-------
tial and commercial-institutional fuel use temporal distribution was
the same. The most extensive data base consisted of hourly natural gas
flow data from the LaClede Gas Company. The analysis of this data pro-
duced a "BTU demand" model to estimate space heating and appliance demands
for fuel consumption over time.
Data-Fuel Usage -
Electrical Demand - Since overall electrical system data was not
available in a form that would allow analysis by grid area and service
type, it was decided to obtain base data from three substations selected
because they served predominantly residential, commercial and industrial
areas respectively. The data obtained was in the form of circular
charts which were digitized. Analysis of the data revealed gaps in the
data and internal inconsistencies which could not be resolved with the
available data. For these reasons, the data proved to be of little value
to the project.
Natural Gas Flows - Hourly gas flows in MCF for the LaClede, St.
Louis gas system were obtained in tabular form for January through Novem-
ber of 1974 and October, November and December of 1973. This was converted
to machine readable form and used extensively in the analysis procedure
described below to develop patterns of usage by various parametric classi-
fications.
Further, the LaClede Gas Company cooperated in supplying a copy of
their billing history for the period specified on some twenty magnetic
tapes to provide both the monthly consumption and the customer classifica-
tions.
16
-------
Meteorology - The hourly gas flow data obtained from LaClede was
accompanied by concurrent measurement of the following meteorological
parameters taken at the LaClede installation at 3950 Forest Park:
radiation in Langley's, wind speed in miles per hour, wind direction and
ambient temperature in degrees Fahrenheit.
Analysis Procedure - Fuel consumption data was summed over each day, mid-
night to midnight. The daily consumption was plotted against the average
daily temperature to investigate temperature dependence characteristics.
The resultant plot is indicated in Figure 2.2. This dependence is
apparent from the plot and is strongly inverse linear below about 68°F:
Above this temperature the usage plot indicates an independence from
temperature effects.
In plotting hourly averages of consumption over the daily cycle, the
temperature, averaged midnight to midnight, was found to introduce an
artificial boundary condition and thereby a discontinuity around midnight.
In order to eliminate this and provide a more realistic model, a traveling
temperature average of the previous 24 hours was utilized. This was the
average daily temperature used in all subsequent analyses (i.e., for
calculating hourly flow, we must know whether the temperature over the
previous 24 hours was less than 68°F or not).
Figure 2.3 shows the hourly average flow for the full 14-month data
period. The cyclical nature of this consumption is due in part to the
diurnal variation in temperature. The consumption exhibits an inverse
proportionality to the average hourly temperatures. The remaining
variation in the daily flow pattern is attributed to a diurnal con-
sumption pattern.
17
-------
OJ
s.
3
+J
to
Ol
Q.
E
OI
1
^>
r
to
Q
CD
01
fO
S-
OJ
.
>
2
o
LZ
(T3
C3
j^
(O
Q
CM
CM
OJ
3
01
Lu
O
O
O
».
O
,
*
**
^
^
«
^
o
o
o
CO
*
4
^
4
*
"-
,-
I*
>*
-
41
«t
« «
K «
c
c
(
I
r
« «
« «
^
R»
« «
4
D
D
O
« «
^
0
**
-. 2 u
-. <* *
^
2 S
tf
0 «
**
^'*
M«
^
* 3
^ *
* * «
o ^
^
*
^
o
o
o
CO
4
«
4
* B
D «>
« <
D tt 4
C.
j
^
D «
4 >
O
U > *
^B
^
J
<
-*
^
Ci
«
< *_»
o
-« 4
* *
^
c
: <
c.
c
c\
CT
«K
u 03
< «l
L> »
»
0 "
Li at
^ 2
CJ
Li
4 41
u
BD L>
»
4
r
^
o
Ul
8,
m
o
m
a
w
CD
0
^
.
«
o
Ul
111
JCO
CL Q.
C C C
J 0 0 O
£ ^ 43
S- t- S-
CU OJ OJ
to 01 in
CO _Q _O o
000
o
i CM CO -M
LU
II II II
10 <*>
(Q/JOW) MOld A1IVQ
18
-------
CM
CM
CO
o
CJ3
>)
r
S-
o
n:
oo
o:
o o
CO
CM
Ol
vr>
CVi
o
CO
co
o
CO
CO
O
CO
CO
o
CO
CO
ft
CO
o
CO
co
n
CSJ
(H/dDW) M01J
19
-------
In order to isolate these effects, the data having little or no tem-
perature dependence were treated separately. The observations with a
traveling temperature average greater than 68°F were separated and analyzed.
Figure 2.4 shows the daily consumption pattern of this "warm" data. This
usage curve was considered to be the base or background consumption that is
always present due to use of fuel for other than space heating.
Flow equations were developed for the two cases: for the baseline
flow on warm days (>68°F); and the "cold" flow on days below 68°F where
the "cold" flow is equal to the total observed flow minus the average
warm day flow. These equations are:
FT(hr) = FC(hr, T, R, WS) + FW(hr) For All Temperatures <_68°F
where the total gas flow (FT) by hour is a function of the cold data
variation (FC) by hour (hr), temperature (T), radiation (R), wind speed
(WS) and of the warm data variation (FVJ) by hour. Thus, it can be seen
that for all periods of average temperature greater than 68°F, this reduces
to:
FT(hr) = FW(hr) For All Temperatures > 68°F.
Figure 2.5 shows the temporal pattern for the cold data. In order
to estimate FC from this, the background consumption FW shown in Figure
2.4 must be subtracted.
The adjusted data is well described by a linear function of temper-
ature, wind speed and radiation. The results of a stepwise multiple linear
regression of the data indicated an excellent fit for the equation:
20
-------
(XI
(XI
c
re
.c:
s-
O)
4->
re
01
CO
3
+J
to
s_
O)
Q.
