
2
Burning:
Open burning
4
4
4
4
Incinerators
4
4
4
4
Agricultural burning
0
0
0
0
Mobile sources:
Highway mobile sources
351
328
287
175
Off-highway vehicles
119
119
119
119
Aircraft
6
6
7
7
Railroads
34
40
61
82
Processes:
Industrial processes
0
0
0
0
Portable batch plants
Fugitive dust:
Unpaved roads
57,300
57,300
57,300
57,300
Agriculture
14,745
14,745
14,745
14,745
Construction
518
518
585
689
Mining
0
0
0
0
Paved roads
1,856
1,876
2,068
2,241
Point sources
1,189
1,190
1,109
1,134
Total
76,343
76,348
76,533
76,769
4-5
-------
are exceeding the annual standards. In this case, a revi-
sion for attainment rather than a maintenance plan would be
developed.
Data for the base year model were collected as des-
cribed in the following sections.
Source Locations
The UTM coordinates for point sources were obtained
from NEDS listings and checked by locating the sites on USGS
quadrangle maps. The coordinates for each source are listed
in the point source summary in Appendix A.
A grid system covering the analysis area was specified,
as shown in Figure 4.2. The area source emissions for Cass
County were apportioned into these 18 grids in order to
distribute the emissions spatially for modeling.
Source Emission Rates and Stack Parameters
Point source emission data for the base year 1974 and
stack parameters for these sources were obtained from file
data in NEDS format, and are summarized in Appendix A.
These data were input directly to the CDM model.
The point source emission projections shown in Table
4.1 were generated from the base year emission estimates and
U.S. Department of Commerce OBERS projections of economic
growth in the Fargo-Moorhead SMSA by industrial classifica-
19
tion. These national projections were used due to lack of
any locally generated estimates of increases in manufactur-
ing activity or employment.
With two exceptions (Fargo Foundry and North Dakota
State University), the point sources are all in the agricul-
ture and food products industrial classification, which has
projected growth as follows for future years:
4-6
-------
Arthur
5210.
Fori
Mopleton
West Fargo
656.
CASS CO.
mCHiANO co.
Figure 4.2. Area source grids for Cass County.
4-7
-------
Food industry earnings
compared to 1974
Metals industry
earnings compared
Year
to 1974
1980
1985
1975
1.00
1.01
1.03
1. 05
1.31
1.64
Emissions for Fargo Foundry were increased according to
projected earnings for the metals industry. Projected
emissions for North Dakota State University for 1980 and
1985 were reduced by 80 percent from the 1974 level to
account for control equipment scheduled for installation on
the lignite-fired boilers. This projection methodology,
20
described in Volume 7 of the AQMA guidelines, probably
overestimates future point source emissions because it does
not account for the generally lower emission rates per unit
production at the new or enlarged facilities providing the
additional activity.
Area source emissions and projections by source cate-
gory were obtained from the previously noted AQMA emission
2
inventory report. The emission estimates by source cate-
gory from this report were summarized in Table 4.1. These
emission totals for Cass County were allocated into the 18
area source grids by the same procedures used for the
McLean-Mercer-Oliver AQMA (see table 3.6), except aircraft
emissions were assigned to the grid containing the airport
(no. 11) and construction emissions were divided uniformly
among the three grids encompassing Fargo and West Fargo.
The resulting area source emissions by grid are shown in
Table 4.2.
Meteorology
Meteorological input data in the format of the day-
night STAR program are available for the Fargo National
4-8
-------
Table 4.2. CASS COUNTY EMISSIONS BY GRID
AND SOURCE CATEGORY, 1974
Source category, ton/yr
nJ
•H
W
rf
T3
¦p
O
¦P
O
+>
Cr>
>1 V
>•,
m
3
4-1
VI
CO
G
(Tj rH
0
a)
3
M
IT]
a
•rl
£ O
S
n
> to
O
4J
U
•a
H
r-H ,Q
G G
A -h
1 A
rH
(0 *0
•H
0)
O
a)
(0
a) e
d) n
tP.C
iw 0*
a fO
u
G
M
>
¦P
Grid
3 0
3
•H
O A
«
D U
<
U
<
a.
E-"
1
1
neg
7
3
1331
326
16
1684
2
neg
neg
8
3
neg
1208
323
43
1585
3
neg
neg
7
3
1
1293
317
28
1649
4
neg
neg
7
3
1
1379
326
6
1722
¦5
neg
neg
7
3
1
1293
326
20
1650
6
2
neg
8
2
1
1198
300
52
1563
7
1
neg
8
2
2
1227
304
40
1584
8
3
neg
8
2
1
1265
300
44
1623
9
1
neg
7
2
1
1132
292
49
1484
10
16
1
11
1
2
105
135
173
283
727
11
41
1
6
neg
48
173
6
149
424
12
123
3
13
1
76
173
338
727
13
neg
neg
8
3
2
1474
326
8
1821
14
neg
neg
7
3
1331
326
15
1682
15
1
neg
8
2
1
1436
309
35
1792
16
neg
neg
7
3
1
1474
326
1811
17
1
neg
7
3
1
1227
326
30
1595
18
1
neg
8
2
neg
1312
309
38
1670
Total
191
5
142
40
16
19809
4871
519
6
1194
26793
4-9
-------
Weather Service Station for the five year period 1970-1974.
The weather station is located at Hector Airport, within the
city limits of Fargo. Based on the reliability of NWS data
and central location of the station, this set of meteorolog-
ical data has been used in the model.
Receptor Locations
Receptor locations in the CDM model were specified at
the corners of the 10 km grids in the rural portions of Cass
County and at 5 km intervals near the urban area, as shown
in Figure 4.3. This configuration gave complete coverage of
the area and concentrated the receptors in the areas where
highest concentrations were anticipated. Also, receptors
were designated at the locations of the four particulate
sampling sites (three in Cass County, one in Moorhead) so
that predicted and measured concentrations could be com-
pared. This receptor array resulted in a total of 47
receptors spaced throughout the analysis area.