O)
O)
re
(1)
3
O
CM
s_
O
o
ro
s-
3
O
O)
CT1
CO
S_
a>
a:
CO
oo
a:
o o
i 3:
CM
cu
S-
3
CT)
CM
O
O
O
ri
CTl
O
O
CM
A
00
o
o
o
o
O
O
CO
(H/JDW) MOU SV9
21
-------
GO
c/i
O)
(O
cu
O.
o o
CM
o
o
o
o
o
10
CM
o
o
CM
ft
o
CM
O
o
00
*l
CO
o
o
(H/JDW) M01J SV9 AlcinOH
22
-------
FC = 3.6798 x 104 - 5.4507 x 102 T + 1.0798 x TO2 R + 97.873 WS
where:
FC = total MSCF/hour attributed to heating use
T = temperature in °F
R = radiation in percent of two Langley's/min*
WS = wind speed in mph
The above regression equation would indicate that space heating demands
increased with an increase in the intensity of solar radiation. This is
contrary to the usual concept of the impact of solar radiation on space
heating. Further investigation indicated that the highest recorded R
values were in the autumn, and that the summer months contained a dis-
proportionate number of low values, presumably due to clouds. In order
to remove this potential problem in predicting fuel usage for a particular
day or hour, the regression was repeated without a solar radiation term.
The results are shown in Table 2.3 and the equation is:
FC = 3.4772 x 104 - 5.0895 x 102 T + 1.0478 x 102 WS
where the variables are the same as defined previously. These results
are similar to those obtained by the LaClede Gas Company in the develop-
ment of a model for different purposes2. Integrated over the entire
* The mean intensity of the solar radiation received at the boundary of
p
the atmosphere ranges from 2.007 to 1.877 calories per cm per minute
(Langley's). Two Langley's/minute thus is the maximum radiation that
could be received. The metered values used in this study thus represent
a percent of this figure.
23
-------
>
o
o
in
cr
o
O O
O O
O O
CD
O
a:
Q_
o in
ra ru
t t-
CC to
_i ru o-
« KI «>
> cr ru
it. -o **
r- ru
r- cvi
03 r-
in in
cr -*
9 o
to in
in -o
Ul
o
o
OL
a.
UJ
n
UJ
o
cc,
o.
UJ
o
X
UJ
o
CD
tr
o
u.
to
CO
o in
»» e\j
oo m
ru r~
CO O-
CO >0
in o^
cc
cc
o
>-
n
o- ru
o co
o -o
ui
_i
31
O -
IU CC
O
«
cc
t
CD
O
>
o
0}
U.
U.
o
ir-
CO
z
!iJ
O
CO
CVJ
Ol
ui
3ET O.
X IU
<. o
X
cr
2: o
O
o a:
_j a.
o z:
«J t-t
QC
O
O
CO
I
CC
ui
o
o
z:
tCl
<
t-l
cc
x:
UJ
a;
O
Z
o
o
CD
<
UJ
oc
=>
o
to
O
z:
L>
»-
U-
o
o
«o
co in ro
CO
o-
ta. -..-
to
o
cc
O-
IU
=> -o »»
_i -o cr
< CO CO
> -o rvi
« CM
in
o
UJ
I
ru »-
UJ
o
z
o
CO
I
to
co o- u-> ru rw -o
M in r-
co 10 cr r-
»
(XI CO 9
«>»«> O
r-in
9
iu
co t-
CO «->
UJ CC UJ
cc ac a cc
z> o ac a
o iu cc o
«o a: iu u>
UI
O
or
o a.
o z co
co » 31.
ac
o
o
24
-------
fourteen months of available data from LaClede (i.e. 9780 hours of
observations), the total predicted cold flow using this equation is 7.1697
x 10 MSCF. Over the same period the non-space heating or "baseline" flow
accounted for 6.7027 x 107 MSCF (approximately 48 percent of the gas flow).
The total 13.8724 x 107 MSCF is within 3% of the total LaClede data.
(Total observed equals 13.5145 x 10 MSCF.)
The baseline flow is most appropriately represented as a composite
average for the time periods when the traveling average 24-hour tempera-
ture exceeds 68°F. As seen in Table 2.4 the above fourteen month base-
line flow reduced to an average hourly baseline flow is 6.854 MSCF. This
table summarizes the hourly baseline flow by the hour and the proportion-
ality factor (PF) by hour related to the average flow. Hourly baseline
gas flow can then be approximated by:
Hourly Baseline = yearly flow x 0.4832 x PF (EQN 2.1)
Flow (HBF) 57515
When the temperature is above 68°F, this equation applies. Hourly
natural gas flow is then (where FC has been normalized by the yearly
predicted flow):
Hourly Gas Flow = yearly flow x [JL849 x 10"4 - 7.0986 x 10'6T +
1.4614 x 10"6 WS + 0.4832 x PF"| (EQN 2.2)
87615J
where the temperature is less than or equal to 68°F.
As these equations represent portions of total yearly quantities,
these equations apply equally well to bottled (LP) gas which has the same
usage characteristics (i.e., its use for cooking, water heating, etc.,
in addition to space heating).
25
-------
Table 2.4 Average Hourly Baseline Gas Flows for the LaClede
Gas Company (temperature >68°F).