Background Air Quality
A background value for this AQMA was determined by the
method specified in the Federal regulations (Section 51.13c)
of using the concentration measured at a nonurban site in or
near the AQMA that is unaffected by nearby emission sources.
The rural Fargo site, located about four miles south of the
city limits, was established as a background site, but has
some source impact from the intensive farming in the area.
3
In 1974, it recorded an annual geometric mean of 34 ug/m
3
and an arithmetic mean of 4 8 ug/m . Previous years' data
showed similar concentrations.
Actual background for the area should be higher than in
3 ^
the McLean-Mercer-Oliver area (20 ug/m geometric, 24 ug/m
4-10
-------
ArgutvllU
MepUton
Co»i»llon
Moorhtod
Figure 4.3. Receptor locations.
4-11
-------
arithmetic) because of higher wind erosion potential, but
probably somewhat lower than measured at the rural Fargo
site. By subtracting the source impact indicated by the
uncalibrated CDM model at the sampling site location, a
better estimate of background would be obtained. This value
3
for source impact was 4 ug/m , resulting in a corrected
3
arithmetic background of 44 ug/m and, using the SGD of 2.2
recorded for the site, a geometric mean background concen-
3
tration of 32 ug/m . These are not much different than the
uncorrected values, but the lower values have been used in
the modeling.
PREDICTED AIR QUALITY AND MODEL VERIFICATION
The CDM model was run with the input data described
above, and the modeling results were compared with measured
particulate concentrations for purposes of verification.
Linear regression analysis of predicted and measured values
produced a slope of 1.236 and a y-intercept of 5.0 for the
line of best fit, and a correlation of 0.49. The data sets
are shown below:
Predicted,3 Measured,
Receptor ug/m3 ug/m3
Fargo rural 48 48
Fargo industrial 59 82
Fargo commercial 52 56
Moorhead 61 52
a 3
Includes background of 44 ug/m .
While this correlation was only fair and not high
enough to use the model results as the sole basis for deter-
mining attainment or non-attainment, three of the four data
sets showed reasonably good agreement. Also, it should be
emphasized that the primary purpose of the modeling was to
4-12
-------
determine whether there were other locations in the AQMA
where concentrations might be higher than the measured
values, possibly exceeding the secondary or even the primary
annual standard. Even with the moderate correlation, the
model should indicate if the sampling sites are in or near
the points of highest concentration in the analysis area.
In order that the model-predicted values match the
measured concentrations more closely, a calibration factor
of 1.236 was applied to all predicted values. Also, the
values were converted from annual arithmetic to annual
geometric means by use of equation 1 on page 3-26, using an
SGD of 2.2 from the Fargo rural and Fargo industrial sites.
The ratio of geometric mean to arithmetic mean was calcu-
lated to be 0.733 for the SGD value of 2.2. Thus, model
outputs for each receptor were converted to predicted con-
centrations by first multiplying by 1.236 for calibration,
3
then adding 44 ug/m to account for background, and finally
multiplying by 0.733 to change the value to geometric mean.
The predicted particulate concentrations are shown in
Figure 4.4. Concentrations are obviously highest in the
immediate Fargo area and, at the scale of analysis employed,
no points of expected higher concentration than the existing
sites were determined.
It is interesting to note that on the basis of emis-
sions in the county and in the analysis area (see Tables 4.1
and 4.2), unpaved roads and agriculture appear to be the
major problem. However, these emissions are so evenly
distributed throughout the analysis area that they add only
a few ug/m"* to concentrations; it is point source emissions
and urban area source categories that contribute to the
higher emission densities and ambient concentrations in the
Fargo-Moorhead area. For example, the four receptors with
highest predicted concentrations had 50 percent of their
concentrations above background resulting from point source
4-13
-------
Arthur
Argusville
Fqrgi
[Moorhead
We»» Fargo
1-94
Kindred
CASS CO.
• ICHIAND CO.
Figure 4.4. 1974 geometric mean particulate concentrations
in Cass County AQMA.
4-14
-------
emissions. Area source emissions for the urban grids in
which these receptors were located showed that only about
five to 10 percent of the total contribution at these
receptors was from unpaved roads and agriculture; reen-
trained dust from paved streets, fuel combustion, and
construction had greater impacts than the former two source
categories.
Although the modeling does not show any violations of
the annual air quality standards, either primary or secon-
dary,3 the short-term standards are probably exceeded
throughout the AQMA. Since no change in emissions is anti-
cipated over the next 10 years in Cass County, the 24-hour
standards will probably continue to be exceeded. In 1974,
3
the primary standard of 260 ug/m was violated on two sam-
pling days and the secondary standard of 150 ug/m3 was
exceeded on one additional occasion at the rural Fargo
(background) site. In comparison, the primary standard was
not exceeded and the secondary standard was exceeded only
once at both the Fargo commercial and Moorhead sites. Two
violations of the primary standard and eight additional
violations of the secondary standard were recorded at the
Fargo industrial site. Therefore, it can be concluded that
the infrequent high concentrations are in most cases due to
some uncontrollable fugitive dust source, either agriculture-
related or from long-range transport. Since maintenance
measures would not be effective in reducing these occasional
high concentrations, it is concluded that no maintenance
plan is required in the Cass County AQMA.
a The 1974 measured geometric mean at the Fargo industrial
site was 62 ug/m3, a marginal violation of the secondary
standard.
4-15
-------
9. Environmental Impact Report: North Dakota Gasification
Project for ANG Coal Gasification Company. Woodward-
Clyde Consultants. March 1975.
10. Guidelines for Air Quality Maintenance Planning and
Analysis. Volume 13: Allocating Projected Emissions
to Sub-County Areas. U.S. Environmental Protection
Agency, Research Triangle Park, North Carolina. Publi-
cation Number EPA-450/4-74-014. November 1974.
11. Meteorological and Air Quality Aspects of a Proposed
Coal Gasification Plant in Dunn County, North Dakota.
Interim Report. North Dakota State University, Depart-
ment of Soils, Fargo, North Dakota. Prepared for
Natural Gas Pipeline Company of America. January 1,
1976.
12. Control of Photochemical Oxidants—Technical Basis and
Implications of Recent Findings. U.S. Environmental
Protection Agency, Research Triangle Park, North
Carolina. Publication Number EPA-450/2-75-005. July
1975.
13. Rasmussen, R. A. and Robinson, E. The Role of Trace
Atmospheric Constituents in a Surface Ozone Model.
Washington State University, Pullman, Washington.
1975.
14. Guidelines for Air Quality Maintenance Planning and
Analysis. Volume 2: Plan Preparation. U.S. Environ-
mental Protection Agency, Research Triangle Park, North
Carolina. Publication Number EPA-450/4-74-002. July
1974.
15. North Dakota Growth Indicators, 1975. North Dakota
Business and Industrial Development Department, Bis-
marck, North Dakota. June 1975.
16. Land Use and Development Policies, Fargo-Moorhead
Metropolitan Council of Governments, Moorhead, Minne-
sota. January 1974.
17. Off to Windy City? Well, Dear Reader, Don't Go to
Chicago. Wall Street Journal. 186:87, October 31,
1975. p. 1.
18. Communication with Mr. Ostenson, unpublished projec-
tions. North Dakota State University, Agricultural
Economics Department, Fargo, North Dakota. October
1975.
R-2
-------
19. OBERS Projections of Regional Economic Activity in the
U.S., Series E Population. Volume 2: BEA Economic
Areas. U.S. Water Resources Council, Washington, D.C.
April 1974.
20. Guidelines for Air Quality Maintenance Planning and
Analysis. Volume 7: Projecting County Emissions,
Second Edition. U.S. Environmental Protection Agency,
Research Triangle Park, North Carolina. Publication
Number EPA-450/4-74-008. January 1975.
R-3
-------
APPENDIX A
NORTH DAKOTA AQMA POINT
SOURCE SUMMARY
-------
ST CC
PL7 PT YR PlAfcT/POJ.NT \!A»*£
LT-< J ,: 2CiX0001 03 70-
'3Lr? LC—fe 70 » 0—91 '< t 0
a)ijLji!OijOOOul O'i 70
bC I i
¦ f.'f 0 ¦
¦jipC21' u 13o0 ju 1—«r-5—7-a-
iTOlxr-
p"L>'
0
tX0^
lfcl.
-w-
206
iAl
-w-
¦Wr
-MH-
¦34-
-W5-
4&i0t?2C>
-wtr
OQitL
. i» 0
,T. ...
—E Y = 4"f|>s
j*022 CI300007 01 70 OOfrOCR ST»TCG P»Vl*iO/CTHC . . .
12
-e-
o
0
0
0
28
60
3
51
0
0
c
0
0
6
13
1
11
120
-t2-
0
0
0
0
1
2
0
2
27
Z+
0
0
0
101
2
4
0
4
0
135
73
0
2
3
39
58
12
91
-9+-
-«r£-
¦SOfr-
¦
G
~0
0
0
229
115
20
609
548 2055
—ee
j >0 tJj i)l j 00i>07 Op 70
¦WJTtiiy Ofty
-e«-
3S02SC13QUL>lt Jl 70 in f HPRV» *0Pi '»LT—1/PT4
H'»fl^Cl30iJ01i OH 70 OTH£fi
pO
I l»J
i 150
SH-
-------
(_
I .,
f » - m
-n \ ST C3 tC PLT ?T VR PLftNT/POINT NAKE
o" r t, jij«t.j * k vi > / >"i e * lv ^ «
HT. [) IA Tctif FLOW PLUME
hOrlZ VERT (FT) (FT) (r) (ACFK) hT PART S02 NOX
EMISS.TPY
HC CO PART S02
r
|^aui!£':t3CiOJll 03 70
dUILER
-23-
-M-
h _ j^»Oga6A-3vV)iiia-ai "?0-»r. ;W>DV. ASPhALT 3/P'
1 |
: 'i^so£^ci3cooia 70
i*'
| ' |'3 »0J J Ci 30J013 01 70 N 11H0r
-+4-S-
-4+5-
I .1
>SO£-aOi300013 0J f-
i »50aJCI30QO'
a 6 >313 ¦
>0*
.,1,CQNVV6HN
>6*
S>v
OPEN stooa
*&-
-e-
¦ iMPav.*sft^A«rr a/r>oT.oav
-?©-
-^e-
H-e-
-«-
-54"
-**-
-S+-
-r&5-
-?+-
-e-
| 3SCtJC13o00 oi.
¦^e omcn-"-
3
. „!
: J35C22C13CC015 01
7A GARDNER FAR*.EL/SHIPEHANO
654.0
5223.0
0
.0 77
0
20E
7
0
0
0
0
21
0
' ^
: 3 5Gz 2C1 ~3 0 3J15 02
74 CLEAN EPS
654.0
5223.0
20 E
77
0
1
0
0
0
0
4
0
-3 5 022 6 13 GOO16-01-
1
-74—A 3S Aft Arte HI LL ER EL / SHP6HN
622.0
5204.0
0
.0 77
0
20 £
10
0
0
0
c
13
0
|35C22C130001o 02
74 3 CLEANS
622.0
5204.0
20E
77
0
5
0
0
0
0
6
0
I-J5C22C13C0016 03
74 HIX SULK FERTILIZ
622.0
5204.0
0
.0 77
0
20E
2
0
0
0
0
1
0
r ...