Composite
Hour Average Flow Proportionality
(MSCF) Factor
1
2
3
4
5
6
7
8
9
10
n
12
13
14
15
16
17
18
19
20
21
22
23
24 (midnight)
Average hourly flow
5727
5490
5339
5301
5262
5550
6312
7378
7899
7948
8387
8214
8062
7892
7640
7360
7245
7267
7315
7094
6769
6632
6433
5971
6854
0.84
0.80
0.78
0.77
0.77
0.81
0.92
1.08
1.15
1.16
1.22
1.20
1.18
1.15
1.11
1.07
1.06
1.06
1.07
1.04
0.99
0.97
0.94
0.87
1.00
26
-------
For other fuels such as fuel oil, coal or coke and wood, the usage
characteristics differ in that these fuels are almost totally used for
space heating. The demands therefore depends only on temperature and wind
speed according to the previously developed equation, without any depen-
dency on a baseline value. Thus:
T £ 68°F
Hourly Demand = yearly total x [4.8499 x 10"4 - 7.0986 (EQN 2.3)
10"6 T + 1.4614 x 10"6 WS]
and for T > 68°F the hourly demand is equal to zero.
27
-------
EVAPORATIVE HYDROCARBON LOSSES
Evaporative losses of hydrocarbons to the atmosphere from dry
cleaning plants, surface coating operations and gasoline marketing were
considered in this study. Spatial allocation of these emissions were
based upon population and commercial land use densities. The temporal
allocations were based on the regular 8:00 a.m. to 5:00 p.m. workday
or upon the diurnal traffic cycle observed in St. Louis.
Evaporative Hydrocarbon Emissions Estimates
Dry Cleaning Emissions
The dry cleaning industry uses two basic types of organic solvents
in the cleaning of clothes. These are petroleum solvents and chlorinated
synthetic solvents (perchloroethylene). Volatile hydrocarbon emissions
occur mainly from the hot air tumbler process of drying the solvent
soaked garments.
According to the St. Louis County dry cleaning plant survey, 89 per-
cent of dry cleaning establishments use perchloroethylene solvent and
63 percent of all clothes are cleaned with this material. Due to the
rising cost of petroleum and increased cost of synthetic solvent (per-
chloroethylene costs approximately $15/gal.), emission control and
recovery equipment has been greatly improved and much more widely used
in recent years. Based on the St. Louis County data, 0.53 pounds of
hydrocarbons from this type of source are emitted per capita per year.
It was assumed that this per capita figure was applicable for the entire
AQCR. The emissions occur during normal working hours and are spatially
distributed according to the percentage of commercial land use per grid.
28
-------
A county-by-county breakdown of evaporative hydrocarbon emissions is
summarized in Table 3.1.
Thus the yearly county data is allotted to the grid areas by a pro-
portion of commercial land in the grid to the total county value. This is
then distributed evenly over the nine hour business day from 8:00 a.m.
to 5:00 p.m. in which cleaning is done over the five workdays a
week.
Paint Emissions
Hydrocarbon emissions from the evaporation of solvents from paint
3
have been found to be significant in previous studies. For residential
and commercial-institutional area source emissions the type of paints
utilized are referred to as "trade-sale" (as distinct from industrial).
Trade-sale paints are distributed through retail stores to the general
public. The most accurate available statistics for estimating the sales
of these paints in AQCR 70 consists of those developed by the National
Paint Coating Association (NPCA) in Washington, D.C. These statistics
indicate that 21.9 and 10.2 percent of the nationally sold trade-sale
paints are sold in the East North Central Region (Ohio, Indiana, Illinois,
Michigan, Wisconsin) and the West North Central Region (Minnesota, Iowa,
Missouri, Kansas, North Dakota, South Dakota, Nebraska), respectively.
The NPCA statistics indicate that in 1973,424,000,000 gallons of trade-
sale paints were sold. Trade-sale paint sales in AQCR 70 were therefore
estimated by the following formulas:
trade-sale paints _ national total x 0.219
sold in Illinois pop. of (Ohio + Ind. +111. + Mich. + Wise.)
=2.32 gallons per capita
29
-------
to
CD
O
S-
^
o
oo
CD
r
S-
o
Q.
to
LU
O
s_
t-
co
r-.
i
S-
o
>)
o
c
CL>
c
HH
CO
o
CO
CO
'E
LU
c
o
JO
S-
tO
0
o
S-
TJ
IE
CO
JO
fO
1
/fcj.i.0 SLnoi }$
>-
z
0
o
SLnon }$
CQ 1 -ipijo * n c
L TJ 4 o
UOSJ8.Lj.9p
Ml 1 \1 M n 1 1
L^ ^
LunossLw
1JOT fill 1 IJ^PM
T . T 1 1
i i p i r\ * T C
. L \J T J
t|d[opuey
aouuo^
uos.Lpew
HOT u i i o
UW4U J- L J
puog
S LOU LI U
^_^
i-
rt3
CTi OO
co co r>^
LO LO 0
i r- CO
tO IO CM
0 0 i
* <* 00
LO LO O
CO CO 1-^
CO CO tO
LO LO O
to to co
CO CO t*^
CO CO to
LO LO O
CM CM LO
r**^ r*** ^^
LO LO r
r-
CO OO ^D
OO OO <*O
r*"* r*1*" ^~
I
O O 0
i i CM
r- r CM
r r CM
LO tO CM
I*"1* r""""» *^^
o o
00 00 tO
00 00 to
CTi CTi CTt
i
i r CM
tO IO CM
z
(-H
^""^ K~
i LU
cnl -i- Qi
C <4- CC
r~ ^*
-l-> CO -E
Ol -^ CO LU
i_ - <:
to 0 00 r >i
CO O S-
CD 1 Q
i CM
LO r~^
oo
0 .
LO to
CVI ^~
t . CT*
CM LO
i
O 1 .
CO 1^
to * o
to cr>
CO
00 LO
^~
^.
to
CM
CO
CO
+-> Ol
r- C
Q. !-
CO 4->
O CO
^ o
JO O
^
-------
trade sale paints national total x 0.102
sold in Missouri = pop. of (Minn.+Iowa + Mo.+Ks. + N.D.+S.D. +Neb.)