1 36022C-130CKH7-01—74—GSAN(>!N -GOL{> i.-45EL— FED/GR-
650.0
5230.0
0
.0 77
0
20 E
45
— 0
c
- 0
0
12 —
0
"
5 02 2C13C0018 01
^ -
!35C22C13C00Id 02
1
74 HARWOOO FAPK.EL/SHIPCHAND
661 .0
5205.0
0
.0 77
0
20E
8
0
0
0
0
11
0
r
74 2 CLEANE
661 .0
5205.0
20E
77
0
1
0
0
0
0
11
0
!:
| 35C2201 JvOOtQ—03^
1
7v LIQUID FEkTL
661.0
5205.0
0
.0 77
- 0
20E
0
0
- 0 -
0
0
0
0
35C2201300019 01
74 HICKSC.M GRAIN CO/SHIPtHAN
661 .0
5170.0
0
.0 77
0
20E
61
0
0
0
0
94
0
35C22C13C0019 02
74 1 CLEAN
661 .0
5170.0
20E
77
0
0
0
0
0
0
0
0
j
35C2291300019 03"
-74 BULK£L!QUID FERTL
661.0
5170.0
0
.0 77
0
20E
0
0
0
0
0
0
0
! .
'
i *•
35022013GG020 01
74 HORACE FARN.EL/SHIPPCHAND
0 60 . 0
5180.0
0
.0 77
0
20E
66
0
0
0
0
96
0
I:
35C22C1300020 02
74 2 CLEANc
660.0
5iao.0
20E
77
0
7
0
0
0
0
0
l-
i -
3502201300020-03-
74 - —- — - 1 NAT.GAS DRY
660.0
5180.0
20 E
0
1
0 -
0
0
0
-
0
35022C13C0020 04
74 BULK&LIQUID FERTL
660. 0
5 130.0
0
.0 77
0
20 E
6
0
0
0
0
2
0
1 „
j
35022C13CCJ21 01
74 HUNTER GRAIN Cu/S hi P&HA.ND
635.0
5228.0
20 E
77
0
20
c
0
0
0
97
0
1. r
p
-35C22C1300021-02
74 4 CLEANS
635.0
5223.0
20E
77
0
21
0 -
0 -
0
0
0
u
k*
35C22C13G0021 03
74 1 DRY
635.0
5228.0
20E
0
3
0
0
0
0
0
**
35022C1300021 04
74 SULK&LIQUID FERTL
635.0
5^2 6.0
0
.0 77
0
20E
13
0
0
0
0
7
0
-------
- ! I.~. NL'^QffS UTM L23~&4 74 BULK FERTL 633.0 5181.0
*"" ' J 35022 C13CG024 01 74 RED RIVES GRAIN /SHI P&HNO 649.0 5195.0
¦ 25C2> C 1300^24 02 74 1 CLEAN 649.0 5195.0
j ,|35«22-e^3eex)25—Ot-^4-p-AGE—etvCO/-S+tlPPlNG£,HANDL 608.0 5223.0
| "!,35C2201200025 02 74 2 CLEANE 608.0 5223.0
i ";35C22C1SCC025 03 74 GSINCI 608.0 5223.0
" } '
! J-35C22 >)'35d22t;13v002o~03 -74 BULK FERTL 651.0 5202.0
], 35G22C1300027 01 74 ROMAN MEAL MILL TNG/SUPCH 669.0 5193.0
"J. 35022C13CG027 02 74 3 CLEA\E o69.0 5193.0
*'35C22Ci3CCj27~Cri _J?2 - FLAKI 669.0 5193.0
"I 35C22C13CQ028 01 74 TGwER CITY GRATN/SHIPtHND 601.0 5197.0
"i 25C22C1300028 02 74 1 CLEAN 601.0 5197.0
** f
3iC22Cl3C0029~01- 7A PEAVf?y-PftOO.SE:-
-------
1
HI.
0 IA
ST CC AQ PLT PT YR
PLANT/POINT NA«E
HCRIZ
VFRT
(FT) (FT)
I
(35C22C13CG032 01 74
REO RIVER GP.AIN/iHIPCHNOL
660.0
5194.0
206
•
•1
35C-22C1300032 02-74
_ 2 CLEANE
660.0
5194.0
20E
•
,
35022 C1300032 03 74
1 PRCP.GR NAT.CAS DRY
660.0
5194.0
20E
•
;
35C2201300032 04 74
BULK FERTL
660.0
5194.0
0
.0
»
»«
-35022-&1300&33-—74" GQCD~LA8EL~ FEEDS/SHIPEHND
660.0
5194.0
2UE
m
35022C13C0033 02 74
HAMP VIL » ROLL,SLNDtPEL
660 . 0
5194.0
0
.0
..
..
35022 CI300034 01 74
INTERSTATE SUNFL&/SHIPCHN
660 .0
5194.0
2CE
•
-35022 el 3 00334-92-"74"
" —- 15 CLEANE
660.0
5194.0
20E
•
-
-
35C22C1300034 03 74
1 PRCP:DRY
660.0
5194.0
20E
•
*
35C22C13C0035 01 74
WHTLNO FRN.UNION GRN/SHPt
626 .0
5196.0
0
.0
-
-3502-2 ei30Ct> 35—02-74
2-CL-EANE
626.0
5196.0
20E
35C22C1300035 03 72
1 HAKKERMI
626 .0
5196.0
0
.0
35022CI300036 01 74
WOODS FRM.COOP EL/SHIPSHN
639.0
5171.0
0
.0
^5C2iO! JU0^36-02 74
3 CLEANE
639.0
5171.0
20E
•
>
1
35C22C13C0036 03 74
1 LPG DRY
639.0
5171.0
20E
•
<1
35C22C1300036 04 74
&ULK FERTL
639.0
5171.0
0
.0
a
"35022
-------
i
l-C. NUMBER
L:T« CUO^D ,.<«
S T A
HT. CIA
(FT) (FT)
i
ST CO AQ PL T P T
YR
PLANT/POINT NA*E
HQS IZ
VERT
.-f
35C22 C13 0004J 03
74
IPG DRYE
635.0
5218.0
20 E
\
\
1
i
41
36C22C13CC040 04
74
GRINOING tRCL LI
635.0
5216.0
0
0
35C22C1300040 05
74
SELL 3ULK FERTL
635.0
52L8.0
0
0
-1
*1
3 5022 C13G0041 01
74
BUFFALO CARG1LL/SHIPfcHNDL
611.0
5197.0
0
0
;? • L
(' ,w i
35C22C13C0041
74
- SELL BULK FERTL
611.0
5197.0
0
0
1
i
i
350 Z2C1300042 01
74
BUFFALO PU PROD.SER/SHPEH
611 .0
5197.0
0
0
- "
..