=2.70 gallons per capita
The Technical Director of NPCA, Mr. Ray Conner, estimated that
nationally approximately 70 percent of all trade-sale paints were water
based.4 The Boston hydrocarbon study indicated that water based paints
contained approximately 3.5 percent by weight volatile hydrocarbons.
Mr. Conner estimated that non-water based paints contained approximately
50 percent volatile hydrocarbons. For this study it was therefore assumed
that each gallon of trade-sale paint contained approximately 17.5 percent
volatile hydrocarbons, by volume or 1.14 pounds of hydrocarbons per
gallon of paint. The density of these volatile hydrocarbons was assumed
to be 6.5 pounds mass per gallon (the approximate weight of mineral
spirits).
Gasoline Marketing Emissions--
Evaporative losses of hydrocarbons in the marketing of gasoline at
local service stations occur in two waysfilling losses from underground
storage tanks and filling losses and spillage from the filling of auto-
mobile tanks.
State totals of gasoline sales were obtained from the Departments of
Revenue for the states of Illinois and Missouri. County figures were
derived by allocating sales per service station to the number of stations
per county as found in the 1972 County Business Patterns.
It was established due to their age that the majority of underground
storage tanks are filled by the gravity drop splash fill method with no
emission control equipment.5 These tanks are filled during normal business
hours. Emissions estimates based on this produce predictions of
31
-------
evaporative losses from gasoline marketing approximately 15 percent higher
than those derived in the St. Louis County Emissions inventory for gaso-
line marketing. This agreement is good considering the differences in
methodologies.
Spatial Allocations
Emissions from dry cleaning plants and from gasoline marketing opera-
tions were apportioned on the basis of the commercial land use within each
grid as derived from the EWGCC commercial land use information. The
formula utilized was:
Grid Annual Dry Cleaning area of land in grid under commercial usage (CLUG)
or Gasoline Marketing = tota] CQunt land area under commercial usage
Emissions
x total county dry cleaning or gasoline
marketing emissions
Surface coating emissions from the evaporation of solvents from
trade-sale paints were apportioned on the basis of population, as follows:
Total Grid Surface _ Grid Population (GPOP) Total County Surface
Coating Emissions = County Population (GPOP) Coating Emissions
Temporal Allocations
The temporal allocations of hydrocarbon emissions are most appro-
priately based upon the 8:00 a.m. to 5:00 p.m. workday for all emissions
except those produced by the filling of automobile gasoline tanks. The
hydrocarbon emissions from automobile tank filling account for fifty
percent of the gasoline marketing emissions.
The other fifty percent of the gasoline marketing emissions are allo-
cated on the basis of being proportional to the measured diurnal traffic
cycle in St. Louis. Figure 3.1 shows this cycle for weekdays, Saturday
and Sunday. Table 3.2 shows the weighting factors to be used for this
apportionment. In addition to these factors, an allowance for the
32
-------
0.06
_i
<
LLJ
UJ
u.
O
O
I-
o
0.04
0.02
1 I ' I 'I
WEEKEND MEASURED, BROADWAY AND LOCUST
0.10
< 008
O
01
UJ
* °-06
O
CC
UJ
>
I °-04
I-
o
0.0?
WEEKDAY
10 15
HOUR OF DAY
20
Figure 3.1 DIURNAL EMISSION PATTERNS FOR ST. LOUIS.
33
-------
Table 3.2 Temporal Allocation Factors for the Filling of" Auto-
mobile Gasoline Tanks.
Hour
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
Weekdays
6.45 x 10"5
5.82 x ID'S
1.62 x 10"5.
3.23 x TO'6
3.23 x ID'S
1.62 x 10-5
9.70 x 10-b
2.20 x TO'4
1.94 x 10'4
1.29 x 10";
1.29 x 10'}
1.46 x 10";
1.46 x 10"4
1.46 x 10'4
1.52 x 10'4
2.43 x ID"4
2.72 x 10'4
2.85 x 1Q-4
2.10 x 10'4
2.07 x 10'4
1.62 x 10'4
1.23 x TO'4
9.70 x ID'5
Saturday
9.70 x ID'5.
9.70 x 10" f.
7.12 x 10"5
2.59 x TO"5
2.59 x 10~5
3.23 x ID'S
4.53 x 10"5
6.79 x 10"5
9.70 x 10'5
1.07 x 10'4
1.23 x 10'4
1.29 x 10"4
1.29 x TO'4
1.16 x 10'4
1.13 x 10'4
1.13 x TO'4
1.20 x lO'4
1.10 x 10-J
9.06 x 10"*
8.73 x 10"5.
9.70 x 10"5
1.10 x 10'4
1.10 x 10"4
Sunday
9.70 x 10~5
8.09 x 10"5
5.82 x 10'5
1.62 x 10'5
1.29 x TO'5
1.62 x 10-5
1.94 x 10"5
2.59 x ID'5
3.23 x lO-5
3.88 x 10"5
5.18 x 10-5
6.47 x 10-5
6.47 x 10"5
7.76 x TO'5
7.76 x 10-5
7.44 x 10"5
7.12 x ID"5
7.12 x TO'5
2.47 x 10-5
6.14 x 10"b
5.82 x 10-5
5.82 x 1Q-5
5.18 x 10"5
34
-------
variation in seasonal driving habits is necessary. Figure 3.2 shows the
monthly gasoline sales for the state of Missouri in 1973 and 1974. Similar
data is available for Illinois. As shown, there are marked differences
between summer and winter gasoline sales. The monthly sales factor shown
in the figure is based upon the average of the 1973 and 1974 values and
the relative fraction of sales for a composite average month.
The temporal allocation is then:
(from 8:00 a.m. to 5:00 p.m.) (EQN 3.1)
Evaporative HC_ -, *
Emissions = 0340" [annual dry cleaning emissions +
annual surface coating emissions +
(0.5 x annual gasoline marketing
emissions x monthly sales factor)]+
(automobile allocation factor x
0.5 x annual gasoline marketing
emissions x monthly sales factor)
(from 5:00 p.m. to 8:00 a.m.) (EQN 3.2)
Evaporative HC_ automobile allocation factor x 0.5 annual
Emissions = gasoline marketing emissions x monthly
sales factor
* This number is based on a workday from 8:00 a.m. to 5:00 p.m., five
days a week and 52 weeks per year.
35
-------
260"
250--
2-Year Average
(1Q6 Gal.)
Jan.
Feb.
Mar.
Apr.
May
June
July
Aug.
Sept.
Oct.
Nov.
Dec.
AVERAGE
Monthly Sales
Factor
240--
230--
220--
210 ..
200 --
190 "
4-
JAN
FEB
MAR
APR MAY
JUNE JULY
AUG
SEP
OCT
Figure 3.2 State of Missouri Gasoline Sales.
36
NOV
DEC
-------
SOLID WASTE DISPOSAL
Emissions from solid waste disposal through open burning and incinera-
tion were considered in this study. Conversations with the St. Louis
6
County Air Pollution Control Board and with the Missouri Air Conservation
Commission indicated that open burning was banned from the populous areas
of AQCR 70. It is allowed by law only where no public or commercial
refuse collection service is available and in places where population
density is less than 100 dwelling units or less per square mile. The
emissions from open burning are considered insignificant in this study.
Information sources on solid waste incineration included a nationwide
study and the emission inventory prepared by St. Louis County. The latter
source of information appeared to be the most reliable as it included
actual emissions data other than nationwide statistics. This study indi-
cated that residential incineration was negligible. According to the
Pollution Control Regulations for the St. Louis Metropolitan Area, only
multiple chamber incinerators may be used and must not exceed 0.3 grains
of particulate per dry standard cubic foot of exhaust gas. Since the
emission standards are stringent and the cost of multiple chambered units
high, residential incineration emissions were considered insignificant in
this study. The only significant source of incineration pertinent to
this study was the commercial-institutional area source category.
Treating the St. Louis County figures as being representative of the
entire AQCR, Table 4.1 summarizes the emissions estimates from solid
waste disposal by county.
The spatial and temporal allocation procedures are similar to those
for emissions from the dry cleaning process; that is, in proportion to
the commercial-institutional land use in each grid square, and an 8:00 a.m.
to 5:00 p.m. workday.
37
-------
Table 4.1 Solid Waste Commercial-Institutional Incineration
Emissions (tons/year) in AQCR 70 for 1973.
County
Missouri
Franklin
Jefferson
St. Charles
St. Louis
St. Louis City
Illinois
Bond
Clinton
Madison
Monroe
Randolph
St. Clair
Washington
Part.
3.21
6.05
5.47
50.00
29.86
0.74
1.52
13.30
0.95
1.63
15.09
0.74
SOX
0.38
0.73
0.66
6.00
3.58
0.09
0.18
1.60
0.11
0.20
1.81
0.09
NOX
0.58
1.09
0.98
9.00
5.38
0.13
0.27
2.40
0.17
0.29
2.72
0.13
HC
2.57
4.84
4.37
40.00
23.89
0.59
1.22
10.68
0.76
1.30
12.07
0.59
CO
5.13
9.67
8.75
80.00
47.78
1.18
2.44
21.37
1.51
2.61
24.14
1.18
38
-------
STRUCTURAL FIRES
Data was not available for wildfires and forest fires. The available
data on structural fires is such that a high degree of reliability cannot
be placed on the emissions estimates. The available data included:
1) the number of fires and dollar damage on a monthly basis for Illinois,
2) statewide statistics on the total number of fires (10,000) doing
9
$45,500,000 worth of damage in 1974; and 3) nationwide statistics indi-
cating that 40 percent of the typical structure is consumed in a fire and
that the average structure contains approximately 17 tons of combustible
10
material.
For Missouri the statewide figures were disaggregated to the county
level on the basis of population. Using the emission factors in AP-42 for
open burning of municipal solid waste, the emission estimates were derived.
These values are summarized in Table 5.1
The spatial apportionment is most appropriately conducted on the
basis of the number of housing units per tract:
Grid = no. of housing units in grid x County
Emissions no. of housing units in county Emissions
The temporal distribution of these fires is random. The apportion-
ment figures can be utilized on an annual basis as a reasonable approxima-
tion. However, for the purpose of the calibration of dispersion models,
the only appropriate approach is for the modeler to be aware of whether or
not a structural fire has occurred that may affect short-term model
results. The fire can be treated as a ground level point source. An
estimate of emissions can be included in the grid area based upon 40 per-
cent of the 17 tons of combustible material being consumed in a four hour
period using the emission factors for incinerator without controls from AP-42.
39
-------
Table 5.1 Estimated Annual 1973 Emissions (tons/year) from
Structural Fires in AQCR 70.
County
Illinois
Bond
Clinton
Madison
Monroe
Randolph
St. Clair
Washington
Missouri
Franklin
Jefferson
St. Charles
St. Louis
St. Louis City
Part.