35C22C13C0042 02
74
1 CLEA!\i
611.0
5197.0
20E
¦
3SC2201300042 03
74
SELL HULK FERTL
611.0
5197.0
0
0
J 35022 CI ^00043 01
74
WALTON EL CO/SHIPNG£HANCL
636.0
5196.0
0
0
3 5022 01300043 02
74
2 CltANE
636.0
5196.0
20F
"3it220 13CQ043-03
-74
636.0
5196.0
20E
1 PhU P•DRY c
35022C13C0O*4 01
74
CHAFFEE FR*.COCP.EL/SHP£H
625.0
5131.0
0
0
.
3bC22CI3C0044 02
74
2 CLEANE
625.0
5131.0
20E
.V
3 i C 2 2C-1-3 eOO44~0 3
74*
- 1 NAT.GAS-DRY
e»25.0
5181.0
20E
>
i
3 5C 22Ci3CC044 04
72
LIO AND BULK FERTL
625.0
5181.0
0
0
m
a
i
350Z2C1300045 01
"35C22 013C0045-02
72
KENT OORAN G^AIN/SH1PCHNO
648.0
5175.0
0
0
A
72
ScLL SULK FERTL
648 . 0
5175.0
t5 """"
0
*j 35C22C13C0046 01
^j35C22C130G046 02
74
OUR91N CCCP EL/SHIPNGCHND
641.0
5166.C
0
0
74
1 CLEAN
641 .0
5136.0
20E
•
3502201300047 Ol
74
OBA- FARM- CRAIN/SHIPNG&HNO
619.0
5184.0
0
0
o
*'
35C22C1300047 02
74
2 CLEANE
ol9. 0
5184.0
20E
35C22CI3CO049 01
72
DAHLGRcN CC/SHI P,HAI\CiCLE
669.0
5 194 . 0
20E
' „
35022C1300U49 02"
72
DRY
C69.0
5194.0
20E
i -
35022C1300050 01
73
D£P TER.M.-FARGC/SHI P&HANO
669.0
5194.0
0
*
35C22C13C0051 01
74
1 NO.MOLASSES/DRYER
649.9
5194.8
36 8
5
-35C22C-1300052--01-74
FAf.GO MILIS/SHIPIKG
666. 8
5193.9
0
0
M
35Ci2C13CC052 02
74
CLEANE
666.8
5193.9
0
0
44
35034C172Q000 OG
0
•
•
c K DAT A ESTIMATED EMISSIONS, TPY ALLGWABLE
temp flow pluke emiss.tpy
(FJ (ACF*) HT PART SC2 NOX HC Co PART S02
04 0 0 0080
77 0 20E 15 0 0 0 0 3 0
77 0 20E 9 0 0 0 0 10 0
77 0 20E 10 0 0 0 0 26 0
77 0 20E 1 0 0 0 0 - 3 0
77 0 20E 18 0 0 0 0 40 0
77 00000010
77 0 20E 0 0 0 0 0 0
77 0 20E 5 0 0 0 0 14 0
77 020C0030
0 0 0 0 0 — 0 1 0
77 0 20E 70 0 0 0 0 114 0
77 0 27 0 0 0 0 22 0
0 10" 0 0 — 0 ' 0 —11 " 0
77 0 20E 0 0 0 0 0
77 0 20 E 5 0 0 0 0 8 0
77 O '20 E 0 0 - 0 — o 0 0
77 0 20E 10 0 0 0 0 24 0
77 020000 0
77 0 20E 11 0 0 0 0 31 " 0
77 02000040
77 000000 16 0
0 1 0 0 0 0 ' " 0
20E 7 0 0 0 0 18 0
200 25184 0 21 0 0 0 0 0
77 0 20F 10 0 0 0 0
77 0 20E 1 0 0 0 0 0
-------
I.D. NUMBER ur.i cooko.km
ST CO AC PLT PT YR PLANT/POINT NAME HCR1Z VERT
r
:V
j35C34C1720000 01 DUNN CCUNTY POINT SCCRC £S
I
? , f 35Ci^Gl72'J0OO~O2 -
' * t
M 35C34C1720001 01 74 DODGE FARMERS ELEV,SHIPPI 712.0 5242.0
3SG240172C001 02 74 GRAIN ClEANl 712.0 5242.0
i ' -5«^Jret72C{r02-Oi—74-CUKN "CENTER GRAIft, SHIPPI 680.0 5247.0
i 25C34C1720002 02 74 GRAIN CLEAN I 680.0 5247.0
...
35C3^C i 720003 01 73 KILLDEER EQUITY EL,SHIPPI 670.0 5248-0
| ;^3SC34vl72<3O03 02-73 GRAIN CLEANI 670.0 5248.0
j 35034C1720003 03 73 FEED POLL 670.0 5248.0
"'[35C54C1720004 OX 74 KILL0EER GRAIN CD.,SHIPPI 670.0 5248.0
*1
^j-S^C34ei72 (ACFMJ HT PART S02 NCX HC CO PART S02
20E
20E
20E
20E
20E
20E
20F
20E
20E
20E
20E
20E
20E
6
1
10
1
12
1
15
6
0
3
14
"0
0
0
0
0
0
0
0
0
0
0
0
0
o
0
0
0
D
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
o
0
0
0
0
0
0
0
0
0
0
0
0
0
0
10
26
3
14
- 3
5
9
4
1
46
~4
2
0
0
0
0
0
0
0
0
0
0
0
0
0
20E . 77 060000 40 0
20 E . 77 0 4 0 0 0 0 20 0
0 .0 77 0 20E 12 0 0 0 0 32 0
20E . 77 02000080
0 .0 77 0 20E 4 0 0 0 0 " 3 0
G .0 77 0 20E 10 0 0 0 0 28 0
20E . 77 0000001C
0 .0 77 t) 20E 27 0 0 O 0 49 C
20E . 77 080000 13 C
0 .0 77 0 20E 31 0 0 0 0 57 C
-------
I.e. NbMBES
V; |ST PLT PT VR PLANT/POINT MAKE
V
rJTicV^cT:
L720005 02 74
¦35C»2Cl-720005-03 74
• I •!
i-t
I-
3 50 720 1720006 01 74 PV CC. PRLD. SER V/S HI PtHAND 318.0 52SQ.0
35C72C1720006 02 74 4 CLEANE 31.
J35C72C1720007 02 74 3 CLfANiE
! 35072C1720009 01 74 EQUITY ELtTRADE CO/SI-FXHN
! j
: iSC"i2C172CC09 02 74 - 4 CLEANE
i "i
: j35C7201720G09 03 74 GRINDI
' '^350 720-1720010-02—74
!35G72C1720010 03 74
>
r
-j
. r3-5C72Ci7
-------
I.e. .M,«S5t
urn cookp.uk
i TACK DAT A
\ST CO AC PLT PT YR PLA.MT/POINT K&V.E
HT. DIA TEMP
HGRII VERT (FT) (FT) IF)
ESTI.-ATEO EMISSIONS, tpy
FlQv. plume
(ACFK) HT PART S02 NGx
;jS5C-,6C1720002 05 74
' 25€ /401-720004-01—74-
• r
'\3SC"i6C1720305 01 74
i350"/6C1720005 02 74
' i-1
I h3^e>6-er7?oeo6-oi-nr4
::i
35C76C1720006 02 74
.'j 35C76C1720006 03 74
'' ;-3 S C f6 « 7 2000T-01- 74
J 35C76C1720007 02 74
35C"i6C172C00C 01 74
^ r^^t"i6 Ct"7 ?etKr8-02 -74-
J^35C16CI 72 0008 03 74
J 35CUC172GOOO 00
BOILER
-UMTEP PV;R.ASS0C/6CiILER
BEULAH FARM.U N. EL/SHIPC H.N 293.0 5238.0
2 CLEANS 290.0 5238.0
6EULAT PV.-CO./SHTPNGEHNOL 290.0 5238.0
2 CLFANE 290.0 5238.0
GRINOEP»ROLLERtHA.^C.ERrtI 290.0 5238.C
FARMERS GRAIN* CG./SX IPCH.N 722.0 5241.0
2 CLEANS: 722.0 5241.0
HAZEN FARM EL/SHIPnGCKsDL 302.0 5241.0
4 "Ctf&NE- 302i0 5241.0
1 FEED ROLL 302.0 5241.0
74 5.0 42 5E 37400
290.0 523fi.(J
3»-T 523T.5
3gO«0 &a»3ie 225 15.0 350 816000
20E
20E
0
20E
0
0
20E
0
20 E
0
.0
.0
.0
77
77
77
77
77
77
77
77
77
77
0
0
206
0
20E
20E
0
0 2C£
0
0 20 E
0
0
218
3822
4
0
9
0
1
7
1
15
Cr
2
346 103
8443 2682
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
c
HC
le
447
0
0
— o
0
0
0
0
0
— o
0
CO
36
694
0
0
0
0
0
0
0
0
0
0
allowable
EMISS.TPY
PART S02
654 1092
2092 20655
28 0
3
25
1
1
22
3
36
O
0
0
0
0
0
0
0
0
0
i."L3S-cft c-i ?2eeeo-e-i
an
-Oirl-VtR—eC— PCINT SOURCES-
135C c6C1720000 02
.! 25C£60l72000l 01 74 MINNKOTA PWR.COOP/801LER
332.'O 5213. 7
aat.o seeoio
250 19.0 330 167000
5740 17863 5128
851 1701 2932 30225
-------
APPENDIX B
AIR EMISSION SOURCES FROM A LURGI
DRY-ASH GASIFICATION FACILITY USING LIGNITE COAL
-------
AIR EMISSION SOURCES FROM A LURGI
DRY-ASH GASIFICATION FACILITY USING LIGNITE COAL
BACKGROUND
This study presents the potential emission sources for
a Lurgi mine-mouth gasification process, producing 250
million scfd of either low or high BTU gas from lignite.
North Dakota and Montana, located in EPA Region VIII, have
vast reserves of lignite, comprising 98 percent of the total
U.S. reserves. Low BTU gas is used for industrial fuels or
power generating turbines. High BTU gas is used as substi-
tute natural gas.
PROCESS DESCRIPTION
The North Dakota lignite has the following character-
istics :
The Lurgi coal gasification process involves mining
lignite, reducing its size to that necessary for gasifica-
tion, followed by pretreating and blending the fuel. After
the lignite is gasified, the low BTU gas is either cleaned
by stripping H2S and tar, or processed through a CO shift
converter and a methane synthesis operation, resulting in a
High heating value
Ash content
2
Moisture content
2
Sulfur content
6500 BTU
7-9%
37%
0.6%
B-l
-------
high BTU product. Previous Lurgi systems have had difficulty
in processing caking coals. Lignite, a non-caking fuel, can
be handled in the gas generator without agglomeration.
Process flow diagrams, illustrated in Figures 1 and 2,
explain the various flow streams and potential pollutants
from the production of low and high BTU gas.
Lignite preparation is similar for both types of gasi-
fication. The mined lignite is crushed and transported to
stock piles. The coal from various storage piles is blended
prior to secondary crushing. The high moisture content of
the fuel (37 percent) will lower uncontrolled emissions.
Conveyor transfer points are equipped with water sprays.
Crushers have either fabric filters or wet scrubbers to
further reduce particulate emissions to the atmosphere.
Lignite storage piles are sources of fugitive dust
emissions. Spraying water in the process areas and roads
eliminates major sources of fugitive dust.
In the Lurgi process (Figure 3), gasification takes
place in a counter-current moving bed of lignite as it falls
from the distributor onto the grate. The gas stream leaving
the gasifier contains coal dust, oil, naptha, phenol,
ammonia, tar, and other constituents. This mixture, after
going through a cleaning unit to remove sulfur compounds and
tar, is the final product, a low BTU gas of approximately 50
to 100 BTU per scf.