0.7
2.1
49.7
2.2
3.8
56.2
1.1
7.0
13.2
12.0
109.3
65.3
sox
<0.1
0.1
3.1
0.1
0.2
3.5
0.1
0.4
0.8
0.8
6.8
4.1
CO
3.8
11.0
264.2
11.6
19.9
298.5
6.1
37.3
70.2
63.6
580.9
347.1
HC
1.3
3.9
93.2
4.1
7.0
105.4
2.1
13.2
24.8
22.4
205.0
122.5
NOX
0.3
0.8
18.7
0.8
1.4
21.1
0.4
2.6
5.0
4.5
41.0
24.5
-------
SUMMARY
General
The Regional Air Pollution Study has unique requirements of its
component emission inventories. Generally it requires degrees of spatial
and temporal resolution that have not been previously achieved. This
document brings together data on fuel usage, land use, sales of gasoline
and paints, the use of dry cleaning fluids, solid waste disposal and
uncontrolled fires for the St. Louis region and develops methodologies
for estimating the pollutant emissions from stationary residential and
commercial-institutional area sources on an hour-by-hour basis for com-
ponents of a spatial grid system developed for RAPS.
The methodologies presented in this document are a series of sub-
elements integrated into a single system for deriving the required
emission data. This system seeks to provide the best emission estimates
possible from the available data. In order to provide the RAPS with as
much flexibility as possible to meet the multiple and varied demands upon
it, the inventory is presented as direct statements of weight of pollutant
emitted by this class of source as a function of location for every hour.
Specific Results
Space Heating - The emissions from space heating were based upon the
emission factors utilized by the National Emissions Data System and from
AP-42 for each fuel. The distribution within the RAPS grid system was
determined by allocating total county fuel use for residential, and, in
a separate calculation, commercial-institutional fuel usage to each grid
area in proportion to the number of units using that fuel in the grid
area. The temporal fuel use variation and variation with meteorological
41
-------
parameters was established by statistically analyzing the detailed data
available on gas use. This analysis identified a base component of
usage which is largely independent of meteorological influence and has
a distinct relationship with the time of day. The remaining component was
shown to be strongly affected by the ambient temperature and wind velocity.
Evaporative Emissions
The major components of residential and commercial-institutional area
sources of evaporated hydrocarbons were identified as surface coating
(primarily painting), gasoline handling, and dry cleaning. Characteristic
emissions on a per capita basis were established and the emissions deter-
mined for each grid area by allocating the total projected county emissions
in proportion to commercial land use in the grid area for gasoline handling
and dry cleaning, and in proportion to population for surface coating.
Structural Fires and Solid Waste Disposal
Emissions from structural fires were difficult to quantify; however,
average data was used and total estimated county emissions were allocated
to grid areas based on the number of homes in the grid. However, it was
realized that significant structural fires would be best handled as a point
source.
Emissions from the disposal of solid wastes is largely restricted to
commercial-institutional enterprises and large incinerators because of the
restrictive air pollution regulations. Thus the NEDS county emissions were
allocated to grid areas in proportion to the commercial-institutional land
use in the grids.
42
-------
REFERENCES
1. Personal communication between Mr. Don McQueen of Environmental
Science and Engineering, Inc. and representatives of the LaClede
Gas Company, May, 1975.
2. Ryan, Richard R., Gas Sendout Forecasting with Weather Sensitive
Loads, LaClede Gas Company, May 1974.
3. Hydrocarbon Emission Sources in the Metropolitan Boston Intrastate
Air Quality Control Region, Vol. 1, prepared for the U.S. Environ-
mental Protection Agency by GCA/Technology Division, May, 1974.
4. Personal communication between Mr. Robert Hoi den of Environmental
Science and Engineering, Inc. and Mr. Ray Conner of the National
Paint Coating Association, May, 1975.
5. Personal communication between Mr. Don McQueen of Environmental
Science and Engineering, Inc. and the Mobil Oil Distributors,
St. Louis, Missouri, May, 1975.
6. Personal communication between Mr. Robert Holden of Environmental
Science and Engineering, Inc. and Mr. Michael L. Farley, Engineer,
Air Pollution Control Branch, Division of Environmental Health
Care Services, March, 1975.
7. Personal communication between Mr. Robert Holden of Environmental
Science and Engineering, Inc. and Mr. Hoffman of the Missouri Air
Conservation Commission, May, 1975.
8. Illinois State Fire Marshal's Office, May, 1975.
9. Personal communication between Mr. Pete Burnette of Environmental
Science and Engineering, Inc. and the Missouri State Fire Marshal,
May, 1975.
10. Derived from data provided by the National Fire Protection Asso-
ciation of Boston, Massachusetts.
43
-------
APPENDIX A
Description of the Annual Emission Inventory Calculation
-------
Emissions from residential space heating were determined by allocating
total county residential fuel usage to each grid base on the number of
homes in the grid using a given fuel type:
SHR1fjfk =
f
NHCi , j
Equations and Input Parameters
Where: SHR-j -j^ = residential space heating emissions
of pollutant _k_
from fuel type j_
in grid number i_
NHGFU-jj = number of homes in grid i using
fuel type j
NHCj,f = number of homes using fuel type j
in the county f in which grid is located
TCFUR-; f = total residential usage of fuel type j
-------
Where: COM-jj^ = emissions from commercial fuel usage
of pollutant k_
from fuel type _j
in grid number i_
CLUG-j = commercial land use (km^) in grid i
CLUCf = commercial land use (km^) in the county
in which grid is located
NHGi = total number of homes in grid i
TCFUCj,f = total conmerical usage of fuel type j
in the county
NCH-: f = is defined above
J >T
NHCTf = total number of homes in county f
Evaporative hydrocarbon emissions for the three processes indicated
below were determined for each grid by allocating total county emissions
from the three processes on the basis shown.
Process Allocation Basis
1) Surface Coating Grid Population
2) Gasoline Handling Commercial Land Use
3) Dry Cleaning Commercial Land Use
GHCEi,-, = CHCE1>f x
GHCEi,2 - CHCE2,f x
GHCEi ,3 = CHCE3,f
A-2
-------
Where: GHCE..- n- = evaporative hydrocarbon emissions
1 »J
in grid i from process j
CHCEj^f = total county f emissions from process j
GPOP-j = population of grid i
CPOPf = population of county in which grid is located
Emissions from structural fires were determined by allocating total
county emissions to each grid based on the number of homes in the grid.