High BTU gas requires additional steps as shown in
Figure 2. In the shift conversion section, water is added
to the CO rich gas, producing carbon dioxide, hydrogen, and
heat. In the methanation reaction section carbon oxides are
added to hydrogen rich gas to produce the high BTU gas.
B-2
-------
tu
I
u>
CONVEYOR
TRUCK OR RAIL
CONVEYOR
CONVEYOR
ELEMENTAL
SULFUR
COAL
LOCK
DIRTY GAS
LURGI
GASIFIER
CLEAN LOW BTU GAS
TO TURBINE OR OTHER SOURCE
EFFLUENT
GAS
OXYGEN
STEAM
ASH
INCINERATOR
SULFUR
RECOVERY
UNIT
COAL
PILES
FOR
BLENDING
SECONDARY
CRUSHER
SIZING
SCREENS
LIGNITE
MINE
PRIMARY
CRUSHERS
INCINERATOR
i
¦
EXHAUST
GAS
Figure 1. Process flow diagram for Lurgi low BTU gasification process.
-------
CONVEYOR
TRUCK OR RAH
CONVEYOR
FLUE GAS
COAL
LOCK
CLEAN GAS
SYNTHETIC
NATURAL GAS
DIRTY GAS
DIRTY GAS
METHANATION _ CLEAN GAS
LURGI
GASIFIER
PIPELINE GAS
DIRTY GAS
EFFLUENT
STREAM
.OXYGEN
STEAM
ASH
FLUE GAS INCINERATOR
SULFUR
RECOVERY
UNIT
COAL
PILES
FOR
BLENDING
GAS
JRIFI CAT IC
POWER AND
STEAM
PLANT
SIZING
SCREENS
SECONDARY
CRUSHER
PRIMARY
CRUSHERS
LIGNITE
MINE
SHIFT CONVERSION
C0+H,0—»»H,+CO,
ELEMENTAL SULFUR
Figure 2. Process flow diagram for Lurgi high BTU gasification process.
-------
COAL LOCK
DISTRIBUTOR
ASH LOCK
Figure 3. Lurgi pressure gasification for fuel gas production.
(Lurgi Clean Fuel Gas From Coal Publication 0 1007/10.71)
B-5
-------
EMISSION SOURCES
The primary potential for emitting particulate matter
from a Lurgi system occurs between the mine and the gasi-
fier. These sources include crushing, storing, sizing, and
all process transfer points. If the power plant fires
lignite or coal fines, it may also be a major source.
Incineration/super heater sections fire a gaseous fuel, and
therefore are not particulate sources. A study of pollutant
control of fuel gasification processes found little or no
3
test data for fuel handling emissions.
4
The Compilation of Air Pollutant Emission Factors
contains no crushing, drying, and sizing information speci-
fically for coal. There are emission factors for primary
and secondary crushing or rock and stone products, these
given as 0.5 lb/ton (0.025 percent) and 1.5 lb/ton (0.075
percent) respectively. Although moisture content is un-
available, one can assume if these values are compared to
lignite crushing, the lignite emissions would tend to be
much lower, due to the high moisture content (37 percent) of
the mined lignite.
Amounts of coal dust lost during handling, crushing,
5
and blending of coals may range from 0.01 to 0.1 percent.
The portion of losses entering the atmosphere is not avail-
able. If it is assumed that one fourth of the total dust
lost enters the atmosphere, particulate air pollutants range
from 0.0025 to 0.025 percent (0.05 to 0.5 lb/ton of coal).
The following uncontrolled emissions are estimated from
the identified sources, based on a total loss of 0.020
percent and relative emission rates for the individual pro-
cesses similar to those for rock crushing:
B-6
-------
Source
Uncontrolled particulates,
% of lignite lb/ton of
processed lignite processed
Primary crusher
Secondary crusher
Screening
Coal piles and blending
Transfer points (total)
Total
0.001
0.003
0.005
0.005
0.006
0.020
0.02
0.06
0.10
0.10
0.12
0.40
The gas stream leaving the Lurgi gasifier contains H-S
6 7
concentrations of 0.32 to 0.45 percent. Various processes
(Stretford, Rectisol, Claus) remove hydrogen sulfide from
the gas stream. Roughly, 94 to 99 percent of the sulfur
that enters the unit is removed and produces high quality
elemental sulfur. The effluent stream from the sulfur
recovery unit contains 770 ppm of S as l^S.6
The sulfur recovery effluent stream must be either
scrubbed, incinerated, or fed through the boiler. In a
plant employing the Lurgi process to produce 250 MM scf of
synthetic natural gas, the incinerated flue gas composition
contains about 295 ppm SC>2 and 58 ppm NOx.6 Flue gas desul-
furization may be required in this example, since the esti-
mated SC>2 level of 1.8 lb/MM BTU is above the allowable
regulations.
The power requirements for a 250 MM scfd gasification
complex are usually produced within the facility. The flue
gas composition coming from a high BTU gas-fired power plant
g
stack includes about 74 ppm SO_ and 128 ppm NO . This
.« X
boiler accounts for 2700 MM BTU/hr, thus producing emission
levels of 0.16 lb of SO- per MM BTU, and 0.2 lb of NO per
£ X
MM BTU. However, emission factors for lignite-fired boilers
are 1.4-1.5 lb/106 BTU for S02 and 0.5-0.7 lb/106 BTU for
NO.
B-7
-------
Hydrocarbons are emitted to the atmosphere due to
incomplete combustion and from leaks in hydrocarbon by-
product and storage. The rate of uncontrolled hydrocarbon
emissions is 0.007 lb/MM BTU measured as methane.^ With
good engineering design, most of the storage tank and by-
product hydrocarbon emissions - can be captured and inciner-
ated or fed into the boiler.