Where: EMSF^ ^ = emissions from structural fires
of pollutant Ik in grid i_
ESFCk f = total county emissions from structural
fires of pollutant k from county in
which grid is located
NHCTf = total number of homes in county
Emissions from the disposal of solid wastes was determined by
allocating total county emissions to each grid based on commercial land use
within that grid.
ESWi k = ESWCk f x CLUGi
lf K'T CLUCf
Where: ESW-j^k = emissions from solid waste disposal
of pollutant k^ in grid 1_
A-3
-------
total county emissions from solid waste
disposal of pollutant k for county in
which grid is located.
A-4
-------
APPENDIX B
Sample Calculation
-------
I. EMISSION FACTORS
Fuel Type
Bituminous Coal (tons/ton)
Domestic
Commercial
Residual Oil (tons/103 gal)
Commercial
Distillate Oil (tons/103 gal)
Domestic
Commercial
Natural Gas (tons/10 cu. ft.)
Domestic
Commercial
Bottled (LP) Gas (tons/103gal)
Domestic
Part
SO,
Pollutant
CO
HC
NO.
0.0100 0.00475 0.0450 0.0100 0.0015
0.0140 0.00475 0.0036 0.0010 0.0046
0.0115 0.1590 0.0020 0.0015 0.0300
0.0050 0.0144 0.0025 0.0015 0.0060
0.0075 0.0144 0.0020 0.0015 0.0300
0.0050 0.0003 0.0100 0.0040 0.0400
0.0050 0.0003 0.0100 0.0040 0.0600
0.00095 7.2xlO"6 0.0010 0.0004 0.0400
B-l
-------
II. COUNTY AND GRID ANNUAL DATA
A. Attributes of County No. 2:*
Population
Commercial Land Use
Number of Homes Using Fuel Type 1 Oil
Total Residential Usage of
Fuel Type
(Table 2.1) (TCFUR)
Total Commercial Usage of
Fuel Type
(Table 2.1) (TCFUC)
Total Evaporative Hydrocarbon
Emissions
(Table 3.1) (CHCE)
Total County Emissions from
Structural Fires Pollutant
(Table 5.1) (ESFC)
Total County Emissions from
Solid Waste Disposal
Pollutant
(Table 4.1) (ESWC)
Total Number of Occupied Homes
in the County
(NHCT)
CPOP
CLUC
1 Oil NHC
2 NG
3 LPG
4 Coal
1
2
3
4
1 Dist-35,000
Residents
2
3
4
1
2
3
1 Part
2 SO,
3 or
4 HC
5 NOV
X
1
2
3
4
5
568100
11.965 km2
16949
186093
2523
3375
13560
26400
16619
18560
0
11100
0
112000
872
3078
151
65 Tons
4
347
123
25
30 Tons
4
48
24
5
= 215479
kK\] units are standard NEDS units (emissions as tons/year, fuel oil as
103 gal/year, natural gas as 10° cu ft/year, LPG as 10^ gal/year, and
coal as tons/year.
B-2
-------
B. Attributes for Grid No. 895, located in the City of St. Louis
which is treated as County No. 2:
Population GPOP 5096
o
Area 1.0 km
Commercial Land Use CLUG 0.15 km2
Number of Homes NHG 1547
Homes Using Fuel Type 1 NHGFU 139
2 1302
3 26
4 41
B-3
-------
III. GRID EMISSION CALCULATIONS FOR POLLUTANT NO. 4 (HYDROCARBONS)
A. Residential Fuel Consumption
1. Fuel Type 1, Fuel Oil
x TCFUR(l) x EF = 139 x 13560(103gal) x
TC3W
0.0015(tons/103gal) = 0.1668 (tons/year)
2. Fuel Type 2, Natural Gas
NHGFU(2) x TCFUR(2) x EF = 1302 x 26,400(106cu.ft.) x
NHC(l) 186093
.0004 (tons/106cu.ft.) = 0.7388 (tons/year)
3. All other Fuel Types similar
B. Commercial and Institutional Fuel Usage
1. Fuel Type 1, Fuel Oil
a. Distillate Oil
CLUG x NHGFU(l) x NHCT x EF x CFT(dist.) =
"ClUf NHG NHC(l)
0.15 x 139 x 215479 x 0.0015 / tons ^ x
11.965 T54T 16949 V103gal;
35000(103gal) = 0.7518 (tons/year)
b. Residual Oil
CLUG x EF x TCFUC (res.) = 0.15 x .0015 (tons/103gal) x
CLUC 11.965
0(103gal) = 0.0 (tons/year)
c. Total Commercial Fuel Oil HC Emissions
Results of "a" + "b" = 0.7518 x 0.0 = 0.7518 (tons/year)
B-4
-------
2. Fuel Type 2, Natural Gas
CLUG x NHGFU(2) x NHCT x EF x TCFUC(2) =
CLUC NHG NHC(2)
0.15 x 1302 x 215479 x .004(tons/106cu.ft.) x
11.965 1547 186093
11100 (106cu.ft.) = 0.5424(tons/year)
3. All other Fuel Types similar to la and 2 above.
C. Emissions from Structural Fires
ESFC(4) x NHG = 123(tons) x 1547 =
NHCT 215479
0.8831(tons/year)
D. Emissions from Solid Waste Disposal
ESWC(4) x CLUG = 24(tons) 0.15 = 0.3009(tons/year)
CLUC 11.965
E. Evaporative Hydrocarbons
1. Surface Coating
CHCE(1) x GPOP = 872(tons) x 5096 = 7.822(tons/year)
CROP 568100
2. Gasoline Handling
CHCE(2) x CLUG = 3078{tons) x 0.15 = 38.5875(tons/year)
CLUC 11.965
3. Dry Cleaning
CHCE(3) x CLUG = 151(tons) x 0.15 = 1.8930(tons/year)
CLUC 11.965
B-5
-------
IV. GRID HOURLY EMISSIONS FOR HC
A. Hourly Characterization
1. Time: 8:00 a.m. = 9:00 a.m.
2. Wind Speed: 15 mph
3. Temperature: 40°F
4. Day: Tuesday
5. Month: February
B. Fuel Oil
[Residential + Commercial & Institutional Fuel Oil Yearly
Emissions] x [4.8499 x 10'4 - 7.0986 x 10'6T + 1.4614 x
10~6WS] = [Results III A.I plus III B.I] x Eqn =
[0.1668 + 0.7518] (tons/year) x [4.8499 x 10'4 - 7.0986 x
10~6 x (40) + 1.4614 x 10-6(15)](yean = (0.9186) x (7.9086 x !Q-4)=
hour
7.2648 x 10'4 tons/hour
C. Natural Gas
[Residential + Commercial & Institutional Natural Gas Yearly
Emissions] x 4.8499 x 10~4 - 7.0986 x 1Q-6T + 1.4614 x 1Q-6WS +
0.483.2 x PF] = [Results III A. 2 and III B.2] x [Eqn with PF
8760
from Table 2.4] = [0.7388 + 0.5424] (t2Dl\ x [4.8499 x 10
v'
~4
7.0986 x 10'6 x (40) + 1.4614 x 10'6 x (15) + 0-483j-.* 1J5]
8760
ari} = (1.2812) x (8.5429 x 10~4) jb= 1.0945 x 10'3 tons/hour
vhburs; nour
D. Other Fuels
1. LP Gas - same as C above
2. Coal - same as B above
B-6
-------
E. Structural Fires
Hourly emissions based on a random distribution may be used.
However, for use in dispersion models these values would be
meaningless. For this source, the modeller must investigate
the nature of any fires. The random distribution is obtained
by dividing the yearly figure by 8760.
F. Solid Waste
[Results from III D] x 4.27 x 1CT4 = 1.28 x 10"4 ^^-
hour
G. Evaporative Hydrocarbons
4.27 x 10~4 x [annual dry cleaning emissions + surface coating
emissions + (0.5 x gasoline marketing emissions x monthly sales
factor*)] + (automobile allocation factor** x 0.5 x gasoline
marketing emissions x monthly sales factor) = 4.27 x 10 x
[1.8930 + 7.822 + (0.5 x 38.5875 x 0.88)] + (1.94 x 10'4 x
0.5 x 38.5875 x 0.88) = 1.14 x lO'2 + 0.33 x 10~2 = 1.47 x
10~2 tons/hour
B-7
-------
TECHNICAL REPORT DATA
(Please read Injunction'* on the icvcrsc bcjore completing)
1 REPORT NO
EPA-45Q/3-75-Q78
3 RECIPIENT'S ACCESSIOWNO.
4. TITLE AND SUBTITLE
Residential and Commercial Area Source
Emission Inventory Methodology for the Regional Air
Pollution Study
5. REPORT DATE
September 1975
6. PERFORMING ORGANIZATION CODE
7 AUTHOR(S)
8. PERFORMING ORGANIZATION REPORT NO.
R. E. Holden and W. E. Zegel
9. PERFORMING ORGANIZATION NAME AND ADDRESS
Environmental Science and Engineering, Inc.
P.O. Box 13454
Gainesville, Florida 32604
10. PROGRAM ELEMENT NO.
11. CONTRACT/GRANT NO
68-02-1003
12. SPONSORING AGENCY NAME AND ADDRESS
U.S. Environmental Protection Agency
Office of Air and Waste Management
Office of Air Quality Planning and Standards
Research Triangle Park, N.C. 27711
13. TYPE OF REPORT AND PERIOD COVERED
Final Report
14. SPONSORING AGENCY CODE
15. SUPPLEMENTARY NOTES
16. ABSTRACT
One of the major objectives of the Regional Air Pollution Study (RAPS) is to
provide data on the emissions of air pollutants, meteorological conditions and
ambient air quality with unprecedented density and resolution as to allow the
testing and development of a spectrum of mathematical models to simulate relation-
ships between emissions of pollutants and air quality. As part of this effort a
methodology for estimating the pollutant emissions from stationary residential
and commercial-institutional area sources on an hour-by-hour basis, and
apportioning them to the RAPS grid system is presented.
17.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b.IDENTIFIERS/OPEN ENDED TERMS
c. COSATI I icld/Group
Regional Air Pollution Study
Emissions
Emission Models
18. DISTRIBUTION STATEMENT
Release Unlimited
19. SECURITY CLAS.S (This Report)
Unclassified
21. NOV
PAGES
2O. SECURITY CLASS (Thispage)
Unclassified
22. PRICE
EPA Form 2220-1 (9-73)
-------
m
D
O
TJ
TJ
O
33
H
C
Z
m
S
TJ
f
O
<
m
3)
|0
O
Tl
O
00
C
CO
rn
5; .. c
.^£s
33
O
H
m
O
H
O
2
^
C
w
i-1
ii
n
>
a
o
!Z!
Z
O
w
2
&
v\
o
oj
-g
v^i
i
O
-J
oo
O
&
Q-
CO
O)
o> ro
Q. '-
- -
CD
CD
Q.
CD
Dl
Q. CD
O
n CD
EH <
CD CD
£=?
» cn
id 3
CD g
_ CD
C
Q.
o
<
TJ
co H ~n
ui m m
O rri
H w
II
>o
o
m
-------