The fuel properties for the fuel in the referenced
report^ are compared to the North Dakota fuel analyses to
obtain emission factors using lignite:
FUEL CHARACTERISTICS
Component
Sulfur, %
Moisture, %
Ash, %
HHV, BTU
Reference
0.95
16. 5
17.3
8872.0
North Dakota lignite
0.6
37.0
8.0
6500.0
1,2
0 6 88 72
S0~ emission conversion factor = ' _ x ,cnn x 0.862
* U. y j 6b0l)
887 2
N0x emission conversion factor = 650Q = 1.36
Figures 4 and 5 display uncontrolled emissions from a
250 MM scfd lignite gasification plant producing low and
high heating value gas using the gas conversion factors.
B-8
-------
0.12 lb PART/TON LIGNITE
lb PART/TON LIGNITE
0.02 lb PART/TON LIGNITE
0.06 lb PART/TON LIGNITE
CONVEYOR
TRUCK OR RAIL
CONVEYOR
CONVEYOR
ELEMENTAL
SULFUR
COAL
LOCK
0.007 lb ch4/mm BTU.
, a.
DIRTY GAS
LURGI
GASIFIER
CLEAN LOW BTU GAS
TO TURBINE OR OTHER SOURCE
664 ppm H,S EFFLUENT
^ GAS
.OXYGEN
STEAM
ASH
INCINERATOR
SULFUR
RECOVERY
UNIT
COAL
PILES
FOR
BLENDING
LIGNITE
MINE
SIZING
SCREENS
PRIMARY
CRUSHERS
SECONDARY
CRUSHER
(
254 ppm SO2
79 ppm NO
EXHAUST
GAS
Figure 4. Uncontrolled emissions from Lurgi low BTU gasification process.
-------
0.12 lb PART/TON LIGNITE
0.1 lb PART/TON LIGNITE
0.02 lb PART/TON LIGNITE
D.06 lb PART/TON LIGNITE
COAL
PILES
FOR
BLENDING
SIZING
SCREENS
TRUCK OR RAIL
CONVEYOR
SECONOARY
CRUSHER
CONVEYOR
CONVEYOR
PRIMARY
CRUSHERS
LIGNITE
MINE
FLUE GAS
174 ppra NO
POWER AND
STEAM
PLANT
COAL
LOCK
0.007 lb CH./MM BTU
CLEAN GAS
SYNTHETIC
NATURAL GAS
SHIFT CONVERSION
C0+H,0—^H-+C0-
DIRTY GAS
DIRTY GAS
ETHANATI0 _ CLEAN GAS
LURGI
GASIFIER
GAS
JRIFICATIC
PIPELINE GAS
DIRTY GAS
EFFLUENT
STREAM
.OXYGEN
STEAM
SULFUR
RECOVERY
UNIT
ASH
FLUE GAS INCINERATOR
664 ppm H_S
254 ppm SO.
79 ppm NO
ELEMENTAL SULFUR
Figure 5. Uncontrolled emissions from Lurgi high BTU gasification process
-------
REFERENCES
1. Steam: Its Generation and Use, 5:11, The Babcock and
Wilcox Company.
2. Personal Communication with Ken Axetell, May, 1975.
3. Control of Pollutants in Fossil Fuel Conversion/Treat-
ment Processes, ESSO Research and Engineering Company,
U.S. EPA Control Systems Division, Contract No. 68-02-
0629, Phase II, 1973.
4. Compilation of Air Pollutant Emission Factors, Second
Edition, U.S. EPA, Office of Air Programs Publication
AP-42, 1973.
5. Evaluation of Process Alternatives to Improve Control
of Air Pollution from Production of Coke, Battelle
Memorial Institute NAPCA Contract No. PH22-68-65, 1970
6. Evaluation of Pollution Control in Fossil Fuel Con-
version Processes, Exxon Research and Engineering
Company U.S. EPA Control Systems Laboratory, Contract
No. 68-02-0629, ROAP No. 21ADD-023, 1974.
7. Cost Estimate for a 500 Billion BTU/Day Pipeline Gas
Plant Via Hydrogasification of Lignite, Office of Coal
Research, Department of the Interior, Contract No. 14-
01-0001-381, Report No. 22, Undated.
B-ll
-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completing)
1. HI t'ORl NO 2.
EPA-908/1-76-009
A. 1 1 7 Lt. AND SUB! 17 l.t
North Dakota Air Quality Maintenance
Area Analysis
3. RECIPIENT'S ACCESSIOf+NO.
5. REPORT DATE
June 1976
6. PERFORMING ORGANIZATION CODE
7. AUTHOR(S)
8. PERFORMING ORGANIZATION REPORT NO.
9. PL Hh ORMING ORGANIZATION NAME AND ADDRESS
PEDCo-Environmental Specialists, Inc.
Suite 13, Atkinson Square
Cincinnati, Ohio 45246
10. PROGRAM ELEMENT NO.
11. CONTRACT/GRANT NO.
68-02-1375
Task Order 19
12. SPONSORING AGENCY NAME AND ADDRESS
U. S. Environmental Protection Agency
Region VIII
1860 Lincoln Street
Denver, Colorado 80295
13. TYPE OF REPORT AND PERIOD COVERED
Final
14. SPONSORING AGENCY CODE
15. SUPPLEMENTARY NOTES
16. ABSTRACT
This report contains air pollutant emissions estimates, air quality data and
dispersion modeling for the base year (present) in AQMA counties in the State
of North Dakota. Projections of emissions and air quality (using dispersion
modeling) are made for 1980 and 1985. The adequacy of the existing North
Dakota State Implementation Plan to provide for the attainment and maintenance
of the National Ambient Air Quality Standards is discussed.
17. KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b. IDENTIFlERS/OPEN ENDED TERMS
c. COSATI Held/Group
Fuel Combustion
Emissions
Mobile Sources
Stationary Sources
Air Quality Data
Dispersion Modeling
Projections
Air Quality Maintenance
Analysis
13. DISTRIBUTION STATEMENT
Unlimited
19. SECURITY CLASS (This Report)
21. NO. OF PAGES
97
20. SECURITY CLASS (This page)
22. PRICE
EPA Form 2220-1 (9-73)
------- |